[INDOLOGY] Dhvanyāloka

Harry Spier vasishtha.spier at gmail.com
Mon Apr 1 02:27:45 UTC 2024

Hi Howard,
I think you are referring to page 20 of the introduction. (see the
following excerpt below).  There are 27 references to aucitya in the
critical edition of Ingalls et al.  I've attached a text version from
archive.org  which you can search for "aucitya". Its a bit garbled since
its from an OCR but you can use it to locate the term and then look at the
print version.
Harry Spier

20 Introduction
. . .

The literary piece must exhibit appropriateness (aucitya). To begin
with, the plot must be appropriate to the emotions, the determinants,
and the consequents which are to produce the intended rasa. In ex-
hibiting the heroism of a human king, for example, one should not
engage him in adventures that could be accomplished only by a god
(3.10-14 A). If a plot as given in the epics and Puranas contains a trait
that is inappropriate to the character of the hero or to the intended
rasa, one must either omit it or add some element to the plot to achieve
the needed appropriateness (3.10-14e A). In this regard Ananda cites
Kalidasa as an example to be followed. His reference is in general terms
only, but we might supply such a specific instance as the Sakuntala,
where in the epic prototype the king abandons with needless cruelty the
heroine whom he has seduced.’ Such action would be inappropriate to . . .

Harry Spier

On Sun, Mar 31, 2024 at 9:53 PM Howard Resnick via INDOLOGY <
indology at list.indology.info> wrote:

> Dear Scholars,
> In the Harvard Press edition of Ānandavardhana’s Dhvanyāloka, I recall
> reading a passage stating, basically, that in the presentation of
> *Mahābhārata* stories, one can employ the principle of *aucitya* (MW:
> fitness, suitableness, decorum) to bring about a suitable rasa that the
> audience will understand and feel. Further, one may adjust or alter certain
> details in the MBh in order to bring this about this effect.
> Presumably a famous example of this would be the various presentations of
> the Śakuntalā story.
> I have been unable to find the passage that, I believe, states this
> principle of aucitya. I would be grateful for any help in tracking this
> down.
> With best wishes,
> Howard
> _______________________________________________
> INDOLOGY mailing list
> INDOLOGY at list.indology.info
> https://list.indology.info/mailman/listinfo/indology
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://list.indology.info/pipermail/indology/attachments/20240331/93fe356a/attachment.htm>
-------------- next part --------------


The pages which follow carry a translation and annotation of two 
Sanskrit texts. The first has been known for many centuries as the 
Dhvanyáloka, or "Light on [the Doctrine of} Suggestion,” and has been 
ascribed to Rajanaka Anandavardhana, a Kashmiri author of the ninth 
century A.D. The fact that this was not the original title of the work 
and the fact that many scholars have recently claimed, | think wrongly, 
that Ananda wrote only the Vrtti or prose portion of it, are matters 
of which I shall speak later. The second text is a commentary on the 
first, called the Locana, or “The Eye,” composed in about A.D. 1000 
by another Kashmiri, the critic, philosopher, and Saiva mystic, Abhi- 
navagupta. These two texts have proven over the centuries to be the 
most influential works of India on the theory and practice of literary 
criticism. For the last thousand years all Indian critics of Sanskrit, and 
many even of those who have written on the literatures of India's mod- 
ern languages, have studied their doctrine, if not directly from the texts 
themselves, at least through the rendering of Mammata. Even when an 
author has disagreed with their pronouncements he has treated these 
works with honor and has taken pains to answer their point of view in 
establishing such other doctrines as he might favor. 

How did Ánanda's view of literature arise? I use this general phrase 
because, as will soon be clear, the doctrine of dhvant, of suggestion or 
suggestiveness in literature, forms only a part of it. And what is there 
about his view and about Abhinava's commentary on it that placed 
these-works in such a magisterial position? In this Introduction I shall 
try to answer these questions, steering a course between the generalities 
of brevity and a full-scale exposition, which would be nothing less than 
a history of Indian aesthetics. 

As both Ánanda and Abhinava were Kashmiris and drew heavily on 
Kashmiri traditions of scholarship, it is to that northern province of 
India that we should look for the historical background of their works. 

2 Introduction 

Kashmir in the narrow sense is a small valley ringed in by the immense 
ranges of the Himalaya and Karakoram mountains. The floor of the 
valley, some eighty by twenty miles in extent, lies a mile above sea level 
and is watered by the Jhelum River, which twists its way northwestward 
into the Wular Lake, issuing westward from which it breaks through the 
mountain barrier. The soil of the valley, helped by winter snows and 
spring rains, is fertile. In those brief periods when the valley was well 
governed and when measures were taken to prevent annual flooding, 
both produce and population rose to a high level, permitting the kings 
of Kashmir to extend their sway well beyond the narrow li its of their 

The frst of our two authors lived toward the end of the longest period 
of strong government that Kashmir ever enjoyed, a period in which for 
once this little valley played a major role in the political history. of 
Asia. This strong Kashmir was the achievement of two long-reigning 
kings of the Karkota dynasty: Lalitàditya, "the World Conqueror" 
(reg. A.D. 725-761), and his grandson, Jayapida, the great patron of lit- 
erature (reg. A.D. 776-807). Kalhana's Rájatararigini gives a delightful 
account of their reigns, mixing history with the romantic adventures 
of folk tales. As the reader may follow the account in Sir Aurel Stein's 
edition and translation of the Rdjatararigini, I shall limit myself to 
selecting the few items which are strictly to my purpose. 

In A.D. 732 or 733 King Lalitaditya borrowed from the Bhütesa Tem- 
ple ten million dirináros to finance an expedition to the south. The 
goal of his expedition was King Yasovarman of Kanauj, who controlled 
through feudatory rights the Punjab and most of the Ganges valley. 
In 733, the date perhaps being confirmed by a solar eclipse, at the 
confluence of the Jumua and the Ganges, 700 miles as the crow flies 
from where he had set forth, Lalitaditya met with and defeated his 

‘ The Gatidavaho, a Prakrit poem deseribing the victories of Ya$ovarman, contains 
a passage (vss. 827-832) mentioning various portents, including a solar eclipse, which 
occurred just before Yasovarmaa's “[royal] position became suddenly weakened" 
(khana-nivvadia-niyaya-paya-bharige = ksano-mirvrtta-nyja/ka|-pada-bharge). Such 
is the reading of three of the four manuscripts of the poem as edited in the Bombay 
Sansknt and Prakrit Series, No. 34, second ed., 1927; see Introduction, p. cclvii. 
Hermann Jacobi took this weakening of the royal position to refer to YaSovarman's 
defeat at Prayaga by Lalitditya (Gott. Gel. Anz. 1888, II, pp. 67-68). A total eclipse 
of the sun would have been visible from PraySga (Allahabad) in a.o. 733. But the 
reading of the latest edition of the Gaüdavaho (Prakrit Text Society, Ahmedabad, 
1975), khana-mvodia-b^ülad-Dhariga-bharigurápórige. eliminates the “weakening of 
the royal position.” 

Introduction 3 

enemy. A temporary peace seems to have been arranged. for in 736 we 
find that Lalitaditya’s ambassador to the Chinese emperor apparently 
refers to Yasovarman as Lalitáditya's ally? Meanwhile Lalitaditya had 
advanced eastward to raid the former adherents of Yasovarman in Ma- 
gadha and Bengal. Before 740 the peace was broken, Yasovarman was 
rooted out of his kingdom, and the revenues of his capital city of Kanauj 
were given to a temple of the Sun in Kashmir.’ Lalitaditya, who was 
now the paramount ruler of northern India, repaid the Bhütesa Temple 
110,000,000 dirináras for the 10,000.000 that he had borrowed. 

The World Conqueror spent the last twenty years of his reign found- 
ing cities and religious institutions in the valley of Kashmir and in 
expeditions against his northern neighbors, the Dards, Tibetans, and 
Turks. On the second day of Caitra (April/May) of an unspecified year 
he won a great victory over the Turks. Three centuries later the Arab 
traveler Alberun! reports that in Kashmir this victory was still the oc- 
casion of an annual celebration. The nationalist pride which Kashmiris 
felt in the victories of their king is reflected in the boast of Kalhana:* 

The rulers of many lands to this very day wear symbols of their defeat, which 
the fierce king forced them to adopt. Clearly it is by his command that the 
Turks, to show that they had been enslaved, still walk with their hands held 
behind their backs and wear their heads half shaved. while he forced the rulers 
of the South, in token of their having been reduced to the state of beasts, to 
wear a tail to their dhotís, which reach to the ground. 

The World Conqueror ended his days on an expedition across "the 
ocean of sand," that is to say, in what is now either Russian or Chinese 
Turkestan, in quest of further glory. After the brief reign of his two sons, 
he was succeeded by a grandson, Jayapida, whom his tutors, following 
the dying instructions of the World Conqueror, had constantly urged 
to "be like your grandfather.” 

Jayüpida, as soon as he reached the throne, attempted to follow their 
advice. He too gathered a force for a great southern conquest. Fate 
turned against him at Benares, where most of his troops deserted. But 
by reckless bravery the young king retrieved his fortune. He allied 
himself with the king of Bengal, whose daughter he married. In the 

? Stein, Rdj T. 1, p. 89, with reference to Chavannes aad Lévy, JA 1895, p. 353 
The ambassador bad been sent to obtain Chinese aid against Lalitáditya's northern 

> Ray. T. 4187. The temple was at Lalitapura, the modera Latpér 

* Réy. T. 4.178-180, my translation. 

4 Introduction 

course of his return he once more subjected Kanauj to the rule of 
Kashmir. Then at the entrance to his homeland he defeated a usurper 
who had arisen in his absence. 

We are told that after his return Jayapida became a great patron 
of learning, attracting so many scholars to his court that there came 
to be a dearth of wise men in other kingdoms. Kalhana lists many of 
the scholars and poets whom the king brought into his service. Among 
them was the grammarian Ksira, from whom the king himself took 
lessons (Ràj. T. 4.489). The person meant is doubtless Ksirasvamin, 
the well-known commentator on the Amarakosa and the Nirukta. “The 
king engaged Bhattodbhata as his sabhápati at a salary of a lakh [of 
dirináros| a day and made the poet Damodaragupta, the author of the 
Kuttanimata, bis chief minister” ( Ráj. T. 4.495-6). I shall have more to 
say below of Bhattodbhata, or Udbhata as he is now generally known, 
for his influence on Ananda and the Dhvanyóloka was great. The min- 
ister Damodara was perhaps the most original of classical Sanskrit po- 
ets. That may be why his work later fel! out of fashion. Fortunately 
nearly the whole of it has now been recovered.* Among other minis- 
ters is mentioned Vamana, doubtless the poetician quoted in our texts, 
and among other poets is mentioned Manoratha (Ràj. T. 4.497), whom 
Abhinava identifies as the author of a verse opposed to the doctrine of 
dhvani, quoted by Ananda (1.1c A). Finally, we are told that the king, 
having dreamed that the sun was rising in the west, was happy to re- 
alize that the Buddhist scholar Dharmottara had entered his kingdom 
(4.498). This will explain how Ananda came to write a commen- 
tary, as Abhinava tells us (3.47 L and see note 6), on Dharmottara's 
Pramànaviniscayatikà. The man had been the leading Buddhist scholar 
in the kingdom in the generation previous to Ánanda's. 

> The frame story, which gives the title to the work ("Tbe Bawd's Advice"), can 
be likened to several other Sanskrit works, but the emboxed stories are sui generis. 
The combination of comic and tragic in the tale of Haralaté and Sundarasena is 
against all tbe classica] conventions of rosa. The tale is told by the bawd with an 
explicitly cynical purpose, but i it the death of Haralata will break the reader's 
heart. D&modera's is the only example I know of in Sanskrit of this double attack 
on tbe reader's sepsibilities. 

* Tbe works of Dbarmottara were aot yet known to the West when Stein wrote 
bis translation of the Rdj T. (1900). Stcherbatsky's edition of the Nyáyabindutikà 
was published only in 1909. This explains Stein's misunderstanding of the passage, 
where he failed to see that dharmottara was a proper name. As for the sun rising in 
the west, Dharmottara, like most visitors to Kashmir, would bave entered the valley 
through one of the western passes of the Himalaya 

Introduction 5 

It was under King Jayápida that the school of literary criticism in 
Kashmir originated. The sabhá to which Udbhata was appointed as 
director (pati) was doubtless a panditosabhà rather than a political 
body. One might therefore translate Udbhata's title approximately as 
Director of the Royal Academy. His salary, by the way, was not so 
immense as it sounds. The dirinára (the word derives ultimately from 
the Roman denarius) had become by this time a money of account, 
of less worth than the smallest coin. A- lakh of dirináras in Jayápida's 
reign would have equalled approximately twenty-five silver rupees of 
Akbar's mintage (Stein, Vol. II, p. 323), a princely but not a fabulous 
daily retainer. 

It is pertinent to our understanding of Ananda and Abhinava to in- 
quire what materials were available for the teaching and research of 
such an academy, or to the individual scholars who might compose it. 
To judge from the quotations and references in the Dhvanyáloka and 
Locana, Kashmiri critics of their time had access to all the epic mate. 
rial and most of the early classical material that we now possess. A 
few works that are now standard in a library of classical authors are 
absent—those of Asvaghosa, as one might expect, and more surpris- 
ingly the Bhattikdvya and the plays of Bhavabhüti—but not many. On 
the other hand, they had much that we have now lost.” [n addition 
they were acquainted with a substantial literature in Prakrit, most of 
which is now lost. The only two works in this category quoted by 
Ananda which have come down to us are the Sattasai and the Gaüda- 
vaho. These account for only about a fourth of his quotations.® 

While the Kashmiri critic thus had access to a substantial library 
of Sanskrit and Prakrit literature, his library of literary criticism, in 
the early years of its development in Kashmir, was extremely modest. 
It consisted of little more than the eighth-century works of Bhàmaha 
and Dandin. There existed also the ancient manual of the theater, the 
Bháratiyanátyasástra (BhNS), but until Udbhata turned his attention 
to it, this work had played almost no part in general literary criticism 

” Of Ananda’s 131 Sanskrit literary quotations, we still possess the originals of 
59, to which one may add 29 more that may still be found in anthologies of later 
date, in reworkings (the Hanumanndtaka), or in the work of later literary critics. 
This leaves 41 for which our only source is Ananda. 

* To be precise, 10 out of 39. Seven verses, not counted in the 10, may be found 
in the supplement to Weber's edition of the Sattasai, but they most probably found 
their way into the manuscript sources of the supplement from the Dhvanyóloka. 

6 Introduction 

Bhamaha and Dandin had spent most of their effort in defining and 
exemplifying the figures of speech, a science which they had developed 
to a point comparable to whet the West has to offer in the Greek of 
Demetrius or the Latin of the Rhetorica ad Herennium. Beyond that, 
they had listed the fauits and the good qualities (gunas) of poetry and 
had spoken of its different styles (ritis). But what they had to say on 
these subjects suffers from two serious weaknesses. The qualities are so 
general that they offer no operable criteria of what is great or beautiful 
in poetry and what is not. The three primary qualities were given as 
sweetness (mddhurya). clarity (prasáda), and strength (ojas). How is 
one to say when a stanza, much less a whole poem, is sweet and when 
it is not, or how judge whether it has strength? In an effort to render 
the qualities more precise, Bhámaba and Dandin made the error of 
identifying them with measurable elements of phonetics and structure. 
Thus, a large number of retroflex consonants and consonant clusters 
and the use of many long compounds were said to exhibit strength. 
The modest use of compounds and the avoidance of harsh phonetic 
combinations gave sweetness. No better was the attempt to associate 
these qualities with regional differences. The sweet style was associated 
with Vidarbha in the Deccan, the harsh or strong style with Bengal. 
A third style, Pafical. fell aesthetically and geographically somewhere 
in between. 

Such concepts and associations died hard. Not until Ananda was it 
pointed out that long compounds are not really necessary for strength 
(Dhv. 2.9). In fact Vamana, who belonged to the first generation of 
Kashmiri criticism, if anything exaggerated and worsened these early 
faults. He extended the number of qualities to ten and defined each 
as of two sorts, depending on whether it was viewed as a quality of 
sound or a quality of meaning. Unfortunately his new qualities, such 
as samatá (regularity), saukumdrya (delicacy), and kdnti (brilliance), 
are as vague and as difficult to define as the original three. It is these 
qualities, he said, which give beauty to a poem, a beauty which may 
then be enhanced by the use of figures of speech (Vàmana 3.1.1-2). He 
continued the old association of certain qualities with regional styles 
and so came to the dictum for which he is chiefly remembered, "Style 
is the soul of poetry" (ritir dtma kavyasya, Vàmana 1.2.6). 

While Vàmana, who may well have come from some older center of 
literary studies outside Kashmir, looked backward for his inspiration, 
Udbhata, whose name indicates that he was a native Kashmiri, may 
be said to have looked forward. We know that he wrote a commentary, 

Introduction T 

unfortunately now lost, on the BANS. It was the first of a series of 
commentaries on that work that were to be written in Kashmir in the 
following two centuries, by Lolla;a, by Saükuka, and by Abhinavagupta. 
The importance of this new inter t is inestimable, for as we shall see, 
it was by bringing Bharata's doctrine of the rasas, the flavors or moods 
of a theatrical piece, into a general theory of literature that Ananda 
arrived at a critique which finally could furnish workable criteria of 
literary excellence." 

Fate has been unkind to the works of Udbhata. His other great com- 
mentary, the Bhámahavivarana, is also for all practical purposes lost 
to us. What appear to be minute fragments of it, written op birch- 
bark in a hand of the ninth to eleventh century, have been lovingly 
reconstructed by the Italian scholar Raniero Gnoli. But while Gnoli 
makes out a persuasive case for Udbhata's authorship of these frag- 
ments, even if his case were fully proved none of them is of sufficient 
continuity to furnish evidence for the new ideas which Udbhata must 
have propounded in that work. We know that it did contain new ideas. 
Ananda and Abhinava ascribe to it, for example, the doctrine that the 
beauty of particular words depends on the rosa that the author wishes 
to achieve (see 3.16m A and L). This would be a major step toward 
their critique. Jacobi goes so far as to say that "Udbhata was the first 
to designate rasa as the soul of poetry" (ZDMG 56, p. 396). But this is 
saying too much. The verse on which Jacobi based his statement is not 
by Udbhata at all, but by some unknown, and doubtless later, author 
quoted by Udbhata's commentator Indurája.'? What is true, rather, 
is tat Udbhata was the first of the literary critics to concern himself 
seriously with the concept of rasa. He was not prepared, however, to 
make it the chief goal of poetry, as Ananda was to do. 

The only book of Udbhata’s that we possess in readable form is the 
Kévydlankérosérasarigraha, “A Compendium of the Most Important 

* The older poeticians had been aware of rasa, but had not shown what | should 
call much interest in the subject. Both Bhámaha and Dandin relegated examples of 
it to the rasddi figures of speech (rasavadalarikdra, preyo‘larikdro, ürjasvin), gures 
where they found the emotions (bhávas) to be strongly or strikingly expressed. 

10 The verse runs: rasddyadhisthitam kávyam jivadrüpatayá yatah / kathyate tad 
rasādīnám kávyátmatvam vyavasthitam // It occurs in Indur&ja's commentary on 
Udbbata's definition of kavyaAetu (6.7; in the numeration of the Viurt: 6.14). That 
it is not by Udbbata is clearly indicated by the words with which Iodurdja introduces 
it: tod huh. The error in identification originated with Col. Jacob, JAOS 1897. 
p. 847, and bas been corrected by P. V. Kane, HSP, p. 128. 

8 Introduction 

Figures of Speech in Poetry.” This little work furnishes definitions 
of forty-one figures, which are then illustrated in verses narrating the 
story of the Kumdrasambhava up to the point reached by Kalidasa in 
the fifth canto of his poem on the same subject. Many of the definitions 
are the same as those of Bhamaha. But we should not let that fact 
obscure Udbhata's innovations. In the definition of the very first figure 
of sense, rüpoka (1.11 Indurája, 1.21 Vivrti), we meet with a distinction 
that was new to Sanskrit poetics and that was destined ultimately 
to transform the analysis of all the figures. This is the distinction 
between the furnishing of a meaning srutyd, that is, explicitly, and 
furnishing it arthena, that is, by the power of the contexual facts, or 
implicitly. The same distinction appears in the analysis of simile (1.16 
Indurája, 1.33 Vivrti). Thus rūpaka (metaphor)! differs from simile 
by the fact that the similarity between the superimposed object and 
its real base is always given arthena, is "understood" from context, 
whereas in simile it is given by śruti, that is, usually by an explicit word 
(e.g., iva, yathà = "like," "as") expressing the fact that the similarity 
is shared. 

This concern with the implications of words appears in many of 
Udbhata's definitions. In dipaka (zeugma), for example, the paired 
properties are said to contain or imply a simile. Thus, where a poet 
writes that "the doom of autumn carried off the beauty of the kadamba 
flowers and all the joy of damsels separated from their lovers," he is 
suggesting an implicit similarity between the beauty of the lowers and 
the joy of the damsels. To follow the concern for the implied or sug- 
gested sense through the whole of Udbhata's book would require a 
more detailed exposition than is justified in this Introduction. It ap- 
pears in his definitions of paryáyokta, aprastutaprasamsá, sandeha, and 

u Rüpaka is not what a Greek would have called metaphor, but that translation 
has come to be used by every Sanskritist. Rupaka is actually a simile in which the 
particle of assimilation has been omitted, e.g., "ber moon face, her cherry lip." la 
a Greek metaphor the object as well as the particle is missiog: “ber stars shone 
upon my face,” meaning that her eyes looked at me. The distinction is noted by 
Gero Jenner, Die poetische Figuren der Inder, p. 68, Ludwig Apfel Verlag, Hamburg, 

Pratihára Induraja, who commented on Udbhata's book some time 
after Ananda’s Dhvanyóloka had become popular,!? concludes his com- 
mentary with a disquisition on why Udbhata had nothing to say of 
dhvani, "which some connoisseurs [i.e., Ananda and his followers] con- 
sider to be the very life of poetry." His answer, in brief, is that Udbhata 
included dhvani in his treatment of the figures of speech. The answer is 
Dot strictly true but it points the way to an important truth. Udbhata 
nowhere uses the word dhvani. He speaks of a meaning's being un- 
derstood (pratiyeména), or implied (gamyate), or of its being included 
(antargata) in another meaning, but he avoids using the more technical 
terms vyajyate or dhvanyate for “is suggested." His avoidance cannot 
have been because he did not know the use of the words in this sense, for 
his contemporary Manoratha laughs at critics of the time "who will tell 
you witb delight that a poem is full of dhvani but cannot tell you just 
what this dhvan: is" (Dhv. 1.1c A). Perhaps Udbhata wished to dis- 
tance himself from the new enthusiasts and to keep as far as possible 
to the old terminology of criticism. But Indurája's remark is justi- 
fied to this extent: Udbbata was fully aware of that type of semantic 
operation that Ananda was later to call suggestiveness (vyarijokatva, 
dhvani) and of the importance to poetry of the suggestions which it 
could bring about. One might fairly say that in Udbbata's mind the 
two main building blocks of Ánanda's critique, rasa and dhvani, were 
present, the first consciously, the second perbaps only subconsciously. 
But the blocks bad not yet been built into a system. 

It is said that in his old age King Jayàpida became ruthless in the 
exaction of taxes, oppressing both his peasants and the brahmins. He 
died after ruling for thirty-one years and was succeeded by a number of 
worthless descendants. For nearly fifty years Kashmir fell back into its 
habitual state of misrule. Then in A.D. 855/856 a strong-willed minister 
set up a young man of a collateral line, Avantivarman, who was to rule 
for nearly three decades. King Avantivarman was descended from Ut- 
pala, an uncle of Jayápida's daughter-in-law, whence the new dynasty 
has come to be known as the Utpala Dynasty. Avantivarman and his 
minister Süra brought the treasury back to solvency. They installed 
major works of drainage and irrigation. Once more the king became 
resplendent through the poets who graced his court. Kalhana gives 
us the names of four of them (Raj. T. 5.34): Muktákana, Sivasvàmin, 

"P. V. Kane puts the date of Prati 
and 950 (HSP, p. 197). 


Anandavardhana, and Ratnakara. The works of Muktákana are lost. 
Of Sivasvamin we possess a Buddhist kávya, the Kapphindbhyudaya. 
Of Ratnàkara we have the longest of all classical kávyas, the Haravi- 
jaya, and a small collection of clever verses, the Vakroktiparicásika. Of 
Anapdavardhana it is now time to speak. 

Of Ánandavardhana's life, beyond the fact that he was patronized 
by King Avantivarman, we know nothing except what can be inferred 
from his two extant works and their colophons. From these it appears 
that he was the son of Nona (Devisataka 101) and that he bore the 
title Rājānaka (Dhv. 4 Conclusion A and note 3). Nothing is known of 
Nona. The title denotes no political position and probably implies no 
more than that he was given a stipend by the king 

Ananda was the author of many books. In the Dhvanyáloka he refers 
to two earlier works, which are now lost: the Arjunacarita, "The Adven- 
tures of Arjuna”; and the Visamabénalilé, "The Sports of the Bowman 
Love.” The first of these was evidently a Sanskrit mahdkdvya. Ananda 
tells us (3.10-14 e A) that he altered the traditional account of Arjuna's 
life so as to include new material of his own invention on Arjuna's ad- 
ventures in the underworld. Abhinava quotes one stanza (3.25 L) of the 
poem, written in an unpleasing meter of unrelieved iambs, but it would 
be unfair to judge the work as a whole from one accidental quotation. 

Of the Visamabánalilà we can say somewhat more. We are given 
four quotations from it and several remarks about its subject matter. 
The quotations show that it was written in Mabarastri Prakrit. It may 
have been in the form of a play (or a narrative work would be possible), 
for 3.15 A refers to "the scene where the God of Love meets with his 
friends (Youth and Springtime] in my Visamabanalilà. The purpose 
of the work, however, was to give instruction in poetry. In speaking 
of the variety which may be achieved by handling an insentient object 
as if it were sentient, Ananda remarks, "This is a well-known proce- 
dure of great poets and has been described in detail for the instruction 
of poets in the Visamabdnalila” (4.7 A). I believe we can be more 
specific. The quotations which we have from the work exhibit vari- 
ous types of suggestiveness. The stanza at 2.1 b 4 exemplifies arthán- 
tarasarikramitavácyadhvani; that at 2.27 b A is of alarikáradhvani. Con- 
cerning the verse quoted by Abhinava there may be some question as 
to the precise type of dhvani intended (see 3.15 L and notes), but it is 
certainly dhvani that is being illustrated. [ would say, then, that the 
Visamabénalilé was Ananda’s first work propounding the new doctrine 
of suggestiveness, in a play or narrative written quite appropriately in 


Prakrit, for Prakrit was the language in which this style of suggestive- 
ness first became popular and it may well have been from Prakrit that 
Ananda's interest in dhvani was first stimulated. The work was most 
certainly not an anthology, as Sten Konow once suggested.'? 

Ánanda also wrote on philosophy. Abhinava twice refers (1.4b L and 
4.5 L) to a work of Ananda's called the Tattváloka, which from the 
context of the references seems to have dealt with both metaphysics 
and literature. Again, Ananda himself speaks of his intention to write 
a book which would examine the doctrines of the Buddhists (3.47 A). 
Commenting on this passage, Abhinava tells us that the book to which 
Ananda refers was his Dharmottarivivrti,!^ an “explanation” of Dhar- 
mottara's commentary (tiká) on the Pramánaviniscaya of Dharmakirti. 
Although the underlying texts here of Dharmakirti and Dharmottara 
are preserved, at least in Tibetan translations (see 3.47 L, note 6), 
Ananda's Vivrti..like the Tattudloka, seems to be irretrievably lost.’ 
It is perbaps natural that Ananda should have chosen Dharmottara as 
representative of the Buddhist viewpoint, for Dharmottara had taught 
his doctrine in Kashmir under the recent reign of King Jayapida. But 
it is unusual for a devout Hindu to have written on such abstruse points 
of Buddhist epistemology and metaphysics as Ananda must have found 
in the Pramánaviniscayadharmottari. 

Ananda was indeed a devout Hindu, as appears from the stanza 
quoted at 3.43b A and from the Devisataka, a poem that has been 
published in the AM Series (ninth gucchaka). The poem consists of 
103 trick stanzas (yamakas and citrabandhas) in praise of the mother 
goddess. It culminates in a sort of crossword puzzle, a great wheel, the 
spokes of which are formed by sixteen stanzas, the outer rim by four 
other stanzas the syllables of which interlock with the spokes. The se- 
cret of the puzzle lies in an "inner rim" which gives the message: “The 
son of Nona has thus performed his worship of the Goddess under the 

? Rajasekbara’s Karpüramafjari, HOS Vol. 4. p. 193. 

“The statements of Kane (HSP, p. 194) and Jacobi (ZDMG 57, p. 328, note 9) 
are based on the incorrect reading of the KM edition of the Locana The correct 
reading 13 uiniscayatikdydm dAormottaryám, not dharmottamáyám. 

*? Ananda seems to have written still other books which are now lost. At least, 
that is what I iofer from the scatter verses of his own which be quotes in the 
Dhv. Some of these verses, like the courtly stanza on bis mistress's face (227a A) 
and the stanza oo discouragement (3.40a A), may be occasional verses (muktakas) 
which never belonged to any larger collection. But others, such as the punning 
benedictions to Krishna and to the Sun God (2.21f A), or to Rukmig (2.21b A), 
would seem by their oature to bave introduced major works 

12 Introduction 

title of "The Goddess's Century,’ as instructed in a dream, a worship 
unsurpassed by reason of her having been the instructress." 

For suggestions as to why Ananda chose to write this citrakávya 
in praise of the Goddess when he casts such scorn on this type of 
composition in the Dhvanyóloka, I refer the reader to my forthcoming 
essay on the Devisataka in the Ernest Bender felicitation volume. I 
shall here remark only on the fact that the yamakas of the poem are 
musical and the citrabandhas extremely clever. [f one may speak of 
better or worse citrakávyas, the Devisataka must rank with the better. 

In the Vrtti of the Dhvanyáloka Ananda proves himself one of the 
great prose stylists of Sanskrit literature. No matter how delicate or 
complex the subject, he is always clear. He varies his expression, so 
that no matter how often he comes back to the importance of rasa and 
dhvani he seems never to repeat himself or become tedious, while the 
thythm of his sentences gives constant delight. In comparison with his 
prose, bis verses for the most part are disappointing. They go against 
his own advice by being too consciously clever. But [ make exception 
of the noble stanza which he gives us at 3.43 b A. I should like to think 
that it represents his view of his life and of his life's work. 

I am weary from much painting of the world, 

for although ] used the new and wondrous sight of poets 
which busies itself in giving taste to feeling 

and used tbe insight of philosophers 

which shows us objects as they really are. 

I never found, O God recumbent on the Ocean, 

a joy like that which comes from love of Thee. 

It is for the content of the so-called Dhvanydloka. however, rather 
than as a poet or a master of prose style, that Ananda has become 
famous. As regards the original title of his great work, one should note 
that none of the manuscripts gives it the name of Dhvanyáloka.'5 The 
colophons of the manuscripts refer to it usually as Sahrdaydloka, "A 
Light for Connoisseurs”; sometimes as Sahrdayahrdaydloka, "A Light 
for the Hearts of Connoisseurs”; and rarely as Kàvyáloka, “A Light on 
Poetry." The oldest commentator on the text whose work has survived, 
Abhinavagupta, refers to his commentary as the Sahrdaydlokalocana, 
“An Eye for the Sahrdayóloka,"'" and this is the title found in the 
colophons of the first three chapters of his commentary in the printed 

* See Kane, HSP, p. 181, and Krishnamoorthy's edition, p. 36, note 4. 
?! The references are noted by Kane HSP, p. 170, note 1 

Introduction 13 

editions. The colophon of the fourth chapter, which derives from a sep- 
arate manuscript tradition, gives the name of the work as Kávyáloko.'* 
In the absence of a critical edition of Ananda's great work’? the colo- 
phon readings may not be decisive, but the weight of Abhinava's tes- 
timony when added to their evidence seems to clinch the matter. The 
name which Ananda gave to his work, I believe, was Sahrdaydloka. 

The Schrdaydloka, then, to give it that title, opens with a spirited 
defense of suggestion as an independent semantic power. As I have 
indicated, the subject of suggestion was not a new one. It had occupied 
the thoughts of Udbhata. The term dhvani that Ánanda uses had 
been laughed at by Manoratha. An important stimulus to discussion, 
it seems to me, must have been the Prakrit literature which formed an 
important part of Kashmiri critical studies. The first five quotations 
in Ánanda's opening defense are all taken from Prakrit. The reason is 
not far to seek. If we look at the verses of the Sattasai, we see that it is 
suggestion upon which the effect of almost every stanza depends. The 
gáthà stanza, in which they are written, is so brief a poetic form that 
it could scarcely attain a powerful effect by any other means. Such 
verses lend themselves naturally to the thesis which Ánanda set out to 

To understand the argument we must cast a brief glance at the tra- 
ditional Indian theories of meaning into which the new doctrine was 
introduced. These theories had been developed over a long period 
of time by the grammarians and the ritualists (Mimámsakas). The 
Buddhists too, in the writings of Dignaga and Dharmakirti, had had 
their say. It was generally agreed that words had two sorts of seman- 
tic power: the power of direct denotation (abhidhd) and a secondary 
power of indirect indication (gunoavrtti, bhakti, upacáro, laksaná).? By 
denotation, a particular group of phonemes in a particular order, say 
9-au-h ("ox"), denotes an animal with horns, hump and tail. But when 
one says gour vdhikeh, "the Punjabi is an ox," meaning that the man 
is stupid, it is the secondary power that is working in the word. The 
secondary power may be elicited by a common property of two objects, 
as in the example just quoted, or it may be elicited by some other 
relation, for example that of possessor with the thing possessed, as in 

1* S. K. De, The Text of the Kàvyólokalocana IV, p. 265 

‘© Krisbnamoortby's edition is helpful. for he gives the variants of a South lodian 
MS from Moodabidre as well as of the NSP MSS and occasionally of other MSS 
from the BORI. But there exist many manuscripts about which he is silent. 

?? For the distinctive uses of tbe Sanskrit terms see 1.1 K. note 2. 

14 Introduction 

nagaram pravisanti kuntáh, “the spears enter the city," 
meant is spearmen.”* 

In addition to these two powers, the school of ritualists founded by 
Kumárila held that there existed a third power which furnished a "final 
meaning” to the sentence as a whole. They called this the tátparyasakti, 
and defended its reality against their opponents, the Prabhakara rit- 
ualists, who claimed that the denotative force in each word kept on 
operating until at the conclusion of the sentence it worked automati- 
cally in harmony with the other words. 

These three powers left no room for what Ananda considered to be 
of all semantic powers the most valuable for poetic expression, a power 
which in its most general aspect he calls vyarijakatva, the power of 
suggestion, or, more literally, the power of revelation (as of a lamp 
which reveals the objects upon which it casts its light). He calls this 
power dhvani when it is in its purest form, that is, when it predominates 
over the other semantic powers in the sentence. He begins the proof of 
its existence by a number of humorous examples. The suggestion may 
be of an act that is the very opposite of what is denoted. One of his 
examples is this (1.4c A): 

Mother-in-law sleeps here, 1 there; 
look, traveler, while it is light. 
For at night when you cannot see, 
you must not fall into my bed. 

What is denoted bere is a prohibition. There is obviously no secondary 
operation of metaphor or the like in the stanza. The tátparyasakti, if 
such there be, merely conveys the syntax of the sentence. The syntax 
is still impeccable if we take the statement as a prohibition. And yet 
we know, as the traveler must have known, that the prohibition is 
unintended and that the woman speaking is inviting him to sleep with 

By the use of examples Ananda builds up a typology of suggestion. 
The type to which the example just given belongs he calls avivaksita- 
vücya, suggestion "where the denoted sense is unintended." The type 
has two varieties. In the variety just exemplified the denoted sense is 
atyantatiraskrta, "entirely set aside." The second variety is where the 

2" n a rough way one may say that a secondary meaning in Sanskrit corresponds 
to the Greco-Latin “trope”; but only in a rough way, because several of the tra- 
ditional! Greco-Latin tropes (e.g., byperbole. allegory) are treated by the Sanskrit 
poeticians as figures of speech. 

Introduction 15 

denotation is not wholly abandoned but is “shifted to something else” 
(arthéntarasankramita). When we say, "The spears enter the city,” 
we are using the secondary power (we are using a trope). The literal 
meaning of “spears,” its denoted object, namely, weapons of a specific 
shape, has been replaced by men carrying spears. But why do we speak 
in this way? Why do we use such secondary or tropical expressions? 
Usually, say Ananda and Abhinava, in order to achieve some sugges- 
tion. In the case of the spears entering, one imagines a more compact 
and injurious force breaking into the city than would be expressed by 
the literal statement. 
The “first of poets,” Valmiki, wrote of the winter: 

The sun has stolen our affection for the moon, 
whose circle now is dull with frost 

and.like a mirror blinded by one's breath 
shines no more. 

Ananda quotes this verse (2.1 c A) for its use of the word "blinded" and 
Abhinava comments both on the secondary usage and the suggestion. 
The word "blinded," he says, is used in a secondary sense here, because 
only sentient creatures can be literally blind. The purpose, though, of 
using this trope is to suggest "numberless properties [of the winter 
moon] such as uselessness, an exceptional loss of beauty, and so on." A 
point that is noticed often by both our authors is that suggestion vastly 
increases the scope of words. The denotation is extremely narrow; the 
secóndary sense includes only things (objects, properties, acts) which 
are closely related. The suggestion opens up a new world. 

So much for suggestion of the first type. In one of its varieties it is 
poetically useful, but in neither does it yet reveal the ultimate purpose 
of literature. That revelation lies within a second type of suggestion, 
which Ananda calls vivaksitányaparavácya, “where the literal sense is 
intended but only as leading on to something further." This type also 
Ananda divides into two varieties, depending on whether or not we are 
conscious of the succession from the literal to the "something further." 
Much the more important of the two varieties is that where we are not 
conscious of any interval between the two senses (asamlaksitakrama). 
for in this variety the "something other" is a rosa or something closely 
allied to a rasa; and rasa in the critique of Ananda is the ultimate aim 
of literature. 

The word rasa in its most literal sense means juice, taste, favor. In 
a technical sense the BANS uses it to express the flavor or mood which 

16 Introduction 

characterizes a play if the author and actors are successful in their 
work. According to BANS 6.15 there are eight possible favors which 
a play may exhibit: the erotic (srrigára), the comic (hásya), the tragic 
(karuna), the furious or cruel (raudra), the heroic (vira), the fearsome 
or timorous (bhayánaka), the gruesome or loathsome (bidhatsd), and 
the wondrous (adbhuta). To these Ananda adds a ninth, the rasa of 
peace (santa), These flavors, as their names indicate, are based on var- 
ious human emotions, the sthdyibhdvas or “abiding emotions,” as they 
are called, which are listed in BANS 6.17.7? Just how the rasas differ 
from the emotions was a question much argued in Ananda's age and in 
the following two centuries. Curiously, Ananda has never a word to say 
on the subject; and unfortunately most students of the Dhvanyáloka 
have inconsiderately filled the gap by superimposing Abhinava's expla- 
nation on the text of Ananda. I propose to come at an answer more 
cautiously by recalling the words of BANS and its oldest commentators 
and then examining Ananda's use of the term. 

BANS 6.31 +3 (the famous rasasütra) tells us that “A rasa is pro 
duced by the combining of the determinants (vibhávas), the conse- 
quents (anubhdvas), and the temporary or transient states of mind 
(vyabhicárinah or vyabhicdribhaves).”? These technical terms require 
explanation. By determinants are meant those factors which make the 
realization of the emotion and the rosa possible. They are of two sorts, 
objective (álambanavibháva) and stimulative (uddipanavibhàva). The 
objective determinants are the objects toward which the emotions are 
felt. In the erotic flavor they will be the lover and his beloved; in the 
tragic, the person or persons who suffer; in the loathsome, the object of 
disgust. The stimulative determinants in the erotic will be such factors 
as the springtime, gardens, or a bridal chamber; in tragedy, such fac- 
tors as separation from dear ones, death, or capture. The consequents 
of the emotions may be regarded by the audience as its symptoms; in 

13 They are: sexual desire (roti), laughter (hdsa), grief (Soka), anger (krodha), 
heroic energy (utsáha), fear (bhaya), disgust (jugupsd), and wonder or amazement 

d Vibhdvanubhávavyabhicárisamyogád rasanişpattih The commentator Lollata 
supplied a genitive, sthdyinah, to go with the ablative compound. That is, he 
interpreted the sütro to say, "A rasa is produced by tbe combining of the abiding 
emotion with the determinants, the consequents, aod the transient states of mind." 
Later commentators found fault with this interpretation, for they restricted the rasa 
to the audience. In the case of the audience, the basic emotions cannot be observed 
without the previous presence of the determinants, etc. But as we shall see, Ananda 
did not restrict the msa in this way. 


the erotic flavor, for example, they will include the sidelong glances, 
smiles, the graceful movements of the limbs. The temporary or tran- 
sient states of mind are listed as thirty-three in number. Among them 
are discouragement, apprehension, jealousy, embarrassment, intoxica- 
tion. Some are appropriate to only one basic emotion, some to several. 
To them are added eight involuntary states (sdttvikabhdvas), which a 
good actor, however, was trained to represent at will: perspiration, 
horripilation, trembling, fainting, and so on. 

An example will make these technical terms clearer. Ananda quotes 
(4.2a A) the following stanza from the Amaru collection as an example 
of the erotic Javor (srrigárarasa): 

Seeing that the attendant had left the bedroom, 

the young wife rose half upright from the bed 

and, gazing long upon her husband's face 

as be lay there feigning sleep, at last took courage 

and kissed him lightly, oaly to discover 

his feint from the rising flesh upon his cheek. 

When then she hung her head in shame, her dear one 

seized her, laughing, and kissed her in good earnest. 
Here the objective determinants are the husband and his bride. The 
stimulative determinant is the bedroom in which the lovers find them- 
selves alone. The consequents of the bride's basic emotion are her 
gazing at her husband's face and kissing him; the consequents of his, 
tbe laughter and kisses with which the stanza ends. Meanwhile we have 
the bride's transient state of shame or embarrassment and the invol- 
untary state evident in the rising flesh on the husband's cheek. It is 
by the combination of these factors that srrigárarosa, here of the type 
“love-in-enjoyment” (sambAogasrrigára), is suggested. 

Bhatta Lollata, the oldest commentator of the BANS whose views 
are known to us,”* said that the rasa was simply an intensified form of 
the abiding emotion, which it assumed after being strengthened by 
the determinants and similar factors; and this is clearly the sense in 
which Dandin had taken the term many years earlier. Lollata also 
stated that the rosa had its place both in the character being portrayed 

They are known from three sources: the Locana on 2.4, ABh. on 6.31 +3, and 
Mammata 4.28. 

?* Tena sthdyy eva wibhávdnubhdvádibhir upacito rasah. 

% Dandin 2275: yuktotkorgarn ca tat trayam, where trayam refers to the three 
rasdds figures of speech. Again, after giving an example of the Ggure ürjasvin, he 
explains (2.283): sty äruħya pardm kotim krodho raudrótmatám gatah. 

18 Introduction 

and in the actor." The most glaring fault of this interpretation, that it 
leaves out the audience, was partially addressed by Sahkuka, the next 
commentator, who may have lived about Ánanda's time. He stated 
that the basic emotion (bhdva), supposed to exist in the character being 
portrayed, was imitated by the actor for the delight of the audience and 
was given a different name, namely, rasa, because it was an imitation." 
This theory too has its drawbacks, which led to still further theories by 
Bhattan&yaka and Abbinava. But those lie beyond the time of Ananda, 
and I have shown enough now to take up Ananda's use of the word. 

Ananda uses the word rasa of a basic emotion that has been height- 
ened,?? sometimes from whatever reason, but most specifically from the 
combination prescribed by BANS. An example of his use of the term 
in the. most general sense is Dhv. 3.26a A: 

The peaceful is indeed apprehended as a rasa. It is characterized by the full 
development of the happiness that comes from the dying off of desire. As has 
been said, “The joy of pleasure in the world / and the greater joy of pleasures 
found in heaven / are not worth a sixteenth of the joy / that comes from the 
dying of desire” (MBA. 12.186.36). 

Here rosa is simply a heightened form of peaceful happiness (sukha). 
Similarly, “For srrigárarosa, as it is regularly the object of the experi- 
ence of humans and is therefore dear to them, is the most important 
[of the rasas)” (3.29 A). Here one cannot argue that the regular object 
of human experience is the aesthetic pleasure of love poetry. What he 
means is a heightened emotion of sexual love. 

Ananda conceives this rasa to abide in the character invented by the 
poet or in the poet himself, as well as in the audience. As for the first: 
“The speaker may be the poet or a character invented by the poet. If 
the latter, he may be devoid of rasa and bhava, or he may be possessed 
of rasa and bhava” (3.6 g A). As for the poet himself, it is when he is un- 
der such a heightened state of emotion as rasa that he becomes capable 
of writing the suggestive poetry that will transfer this rasa to his hear- 
ers. The process is illustrated by the story of the first poet, Vàlmiki, 

?" Sa cobhayor apy anukárye ‘nukartary api. 

?! Sthayi bhávo mukhyarámádistháyyanukaranarüpo 'nukaronarüpatvád eva ca nã- 
mantarena vyapadisto rasah; ABh. on BANS 6.31 (Vol. 1, p. 272, two lines from 
bottom of page; given also by Gnoli, The Aesthetic Experience, p. 4, lines 8-9) 

7° Even after the time of Ananda, Indur&ja considered a rasa to be simply the 
basic emotion which had undergone strengthening (Indurdja on Udbhata 4.3-4) 
The Candrik&kàra seems to have held this same view; see 3.4a L. note 4 

Introduction 19 

who was so saddened by the wailing of the curlew bird who had lost 
its mate that Valmiki’s grief (Soka, the basic emotion) was transformed 
into the tragic rosa of the Rémdyane (Dhv. 1.5 K and A). The notion of 
Abhinava that Válmiki ruminated on the determinants and consequents 
of the bird's bereavement and so developed his rasa in the scriptural 
way strikes me as an addition quite foreign to the view of Ananda. 

In most cases, of course, Ánanda's rasa is indeed produced in the 
scriptural way by the poetic use of determinants and consequents. The 
examples of this use, as in the verse of Amaru quoted above, are legion 
throughout the book. I wish to emphasize, however, that Ananda's 
sense of rasa has none of the aesthetic removal, the impersonality and 
generalization, which we shall see Abhinava give to the term. 

Ananda was the first Indian critic to state that a rasa cannot be 
directly expressed. If we say, “A young man and his bride were very 
much in love,” we give the hearer no flavor at all of what the love was 
like. This can be done only by suggestion. Accordingly, rasa is as 
important in poetry and literary prose as it is in plays, for there is no 
other way of enlisting the sympathy of the reader. By suggestion the 
rasa arises without any conscious realization that our experience has 
been preceded by a perception of the determinants, coasequents, and 
transitory states of mind. These have been denoted literally and are 
not unintended by the author. They are intended, however, only as 
being productive of the rasa. 

I shall not describe in this Introduction the second variety of vivaksi- 
tüányaparavácya, the variety where we are forced to think about the 
literal sense for a moment before we perceive the suggestion and are 
therefore conscious of the interval between the literal and the suggested 
sense. The reader may examine that variety with all its subvarieties in 
the translation which follows (2.20ff.). Here I wish to speak of matters 
more strictly pertinent to my purpose. 

Ananda tells us that dhvani, that is, suggestion, or more specifically 
suggestion acting as the primary sense of a passage, is the soul of poetry 
(1.1 K). But that is only half the story, for his critique is one which 
explains the goal of poetry to be rasa, and dhvant to be its means. 
Now the concept of rasa. it seems to me, is more important than that 
of dhvani in furnishing a criterion of beauty. For not all dhvant leads 
to rasa, nor does all dhvani lead to beauty.” And it is as the discoverer 

2° Only the most ardent enthusi ill &nd beauty io the punning suggestions of 
the verses quoted under 2.21 f. 

20 Introduction 

of a workable critique of beauty in literature that Ananda merits the 
fame which has long been accorded him. The works of previous poet- 
icians in India, although of interest for their analysis of language, are 
almost useless for this, the chief goal of literary criticism. One might 
write a poem embodying all the figures of speech listed by Bhámaha 
and Dandin and compose it in a style calculated by its phonemic and 
word-joining form to produce sweetness or strength. With a modest 
amount of care in syntax one could add clarity. All this would not en- 
sure the poem's being beautiful, delighting its hearers. Indeed. if one 
followed the definitions too closely, the composition would more likely 
bore them. One gets to the reader only through the flavors deriving 
from the basic emotions. To do that, suggestion is necessary, but the 
test lies in the flavor. 

Te might be thought that for purposes of furnishing a criterion of 
beauty or of literary excellence the achieving of a rosa suffers from a 
touch of the vagueness and lack of precision of which we complained 
in speaking of the “qualities” of poetry. On what basis is one to say 
that one verse achieves rasa and another does not? Of course the final 
test will be the judgment of the heart. But to help the reader and 
the composer—for Ananda always writes with both types of student in 
mind—he speaks of several supplemental tests 

The literary piece must exhibit appropriateness (aucitya). To begin 
with, the plot must be appropriate to the emotions, the determinants, 
and the consequents which are to produce the intended rasa. In ex- 
hibiting the heroism of a human king, for example, one should not 
engage him in adventures that could be accomplished only by a god 
(3.10-14 A). If a plot as given in the epics and Puranas contains a trait 
that is inappropriate to the character of the hero or to the intended 
rasa, one must either omit it or add some element to the plot to achieve 
the needed appropriateness (3.10-14e A). In this regard Ananda cites 
Kalidasa as an example to be followed. His reference is in general terms 
only, but we might supply such a specific instance as the Sakuntala, 
where in the epic prototype the king abandons with needless cruelty the 
heroine whom he has seduced.’ Such action would be inappropriate to 

‘In the Critical Edition of the MBA, after being forced by a voice from heaven to 
recognize his son, the king says to Sakuntald, “It was to purify you jie., to convince 
my people of your purity] that I did this” (MBh. 1.69.40). The Southera version, 
however (MBA. 1.627* S), says that he had simply forgotten her 

Introduction 21 

true love and to the noble character of King Dusyanta as Kālidāsa con- 
ceives it. So Kālidāsa invented the story of the ring of recognition,? by 
losing which Sakuntal3 unhappily brings upon the king his involuntary 

All the sandhyafigas, the plot-components which are prescribed act 
by act for a play in the BANS, are to be employed only insofar as they 
are consistent with the rosa which the author intends to display. In 
this regard Ananda very justly praises the Ratnávali and reprehends 
the Venisamhüra. In the latter play Duryodhana suddenly exhibits 
amorousness (tilása) in the second act, which is otherwise filled with 
preparations for war and vengeance, simply because Bharata prescribed 
uildsa as a sandhyariga of second acts. As Abhinava puts it, the au- 
thor should have taken the word vilàsa in a wider meaning and have 
depicted in Duryodhana a yearning not for sex but for some goal more 
appropriate to the spirit of the play. 

The concept of appropriateness was further elaborated by later critics 
of the Kashmir school. Ksemendra in the generation following Abhi- 
nava wrote an entire treatise on the subject. What is characteristic of 
Ananda's treatment, and what | would emphasize in taking a view of 
his work as a whole, is that he always associates his appropriateness 
closely with rasa. [n great literature the words must be appropriate 
to the plot, the characters, the immediate situation, but they become 
appropriate only through their enabling these factors to build up to the 
intended rasa. 

Ananda brings also other, older elements of the critical tradition i 
a subservience to the same final goal. He redefines the old qualiti 
of sweetness and strength by treating them as ornaments of partic- 
ular rosas. Sweetness is what ornaments srrigára, whereas strength 
ornaments the rasa of fury (2.6 A to 2.9 A). Style (sarighatand, riti) 
also is influenced by the rasa, as it is in intimate connection with the 
qualities, but a sparingness or frequency of compound word structure 
is no sure guide, in Ananda's opinion, to the presence or absence of 

? Wipternitz and others supposed that Kālidāsa bad taken the story of the ring 
from tbe Sratukhanda of tbe Padmapuràno. The Srstikhanda occurs only in the 
Bengali version of the Pad.P. and is a late Vaignava reworking of an earlier text. 
The reworking is later even than the Muslim conquest of Benares. The Sakuntalà 
story occurs i0 chapters 1-6, which are part of the Vaisnave reworking. They are 
therefore likely to be a derivative of Kalidasa's play and are certainly aot its source. 
See Asoke Chatterjee. Padma-Purdna: A Study. Calcutta Sanskrit College Research 
Series LVIII (1967), pp. 104 8., especially p. 114. 

22 Introduction 

sweetness or strength. Furthermore style is influenced by genre. The 
single stanza (muktaka) offers less scope for a heavy style than linked 
stanzas (e.g., the kulaka). The heavily ornamented compound style is 
especially appropriate to the prose romance (3.8). 

Ánanda's critique with its emphasis on rasa offers for the first time a 
criterion for the figures of speech. A figure of speech is well constructed 
when it strengthens the rasa. To do this it must not be overworked. 
Ananda quotes a verse (2.18-19e A) which begins, “In anger she has 
bound him tightly in the noose of her soft arms." Abhinava remarks 
that "were one to continue the metaphor furnished by the woman's 
creeper-like arms acting as a noose for binding, the woman would be- 
come a huntress, the bedroom would become a prison or cage, and 
so on, all of which would be most inappropriate." In general, figures 
should never be so elaborate as to take either the poet's mind or the 
reader's mind off the main goal, which is rasa. As Horace would say, 
there should be no purple patches. 

In several passages (e.g., 2.38. and 3.348.) Ananda distinguishes be- 
tween dhvani, as a suggestion which furnishes the predominant mean- 
ing of a sentence, and a subordinated form of suggestion which he calls 
gunibhütavyarigya. Among his examples of the latter type is a stanza 
which he quotes twice (at 3.34 A and 1.13e A). 

The sunset is flushed with red; the day goes ever before. 
Ab, such is the way of fate that never the two shall meet. 

As the Sanskrit word for sunset is feminine (sandhyá) and the word for 
day masculine (divasah), the suggestion arises of two lovers prevented 
by adverse fate from ever joining. But the stanza is obviously from 
a description of sunset. The literal sense remains predominant. The 
suggestion functions as a figure of speech.? 

This distinction has often been misunderstood by modern Sanskrit 
scholars and among them by some of the best. Jacobi, in the introduc- 
tion to his admirable translation of the Dhvanyáloka (ZDMG 56 [1902], 
p. 400), speaks of the poetry of subordinated suggestion as "eine Poe- 
sie zweiter Güte," a phrase repeated by Winternitz twenty years later.* 
S. K. De, in his History of Sanskrit Poetics (II, p. 162), uses the same 
pejorative. "By the side of dhvani kávya,"? he writes, "in which the 

J There was argument over whether to call the figure samásokti or dksepa. See 
notes on tbe passages where the stanza is quoted 

* Geschichte der indischen Literatur III, p. 18. 

? A phrase, by tbe way, which Ananda nowhere uses. 

Introduction 23 

suggested sense is predominant, we have poetry of second-rate excel- 
lence, designated guni-bhuta-vyarigya kávya, in which the unexpressed 
plays a subordinate part.” 

This error should be corrected, for nowhere in the Dhvanyáloka does 
Ananda characterize the poetry of gunibhutavyarigya as second-rate. 
That characterization appears first in Mammata (1.4-5), who speaks 
of madhyamam kávyam as opposed to uttamam. Mammata threw all 
cases where the suggestion was obvious (agüdha) or not beautiful (asun- 
dara) into the category of gunibhütavyarigya. Ananda, on the other 
hand, refers to subordinated suggestion (3.36 b A) as having been used 
by the great poets and states that it can be extremely beautiful and 
should be studied by sensitive readers. The very examples which he 
gives of gunibhutavyarigya (e.g., at 1.13d, e, 3.39, 3.40) should inform 
the reader of his evaluation, for they are among the most beautiful 
stanzas in the. whole book. I shall not quote them here, as the reader, 
if curious. can look them up in the translation. 

What prompted Ánanda to make this distinction between dhvani 
and subordinated suggestion was an historical fact, not an aesthetic 
judgment. Many cases of suggestion had been preempted by the older 
poeticians, especially Udbhata, under their definitions of the figures of 
speech. Thus, according to Udbhata, the figure sarnásokti (compound 
statement) occurs where from a description of the object-in-hand (pra- 
stutártha), that is, the primary object of the sentence, one understands 
some other object (Induràja 2.10; Vivrti 2.21). An example would be 
the little stanza which I have just quoted, “The sunset is flushed with 
red." The figure aprastutaprasamsá (reference by means of the ex- 
traneous; in some instances equivalent to allegory) occurs where from 
ap extraneous object (aprastutártha) we understand the object that 
the poet really has in mind (Induraja 5.8; Vivrti 5.14). Now it was 
Ananda’s goal to break away from the tradition of figures of speech, to 
set up suggestion (dhvani) as an independent power of words, and to 
establish the suggested meaning as the soul of poetry. As the Sanskrit 
term for a figure of speech (alarikdra) means literally an ornament, 
Ananda was also faced with the logical problem of how the soul could 
act as an ornament. One might conceive of the soul's being orna- 
mented, say by its body or its virtues, but by what sort of logic could 
a primary element, the thing-to-be-ornamented (alarikárya), itself act 
as an ornament? 

Ananda’s solution to the problem was to relegate all instances of 
suggestion which had been included in the figures of speech by the 

24 Introduction 

older critics? to a subordinate position. In that position they could 
very well serve as ornaments. An example will show the method which 
he followed and its success in achieving his goal. 

There is a famous stanza, written perhaps by Bana but included in 
the collection of Amaru, which likens the purifying power of God to 
the fire by which he destroyed the triple citadel of the demons.’ 

The women of the Triple City wept from lotus eyes 

as Sambbu’s arrow-flame embraced them; 

but still, though shaken off, the fire caught their hands, 
though struck, did pluck their garments' hem, 

denied, it seized their hair, and, scorned 

like lover who has lately loved another, lay before their f t. 
May this same fire burn away your sins. 

In this stanza. as Ananda remarks, the description of the demon 
women suggests that unhappy variety of $rrigára which is so close 
to tragedy, namely love-in-separation, here brought about by jealous 
anger. But this flavor (irsyávipralambhasrrigárarosa) is not the fi- 
nal aim or meaning of the stanza, which is rather the extraordinary 
power of God. As the suggested rusa of love is not the final aim, he 
characterizes it as subordinated suggestion, not dhvani in the strict 
sense. This subordinated element can logically act as an ornament. 
Bhámaha's definition of rasavadalarikára is thereby maintained witb- 
out injury to Ananda’s new doctrine of dÀvoni. But there is nothing 
second-rate about such instances of subordinated suggestion. He calls 
such instances "derivative of dhvani" (dhvaninisyanda, 3.36b A and 
3.41-42b A) and remarks under 3.40 that such instances “may again 
turn into dhvani when regarded from the viewpoint of (the final] rosa." 
In the stanza just quoted one may take the final meaning to be the 
rasa of God's heroism or the rasa of the worshiper's love of God. The 
historical reasons which prompted Ananda to make the distinction be- 
tween predominant and subordinated suggestion are no discovery of 
mine. They were noticed by both Jacobi and S. K. De. 

One final question must be raised and answered concerning Ánanda 
before [ move on to the period of his commentator Abhinava. This is 
the question whether he was the author of the whole of the Dhvanyáloka 
or of only a part of it. 

Introduction 25 

The text of the Dhvanydloka consists of 138 Kárikás,! written in 
simple verses (almost all in sloka or Gryd meter) and intended to be 
memorized. Expatiating on the verses is a prose commentary ( Vrtti), 
approximately twenty times their length. Within this prose commen- 
tary again are some twenty-nine simple verses, indistinguishable in style 
from the Karikds, but usually introduced with some such remark as 
"This is a supportive (parikara) stanza," or “Herewith a summariz- 
ing (sanksepa, sarigraha) stanza.” These supportive and summarizing 
stanzas can also be distinguished from the Kdérikés by the fact that 
the prose Vrtti never comments on them. In the case of two of them 
(3.41-42a A), which are not introduced by the usual remark but by 
the simple phrase, "it is stated," Abhinava tells us specifically that the 
matter is stated by the author of the Vrtti. 

Until the mid-nineteenth century it was always supposed in India 
that all this material was the work of one man, Anandavardhana. And 
this is quite in keeping with the form in which other Sanskrit treatises 
on literary criticism have been handed down. Kuntaka's Vokroktijivita, 
written in the age of Abhinava, comes to us in precisely the same 
combination of Kdnkds, prose commentary, and supplemental stanzas. 
Except for the absence of supplemental verses the same is true of the 
text of Mammata, Ruyyaka, Hemacandra, and Vigvanatha. Not only 
literary criticism but Sanskrit treatises on most scientific or philosoph- 
ical subjects tended to be composed in this form. One may instance 
the works of Bhartrhari the grammarian, Kumarila the ritualist, Dhar- 
makirti the Buddhist. 

Then in the mid-nineteenth century a remark of Georg Bühler? sug- 
gested that the Karikds of the Dhvanyáloka might be by an older au- 
thor and only the Vrtti by Ananda. Thirty years later Hermann Jacobi 
took up the question in earnest.’ He pointed out that the question 
of dhvani must have been discussed for many years prior to the time 
of Ananda. This follows from the variety of opinions on the subject 
of dhvani which Karikà 1.1 ascribes to "others." So far so good. But 

® In the Kashi Series text, upon which we bave based our translation, the Kárikds 
come to a total of 142. But three of them (4.4, 4.9, and 4.10) are almost surely not 
intended as Kdnkds but as summary or supportive stanzas, while 3.5 is a quotation 
from some other author. See notes to tbe translation of those passages. 

* "Detailed Report of a Tour in Search of Sanskrit MSS in Kashmir, Rajputana 
and Central India,” Journal of the Royal Asiatic Society of Bombay, extra oum- 
ber 1877, p. 69. 

10 See especially the i 

26 Introduction 

Jacobi went on to claim that Abbinava, the oldest commentator that 
we have on the text, furnishes evidence of a dual authorship. 

With the adherence of other Sanskritists to the theory of dual au- 
thorship, a vast amount of ink has flowed on the question. It was 
pointed out that the Abhidhavrttimátrká, written in about A.D. 900- 
925, attributed the doctrine of dhvani to sahrdayàh ("connoisseurs," 
or possibly "the honorable connoisseur"), whereupon one misguided 
scholar claimed that the name of the author of the Kdrikds must have 
been Sahrdaya. On the other hand, it was established that Rajasekhara 
about the same time attributed one of the supplemental slokas of the 
Dhvanyáloka to Ananda and, in a verse preserved in the Süktimuk- 
távoli (4.78), attributed to Ananda the whole introduction of dhvani 
into poetics. 

The student who would examine all the arguments which can be ad- 
duced for dual authorship should consult Kane's HSP, pp. 153-190, 
where that great scholar, like the lawyer that he was, gives a full-dress 
argument in its favor. Almost all of Kane's arguments, like those of 
Jacobi, are based on the remarks of Abhinavagupta, who frequently 
supplements such a phrase of the Vrtti as idam pratipáditam ("this 
has been stated") by some such addition as urttikarena ("by the au- 
thor of the Vrtti") or by asmanmülagranthakárena (“by the author of 
the basic text [i.e., of the Kdrikas|”). From such passages Kane ar- 
gues that Abhinava regarded the two portions as written by different 
authors. Against his view, Dr. Satkari Mookerjee (B. C. Law Vol- 
umes, I, pp. 179-194), followed by Dr. K. Krishnamoorthy (JHQ 24, 
pp. 180-194 and 300-311), has argued that Abbinava is merely dis- 
tinguisbing the different functions of one man. To me the arguments 
on both sides are inconclusive because Abhinava is so inconsistent. 
There is no doubt that he regarded Ananda as the author of the Vrtti. 
There are passages in the Locana, I admit, where he seems to regard 
the author of the Kérikds as someone else. But then in his Abhinava- 
bhárati (Vol. 2, pp. 299-300) he explicitly ascribes two of the Kdrikds of 
the Dhvanyáloka to Ananda.!! I am not at all confident that Abhinava 
had any historical knowledge on the matter. A man who could speak 
of Manoratha as a contemporary of Ananda (1.1c L) and who confuses 

?! Kane tries to explain away this inconsisteacy by saying that Bhattatauta was 
Abbinava's teacher in BANS, whereas Bhattendurája was his teacher in Dhu., and 
that in both cases Abhinava merely followed bis teacher's opinion. But that argu- 
ment is destructive of Kane's goal. Bhattatauta's opinion would be more valuable 
than Abhinava's and no less valuable than Bhattendurdja's 

Introduction 27 

Abhinanda with his father Jayantabbatta (3.7 L, but cf. note 3) is not 
to be much trusted in matters of history. 

Two considerations persuade me of the single authorship of the Dhva- 
nyáloka. First, there is not a single instance in the Vrtti of substantial 
disagreement with the Karikds. There is not even a case where the Vrtti 
interprets a Kérikd in a forced or unnatural manner. This is rarely the 
case where one Sanskrit critic comments on the work of another. There 
is much matter and long arguments in the Vrtti which are not in the 
Kánrikás, it is true. If there were not, there would have been no purpose 
in writing the Vrtti. But these matters and arguments are auxiliary. 
They do not change the basic system. 

Second, if some earlier genius had established the system of dhvani 
and the general critique of literature in terms of dhvani and rasa which 
is found in the Karkés, I find it inconceivable that a later author should 
not have given some praise, some respect, to him. indeed that be should 
not even have mentioned bis name. Lmportant texts are never treated 
by the Sanskrit tradition as anonymous. They always carry the name 
of an author, even if modern scholarship may prove that the name is 
mistaken or fictitious. If the Kárikás are not by Ananda, his silence 
regarding their authorship would be an instance of disrespect to an 
intellectual master without parallel in Sanskrit literature.!? 

N Professor Patwardhan has called my attention to an article by Dr. Senarat 
Paracavitana, "The Dhvanyóloko in Gfteeath century Ceylon,” JAOS 94 (1971), 
pp. 131-133. The article contains the text and translation of a Sanskrit inscrip 
tion giving a thesis (sthópano) upheld in debate by a scholar at the court of King 
Parákramabáhu VI (a.D. 1412-1467) to the effect that the Dhvanikdrikds were writ- 
ten by a Buddhist named Dbarmadása. The debater's evidence consists in bis 
statement that a manuscript of the Dhvanskdrikds in the library of tbe King of 
Suvarnadvipa (Sumatra) bore on its last page the statement " Dharmnodása-pandita- 
viracitam." The debater seems not to have seen this manuscript bimself, but to 
have heard of it from a RAdhákrspa-pandita, who gave the further information that 
the first twenty Kárxküs of the manuscript were not to be found in copies of the 
Dhvonikárikós in India. In these verses the author divided semantic powers into 
arthasakti and vyarjanasokti from the latter of which springs dhvant The debater 
argues that Anandavardhana left out these verses because they were too obviously 
connected with the tradition by which the Buddha is said to bave taught sdrthah 
savyañjano dharmah. 

Professor Patwardhan, | think, gives more credence to this thesis than I do. If 
there was indeed such a manuscript in a royal library of Sumatra, I sbould think it 
must bave been a Buddhist reworking of the Dv. verses. The norma) meaning of 
arthato vyarjanato dharmadesond in Buddhist texts is “the teaching of the letter 
and spirit of the Law." | much doubt that these terms would have been applied 

28 Introduction 

Soon after the death of King Avantivarman (A.D. 883) literature seems 
to have lost its royal patronage in Kashmir. We are told of one learned 
brahmin, Nayaka,'? who was given charge of a newly erected temple 
(Ràj.T. 5.159), but the court poets, such as Bhallata (5.204), fared 
badly. The favorites of the new king Saàkaravarman (A.D. 883-902) 
were men of low birth. Kalhana speaks with scorn of Saàkaravarman's 
ignorance of Sanskrit, claiming that he spoke "an Apabhramsa dialect 
worthy of a drunkard" (5.206). 

With the death of Saükaravarman things went from bad to worse. 
The history of Kashmir in the tenth century falls roughly into two 
parts. The first half saw the breakdown of royal administration and 
power under the demands of the Tantrin footsoldiers who time and 
again sold the throne to the bigbest bidder. Then from about the 
middle of the century the guidance of political affairs passed into the 
hands of the terrible Diddà. Diddà was born a Khasa princess. Her 
father beld the fortress of Lohara on the main route from Kashmir to 
the Punjab. On her mother’s side she was descended from the Shahi 
kings of Und and Kabul. Outliving her royal busband, Diddà governed 
for some years in the name of her child son, securing her own safety 
by fomenting discord among the military and political factions. When 
the son died not long after coming of age, she established a grandson 
in bis place. There were three of these little grandsons whom the 
unnatural Didd& placed on the throne only to murder each child after 
his enjoying for a few years the titular sovereignty. Finally, from 980— 
1003 she assumed the royal title in her own right, governing with the 
aid of her paramour Tu&ga whom she had elevated from the peasantry. 
In the end she left a strong kingdom to a nephew whom she had chosen 
by carefully testing bim against other candidates. And so began the 
Lohara Dynasty with a return of prosperity under two long-reigning 

Because of the withdrawal of court patronage, court literature vir- 
tually disappears from Kashmir during the tenth century. From this 
century in Kashmir we have no plays, no Sanskrit lyrics. The only 

to the technical study of semantics without a stimulus from non-Buddhist sources. 
And the total silence of India about Dbarmadasa's authorsbip of the work strikes 
me as strong evidence against the thesis. Kashmiri brahmins of Ananda's time, 
includiog Ananda himself, showed no prejudice against Buddhist authors nor any 
desire to hide Buddhist ideas. 

1 The name and the date suggest an identification with Bbaan&yaka, of whom 
I shall have more to say. But "Náyaka" was not an uncommon name. 

Introduction 29 

mahakavya that we have from this period is Abhinanda's Kádambari- 
kathdsdra, a work which retells in verse what Bana in a former century 
had told better in prose.!* The traditions of Sanskrit scholarship, how- 
ever, were not broken. The brahmins living in the capital or on their 
tax-free grants of land saw that their sons were taught Sanskrit gram- 
mar and the traditional Sanskrit sciences, in many cases teaching their 
sons themselves. The tradition was especially well maintained in Saiva 
philosophy and literary criticism. 

Saiva philosophy owes its origin in Kashmir to two sages of the ninth 
century, Somananda and Vasugupta. The views of the former were 
developed by his son Utpala into the doctrine of recognition (praty- 
abhijná); those of the latter, with the help of his disciple Kallata, 
into the doctrine of cosmic vibration (spanda). These two branches 
of philosophy were preserved during the difficult years of the tenth 
century, the former by Utpala's son Laksmanagupta, who became one 
of Abhinava's teachers. The school of vibration had a more checkered 
career, for Kallata's son, Mukula, seems to have turned away from 
philosophy toward literary criticism. His one surviving work, the Abhi- 
dhávrttimátrkà, is concerned with the nature of denotation and the 
secondary use of words. His son, Pratihara Induraja, followed in his 
steps and wrote a commentary on Udbhata. Meanwhile, the school of 
vibration was carried on in the family of a scholar named Bhütirája, 
who also taught the Krama Tantras and, as an old man one presumes, 
taught tantrism to Abhinava. Bhütirája's son, Bhattenduraja,'® also 
taught Abhinava in other subjects, notably in the Bhagavadgita and 
the Dhvanyaloka.'* 

Among these brahmin scholars with their thoughts turned away from 
politics to mystic philosophy and literature the Dhvanyáloka was much 
studied. The first commentary on the Dhvanyéloka, now lost, was 
called the Candrikà. It was written by some member of Abhinava's 
family, to whom Abhinava often refers but never by name. Sometimes 
he calls him "the author of the Candrikà" (3.26b L), sometimes "a 

'* At the end of the ninth century, Abhinanda's father, Jayantabhatta, had writ- 
ten an admirable work, the Nydyamarijart Jayanta not only gives a lucid account 
of the Ny&ya system; he writes with style and with the true Kashmiri gift for satire. 

'? Bhatteoduréja, Bbütirája's son and the teacher of Abhinava, must not be con- 
fused with Pratibàrendurija, the son of Mukula and the oldest commentator on 
Udbbata. See Kane, HSP. 

'* A brother of this Bhattendurdja, it appears, was Helarája, the well-known 
grammarian. It was a learned family iodeed. 

30 Introduction 

commentator, an older member of my family” (3.24 a L, 3.40 L). Always 
his references carry a criticism: "The author of the Candriká who 
could easily fail to see an elephant in front of bis eyes” (3.33b L): 
“A certain commentator now enough of arguing with persons who 
think themselves wise but whose references are wrong” (3.4a L). If 
one glances through all such passages, one will find that the author 
of the Candrikà usually chose the simple or natural meaning whereas 
Abhinava gives a more subtle interpretation. In several cases I think 
the Candrikà came closer to what Ananda actually meant (e.g., 3.33b L 
and 1.1b L). 

Another scholar before Abhinava's time occupied himself with the 
Dhvanydloke but for a different purpose. Bhattanàyaka's intention was 
to demolish the concept of dhvant. He seems to have gone through the 
book systematically, examining Anaoda's examples, showing how each 
one might be explained without reference to the new concept. Bhatta- 
nàyaka's work was called the Hrdayadarpana, and it too is now lost. 
But Abhinava describes for us, largely in his opponent's own words, 
the theory which Bhattanayaka hoped to substitute for that which he 
destroyed. I shall treat of it in dealing with Abbinava's doctrine of rasa. 
for Abhinava, although he vigorously opposed Bhattanayaka, borrowed 
from him not a little. 

Abhinavagupta was born about the middle of the tenth century into 
a learned family that descended from a brahmin named Atrigupta, who 
had been brought to Kashmir from Kanauj by King Lalitaditya after 
his conquest of that city." The king had given him a dwelling place 
in bis capital of Pravarapura (the modern Srinagar) on the bank of 
the Jhelum river facing the Saiva temple of Sitamsumaulin.'® From his 
loving description of its environs one infers that Abhinava had lived 
in that ancestral mansion at least as a child. His father, whose proper 
name was Narasimha, but who was popularly known as Cukhalaka, was 
an ardent devotee of Siva.!? Becoming a vairágin by strenuous asceti- 
cism, he overcame the miseries of worldly existence. Before departing 
from the world, however, he introduced Abhinava, and presumably his 
younger brother also, to Sanskrit grammar. The brother, Manoratha, 
was to be the first of Abhinava's disciples.?? 

Introduction 31 

The number of Abhinava's works is large. Even the preserved works 
very nearly fill a half shelf in my library. To these must be added a 
considerable number of lost works, of which we know tbe title or subject 
matter from references in the works which are preserved. As a complete 
bibliography is available in V. Raghavan's Catalogus Catalogorum, and 
as details on the subject matter of most of the works may be found in 
Pandey's Abhinavagupta, | shall give only an outline. 

In general Abbinava's oeuvre falls into three parts. (1) Commen- 
taries on the Tantras and surveys of their doctrines. The greatest of 
these works is the huge Tantráloka, published in twelve volumes in the 
Kashmir Series of Tezts and Studies. and the summary, Tontrosára, 
published in the same series and now available in an Italian translation 
by Raniero Gnoli." (2) Works on literary criticism. The first of these 
was a commentary on the Kdvyakautuka of his teacher Bhattatauta. 
Both it and the work on which it commented are known only from ref- 
erences and quotations,?? chiefly from the Locana and the Abhinava- 
bhdrati. The Locana?’ must have come second, for Abhinava refers 
to it in the 4bhinavabhároti. The latter work is Abhinava's commen- 
tary on the BANS. An almost complete, though sadly corrupt, text 
of the ABA. is now available in the GOS.?* (3) Commentaries of the 
Recognition (pratyabhijriá) School of philosophy. There are three of 
these. On Somànanda's Porátrimsikàvivrti (the Pardtrimsikd, or verses 
on tbe ultimate, form the final portion of the Rudrayamalatantra), 
Abhinava wrote the Parátrimsikátattvavivarana, or Anuttaratrimsikd- 
vitrti (KSTS 7, 18). On Utpala's /[svarapratyabhijnásütra he wrote 
the Isvarapratyabhijniávimarsini, also called "the Small Commentary" 
(Laghuvivrti), published as KSTS 22 and 32. On Utpala's /svara- 
pratyabhijridtikd he wrote the Isvarapratyabhijnávivrtivimarsini, called 
also "the Great Commentary" (brhati vimarsini), KSTS 60 and 62. 

?! “Essenza dei Tantra," Encyclopedia de autor cl sici, No. 38, Boringhieri, 
Turia, 1960. 

2 Gathered by P. V. Kane. HSP, pp. 209-212. 

a The Locona bas been printed several times. We bave used the Kasbi text 

the basis for our translation. Although it is marred by sumerous misprints 
it carries the valuable Bálapriyó commentary, which supplements the still more 
valuable Kaumudi, available only on Chapter One. For particulars of these texts see 
Abbreviations and Works Cited. 

™ One should use the second edition of tbe first volume, as it contains numerous 
improvements in the text. Four chapters (6, 7, 18, 19) have also been edited by 
R. P. Kaágle with a Marathi translation aod commentary. His emendations of the 
text will be useful even to those who cannot read Marathi. 

32 Introduction 

In addition to writings on these three subjects Abhinava was the au- 
thor of numerous religious verses (stotras); of the Paramárthasára,"* a 
Saiva reworking of a Vaisnava text; of the Bhagavadgitarthasarigroha,”* 
a brief collection of notes on important passages of the Bhagavadgità; 
and of a commentary on the Ghatakarpara. 

The order in which Abhinava wrote these works is not quite cer- 
tain. He gives the dates of completion of only three of them. The 
Ksemastotra was completed in A.D. 990, the Bhairavastotra in 992, 
and the Jsvarapratyabhijtiávivrtivimarsini in 1014. K. C. Pandey in his 
monograph on Abhinavagupta believed that a period of Tantric studies 
in Abhinava's youth was followed by his work on literary criticism and 
this in turn by a final period of concern with Pratyabhijàà philosophy. 
The linchpin of his belief was a reference in the Locana which he be- 
lieved was to the Tantraáloka. But this linchpin has now broken. The 
reference exists only in the false reading of the NS edition (p. 19) of 
that work.?” Furthermore, the hypothesis does not agree with the gen- 
eral order of Abhinava’s intellectual interests which seems to be given 
in the biographical information at the end of the Tantrdloka: 

[The author] was introduced into the forest of grammar by his father, had his 
mind clarified by a few drops of the sea of logic, and, when intent on enjoying 
the full rosa of literature, was seized with an intoxicating devotion to Siva. 
Being wholly filled with that, he no longer cared for any worldly pursuit, until, 
in order to increase his enjoyment of that devotion, he went to serve in the 
houses of [religious] masters. ( Tantrdloka 37.58-59) 

There follow the names of a great many of his teachers, among which 
is the name Bhütir&jatanaya, "the son of Bhitiraja,” that is to say, 
Bhattenduràja, Abhinava's master in the Dhvanyáloka, as also the 
name of Laksmanagupta,” his teacher in Pratyabhijiia philosophy. The 

33 KSTS 7. The work bas been translated by L. D. Barnett, JRAS 1910, pp. 718- 
47, and by L. Silburn, Paris 1957. 

™ Translated by Arvind Sharma, Leiden: E. J. Brill, 1983 

?' The false reading is as follows: toduttirnatve tu sarvam paramesvarddvayam 
brahmety osmocchástrünusárena viditam tanirdlokagrantham wicdraya The cor- 
rect reading, given by the Kashi edition, p. 67, carries a very diferent meaning: 
taduttimatve tu sarvam brohmety asmacchastrakdrena na na viditam tattudloka- 
grantham viracayatà. Thus the reference is to a work of Ananda called the Tattvá- 
loka, not to the Tantráloka of Abbinava. The Kashi reading is substantiated by 
Abhinava's later reference to this Tottvóloka under Chapter 4, Karikd S of the Lo- 
cana (Kasbi edition p. 533, line 5 of Locana): etoc ca granthakdrena tattudloke 

?! Laksmanagupta is also praised in TA 1.11 

Introduction 33 

Tantráloka and Tantrasára should therefore be placed among the last 
of his works. His initiation into Kaula tantrism by Sambhunátha and 
his düti Bhágavati (see TA 1.13 and commentary) seems to have given 
the final spiritual increment to his life. By it he became a siddha and 
there was nothing more to learn. 

Whether the literary studies followed or, as I should think, preceded 
the works on the philosophy of recognition, is not certain. In the Loca- 
na he quotes one verse of Utpala's /svarapratyabhijnid (see 1.8 L and 
note 3), but this proves only that he was acquainted with the works of 
Utpala at that time, not that he had already commented on them. In 
the Locana he comments on a verse from the Gità ( Bh.G. 2.69), giving 
much the same interpretation as that given in his Bhagovadygitórtha- 
sarigraha (see Dhv. 3.1 b L and note 3). This likewise does not prove 
that he had already written tbat work. 

It would take more space than I propose to allow myself and more 
knowledge than I possess to give an account of Abhinava's entire con- 
tribution to Indian thought. What I shall have to say in the following 
pages is limited to the contribution furnished by the Locana. 

Abhinava chose the title Locana for his commentary on the Sohrdayd- 
loka (Dhvanydloka) because he intended it to serve as an "eye" by 
which one could see the “light for connoisseurs” which Ananda had 
furnished. If we are to be fair to him, it is by his achievement of that 
purpose that we must judge him, not by modern standards of historical 
or philological accuracy, nor even by the criterion of originality. 

+ Like all Sanskrit commentators Abhinava had no interest in history. I 
have already noticed two of his historical errors. Just how uninterested 
he was in history appears still more clearly from his comment on the 
first Kàrikà In order to justify the perfect tense of the verbs there 
used (tasydbhdvam jagadur apare, etc.) Abhinava says, “The author 
had not actually heard the alternative views of those who deny the 
existence of suggestion. Rather, be will imagine such ideas and then 
refute them. Hence his use of the perfect tense,” for Pānini prescribes 
the perfect to be used in the Bhdsd only for that which occurred in the 
past outside the range of one's experience. Abhinava would rather allow 
his author to have been obtuse than to have committed a grammatical 
error. Frequently throughout the Locana Abhinava will depart from 
& natural interpretation of Ananda's words in order to save him from 
some inconsistency (see, for example, 1.13i or 2.26 A, note 2). He will 
even exclude an example as being spurious if he believes it has been 
improperly adduced (see 2.27a A, note 2 and 2.27 a L). 

34 Introduction 

As against such instances which are faults by a standard foreign to 
the Sanskrit tradition of scholarship, Abbinava exhibits a high degree 
of skill in those abilities which his own tradition sought to develop. He 
is an impeccable grammarian, possesses a sound knowledge of Nyàya 
and Mimàmsá, and has the works of the older poeticians by heart. He 
uses his formidable education not in order to show off but to give the 
reader an accurate understanding of Ananda's critique. In example af- 
ter example he points out just where the suggestion lies, the range of 
its meaning, and often just what it is that gives to it its camatkrti, its 
sudden effect of delight. For examples see his comments on the stanza 
"White herons circle against dark clouds” (snigdhasyámalakantilipta- 
viyuti, 2.1a), "Why do you laugh" (kim hásyena, 2.5 b), "Say, happy 
friend, if all is well" (gopavadhüvilásasuhrdám, 2.5f). Such careful aes- 
thetic ezplications de tezte had just come into vogue. We find the fash- 
ion also in Abhinava's contemporary Kuntaka. I know of no examples 
in the older literature. But, once established, it became characteristic 
of Sanskrit literary criticism and is what gives to that tradition of crit- 
icism its great strength. [n our Western classical tradition there is 
nothing to compare with it except pseudo- Longinus. 

This careful analysis of Ananda's examples leads Abhinava at times 
to remarkable discoveries. I may here point out just one of these. 
In 2.9 Ananda cites two examples of the rasa of fury (raudrarasa), 
the first composed in the style of long compounds traditionally as- 
sociated with that rasa, the other in a non-compounded analytical 
style. The point that Ananda would make in the passage is that 
either style may serve the purpose of suggesting fury. He adduces 
the two stanzas as examples and leaves it to the reader to savor the 
effect. Abbinava analyses the examples. Both examples are taken 
from that drama of vengeance, the Venisamhára. In the first it is the 
hero Bhima who speaks. He vows to crush the thighs of Suyodhana, 
who had dragged Queen Draupadi through the Kuru assembly. Then 
he will deck Draupadi's hair with his “hands new-reddened in that 
fresh-congealing blood." The stanza begins with an immense com- 
pound: caricadbhujabhramitacandagadábhighatasancürnitoruyugalasya 
suyodhanasya (literally, “of the by-my-whirling-arm-held-brutal-war- 
club-stroke-crushed-thighed Suyodhana"). After commenting on the 
suggestions of the stanza, Abhinava remarks, “From the long com- 
pound, flowing in an uninterrupted stream and allowing the hearer no 
Pause in all its course, there results an apprehension of the whole scene 
as a unity up to the representation of the broken-thighed Suyodhana. 

Introduction 35 

This serves to intensify the impression of Bhima's violence."?? On the 
other hand, in the second, analytical, stanza adduced by Ananda, Abhi- 
nava sees the effect to lie in a protracted climax. He says, “Here the 
anger of the speaker rises to the highest pitch by a progression from 
word to word through meanings which, being presented separately, are 
reflected upon by the hearer in succession. And so the very absence of 
compounds acts as a cause of rising excitement.” When one reads such 
verses over after reading Abhinava's comments, one reads them with a 
new appreciation and with some degree of the excitement and delight 
that be found in them himself. This is the highest gift that a literary 
critic can possess in any tradition. 

In only one important respect did Abhinava change what he found 
in Ánanda's text. I refer to the new explanation he gave of rasa and 
of the psychological process by which it appears in the beart of the 
reader or the poet, for with Abbinava rasa is sbarply excluded from 
the character invented by the poet or portrayed on the stage by an 
actor. The experience of Rama on losing Sita is the emotion of grief 
(Soka). If the hearer of a poem or play were to experience the same 
emotion, he would close his book or leave the theater. There must be a 
qualitative difference between the stháyibháva and the rasa to explain 
how we can relish the tragic or the rasa of fear. 

Most of the components of Abhinava's new theory are borrowed, 
strange to say, from Ananda's chief critic. Bhattan3yaka had insisted 
that we do not perceive rasa as belonging to someone else, for in that 
case we would remain indifferent. Nor do we perceive it as belong- 
ing to ourselves—here Abhinava was to disagree—for the factors which 
Bharata says are productive of rasa, the determinants, as for example 
Sita and the abduction of Sita, work on Rama, not on us. Indeed, said 
Bhattandyaka, rosa is not perceived at all, it is simply enjoyed. What 
bappens is that in a great poem a second semantic operation comes into 
play by which denotation assumes a new dimension. Bhattanayaka gave 
this operation the name of bhdvand (aesthetic efficacy). He borrowed 
the term from the technical vocabulary of Mimamsa. where it is used 
of the efficacy residing in the verb of a Vedic sentence, which explains 
how that verb can bring an actor to pursue a given aim. By bhdvand 

?? Some years ago | made the following remarks on this stanza with reference 
to Abhinava's interpretation in a paper published in The Harvard Advocate (CXV, 
No. 4, summer 1982, p. 126). "The whole scene is before our eyes as soon as we 
understand the words at all. We are not allowed to dilute the effect by relishing it 
bit by bit. It hits the aesthetic sense not like pebbles but like a rock." 

36 Introduction 

the ritualists meant the efficacy of a Vedic command. Bhattanayaka 
then applied the term to poetry as the aesthetic efficacy of a particular 
combination of determinants and consequents. This aesthetic bhdvand, 
he claimed, has the effect of universalizing the determinants and other 
factors, so that they may bring about or realize a rasa. Upon the real- 
ization of the rasa, a third stage of the aesthetic process begins, namely 
enjoyment (bhoga), which Bhattandyaka regarded as springing from a 
third semantic power, bhogakrttva (enjoyment-efficacy). We enjoy the 
rasa in a manner different from our enjoyment of direct experience 
or of apprehensions derived from memory. This enjoyment takes the 
form of melting, expansion, and radiance, and is like the bliss that 
comes from realizing one's identity with the highest Brahman. This 
is the purpose of poetry. Any instruction that poetry may furnish is 

One should bear in mind that we know of Bhattanáyaka's theory 
only through the writings of his opponents, Abhinava and Abhinava's 
follower Mammata. The questions that arise in one's mind as to the 
exact nature of bhdvand and why it should work to the effect claimed 
for it are ones for which he may have given answers of which we are 
not told. Abhinava's most telling criticisms are two. First, that the 
newly extended use of the word bhávaná refers to notbing other than 
the suggestion of rosa (rasadhvani) described by Ananda. Second, 
that it cannot be that rasa is never perceived. We must perceive it, 
or we should be unable to discuss it. And granted that this perception 
may be of a different sort from sense perception, we perceive a rasa as 
belonging to us. 

Beyond these specific criticisms, however, the reader will be struck by 
how much of his rival's theory Abhinava incorporates into his own. He 
too excludes the actor and the portrayed character from enjoyment of 
rasa. What the character experiences is the basic emotion. As for the 
actor, if he experienced either the emotion or the rasa, he would forget 
his lines. Then too, although Abbinava holds to Ánanda's term of 
suggestion or dhvani, he sees in rosadhvani a transforming power that 
bears a close resemblance to Nàyaka's bhdvand. The grief, for example; 
that the observer perceives in the poetic character or the actor, by the 
observer's ruminating on its determinants and consequents, meets with 
a response from his heart in which he identifies that grief with the griefs 
in his own memory. Rasa is not simply the apprehension of another 
person's mental state. [t is rather a supernormal relishing based on 

Introduction 37 

an involved sympathy.°° By this sympathy, one might say, the reader 
or audience loses his own grief in the larger dimensions of compassion. 
Abhinava, like his opponent, sees rosa not as an object to be enjoyed 
but as the ongoing process of enjoyment itself. He too uses the word 
"melting" as one of its characteristics (see 1.5 L, note 3). He too is 
struck by the similarity of rasa to the relishing of the ultimate Brahman 
(2.4 L). One might say in sum that Abhinava has taken over most of the 
new ideas of Bhattanáyaka but trimmed them here and there so that 
they may fit into the terminology and the general view of Ananda. He 
even agrees at one point with the statement tbat enjoyment (he calls it 
bliss) is the chief goal of poetry, in comparison with which instruction 
is a far lesser goal (1.1e L). But later, in Chapter Three (3.10-14f), 
he brings this admission into harmony with a more orthodox view by 
attempting to show that delight and instruction are not different in 

Abhinava adds much that is not in Ananda. He adds arguments 
against all the opponents of dhvani not only against Bhattandyaka 
but against both schools of Mīmāmsā. He gives numerous examples 
of varieties of dhvani and of subordinated suggestion which Ananda 
had passed over. Interesting are his remarks on rosábhása, false or 
improper rasa. According to the tradition going back to BANS, a love 
that is not mutual. a love implemented by violence such as Ràvapa's 
love for Sità, or an adulterous love, is productive only of rasdbhésa. If 
one guides one's criticism strictly by the words of Bharata (BANS 1, 
p. 295) such false love should lead to comedy. In fact, says Abhinava, it 
may lead to comedy only at a time long after our experience. When we 
hear Ravana’s passionate words in the Rávanakávya (alas, now lost) 
there is no occasion for relishing comedy (2.3 L). This qualification 
opens up to favorable evaluation much that would have been rejected 
or reprehended by older standards. 

Since Sanskrit criticism has been accused of concentrating too much 
on the individual verse or stanza, one will take a special interest in 
Abhinava's tracing of the predominant rasa throughout a whole play, 
as he does with the Táposavatsarája (3.10-14 g L). His survey should 
impress the Western reader with the basic difference of movement in 
Sanskrit works of literature from that in European works. This differ- 
ence, between a tortuous or cyclical movernent, a periodic distancing 
from and return to the predominant theme, as against the climactic 

inava's argument with the Mimámss at 1.18 L. 

38 Introduction 

achievement of a final result,” is found, I believe, in all the traditional 
Indian arts. The Indian style is found only rarely in the West, as in 
the music of César Franck or in Joyce’s Finnegans Wake. 

To continue with a listing of interesting or illuminating passages in 
Abbinava’s commentary would be otiose. The translation is here for 
the reader to find them on his own. [ would sum up my opinion of 
Abhinava's work in the Locana by saying that he accomplishes ad- 
mirably what he set out to do. He expresses Ananda’s views more 
logically than Ananda expressed them; he defends them against those 
who had opposed them. Under the provocation of Bhattanáyaka he 
develops the concept of rasa into something different from what I be- 
lieve Ananda envisaged, but the new concept is more consistent than 
Apanda’s and was to become, after Mammata's incorporation of it in 
his Kávyaprakása, the leading view of rosa in Indian criticism. More 
than all this, Abhinava's analysis of Ananda’s examples will give the 
sensitive reader a hundred new insights into the beauty of Sanskrit 

There is nothing in our Western classical (Greek and Latin) tradition 
of criticism that corresponds to rasa and nothing that corresponds to 
dhvani in the grand dimensions in which Ananda and Abhinava con- 
ceived it. Our classical rhetoricians, all but one of them, chose the path 
that Bhàmaha and Dandin chose: they included such instances of sug- 
gestion as they recognized in the tropes and figures of speech.?? One 
man of this tradition, however, the author of the treatise On the Sub- 
lime, bad an uncanny skill at recognizing passages of literature which 
excite the reader or, as he put it, drive bim to ecstacy. I have noticed 
that almost every instance of what pseudo-Longinus? cites of what he 
calis "the sublime" in literature, is actually an instance of suggestion. 

M Sanskrit dramaturgy speaks of the achievement of a result (phalayoga), it is 
true. But the pha/ayoga, for example in the Tépasavatsardja, namely Udayana's 
retrieval of V&savadattà and acquisition of universal sovereignty, is aesthetically far 
less important than the recurring manifestations of frrigárorasa throughout the play. 

7 [n the Greek cudagcc, the significatio of the Rhetorica ad Herennium and of 
Quintilian, it is the figure itself that is usually striking. Only under irony and 
allegory does Greco-Latin rhetoric come to what would qualify with Ansnde as 
dhvani, and at that only as vastudAvani. 

» Manuscript P of the [Tepi vwov, from which apparently all our texts descend, 
refers to bim as “either Dionysios or Longinos.” He appears to have been a Jew and 
perbaps for that reason bad no following among the Greek and Roman rbetoricians. 
His fame began only with the rediscovery of bis work in Renaissance times. 

Introduction 39 

I translate here from just one of his instances, the famous gaiveral pou 
rivo ioo Üéo.aw : 

He seems to me the equal 

of the gods who sits beside you, 
listening to your sweet speech 

and to your lovely laughtet. 

which has quickened the heart in my breast. 
When I see you, Brochea, my voice 

leaves me, 

my tongue is broken. a thin Gre 

Tuns over my flesh, 

my eyes have no stren 

my ears ring. 
Longinus, if that was his name, says that the beauty of Sappho's poem 
comes from its arrangement and that the result is the sublime. Ananda 
would have said that the beauty comes from dhvani and that the re- 
sult is srrigárorasa. If only Longinus had had followers. they might 
have worked out a critique of literature not unlike that of Ananda and 



A Of Madhu's foe 
incarnate as a lion by bis will, 
may the claws. which put the moon to shame 
in purity and shape, 
by cutting off bis devotees' distress 
grant you protection.! 

1. In this benedictory verse which introduces his work Anandavardhana 
takes as subject an attribute of his chosen divinity (istodevotà) Nrsimba, the 
man-lion incarnation of Visnu (Madhu's foe). Vispu became incarnate as a 
roan-ljon in order to destroy tbe demon Hiranyakasipu and thereby remove the 
distress of the Vaisnava devotee Prahrada. For the story, see Bhag. Pur. 7.2-8. 
Benedictóry verses to Nrsimha usually take bis claws as their subject; see 
SRK 116, 128, 141. As is appropriate to a verse introducing a work on 
“suggestion (dAvani), the present verse contains numerous instances of dhvoni. 
Abhinava points out one occurrence of rasadhvani, five of vastudhvani, and 
three of alarikéradhvani. 

L Victorious is the Muse's double heart, 
the poet and tbe relisher of art: 
which has created brave new worlds from naught 
and even stones to flowing sap has brought, 
imparting beauty to all within its reach 
by successive flow of genius and of speech.’ 

At Bhattenduraja’s lotus feet | beard 
all that I know and love of letters; 
from tbat, echoing but a little portion, 
I, Abhinavagupta, shall explain 

with my own Eye the Light of Poetry. 



44 [$ 1.1 Introduction L 

Although the author of the commentary (vrttikára)? has himself ac- 
complished his own aim in life by his continuous and perfect devotion 
to the highest God, in order to achieve a desired purpose, namely, that 
teachers of his work be unhindered in their explanation and students be 
unhindered in their audition, he enlists God's attention to this purpose 
by publishing an appropriate benediction. "May the claws of the foe of 
Madhu protect you," that is, may they protect you teachers and stu- 
dents of this work; for teachers and students are the only appropriate 
objects of address here and the word "you" implies direct address. To 
“protect” is to furnish help toward attainment of a desired goal and 
that help comes about by such means as the opposing of hindrances. 
Such is the protection that is here meant. 

The heroic flavor (virarasa)* is here suggested by our apprehension 
of energy (utsdhe), an apprehension furnished by the association of 
God, who is constantly exerting himself [on behalf of mankind], with 
the characteristics of clarity of purpose aod diligent resolve. Again, 
as claws are weapons and as protection requires a weapon for instru- 
ment, the extraordinary power of these claws, which are identical with 
instruments, is suggested by treating them as agents. Also suggested* 
is the absence of dependence on extraneous instruments on the part of 
God. Tbe word “Madhuripu” conveys the fact that he is always active 
in removing whatever is a menace to men." 

What sort of Madburipu ("Madbu's foe”)? “Incarnate as a lion by 
bis own will," not because he is subject to his earlier deeds (karma), nor 
to tbe will of another; rather, he took the form of a lion in conformity 
with his entertaining a desire to do so, a desire which is appropriate 
to the killing of a particular demon. 

What sort of claws? Those which "cut off his devotees' distress." 
The power to cut is appropriate to claws and while it is impossible 
for [ordinary] claws to cut away mental anguish, it becomes possible 
in the case of these claws because God's creations are conformable to 
his desires. Or [we may take it as follows): Hirapyakasipu was the 
thorn of the three worlds, a torment to everyone, and so in reality he 
himself was pain in concrete form to those who come to God as their 
sole refuge. When these claws destroyed him, it was the very pain (of 
God's devotees] that was rooted out; so this shows how God, even in 
such a state? [i.e., even while engaged in an act of destruction], is still 
exceptionally compassionate. 

Furthermore, the claws by their svaccha, that is, their property of 
being pure, their purity—for words like svaccha and mrdu in their 

§ 1.1 Introduction Z } 45 

primary sense refer to abstract properties!°—and by their shape, which 
is curved and lovely, have “shamed,” that is, have distressed, the moon. 
It is here suggested by a suggestion based on the power of meaning’! 
that the moon is new. By the moon's being afflicted, it is suggested 
that it appears pale and devoid of charm in the presence of the claws; 
and it is generally well known that claws can cause pain. But in the 
case of the man-lion this property of causing pain has been presented 
in superhuman form [viz., by saying that His claws are pure and lovely 
although causing pain]. And so, when the new moon looks at the pu- 
rity and curved shape of these claws, he feels an inner pain [as follows]: 
"Although I may be equal so far as purity and my curved shape go, 
tbese claws are skillful in removing the distress of devotees, whereas I 
am not." [n this manner the figure of speech known as contrast (vyati- 
reka)'? is suggested. “Moreover, in tbe past I alone was desirable to 
all people because | was possessed of unequalled clarity and beautiful 
shape, but: now all these ten claws have the same sbape as the new 
moon and in addition to that they are skillful in removing both mental 
and physical pain and so now people regard them and not me with 
the respect due to the new moon.” Thinking in this manner, the new 
moon experiences, as it were, continual torment and so the two fig- 
ures of speech, fancy (utpreksá)!? and denial (apahnuti),'* have been 
Suggested. Thus in this verse the three varieties of suggestion, namely 
vastudhvant, alarikdradhvani, and rasadhvani, have been explained by 
my teacher. 

1. The inner glory of Lterature is here derived from its three characteristic 
abilities: to create (prothayitum) new worlds; to impart a relish (sárayitum) 
to even the dullest objects in the actual world, so that they may elicit ses- 
thetic response (rasa); and to render both these areas bright (udbhásayitum) 
with a beauty constructed out of the poet’s genius and the words with which 
he communicates. This much is a fine characterization. What is even finer 
in my opinion is Abbinava's realization that the beauty of poetry, or of art 
in general, depends upon the audience as much as on the artist. One may 
find echoes of this verse in the benedictions of later critics. Compare, in 
Mammata's benedictory verse to his Kdvyaprakdga, niyamarahitam with vind 
káranakalàm of our verse, or navarasarucirém with nijarosabharàt sdrayat. 
2. Here the oame of Anandavardhana's work is given as "The Light of Poetry” 
(Kavyáloka). Elsewhere Abhinava refers to it as the Sahrdaydloka; see Intro- 
duction, p. 12. Abhinava gives to his own commentary the title "The Eye," 
(locana) and by his words here indicates, according to the Kaumudi, that he 
will comment on only parts of the text, the parts that are difficult or subj t 

46 [$1.1 Introduction L 

to doubt. In the fourth line of the verse, janasya is proleptic genitive: I shall 
explain “to the world." — 3. Abhinava attributes the introductory stanza to 
the author of the Vrtti, not the author of the Kániàs. That the two were the 
same man bas been argued in the Introduction, pp. 25-27. — 4. BP notes that 
as Ànandavardhana proposes to show that suggestion is the soul of poetry, 
his introductory verse migbt naturally be supposed to contain suggestions; 
accordingly, Abhinava points out the three types of suggestions which it con- 
tains, beginning with the most important type, rasodhvani. For the technical 
terms rasa, stháyibháva, vibhàva, anubháva, and vyabhicánbhóva, see Intro- 
duction. Abhinava here discovers a suggestion of the heroic rasa. Of its com- 
ponents be points out the sthdyibhdva as the utsáha (energy) of Nrsimha. For 
utsáha as the stháyibháva of virarasa see BANS 6.67. The álambana-vibháva 
of course is Nrsimha himself. The vyabhicdribhdvas are given by Abhinava 
as the god's asammoha (clarity of purpose) and adhyavasáya (diligent re 
solve). According to BP, God's constantly exerting himself (nityodyogitva) is 
the anubhávo. $. The point is this. Usually the karana (instrument) and 
the kartr (agent) are kept distinct and given different case endings. NakAa 
is usually a korano, the means of accomplishing something, and not a kartr. 
Now a kartr is the chief of all the károkos; all other károkos are dependent 
(svotontrah kortá). Therefore by presenting the nakhas themselves as the 
agent, that is, by placing nokháh in tbe nominative instead of the instru- 
mental case, the extraordinary power of these claws is suggested. This is an 
instance of vastudhvoni. — 6. Dhvanitas co refers to a second vostudhvani. 
7. Tasya sadoiva jagattrdsa, etc., is the third vastudhvani. Abhinava has 
used the word ukta here to mean sūcıta Kaumudi, p. 19: vyorijita iti vak- 
tavye sati ukta its vacanam abhidhávyópóáragocaravat prakatapratipattikatvam 
vyarigyórthasya prodarsaystum; that is, ukta has been used in order to show 
that the matter is as clear as if it were conveyed by abhidhd itself. — 8. In 
the long compound uisistaddnava-...1 have taken the word ucita to modify 
icchà; it might equally wel) modify parigraha. The point is that the form of 
a man-lion was required for the slaying of this particular demon, who could 
not be slain by a god, a man, or an animal. Had Visnu wished to slay an- 
other demon, he would have wished for some other form. 9. Tasydm apy 
avastháyám means that he is compassionate even when he is engaged in the 
act of killing. This is Abbinava’s interesting explanation of the apparent con- 
tradiction in a verse like ksipto hastávalagnah (2.5c A), where it would appear 
that Siva is being cruel. Abbinava would argue that he is really acting out 
of compassion since he is acting for the sake of the world. not for the sake of 

10. It is a doctrine of the grammarians that adjectives (like fukla ^white") 
denote primarily a quality (e.g., suklatá "whiteness"). Only secondarily do 
they come to denote a substance qualified by a quality (as in "a white horse"). 
See Patafijali on Pao. 2.1.30. Accordingly, Abbinava is here analysing the 

§11 K] 47 

compound as "pained by their pure-ness and by their own (curved and lovely) 
shape,” rather than taking svaccha to modify svoccháyà. 11. This is an- 
other case of vastudhvani. The idea that the moon is new, i.e., at its most 
beautiful, on the first day of its appearance as a slender crescent, is not con- 
veyed by slesa and so is not an example of suggestion based on the power 
of a word. It is conveyed by the meaning of the words and so is an instance 
of arthasaktimüladhvani (suggestion based on the power of meaning). See 
220 K. 12. This is the first of the three cases of alarikáradhvani discov- 
ered by Abhinava in the verse. For definitions of vyatireka see Bhámaha 2.75, 
Dandin 2.180, Al.Sarv. p. 101. In vyatireka the upameya (the thing itself) is 
usually shown to be superior in some respect to the upamána (the thing to 
which it is likened); for example, "her face by being spotless puts to shame the 
moon with its spot.” 13. For utpreksd see Bhamaha 2.91, Dandin 2.221, 
Al.Sarv. p. 69. In utpreksá the possibility which one fancies is usually in cold 
fact an impossibility. It is of course impossible that anything should pain or 
shame the mogn, for the moon is an insentient obj t. — 14. For apahnuti see 
Bhamaha 3.21, Dandia 2.304, Al.Sorv. p. 63. Bhamaba’s example is: "The 
bee is not buzzing; this is the sound of Love's bow." In Ananda's verse the 
suggested apohnuti would take the form: "People deny me the status of new 
moon aad look only on these claws in that way." Note that all these figures 
in the present stanza are suggested, not directly conveyed. Thus they are 
examples of olarikáradhvani. 

K Some have said that the soul of poetry, which has been handed 
down from the past by wise men as “suggestion” (dhvani).' does not 
exist; others, that it is an associated meaning (bhdkta);? while some 
have said that its nature lies outside the scope of speech: of this (sug- 
gestion} we shall here state the true nature in order to delight the hearts 
of sensitive readers. 

1. The key word of the book, dhvani, is used in many different senses. Of 
the senses used by literary critics, as opposed to the grammarians (for whom 
see 1.131 A and L) Abhinava specifies the following (1.13 L): (a) fabda^: a 
word which gives rise to a suggestion; (b) arthah: a meaning which gives rise 
to a suggestion; (c) vydpdrah: the operation, the suggesting of the implicit 
meaning; (d) vyarigyam: the suggested meaning itself; (e) samuddyah: the 

48 [$11K 

group; or a poem which embodies all the above factors. K and A restrict 
the senses of the word to (c), (d), and (e) and. wherever they would be 
precise, specify that the suggestion involved must be the primary suggestion 
of the sentence. It is not immediately clear in which sense K is here using 
the word. The Vrtti in setting forth the frst argument (1.1a A) seems to 
take it in sense (c) or (d). In the second argument (1.1b A) it clearly takes 
it in sense (e). — 2. Ananda uses the word bhákta to cover both types of 
secondary or associated meaning: the metaphorical (which Abhinava calls 
gauna), as in “The boy is a tion,” where "lion" takes on secondary sense 
because of the boy's similarity in some respects to a lion, and the relational 
(which Abhinava calls láksanika), as in “Bring in the spears," where "spears" 
is used for "spearmen" because of some relation other than similarity between 
the two objects. Ánanda's general words for secondary usage or the secondary 
operation of a word are bhakti and gunavrtti. In 3.33h Ananda distinguishes 
the two types of gunavrtti as upacdra (metaphorical) and laksand (relational). 
For making this distinction Abbinava prefers the terms gauni and ldksantki 
He uses the words laksand often and gunavrtti occasionally in a very general 
sense for any sort of secondary operation. 

A By wise men, tbat is, by those who know the essence of poetry 
named "suggestion" (dhvani), which has been handed down from the 
past through a succession [of wise men], that is to say, has been made 
fully known far and wide:' this [entity], in spite of its being clearly 
apparent to the hearts of sensitive readers, some have claimed to be 

The following alternative ideas are possible for those who deny the 
existence of suggestion. 

1. [n glossing samámnáta, A takes sam to mean samyak, "fully, thor- 
oughly”; and he takes à to mean samantàt, “far and wide." One can harmo- 
nize this interpretation with the verse at the end of 3.33p A by supposing that 
it bad been long held by some critics that suggestion was the finest part of po- 
etry but that suggestion had nevertheless been aviditasattvah, "not precisely 

L (Commentary on the Kárikà:] Now the author states directly 
the nature of the subject matter [viz., suggestion, the soul of poetry], 
making it the predominant element of his sentence. Through a sub- 
ordinate element of the sentence [“to delight the hearts of sensitive 
readers") he states directly the purpose of the purpose of the book [the 

$1.11] 49 

purpose of the book is "knowledge of the nature of dhvani”; the purpose 
of attaining this purpose is "to delight the hearts of sensitive readers"]. 
By implication he shows us the purpose of the book, that purpose being 
connected [with the second purpose].' Thus he says, "[some have said 
that} the soul of poetry,” etc. 

(Commentary on the Vrtti:| In view of the proximity of the term 
“soul of poetry,” the term “wise men” in the verse must be taken in 
the sense of those who give instruction in the soul of poetry. With this 
in mind, be explains "wise men" by those who know the essence 
of poetry. Explaining the meaning of the word "soul" by the word 
"essence," he shows the prime importance of suggestion and the fact 
that what it produces lies far beyond the reach of other word-powers 
[namely, the literal and the secondary]. 

It may be objected that the word iti refers to the phonetic form of the 
word dhvani;? for the denotandum of the word cannot be the referent 
as this is still a matter of controversy and has not yet been decided. 
He explains away this difficulty by the word named. In truth the 
word dhvani bas not been used [in the Kdrikd] as a mere name; rather 
there is a thing called by the name of dhvani and it is the essence of 
all poetry, for otherwise the wise would not have taken pains to hand 
down a mere name. Thus he will go on to explain, “in spite of its being 
clearly apparent to the minds of sensitive readers.” But the following is 
a better explanation. The word iti is used out of normal order. It must 
refer to the sense of the clause as a whole [and not to the word dhvani 
alone}. So we should construe the clause as follows: “The thing, namely 
dhvani, which has been traditionally called the soul of poetry." For if 
the word iti referred only to the word dhvani, how could we reconcile 
this with the commentary's speaking of a thing named dhvani? For if 
that were the case [i.e., if iti governed the word dhvani alone], we should 
have to understand that “the word dhvani is also the soul of poetry," 
as when we say, "He says 'ox'."? That which has no reality [viz., a 
mere name] cannot be a matter of controversy. Only when a thing 
(dharmin) exists [distinct from its name], do its properties become a 
matter of controversy. But now enough elaborating on an irrelevant 
topic that will only annoy sensitive readers. "kd 

Should only one scholar have made such a statement [viz., that sug- 
gestion is the soul of poetry], it is possible that he could have been 
mistaken, but that is not likely where many have been involved. Ac- 
cordingly, [the Kariká] uses the plural: "wise men." [The commentator] 
expatiates on the matter: through a succession. That is, they have 


said this in an unbroken succession, without, however, putting it down 
in specific books. This is what he means. For many scholars could 
not teach, with great respect, something that really did not merit such 
respect. And this they did teach with respect. Therefore [the Káriká] 
says “which has been handed down from the past,” for the phrase “from 
the past” shows that one should not imagine this to be the first mention 
of dhvani. And [the commentator] explains by saying: this [entity] 
has been made fully known far and wide. 

How can one entertain the thought that an entity which one should 
strive to understand can be non-existent? What can one say? The 
sense is that those who deny its existence exhibit an extraordinary 
stupidity. Also [it is implied by the author of the Kánikà that] he 
has not actually heard the alternative ideas of those wbo deny the 
existence of suggestion; rather he will imagine such ideas and then 
refute them. Hence [his use of the perfect tense (jagaduh), which 
implies] absence of direct perception [by the subject of the verb].* 
[He could not have used the future tense, as] it is not proper to re 
fute something that is in the future, for the simple reason that it is 
not yet there. If it be argued that a hypothetical fact can be re 
futed, we say that its futurity would be abrogated by its being [al- 
ready| hypothesized. Thus, because these views are imagined to be 
in the past, because they are beyond the range of direct perception, 
and because they have not been specifically characterized as belong- 
ing to present time, the perfect tense can be used. Accordingly, the 
alternative views will first be imagined and then refuted. But even in 
imagining something, it is improper to imagine an implausible thing; 
one should imagine only a plausible thing. Otherwise there would be 
no end to the products of imagination and their refutation. And so, 
in order to give substance to these hypotheses, to be stated presently, 
he has said in advance that they are possible. Had the term "are 
hypothesized" (sambhávyante) been used, there would be tautology 
[in the use of :he optative of hypothesis in what follows]. And what 
is possible is not only hypothesized here but appears to him clearly 
as a present reality; hence his use of the present tense [in "they are 

Fearing that someone might object that it is impossible to criticize 
that which has been predicated by a hypothesis based on something 
impossible, he says, "ideas" (vikalpdh). There is no real thing to which 
these hypotheses refer; tbey are merely ideas.* Furthermore, they might 
have occurred through an ignorance of the true nature [of poetry]; and 

811L] 51 

so he uses optative forms like ácaksiran which refer to possibilities and 
here amount to a reference to past time [i.e., "they might have said”). 
The same usage can be instanced in the following verse: 

If what is within the body 
had been outside, 

people would need take sticks 
to drive off dogs and crows.* 

The meaning is that if the body had been visible in such a way, such 
[a result] would appear. The sense of past time is here implied. The 
same in the following [negative condition]: 

(But) if it were not thus, 
what would be the case?" 

The meaning is: what would be the case if there had been no hypothesis 
as above of (the body's] being [inside out]? Here we have the same 
[preterite] sense. But now enough of dilating upon an irrelevant topic. 

Here in brief are the vikalpas, the alternative ideas, that might be 
put forward against the concept of dhvani. (1) Words transmit mean- 
ings because of the conventions (samaya) which we have assigned to 
them. Accordingly, there can be no suggested sense over and above the 
denoted senses [which we have assigned]. Or, (2) granted that there 
are extra senses, such a sense will be implied by the denotative opera- 
tion and will be merely an associated sense (bhákta) since it has been 
drawn into our mind by force of the understood meaning of the word. 
Or, (3) the suggested sense is not implied in this way, but is impossi- 
ble to describe, just as the happiness of having a husband cannot be 
described to virgins who have not experienced it. These are the three 
main varieties of disagreement [with our doctrine]. But of them, the 
argument that the suggested sense does not exist may be divided into 
three sub-arguments. The first of these may be put thus. (a) As it 
is the qualities (gunas)® and figures of speech (alarikáras) that impart 
beauty to words and their senses, and as a poer consists of words and 
senses which are more beautiful than those used in conversation or in 
scientific works, there can be no source of beauty of which we have not 
already taken account [in our definitions of the qualities and the figures 
of speech]. The second is this. (b) Whatever we have not taken into 
account is not a cause of beauty. [Now the third:| (c) If suggestion is a 
source of beauty, it must fall under either the qualities or the figures of 
speech, in which case it shows no great scholarship to give it another 
name; while if it is not actually included in the qualities or figures. 

52 [511L 

still it will be only by reference to some minute differentiation that 
you give it another name, such differentiation being possible because 
of the endless varieties of loveliness in simile. The fact remains that 
suggestion is not really outside the area of the qualities and figures of 
speech. By giving it a new name what have you accomplished? It is 
always possible to imagine a new shade of beauty. For we see that the 
ancients, such as the sage Bharata, accepted only two figures of speech, 
the yamaka? for sound and the simile for meaning. What later writ- 
ers on poetics have done is only to show the direction in which these 
figures are to be multiplied [by giving independent names to the differ- 
ent forms of strikingness which they possess]. Suppose a man, familiar 
with the grammatical rule karmany on (Pan. 3.2.1) and hearing such 
examples as kumbhakára "pot-maker," were to invent such a word as 
nagarakaro "city-maker," would it not be foolish for him to feel proud 
on this account? The same principle applies to the topic under discus- 
sion. This then is the third subvariety. And so, the first view having 
three subvarieties and being joined to the other two [major views], we 
get five alternative views in all. Such is the overall meaning. 

1. Abhipava here examines 1.1 K for information on the anubandhas, 
those “pertinent points” concerning a work which commentators on Sanskrit 
philosophical texts always try to make clear, for it is by them that an intelli- 
gent mao will Cecide whether or not to study the text; see also 1.1e L, p. 69. 
The traditional anubandhas are four: abhidheya, prayojana, sambandha, and 
adhikéro. Abhinava finds the abhidheya “the subject to be treated” clearly 
expressed in the subject of the sentence, viz., “the soul of poetry called ‘sug- 
gestion'.^ The prayojana “purpose” of the work, he states, is given only 
by implication. It is of course “knowledge of the nature of suggestion,” as 
implied by the statement, in a subordinate clause, of the purpase of the 
purpose, as this is “connected with the purpose.” We take the reading of 
both Chowkhamba and Kashi editions tatsambaddham prayojanam ca. The 
Kaumudi, wishing to bring in a third anubandha, reads tatsambandham pra- 
yojanam ca and supposes that sambandha refers to the connection between 
the subject and the purpose. One may add to Abhinava's information that 
the adhikáro “the qualification required of the reader” is that he be a "sen- 
sitive reader.” 2. The difficulty which Abhinava here discusses arises from 
the placement of the word iti in the verse and from the gloss sañjñita in the 
Vrtti. The word iti functions like quotation marks in English to shift the 
denotandum from thing to word. An ox (gauh) is an animal; “ox” (gaur iti) 
is a word, beginning with 'o' in English and with 'g' in Sanskrit. This shi ing 
power (viparydsakarana) of the word iti is often noticed by the grammarians; 
cf. Nyàsa on Kas. 1.1.44 and on 1.1.66. Now if we take 1.1 K to mean that 

$11 L] $3 

wise men bave called the soul of poetry "suggestion," but some have seid 
that it does not exist, we are in danger of making the verse say that “some 
have said that the word ‘suggestion’ does not exist,” which is nonsense. So 
Abhinava first claims that dħvanir iti means more than just the word. Next 
he gives a better explanation, namely that the word itt is ou: of place. It 
really goes with kévyasydtmd, the sense being, "wise men have said, ‘dhvant 
is the soul of poetry.'” A more radical solution of the difficulty would be to 
change the wording of the verse. Mahimabhatta suggests changing the first 
line to read: kávyasyátmety amalamatibhir yo dhvanir náma gitah ( Vyakti- 
viveka, p. 397). This would be in fact a considerable improvement. 3. See 
Mahábhásya 1.1.44, Värt. 3. — 4. The perfect tense is to be used for an act 
in past time which one has not directly perceived (porokse lit, Pan. 3.2.115). 
The remarks which follow are occasioned by Abhinava's desire to reconcile 
the perfect tense (jagaduh) used by the Kdrikd with the optative (dcaksiran) 
about to be used by tbe author of the Vrtti (beginning of 1.1a A). He claims 
that the perfect is used in strict accordance with Pànini's prescription to re- 
fer to past time which has not been directly experienced and he will claim 
of the optative that it here refers to past time rather than present (“might 
have said” rather than “might say"). Such an interpretation of the optative 
is grammatically justifiable, for Kas, on Pn. 3.3.154 states that in an by- 
pothesis the optative shall take precedence over all other tenses and moods 
(sarvalakáránám apavàádoh). If the author of the Karikd actually had heard 
the criticisms of dhvani, he should have used the form agadan (imperfect) 
rather than jogaduh. Abhinava is so little interested in the historical data on 
which the Kánikà and the commentator based their criticism that be is willing 
for the sake of justifying a grammatical inflection to deny that such data ever 
existed. And yet later (1.1c L) he will speak of Manoratha, who ridiculed the 
concept of dhvani, as having been a contemporary of Ánandavardhana. [t 
seems to me that Abhinava's strict grammatical interpretation of the perfect 
tense is here quite misguided and that Karikd and commentary are both refer- 
ring to views which their author had actually read or heard expressed. — S. If 
someone says "the flowers that grow in the sky are red," an opponent cannot 
validly criticise him by saying that the flowers that grow in the sky are not red; 
but he can enticise the idea as a whole by saying that flowers do not grow in 
the sky. 6. The verse is a paraphrase of Brh. Ar. Up. 3.9.25. Sanskrit makes 
no formal distinction between past time and present in conditions expressed 
by the optative (see note 4 above). So Abhinava is grammnarically justified 
in taking the optative inflection here to refer to past time. He has thus rec- 
onciled the tense used by the Karka with the mood used by the Vrttikara. 
7. "Yadi na syát tatah kim syát" forms another quarter-verse that follows 
the half-verse just quoted. Abhinava forces the same preterite interpretation 
upon the condition. The moral of the three quarter-verses :aken together, 
according to BP, is this. If one were to imagine the body's being constructed 

54 [511L 

inside out, one would be disgusted by the thought of dogs and crows rushiog 
toward such filth. But even without such a repulsive hypothesis "the body 
remains a disgusting collection useful only for the pursuit of worthless sense 
enjoyments." 8. See below, 2.6. — 9. See below, 2.15 K, note 1. 

A Here some might contend! that poetry is nothing more than 
what is embodied in word and meaning.? The means of beautifying this 
pair that lie in sound, such as alliteration, and those that lie in meaning, 
such as simile, are well known. Also well known are (those qualities] 
such as sweetness, wbich possess certain properties of phoneme and 
arrangement. The vrttis,! which have been described by some writers 
under such names as upanágariká, and which are not different in func- 
tion from these [figures and qualities?] also have reached our ears. So 
also the styles (ritis) such as the Vaidarbhi.® What is this thing called 
dhvani that it should differ from these? 

1. Abhinava interprets as referring to past time: "some might have con- 
tended”; see 1.1 L, noted. — 2. See Bhárnaba 1.16, Dandin 1.10, Vàmana 1.1, 
and compare the opening of the Roghuvamsa: udgarthdv iva samprktau 3. [ 
follow the reading of the Kashi Sk. Ser. edition: varnasarighatanádharmnáf ca. 
Abhinava, the Kaumudi, and Krishnamoorthy's MS MB drop the word varna. 
My preference is based on the following consideration. The old view of the 
qualities mádAurya, ojas, etc., was that they were based solely on the choice of 
certain phonemes (varna) and certain degrees of compounding (sarighatand). 
Ananda will insist later in this book (2.7) that factors of meaning must also 
be taken into account. Here, however, where he is giving the obj tion of 
an old-fashioned opponent of dhvani, it is likely that he would give it in the 
old-fashioned terms. But Abhinava, noticing the discrepancy with Ananda's 
later pronouncement, would have been glad to drop the word varna. By do- 
ing so he could supply "abdárthayoh," thus making the qualities depend on 
both sound and meaning. For the technical term sarighatand (arrangement, 
texture), see 3.5 K and 3.5 A, note 1. 4. The word vrtti, literally manner 
of employment or place of employment, bears two different technical mean- 
ings in this book. the one derived from Udbhata, the other from BANS. The 
word is here used in Udbhata's sense, who applies this term (Indurája 1.7) 
to the three varieties of simple alliteration, that is, to what later writers call 

$1.1a L] 55 

vrttyanuprása. He calls the three types parusà (harsh), upanáganriká (polite), 
and grémyé (rustic or vulgar). He calls the third type also komalá (soft). For 
further details see 1.1a L and note 4. On the other hand, BANS applies the 
term urtti to four modes of gesture and speech (BANS 18.7 and 20.8f.); see 
below, 3.6g L, and note 1, as also 3.33 K and A. — S. As will be seen from L, 
the vrttis are not different from the figure alliteration, while the styles (riti) 
are not different from the qualities sweetness, etc. 6. Three styles (riti), to 
which are given the geographical names Vaidarbhi, Gaudi, and PSacall, are 
described by Dandin 1.408. Dandin assigns all the good qualities of poetry to 
Vaidarbhi, which he further characterizes as employing few compounds and 
avoiding harsh sounds. In these last two respects the Gaudi style is its op- 
posite, while the P&Rc&Il lies intermediate between the two. The differentia 
are similar to tbose used by Udbhata in distinguishing the three vrttis. But 
the evaluation was different. Udbhata regarded the intermediate vrtti as best, 
whereas Dandin chose as best the soft riti. The concept of riti was emphasized 
by Vàmana to the point of calling riti the soul of poetry (1.2.68.). On riti in 
general see P. C. Lahiri, Concepts of Rit: and Guna in Sansknit Poetics. 

L He now describes these views in sequence. Nothing more 
than what is embodied in sound and meaning: By saying “noth- 
ing more" he shows that no one will object [to this definition]. Now 
this sound and meaning themselves cannot be dhvani, for what ad- 
vantage would arise merely from giving them an extra name? But 
perhaps you suppose that dhvani refers to some special beauty of the 
sound or meaning. Very well; we may speak of two sorts of beauty: 
the beauty of a thing in itself and the beauty that arises from its ar- 
rangement (sarighataná). Of these sorts, the beauty that belongs to 
the sounds themselves derives from the figures of sound (such as allit- 
eration], while that which arises from their arrangement derives from 
the qualities (gunas) of words [when they are so juxtaposed as to pro- 
duce sweetness, harshness, etc.]. So also of the meanings: the beauty 
in themselves derives from such figures as simile, while the beauty that 
arises from their arrangement derives from the qualities of meanings. So 
there is no such thing as dhvani distinct from the figures and qualities. 

Which possesses certain properties of arrangement: that is, 
arrangement of sound or meaning. The proof by negative probans will 
run thus: Whatever is other than the figures and qualities is not a cause 
of beauty, as for example the absolute faults such as solecism and the 
relative faults such as harshness;? and dÀvani [you say] is a source of 
beauty; therefore, dhvani is not other than the figures and qualities. 

56 [81.1aL 

To this argument it might be replied: Just as the vrttis and the styles 
(ritis) are distinct from the figures and qualities and at the same time 
are causes of beauty, so it may be that dhvani is distinct and is also 
a cause of beauty. lt is with this reply in mind that he [who denies 
the existence of dhvani] continues: the vrttis also ... which are not 
different in function from these. It has not been proved that the 
urttis and styles are distinct from the figures and qualities. For it is 
merely three types of alliteration that are called by the name vrtti in 
order to group these alliterations under three broad classes to be distin- 
guished as harsh (parusd), graceful (lalitá), and intermediate, so called 
because of their utility in describing three types of subject matter, 
namely tbat which is fiery (dipta); smooth (masrna), or intermediate. 
(The literal sense of the word vrtti is that] the alliterations occur (vrt) in 
these forms. As has been said (by Udbhata, Jndurdjo 1.7, Vivrti 1.12]: 
"The wise hold that alliteration (anuprósa) is the placing of homoge- 
neous consonants, separately for each separate class, in these three ways 
(vrtti).^ “Separately for each separate class": thus, harsh alliteration 
is the ndgariké (citified) vrtti; smooth alliteration is the upanágarikà 
(polite), so called on the analogy of a sophisticated lady of the city (ná- 
garikà);? tbe intermediate variety is between gentle and barsh; such is 
the sense: it is called grámya (rustic, vulgar) on the analogy of a coun- 
try woman, who lacks sophistication and is neither gentle nor harsh by 
nature. The third of these varieties is also called komalánuprása.! Thus 
the urttis are simply classes of alliteration. We are not using the word 
vrtti here in the sense in which the Vaisesika philosophers use it, for 
[by Vaisesika rules] there could be no occurrence (vartamánatva) of a 
member in its class;? all that is meant is that tbe alliteration functions 
by means of such and such a vrtti. Just as it is said: "Kings function 
(vartante) on a plane of insight above that of common mortals." 

So the vrttis do not act as anything other than alliteration. They 
have no extra function; and as they have no separate function, they 
should not be counted separately. Thus we see that in the compound 
anatiriktavrttayah the meaning of the word vrtti is function (vydpdra). 
It is because the urttis are nothing other (than alliteration] that Bha- 
maha and others made no mention of them. Although Udbhata and 
others did mention them, they nevertheless convey no additional mean- 
ing [beyond that of alliteration] to our minds. It is with this thought 
that he says have reached our ears.® 

So also the styles (riti): One is to construe these words [with the 
preceding sentence] thus: So also the styles, which are not different in 

$11aL] 57 

function from these, have reached our ears. The word “these” (as here 
supplied”) refers to the qualities. Now in giving the appropriate vrtti 
to a passage? these gunas may be combined, through their capability 
of mixing with one another, in the form of a texture (sarighatend), just 
as the tastes of sugar, pepper, and the like may be combined in a drink. 
Such a combination is called a riti It is of three sorts, as it applies toa 
subject matter that is fiery, or delicate, or something intermediate, in 
accordance with what we see to be most eagerly sought in the country 
of Gauda, or Vidarbha, or Paücála. The class is here not different from 
its members and the whole is not different from its components. So the 
vrttis and styles are not different from the figures and qualities; and 
the negative probans [which we just gave] is valid. 

So he says: what is this thing called dhvani that it should be 
different from these? It is not a substrate of beauty, as it is neither in 
the form of sóund or meaning. Nor is it a cause of beauty, as it is other 
than a figure of speech or a quality. If we examine poetry analytically, 
despite the fact that it should be appreciated synthetically,? even so we 
find no distinct thing which cap be called dhvani. He says as much by 
using the phrase "this thing called" dhvani. 

1. Abhinava omits the term varna (phoneme) in the compound varna- 
sarighatanddharmah. 2. The faults (dosas) have been treated by almost 
every Sanskrit literary critic. For accounts in English, see Raghavan, Bhoja's 
ŚP., pp. 203-243, and Krishnamoorthy, "Doctrine of Dogas,” IHQ 20.217-232. 
The distinction of nityà doséh, usages which are always faulty, i.e., absolute 
faults, from anityá dosdh, usages which are faulty oaly in certain contexts, 
i.e., relative faults, is known to Ananda (see 2.11 below). An example of the 
latter type is harshness, which is reprehensible in a passage of love but may be 
praiseworthy in a passage of heroism or cruelty. — 3. Abhinava takes the term 
(by Saunága Várttikà 7, see Mahàbhásya 2.2.18) as ndgarikayd vidagdhayá 
upomitd: given its simile by, i.e., similar to, a nāgarikā, that is, a sophisticated 
lady. The same explanation was given by [ndur&ja on Udbhata 1.5 and is 
grammatically unexceptionable. But it leaves us wondering why the harsh 
variety of vrtti should be called nāgarikā, a term not found in Udbbata and 
which seems to originate with Abhinava. To that the Kaumudi says that the 
harsh variety is like a city lady in brilliant costume. — 4. It will be seen that 
Abhinava has changed the distribution as given by Udbhata. Udbhata gives 
three vritis: (1) harsh, which alliterates by means of 5, s, the retroflexes, and 
certain harsh conjuncts; (2) upandgariké, which alliterates by means of the 
Stops other than the retroflexes and by conjuncts of which the first member isa 
nasal; and (3) vulgar (grámyá) or soft (komalá), which alliterates by means of 
the remaining consonants, i.e., the semivowels and ^. Obviously his urttis are 

58 [$11aL 

ordered according to a decreasing degree of harshness. What is reprehensible 
in the third type is its excessive softness or liquidity. But Abhinava wants to 
harmonize the three vrttis with the three ritis given by Dagdin and Vàmana. 
Accordingly, he keeps the first urtts as harsh, makes the second vrtti soft, and 
makes the third urtti a mixture of the two. He is left with the contradiction 
that the third, mixed. urtti bears the traditional name of komald (soft). The 
commentators (BP and Kaumudi) are forced to say that the term komala 
is here ridhd, i.e., used without regard to its etymological meaning. Such 
was Abhinava's authority, however, that his scheme is followed by Mammata 
(9.80 and cf. 8.74-75) and other later authors. 5. Udbhata had said that 
alliteration occurs in three vrttis, to which Abhinava has added the statement 
that these three urttis are classes (varga). Now in the Vaisesika system a 
member cannot occur (vartate) in its class; the class occurs in, or inheres in, 
its members. So it is necessary to specify that “alliteration occurs ip a given 
vrtti" means simply that it functions in a certain way. — 6. That is, these 
terms have reached our ears but not our minds. 7. When the compound 
tadanatiriktavrttayah explicity modifies the word vrttayah, one will take tad to 
refer to the figure anuprása. When the compound is supplied as a modifier of 
ritayah, the tad wil refer to the gunas. 8. samucitavrttyarpane: Abhinava's 
conception seems to be that a combination of gunas, say of mádhurya and 
prasáda, transfers to a particular urtti, say the uponóganikà vrtti, its ability 
to express an appropriate rasa, say, srrigdra. The word yad which follows 
in the sentence modifies sarighátarüpatàágamanam. 9. The notion that the 
dsvdda of poetry is okhanda belongs really to the siddhántin, to Abhinava 
himself. The idea might have been borrowed from Bhartrhari, who says that 
we understand the sense of a phrase without analysing or dividing it, but as 
an indivisible whole. We achieve this by means of pratibhà See Vakyapadiya, 
ed. Abhyankar aad Limaye, 2.143-147 and Filliozat, p. xviii. 

A Others might say: There is no such thing as dhvani.! For a 
type of poetry that falls outside our well-known system would no longer 
be poetry. The correct definition of poetry is that which consists of 
sounds and meaning which delight the heart of a sensitive audience 
To a method which differs from the system which has been laid down 
this [definition] is inapplicable. Moreover, if you were to confer the 
title of "sensitive audience" on some few persons who belong to your 

§11bL] 59 

persuasion and on that basis assign to dhvani the title of poetry, you 
would not thereby gain the assent of the general body of educated men. 

1. Here used to mean a type of poetry; see 1.1 K. not« 1. 

L  Nowit might be granted that dhvani does not consist of sound 
or meaning; furthermore that it is not a cause of their beauty; aod 
yet it exists and for the aforesaid reason is different from the figures of 
speech and the qualities. In order to combat such a position the author 
introduces a second type of persons who deny the existence of dhvani: 

[Others might say:] It may well be [that such a thing as dhvani exists. 
distinct from the figures and qualities]. But this dhvant is not such as 
you would have it in your definition. For you were going to speak of 
something related to poetry; and this [dhvani] is no more related to 
poetry than are dancing and singing and instrumental music. Poetry 
must be spoken; therein lies its nature. One cannot claim that dancing 
and singing are spoken. 

Well-known: The well-known system is one of sound and meaning 
and of the figures and qualities. “System” means the path which people 
follow in a continuous tradition. Type of poetry: this method of 
yours is intended by you to be a type of poetry, for you have called it 
“the soul of poetry.” Why is it not poetry? Because of the definition 
“,.lof a sensitive audience." To a method: what he means is such 
a method as is used in dancing, singing, motion of the eves, etc. This: 
supply “definition” (viz., that which consists of sounds and meanings 
which delight the heart] of a sensitive audience. 

Now it might be argued that the only sensitive auditors are those 
who recognize this novel form of poetry; and as that which lies outside 
(the well-known system] is approved by them, it may well serve as a 
definition of poetry. With this in mind he says, moreover. The case 
is similar to a man's saying, "I will define a sword" and proceeding to 
do so as follows. "A sword has length and breadth, it can be worn. 
it covers the whole body, it is soft, it is woven of variegated threads, 
it can be spread out or rolled up. it cannot cut. but can itself be cut. 
This is the best kind of sword." Then when someone objects that he 
bas described a cloth, not a sword, he says. "This is what I consider to 
be a sword.” Just so is this [definition you are giving of dhvani]. What 
the author means is that what we dcfine should be what it is known to 

60 [$511bL 

be, not a figment of the imagination. So he says, the general body 
of educated men. As a few educated men might hold this strange 
opinion, he rules them out by saying "the general body." For even 
should a few hold it, what difference does it make? It merely proclaims 
their insanity. This is his meaning. 

Another commentator has taken this passage otherwise. He would 
have the denier of dhvani to argue thus. "What you recognize is a 
dhvani that is the very life of poetry. Now such a life of poetry lies 
outside our well-known system because it has not been mentioned by 
the experts on the figures of speech. Furthermore, this word dhvani is 
not an accepted term for poetry in common usage.” This interpretation 
runs counter to everything the text has said. For if dhvani as a vivifying 
principle of poetry is admitted by the denier of dhvani, the fact that it 
has not been mentioned by the ancients should rather be a reason for 
now defining it. Therefore, the meaning should be taken as we have 
taken it above. 

1. This interpretation is simpler than Abhinava's. It amounts to this. 
The frst argument of the dhvani-deniers was against dhvani as the content 
of poetry or as the cause of the beauty in poetry. This second argument 
is against dhvant as the highest type of poetry itself. Dhvani is not that, 
because the ancients made no mention of it as such and because people do 
not commonly use the word dhvani in that sense. Such.an interpretation is 
too simple for Abhinava, for it would leave the proponent of dAvani with the 
reply: “That the ancients failed to notice this type of poetry is all the more 
reason why it should aow be defined.” 

A Still others might argue for its non-existence in another way. 
Dhvani simply cannot be something entirely new because, being some- 
thing that falls within the area of beauty, it must be included in the 
means of beautifying poetry that have been mentioned [in earlier works 
on poetics]. It is trivial to single out one of these means and merely 
give it a new name. Moreover, as the possibilities of speech are lim- 
itless, there may well be some small variety that has not been dealt 

ith by the well-known compilers of definitions for poetry.! Even so, 

$11eL] 61 

we cannot see any justification here (for the proponents of dhvani] to 
close their eyes in the fond imagination that they are sensitive critics 
and to dance about chanting "dhvani, dhvani.” Others, great men too, 
have shown in the past different varieties of beauty in poetry? by the 
thousand and continue to do so.? But we do not find them acting in 
this indecorous fashion. As a matter of fact, dhvani is mere prattle. It 
is simply not possible to put forward anything as a definition of dhvani 
that can bear critical examination. [n this vein someone* has written 
a verse on the matter: 

A fool will take a poem that has no content 
to make the heart rejoice, no ornament, 

no words to show the autbor's skill, 

no striking turn of speech; 

and tell you with delight 

that this same poem is full of dhvani. 

If you who are wise should ask him, | am sure 
he could not tell you what this dhvani is. 

1. What are meant are those who have defined the figures of sound 
and meaning, the qualities and styles, e.g., Dandin, Bhamaha, and Vamana. 
2. Alorikdra here does not mean specifically a figure of speech, but rather 
beauty of poetry in general. Cf. the opening of Vàmana's KA: kávyam 
gráhyam olarikárát, and the second sutra: saundaryam olarikároh and its vrtti: 
karanovyutpottyà punar olarkárasabdo 'yam upamédigu vartate. — 3. Cf. 
Dandin 2.1: te códyópi vikalpyante kas tán kórtsnyena vaksyati — 4. Ac- 
cording to Abhinava (see below) the author of the verse was Manoratha. 

L But suppose that dhvani is a cause of beauty and that it 
can be included under the figures or qualities either of sound or sense; 
still, this entity has not been spoken of as “dhvani,” nor has anyone 
heretofore spoken of it as the life [of poetry]. In order to combat such a 
position, the author now introduces a third variety of those who deny 
the existence of dhvani: still others. 

"Beauty" (kdmaniyaka) is the passive abstract of "beautiful" (kà- 
maniya),' i.e., the source of our notion of cárutva (dearness, beauty). 
But now, as there is an endless number of charming things, [the pro- 
ponents of dhvani might argue that| they have discovered a certain 
charming thing that cannot be included under such (figures) as allit- 
eration, or under such [qualities] as sweetness, as defined by earlier 

62 [§ lle ZL 

writers. The opponent refutes this after first accepting it for the sake 
of argument: the possibilities of speech (vāc). Vdc can have any 
of three meanings: (1) that which expresses, namely a word; (2) that 
which is expressed, namely a meaning; or (3) that by which a meaning 
is expressed, namely the denotative function.” 

A small variety: for such a source of beauty will be either a quality 
or a figure of speech and so will be included under the general definition. 
As has been said: "The factors which make for beauty in poetry are 
the qualities. The figures are what add to this beauty."? Also: "An 
unusual or striking turn of word or meaning (vakrokti) is considered an 
ornament of poetic utterances."* 

Dhvani, dhvani: the repetition suggests the excitement [of the 
proponents of dhvani] and shows the awe [with which they regard their 
concept). Dance about: one may supply as subject "those who define 
dhvani, those who compose poems that use it, and those who experience 
a thrill on hearing it."* Why, he means, should there be such reverence 
for this word "dhvani^? Dhvani is mere prattle: a general view, 
summing up all [three] positions of those who deny the existence of 
dhvani. It is mere prattle (1) because if it is a cause of beauty, it is not 
different from the figures and qualities; (2) because if it is different, it 
is not a cause of beauty; and (3) because even if it is a cause of beauty, 
it is not worthy of our serious attention [as it has already been included 
under the general definition of the figures and qualities]. This is the 

Now this imagining of the position of those who deny dhvani cannot 
be charged with being [historically] baseless. Thus he says, in this 
vein someone. The reference is to a poet named Manoratba who 
lived at the same time as the author of this book.* 

[To explain Manoratha's verse:} As it has "no ornament,” therefore it 
does not “make the heart delight.” This shows that the poem lacks the 
figures of meaning [simile, etc.]. "No words to show the author's skill”: 
this refers to the figures of sound (alliteration, yamaka]. “Striking 
turn of speech”: elevated style or arrangement (sarighatand). That 
it lacks this implies that it lacks the qualities of sound and meaning. 
Some have taken this phrase [viz., "no striking turn of speech"] to 
mean that since the poem lacks this general characteristic of the figures 
of speech, it must lack every figure.’ But by this interpretation one 
could not avoid involving the author in tautology. I shall not argue 
the point further. “With delight”: he means a passion for following by 
rote the example of others. “Who are wise": for if a fool asked him, 


he could reply with such [silly gestures] as r 
rolling his eyes. 

1. By Pan. 5.1.132 the suffix vuri (vrddhi  -aka) is added to certain 
noun stems in the sense of bháva or karma. Such formations are nouns. 
Bhéva refers to a nominal abstract, the state of being beautiful; karma to 
à passive nominalization, "that which has been beautified.” In English we 
may render both concepts as “beauty.” But note that in the phrase kávyasya 
kámaniyakam (the beauty of poetry) Abhinava takes vuri specifically in the 
sense karmani the whole word kdmaniyakam denoting something that has 
been made beautiful. 2. Vàc is here etymologized as the verbal root voc 
plus the null suffix kvip with irregular urddhi by Unádisütra 225. Normally 
kvip is used actively, by Pan. 3.4.67: thus, vaktiti udk or ucyate 'nayeti udk. 
But by drawing down karmani from Pán. 3.2.1 it may be taken passively: uc- 
yata iti väk. — 3. This is from Vàmana's Kàvyólarkórasütro 3.1.1-2. Under 
guna Vàmana includes the fabdagunos and the arthagunas, viz., ojas, prasáda. 
etc.: ye khalu sabddrthayor dharmáh kévyasobhém kurvanti te gunāh Un- 
der olarikára he includes the sabdélarikdras and the arthdlarikéras: alarikárás 
ca yamakopamádayah. 4. Bh&maha 1.36. Vakrokti here has therefore a 
very general application, as it has in Dandin as well. See Dandin 2.363, 
Bhàmaba 1.30, 1.34, and especially 2.85: saisá sarvaiva vakroktir anayár- 
tho vibhávyate / yatno 'syóm kavind káryoA ko "lankóro 'nayé vinà — S. It 
is not fully clear whether "it" (the tat in tac-chravana) refers to the po- 
emg that make use of dhvani or to the word "dhvani" itself. — 6. Abhinava 
uses the term granthakrt (Kane, HSP, p. 156) as synonymous with vrttikára. 
He uses the term asmanmuülagranthakrt for the author of the Kdrikds (ibid., 
P. 188). Thus. whether or not we suppose that he is following a tradition of 
dual authorship, we must charge him with here saying that Manoratha was 
a contemporary of Ánanda. For this error see Introduction, pp. 4, 9, 26-27. 
7. This is certainly the more natural interpretation. — 8. Cf. the similar pas- 

refers to the Buddhist definition of pramána: anyathá mukAabhafigamurdha- 
kampárigulimotanddimátratattvam tat. 

[§11d A 

A “Others say that it is an associated meaning (bhákta)": others 
say that this soul of poetry which we call dhvani is [merely] secondary 
usage (gunavrtti).! And although the authors of definitions for poetry 
have not given the specific name dhvani to secondary usage nor to any 
other sort of thing, still, in showing how secondary usage is employed 
in poetry, they have at least touched on the process of dhvani even 
if they have not actually defined it. It is with this in mind that [the 
Káriká] states, "Others say that it is an associated meaning." 

1. See 1.1 K. note 2 

L The (three| alternative views of the non-existence of dhvani 
were presented successively and not without connection of thought 
That is why he used the phrase “still others” in introducing his de 
scription of the third alternative; and (with this close connection] the 
unitary conclusion [viz., “dhvani is mere prattle”| is in keeping. As the 
doctrine of the non-existence of dhvani is completely hypothetical, he 
used the perfect tense! in reporting it. On the other hand, the view 
that dhvani is no more than an associated meaning can be found in 
written texts; and so he now employs a present tense: [others] say 
that it is an associated meaning, using the present tense for the 
expression of that which takes place constantly.? 

[Abhinava now gives four different etymologies to explain the mean- 
ing of the word bhākta.) (1) Bhakti ("association") is a property that 
is associated (bhajyate), or is in company with,’ or is regarded [by 
the speaker| as commonly recognized in, a word-object (let us say, 
river-bank]—e.g., a property such as proximity to the direct object 
of the word (here, river]. Now the relational secondary sense (láksaniko 
'rthah) that derives from [i.e., that is made possible by] this property is 
called the bhdkta ("associated") meaning. (Thus, river-bank is an as- 
sociated or relational meaning of "river," as may be seen in the phrase 
“a village on the river."| As has been said, 

Loksaná (secondary usage) is held to be of five sorts, as it is based on the 
proximity of the secondary object to the direct object, on its similarity to it, 
its involvement i it, its opposition to it, or its being connected with the same 


$11d L] 

(2) Bhakti (“portion”) is a portion (bhága) of the meaning of such a 
word as is used of a group of properties, e.g., the portion of the word 
“lion” that means fierceness. Now a metaphorical meaning (gauno 
"rthah)* that arises thus is called a bAákta ("partial") meaning. Such 
a meaning may be seen in the sentence "the boy is a lion," meaning 
that the boy is fierce. (3) Bhakti ("attachment, love, affect") is the 
intense desire one may have to express such a concept as proximity [to 
a holy river] or fierceness [in a young man]. The meaning that arises 
from such eagerness is called a bhdkta (“affective”) meaning and may 
be either a metaphorical (gauna) or a relational (láksanika) meaning. 
(4) Bhakti is the breaking or blocking (bhariga, from bhariju ámardane) 
of the primary meaning. Hence a meaning that arises from blocking of 
the primary sense is a bhákta meaning. 

These etymologies will show that the presence of three factors forms 
the seed from which secondary usage (upacdra) arises. They are: the 
blocking of the primary meaning;® a cause [e.g., proximity to or sim- 
ilarity to the primary object]; and a purpose [e.g., one's eagerness to 
express forcefully tbis proximity or similarity]. 

What lies back of the apposition between soul of poetry and sec- 
ondary usage is this." Although secondary usage is [sometimes] found 
in that variety of dhvani where the literal meaning is not intended, as 
in "like a mirror blinded by breath,” nevertheless dhvani is not iden- 
tical with it, for we find dhvani without secondary usage in such a 
variety of dhvani as that where the literal meaning is subordinated to 
a second meaning And we shall show? that even where the literal 
meaning is not intended there may be secondary usage without dhvani. 
So our author will say: "This dhvani is not identical with bhakti (sec- 
ondary operation), because it differs from it in form; nor is it defined by 
that, because the definition would be both too wide and too narrow." 
(1.14 K) and “It might, however, be an adventitious mark (upalaksana) 
of a certain type of dhvani"!? (1.19 K). 

(Abhinava now takes the part of the opponent, etymologizing the 
word gunavrtti to show that just like dàvani it may denote a word, a 
meaning, or an operation or usage.]'! 

Gunas are properties such as proximity or fierceness. Gunavrtti (Lit- 
erally, “of which, or in which, there is an occurrence because of prop- 
erties”) may be either a word or a meaning (depending on whether we 
analyse the compound as a genitive or locative bahuvrihi], viz., “that 

66 [§lad L 

(word) of which there is an occurrence in (or, application to) a mean- 
ing other than the primary meaning because of these properties,” or 
“that (meaning) in which there is an occurrence of a word by these 
means." Or: gunavrtti may be the occurrence, that is, the operation, 
of the secondary power of meaning through these properties. This is 
as much as to say: in whatever sense we understand the word dhvani, 
viz., as that which suggests, or that which is suggested, or the oper- 
ation of suggesting, still it is nothing different from a word used in a 
secondary meaning, or the secondary meaning itself, or the secondary 
operation. For, as denotation is the only operation in conveying the 
primary sense, we are left with only one possibility: that dhvani is the 
secondary operation, as there is no third. 

But who ever said that dhvani was gunavrtti? With this objection 
in mind, (the denier of dhvani] says, and although, etc. Any other 
sort of thing: e.g., any sort of quality or figure of speech. In show- 
ing: he is referring to such authors as Bhattodbhata and Vamana. 
For where Bhámaha says, “Words, meters, designations (abhidhana), 
meanings," '? Bhattodbhata explains the difference between words and 
designations as follows: "Designation means the denotative function 
of words, which may be primary or secondary (gunavrtti)."? And 
Vamana has said, "Vakrokti is secondary usage (laksaná) based on 

Have at least touched on: that is, they showed the direction in 
which dhvani lies, but being men who read literally, they gave no defi- 
nition of its true nature, as they were unable to distinguish it. Indeed, 
they scorned it, merely taking up as they found them the words which 
contained it [without examining their precious inner meaning], as a 
man might take up a coconut [with no conception of the delicacy of 
its inner meat].'* Hence [the Vrtti] says that it is wich this in mind 
that the Karikà states .... If one fails to interpret the passage as we 
have done, the statement of the dhvant-opponent that [former writers] 
have "at least touched on" the process of dhvani will be contradicted. 

1. Viz., “jagaduh”; cf. 1.1. L, note 4. 2. The word dhuA "they say,” 
although it carries a perfect suffix, is regarded by Panini as a present (Pán. 
3.4.84). “The present tense for the expression of that which takes place con- 
stantly” (nityapravrttavartamáne lat) is an expression taken from Vàrttika 1 
on P&n. 3.2.123. It is given by Bhoja (SP 5, Josyer's edition, p. 164) as one 
of the six uses of the present tense. — 3. The wording sevyate padarthena 
... dharmah is awkward, as may be seen by the variant reading pràjnena for 
padàrthena noticed by the Kaumudi and accepted by Kane in his notes on 

$11e4] 67 

SD, p. 320. But the variant is surely wrong. Abhinava is forced to use sevya- 
te in glossing bhajyate because that is ita gloss in the DAP (1.1047 b^aja 
sevdyám). 4. The quotation is from a Mimámsá author Bhartrmitra whose 
works are | t. See Kane, SD, p. 320, footnote. Abhinava uses a different 
version of it in commenting on 1.18. The word sdripydt is there replaced by 
samyogát, which would exclude the metaphorical variety of secondary usage 
from loksaná. The verse is quoted in the Abhidhávrttimátrka on Kárikás 9-10 
in still a different version, but one that does include the metaphorical variety 
(abhidheyena sambandhát sddrSydt samavdyatah). The Abhidhdurttimdtrkd is 
about a century older than the Locana. — 5. For the distinction of gauna and 
làksanika see 1.1. K, note 2. — 6. By blocking is meant that the context ren- 
ders the primary meaning impossible. Boys, for example, are not really lions. 
T. Abhinava is here speaking in propria persona, not in the role of the de- 
nier of dħvani. He admits that there can be sámányádhikaronya (apposition, 
syntopicity) between dhvani and gunavrtti. This is very different from saying 
that they are identical. "An oak is a tree" is a sentence which exhibits syn- 
topicity, but a tree is not an oak. For the distinction see NVTT 1.1.4 (Kashi 
ed., p. 110, lines 168.; Calc. ed., p. 96, lines 2f.) 8. One must correct 
the punctuation of the other editions by the Kaumudi, thus: tadvyatirekenápi 
bhávát vivaksitányaparovácyaprabhedádau / avwaksita For the two vari- 
eties of dhvani here mentioned see below, 1.13m A and 2.1 [ntroduction A. 
9. Cf. 3.33). 

10. An upalaksana is a characteristic that helps define a term only 
temporarily or under certain conditions. The laksana (definition proper) 
of the washerman's house might be "the first house east of the lake" An 
upolaksana of this house might be “the house on the roof of which a crow is 
sitting.” — 11. Cunavrtti is actually used almost exclusively in the third of 
these senses. But the point of the etymologies is to show correspondence be- 
tween gunavrtti and dhvani which might be thought to support the opponent's 
view. 12. Bhámaba 1.9. 13. See J. Masson, “A Note on the Authen- 
ticity of the Bhámahavivorana Attributed to Udbhata," JIJ 13, pp. 250-254. 
14. V&mana 4.3.8. — 15. Vidyanátha speaks of dráksápáka "grape-taste," 
where enjoyment is easy and immediate, and nánikelapáka "coconut-taste," 
which is more difficult to obtain, but finally gives unsurpassed pleasure. See 
De, HSP, II, p. 242. See also Agnipurána 346.22-23. 

A Finally some, whose minds have shied away from attempting à 
definition, have declared that the true nature of dhrani lies outside the 

68 [§11eA 

realm of speech, that it can only be felt and that only by a sensitive 
reader.’ Therefore, in view of such disagreements, we shall state its 
true nature in order to delight the hearts of sensitive readers. For the 
nature of this dhvani which is the secret of all good poets' poetry, 
which despite its extraordinary beauty has not been opened to view 
by the subtle minds of the ancient makers of definitions of poetry, 
which, moreover, is clearly seen to be at work in such great poems as 
the Raémdyana and the Mahabharata, will here be revealed so that the 
bliss [which arises| in the hearts of sensitive readers on their noticing 
it in [the poems that form) the object of their attention, may take firm 
hold in their hearts. 

1. One may note that the Karikés, while they refute both the abhávaváda 
and the bhàktaváda, are silent about the andthyeyavdde This may be, as the 
Vrtti claims at the end of the chapter, because the author felt that the very 
statement of his own theory was sufficient answer. — 2. This is a difficult 
sentence to translate. Lf one breaks it up, one loses the tight connection that 
subsists between its parts. The bohuvrihi compound prasiddha-vyavahdram 
is in a double construction (kdkdksigolakanydyena). The nature of dhvani, 
which is clearly at work in all great poetry, will be revealed for the delight of 
those readers who notice that it is clearly at work in all great poetry. Their 
initial observation of dhvani has brought them bliss, but this bliss will take 
permanent hold on their hearts (or minds) only by the clear definitions of 
dhvani that the author is about to reveal. Abhinava gets carried away by the 
discovery of latent implications in the sentence. To him it suggests arguments 
against all the preceding wrong views of dhvani [n order to find an argument 
against the last of the wrong views, viz., that dhuani cannot be defined, he 
takes the word laksayatàm here to mean “who are defining" rather than “who 
are noticing.” This seems to me (D.I.) to mistake entirely the intention of the 
sentence. If sensitive readers are already defining the nature of dhvani in the 
poems of their reading. what need would there be to write the DAvanyáloka? 
Abhinava's remark on the word dnanda should be accepted. The author is 
here playing on the proper name. The effect of his book will be to give firm 
footing in the hearts of sensitive readers not only to the bliss of understanding 
dhvani but to the fame of Anandavardhana. 

L Whose minds have shied away: who were of ti id intel- 
lect. The three [sets of critics, viz., those who deny the existence of 
dhvani, those who say that it is merely an associated meaning, and 
those who say that it cannot be defined,| are such that each one later 
mentioned is of sounder judgment than the preceding. Those of the first 

$11eL] 69 

set are completely wrong; those of the second, while they recognize its 
nature, deny it [to be dhvani) because of indecision;' those of the last 
set do not deny it but know not how to define it. So what characterizes 
the three groups is in turn error, indecision, and insufficient knowledge. 

Therefore (tena): He uses the singular (tena, lit., "because of this") 
since any one of these (three] statements of divergent view might serve 
as justification for the following description [of dhvani]. 

Such disagreernents: locative of limitation; the sense is, because 
of any one type among these divergent opinions.? 

We shall state its true nature: These words imply that the sub- 
ject matter of the book is the true nature of dhvani, that the relation 
of dhvani to the book is a relation of subject matter to speech, that the 
relation of speaker to bearer is a relation of instructor to instructed, 
that the purpose of the book is (to give] a knowledge of the true na- 
ture of dhvani by refuting wrong opi ions on the subject, and that the 
relation of the book to this purpose is a relation of means to goal.’ 

Now in order to explain the portion "to delight the hearts of sensitive 
readers," a portion which sets forth the purpose, resident within the 
hearer, of the purpose [of the book], be says: for the nature of this, 
etc. The meaning of "this" is "this dhvani which has become a matter of 
controversy." The structure of the sentence is as follows. The nature of 
dhvani will be revealed in order to effect a purpose, namely, so that bliss 
(Gnanda), which is a sort of delight (nirvrti) also known as “rapture” 
(camatkára), may assume a firm stance—firm enough not to be shaken 
by other critics who suffer from error, (indecision, and ignorance|— 
within the minds of those who define (laksayatám, see 1.1e A, note 2 
and remarks of L below) (this nature of dhvani|. As one understands 
the purpose [viz., the giving of delight to sensitive readers] to be that 
which ultimately prompts [the author to furnish] the matter [viz., the 
definition of dhvani] which achieves the purpose, this explanation takes 
the words [of the last line of the Karika, viz.,] pritaye tatsvarupam 
brümaA as part of a single complex sentence.‘ 

In explaining the words “its true nature,” the Vrttikara indicates 
briefly his rebuttal of the five divergent views which he has mentioned 
above) All: By the word “all,” combined with good poets, he re- 
futes [the view that dhvani might consist] "in some small variety" [that 
bas not been dealt with previously; cf. 1.1c A]. By its extraordinary 
beauty he shows its difference from associated usage |cf. 1.1d A], for 
there is no particular beauty in such instances of associated meaning 
as "The boy is a lion," or "A village on the Ganges." By calling it the 

70 [$11eL 

secret [of all good poets' poetry], he refutes the view that it is merely a 
new name [for something already defined by earlier critics; cf 1.1c A]. 
By speaking of [dhvani as not having been opened to view] by the sub- 
tle minds [of the ancients], he shows that it cannot be included in the 
qualities or in the figures of speech (ef. 1.1 a A]. By the passage stating 
that moreover (dÀvani is clearly seen to be at work in great poems], 
he refutes the suspicion of cliquishness that [was brought against pro- 
ponents of dhvani when the objector] spoke of "some few persons of 
your persuasion" [cf. 1.1 b A]; and by mentioning the Rámáyana and 
the Mahābhārata he shows that dhvani has been revered by every one 
from tbe time of the very first poet." 

By the participle laksayatám (which in translating A we bave ren- 
dered as noticing it] he sets aside the objection that dhvani lies outside 
the realm of speech. The noun loksa means that by which something 
is recognized, that is, a definition (laksana). The denominative verb 
laksayati means to describe something by defining it. So the participle 
laksayatém means "describing it by means of definitions." 

Of sensitive readers (sardayánàám): The word sohrdaya (lit., 
"having their hearts with it") denotes persons who are capable of iden- 
tifying with the subject matter,’ as the mirror of their hearts has been 
polished? by the constant study and practice of poetry, and who re- 
spond to it sympathetically in their own hearts. As has been said, 
“The realization (bháva) of that object (e.g., vibhàva, etc.) which finds 
sympathy in the heart is the origin of rasa. The body is pervaded by 
it as dry wood by fire" (BANS 7.7).!° 

The bliss: In showing the primary object to be bliss, which is noth- 
ing more than the relishing of rasa, he shows that [of the three types of 
dhvani] it is the suggestion of rasa that is the most important and is 
the real soul of poetry. Hereby the following verse [of Bhattanáyaka!!] 
is given a mortal blow: “Supposing that one could prove dhvani to 
be a separate verbal operation. whose nature is suggestion, it would 
still form only a part of poetry, not its very self."!? For you [i.e., 
Bhattaenàyaka| have admitted that while poetry consists of the three 
parts, designation (abhidhá), aesthetic efficacy (bhdvand). and the rel- 
ishing of rasa (rasacarvaná), it is the relishing of rasa that gives it its 
life. As you yourself have said, "It is the man who relishes [what he 
reads], not he who learns it nor he who obeys it, [who is eligible] for 
[reading] poetry." '? 

If your saying that dhvani is only a part of poetry is a statement 
made with reference to vastudhvani (the suggestion of a fact) or to 

§lleL) 1 

alarikaradhvani (the suggestion of a figure of speech), you are merely 
confirming what we regard as already confirmed;!* but if it is made 
with reference to rasadhvani, your statement stands in contradiction to 
the experience proclaimed in your own admission.'* 

In this matter’® (of the primary goal's being bliss, one may make a 
distinction]: For the poet, delight is certainly his goal, but it may be 
achieved also by fame, as the verse proclaims: "for they say that fame 
has heavenly reward."!" For the auditors (or readers), it is true that 
both instruction and delight are goals, for it has been said, "The study 
of good poetry imparts skill in dharma, artha, káma, moksa, and the 
arts; it gives both fame and joy" (Bhámaha 1.2]. Nevertheless, of in- 
struction and joy, joy is the chief goal. Otherwise, what basic difference 
would there be between one means of instruction. viz.. poetry, which 
instructs after the fashion of a wife, and other means of instruction, 
such as the Vedas which instruct after the fashion of a master, or his- 
tory which instructs after the fashion of a friend? That is why bliss 
is said to be the chief goal. In comparison with ‘poetry's| instruction 
even in all four aims of human life, the bliss which it renders is a far 
more important goal. 

“Ananda” (“bliss”) is also the name of the author. Therefore [the 
concluding sentence of the Vrtti on 1.1e] also means: may the teacher 
Ánandavardhana attain by means of this book an imperishable place 
in the hearts of sympathetic readers, as [the statue of a god, properly 
consecrated, attains such] a place in a temple. As has been said: 

The authors of great works, 
even after death, 
leave with us in their poems 
a body of undiminishable beauty. 
[Bhàmaha 1.6] 

From the place he attains in their hearts, one may judge the heart of 
the author himself: the sense is that he is a prince of connoisseurs. 
The same use of “place” is found in the line of verse: "In battle the 
highest place was Arjuna's." The mention of his own name is simply 
to encourage his readers [to study the work} by arousing their respect 
and their trust, as we shall explain at the end of the book. Thus (the 
word "bliss" hints at] the primary goal of the author, of the poet, and 
of the reader. 


1. Sandeha (indecision) is that type of doubt which cannot decide which 
of two identifications is correct, e.g., “Is it a man or a tree? Is it a cloud or 
smoke?" Here the sandeha would take the form “Is this an instance of dhvani 
or of associated meaning?” 2. Abhinava does not take the locative evam- 
vidhásu vimatisu as a normal locative absolute (laksanà-saptami), as we have 
taken it. He assigns it rather to the locative of limitation (Páp. 2.3.41). So 
the literal sense by his interpretation would be: "because of this or that mis- 
understanding among (or, within the limit of| these divergent opinions, wef | 
compelled to set the matter straight.” One should place a danda after tenaiva 
hetund. 3. This sentence elicits the anubandhas essentially as Abhinava has 
already given them; see 1.1 L and note 1 thereon. 4. The first Kàriká says 
in essence: "Dhvani has been understood in various ways; therefore, we shall 
state its true nature for the joy (pritaye tatsvarüpam brümaA) of sensitive 
readers.” Other things being equal, two methods of interpreting the Karkd 
would be possible: (1) by vákyabheda, as furnishing two parallel clauses, in 
which case the sense would be: "Because dhvani has been understood in var- 
ious ways we shall state its true nature and we sball state its true nature for 
the joy of sensitive readers"; (2) by ekavdkyatd, as furnishing a single complex 
sentence, in which one clause is subordinated to another, viz., "Since dhvan: 
has been understood in various ways, we sball state its true nature for the joy 
of sensitive readers." Of these two possible interpretations, only the second 
is correct because the purpose of stating the nature of dhvani is to furnish 
delight to the readers; the need of refuting the various wrong opinions is sub- 
ordinate to this purpose. We have followed the preference of BP in taking 
vydkhyeyom as vydkhyd + iyam. It could of course be understood as a single 
word, a gerundive modifying the iti-clause pritaye tatsvarüpam brümah. One 
would then translate: "The words pritaye, etc., are to be interpreted as part of 
a complex sentence." — 5. What follows is an instance of Abbinava's passion 
for discovering hidden indications and suggestions in Ánanda's text. He not 
only finds suggestions to refute the five arguments against dhvani, he assigns 
special suggestions to no less than nine words or phrases in tbe brief passage 
comprised by 1.1e A. His enthusiasm pushes him into a false interpretation 
of the word laksayatám, as it is only by this false interpretation that he can 
find in this passage a reference to the last anti-dhvani argument, viz., that 
dAvani cannot be described in words. Cf 1.1e A, note 2. — 6. If dhvani is 
the secret of all good poets' poetry, it must be more important than any of 
the components of poetry mentioned by the older authors. We cannot call it 
merely a new name for one of those anciently defined components. 7. Vàl- 
miki was the first poet. The Ramdyana according to Indian tradition is older 
than the Mahábháruta. 8. In ABA. Vol. 2, p. 339, Abbinava defines sar- 
dayatva (literary sensitivity) as the faculty of entering into identity with the 
heart of the poet (kavihrdayatddatmydpattiyogyatd). The passage has been 
pointed out by Gnoli, p. xliv. — 9. The polished mirror is a favorite image 

§lleL) 73 

of Abhinava. Cf. IPV, beginning of Vol. 2: vitatavisadasyatmadrée svosakti- 
rasoyjvalam, and Tantrdloka, Vol. 2. p. 4, vs. 4: nirmale mukule yadvad bhdnti 
bhümijalddaya / amisrás tadvad ekasmims cinnáthe visvavrttayoh. Mirrors, 
which were of metal, were polished with ashes; cf. Vajjálagga 33. 

10. By this quotation Abhinava seems to indicate that it is only within 
the soArdaya that rosa can arise. Unfortunately the MSS of ABA. break 
off just before this passage of the BANŚ. In its place one may use the Kau- 
mudi's remarks, for its author bad probably read an undefective MS of the 
ABh. and is most likely following it in commenting on this verse. "Whatever 
thing,’” be says, "that is, whatever form of vibháva, etc., occurs in a good 
poet's description. ‘Finds sympathy in the heart,’ that is, is such as to become 
the object of the heart's sympathy. "The realization (bhóvo) of that thing,’ 
in other words, its blossoming within the frame of the spotless mirror of the 
heart: that is, the origin, or more strictly, the cause of tbe rise of, rasa. 'By 
it,’ that is, by a vibhóva, etc., of such a sort, the heart is pervaded. For 
this sudden and uniform pervasion he gives an example: 'as dry wood by 
fire.’ It is dry wood that is so pervaded, not stone or some other substance. 
Accordingly, to speak of what is exemplified, Vedic scholars (srotriyàh) and 
such like persons have no poetic sensitivity, for their hearts lack any proclivity 
(vàsaná) toward such emotions as love. By the wood's being ‘dry’ he indicates 
the purity of the heart achieved by its study of poetry, while by ‘fire’ he shows 
that the property of being a mbhdva depends on the beauty of the poetic 
qualities and figures of speech by which it is expressed." 11. Abhinava 
later refers to the author of the verse by name (1.42 L). 12. Rüpotà is here 
used, metn: causa, for svarüpatà. K glosses it by svarüpatvam, dtmatvam. 
13. To explain the verse Kau supplies the words sarva eva kduye 'dhikriya- 
te. As Kau. elsewhere quotes Bhattanáyaka independently of Abhinava, its 
author must have known the text of the Hrdayadarpana at first hand. He is 
therefore a reliable guide in such instances as tbe present. Bhattanayaka is 
here envisaging three types of reader to fit the three types of literature, viz., 
tbat which delights (poetry), that which instructs (history), and that which 
commands (the Veda). 14. In Abhinava's system vastudhvant and alarikdra- 
dhvan: are merely parts of poetry, being superior to direct designation but 
not being the real soul of poetry, which is rasadhvani. 15. We have taken 
the reading sudbhyupagama-prasiddha-samvedand, preferred by Kau., rather 
tban the reading with prosiddhi. The latter, given by both Chow. and Kashi, 
makes an awkward dvandva: "stands in contradiction to your admission, to 
what is well known, and to inner experience." Furthermore, the point is that 
Bbattanàyaka has admitted rosacarvaná to be the essential delight of poetry, 
so how can he make rasadhvani a subordinate part? — 16. Abhinava does aot 
spell out the connection of his thoughts here. It seems to be this. Ananda has 
chosen the word bliss to express the final goal of his work: it will give bliss to 
his readers. His choice of words is appropriate because it hints at the thesis 

74 [$11eL 

that the primary purpose of poetry itself is the bliss that it gives, rather than 
any instruction one may gain by it. For the relative importance of enjoyment 
and instruction in the reading of poetry, see also 3.10-14 f L and Introduction, 
p.36. 17. One may take svarga either literally, understanding that fame 
leads one to heaven, or metaphorically, understanding that fame gives to its 
possessor a delight equal to heaven. 

A At this point, although it is only dhvan: that the author has 
undertaken to define, he states the following in order to lay a ground- 

L Now the reader might ask what the train of thought can be, 
for after promising to "state the true nature of dhvani,” the next Kàrikà 
goes on to speak of the literal meaning, telling us that "there are two 
varieties of meaning, the literal and the implied." To sbow what the 
trend of thought is, (the Vrtti] furnishes an introductory remark. 

At this point: that is, the subject matter [of the book| and its goal 
being as stated.! Groundwork: anything similar to a ground or basis. 
Just as when one wants to build something new, one first prepares the 
ground, so also when one is about to describe the true nature of dhvani, 
which is none other than implied meaning, one takes as groundwork the 
literal meaning, whicb is undeniable and known to everyone, because 
the implied meaning will be more clearly noticed when placed beside 
it. Its being placed here on the same level with the literal meaning is 
in order to convey the fact that it also is undeniable. 

1, The subject matter is dhvani and the goal is to give delight. Both 
factors seem to be inconsistent with a mention of the literal meaning. 

K Meaning, which has been praised by sensitive critics and de- 
termined to be the soul of poetry, is traditionally held to have two 
varieties, the literal and the implied." 

$1.2 L] 75 

1. This is a badly constructed verse, as many Sanskrit critics have noted 
Taken literally, the relative clause is restrictive, for the anaphoric pronoun tad 
must take both subject and predicate of the relative clause into its reference 
(tasyeti tatpadenoddesyavidheyasamuditdrthavisesasyaiua tatra parámarsah). 
Accordingly, a literal translation would be: “That meaning which has been 
called the soul of poetry is held to have two varieti ." This cannot be 
what the author intended, for it flies in the face of later statements by both 
the Kdrikds and the Vrtti. The soul of poetry is limited to the implied or sug- 
gested sense alone. It is meaning in a general sense that has two varieties. The 
contradiction is pointed out by Mahimabhatta (p. 89 Benares ed.), Visvanátha 
(Book I, p. 29, just before the first half of verse 3), and others. In our transla- 
tion we have given the sense that the author seems to have intended by mark- 
ing the relative clause off with commas as if it were descriptive. Abhinava, by 
the use of considerable ingenuity, arrives at much the same conclusion. 

A Meaning, which is praised by sensitive critics as being essential 
to a poem and therefore what the soul is to a body already charming by 
the configuration of graceful and appropriate parts, has two varieties, 
the literal and the implied.’ 

1. This prose sentence suffers from the same fault as does the Kariké: 
the relative clause is properly restrictive. "Charming by the configuration of 
graceful and appropriate parts” goes with the word “poetry” as well as with 

L (Comment on the Kārikā:) Is traditionally held to have: 
this reinforces the statement [of 1.1 K] that the concept of dhuani has 
been handed down from the past. 

When it is said that "poetry is embodied in word and meaning" (see 
1.1a A], we infer from the reference to a body that poetry must also 
have something as a soul to give it life. Of the two elements [word 
and meaning], word falls wholly within the category of body, for it has 
properties sensible to everyone, just as fatness and leanness [are sensible 
in the human body). Meaning, on the other hand, is not sensible to 
everyone, for we do not call something a poem solely from its having 
meaning. Both everyday sentences and Vedic sentences have meaning 
without being poems. So he specifies: which has been praised by 

76 [51.2 L 

sensitive critics. The one general concept “meaning” is distinguished 
in the mind of discri inating critics into two branches. Now, both 
of these being equally “meanings,” why should sensitive critics praise 
just the one? There must be something special about it. This special 
something is the part of meaning that is implied and that is determined 
by discriminating critics to be the soul of poetry because it is a cause of 
the special property [of poetry]. Other persons, however, whose minds 
are confused by the close connection (of the implied) with the literal 
meaning, dispute its separate existence, just as the Carvākas dispute 
the separate existence of the soul. Accordingly, while he begins with 
the word “meaning” in the singular, he goes on to say that there are 
two varieties or sorts of this meaning, giving the reason for this by 
mentioning the distinction [enjoyed by the implied meaning] of being 
“praised by sensitive critics." He does not mean that there are two 
souls of poetry. 

(Comment on the Vrtti:] To explain the word "poetry" as used in 
the Kérika, the Vrtti says, to a poem, etc. By the word charming 
he indicates that the qualities and figures of speech impart this charm 
to it. By the word appropriate he hints at the fact that rasadhvani 
is the real life of a poem because he will show that propriety is always 
with respect to the rasa. For if the rasa is absent, with respect to what 
could one use this word "propriety" that has become so popular? 

Meaning, which is... By the pronoun "which" (yad)! he picks 
up as subject of the relative clause a fact already known. Thus he 
shows that this fact at least (viz., that artha, "meaning," is admired by 
connoisseurs) is accepted even by the opponent. By the main clause, 
viz., tasya, etc. ("of it there are two varieties," etc.), he shows that 
this acceptance is possible only if there are two varieties (of meaning]. 
[Furthermore] he hereby demonstrates that the argument that "dhvani 
is not different from the gunas and alarikdras because it is a cause of 
beauty” suffers from a falsely assigned reason, because dhvani [is not a 
cause of beauty but] is the soul itself of poetry. For we do not say that 
the soul is the cause of the body's beauty. Or, even if we grant that we 
might say so, the objector's reason becomes inconclusive when applied 
to the literal meaning, for at least that [portion of tbe literal meaning] 
which is to be ornamented cannot be itself an ornament (alarikára), nor 
can that (portion] which possesses a poetic quality be itself a quality. It 
is for this reason that the author has brought in the literal meaning.? 
And that is why he will go on to say, "the literal meaning is well 
known," etc. 

§1.3 K) 77 

1. yadānuvadan: yodd is the instrumental of the word-stem yad 2. The 
cause of the difficulty in this passage is that Abhinava is trying to justify a 
statement of Ánanda's that is really not justifiable. We have noted Ananda’s 
fault (1.2 K, note 1, and A, note 1), viz., the bringing in of the literal meaning 
to share in the designation “soul of poetry.” In the present passage, from yo 
‘rtha iti to véeydmsopaksepah, Abhinava seeks to justify what Ananda has 
done. It is necessary, Abhinava says, to bring in the literal meaning in order 
to exhibit a logical fault in an argument of the piirvapakga. The purvapaksa 
has argued that dhvani must be the same as the qualities (guna) or ornaments 
(alankára, figures of speech). The parts of the syllogism may be identified as 

Paksa: dhvanth = pratiyaméno 'rthah (the implied meaning) 
Sádhya: gundlarikdrdnatiriktah (is not different from the gunas or olarikáros) 
Hetu:  cárutvaAetutuót (because it is a cause of beauty). 

The first and most obvious fault of this syllogism is that it suffers from asid- 
dhohetutva, that is, its hetu is not true of the paksa. Dhvani is not a cause 
of beauty; it is the beauty of poetry. Now the purvapaksin might reply that 
while this is strictly true, one might, by metonymy, speak of dAvani as a cause 
of beauty in poetry. A result may be referred to as a cause; we might allow 
that the soul is a cause of the beauty that is found in the complex of body and 
soul called man. Very well, says Abhinava, in that case your syllogism suffers 
from another fault. If we substitute vdcyo ‘rthah (the literal meaning) for 
pratityamáno ‘rthah as the paksa, the Aetu will be inconclusive (anaikdnttka). 
The substitution is permissible because the vdcyo 'rthah is inextricably bound 
up with the pratiyamdno 'rthoh. But now the hetu will occur in the absence 
of the sédhya as well agi its presence. Being a cause of beauty will occur io 
that portion of the vàcyo ‘rthah which is to be ornamented (alarikárya) as well 
as in that which is an ornament. And the olarikárya cannot be an alarikára. 
Similarly that which is a gunin cannot be a guna Here is Abhinava's pièce 
de résistance. The vdcyo 'rthoh must be brought in in order to show the in- 
conclusiveness of the proving an artha to be a guna or an olorikára from the 
fact of its being a cause of beauty. 

K Of these (two varieties] the literal meaning is well known and 
has been analysed by others into many figures such assi. ile. We shall 
therefore not expatiate upon it here. 


A By others, viz., those who have made definitions of poetry.* 
[Shall not expatiate: i.e.,] we merely mention it whenever there is need. 

1. Many MSS here add the phrase bAattodbhataprabhrtibhth "such as 
Bhattodbhata.” See Kaumudi and Krishnamoorthy ad loc. 

L Of these: the sense is, "although there are two varieties." 
Well known: he means such things, well known in the world, as 
a ledy's face, a garden, moonrise. The construction is: “has been 
analysed in many ways into the figures simile, etc." The Vrtti explains 
the word "others" of the Karikd by "those who have made definitions,” 
etc. We shall therefore not expatiate: the Vrtti shows that by this 
particular negation the remainder remains unpegated; so it says “we 
merely mention it," etc.? 

1. Abhinava here lists objects that might serve as dlambana or uddipana- 
wibhávas for the production of rasa — 2. The negation is a particular one. It 
does not negate all mention of the literal meaning. 

K On the other hand, the suggested is something different, found 
in the works of great poets. It is that which appears as [something] 
separate from the well-known elements (of poetry}, just as charm in a 
woman (is something that appears different from the well-known indi- 
vidual parts of ber body]. 

A The suggested, on the other hand, is something which is found 
in the speech of great poets, different from the literal meaning. It is that 
which is well-known to sensitive readers and is separate from the known, 
ornamented, elements [of poetry], after they have been examined, being 
thus like charm in women. For just as charm is a certain something 
in women, a feast to the eyes of the discri inating, distinct from all 

$14L| 79 

the parts of the body after they have been examined, just so is this 
[suggested] meaning. 

L Something different: The word punar (“on the other 
band”) reinforces the difference [of the implied meaning] from the lit- 
eral. What he means is both “different from that” and at the same time 
"the very essence [of poetry].” The plural in great poets conveys the 
fact that this [suggested meaning] extends throughout their works. In 
fact, the title of “great poet” is used only of a poet who has the in- 
spiration needed to produce poetry which is enlivened with suggested 
meanings such as we shall explain in this work. It is because [a sug- 
gested meaning] of this sort exists that it is apprehended. For it does 
not stand to reason that something completely non-existent should be 
apprehended. Even the silver [for which we mistake the mother-of- 
pearl] is not wholly unreal.’ The apprehension of something is due to 
its actual existence; and so from apprehension we infer existence. This 
is as much as to say that what appears is such as it appears. For the 
purpose of syllogistic demonstration [one may say that] the well-known 
literal meaning is that which will be shown to have a property [i.e., is 
that which forms the paksa of the inference]. It has that property by its 
being accompanied by an implied meaning distinct [from the literal]. 
It is thus because it so appears, as do the limbs of a woman endowed 
with cbarm. 

“Well-known: The word prasiddha has the two senses "well-known 
to all” and "ornamented."? 

That which (yad tad): [By its indefiniteness|? the double pronoun 
Shows two characteristics of both the example [viz., charm in women] 
and that which the example illustrates [viz., the implied meaning}: 
namely, that neither can be precisely described—this serves to empha- 
size the aesthetic effect—and that each is readily mistaken for that with 
which it is intimately combined, (charm being confused with beauty of 
the limbs and the implied meaning being confused with the literal]. The 
Vrtti renders this by a certain something. For charm is revealed by 
the configuration of the limbs, but is a special property different from 
(that of] any particular part. Charm does not consist in the mere fault- 
lessness of the limbs or in their association with ornaments. For we find 
that discriminating critics will say of a woman, “She is not really beau- 
tiful,” even though the parts of her body on being examined are found 
to exhibit no fault, such as dullness of the eye, and even though her 

80 [81.4L 

limbs be ornamented with jewels. On the other hand, of a woman who 
is not such, they may exclaim that she is the very paragon of ambrosial 

1. Abhinava is bere attributing to Ananda the anyathákAyáti theory of 
error. In our errors we do not invent objects. The erroneously perceived 
Object is not unreal, but is merely in a place or relation other than that in 
which it is perceived. This is the standard Nyàya-Vaisesika theory of error. 
2. For prasiddha in the sense of ornamented see Kum.Sam. 5.9 and 7.16. 
on which Mallinatha quotes Amoruko$a: prasiddhau khyótabhüsitau; so also 
Ragh. 18.41. 3. The basic difference between the single pronoun yod and 
the compound yat tad is that the latter is indefinite whereas the former is 
definite. See Speyer 287c. Abhinava goes on to specify two ways in which 
the double pronoun is here indefinite. — 4. The distinction is admirably put 
by an ancient Roman: non est formosa cuius crus laudatur aut bracchium 
sed illa cuius universa facies admirationern partibus singulis abstulit (Seneca. 
Epist. 33.5): "She is not formosa [= Sanskrit lávanyavati; the word is opposed 
to bella or pulchro] whose thigh or whose arm is praised, but she whose whole 
configuration steals our admiration from the individual parts." 

A For this meaning, implied by force of the literal sense, will 
be shown to be divisible into several categories: a simple thing (vastu- 
mátra), a figure of speech, a rosa, etc.' [n each of these varieties what 
is suggested is different from the literal meaning. Thus, even the first 
variety [viz., vastudhvani] is totally different from the literal sense. For 
sometimes where the literal meaning is an injunction, the suggested 
meaning takes the form of a prohibition. 

1. Throughout the book reference will be made to “rasa, etc." (rosádi). 
The term refers to all the elements that belong to rosadhvani: not only rasa 
but bhávo, rosábhása, bhdvdbhdso, bhàvodaya, bhávasandhi, bhávasabala, and 
bhdvaprasama. For definitions see 1.4g L and for examples 2.3 L. 

L Now it may be objected that charm is widely recognized to 
be different [from beauty of the limbs], but we do not koow what a 

§1.4a L] 81 

suggested meaning is; much less is there any general recognition that 
it is different [from the literal]. So the hetu of the syllogism |viz., 
"because it so appears"| is untrue. Anticipating such an objection, 
[the Vrtti] states the nature [of this suggested meaning] in the words 
for this meaning, etc.; and in the words in each of these vari- 
eties, etc., it will establish the fact that it is generally recognized as 

Here we may begin by distinguishing two varieties of suggested mean- 
ing, a workaday variety and a variety that is found to operate only in 
poetry.' The workaday sort is that which may take the place of a literal 
form of expression;? and its types, such as injunction, prohibition, etc., 
are designated by the term vastu ("things"). This workaday dhvani 
is in turn twofold. One type, a sense which enjoyed the nature of an 
alarikàro, being in the form of a simile, etc., as it was exhibited in some 
previous [literal] sentence-meaning, is now (in the suggestive mode of 
speech] no longer an alarikára because there is no other factor to which 
it can be subordinated. But because of our recognition of it from the 
past, it is still called alarikaradhvani, much as a sramana (Buddhist 
monk) who was once a brahmin is called a brabmin sramana.* On the 
other band, what lacks this special form is called [suggestion of] a sim- 
ple thing (vastu- mátra). By the word simple (mdtra) the other form is 

On the other hand, rasa is something that one cannot dream of 
expressing by the literal sense. [t does not fall within workaday ex- 
pression. It is, rather, of a form that must be tasted by an act of 
blissful relishing on the part of a delicate mind through the stimula- 
tion (anurága) of previously deposited memory elements which are in 
keeping with the vibhávas and anubhdvas, beautiful because of their 
appeal to the heart, which are transmitted by [suggestive] words [of 
the poet|.* The suggesting of such a sense is called rasadhvani and is 
found to operate only in poetry. This, in the strict sense of the word, 
is the soul of poetry. 

When Bhattanáyaka says that dhvani could form “only a part of po- 
etry, not the very self” (cf. 1.1 e L), if by any stretch of the imagination 
this could be considered a valid reproach, it would be so only in respect 
to vastudhvani and alariküáradhvani. He himself has (in effect] admitted 
that rasadhvani is the soul of poetry by his setting the third mode 
of speech, which he identifies as the relishing of rasa, far beyond the 
mode of designation (abhidhd) and aesthetic efficacy (bhdvand). That 
vastudhvani and alarikdradhvani lead to resadhvani alone is a matter 

82 [$1.4a L 

that we shall illustrate from time to time in what follows.5 So we let 
the matter rest for now. 

Implied by force of the literal sense: this characterization holds 
for all three varieties of dhvani, for although suggestion is an operation 
of the word, the force of meaning never fails to act as an auxiliary cause; 
so we may speak of the suggestive operation as being implied by force 
of the literal sense. Even in that variety of dhvani called sabdasakti- 
mülánurananavyarigya (that form of suggestion which is similar to a 
reverberation and which is dependent on the suggestive power of words; 
for this type see 2.20-21 below), we shall show that our understanding 
of the implied meaning comes from the force of the literal sense, the 
power of words being only a subordinate auxiliary. 

Is totally different: no one will gainsay the fact that injunction 
and prohibition contradict one another. That is why he illustrates them 

1. The following remarks of Abhinava will be more easily understood 
if the reader will keep in mind the three traditional types of suggestion: 
(a) vastudAvani: suggestion of a thing. What is covered by the term “thing” 
is extremely various: a prohibition may be suggested, or an injunction, or a 
fact, or a situation. (b) alarikdradhvans; what is suggested seems to be a figure 
of speech. (c) rasadhvani: what is suggested is a rasa, bhávo. etc. Abhinava 
magnifies the value of the last type. To him it is vastly more beautiful than the 
other two and it alone forms the real soul of poetry. So he begins his analysis 
of suggested meaning by setting forth just two categories: poetic suggestion 
(= rasodhvani) and workaday suggestion (= all other forms of dhvoni). All 
workaday suggestion, he says, is really the suggestion of things (vastu), but 
one particular type of thing occasions a special designation. When the thing 
suggested seems to be a figure of speech, we call it olarikórodhvani — 2. This 
is the essential characteristic of workaday dhvan: The sense given by po 
etic dhvani (= rasadhvant) cannot be expressed by any other verbal means. 
On the other hand, prohibitions (cf. the suggested prohibition in the verse 
quoted under 1.4 b), injunctions (cf. 1.4¢), figures of speech—all these can be 
expressed by either a literal or a suggested mode of speech. 3. Alarikára- 
dhvani is discussed in Chapter Two of the present work, esp. 2.21 e, but some 
anticipation of what is there said is needed if the reader is to understand 
what Abhinava says here. Take the illustration of rūpaka-dħvanı: atrdntare 

ajrmbhato grismábhidhánah phullamallikddhavaldttahdsa mahakdlah. Here 
the litera] meaning, as demanded by the context (a description of the passage 
of time), is: "Then expanded the long season called Summer. in which there 
was a blossoming of the market stalls which were white with jasmine." But 
the power of words suggests a non-contextual meaning, viz.. "Then yawned 

$14b A] 83 

the God of Destruction, whose terrible laughter is white as jasmine.” The 
confrontation of the second meaning with the first suggests a figure of speech, 
viz. rüpaka: the long season of summer is a god of destruction. Now this 
figure of speech can perfectly well be furnished, and in another context might 
well have been furnished, by the literal sense of a sentence rather than by 
a suggestion. Note further that the suggestion is not, strictly speaking, an 
alorikára, for Ananda will define an alarikára as something subordinate. Just 
as jewelry is subordinate to the limb or body on which it is worn, so a figure 
of speech is subordinate to a sentence meaning, a rasa, bhdve, etc. But in 
the rüpoka-dhvani just instanced there is nothing to which the dhvani is 
subordinated. It is itself the sentence meaning. Accordingly, it is only by 
a fashion of speech that this can be called alarikáro-dhvani. 4. Strictly 
speaking, a sramana cannot be a brahmin, for his initiation will have forced 
him to give up all marks and distinctions of caste. — S. The long compound 
would be easier to understand if we read samudita in place of samucita, as 
does the KM edition. The more careful editions, though, give only samucita. 
6. At 1.5 L, etc. ` 

A For example: 

Go your rounds freely, gentle monk; 
the little dog is gone. 
Just today from the thickets by the Goda 
came a fearsome lion and killed him. 
[Sattasai 2.75]' 

1. The text of the verse is given with better readings in Weber's edition 
of the Sattasai: 

bhama dhammia visaddho so sunaho ajja mario tena / 
golàadaviadakudarigavásina dariasthena // 
In Sanskrit, if we disregard meter, this would be: 
bhrama dharmika visrabdhah sa sunako 'dya máritas tena / 
goddvaritatavikatakunjavdsind daryasimhena // 

The word kudariga is given by Hemacandra in the Abhidhdnacintémant (1115) 
in the sense of kurja. For daria (Sk. darya) “fearsome. causing fear,” see 

84 [$1.4b A 

gana on Pap. 5.1.2. The traditional rendering by drpte is incorrect: drpta 
would become ditta in Prakrit. As regards the meaning of the verse, one 
may correct Abhinava's comment by the remarks of Mammata p. 253 (5.139) 
and his commentators as well as those of Hemacandra (KA 1.19). A religious 
mendicant has hitherto been frightened away from a certain house by the 
family dog, but has wandered along the riverbank nearby, gathering flowers 
for pujà. Now, the young wife of the house has been accustomed to steal out 
and meet her lover in a grove by the riverbank. She fears interruption by 
the mendicant and furthermore begrudges him the loss of flowers from her 
trysting spot. So she tells him that the dog has been killed by a lion who 
lives in the thickets by the river. As Mammata puts it, the invitation to the 
mendicant, through the death of the dog, to make his rounds at the house 
suggests that he will no longer make visits to the riverbank when he hears of 
the lion. 

L  [Abhinava begins by translating the Prakrit verse into San- 
skrit. He renders the second half as godávarinadikülalatàágahanavásind 
drptasimhena: “by a proud (or fierce) lion dwelling in the thicket of 
vines by the bank of the Godavari River.” He then explains the mean- 
ing of the verse.] 

These are the words of a certain woman spoken in order to save a 
trysting place, close to her heart, from the intrusions of a mendicant 
and from his spoiling its beauty by plucking its leaves and flowers. His 
walking in that place! is a natural activity that has been inhibited by 
fear of a dog; so the injunction bere is merely the absence of preven- 
tion that arises from lifting of a ban and is not an original command; 
for the imperative inflection here conveys the sense [not of command 
but] of permission (atisarga) or of “the proper time to do something" 
(pràptakólo) (Pan. 3.3.163]. As there is contradiction between an ac- 
tivity [e.g., walking] and its absence, both cannot be directly expressed 
(by the same word] simultaneously. Nor can they be expressed by this 
word successively, for the dictum, “Abhidhd cannot express the indi- 
vidual,” etc.,? states that after the denotative function has once ceased 
to operate. it cannot operate again. 

[Objections from the Abhihitanvaya-vàda and their rebuttal.) 
Now according to the view of the abhihitánvaya-váda, the expressive 

power of the sentence (tátparya-sakti), without coming to rest [in a per- 
mission], produces a notion of prohibition as the sense of the sentence.’ 

§1.4b L] 85 

It does this by the cooperation of the speaker's intention (vivaksá) with 
the secondary assignment of reversed sense (viparita-laksand) to the 
words, brought about by a contradiction, that is, by the blocking of 
the primary sense (here in the form of “you may wander fearlessly"| 
which is not construable with the sense of the words “fierce (lion],” “pi- 
ous monk,” and “that (dog].” Thus the final meaning is entirely based 
on the [denotative] power of the words. And this is what we find in 
actual communication. People say, “This is what he said” (not “This 
is what he suggested"). So there is no other sort of meaning here than 
the literal [i.e., there is no dhvani]. 

But this is not true, for one can observe three semantic operations 
in this verse.* The operation of denotation (abhidhá) conveys senses 
that are of a general nature, for denotation is a semantic power which 
depends on convention, and convention is tied to the general; it lacks 
reference to the specific individual, for otherwise there would be no 
end [to the conventions that would have to be made for each word] 
and there would be failure of a word connected with one (individual 
to refer to other individuals of the same class]. After abhidhà, the 
power of tátparya conveys the sentence sense, in which [the general and 
unconnected] word senses are particularized and mutually connected 
according to the maxim, "The general senses [of the words| lead to 
& particularized sense, for if that were not the case, no effect could 
ensue."? Now in the example under discussion, in the second stage |i.e., 
in the moment when tátparya operates|, nothing more is understood 
than the injunction “you may go,” for this is what is furnished by the 
mere syntax of the words [i.e., by the syntax as opposed to the otber 
semantic factors]. Such is not the case in the examples “There is a 
village on the Ganges," or "The boy is a lion." For in those examples, 
the syntax [i.e., the logical connection between the sense of the different 
words], while it is on the point of taking place, is thwarted because of 
the inherent absurdity.* But in the case of the present stanza there is no 
such difficulty’ with regard to the logical connection between the senses 
of the different words: "That dog, which prevented you from going, has 
been killed by the lion and therefore, because of the absence of tbat 
which prevented your going, it is now proper for you to go." Therefore 
no blocking of the primary sense can be suspected and accordingly 
there is no occasion for laksand (secondary usage, metonymy) to give 
a reversed sense. [What we have here is rather suggestion (dhvani).| 

Or we may even admit [for the sake of argument] that there is laksa- 
nā here. Still, it cannot be said to occupy the second stage (i.e., to 

86 [$1.4bL 

operate simultaneously with the force of tátparya]. For laksaná takes 
place when there is a blocking of the primary sense. This blocking 
takes the form of an apprehension of inconsistency. In the case of the 
present stanza, there is no inconsistency in the sense of the words them- 
selves. Should you argue that they are inconsistent with each other, 
that must be understood as an inconsistency with regard to the syntax. 
Now there can be no apprehension of this inconsistency until the syn- 
tax is understood, and the understanding of the syntax does not come 
about through abhidhd, for abhidhá exhausts itself in conveying the 
[individual] word-meanings and bas no power to function further. Our 
understanding of the syntax comes about only through tátparya-sakti. 

It has been objected that if that were the case, there would be an ap- 
prehension of syntax even in the phrase, "There are a hundred elephants 
on the tip of my finger."? Well, we reply, is there not an apprehension of 
syntax in that phrase, as there is not in the phrase, “Ten pomegranates. 
six pancakes,” etc.?!? (Of course there is,] but this syntax, although it 
has been understood, is countered by other valid means of cogaition, 
such as perception, as in the cognition "There is silver" in regard to 
mother-of-pearl. Hence, the sentence which conveys such a meaning 
is not valid. In the example "The boy is a lion,” on the other hand, 
there arises a third power called /aksand, which is diferent from both 
the power of abhidhá and the power of tatparya. It arises immediately 
after the emergence of the factors repugnant to the syntactical connec- 
tion conveyed by the power of tátparya belonging to the second stage 
and it is able to neutralize those repugnant factors. 

Our opponent may object: “If this were so, such examples as ‘The boy 
is a lion’ would be poetry, because, as you will shortly say, the soul of 
poetry, which you define as suggestion, is found in such examples as well 
[as in poems which exhibit rasa].” To which we answer no; one might as 
readily say that a clay pot is alive, because, as the soul is omnipresent, 
it must be in the pot as well. Should you try to reply to this answer 
by saying that it is only when the soul is present in a body that serves 
as basis for particular [sense faculties and the like], and not when the 
soul is present in any other sort of locus, that we speak of life, very 
well, we will employ the title “poetry” only when dhvani is embodied 
in a composition containing gunas, figures of speech, propriety, and 
beautiful words and meanings. But in neither case does the soul (or 
dhvani] lose its precious nature.!! 

One cannot say that dhvani is simply bhakti (associated usage),'? 
for bhakti is the same as the operation called lakgand and it belongs 

§1.4b L] 87 

to the third stage [of verbal understanding], whereas the operation of 
suggestion belongs to the fourth stage. We may put the argument thus. 
You have agreed that laksanà comes into play on the concurrence of 
three conditions. Of these the first, which is the blocking of the primary 
sense, is based on other means of cognition, such as perception. (The 
second,] which is known as the cause, i.e., such relations (between the 
primary and secondary objects] as proximity, etc., can also be under- 
stood from other means of cognition. But the purpose [for which the 
secondary sense has been employed] are notions such as the extreme 
holiness, coolness, fitness to be visited, etc., in the case of the village, 
or the extreme courage in the case of the boy, notions which cannot be 
expressed in other words and for which there is no other valid means 
of cognition. In our cognizing of these [notions] the process cannot be 
other than verbal, [as we shall now demonstrate.| 

[The process cannot be inferential.] To infer the existence of qualities 

[like holiness in the village] from its proximity (to a holy river] would 
be to draw an inference faulty because of an ambivalent hetu." In 
the case of the boy, the fact that he is referred to by the word “lion” 
will constitute only an illusory (asiddha) hetu.'* Or, if the inference 
appears in the following form: “Wherever there is use of a word (e.g., 
"lion'] in such a way [viz., as in ‘the boy is a lion,’ where the word 
‘lion’ is not used as denotation but by laksaná], there is the existence 
of those qualities," it will be necessary to furnish another supporting 
means of cognition [such as perception] at the time of comprehending 
the vyápti.* And there is no valid supporting cognition. 
* Nor is this [knowledge of the purpose, viz., to furnish a suggestion 
of holiness, etc.) a case of remembrance, because it is not possible to 
remember something we have not experienced. And there is no rule 
of association by which we could determine what the speaker intended 
(namely, just this property and not some other]; so there would be no 
way of determining the meaning. Therefore, the operation [of under- 
standing] in these cases must be verbal. 

The verbal operation cannot be the operation of direct designation 
(abhidhà) because there is no conventional association here.'® It cannot 
be the operation of sentence meaning (tátparya), because that opera- 
tion exhausts itself in giving us our apprehension of the syntax. It 
cannot be the operation of secondary usage or metonymy (laksan4), 
because it lacks the stumbling gait (skhaladgati) that laksand assumes 
due to (the blocking of the direct object], a reason already mentioned.!" 
For if this operation too ran a halting course, it could only be because its 

88 [§1.4bL 

primary goal was blocked, which could only be because of some further 
purpose or intended goal (prayojana), so that an infinite series of in- 
tended goals would ensue. Accordingly, the name laksita-laksaná (sec- 
ondary operation arising from a secondary operation), given by a cer- 
tain author to this type of operation, is a piece of stubborn perversity.!* 
We are thus forced to admit that this is a fourth type of operation, dis- 
tinct from abhidhá, tátparyo, and laksand, one which has been described 
by such closely related terms as suggesting (dhvanana), indicating (dyo- 
tana), hinting (vyarijana), giving a notion (pratydyana), and giving to 
understand (avagama).'® As will be said: 

When a word abandons its primary operation and reveals an object by 
secondary usage, the purpose for which this is done is one to which the word 
moves without interruption (1.17 K). 

So then: the power of denotation is the power, regulated by conven- 
tion, to convey the literal sense [of the individual words]; the power 
of sentence-operation is the power to convey a sense [of the whole], a 
power which is aided by the impossibility of the literal sense without 
it? the power of secondary usage is the power to reveal a sense as 
regulated by such cooperating factors as the blocking of the primary 
sense; the suggestive power is the power to suggest, a power which has 
its origin in one's understanding of objects revealed by the first three 
powers, and which is then assisted by the imagination of the listener 
which has been prepared by these revelations. 

This suggestive power, this suggestive operation, overshadows the 
three operations which precede it and is the very soul of poetry. This is 
the author's intention; and although this power has for its object in this 
verse the purpose (for which the metonymy was used, viz., the saving of 
the trysting place and the adulterous intentions of the speaker], still, as 
these notions are introduced by the notion of a prohibition, the author 
has spoken simply of its having a prohibition for its object. 

The preceding?! has been said merely for the sake of argument. In 
truth there is no secondary usage (laksanaá) in the verse in question, 
for neither is the primary meaning entirely set aside here, nor is it 
shifted to another meaning.?? In fact there is never any operation of this 
[secondary power] in the type of suggestion which is based on the power 
of meaning." And it is obvious that from a difference in cooperating 
causes one may have a difference of power. One and the same word may 
operate, when aided by the memory of a concomitance (vyápti), as an 
inferential mark for the apprehension of the speaker's intention, and, 

$14bZL] 89 

when aided by sense perception, as that which renders the perception 
determinate.?* So much, then, is incontrovertible by those who hold to 
the abhihitánvaya-váda. 

(Objections of the Anvitabhidhana-vada and their rebuttal.] 

Now the school of anvitábhidhàna?* holds dearly to the doctrine that 
"the word's meaning is that to which the word [finally] leads," and 
would have it that the denotative operation continues longer and longer, 
like the course of an atrow.?* We ask them: if the operation continues so 
long, how can it be one, for its objects will be various? And if it is more 
than one, it stands to reason that it consists in heterogeneous elements, 
because both its objects and its cooperating causes are various.”” Fur- 
thermore, if its effects were homogeneous, it would have to pause at 
each object and then operate again. But such repeated operation of a 
word, an activity, or a cognition is ruled out by [all] metaphysicians,”* 
while if you admit that its effects are heterogeneous, why, this is our 
very position. 

But perhaps our opponent, in speaking of longer and longer opera- 
tion, means only that the meaning found in the last stage of appreben- 
sion is expressed so rapidly by the sentence (that this final, suggested 
meaning appears to be furnished by the initial semantic operation]. 
But how can this meaning possibly be understood when there is no 
convention connecting it [with word or sentence|? Our opponent may 
apswer that conventions subsist between the causal factors (namely, the 
individual words and their meanings] and are therefore unnecessary be- 
tween the result and its meaning. Now, look at the skill of this Vedic 
scholar! Here he is saying that the later understanding of the individ- 
ual words—for according to his theory it does come later—becomes a 
cause of the meaning which occupies the final stage, a meaning which 
[according to his theory of semantics] enters the apprehension first. 
Why, this Mimàmsaka might claim to be the descendant of his own 

Our opponent might claim that such understanding [viz., of the final 
meaning] occurs only to one who has previously been initiated into the 
conventions [of the initial, denotative meanings]; and because the mat- 
ter stands thus, the [initial] meanings do act as a cause. But by recourse 
to this argument he would not be saying anything of use.?? Further- 
more, in our opponent's theory there is no previous understanding of 
the individual word-meanings, for they are invariably used in sentences 

90 [§1.4b L 

[which are understood as wholes before the meaning of their compo- 
nents can be inferred]. If he says that such understanding does in- 
deed come about by insertion and removal (dvdpodvdpdbhydm),° this 
is tantamount to saying that the convention applies to individual words 
[which are general) and that the understanding of the specific [sentence 
meaning] comes later. 

Now [the anvitabhidhdnavadin] may say, “The final sentence meaning 
occurs to us immediately; there is no way around it." This is a fact 
that we too are not unwilling to accept. Our author will go on to say: 

Just so does the suggested sense flash forth in an instant in the minds of 
the intelligent auditors who are averse to the literal sense and in quest of the 
teal meaning (1.12 K). 

But this is because the auditor has considered the subject so often that 
the succession, which must be hypothesized,”! is not felt, because there 
is no overt manifestation of succession among notions that belong to 
the same category, just as we are unaware of succession in our memory 
of concomitance and verbal convention.?? A relation of cause and effect 
[between the initial meaning and the final meaning] must be accepted 
if we are to keep the secondary sense, of either metaphorical (gauna) or 
relational (làksanika) type,” distinct from the literal sense, or [if our 
opponent is] to avoid impugning the doctrine that "of the six exeget- 
ical criteria—direct statement (sruti), implication (liriga), etc.—each 
that follows in the list is weaker than those which precede" (Mimámsà 
S. 3.3.14), for this can only be justified by the causal efficacy (of dif- 
ferent sorts of meaning]. And if you accept a variety of causal efficacy, 
what point is there in your ill will toward us? 

[Remarks on the Sphotaváda.] 

Those too who claim that both sentence and sentence meaning are an 
indivisible entity called the sphota, when they descend into the world of 
communication,” follow our system in all respects. Above that world, 
of course, everything is brohma, which is identical with God Supreme: 
a point of view not unknown to our author, who also wrote a work 
called Tattvaloka.?* So now enough. 

(Bhattandyaka’s interpretation refuted.] 

Bhattanáyaka has said: "In this verse our understanding of the pro- 
hibition is brought about by the entrance of bhaydnaka-rasa (the flavor 

§1.4b L] 91 

of the timorous) through the use of the words ‘fierce lion,’ etc., with 
the use of the word ‘pious,’ for there would be no understanding of 
the prohibition in any other way if we lacked an understanding of the 
two characters [portrayed] here as fierce and as ti id respectively. So 
it is not simply the suggestive force of the situation tbat causes this 

To this we reply [as follows]. Who ever said that without an under- 
standing of the particular speaker and the particular person addressed, 
and without the operation of suggestion that belongs to the words, 
there could be an understanding of the prohibition? We have said 
that it is essential to suggestion that it be helped out by the imagina- 
tion of the hearer. And we do not [even] rule out the entrance of the 
bhaydnaka-rasa, for we admit that it may arise from a simple [emotion 
(bhàva) of] fear. And this rasa may enter the hearer?” if the rasa is 
manifested [in the verse]. But the rasa must be suggested. Its being 
directly denoted is not admitted even by Bhattanayaka; so it must be 
suggested. Furthermore, this rasa does not necessarily enter the hearer, 
for the sensitive reader is not necessarily similar to the timid monk. Or, 
if Bhattanáyaka supposes a special nature of the reader [viz., that he 
must be aesthetically sensitive to fear] to be a cooperating cause [in 
producing rasa], why should he be so opposed to an operation of sug- 
gestion enlivened by the imagination of both speaker and hearer? What 
is more, by trying to deny vostudhvani in the verse, he has made out a 
case for rasadhvani What a powerful critic of dhvani he turns out to 
bė! As has been said, "Even the anger of a god is like a gift." 

If he should claim that all that has been shown [by this example] is the 
supremacy of rasadhvani, who would deny it? But then, he might say, 
it was not right to adduce this verse as an example of mere vastudhvani 
To this we reply that as this example is of poetry, let it exemplify two 
types of dhvani; what harm is there??? But if he insists on the mixture 
with rasa, know that a mixture with the bhayánakarasa does not sit well 
in the mirror of a connoisseur's heart. Rather (the connoisseur will feel 
that] in this verse there is the erotic rosa, which arises in the manner 
we have described,’ from a mixture of vibhávas and anubhávas: we 
have the trysting place serving as a vibhdva (stimulating determinant) 
of the [basic emotion which is] the desire for intercourse; and we have 
such anubhávas (symptoms) as a specific tone of voice appropriate [to 
the mention of a trysting place]. 

It is because rasa is unworldly and cannot be understood straight off 
that the author has begun with this example of vastudhvani, intending 

92 [514b L 

thereby to exhibit [an instance of literal and suggested meanings that 
are] indisputably distinct, [viz.,] injunction and prohibition. 

As for him who set himself up as an explainer of dhvani and said 
that it was nothing but the power of sentence meaning (tátparya-$akti) 
or inference of intention (vivaksó-sücakatva),' he does not appeal to 
us. As they say, "Each to his own taste."*! As we shall deal with this 
later in the book as occasion arises, let the matter rest for the present. 

[Glossing the words of the verse.) 

Go your rounds: you are permitted; it is time for you to wander. 
Pious monk: it is appropriate for you to wander about gathering 
flowers for pujé. Freely: because of the removal of the cause of your 
hesitation. The little dog: viz., he who caused your slender little 
body to tremble with fear. Just today: the sense is that you have 
had a stroke of luck. Killed: so he will not appear again. Lion: that 
lion whom you heard about from hearsay and who lives in the thickets 
along the Godavari. For, to protect (her trysting place] she had already 
seen to it that be was told (of such a rumor].‘? But now, because of 
his fierceness, he has emerged from the thickets. So the monk's going 
anywhere near the bank of the Godavari is out of the question; how 
much more his entering the thickets. 

1. The word tatra ("in that place") is misleading. If the dog had fright- 
ened the mendicant from the trysting place, there would be no reason for 
the woman to invent a lion by which to terrify him further. We prefer the 
interpretation given above (1.4b A, note 1). 2. The quotation is said to be 
from Mandana Misra (so Jbalkikar on KP, p. 44). The full line is: visesyam 
nábhidhà gacchet ksinasaktir visesane, "as the power of direct designation is 
exhausted in [denoting] the classifying character, it cannot operate on the 
classified individual." Mandana followed the Mimárnsà theory that words de- 
note class characters or universals. When we say gàm ānaya, "bring ... cow" 
(Sanskrit lacks the definite article), the word gam refers directly to gotvam 
(bovinity). The sense of a particular cow, or the cow, characterized by bovin- 
ity, is given only by the sentence meaning, which depends on context. The 
verse here quoted is Mandana's refutation of an opponent who argues that the 
word gdm might denote both the universal and the particular. [t cannot do 
so, he says, because the abhidhd (power of direct designation) in a word dies 
after it has once operated. Abhinava in the present passage uses Mandana's 
dictum to show that we cannot let bhama ("go" or "walk") in the exem- 
plar verse designate directly two different things (“go your round" and "do 
not go into the thickets"). It can mean directly only one of these; the other 

$14bL] 93 

meaning must be furnished by a different semantic power, viz., suggestion. 
3. The abhihiténvaya doctrine, held by the Bhátta- MImàmsakas, or followers 
of Kumárila, holds that the final sentence meaning (/átparya) is furnished 
by the syntax (anvaya) of the directly expressed (abhihita) meanings of the 
individual words. The meanings of the individual words are universals; the 
tátporya is specific. The doctrine is directly opposed to the anvitábhidhána 
view of Prabhákara, which argues that there is no need for two semantic pow- 
ers here. All meanings, according to Prabhákara, are specific, the signification 
(abhidhána) of words being understood only with reference to the specific acts 
and situations in which they are involved (anvita). — 4. The three operations 
that Abhipava here points to are abhidhd, tátparya, and vyafijana (dhvanana). 
But he wil) go on to allow, for tbe sake of argument, the possible presence of 
another operation, laksand — 5. If you tell a boy “gdm dnaya,” he cannot 
direct his action to the class character bovinity. He can only act with regard 
to a particular cow. 6. Accordingly, we are forced to understand these 
sentences by the semantic power of loksaná (secondary usage or metonymy) 
They mean, respectively, “There is a village on the bank of the Ganges,” 
and "The boy is brave as a lion.” Note that the Sanskrit sentence garigdyém 
ghosah, unlike the English translation, “a village on the Ganges." is literally 
impossible. The locative case does not have as wide a span of meaning as 
the English preposition “on,” which may mean “by the side of” as well as 
physically “on top of." The Sanskrit phrase means Literally ^a village situ- 
ated in the Ganges,” so if we take the phrase literally, we will suppose that 
the inhabitants are drowning. 7. The words na hi (Kashi ed. p. 57, line 2; 
Vidyabbavana ed. p. 55, line 2) construe with kdcit ksatih (Kashi 57.5; Vidy. 
55.5). 8. If one admits that latgand is at work in the verse *bhama dham- 
mia," the nature of the verse's suggestion (dhvoni) will differ from what it was 
taken to be when the operation of latsand was denied. Without laksand the 
suggestion in the verse is a vastudhvani that takes the form of a prohibition, 
viz., “you must not wander into the thickets by the river.” With laksand, the 
prohibition is furnished by laksand and the suggestion becomes a rasadhvanı, 
viz., a suggestion of the love between the speaker of the verse and the man 
she hopes to meet at the tryst. 9. A standard example of absurdity; see 
Jacob's Handfull of Marims III. p. 4. It can become reasonable under cer- 
tain circumstances; see 3.331 L, note 9. But the point here is merely that an 
absurd sentence has syntax and is meaningful. 

10. Literally, “Why should there be no apprehension of syntax here, as 
in ‘ten pomegranates,’ etc.?" BP says that Abbinava here furnishes an ex- 
ample by giving the opposite (vaidharmyena drstantam dha). "Arigulyagre," 
etc., does have syntax; “dasa dddimans,” etc., does not. The latter quota- 
tion is from Mahábhásya on 1.1.3, Várt. 2 (repeated on 1.245). It became 
a standard example of word groups that are meaningless because of lack of 
Syntax. The full quotation is: dasa dádimáni gad apiipdh kundam ajájinam 

94 [$14bL 

palalapindah adharorukam etat kumáryah spháyokrtasya pità pratisinah, “ten 
pomegranates six pancakes basin goatskin sesamum seed-cake petticoat this 
of a girl of sword-maker's-son the father curdled." The individual words have 
meaning and one can even make sentence meanings out of some of the compo- 
nent parts by supplying the verb to be (e.g., “there are ten pomegranates”), 
but there is still no syntax of the whole. The whole is not absurd but mean- 
ingless. 11. The fact that suggestion is found in many utterances—it is 
found in connection with all tropes and metonymies unless they have become 
frozen—does not cheapen it. It remains the central and most essentia] element 
in poetry. lf one seeks a more precise nomenclature, one may call the unpo- 
etic uses of suggestion vastudhvani, saving the term rasadhvani for the type of 
suggestion that is poetically effective. 12. Abhinava here gives his clearest 
proof of the difference between dhvani and laksand; the matter is not treated 
by Ananda untii later on (1.14) and his distinction is not so clear. 13. The 
inference "That village possesses holiness because it is close to the Ganges" 
suffers from ambivalence because the Aetu "proximity to the Ganges" occurs 
in mpaksa objects (e.g., unholy objects such as filth and dead bodies) as well 
as sapaksa objects (holy objects such as temples or the village in question). 
For the fault of anaikántikotva in inference, see N.S. 1.2.5 aud the comunen- 
taries thereon. 14. An illusory probans (asiddha-hetuA) is one which does 
not really occur in the minor term (poksa). For example, if we argue Arado 
vahnimán dhümát, ‘the lake contains fire because it has smoke," the probans, 
smoke, is asiddha, for there really is no smoke on the lake. Now there may be 
mist or fog on the lake, but a probans in the specific form (rüpa) of smoke is 
absent. Hence the probans in such an example may be more precisely termed 
svarüpásiddha: "not found in that specific form in the paksa." To come to the 
case at hand: the Mimamsaka wishes to argue that we arrive at the notion of 
the boy's extraordinary courage by an inference: the boy has extraordinary 
courage because he is simhasabdavdcya, "denoted in a primary sense by the 
word ‘lion’.” But that is just the point. The boy is not so denoted; he is 
denoted, that is, spoken of in a primary sense, by the word "boy" So the 
Mimamsaka must substitute a different inference. — 15. An inference must 
always be backed up by perception and memory. If someone says, "The boy is 
courageous (parákramaván ménavekah),” we can infer courage to be a prop- 
erty of the boy, because we have perceived courage in a number of persons who 
were directly denoted to us as parükramaván and we have a memory of those 
perceptions. But in the case of a metaphor no such background exists. We 
will have observed "lion" used metaphorically of persons who are cruel, royal, 
proud, or courageous. Without any rule of association (niyama) we cannot 
determine by inference what the speaker's intended meaning (vivaksita) may 
be. We can arrive at that knowledge not by inference, aor by the previously 
mentioned powers of the word and sentence. but only by a separate power. sug- 
gestion, as Abhinava proceeds to demonstrate. — 16. The samaya is between 

$14b L] 95 

the word garigd and a river, not between garigá and purity. etc. 17. The 
operation of abhidhá that runs from the word garigá to a river runs a direct 
course. The operation of laksand starts out from garigá for a river but then 
shifts course to a river bank or some such nearby object. Accordingly, it runs 
an interrupted course. The course that runs from garigd to the suggested 
meaning of purity again runs a direct course, being in this respect like the 
denotative operation and unlike the operation of metonymy. 18. Abhinava 
means that the term has been used only in order to avoid using the correct 
term "suggestion." The term laksitaloksand misrepresents the basic nature of 
the operation, which is not a loksand at all because it lacks skhaladgatitva. 
19. The substance and often the very words of the foregoing paragraph are 
repeated in Mammata's Sabdavyápáravicàra, pages 5-6. 

20. The words tadanyathánupapattyá have given the commentators trou- 
ble. The Kaumudi's explanation, which takes tad to refer to abhidhdsakti, 
seems the simplest. The power to convey a sentence meaning, a meaning 
which is of a specific situation that exists in the external world. is helped out 
by the fact that the individual word-meanings cannot be found in the world 
without it. In gém dnaya. "bring the cow," the vdcydrtha, bovinity, cannot 
be found except as characterizing an individual cow such as we find in the 
sentence meaning. Hence we are forced to go on from the litera! sense of the 
individual words to the specific sense of the sentence. 21. "The preced- 
ing” (etad) refers to everything that has been said in the English translation 
from “Or, we may even admit,” page 85, up to the present point; in the San- 
skrit, from bhavatu vdsau (Kashi ed. 57.7 to 63.2; Vidy&bhavana ed. 55.7 to 
61.3). 22. The reference is to the two types of suggestion which are based 
on igksand: atyantatiraskrtavácya and arthántarasankramitavácya. They are 
described and identified later on (2.1a-c). — 23. Having given up the "adruis- 
sion for the sake of argument," Abhinava comes to what he believes the nature 
of the suggestion truly to be in the verse in question. It belongs to the second 
great class of suggestion. called vivaksitányaparavácya (2.1 Introduction and 
2.2 A. note 1). Within this class it belongs to the type samlaksyakramauyarigya 
(2.2 K) and within that type to the sub-group arthasaktimüia (2.20 K). In 
Other words. he has now given up the interpretation by which the object 
suggested was taken to be the saving of the trysting place and the adulterous 
intentions of the speaker. The object is now taken to be the prohibition of the 
monk from wandering into the thicket. This is vivaksitányaparavácya because 
the literal meaning (a permi ion) is subordinated to something else which is 
primarily intended. It is samlaksyakramavyarigya because we are aware of an 
interval between our understanding of the literal meaning and our understand- 
ing of the suggestion, as is not the case in rasadhvani. It is arthasaktimula 
because the suggestion is based on the force of the situation rather than on 
that of an ambiguous word. 24. A difficulty has arisen. which Abhinava 
Seeks to solve by an analogy. The difficulty is this. How is it that the very 

96 [§14b L 

same words, "go your rounds freely, pious monk,” may be said according to 
one theory to have the power and the operation of laksaná and according to 
another theory may be said to have the power and the operation of sugges- 
tion? This is possible, says Abhinava, by a change of cooperating causes. BP 
explains. [f we suppose a blocking of the primary meaning to occur because 
that meaning is incompatible with "fearful lion," etc., a power of laksaná will 
arise in the words “go your rounds"; and the laksita meaning will be “do not 
go your rounds.” On the other hand, if we are not aware—as we ought not 
to be, according to Abhinava—of this blockage and if, instead, we are aware 
of some special characteristic of the speaker (an adulteress) or of what she 
seeks to convey, a power of suggestion will arise in the words. An analogy is 
furnished by the word "Devadatta," let us say, in the sentence "This is Deva- 
datta." The word may operate in the realm of inference, if we are seeking to 
infer that the speaker has an intention to convey certain information. It may 
operate within the realm of perception if we are seeking to form a determinate 
perception of the indeterminate thatness in front of our eyes. The powers that 
arise in words depend on the causes that cooperate with words in giving us 
our cognitions. 25. That is, the followers of Prabhákara; see note 3 above. 
26. The Prabhakara doctrine of word-meaning is brought up again at 3.33d L 
(the long operation). The Kaumudi here gives the following explanation of 
the simile of the arrow. Just as a swift-handed bowman might shoot an ar- 
row that would pierce his enerny's armor, then take the man's life and &nally 
enter the earth, just so a single denotative operation may run on to the final 
stage of our comprehension, leaping through the intermediate stages. As so 
much of what Abhinava says in this section, this too has been taken over by 
Mammata (KP 5, Jhal. ed. p. 225, and Anand. ed. with Govinda and Nagoji, 
p.213) 27. Its objects: the literal meaning, the secondary meaning, the 
suggested meaning. Its cooperating causes: the convention, the blocking of 
the literal meaning, the special properties of speaker or context. — 28. Cf. 
Sabara 1.1.25: padáni hi svam artham abhidhóya nivrttavyāpāräni and cf. 
note 2 above. But the doctrine extends farther than to words. A given action 
carries only one result. If we do one good deed, we reap the benefit of that 
good deed only once; we do not continue to enjoy the benefit time after time 
We make a valid cognition only once; it is valid only for the time at which 
we make it. We may perceive smoke on the mountain and infer that there 
is fire there now. We may not, after perceiving smoke today, infer tomorrow 
that there is fire there. 29. BP: "Because the person who has learned the 
conventions would understand the meaning that is conventionally associated 
with the denotative meaning (i.e., the meaning of the first stage). How would 
he come to understand the meaning of the final stage. for which there is no 

30. Insertion and removal (dudpa-ududpa) is the method, according to the 
anvitàbhidhána-váda and other schools, by which a child learns the meaning 


of words; see KP ed. Jhalkikar, p. 221; ed. Anand. with Govinda and Nagoji, 
p. 210. A child hears an older man say, "Devadatta, bring the cow," and ob- 
Serves a younger man go and bring a cow. Later the child hears such sentences 
as "Caitra, bring the cow," “Devadatta, bring the horse." By the removal of 
words from, and the insertion of words in, the various slots of a sentence, the 
child gains a knowledge of the meaning of the individual words. 31. Because 
cause must precede result. 32. “The same category": verbal, inferential, 
etc. We jump from the notion of Ganges to purity and holiness without aware- 
ness of the succession of our ideas, just as on seeing smoke we almost instantly 
conceive of fire, without being conscious of the concomitance, “where there is 
smoke there is fire”; or just as, on hearing the word “cow,” we understand the 
object cow without consciously remembering the convention, "the sound 'c-o- 
w' sball represent the class notion underlying an object with horns, boofs, tail, 
etc.” 33. Cf. 1.1 K, note 2. 34. Abhinava is not fair here to the sphota- 
vdde. The grammarians conceived of sphota in the world of ordinary commu- 
nication (vyavahdra, avidyd) as well as in the rarified metaphysical world of 
param brahma. For sphota, see John Brough, “Theories of General Linguistics 
in the Sanskrit Grammarians,” Transactions of the Philological Society, 1951, 
and "Some Indian Theories of Meaning,” ibid., 1953; also K. A. Subrabma- 
nia lyer, "The Doctrine of Spbota," Journal of the Ganganatha Jha Research 
Inst, Vol. $, Pt. 2. 35. Presumably, this lost work dealt with metaphysics. 
The only other reference to it, so far as we know, is again by Abhinava; see 
45 Landnote9. 36. See Corrections of the Kashi Text. Here, as elsewhere, 
Bhattanàyaka's effort is to deny the need of positing dhvam in order to explain 
the verse. He supposes that a feeling of rasa is brought about by the verse's 
bhávokatva, its possession of bhávaná or aesthetic efficacy. See Introduction, 
pp. 35-36 and 2.4 L. [t here consists of the poet's baving so arranged the 
words as to impress on us the terror of the pious monk. Once we relish aes- 
thetically the monk's emotion, we shall understand ipso facto that the words 
of the verse amount to a prohibition. The term artha-sdmarthya ("suggestive 
force of the situation") is approximately equal to vastudhvani — 37. It isa 
cause of some confusion in this passage that the same word, pratipattr, is used 
for the hearer in the verse, i.e., the pious monk, and the hearer of the verse, 
i.e., the reader or connoisseur. In the phrase vaktrprotipattrvisesávagama the 
former must be meant. In pratipattrpratibhd and in pratipattus ca rasdvesah 
the latter is meant. 38. The point is this. In a work of philosophy it 
would be considered a fault to give an example tbat illustrates two principles 
at once when you are concerned oniy with one. But in poetry to do so is 
inevitable, since so many verses contain more than one excellence. So Kau- 
mudi, p. 129: bohuvisayatvád ekasydpi kávyasya niyatagocaratudt. 39. By 
“the method we have mentioned” is meant the method of suggestion fol- 
lowing upon the literal sense. The thirty-nine syllable compound beginning 
sambhoga- is curiously compressed. Abhinava means that the Srrigdrarasa 

98 [814b L 

arises from a combination of anubhávas appropriate to a vibháva of the stháyi- 

40. None of the commentators has identified the person to whom Abhi- 
nava is referring in this passage. Tátparyasakti for dhvani would be a likely 
substitution for a Mimamsaka. The Kaumudi finds the substitution of vivaksá- 
sticakatva to be characteristic of a Buddhist. 41. Ragh. 6.30. If this was not 
a proverb before Kálidása's time, it has become one since. 42. Abbinava's 
incorrect interpretation of the verse here leads him to a farfetched hypothesis. 

A Sometimes when the literal meaning is a prohibition, this (sug- 
gested meaning] takes the form of an injunction (or invitation), as in: 

Mother-in-law sleeps here, I there: 
look. traveler, while it is light. 
For at night when you cannot see 
you must not fall into my bed.* 

1. The verse is a variant of Sattasai 7.67, which has been imitated by 
the Sanskrit verse SRK 812. As in the case of the verse in 1.4b above, one 
may analyse in either of two ways. If one finds no laksaná in the verse, the 
suggestion will be simply an invitation to the traveler to come to the woman’s 
bed. Presumably this was Ananda’s understanding. If one takes the invitation 
to be conveyed by laksand, the suggestion will be of the woman's love of the 
traveler. Vi$vanàtha gives this the title rasdbhàsa rather than rosa because 
the underlying love is adulterous; see SD, p. 26, prose following the 9th verse 
quoted after 1.2. 

L [After translating the Prakrit stanza into Sanskrit, Abhinava 
continues:] In the Prakrit, maha is an irregular form used in many 
senses, Here it has the sense of the genitive plural (“our”), not the 
genitive singular.! Had she referred specifically to herself, she would 
have aroused suspicion? and so have been unable [later] to receive him 

[The situation is this] The sprout of love has suddenly arisen in 
a traveler as he looks at a young woman whose husband is away from 

§ldeL) 99 

home. By means of this prohibition she gives him permission. So, what 
we have here is an injunction that consists in the absence of prohibition. 
It is not a command, setting someone to do that which he has not set 
about, for such would be insulting to her opinion of her own charms. In 
keeping herewith is her hint in the word rátryandha ("blind at night") 
that he will be out of his senses with the desire that will come over 
him at that opportune time. As an action and its absence are self- 
contradictory, it is clear that the suggested sense is here different from 
the denoted sense. 

Bhattanáyaka has said: "In this verse too, as in the preceding verse, 
the meaning is furnished verbally,” by the woman's conveying her state 
of desire by the use of the word 'I' (in ‘I sleep there’) accompanied 
by particular gestures.” We reply that the word "I" does not directly 
denote this sense [of sexual eagerness]; while if, in conjunction with a 
tremor of the voice, it ray hint at this sense, that may count as a help 
to the theory of dhvyani. not a hindrance. 

From the word "mother-in-law" it follows that he must make love 
quietly so that it may not be known. And in speaking of “this miserable 
day" (divasaka)* she suggests, "I know that your heart is being shot 
in pieces by volleys of Love's arrows and that I should take heed of 
you, but what can [ do? The contemptible daylight is still with us." 
It is [called] contemptible because it is unsuitable for love. In Prakrit 
the distinction of masculine and neuter does not hold.* "Nor do I fail 
entirely to take heed of you, as I remain right here. So look at me. I am 
not leaving you. We can get through the day with the solace of looking 
at each other's face.” Such is the meaning. And there is a suggestion 
that “you should not join me in bed, being blinded [by passion], the 
very minute it grows dark, but should be very secret and wait until you 
have discovered that sleep has overtaken this thorn in my flesh called 
a mother-in-law.” 

1. Maha, or, according to the Kaumudi reading, maham, is irregular and 
is used for v ious cases of the singular first person pronoun: accusative and 
genitive, the latter of which may also substitute for the dative. But it seems 
never to be used for the plural. The reason that prompts Abhinava to this 
interpretation is his overrefinement of the woman's character. 2. Abhinava 
supposes that the words are spoken in the presence of the mother-in-law. 
3. Ananda has quoted the verse as an example of vastudhvani, a sugges- 
tion which arises artha-sámarthyát ("from the capability of the situation"). 
Bhattanàyaka is saying that the suggestion here does not arise from the sit- 
uation; it arises from a skilful use of the word “aham.” 4. Abhinava is 

100 [514cL 

interpreting the word diasaam of the Prakrit as though it formed an elliptical 
sentence, standing for divasako 'yam, "This is miserable daytime." He takes 
the suffix to be the -ka of contempt given in Pan. 5.3.74. The interpretation is 
wrong on both counts and forces him to find a reason for the supposed neuter 
gender of the word. It is not neuter, of course, but accusative masculine: 
"during the daytime." — 5. This is to explain how diasaam has been used in 
the neuter (see preceding note). Actually, the grammarians permit divasa to 
be used as a neuter even in Sanskrit (Cana on Pàn. 2.4.31 and AK, 
but we do not remember ever having seen it so used. 

A Sometimes the literal meaning is in the form of an injunc- 
tion,while the suggested meaning takes a form that is neither [injunc- 
tion nor prohibition]. Thus, 

Go, and let the sighs and tears 

be mine; nor let them rise 

from you as well, tortured, 

being without her, by your hateful courtesy.’ 

1. Found in a non-Vulgate version of the Sattasai (Weber 944). The 
literal sense of cd is probably "May they not arise from you, being without 
her, destroyed by your courtesy" But dakkhinna-hoassa could (it is just 
possible) stand for hata-daksinyasya, “possessing hateful (damned) courtesy.” 
Weber's suggestion that Aaasa may represent Artasya is improbable. The 
point of the verse lies in the lady's fury at her lover's affectionless politeness. 
Whether she says that the politeness is damned or that he is damned is not 
important. But it would ruin the verse to say that he is “carried away (Arta) 
by politeness.” 

L Here the word "go" is an injunction. We understand from 
the verse the intention of a woman who has been slighted! and whose 
pride has been deeply wounded. Her intent is: “Your union with this 
other mistress was not a careless adventure, but arose from the deepest 
love, as may be seen from your change of color and from your having 
inadvertently called me by her name. You remain here only out of the 

$14e4] 101 

courtesy [of pretending] to maintain our former relationship. You are 
a complete hypocrite.” There is no [suggested] prohibition here in the 
form of not letting him go, nor is there a non-prohibition in the form 
of some other injunction. 

1. Khanditá (“slighted”) has been defined by BANS 22.217 as a woman 
whose lover fails to visit her at the accustomed time. What is there meant is 
probably a lady of the harem who misses out on her "turn." SD defines the 
word as one whose lover arrives bearing signs of having enjoyed anotber. In 
the present instance the sign of the lover's faithlessness seems to bave been 
his calling the speaker by another woman's name (gotraskhalana). 

A Sometimes the literal meaning is in the form of a prohibition, 
while the suggested meaning takes a form that is neither (prohibition 
nor injunction). Thus, 

Turn back, I beg you. You are making trouble 
for other ladies stealing to their lovers. 

The moonlight of your countenance destroys 
their covering darkness, wretched woman.' 

1. Supplement to Sattasai (Weber, No. 968). Cf. also Hemacandra 
AC 1.22 (K.Anu. p. 55) and Mahimabbatta p. 747. The verse is addressed 
to an abhisdrikd, a woman who steals forth at night to visit ber lover. The 
simple explanation of the stanza is that it is merely complimentary. One may 
remark further that much of the charm of the verse comes from its having 
hatáse for the] t word, using it only after the pretty compliment bas shown 
tbat the lady in fact is far from being what that term implies. Hatáíe has 
much the same double sense that “wretched woman” has in English. It can 
be a term of compassion, if used of a woman who is truly wretched, or a term 
of reproach. if used of a woman who is vicious or cruel. The lady of our verse 
is shown, on the other hand, to be both beautiful and loving. But the simple 
explanation meets with a difficulty. In 2.4 and 2.5 our author will distinguish 
true dhvant, where the predominant meaning is a suggested rasa, from figures 
of speech like preyo larikára ("a figure of complimentary address") which in- 
volve a subordinate use of dhvani. Now by the simple explanation tbe present 

102 (§14e4 

verse would exemplify preyo larikára rather than true dhvani. I doubt that 
this would have troubled Ananda, who is not concerned at this point with 
whether dhvani is used for final meaning or as a subordinate element. But 
the difficulty did trouble Abhinava and is the cause of his whole comment. 

L "De" is a particle used in making a request. “A” has the 
sense of "tàyat"; so the meaning is: “Just turn back please," etc. As 
we understand the stanza to say “turn back” from your intended going, 
the literal sense is a prohibition. 

[One might explain the suggested sense as follows:| A lady had come 
to her lover's house, where he had slighted her in some such way as 
addressing her by another’s name, whereupon she had started to go 
home. He now turns her back with this clever piece of Battery: “You are 
putting di culties in the way not only of your own pleasure and mine, 
but of those other ladies. You will never attain a drop of happiness. 
So you are a most ‘wretched woman.'" Here the suggested sense is a 
particular compliment that represents the true feeling of the lover (BP: 
viz.. that no other woman is her equal]. 

Or, [we might say that] a lady has been warned by her female friend 
not to go, but scorns the warning. Now the friend tells her, "Not 
only are you making difficulties for yourself, cheapening yourself by 
this light conduct, and so are a ‘wretched woman,' but you are making 
difficulties for other women, stealing out to visit their lovers, by your 
lighting up the street with the moonlight of your countenance." Here 
the suggested sense is a particular compliment representing the feeling 
of the friend.' 

But in both these explanations, [the suggestion| comes back to rest 
in the literal sense, namely, a request to desist: from the intention of 
going back home, or from her going to a lover's house. And so this 
verse would be an example of a subordinated use of suggestion, that 
is of an alarikára, either preyo'lankdra or rasavadalarikàra, and not of 
[what our author calls} dhvani. So let us explain as follows. A certain 
lady is hurrying to her lover at night, who in turn is on his way to her 
and meets with her on the way. Pretending not to recognize her, he 
addresses her with this stanza. That is why he adds “wretched woman” 
as a joke at the end, to let her know who he is: “You are causing 
difficulty for other women too, so how can you hope to receive your 
own desire? So either come to my house, or let us go back to yours." 
So the suggested meaning is a clever compliment that represents an 

814fL] 103 

intention on the part of the lover, an intention that is in the form of 
neither [injunction nor prohibition| because the final sentence meaning 
allows of both. 

Others have explained the stanza as being the words of certain gen- 
tlemen of taste who happen to be present [as the lady passes by]. But 
I ask persons of taste whether it would be at all proper in such a case 
to use an expression like "wretched woman." 

1. This is, essentially, the explanation that Mahi 
(p. 474) 

A Sometimes the suggested meaning is made to be directed to a 
person (or persons) different from that (or those) to whom the literal 
meaning is directed. Thus: 

Who wouldn't be angry to see 
his dear wife with her lower lip 
You scorned my warning to smell 
the bee-holding lotus. Now you must 
[Non-Vulgate Sattasai Weber No. 886]? 

1. The stanza is quoted by Abhinava in Abh. on BANS 18.123; by Mam- 
mata 5.135; by Hemacandra K.Anu. 1 vs. 25, who in his AC repeats the 
comments of the Locana; and by SD on 5.2. 

L In the previous examples there has been shown to be a dif- 
ference between the literal and the suggested meanings even when the 
two meanings were addressed to the same person, viz., the monk, the 
traveler, the lover, or the abhisdrikd. Now he shows that the suggested 
Meaning may differ by its being addressed to a different person (or 

104 [(514fL 

persons) [from the person addressed by the literal]: sometimes the 
suggested meaning, etc. 

Who wouldn't: that is, even a man without jealousy would be 
angry if he even [thought that he] saw it, that is, if he noticed her lip 
as being wounded because it appeared different for some reason even 
if it had not been [wounded].' 

Is made to be directed: What he means is that although various 
persons lie in the direction of application, a sensitive reader can make 
out the correct direction. [Abhinava here gives a Sanskrit translation 
of the Prakrit verse, literally:] Who wouldn't be angry on seeing his 
dear wife's lower lip with a wound? O you whose habit it is to smell 
bee-concealing lotuses, you who are averse to being prevented, now you 
must suffer. 

O you whose habit it is to smell bee-concealing lotuses: be 
cause a person's habit cannot in any way be prevented. Averse: un- 
willing to accept. Being prevented: prevention. Now you must 
suffer: viz., a long and severe scolding. 

The meaning of the stanza is as follows. An unfaithful wife has had 
her lip bitten by a lover. To save her from her husband's reproaches 
she is here addressed by a clever female friend, who knows that the 
husband is nearby but pretends not to see him. Now you must suf- 
fer: the literal sense is directed to the adulterous wife. The suggested 
sense, on the other hand, is directed to the husband and informs him 
that she is not guilty of offense.? There is also a suggestion directed to 
the neighbors who, if they hear the wife being roundly abused by the 
husband, may suspect her of misconduct. The suggestion in this case 
is the assurance provided by this concealment of her adultery. There is 
a suggestion directed to her fellow wife, who would be delighted by the 
abuse of her rival and by [the news of} her adultery. The suggestion lies 
in the word dear (“dear wife^),? which shows that the wife addressed 
is the more attractive. There is a suggestion to the adulterous friend of 
the speaker, informing her, "You should not take on humiliation at the 
thought of being accused of bad character in front of your fellow wife; 
rather, you should take to yourself high esteem and now shine forth (sa- 
hasva).* To the wife's secret lover there is a suggestion, telling him that 
"Today I have thus saved your heart's beloved who loves you in secret, 
but you must not bite her again in a place that is so obvious." To any- 
one clever who is standing nearby the speaker's cleverness is suggested, 
[as though she were to say,] "This is the way I have concealed things." 
All of this is indicated by the expression is made to be directed. 

§14g A] 105 

1. akrtud, even if it had not been wounded: this appears to be the inter- 
pretation of the Kaumudi and of BP. The grammatical interpretation, “see 
ing her lip wounded even if he had not done it," makes no sense. Obvi- 
ously the husband would not be angry if he had bitten his wife's lip himself. 
2. Most printed texts insert here the brief sentence: sahasvety api ca tad- 
vigayom vyarigyam, “There is also a suggestion that he must suffer.” The 
sentence is missing, however, from the MS used by the Nirnayasagar edition. 
The trouble with it is that such a suggestion, if present, would be directed 
to the reader, not to the husband as portrayed by the poem. Pathak omits 
the sentence in his Hindi translation. 3. What I have translated as “dear 
wife” is in Sanskrit a single word, priydydh, a word which has two meanings, 
"wife" and "beloved." The literal meaning here is simply wife, but a sug- 
gestion arises sabdabolàt, from the verbal force, to the effect that this wife 
is also her husband's beloved, that is to say, the one of his wives that he 
finds most attractive. — 4. [n assigning the sense of Sobhasva as a second 
meaning to soAgsva, Abbinava is probably following some Prakrit grammar. 
Hemacandra (8.)4.100 gives the root sah as a synonym of rój (rójati). The 
Kaumudi prefers the ancient Vedic meaning of sah, to win or overcome: “sa- 
honom is here used in the sense of overcoming her fellow wife." That any 
second meaning is intended in sahasva seems to me most unlikely. 

A Other differences of the suggested meaning from the literal 
meaning are possible along these lines. We have merely indicated the 
general direction. How the second variety of suggested meaning, viz., 
alarikáradhvani, differs from the literal will be shown in detail in what 
follows. But the third variety, involving rosa, etc., which appears as 
something implied by the inherent capability of the literal sense but as 
an object on which no words can operate directly, must necessarily be 
different from the literal. This may be shown [more formally]. For if 
such states as-rasa are to be denoted, it must be either by reporting 
them under their own names, or through conveying them by means of 
the vibhávas, etc.! If the former were the case, it would follow that 
wherever the rasas, etc.. were not reported by name there could be 
no apprehension of them. But it is not true that they are everywhere 
reported by name. Even where they are, our apprehension of them 

106 [§14g A 

is through their being conveyed by means of particular vibhávas, etc. 
This apprehension, while it may later be referred to by name,” is not 
produced by the naming, because in other cases we do not find it. For 
in a poem which merely uses such words as “erotic,” etc., but fails to 
convey the vibhávas, there is not even the slightest apprehension that 
the poem contains any raso. And since there is the apprehension of 
rasas, etc., from particular vibhávas without any naming of these rosas 
and there is no apprehension of them from the mere naming of them, it 
follows by the application of positive and negative concomitance that 
the rasas, etc., are implied by the force of things that are literally 
denoted and are in no way denoted themselves. So it stands proven 
that the third variety [of suggested meaning} also is diferent from the 
literal meaning. That we apprehend it as though it were simultaneous 
with the literal meaning will be shown in what follows. 

1. This sentence lacks Ánanda's usual clarity, for the conveying of rasa by 
means of the vibhávas is not an instance of vácyatva (“being denoted”) at all, 
but of vyarigyatva (“being suggested"). Abhinava tries to exculpate our author 
by supplying tátparyasaktyà. — 2. This concession is worth remarking on. for 
jt is generally overlooked by later àlankárikas. Later authors were generally 
of the opinion that to use a word denoting the actual emotion (bháva) or rasa 
copstituted a major fault, so much so that such cases could not be considered 
examples of dhvani. Aaanda’s concession allows for such words if used as an 
anuvdda (mere reference). The concession allows many fine poems to pass 
muster which are lowered in value by the later critics. 

L In what follows: viz., in Chapter Two, where suggestion 
in which the literal meaning is intended but is subordinated to a sec- 
ond meaning (vivaksitányaparavácya) is said to be of two types, “one 
where the suggestion is produced without apparent sequence [i.e., im- 
mediately, together with the primary meaning], the other where the 
sequence is apparent," (2.2). There, in describing the second of these 
types, [the variety bere referred to, namely alarikdra-dhvani, is dealt 
_with in detail; see 2.20-21 and 25-26]. While it is easy to summarize 
vastu-dhvani under the heads of injunction, prohibition, and neither in- 
junction nor prohibition, it is not easy to summarize alarikára-dhvani, 
because the figures of speech (alarikáras) are so numerous. And so he 
says: in detail. 

But the third variety: the word "but" is used to point a contrast. 
In the first place, the property of being expressible by the denotative 

§ 1.4g L] 107 

force of words (as well as by suggestion} attaches to a situation (vastu) 
or to a figure of speech (alarikāra). On the other hand, a rasa, an 
emotion (bháva), an improper rosa or emotion, or the cessation of a 
rasa or emotion, are never directly denoted. They appear rather as 
matters that come to life in the process of being relished (dsvddya- 
mäna), and for this there is no explanation other than the operation 
of suggestion. For we cannot suspect as being here at work any of the 
conditions of laksaná, such as blocking of the primary meaning, because 
there is here no halting gait in the journey [from word to meaning; cf. 
144b L, note 17]. 

[Definitions of rasa, rasábhàsa, etc.) 

Rosa appears when a stable state of mind (cittavrtti), constantly di- 
tected toward a proper object,! is aesthetically relished. Bhdva appears 
when a transitory state is so relished. The improper variety (dbhé- 
sa) of rasa or bháva appears when either of them is directed toward 
an improper object, as when Ravana's love is directed toward Sita.” 
While that case really belongs to the comic flavor, in accordance with 
[Bharata's] dictum that “the erotic leads to the comic,”? that stage of 
realization overtakes the audience only later. Since the relish one ex- 
periences in the stage where one is identifying (the portrayed emotion 
with one's own] is of love, the rasa will appear to be the erotic rasa as 
long as we overlook the broader context, as we do when hearing: 

I merely heard her name 
and it acted as a magnet or a maddening charm.* 

This is therefore a case of the improper or spurious erotic, (not of the 
comic}. An emotion (bhdva) which goes to form an improper rasa is an 
"improper emotion" (bhàvábhása). As the cessation or checking of an 
advanced emotion is especially delightful to the heart, it is separately 
mentioned [in the list that we just gave], although it is actually included 
[in the term 5hàva].5 An example is: 

They lay upon the bed each turned asi 

and suffering in silence; 

though love still dwelt within their hearts 

each feared a loss of pride 

But then from out the corner of their eyes 

the sidelong glances met 

and the quarrel broke in laughter they turned 
and clasped each other's neck. 

108 [$1.4g L 
Here we have the cessation of a pride which has taken the form of 
jealous anger.’ 

Now this suggested entity, rasa or the like, is not generated within 
us after the fashion that joy is generated from [the direct force of] 
the words "A son is born to you.”® Nor does it come from the sec- 
ondary power of the words. Rather, it makes itself felt (parisphurati) 
as something the whole life of which consists in the ongoing process 
of relishing and which thereby differs from something like joy or grief 
that is a finished or frozen state? This process of tasting arises in a 
sensitive person through his empathy upon apprehending the vibhávas 
and anubhávas, an empathy made possible by his heart's being in tune 
with [the poetic message]. Our author states this: which appears as 
(something, etc.). And so’? in these [instances of rasa, etc.) suggestion 
is an operation of a word as helped out by [that word's literal] mean- 
ing. But this [literal] meaning, which will be a vibháva or the like, 
does not generate an emotion like the joy generated by the birth of a 
son. So suggestion is said to be an operation different from generation 
(janana), an operation which belongs to meaning as well [as to word].!! 

Under their own names: reporting them by the operation of de- 
notation by using the words srrigára (“the erotic”), etc. By means of 
the vibhdvas, etc.: He means “through the sentence meaning.” Here, 
by ruling out rasa, which consists essentially in the process of relishing, 
by the use of positive and negative concomitance, from the use of the 
very words which denote it, he shows that these concomitances belong 
to suggestion. 

It is not true that they are everywhere (reported by name]: 
for example, in this stanza of Bhattenduraja, 

A tremulousness of the eyes, 

hesitating in mid-glance; 

limbs daily growing thinner 

like severed lotus stems 

and cheeks so pale they seemed 

to imitate white dürvá grass: 

such was the costume put on by the gopis 

as they and Krishna came of age. 
Here, after we become aware of the anubhdvas and vibhávas and have 
joined ourself to them by empathy, the meaning, in the form of a rusa, 
makes itself felt (sphurati) as that which is blissfully relished by the self- 
consciousness, which is colored by latent impressions (vdsand, see 2.4 L, 
note 6) responsive to these vibhdvas and anubhdvas; all of this without 

§1.4gL] 109 

the use of any such words as abhilása (desire), cinté (worry), autsuk- 
ya (eagerness), nidrà (sleep), adhrti (frailty), gléni (drooping), dlasya 
(languor), Srama (weariness), smrti (remembrance), vitarka (specula- 
tion), or the like.!? 

Having thus shown the failure of a negative concomitance,'! he goes 
on to show the failure of the positive concomitance:'4 even where 
they are. "They" refers to words that directly name a rasa, etc. 
Through their being conveyed: by the conveying of vibhávas, etc., 
through tbe use of words. It may be merely (referred to by its 
name]: as in the following stanza: 

When Madhu's foe had left for Dvárak& 

his R&ádhá hugged the slender tree 

on KAlindi’s bank from whose wealth of frondage 
he had in time past given her gifts. 

With bigh-pitched voice and heavy falling tears 
she sang a song with longing, 

to which the birds who swam upon the wave 
gave back a yearning cry.'* 

In this stanza the vibhávas and anubhdvas are clearly apprehended!5 
and longing is [thereby] conveyed as the object of one's relisb.!" The 
word sotkanthé (“with longing") gives us no more than has already 
been given. But although it is merely a reference to the anubháva 
[which has been learned through non-direct means|, the word is useful, 
as it is employed to draw together the stated anubhdvas [e.g., the sad 
song and tears of the heroine] with the word “yearning” [applied to 
the cry of the birds]. For if the poet had conveyed the whole set of 
anubhávas all over again, the stanza would suffer from tautology and 
we should not empathize. 

Is not produced by the meaning:!* he gives the reason for this 
with the words, because in other cases, etc.; for example, in the 
stanza, “A tremulousness of the eyes," (where there is no naming of 
the rasa or its components]. The sense is that A cannot be produced 
by B if A comes into existence in the absence of B. He strengthens [the 
statement of] our not finding rasa by the next sentence, beginning na 
hi. He clarifies the expression such mere words, etc., by (adding the 
condition that the same poem fail to convey] the vibhávas, etc. 

In a poem: whereas in your opinion [i.e., according to the opponent 
who claims that naming the rasas should give rise to aesthetic relish], 
it should become poetry. Not even the slightest: as in the following 

110 [§14gL 

stanza there is not the slightest aesthetic relish although it names all 
the rasas. 

The erotic, comic, tragic, and heroic, 
the flavors of fury, fear, disgust and wonder: 
such are the rasas, which number eight, 
in our tradition of the drama. 
{BANS 6.15} 

Having thus shown by a persuasive argument employing negative and 
positive concomitance that the rasas, etc., are absent [from a verse] 
when they are directly named in it, he now sums up the matter in 
similar fashion!? in the passage beginning with and since there is 
and ending with and are in no way [denoted]. 

[Explanation of the phrase by the force of things which are lit- 
erally denoted.|? When the suggestion of rasa is ascribed to a word, 
the force (sámarthya), that is, the cooperating force, viz., the vibhávas, 
etc., is the directly denoted meaning. When tbe suggestion of rasa is 
ascribed to the directly denoted meaning—inasmuch as the suggestion 
of rasa is not a case of one thing's begetting another, because of the 
different nature (yogaksema) of joy at the birth of a son; and not a 
case of one thing's being inferred from another, because of the distinct 
nature of the inference of a man's eating at night from the premise of 
his being fat compounded with his not eating in the daytime?! —then 
the force (sáàmarthyo, Sakti) of this meaning is the totality of denota- 
tive words arranged in their particular way.?? Thus the suggesting is an 
operation of both word and meaning. And so, in addressing the alter- 
natives [A, that rosa, etc., can be conveyed through tbe mere naming 
of a rasa, etc.; and B, that rasa, etc., are conveyed by one's furnishing 
the vibhávas, etc.], the former bas been refuted, while the latter has 
been partly refuted and partly accepted. If taken as meaning that the 
operation (by which the vibhávas lead to rasa, etc.| is a begetting or 
inferring, that is refuted; if taken as meaning that the operation is a 
suggesting, that is accepted. 

He who thinks that even here suggestion is nothing more than tátpar- 
yasakti (the power of the sentence meaning) does not know the truth of 
the matter. For in a sentence that conveys the vibhávas and anubhávas, 
the tátparyasokti exhausts itself in giving the syntax (samsarga) [of the 
sentence] or its difference [in meaning from that of other sentences]; 
it does not concern rasa, the essence of which consists in the process 
of relishing. Let us say no more. 

§ 14g L) 111 

The word so is used in the sense of cause. The connection is: and for 
this cause also, the third variety too [of suggested meaning] is different 
from tbe literal. 

As though it were simultaneous: By saying “as though” he shows 
that although there is really a succession, the succession is not noticed. 
In what follows: in Chapter Two (2.20-21 and 25-26). 

1. cittavrtteh sthdyinydh: the more usual term would be stháyi- 
bhávasya. The phrase aucityena provrttau is taken from Udbhata (Indurája 
4.5, Vivrti 4.9). — 2. The rasa is improper because Sita is another's wife 
and because the emotion is not reciprocated. On the concept of rasábhàsa 
see Sivaprasad Bhattacharya, Calc. Or. JL 2, pp. 237-247, and J. L. Masson 
and M. V. Patwardhan, Aesthetic Rapture I, p. 42 and II, pp. 57-58. The 
concept of rasdbhdsa is highly restrictive of literature. If we are to limit rosa, 
the sole aim of literature, to only such subjects as conform to propriety and 
even to the 5^ tros, as Udbhata would have it, not a little of Sanskrit litera- 
ture and surely the greater part of Western literature will be judged to be of 
little worth. Abhinava seems to have been the first Indian critic to face this 
problem and find an answer: the dbhdsatva, the impropriety, of such experi- 
ences is something we realize only later; during the actual experience we are 
absorbed. 3. BANS 6.39. The next verse specifies that it is when the erotic 
is parodied (rrigáránukrti) that it becomes comic. — 4. A larger fragment 
of this stanza is introduced at 2.3 L by the identification "Rávonakáwye" (see 
also Abh. 6.40), but whether this means "in a poem called the Rávanakávyo," 
or merely "in a poem about Rávana" is not clear. The full stanza is given in 
Hemacandra's AC on K.Anv. 2.55 as dirékarsanamohamantra iva me tan- 
némni yáte írutim, cetohkálokalám api prosahate ndvasthitim tám vind / 
etair àkulitasya wiksatarater arigair anarigdturaih sampadyeta katham tadép- 
tisukham ity etan na vedmi sphutam. 5. One might regard all the conditions 
of bháva, such as bhávodbhava, bhávasandhi, bhdvasabalatd, and bhavaprasama 
(see 2.3 L) as being included in bháva. 6. Amaru Sat. 23, quoted in nearly 
every anthology of Sanskrit. (a) vitottaram: probably, “without reply, in si- 
lence.” BP's interpretation “without aay of the action that [normally] follows 
lying down in bed” seems to me farfetched. (b) The MS and anthologies 
vary between kanthagroham, adverb, and kanthagrahah, bahuvrihi The for- 
mer makes for clearer syntax. 7. Whether one considers such verses as 
examples of bhávaprasama or bhávodaya depends on whether one finds more 
charm (camatkdra) in the description of the ceasing emotion or the originating 
one. Mammata (4.51) quotes Amaru 22. a verse similar to the present one, as 
an example of bhávodaya. 8. Abhinava would here make another. radical, 
distinction between suggestion and the other powers of words. The denotative 
and secondary powers (abhidhá-sokti and laksaná-fakti) are able to give us 
only cognitions or concepts. The joy that may follow from "You have a new 

[514g L 

son," or the grief that may follow from “Your unmarried daugher is. pregnant" 
(KP ed. Jbalkikar, p. 229) is a subsequent development growing out of the 
word-meanings or concepts. [n the case of suggestion, on the other hand, the 
meaning itself is the rasa, the flavor that we relish. Aesthetic pleasure is not 
the result of a meaning; it is the meaning itself. — 9. Siddho-svabhàávo: the 
comment ies, | think, fail to understand this term. BP's Girst explanation 
misunderstands the syntax, taking the whole compound as "different from 
sukha, etc., by being of a siddhasvabháva," as if it were sukha, etc., that were 
sddhya Its second explanation agrees with the Kaumudi and supposes that 
sukhádi stands for rati and the other sthdyibhdvas. The correct interpreta- 
tion surely will connect sukhddi with the put janmeharga just referred to 
and shortly to be mentioned again. The contrast is between (a) the denota- 
tive force of words which produces a meaning, which in turp generates a fixed 
mental reaction, pleasure or grief, and (b) the suggestive force of words which 
produces an ongoing process of relishing or enjoyment. 

10. The Vrtti has stated that rasadhvani is vácyasárnarthyóksipta. And 
so (tena) the suggestive power of the word must be helped out by a meaning, 
viz., the vdcyo meaning. 11. The literal meaning should not be said to help 
generate (janayati) the rosa; it should be said to help suggest (dhvanayatr) 
the rasa. Abhinava is merely distinguishing the primary production of the 
rasas from the secondary production of pleasure and pain. Later, on 2.4, he 
will admit, even insist, that the rusas are produced (utpddyante). 12. Of 
the ten words on the list, abhtldsa probably represents the sthdyibhdva, rati; 
the five words gláni, srama, cintd, autsukya, and nidrá are listed by BANS 6.45 
as denoting anubhdvas of vipralambhasrrigáro; the three words dlasya, smrti, 
and vitarka denote vyabhicdribhdvas, listed in BANS 7.47, 53, 91. Only of 
Gdhrti can I not furnish a technical assignment. 13. The negative con- 
comitance would be: “Where there are no words directly namiog the rosa 
or its components, there is no rasa-experience.” 14. This would be in the 
form, “Wherever there are words directly naming the rasa or its components, 
there is a rasa-experience.” 15. The author of the stanza is unknown and 
the text of the first line is in question. All the printed texts write taddat- 
tayhampanatam, “bent down by the leap which had been given by him." It 
is certainly odd to speak of “giving one's leap to a tree." BP tries to make 
out that this was the tree from which Krishna leaped into the Kálindi. But 
the tradition is unanimous that that tree was a kadamba (Jfarivam$a 55.57, 
Vignu P. 5.7.10, Bhág.P. 10.16.6), not a vañjula. Neither is it clear just what 
tree is here meant by vafijula, except that it cannot be a kadamba. The word is 
used for an asoka, or a syandana (=tiniga, the Anglo-Indian sissoo), or a reed. 
Whichever tree is meant, I prefer the reading of the Malayalam MS quoted by 
Kuppuswami Sastn in his edition of the Kaumudi, viz., taddattasampannatám, 
and have translated accordingly. — 16. The dlambana-vibhdvas are Krishna 
and Radha: the uddipana-vibháva is the bank of the Kālindī; the anubhávas 

§1.5 K} 113 

are embracing the varijula, shedding tears, etc. 17. This is as much as to 
say that longing is thereby suggested. 18. One should place a danda after 
the words atanmayibhávo và. In the text the words na tu tatkrta should be 
printed in boldface. 19. Viz., again by the use of positive and negative 

20. The complicated and highly improbable interpretation which follows 
is occasioned by Abhinava's desire to bring this statement, which ascribes 
the suggestion of rasa, etc., only to meanings (viz., to vibhávos, etc.), into 
line with the opinion elsewhere expressed by Ananda that the suggestion of 
rasa derives from both meaning and word. To accomplish this aim, Abhinava 
takes abhidheya-sámarthya first as a karmadhdroyo compound and next as a 
gosthi-tatpurusa. The meaning assigned to sdmarthya differs in the one case 
from the other, 21. The stock example of the arthdpatti of the Mimàmsá. 
here reduced, as it is by the Nydya, to an inference. 22. BP: put together 
with such gunas and olarikáros as are conducive to raso. 23. That relating 
and differentiating are the two functions of the sentence is a notion first found 
in Mohábhásya 2.1.1. Värt. 2 (Kielhorn ed., 1.364.24; S. D. Joshi, ed. and 
trans. of 2.1.1, para. 84). To explain: the sentence “Göm dnaya” not only 
relates the object cow to the action of the addressed person; it differentiates 
the command from one concerning horses or concerning some other action. 

K lt is just this meaning! that is the soul of poetry. And so it 
was that, long ago, grief, arising in the first poet from the separation 
of the pair of curlews, became verse. 

1. In order to make sense of the Kárikà, we must take “this meaning" to 
refer not to the suggested meaning in general that was mentioned in 1.4 K, but 
specifically to rosa, etc., the third type of suggested meaning, which has been 
mentioned only by the Vrtti on 1.4. It was this element arising in Valmiki, 
whether one regard it with Ananda as the bhdva, soka, or with Abhinava as 
the karunarasa, that produced the first poem, for it is rasa, etc., that gives 
life to poetry as the soul gives life to the body. Note that Ananda’ 's concept 
of bhéva and rosa is much simpler than Abhinava's. To Ananda rasa is no 
more than the sharpening of VAlmiki's emotion of grief. See Introduction, 
Pp. 15-19. The quarter stanza slokah fokatvam dgatah is quoted from Rém 


A It is just this [inner] meaning that is the essence of a poem, 
which has [outward] beauty in its wealth of direct meaning, word, and 
structure.! And so it was that the grief (Soka) of the first poet, Valmiki, 
born of the wailing of the cock curlew desolated by loss of its slain 
mate,” turned into verse (sloka). For grief is the basic emotion of the 
flavor of compassion (karunaresa) [which, as has been said, appears 
only as suggested). Although other types of suggested meaning may 
be found, they can all be supplied from the mention of rasa and bhava 
because those are the most important. 

1. The phrase vácyavácokaraconá recurs at 1.84. What is here meant is 
the choice of word, direct meaning, and structure (degree of compounding and 
degree of phonetic harshness) appropriate to the rasa that is to be suggested 
and that forms the inner or essential meaning. Vécya is used in distinction 
from vyarigya. 2. Both the reading and the sense of the passage have been 
questioned, wrongly. The reading nihatasohacari is found in the Kerala MS, 
in Krishnamoorthy's MB MS (see p. 311 of his ed.), in the text of the Locana, 
and in the semi-quotations by Rajasekhara's Kàv. M. p. 7 and by Candidása's 
Dipikà (see Krishnamoorthy loc. cit.). The reading sannihitasohacari occurs 
only in two of the Nirnaya Sagar MS (KM ed.). The difficulty with the sense is 
that in the form of the legend given in the Rámáyana, a form that every Indian 
schoolboy used to know, it is the male bird that was killed (Ram. 1.2.10). It 
was the grief of the female bird that Valmiki transformed into verse. In 
order to reconcile these traditions the learned Kuppusvami Sastri (Upalocana 
pP. 163-164) proposed an unnatural analysis of Ánanda's compound, taking 
nihata by a frog's leap with krourica instead of with sahacari. Pt. Badari Nath 
Sarmi in his Didhits emended the text. All needlessly. Ananda has altered the 
legend to suit his purposes. See J. L. Masson, “Who Killed Cock Kraunca," 
J.O.I. Baroda 18 (3), March 1969. 3. The phrase pratiyamánarüpa evet 
pratipáditam, translated above by the words placed in brackets, appears in 
most of the MSS, but Abhinava makes no mention of it. One cannot say with 
certainty whether it has crept into the text from a marginal annotation or 
whether it has dropped out of an early copy of the text by haplography, the 
eye of the scribe having jumped from the initial word of pratiyamánarüpa to 
the pratiyamdnasya of the next sentence. It is missing from the Kashi and 
Vidyàbhavan texts. 

L {Comment on the Karikà:] So far, by stating that the 
suggested, on the other hand, is something different (1.4 K), 

§1.5 L] 115 

the nature of suggestion has been explained. Now, by making use of a 
well-known legend, he will show that it is the soul of poetry: the soul 
of poetry. 

It is just this: while the antecedent is suggested meaning in general, 
what we are here to think of is the third variety, namely suggested rasa 
(rasadhvani), for that follows from the use of the legend and from the 
immediately preceding passage of the Vrtti. So it is rasa that is the real 
soul of poetry. Vostudhvani and alarikáradhvani, however, regularly end 
up in [producing] rasa, and it was in order to mark their superiority to 
the literal sense that he said [in 1.1 K] that dhvani in general was the 
soul of poetry. 

Grief: That grief which arose from the separation of the pair 
of curlews, that is, from the destruction of the mating arising from 
the killing of the bird's mate, a grief which was a basic emotion differ- 
ent, because of its hopelessness,! from the basic emotion of love found 
in love-in-separation: that grief, by the poet's ruminating upon its 
[álambana-] vibhávas [i.e., the birds] in their [unhappy] state and on 
the anubhávas arising therefrom, such as the wailing [of the surviving 
bird], met with a response from his heart and with bis identifying [of 
the bird's grief with the grief in his own memory] and so transformed it- 
self into a process of relishing.” It thus became the flavor of compassion 
(karunarusa), which differs from ordinary grief by its being experienced 
primarily as a melting of one's thoughts.’ Then, like the spilling over 
of a jar filled with liquid, like the pouring forth of one's emotion into a 
cry of lament, this (grief now transformed into the rasa of compassion] 
found its final form in a verse cast into fixed form of meter and into 
appropriate words, for cries of lament and the like are suggestive of 
a state of mind without the need of semantic convention; appropriate 
also because Valmiki was wholly engrossed and the words came from 
him naturally. His words were: 

May you never find honor, Nisáda, 
for everlasting years, 

who have shot the loving mate 
from this pair of curlew birds.‘ 

But we must not suppose that the sage experienced grief, for if that 
were the case there would be no occasion for calling rasa the soul [of 
Poetry}, as the poet would actually be in pain, pained by that grief. 
Nor does such a state [BP: the exalted state of being able to pronounce 
a curse, or to write a sloka] belong to one who is afflicted with pain. 

116 [§1.5 L 

Thus, since it forms the nature of this overflow [viz., the verse just 
quoted] of the flavor of compassion, of which the abiding emotion is 
a grief amenable to relishing, this rosa is therefore the soul of poetry, 
that is, its essential nature, that which produces a result beyond the 
reach of any other word-powers [than suggestion]. It has been said (by 
Bhattan&yaka] in the Hrdayadarpano, 

Until he is filled with this rasa 
the poet does not spill it forth. 

[In the quotation given above from the Ramdyana] the form agamah 
shows Vedic retention of the augment. 

It is just this: by the word “just” he would say that there is no 
other soul [of poetry]. Accordingly, Bhattandyaka is wrong when he 
writes: "One may distinguish the Sdstras by the prominence they give 
to the word. One knows that stories are wedded to meaning. One 
forms a just notion of a poem by subordinating these two, viz., word 
and meaning, and making the operation (vyépéra) paramount." For 
if the "operation" he speaks of is essentially suggestion and consists 
in relishing, he is saying nothing new, while if he means the operation 
to be denotation (abhidhd), we have already shown that it holds no 

[Comment on the Vrtti.] 

The Vrtti comments on the (Kérika’s} stanza. In its wealth of 
direct meaning, word, and structure: that is, because a poem is varied 
in accordance with whatever rasa is to be suggested. [Only such a 
composition is called a poem and] therefore, although suggestion occurs 
everywhere, we do not speak of [poetry being everywhere], just as we 
speak of life only in some places although the soul exists everywhere, 
as we have said before [1.4b L]. So there is no occasion for what is 
objected in the Hrdayadarpana, that “we should have to use the term 
poetry everywhere."* 

Its slain mate: here we have the [dlambana-| vibháva; the wailing: 
with this word, the anubhdva. Born: one must supply, “through being 
the object of his relish."? 

But if verse (sloka) arose from relished grief (soka), why is the thing 
that is suggested [i.e., the raso] said to be the soul of poetry [rather than 
grief]? It is with a view to this objection that he says: for grief is, etc. 
Grief is the basic emotion of the rasa of compassion, for compassion 

§15 1L] 117 

consists of relishing (or aesthetically enjoying) grief. That is to say, 
where we have the basic emotion grief, a thought-trend that fits with 
the vibhàvas and anubhàvas of this grief, if it is relished (literally, if it is 
chewed over and over), becomes a raso and so from its aptitude [toward 
this end] one speaks of [any] basic emotion as becoming a rasa.!? For 
the basic emotion is put to use in the process of relishing: through a 
succession of memory-elements it adds together a thought-trend which 
one has already experienced in one's own life to one which one infers 
in another's life, and so establishes a correspondence in one's heart.!! 

It may be objected that it is anything that takes the form of a sug- 
gested meaning that forms the soul [of poetry] and that three varieties 
of this [suggested meaning] have been stated, not simply that which 
takes the form of rasa, whereas the Valmiki legend seems to say that 
only rasa is the soul. Our author foresees this objection and accepts 
it, saying although other types of suggested meaning: the other 
types are vostudhvani and alarikáradhvani. 

The inclusion of the word bhava indicates that even a transitory 
State (vyabhicáribháva) may form the life of a verse although the rel- 
ishing of it is not complete in itself [but will go on to a relishing of 
a rasa) and although it never achieves the position of a rasa belong- 
ing to the final state of relishing a basic emotion." An example is the 

Rubbing one nail with the tip of another, 
turning about her loose bracelet, 

slowly drawing a line on the earth, 

her anklet softly jingling ... 

Here we have (the transitory state of] shyness [forming the life of the 

Furthermore, by the words rasa and bhava there are included the 
improper varieties (dbhdsa) of these as well as the termination (pra- 
sama) of these, for although there are many sub-types, a single form 
rups through them all. 

Because these are the most important: they are so because the 
other types end up in or lead to rosa. Vastudhvani and alarikáradhvani, 
while they are not complete in themselves, can be called the soul of 
a verse from their aptitude, that is, because of their ability [also] to 
furnish [a delight] that lies beyond the reach of other word-powers. 

118 ($1.5L 

1. Grief characterizes not only karunarasa (the flavor of compassion or 
tragedy), but also that variety of tbe erotic flavor that is based on the sepa- 
ration of lovers (vipralambhasrrigára). Between the two sorts of grief is this 
difference: the grief of tragedy, as in the present instance, expects no relief; the 
grief of separated lovers looks forward to reunion. The term nirapeksabháva 
for the hopelessness of tragic grief is taken from BANS 6.45, near end of 
prose. 2. The all-important transformation from the emotion, grief (bhdva. 
Soka) of the character portrayed, to the relish (dsvdda) or flavor of compassion 
(karunarasa) of the poet or of his audience is here passed over very rapidly. 
Abhinava furnishes more detail at the end of his comment on the present pas- 
sage and in commenting on 1.18 and 2.4. The sympathetic response (Ardaya- 
samvdda) to the vibhávas and anubhávas is said to “transcend the experience 
of the workaday world" (2.4 L). Where the Westerner may think of empathy 
as rendering Hamlet's griefs and problems his own, Abhinava thinks of the 
process of empathy with, say, Rama, or witb the grieving bird, as Liberating 
one’s personal memory of grief into a universal, impersonal flavor. 3. This 
melting of the mind (druti) is one of the symptoms assigned to the relishing 
of rasa by Bhattanàyaka (2.4 L). The others are expansion (vistara) and ra- 
diance (vikdsa). At tbe end of 2.4 L, Abhinava somewhat grudgingly accepts 
these characterizations from his rival, but insists that they are not exhaustive. 
4. Rámáyano 1.2.14. The pair of birds had been mating as the Nisáda shot, a 
fact that doubtless would bave brought a curse upon him even if Valmiki had 
Dot been present to versify it; compare the curse of Pandu, MBh 1.109. The 
legend is built on a folk etymology deriving sloka from soka, and from the 
despised status of the Nisáda caste. As the incident is told at tbe beginning 
of the Rámáyana, it has been taken as introducing the tragic flavor of that 
work. If we accept Ánanda's alteration of the story, one may take the bunter 
to foreshadow Rávana; the slain ben-bird, the kidnapped Sita; and the heart- 
broken survivor, Rama. 5. One must remember that Abhinava regards rosa 
as a form of bliss. Naturally it must be different from grief, which is a painful 
emotion. Here Abbinava is writing of the poet. On 2.4 he brings out the same 
contrast in tbe case of the audience: if they felt pain at a representation of the 
Ramayana story, for example, they would not return to the theater. Masson 
has written of the Indian recognition of the poet's need to distance himself 
from his emotions before writing of them (Séntarasa, p. 84). Ingalls would 
add that if we follow Abhinava's account strictly, we must say tbat the poet 
in fact never writes of his griefs. He writes only of the griefs of others, which 
he has relished. By relishing them it is implied that he has lost his own griefs 
within them. This is a far more refined view than that of Ananda, who writes 
quite unconcernedly of the "grief of the first poet.” And Abhinava's view, 
as Masson points out in the passage just referred to, differs from that of the 
Ramayana itself, whicb in narrating the incident speaks time after time of the 
poet's grief and pity. 6. According to the Kdsiké, such cases are covered by 

81.8 A] 119 

the word bohulam in Pan. 6.4.75. They are noticed by Whitney, para. 579e. 
7. What Bbattanáyaka meant doubtless was neither. He must have meant the 
operations (bhávaná, bhoga) of the word's powers of bhdvakatva and bhogakrt- 
tva, these being the special terms by which he explained the nature of poetry; 
see Introduction, p. 36, and 24 L, near end. One may note that Vidyádbara 
in his Ekàvoli (pp. 13-15) combines this triple distinction of Bhattanáyaka's 
with Abbinava's doctrine of the three vyutpattihetavah (means of iostruction; 
see p. 71), but in doing so substitutes dhvanipradhéna for Bhattanáàyaka's 
vyápárapradhána. 8. Bhattandyaka must have singled out 1.5 for criticism, 
saying that if suggestion in general is to be called the soul of poet , the title 
of poetry will be assigned to almost every sentence, as suggestion is found in 
every metaphor and trope. There are two ways to silence this objection: by 
showing that it is only resadhvani that really qualifies as the soul of poetry; or 
by specifying other properties that poetry must have. Abhinava used the first 
way in commenting on the Kdrikd. He now uses the second in commenting on 
the Vitti. 9. See how subtly Abhinava alters the meaning of his text. We 
are not to think of the grief as belonging to Valmiki. The grief is the bird's. 
It gives: birth in Valmiki not to grief but to a relishing of the bird's grief. 

10. Note that this statement is metaphorical, oot exact. It is one's 
own cittaurtti (thought-trend, state of mind), not the basic emotion, that 
becomes the rasa. How it does so is indicated in the next sentence. 11. It is 
this hrdayasamváda (response of the heart) which permits the expansion and 
depersopalization of one's own emotions. — 12. It seems highly improbable 
that Ananda meant any such thing. By bhdva he probably meant stháyibháva 
and be probably intended such a sthdyibhava, grief (suka), to be the meaning 
suggested by Vàlmiki's first verse. 

K Sarasvati. [working] within great poets, in pouring forth tbis 
sweet matter (arthavestu) [viz., the emotions and flavors] reveals a 
special, vibrant, genius (pratibhd), which is superhuman. 

A The divine speech of great poets, in pouring forth this essen- 
tial matter (vastu-tattva), reveals a special, vibrant. genius, whicb is 
superhuman. Thus it is in this world, where there has been a long 

120 [$1.64 

succession of poets of every possible kind, that only two or three, or 
maybe five or six, such as Kālidāsa, can be counted as great poets. 

L Having thus shown by means of a legend that the suggested 
meaning is the soul of poetry, he now shows that this is also a matter 
of one's own experience. Sarasvati: He means that goddess in the 
form of speech. For the components of the compound arthavastu in 
the Kdrikd, the Vrtti substitutes vastu for artha and tattva for vastu. 
Pouring forth: giving forth from her very self the divine rasa of bliss.’ 
As Bhattanáyaka puts it: 

Prompted by the thirst of these children,? 
the cow of speech 

gives forth this rasa as her milk; 

to which the experience milked by yogis 
bears no comparison. 

For without the afflatus of this rasa? what the yogis milk they milk by 
force. [How different from the yogis are those who are found worthy to 
receive the gifts of a goddess will appear from the following lines:] 

The mountains made Him&laya their calf; 

then with Meru playing the skillful milkman 
and Prthu giving instruction to the mother, 
they caused to flow for him from Mother Earth 
her milk of mighty herbs and shining gems.‘ 

So runs the stanza [of Kalidasa] which shows the worthiness of Himalaya 
to receive the most precious things. 

Reveals a vibrant [genius]: The poet's genius is not inferred by 
the audience, but shines forth with immediacy because of his inspiration 
with rasa. As my teacher Bhattatauta has said, “This is why the 
experience of hero, poet, and audience is the same."* 

Genius is an intelligence capable of creating new things. The spe- 
cial genius here is one which is capable of composing pure and beautiful 
poetry because of the inspiration of rasa. As the sage [Bharata] bas 
said: "[They transmit] the inner mental state of the poet."" 

Thus it is: The sense is that the number of great poets is arrived 
at by counting those who reveal this special, vibrant genius 

1. ánandarasam: that rosa which is bliss. The association of the two 
words is ancient. Cf. Brahmasutrabhásya 1.1.12, where Sankara in explaining 

§ 1.7 Introduction L | 121 

änandamaya quotes from Tait. Up. 2.7 raso vai sah rosam hyeváyam labdhvá- 
nandi bhavati. 2. By "these children” is meant men of taste. connoisseurs. 
3. tad-dvesena vind: without the afflatus, the divine inspiration, of this rasa. 
Sarasvati gives freely to the sahrdaya or rasika, as the earth gave her gifts 
freely to the calf Himálaya. Yogis, on the other hand. must withdraw their 
mind and senses from all obj ts in order to force their way to their goal. For 
passages from the ABh bearing on the comparison of aesthetic and mystic 
bliss see Aesthetic Rapture Vol. Il, p. 45 (note 263). 4. Kum. 12. The 
stanza furnishes the mythic explanation, drawn from the Visnu Purdna, of 
how the Himalaya came to be the possessor of jewels and of the herbs to cure 
all diseases. 5. An extraordinary statement for Abhinava to quote with 
approval in view of the careful distinction which he makes elsewhere between 
the emotions of the hero and the aesthetic relish of the poet and audience. Of 
course the only point of the quotation here is to show a similarity of experience 
between poet and audience because of Sarasvati's gift of rasa. [f only we could 
take náyakasya as an objective genitive, all would be well; the poet's and the 
audience's experience of the hero is the same. 6. This definition is close to 
Bhattatauta's: “ao intelligence which keeps blooming with ever new things 
is called genius” (prajñā navanavonmesasálini pratibhd matá); see Séntarasa, 
p.18. 7. BANS 7.2; it forms part of a verse explaining the etymology of 
bhdva The bhávas (the word is used in a very broad sense to include the 
vibhéves, anubhdvas, and vyabhicdribhdves as well as the bhàvas proper, all 
of which are to be described in BANS 7) are so called because they transmit 
(bhávayan) to the audience the inner state (bháva) of the poet. Here Abhinava 
takes the words to substantiate the statement of 1.6 that the words of great 
poets reveal their genius. The phrase is quoted again 3.41-42a L. 

A Here is another proof of the existence of a suggested meaning: 

L Here is: It is not only, as indicated in 1.4 K when it spoke of 
the suggested as something different, that literal and suggested sense 
Tay differ in nature and in the person to whom they are directed 
There is proof that the suggested sense differs from the literal in that 
that it is understood through a wholly different set of causes. 


K It is not understood by the mere knowledge of grammar and 
dictionaries.! It is understood only by those who know tbe true nature 
of poetic meaning. 

1. sabddrthasdsana: We take Sabdasdsana to equal sabddnusdsana, as 
in the beginning of the Mahdbhdsya; arthasásana will then be the teaching 
imparted by dictionaries. 

A Because this [suggested] sense is understood only by those who 
know the nature of poetic meaning. If this meaning were denotative, 
one would get to it by a knowledge of literal, denotative meanings and 
the words that convey them. But this meaning is beyond the range 
of those who bave taken pains only on the definitions of words and 
who have paid no attention to the study of poetic meaning, just as the 
character of the notes (svaras) and srutis, etc., is beyond the range of 
those who know tbe definitions of music but are not good singers.' 

1. Text and meaning are doubtful. The reading iva pragitánám (“just 
as of good singers”) is found in the KM ed. and in the three MSS on 
which it is based. All other MSS seem to read ivápragitànám ("just as ... of 
not good singers"). The Locana says nothing of the negative, although the 
commentaries on the Locana infer or supply its presence: Kuppusvami Sastri 
ad Kaumudi pp. 173-174, BP p. 95, and, most ingenious of all, Pathak (Vidyà- 
bhavana ed., foot of p. 95), who supposes that the Locana's two explanations 
are furnished the one to fit apragito, the other to fit pragito. Despite all 
this ingenuity it is unparalleled for Abhinava to gloss the second half of a 
negative compound without mentioning the negative. Jacobi, who had only 
the KM ed. to work with, translated the passage: "wie solchen, welche nur 
die Theorie der Musik kennen, die individuelle ganze und Zwischentóne guter 
Sänger unkennbar sind.” This destroys the parallelism of the sentence. We 
have chosen the reading with the negative and a translation essentially the 
same as Krishnamoorthy's. If one accepts this reading of the text, one will 
explain the passage as follows. Most would-be poets know only the literal 
meanings of words; only a few, like Kalidasa. are capable of using words in 
their full suggestive meanings. In this respect they are like singers. Those 

§1.7L] 123 

who know merely the definitions of the books on music, if they are not good 
singers, are incapable of producing the notes and srutis of the various melodies 
(grdmas; cf. 1.7 L below). I owe this explanation to Dr. Gary Tubb. 

L Is understood: it is not [to be left as] not understood, by 
which (one might suppose] it does not exist. That is his intention. 

Who have paid no attention to the study, that is, to a repeated 
reflection on matters other than the literal, of the meaning which is the 
nature of poetry. 

The notes: of these there are seven, beginning with the tonic 
(sadja). A $ruti is a change [of pitch] of such size as to make any 
alteration of a note. There are twenty-two of these srutis, formed of 
the notes, note-intervals, or both.' By the word “etc.” he would in- 
clude the [grámas or melody types] such as the játyamsaka, grámarága, 
bhásá, vibhàsà, antarabhásà, desi, and márga.? 

Good singers (pragitàh): Those of whom the singing is good are 
called pragitáh. Or, those who have begun to sing are pragitàh, the 
past passive participial suffix being used in the sense of beginning an 
action (P&n. 3.4.71). By the beginning is here indicated everything up 
to the final result. 

1. (Note furnished by Dr. Gary Tubb.| What Abhinava is referring to 
is the classical system of twenty-two srutis, described by Nijenhuis, p. 10, 
as “micro-intervals used to describe interval arrangements" and by Capwell, 
p. 780, as “modally diagnostic microtones.” It is by the srutis, the minimum 
units of measure of pitch interval, that the basic notes (svara) of a melody 
type are defined. If any of the notes associated with the melody type is given 
a new assignment differing in pitch by a single Sruts, the type will fall under a 
different designation. 2. [Dr. Tubb] Játyarnáaka is a term found frequently 
in BANS (chapters 28-33 in the GOS edition), while the other terms are 
discussed in Matanga's Brhaddesi, (rom which Abhinava most likely took 
them. Part Three of the Brhaddesi discusses the grámarágas. Bhásó, vibhàsá, 
and antarabhásá (so-called because these melodies were used in dialect songs) 
are discussed in Part Four. It is of interest to note that the Brhaddesi (p. 105 
in Sambasiva Sastri's edition, near the beginning of Part Four) uses the term 
pragita exactly as Ananda and Abhinava have used it: prakáam na ca laksyate 
yatnahinais tu gdyakaih / pragitds tu prosiddhyanti susvaránám vifegatah. 
“(The melody type called bhds] is not manifested clearly by singers who 
have not practised hard. Good singers, however, succeed, especially those of 
perfect pitch." 

[ § 1.8 Introduction A 

A Having thus proven the existence of a suggested meaning 
which differs from the direct meaning, he goes on to show the greater 
importance of the suggested: 

L Thus: that is, he has proven it by the difference in the na- 
ture and person addressed of the suggested and by the fact that it is 
apprehended through a different complex of causes. 

K This meaning and whatever particular word has the capability 
of conveying it are the meaning and the word which should be carefully 
scrutinized (or recognized, pratyabhijrieyau) by a great poet.' 

1. Jacobi has taken mahákaveh as possessive genitive. One will then 
supply some such word as sahrdayath with pratyabhijteyau. But the Vrtti by 
rearranging the word order seems to take mahdkaveh as subjective genitive 
(Pag. 2.3.71) and Abhinava takes it definitely in that sense. 

A The suggested meaning and the particular word that bas the 
capability of conveying it, not just any word: this word and meaning 
should be scrutinized (or recognized) by a great poet. It is by the 
proper use of the suggested sense and the word that suggests it that a 
great poet deserves his name. not by mere structuring of the denoting 
word and the denoting meaning. 

L Should be scrutinized (or recognized): the gerundivesu x 
is here used in the sense of “should,” ? for the fact that everyone strives 

§18L] 125 

in this way |viz., for the suggested word and meaning] in itself furnishes 
proof that they are well known to be more important (than the denota- 
tive word and sense]. And by the suffix's use in this mandatory sense he 
indicates that this [seeking out of the suggestive word and sense] forms 
part of [a poet's] education. By using the word pratyabhijneyau he 
would indicate that although poetry may flash forth (parisphurati) of 
its own accord in the way described [by Bhamahba 1.5], “Poetry comes 
to the man of genius, and at that only sometimes,” still, it increases in 
a thousand ways if that man will keep considering his poem carefully, 
thinking, “this should be like this," i.e., “I should say such and such, 
not such and such.” in this way always seeking the suggestive word 
and sense.? The matter was put as follows by my teacher's teacher, the 
renowned Utpala: 

As some lover brought by many prayers 

tora lady's side, only to find 

that she does not recognize him when he is come 
and so all hope of making love to her is gone; 
just so is God, although he be 

our very soul, misprised within us 

and cannot share with us his glory. 

Therefore I have written this book 

called “Recognition.”? 

One sees from this that pratyabhijnd (recognition, scrutiny) is a care- 
ful inspection of and continuous reflection upon an object although 
that object is already [in some sense] known. This is what is meant by 
protyabhijnd and not the mere recognition that consists in noting that 
"this is the same thing I saw before." 

A great poet: One hopes that one also may be a great poet.* 

By his speaking thus of the importance of the suggestive word and 
the suggested meaning he has implied an importance also of the relation 
between the suggestor and the suggested. Thus he has shown that the 
three [senses of dhvani] will fit: that which suggests, that which is 
suggested, and the operation of suggesting.* 

1. More literally, “in the sense of the worthiness or desert of the subject" 
Pàn. 3.3.169. The suggested word and its meaning deserve to be scrutinized 
2. The contrast of effort and genius (inspiration, imagination) io the making 
of poetry is noticed in Ariguttara Nikdya 4.230 and Dandin, KA 1.103. See 
also J. L. Masson, "Imagination vs. Effort,” JIP 1. — 3. The book is Utpala's 
Isvaropratyabhijià, "Recognition of God,” where this stanza occurs at 4.1.17 

126 [$1.8 L 

(p. 313 of the Bháskari, Vol. 2). It is also found in the /svarapratyabhijriávivrti- 
vimarsini 4.4.2 (Vol. 3, p. 403). The lady, presumably, has sent a go-between 
to the potential lover, whose reputation she had heard of. He then steals to her 
garden some night only to find that she mistakes him for a stranger and wi!) 
not come forth or allow him to enter. This simile of God's lying unrecognized 
within us suggests to Abhinava another simile. Just as God, if unrecognized 
within us, cannot impart to us his glory, just so our poetic genius, if we do not 
recognize or scrutinize it—and he goes on to give a very special sense to this 
term—cannot produce the great poetry of which it is capable. 4. Thus the 
study of the suggestive word and its meaning will form part of the education 
of every poet. — S. These are the first three senses that we listed in 1.1 K, 
note l. 

A Now, although a correct choice of suggested meaning and sug- 
gestive word is more important, it is right that poets should first turn 
their attention to the correct choice of denoted meaning and denoted 

L The author anticipates that an inference might be drawn of 
the greater importance of the denotative word, meaning, and operation 
from the fact of their being taken up first for consideration.’ So he 
shows, with now, etc., that as a reason (or middle term) this [fact 
of being taken up first] is contradictory to what is here sought to be 
proved, viz., greater importance, for he takes the view that it is the 
means that are first taken up, [not the all important goal]. 

1. The inference would appear as: pradhdnd vdcyavdcakatadbhduah pro- 
thamopadiyamdnatudt. But here prathamopádiyamánatva is a viruddho hetuh 
because pradhdnavastusu prathamopddiyamdnatudbhava eva. 

§ 1.10 K } 

K Just asa man who wishes to see will take pains with the flame 
of the lamp as a means thereto, just so will a man who cares for this 
[suggested meaning] take pains [first] with the denoted sense. 

A For just as a man, although the object of his wish is to see, 
ill take pains with the flame of the lamp as a means thereto, for it is 
impossible to see without the flame of the lamp, just so will a man who 
cares for the suggested meaning take pains with the denoted sense. 
So far the author has described the communicating poet’s engage- 
ment with the suggested meaning. In order to describe the engagement 
of the recipient audience he goes on to say: 

L To see: seeing. The reference is to seeing such things as the 
lotus-like face of one's beloved, and for that the flame of the lamp is a 

K Just as the sentence-meaning is apprehended through the 
meaning of the words, just so is the apprehending of this matter pre- 
ceded by the denoted sense.’ 

128 [$110 K 

1. This analogy is later qualified by Ananda (3.33f A). It is intended 
merely to show that the denoted sense is a means, and occurs at a time pre- 
vious, to the suggested sense. In other respects the relation of word meaning 
to sentence meaning differs from the relation of denoted sense to suggested 

A For just as the sentence meaning is understood through the 
meaning of the words, just so is the understanding of the suggested 
meaning preceded by an understanding of the denoted sense. 

L [Comment on the Karika.] 

The word pratipat (apprehending) contains the null-sufx kvip used 
to form an action noun (Värt. 9 on Pan. 3.3.108). Of this matter: 
viz., of the essential, that is, the suggested meaning. 

This verse shows! that the sequence [of meanings, as first denoted and 
then suggested.) is clearly noticed only by those who are not sensitive 
to poetry, just as the sequence word meaning, sentence meaning, is 
noticed only by one who is not knowledgeable in the use of words. On 
the other hand, to one whose sensitivity is at a maximum, just as to 
one who is really skilled in the use of words, the sequence, although 
it exists in fact, is not noticed any more than one is aware of one's 
memory of the concomitance in an inference that has been frequently 

1. The words anena $lokena construe with iti dorsitorn at the end of 
the comment. 2. When we see smoke, we infer fire without being aware of 
remembering the rule “wherever there is smoke there is fire.” 

A Now the author shows that the greater importance of the sug- 
gested meaning is not impugned by the fact that it is apprehended after 
the apprehension of the denoted meaning. 


L Is not impugned: Since persons [of training or sensitivity] 
hasten with eagerness toward the end [viz., the sentence meaning or 
suggested meaning| because of its importance, and do not pause with 
pleasure along the way, they fail to notice a succession of meanings 
altbough it actually exists. This failure is thus a proof of the importance 
lof the final meaning]. 

K Just as the meaning of an individual word. by force of its 
capability, acts toward conveying the sentence meaning, but is no longer 
distinguished after its activity is completed 

A Just as the meaning of an individual word, by force of its 
capability, acts toward revealing the sentence meaning, but is no longer 
distinguished apart [from the sentence meaning| after its activity is 
completed ... 

, L Capability: The capability of a word is its dkanksd (the 
‘expectancy’ of its meaning's being completed by other words in the 
sentence), yogyatd (its compatability with those other words), and san- 
nidhi (its contiguity to those other words). Distinguished (vibhàv- 
yate): The prefix of the word denotes separation; the sense is "is not 
noticed as being separate." This states that the succession [of sentence 
meaning to word meaning], although it exists, is not noticed. In con- 
tradiction with this statement is what the grammarians say, speaking 
according to the theory of sphota, namely that the succession does not 

1. The view of Bhartrhari is that the sentence as a semantic symbol 
(sphota) has no parts; it is only the sentence which ve hear that has parts. 
See Vakyapadiya 1.73 vàkyát padánám atyantam praviveko na kascana. 

(§ 1.12 K 

K just so does the suggested sense flash forth in an instant in 
the minds of intelligent auditors who are averse to the literal sense and 
in quest of the real meaning. 

L Who are averse to the literal sense: whose selves or hearts find 
in the literal sense no satisfaction that could arise from dwelling on it. 
This brings out the force of the word sacetosám ("intelligent," but 
literally, “possessing a mind or heart"). One might suppose that this 
[rapid appearance of the suggestion| lies in the brilliance of sensitive 
auditors and [re&ects| no special excellence of the poem. So he says: 
flashes forth. Because of this [rapid scintillation] the literal sense does 
not appear as something separate, but this does not mean that it does 
not appear at all. So there is no contradiction of this passage with that 
in Chapter Three (3.33f A] where he will state that our apprehension of 
the literal does not disappear when we apprehend the suggested sense, 
any more than the lamp disappears when [by its light] we perceive the 

A Having thus shown the existence (sadbhàva) of the suggested 
meaning as distinct from the literal, he puts it to use in the matter at 

L Existence: the word sadbhdva has [also] the meanings of 
excellence and predominance, both of which he wishes to convey. Puts 
to use: gives one to understand its use.' In the matter at issue: 
viz., in the de&nition [of dhvani]. 

1. What A and L mean is that the use of proving the existence of dhvani 
is that only then can one proceed to define it. 

$143 L| 

K The type of poetry which the wise call dhvani! is that in 
which sense or word, subordinating their own meaning, suggest that 
[suggested| meaning. 

1. See 1.1 K, note 1. Dhvani is here used in the fifth of the senses there 
listed, viz., of a type of poetry. But this sense is not sharply distinguished 
from the fourth sense, viz., the suggested meaning. The Vrtti on this verse 
slides very easily from the one sense to the other. 

A The type of poetry which the wise call dhvani is that in which 
sense, viz., a particular literal sense, or word, viz., a particular deno- 
tative word, suggests that meaning. 

L [Abhinava here comments only on the Kárikà. The Vrtti is 
so similar as to need no separate comment.) That meaning: here he 
puts [the proven existence of the suggested meaning] to use.' (The com- 
pound upasarjanikrtasvdrthau (“subordinating their own meaning") is 
to be analysed as follows:?] sva = self; svárthau = self and meaning; 
upesarjanikrtesvarthau = subordinating itself and its meaning. Here 
we must pair off the terms in order, viz., the meaning subordinates 
itself and the word subordinates its meaning. That meaning: viz., 
the meaning that he has already referred to in speaking of "Sarasvati, 
pouring forth this sweet matter” (1.6 K). 

Suggest: i.e., indicate. Here he uses the dual form, [rather than the 
singular], for while it is true that in the avitakgitavácya type of dÀvani 
a word is the suggestor,? the cooperation of its [literal] meaning cannot 
be wholly dispensed with; otherwise, a word of whose meaning we are 
ignorant might be a suggestor. And in the vivaksitányaparavácya type* 
there must be the cooperation of words, because the meaning [which is 
there predominant] could not be suggested if the denoted sense were not 
furnished by a word or words. Accordingly, the operation of suggestion 
always belongs to both word and meaning. So when Bhattan&yaka 

[5133 £ 

criticizes the dual here, he is overlooking the obvious facts.* But [we 
must remember that] in stating the alternative “word or sense” our 
author means [to include the notion of] predominance.* 

The type of poetry: One may analyse the compound (kávyavisega) 
as a karmadháraya or as a genitive tatpurusa." By using the word "po- 
etry” he shows that that soul which has been characterized as dhvani 
falls in the area of words and meanings embellished by the poetic qual- 
ities and figures of speech, so that there is no occasion for applying the 
word dhvani to the “material inference” (arthápatti) [of the Mimamsaj.° 

As for what bas been said that “then the apprebension of beauty 
(cárutvapratiti) will be the soul of poetry,” we are quite willing to 
accept it. The only dispute is about a name (viz., whether to call the 
soul of poetry cárutvapratiti or dhvani]. But when it is said that "If the 
soul of poetry is (nothing more than) the apprehension of beauty, the 
soul of poetry could arise from any means of cognition, such as visual 
perception and the like,” we reply that the statement is nonsense. The 
context is an effort to define the soul of poetry, which is an entity 
consisting of words and meanings. How would there be any occasion 
[for bringing in visual perception]? 

[The five meanings of dhvani] 

That in which: we may consider the reference to be to the sense, 
or the word, or to the operation [of word and sense]. And the sense 
may be either the literal sense, for it suggests (dAvanati), as does the 
word, or the suggested sense, for it is suggested (dhvanyate), [while| the 
operation is an alternative (because it is| the suggesting (dhvanana) of 
word and sense. But the Kérikd would convey by the word dhvani 
primarily the sum total of these elements in the form of poetry.? 

1. The pronoun "that" (in "that meaning") can refer to the suggested 
meaning because he has already established the existence of such a meaning. 
2. The reason for Abhinava's odd analysis of the compound is that if we take 
it naturally, the element artha is meaningless; upasarjanikrtasvau would have 
been sufficient. — 3. In that type of dhvani “where the literal meaning is un~ 
intended," what is important is the word. For example, in gangáyám ghosah, 
“a village on the bank of the Ganges." if we were to substitute the name of 
another river, the suggestion of holiness, etc., would disappear. 4. Many 
examples of this type, "where the literal meaning is intended but is subordi- 
nated to a second meaning," will be given in what follows. It includes all cases 
of rasadhvani. This whole paragraph of the Locana is quoted and subjected 

$1.13a A] 133 

to criticism by Mahimabhatta in his Vyaktiviveka, p.91. 5. Gajanimilikó: 
literally, blinking at an elephant. The expression is used in Raj. Tar. 6.73, and 
frequently in modern Sanskrit (e.g., Abhyankar's commentary on SDS, pp. 52, 
86). BP gives an incorrect explanation, confusing the term with gajasnáno. 
6. By adding this sentence Abhinava justifies the use of the dual inflection. 
If the author had intended to speak of either word or sense acting as sole 
agent, he should have used the singular. But his intention was to speak of 
word and sense cooperating to suggest, with either word being predominant 
or sense being predominant. His intention therefore required the use of the 
dual. 7. BP explains the difference of meaning effected by these analyses 
as follows. In kávyam ca tad visesarn cásou it takes vifesa in its normal sense 
(vi + Sis +ghori, Pin. 3.3.19), where the suffix gives the root passive sense: it 
is poetry and it is that which is distinguished [from sdstra, etc.]. One should 
then translate from the Kürikà: “Poetry as a distinctive type of literature is 
called dhvani by the wise.” In the tatpuruso analysis BP seems to take visesa. 
as formed with the upapada suffix an (Pan. 3.2.1), for it takes the verbal root 
in an active sense. One should then translate: “The distinguisher of poetry 
(namely, the soul, suggestion) is called dhvani by the wise." Neither of these 
analyses would permit a sentence lacking in dhvani to be called poetry. On 
the other hand, I prefer to take the compound as tatpurusa but to give visega 
its normal sense, as formed by ghari: "That type, as distinguished from other 
types, of poetry is called dhvani ...." This is because | suppose that Ananda 
would have considered poetry even not of the best to be still poetry. — 8. See 
1.4g L, note 21. 9. It may be found useful to memorize BP's convenient 
summary of the five meanings: tathd ca tathdvidhah $abda-vácya-vyarigya- 
vyerijana-samuddydtmakah kávyaviseso dhvanir iti kathitah. For an English 
rendering, see 1.1 K, note 1. 

A This shows that the habitat of dhvant is different from that of 
causes of beauty in the denotative sense and word, such as simile and 
alliteration.’ " 

The objection that "there is no such thing as dhvani because what 
falls outside our well known system would no longer be poetry" (1.1b A) 
is injust, for it is only to the makers of definitions [of poetry) that 
dhvani is not well known. When poetry itself is examined, one finds 
that dhvani is the poetic essence that delights the heart of the sensitive 

(§1.13a A 

ience. Whatever differs from it is mere citra (display),? as we shall 
in what follows. 

1. "Such es simile” stands for all the figures of meaning; “such as allit- 
eration” for all the figures of sound. 2. Citra: lit., a bright picture, but in 
the technical sense, a display of mastery in poetic figures and meters. Citra 
verses are discussed in 3.41-42. 

L Is different: dhvan: cannot be included in them because 
the life of poetic qualities and figures lies in the nature of word and 
literal sense. whereas the essence of dAvani lies in the nature of the 
suggestor and the suggested, which are different from that. The word 
habitat (visaya) shows that it exists in no other place. In this way 
the deprecation is silenced that took the form of “What is this thing 
called dhvani that it should differ from these?" (1.1a A). 

The makers of definitions: (In inferring tbe non-existence of 
dhvani] its being not well known to the makers of definitions is a reason 
contradictory to what is sought to be proved (viruddho hetuh).' In fact 
this is all the more reason for trying to frame a definition. On the other 
hand, its not being well known in the poetry itself would be a “falsely 
assigned reason” (asiddho hetuh).? And the suggestion that it might be 
something like dancing and singing (cf. 1.1b L) [is nonsense because 
tbat| would have nothing to do with poetry. 

Citra: it is called "display" because its use of meters and other 
[embellisbments] causes admiration, while it lacks the exudation of that 
nectar of true beauty that is sought by the sensitive audience. Or it 
may be called citra in that word's sense of “picture,” asani itation of 
poetry, or because it is a mere written design,’ or because it is simply 
one of the arts.* 

In what follows: — ., in Chapter Three [verse 41, which Abhinava 
here quotes]. 

1. Cf. 19 Intro. L and note 1. 2. Cf. 12 L, note 2. 3. This would 
apply to the topiary verses, e.g., verses composed in the shape of a sword or 
lotus, with which the lndians, like the ancient Greeks, amused themselves. 
4. The last suggestion is not quite clear to me. Perhaps Abhinava feels that 
any of the sixty-four arts may be called citra in the sense of bright, interesting, 

§1.13¢ A} 

A The objection is also wrong which said that “Dhvani cannot 
be something entirely new because, being something that falls within 
the area of beauty, it must be included in the types of figures of speech, 
etc., which have been recorded” (cf. 1.1¢ A). For how can dhvani, 
which is found to occur always in dependence on suggestive word and 
suggested meaning, be included in a system that depends on only the 
literal word and meaning? Furthermore, the causes of beauty in the 
literal word and meaning are subordinate to it whereas it is principal 
to them, as will be shown in what follows. The following couplet will 
give support (parikara) [to our position]: “As dhvani depends on the 
relation of the suggestor and the suggested, how can it be included in 
the causes of beauty belonging to the denoter and the denoted?” 

L A supportive couplet is a verse (sloka) intended to supple 
ment the Karikds in order to fortify (parikara) their argument. 

A Now an opponent of dhvani may allow that [a figure of speech] 
where a suggested meaning is not clearly apprehended falls outside the 
habitat of dhvani. But where it is perceived, as it is in such figures as 
samásokti, dksepa, the type of visegokti where the reason is not given, 
paryáyokta, apahnuti, dipaka, sarikara, and the like! he will say that 
dhvani must be included. It is in order to refute such a suggestion 
that it was specified [by the Karika] that “word and sense subordinate 
themselves.” It is where a sense by subordinating itself, or a word 
by subordinating its literal sense, reveals another [suggested] meaning. 
that we have dhvani. That is to say, since dhvani is found only where 
the suggested meaning is predominant, which it is not in samásokti and 
the like, how can it be included in them? 

136 [$113c A 

1. For these figures of speech see the Index. 

L Where: viz., in a figure of speech. Clearly: he means both 
with beauty and clarity. Specified: The Vrtti uses the past tense 
because it has already dealt with the word “suggest” [in 1.13b].! Sub- 
ordinating itself: the Vrtti explains the sense of sva ("own") of the 
Kürikà by átman ("itself"). Which it is not: this predominance of 
the suggested sense is not. That is to say, its predominance does not 
appear as we are understanding the verse, because we spend our enjoy- 
ment in an unbroken relishing [of both literal and suggested meanings] 
on the principle of [what was said in 1.12 K]: “in the minds of those 
io whom the real [meaning] appears."? But [in a second stage], when 
a man of discri ination seeks the enlivening element, since it is the 
suggested meaning that gives life to the literal, be will decide that he 
is presented with a figure of speech, because the suggested is helping 
out that (literal sense]. And so he will say that be received bis delight 
from the literal meaning as helped out by that [suggestion]. Although 
in a final [third] stage, there is indeed rasadhvani, nevertheless this 
suggested meaning in the second stage does not point to rasa; for its 
own part it simply hastens to ornament the literal meaning. And so 
the Vrtti speaks of the suggested meaning as being subordinated.? 

1. The usual word for the Vrtti to use in such references is dha, which 
the Paninians take to be a present tense (PAn. 3.4.84). Abhinava feels that he 
must give a reason why abhiAitam, a past tense, is here substituted. His reason 
is that the Vrttitdra has already explained the word vyariktah, which occurs 
in the second half of the Káriká He now turns back to explain upasarjani- 
krtasvárthau, which occurred in the first balf of the verse. So in referring 
back to it he uses a past tense. 2. Abhinava is quoting from memory. The 
Káriká actually used the word tattvárthadarsinyám, not tattvávabhásinyám. 
3. The complication of Abhinava's thought in this passage is caused, it seems 
to me, by the fact that he is trying to reconcile matters that are irreconcilable. 
There is no doubt that the figure samásokti involves suggestion, as the exam- 
ple about to be quoted io 1.13d A will show. The old poeticians were aware 
of this fact and defined this sort of suggestion within their system of figures of 
speech. Ánanda, seeking to make suggestion into a wholly new semantic cate- 
gory but unwilling to defy the definitions of the past, abandoned the group of 
suggestions that had already been categorized by the older writers under the 
figures of speech and tried to find an area where suggestion could form its own 
independent species. To do this he invented the distinction between predomi- 
nant and subordinate suggestion. Where it was predominant it could form its 

$1.13d A] 137 

own species. All the cases of suggestion contained in the old, defined, figures 
of speech could be left out as instances of subordinate suggestion. See Intro- 
duction. pp. 22-24. There would still be room for an important new species: 
predominant suggestion. The only trouble with the innovation is that many 
of the old type, many instances of samdsokti, dksepa, etc., move us as deeply, 
both emotionally and aesthetically, as any new examples which he can adduce. 
How can one explain this fact, if true dhvani exists only where it is predomi- 
nant? So Abhinava, like Ptolomy inventing new pericycles to rectify a system 
that is basically wrong, invents still another distinction: the three stages in 
our response to these long-recognized instances of dhvani. [n the first stage 
we respond automatically without distinguishing what is predominant and 
what is subordinate. [n the second stage we realize upon reflection that the 
suggestion is belping out a defined figure of speech. Finally in a third stage we 
relish the rasadhvani despite our intellectual decision that it does not belong 
in the same category as that which is produced by a predominant suggestion. 

A Let us begin with the igure samdsokti (compound statement): 

The reddening moon has so seized the face of aight 
with her trembling stars, 

that all her cloak of darkness in the east 

falls thus unnoticed by her in confusion.’ 

In this. as in similar verses, the literal meaning, although it is accom- 
panied by a suggestion, is apprehended as the more important, for the 
main purport of the sentence concerns the moon and the night,? on 
which bave been superinposed the behaviors of a lover and his lady. 

1. In pàdo d I bave taken the reading of the oldest version, mohdd, in 
place of rdgdd, which avoids the awkward repetition of rága. The words of the 
Stanza have double meanings, which are immediately brought to our attention 
by the obvious pun in rdga: redness, or love, and by the masculine gender of 
the word for moon in contrast with the feminine gender of the word for night. 
Proceeding from these hints, the mind soon finds a suggested meaning for 

138 [$ 113d A 

each word (see Abhinava's comment below). The suggested meanings, when 
combined, furnish a sense as follows: “The lover, with aroused passion, kisses 
the face of his beloved, whose eyes (tdrakd = ‘pupil’ as well as ‘star’) tremble, 
so that she drops her robe entirely before him (purah = ‘in front of’ as well as 
‘in the east’) without noticing what she has done in her confusion.” The tradi- 
tional interpretation of such verses insists on keeping the two versions distinct, 
allowing one message to suggest the other, but refusing to mix the images. 
Masson writes: “When I attempted to mix the two, speaking of the moon 
kissing the face of the night, which is after all rather poetic, Pandit Srinivasa 
Shastri of the Deccan College looked astonished: “Katham § 7 cetanavastu- 
van nigém cumben nanu? Tathá násti, sarvathé asambhavam." The verse is 
quoted in most of the anthologies. usually without variant, and invariably at- 
tributed to the poet Panini, for whom see Peterson, JRAS 1891, pp. 313-316. 
Thus it is in Sárrig. 3634, SuktiM. 72.5, Sodukti. 412 (where we find the bet- 
ter reading mohàd in d). The first word and last quarter of of the stanza are 
preserved among the fragments edited by-Gnoli of Udbhata's Commentary on 
the Kavyélarikéra of Bhàmaho, fragment 37, pp. 34-35, where the reading of 
d appears as puro ‘pi mohdd galitam na raksitam. The word mohád has been 
changed in the later versions, presumably because of the difficulty of finding a 
puninit. 2. It is context which tells us which of the two possible meanings 
is the main purport (vdkydrtha). Even without other verses which may once 
have accompanied it, we may be sure that the context of the present stanza 
is a description of moonrise, that being one of the favorite topoi of Sanskrit 
poetry. The lover and his beloved are a secondary suggestion. 

L Samásokti: 

Where in a statement a second meaning is understood because of epithets 
common [to both meanings]: the wise call that samásokti, because the mean- 
ing is composite. (Bhámaha 2.79) 

In the four quarters of this verse the author has given successively the 
basic characterization of samdsokti, the reason for it, its name, and the 
explanation of the name. 

Reddening (literally, possessing redness): [in the case of the moon 
this means] assuming the red color of twilight; and [in the case of a lover 
it means] assuming the feelings of love. With trembling tārakā: in 
which the lights of heaven are trembling and in which a portion of the 
eye is trembling. So: i.e., suddenly (of the moon] and with a rush of 
love [of the lover]. Seized: illumined (of the moon with respect to 
the night] and seized in order to kiss [of the lover with respect to his 
beloved]. The face of night: the beginning (of the night] and the lotus 

$113d L] 139 

face [of the woman]. Thus: suddenly (of the night] and with a rush of 
love (of the woman]. Cloak of darkness: Of the night timirdmsukam 
means timiram (darkness) and süksmám£avah (feeble rays),' i.e., her 
mass of darkness spangled with a few rays [of the stars]; of the beloved 
timirámsuka means the dark veil appropriate to a newly wed bride 
who is shy. [In place of mohdd (in confusion) Abhinava reads rágád, 
which he explains as follows:] Rágád (in the case of the night] means 
"from redness," that is, immediately following the redness produced by 
twilight; and [in the case of the woman] it means "from love," that 
is, because of love. In the east (in connection with night) and (in 
connection with the woman| in front of her. Fallen: vanished in the 
one case and fallen in the other. By her: by the night as instrument 
(Pan. 2.3.18) the mass of darkness is spangled (with starlight], or we 
may take the pronoun to be an instrumental of identification (Pan. 
2.3.21).4 Unnoticed: it was not realized that this was the beginning 
of night, for people recognize the face, or beginning, of night when they 
see a mass of darkness spangled with starlight, but not when a clear 
light [as of the moon] is present. In the case of the woman, on the 
other hand, "by her” will be an instrumental of agent (“unnoticed by 
her”). In the case of the night the word ap: [in puro 'pi = even in front 
of her] must be transposed to follow upalaksitam |i.e., “was not even 
noticed by people”). And here [in the case of the woman] the veil falls, 
or drops, in front, as the lover coming from behind begins to kiss her. 
Or, we may take the syntax to be that the lover [standing] in front 
seizes her face. 

So although we understand a suggested sense in this stanza, it is not 
predominant. That is to say, the [superimposed] behavior of lover and 
beloved in ornamenting the moon and night, which thus take the form 
of vibhávas of the erotic flavor, acts as an ornament or figure of speech. 
But then from the literal sense [of moon and night] which has been 
turned into a vibhdva, there issues forth a steam of rosa. 

Here someone has said, “The word ‘by her,’ viz., by the night, ex- 
presses agency, and since agency is impossible on the part of an insen- 
tient being, the behavior of lovers which we infer is given by the deno- 
tative force of ‘night’ and ‘moon’ and not by a suggestion. That is why 
the stanza forms a compouind utterance (samásokti).^* This explicator 
has ignored the clear sense of the text with which we are concerned, 
(for the Vrtti has clearly said) "accompanied by a suggestion." At that 
rate the figure would be ekadesavivarti rüpakam (a partial metaphor) 
like the couplet: 

The pond kings were fanned 
by autumn with her wild geese 
(Udbhata, 1.*12 Indurája = 1.24 Vivrti)" 

It would not be a samásokti because it would not contain epithets 
that apply [in each instance] equally to both [the denoted and the 
suggested sense]. Furthermore, the denotative function is ruled out 
[in Bhámaba's definition of samásokti just quoted] by the phrase “is 
understood” (rather than “is stated”]. But let us not run on at too 
great length on a subordinate matter 

[Here it might be thought that Ananda should have written samé- 
ropitandyaka-vyavahdrayoh rather than samdropitandyikd-ndyakayoh® 
However, there is no need for an. ekasesa compound [viz., náyakayor| 
if we explain the meaning to be that the behavior of a lady toward her 
lover is superimposed on night and the behavior of a lover toward his 
lady is superimposed on the moon. 

1. He is taking the compound as a dvandva rather thao as a karma- 
dhåraya In om$uko he is taking the suffix as the diminutive -ka (Pan. 5.3.85) 
appended to the stem améu-, "ray" 2. Abhinava's interpretation of the 
second meaning of timirdmsukam seems to me as faulty as his explanation of 
the first. [n his interpretation the word "all" (samasta) will have no force. 
Nor does the falling of a mere veil indicate much passion (rága). What is 
in fact suggested is the falling of her entire garment. [t is a commonplace 
of Classica! Sanskrit poetry that the knot of the beloved's garment opens 
of itself and her dress falls as her lover embraces her; see Meghadüta 73, 
KumSam. 8.4, Sisupála 10.45 and 50, Kiróta 9.47-48. 3. Abhinava takes 
the ablative inflection of rága as denoting cause only in the suggested meaning, 
which concerns the woman. To explain the ablative in the first meaning 
concerning night, he supplies the word anantaram, "immediately afer.” This 
word governs an ablative on the analogy of Pan. 2.3.29. — 4. This is the usage 
found in such phrases as api bhaván kamandalund chátram adráksit. “Did you 
see the pupil with a water pitcher?” where “with a water pitcher” serves 
to identify the pupil who is intended. Here the mass of darkness spangled 
with stars that was “with the night," i.e., that characterized or identified the 
night, wes not noticed. Both interpretations are wildly improbable. Tayà 
must be taken as instrumental of agent with laksitam in the case of the night 
as well as in the case of the woman. If it is objected that an insentient thing 
like night cannot “notice” anything (this difficulty doubtless led Abhinava 
to his interpretation), one may have recourse to the older reading raksitam. 

§1.13e A] 141 

5. The anonymous opponent holds that words have only one operation, abhi- 
dhá (denotation). When context forbids our taking the usua] meaning of 
the word, we choose a second meaning. But the opponent refuses to call this 
second meaning laksita, any more than he will call it vyarigyo. As night cannot 
be supposed to “notice” anything, we are forced to take “night” to mean a 
woman, whereupon we shall take “moon” to mean a lover. These meanings 
are denotative, just as night and moon are denotative. It is because of the 
double denotations in the verse that the figure is called samdsokti — 6. This 
is a sure refutation if the explicator was commenting on the Dhvanydloka. 
But if he was attacking it, the statement of the Vrtti would carry no weight. 
7. Chowrie-bearers fan kings and one can express ponds metaphorically as 
kings, but autumn does not perform the function of such intelligent beings as 
fan-bearers. The example is quoted again at 3.36 L. — 8. Lf we take Ananda's 
compound, as 1 have done, to mean “on which has been superimposed the 
behaviors of a lover and his lady,” we make Ananda guilty of disregarding Pan. 
1.2.67, which states that in such cases the feminine component of the pair is 
dropped and the masculine component, as an ekasesa compound. suffices for 
both. Abhinava exonerates Ánanda by giving the compound a meaning other 
than “the behaviors of a lover and his lady." 

A In Gksepa (a bint, often in the form of a denial) also, while 
it bints at a particular suggestion, the literal sense is charming. The 
literal sense! is known to be predominant from the very fact that [the 
figure) is called aksepa. Thus, it is the hint itself, in the form of a denial 
explicitly stated with a view to expressing some particular,? that forms 
the principal body of the poem, even though it hints at some particular 
suggestion. This is because the decision whether Literal or suggested 
meaning is the more important depends on which is the more charming. 
An example is this couplet: 

The sunset is flushed with red, the day goes ever before; 
Ab, such is the way of fate that never the two shall meet.” 

Although we apprehend a suggestion in this verse, it is this [literal 
sense] that holds the greater charm. So [we should take it that] the 
literal is intended to be predominant. 

142 [§1.13e A 

1. There is a problem of reading here. The Kaumudi, and presumably all 
but the two KM MSS which it mentions as being in disagreement, reads vã- 
cyórtha for vàkyártha. The KM edition, however, chose the reading vákyártha 
from those two MSS and has been followed by all the printed editions. If one 
accepts that reading, one must understand the word as a locative and take 
the whole passage from àksepe "pi to jridyate as a single sentence: "In áksepa 
also, while it hints at a particular suggestion, the fact that the beauty of the 
literal sense is predominant is obvious from the force of the expression of the 
hint in the final or sentence meaning." This is awkward, to say the least. I 
prefer to read vàcyártha, which I take as a nominative, and to put a danda 
after cdrutvom. The sense, as I have given it in the translation. is then clear 
and to the point. 2. [n the phrase visesábhidhánecchatà one must not take 
abhidhdna too narrowly. BP glosses the word by vyarigyabhütavisesapruti- 
pádayisayá. 3. What is hinted at in this anonymous stanza is a pair of 
lovers who are prevented. by their parents or by social differences or by rea- 
sons of state, from ever uniting. Clearly Ananda considered this couplet to be 
an example of dksepo. Other Alahkarikas who quote the verse consider it to 
exemplify samdsokti or visesokti or sarikara; see KP 9.382 (p. 526). SD 10.99 
(Sanskrit p. 63 Kane), and the Locana in what follows. To explain this differ- 
ence of judgment some historical remarks will be helpful. The word àksepa is 
post-epic. Whether its earliest meaning was ^hint" or "denial" or *censure" I 
cannot say, for it bears al) these senses in the later literature. As a technical 
name of a figure of speech it is defined by all the Ála&kàrikas starting from 
Bhamaha and Dandin. Usually it is defined as a denial (pratisedha, nisedha) 
which hints at something unexpressed. The denial may be of a fact or of a 
word that has been spoken, or it may be mere reticence, the refusal to say 
something that would be painful. For examples, see Abhinava's comment be 
low and 2.27c A with note 2. Vamana, however, has a definition of àksepa 
that differs from that of all other authors. The meaning of his laconic defini- 
tion: upamdndksepas cóksepah is shown by his examples to be " Aksepa is the 
censure of. or hint of, a simile." His examples will be found quoted by Ab- 
hinava in his comment on this passage. [t is the second half of the definition 
(hint of a simile) that Ananda must have in mind here. Neither Vàmana nor 
Ananda tells us how to distinguish this second type of dksepa from samásokti, 
so we are left to speculate. Presumably samasokti was allowed to be the more 
general figure, dksepa preempting to itself only those instances where the hint 
was specifically of a simile (upamána). 

L In aksepa: [The figure has been defined thus:] 

The denial (or holding back) of an intended [statement] out of a wish to 
convey some special [suggestion] is àksepa (denial, hint, censure). which is of 
two types depending on whether the statement was about to be spoken or has 
been made. (The first half of this verse is Bhamaha 2.68)' 

§1.13e L] 

An example of the first type is this: 

If in my longing 
I should lose sight of you but for a moment 
But I say no more. 
Why say that which would pain you? 
(Bhámaha 2.69) 

Here we have àksepa in the form of withholding the death that the lady 
was about to speak of. The words "But I say no more," while they sug- 
gest the statement “I will die," are themselves the cause of the beauty 
in the verse. So [we must say that] the dksepa (the hinting reticence) 
as embellished by what it hints at is the predominant element. An 
example of [the second type, where] a statement that has been made 
[is denied or censured] is a verse of mine: 

“My dear traveler, what causes this sudden collapse?" 
“What else can I do who have such thirst, 

when the miserly road here hides its water?" 

“Your thirst is mistimed, good sir, and misplaced 
Vent your anger on it. The glory and greatness 

of the desert road are famous throughout the world.”? 

In this verse a servitor is present. His heart is torn by desire and he 
asks why he receives nothing from his lord. Some one puts him to a 
right way of thinking by this censure. Here it is the literal meaning, 
expressing distress at the lack of recompense from service with a bad 
master, being transformed by a censure in the form of a denial into the 
vibhàva indifference (nirveda) which is the basic emotion of the flavor 
of peace (sántarasa), which gives beauty to the stanza. 

Vamana, in different fashion, defines áksepa as censure of the simile. 
It amounts to saying, "When this is present, of what account are you?" 
He gives an example: 

When we have the fair clarity of her face, 

who would care for the full-orbed moon; 

or who would care for water-lily flowers 

before the vast beauty of her eyes? 

What price would you pay for the tender, lovely 
opening of a leaf, when you see her lower lip? 
Ah me! but God shows a stubborn zeal 

for tautological creation 

144 (§113e L 

In this stanza the simile, although it is suggested, merely serves the 
literal sense. The expressed censure, in its casting away (of the simile] 
with a “who would care for it?” is the source of all the charm. 

Or again, áksepa [here simply hint, not censure] of the simile may be 
brought in by the inherent capability [of the literal sense], as in this 

Lady Autumn beautifies the moon 
although his face is made imperfect by its mark, 
and bearing on her cloud a rainbow 
like a wound left by a lover's nail, 
has made the sun grow hot. 
(Vamana, ibid. vs. 2)? 

In this stanza, although a simile is binted at, namely a lover who is 
pained by jealousy, it serves only to beautify the literal meaning. 

But the stanza [which Ananda] here (quotes, viz., “The sunset is 
flushed with red,” etc.] is really an instance of samásokti [if one follows 
the normal definition, as of Bhámaba]. 

So he says: "[depends on] which is the more charming," and apropos 
of this quotes a well-known example: The sunset is flushed with 
red. Note that he has not yet finished his consideration of áksepa and 
that it is as an example of dksepa that he quotes this samdsokti couplet. 

Ab, but such is the way of fate: The (suggested) sense is that 
there is no union [of the lovers) because of some such obstacle as their 
subjection to their parents. This: i.e., the literal sense. 

Realizing that the figure of speech here is dksepa according to Vā- 
mana's definition and samásokti according to Bhamaha's, our author 
by joining the two figures has given this single example. Let it be an 
example of samásokti or of dksepa, what does it matter to us? All that 
we are trying to demonstrate is that in figures of speech the suggested 
sense is always subordinate to the literal. This is how my teacher 
(Bhattatauta] explained the intention of the present passage. 

1. This is the definition that most later authors have followed. It is 
unfortunate, that several of the best Indian scholars have adopted the term 
“paraleipsis” as a translation, for that is quite a different figure both in form 
and intent. An example from the Rhetorica ad Herennium is: "I would speak 
of your vices, had | time. But I pass over them and over the fact that you 
often left the army without leave" (4.27). Here one neither denies a statement 
made nor suppresses a statement about to be made. Rather. one passes over 
a fact (this is the literal meaning of paraietpsis) lightly in order by minimizing 

§1.13fL) 145 

it to increase the effect of one’s major charge. If one wants a Greek or Latin 
translation for Bh&maha's first type of dksepa, the Greek would be aposiopesis, 
which the author of the Rhetorica ad Herennium calls praecisio (ad Her. 4.30) 
and Cicero calls reticentia (de Or. 3.538). 2. What the verse suggests is a 
conversation between courtiers in which the first complains that his service 
is not rewarded and the second censures his complaint. Accordingly, most 
Álaükárikas would take the figure of speech here to be aprastutaprasémsd. But 
Abhinava is thinkng of Bhamahba’s second type of áksepa and he is influenced 
by the sense of censure that Vàmana, whom he is about to quote, attributes 
to the name. We have here clearly a censure of what has been said. In 
general. when we can assign a figure of speech to either of two defined types, 
we should assign it to the narrower type; see 1.13e A, note 3, ead. There is 
room for aprostutaprasámsá where an allegory suggests the matter in hand 
without a censure of what has been said. One may add that to a tenth-century 
Kashmiri the desert road was still famous as a road to wealth; compare the 
expeditions across the northern desert of King Lalitaditya recorded in Rdj. Tar. 
4.172, 277 f., and 337 ff. 3. The literal meaning describes the phenomena of 
autumn, when the moon appears at its loveliest (our “harvest moon”) despite 
the dark birthmark (kolariko, our “man in the moon") on his face. In an Indian 
autumn, which comes directly after the cooling season of the monsoon, the sun 
again grows hot. Several of the words of the stanza are puns. Prosádayanti: 
making beautiful or granting favors to; payodhara: cloud or breast; tapah: 
heat or pain. Hence the suggestion of a courtesan favoring a lover of bad 
Character and so making her noble lover jealous. 

A And as in dipaka, apahnuti,' and the like, although we appre. 
hend a simile as something suggested, we do not call these figures of 
speech by that name because the simile is not intended to be prominent, 
so the same applies here [viz., we do not call a figure of speech by the 
name of dhvani because the dhvani is not intended to be prominent]. 

1. These figures of speech are defined below in L and the notes thereon 

L Having thus provided an example of where (the literal] is in- 
tended to be predominant, he now gives an example, acceptable both to 

146 (51.13f L 

his followers and his opponents, of how the name [of a figure of speech] 
derives from its predominant element: and as in, etc. (Although 
we apprehend] a simile: he means that we apprebend a relation of 
subject and image. By that name: viz., "simile." For example, in 
dipake—of which the definition runs thus: "Dipaka is held to be three- 
fold, as it falls at the beginning, middle or end”! (Bhamaha 2.25)—the 
beauty of the verse is occasioned by the operation of the dipaka, (not 
by the suggested simile,] as in this example: 

A jewel placed against the whetstone, 
a victorious warrior wounded by the sword. 
the moon when left with its last digit 
a young woman thin from exercise of sex, 
an elephant io rut, a river 
drawn back from its lovely sandbanks in the autumn. 
and the rich who spend their wealth by giving to the poor: 
all these by lessening grow resplendent 
(Bhartrhari, Nitis. 11) 

Apahnuti has been defined: “The denial of what one really accepts, if 
it contains to some extent a simile. is apahnuti? (Bhamaha 3.21)? In 
this figure it is the denial itself that is charming, as in: 

This is not the buzzing of a bee, 
busy in ber drunken joy; 
it is the twanging of the string 
as Cupid pulls his bow. 
(Bhamaha 3.22)? 

1. Bhámaha's "definition" is in effect no definition at all. [t merely di- 
vides the figure into three types. Dipaka, "the lamp,” is so called because a 
single verb or property serves to illuminate more than one object. Among 
early authors dipeka is the same as the Graeco-Latin zeugma. Mahābhá- 
rata 2.5221: Dhrtarástrena cáhütah kálasya samayena ca, "challenged by 
Dhrtarástra and by the doom of fate." Ovid, Met. 7.133: demisere metu 
vultumque animumque Pelasgi, "the Greeks lowered their faces and spirits in 
fear." From the time of Udbhata the definition was further particularized. 
Udbhata (1.14 Indurája = 1.28 Vivrti) distinguished dipaka from tulyayogita 
(originally quite a different figure, which had come to encroach on dipoka) by 
claimiog that in dipaka the common verb or property must join the matter-in- 
hand (pradhána, prákaranika, prastuta) with some extraneous matter (gauna, 
aprákaronika. aprastuta), whereas in tulyayogità the combined items are all 
within the subject of discourse or all outside it. Thus, an example like the fol- 
lowing, which BANS 16.55 gives of dipaka, becomes an instance of tulyayogitd 

$113g A] 

accordi to the new view: sarāmsi hamsath kusumois ca vrksaih ... tasmin 
na $ünyàni sodà kriyonte, "In autumn lakes are ever filled with geese, the 
trees with flowers, etc.” for all the objects joined by the verb are within the 
subject of discourse, autumn. Later authors usually follow Udbhata's distinc- 
tion. 2. The printed versions of Bhamaha all bave the reading apahnutir 
abhistd ca. If we take Abhinava's reading, we must use the technique of àvrtti, 
that is, we must read apahnutir twice. One might best render apohnuti in 
English as “feigned denial." Its definition remains essentially the same in all 
Alaükàárikas except Dandin, whose view does not here concern us. 3. Note 
that there is an appropriateness to the simile based on sound. The bee is a 
sign of spring, when Cupid annually renews his archery. 

A Also in tbat type of visesokti (here — cause without effect) 
where the reason is not expressed, as in such verses as 

Although bis friends have waked him, 
although he answers, "yes," 
although his mind tells him to go, the traveler 
does not uncurl bis limbs. 

we merely apprehend the suggestion from force of the context, but there 
arises no particular beauty from the apprehension; so the suggestion is 
not predominant. 

1. The verse is by a once famous poet, whose odd name Bharscu appears 
also as Bharvu, Bhascu, Bbatsu. A handful of his verses are preserved by the 
anthologists. What little is known of him may be found in V. Ragbavan's 
Bhoja's SP, sec. ed., pp. 817-818. Bharácu served the Maukhari kings of 
Kanyakubja, was Bàna's teacher (see Kadamban, Introductory verse 4), and 
accordingly belongs to the early seventh century. The verse here quoted is 
found in the anthologies (Sarrig. 3932, Stikti.Af. 63.23, Subh. A. 1838) under 
descriptions of winter, which shows that they followed the interpretation which 
Abhinava (see below) ascribes to Ubdhata, viz., that the unexpressed reason 
for the result not to occur is the fact that the traveler was cold. Abhinava's 
Locana is the earli t of our preserved texts to mention a more romantic 

(§ 113g L 

L After considering áksepa in the above manner, he now speaks 
of the subject that immediately followed it in the order of his initial 
statement:! and in that type of viseasokti. (The Ggure in general 
is thus defined:] “The praise of one quality in the absence of others, 
if made in order to reveal some excellence, is traditionally known as 
visesokti" (Bhamaha 3.23).? An example is the following: 

Alone the god of the lowered bow 
conquers all three worlds, 
whose body Siva destroyed 
but left him still his strength. 
(Bbàmaha 3.24) 

As the reason in this case [viz., the reason why incineration should not 
lead to impotence] is inconceivable, there is no suggestion [of a reason]. 
Even where the reason is explicitly stated, since the statement amounts 
to no more than giving the nature of the object, there is no question 
of a suggested sense. For example: 

I give my praise to him 

who like camphor grows stronger with burning, 

against whom no man prevails, 

the god of the Bowered bow. 

( Bélaréméyane 3.11)? 

Accordingly, our author passes over these two varieties and examines 
only a third type [as a possible case of a predominant suggested sense]: 
when the reason is not expressed. 

[We merely apprehend] the suggestion: viz., according to Bhat- 
todbhata,* the discomfort caused by the cold, this being the reason 
[for tbe traveler's remaining curled up in his bedding]. Comformably 
with this interpretation, our author says, there arises no particular 
beauty here. Some men of taste have imagined a different reason here, 
namely that the traveler does not uncurl his limbs because he wishes 
to bring back sleep, thinking that a dream would be a quicker means 
of union with his beloved than setting forth (on his farther journey]. 
But the experts in figures of speech have not taken even that reason 
as a source of beauty here, but have taken the source of beauty to be 
the words “does not uncurl his limbs," which form part of the visegokti 
itself, as embellished by the suggested reason. Otherwise this would not 
be an example of visegokti at all.* So our author here accommodates 

§1.13h A] 149 

both views by speaking generally [i.e., by not mentioning the precise 
nature of the suggested reason]. It should not be thought that the text 
is based exclusively on the view of Udbhata. 

1. In the initial list (1.13c A) anuktanimittd visegokti followed directly 
on dksepa. The subject of apahnuti and dipaka has intervened merely for the 
sake of an example. Ananda will come back to them again later in their proper 
sequence (1.13h A). 2. The name visesokti (statement of the excellence) 
finds its explanation in the old concept of the figure as seen bere and in 
Dandin. where a visesokti was a statement of deficiency in certain respects, 
made in order to give special prominence or praise to an excellence or efficiency 
in some other respect. Thus, Dandin's example (2.328): "He has a one 
wheeled car, a crippled driver, an odd number of horses, and yet the sun in 
bis glory travels the whole sky.” But Udbhata (5.4 Indurája = 5.5 Vivrti.) 
changed the definition to “A statement that the result fails to arise when all 
the causal factors are present, if made from a desire to point out a [particular] 
excellence." He further divided the figure into two types, one where a reason 
is expressed for the failure of the result to occur, the other where it is not 
expressed. Later writers al) follow Udbbata. In Bhámaba's definition, here 
quoted, samstuti bas its natural meaning "praise," not the watered-down 
meaning of "mention" by which it is glossed by commentators seeking to 
broaden the definition. 3. The reason why Kama remains strong despite 
his incineration is here stated: because he grew stronger with being burned. 
as does camphor. We remain with a literal description of K&ma's nature. 
Liquified camphor gains not only in scent but in refrigerating power, a fact 
that is again used as a simile in Naisadhiyo 7.25. — 4. Presumably Udbhata 
quoted the verse in his commentary on Bhamaba. It is not found in his Kavyd- 
lankdrasitrasarigraha. 5. Obviously, if the suggestion were the chief cause 
of beauty, we should cease to have a figure of speech; we should have dhvani 
i tead. 

A In parydyokta (statement of periphrasis),! if the suggestion is 
predominant we may well include it in dhvani. But by no means may 
we include dhvani in it, for as we shall demonstrate, dhvani is of much 
wider range and is always the predominant element. Furthermore, in 
examples such as that adduced by Bhàmaha, the suggestion is not 

150 [5113h A 

predominant, because there is no intention there of subordinating? the 
literal sense. In apahnuti and dipoka, on the other hard, the literal is 
[always] predominant and the suggested [simile] merely follows along 
with it, as is well known. 

1. See note 1 on the Locana on this passage. 2. Read upasarjani- 

L In paryéyokta: It has been defined thus: “Parydyokta is 
when something is said in a different manner, namely through an under- 
standing that lacks the operation of denoter and denoted” (Udbhata, 
4.6 Indurdja = 4.11 Vivrti).! An example is this: 

The sage Rama, who had strayed i 

in his rage to cut down his foes, 

was instructed in the path of duty by this bow. 
In this stanza, although we understand (the suggestion] that the might 
of Bhisma overcame the might of Parasu RAma, it is merely as helped 
out by this suggestion that the literally used phrase “was instructed in 
the path of duty" ornaments the final meaning of the verse. 

When what is said is distinguished by a parydya (periphrasis), that is, 
speaking in a different manner, which consists in a giving to understand, 
[that is, when it is distinguished] by a suggestion, then the literally used 
words themselves form a parydyokta (statement of periphrasis). Here 
"when something is said" forms the definition, "statement of periphra- 
sis" is the thing to be defined, and the general characteristic of this 
thing is as a figure of speech based on meaning (arthálarikára). And 
so everything is here in order. On the other hand, if one forces on the 
phrase "when something is said" the unnatural interpretation that it 
means "when something is apprehended as the chief element," and if 
one offers as an example such verses as “Go your rounds freely, pious 
monk" (cf. 1.4b A), then the statement of periphrasis will cease to 
be a figure of speech at all, for it will end up as the soul (of poetry, 
namely dhvani]. In that case it should not be included in figures of 
speech. Furthermore, in that case we should have to list subvarieties 
of it [just as injunction, prohibition, etc., were listed as subvarities in 
14b, c, d, e AJ. Our author says this in the words if the suggestion 
is predominant. That is, by including it in dhvani, parydyokta would 
be the soul of poetry, not one of the alarikdras (ornaments, figures of 
speech). In it: dhvani cannot be included in that sort of statement 

§113h Z] 151 

which is intended as a figure of speech. We have not defined dhvani 
as something [subordinate] of that sort. For dhvani has a wide range. 
As it exists in all [sorts of statements], it is widely spread (vydpaka), 
being the principal element on which all the [gunas and alarikdras| are 
placed. A figure of speech is not widely spread any more than any 
other ornament, [e.g., a bracelet is worn only on the wrist, an anklet 
on the ankle]. And it is not the principal, as it is subordinate to the 
Object which it ornaments. Or, if you assume that the suggestion here 
is both of widely spread type and principal, and that there is no figure 
of speech here, why then you would be accepting our own position but 
out of mere spite continuing to call it a paryáyokta. 

Our author now shows that even this much [viz., that the suggestion 
can sometimes be predominant) was not understood by the ancients but 
was first revealed by himself: furthermore. Bhamaha furnished his 
illustration of paryáyokta in accordance with the nature he conceived 
the figure to have. Now in his illustration the suggested element is 
not predominant because it is not the source of any special beauty. 
Accordingly, we may agree that in such other illustrations as may be 
composed along the same lines, there is likewise no predominance of the 
suggested sense. If you disregard the illustration just given and give as 
an illustration some such verse as “Go your rounds freely, gentle monk,” 
you will have become our author’s pupil (for you will be talking about 
dhvani]. But one must say that you have behaved in an unmannerly 
fashion by éducating yourself in his doctrine with disregard of the rules 
and by an illicit hearing of it. The experts in sacred history say: 

He who shows no respect to the teacher 
but listens in hiding to his teaching 
goes straight to bell.” 

The illustration which Bhamaha gives is this: 

Neither at home nor when abroad 

do we eat food that is not eaten 

by learned brahmins 

(Bhamaha 3.9)* 

This statement of the blessed Krishna by a periphrasis averts his being 
given poison. As Bhàmaha says himself, "This is to avoid his being 
given poison." Now there is nothing charming in this suggestion of the 
averting of poison by which we might suppose it to be predominant. 
Rather, the statement of periphrasis itself, viz., that he does not eat 
without the prior eating of brahmins, as embellished by this suggestion, 

152 [$1313h L 

ornaments the matter under discussion, namely his eating of food. The 
intention of the statement is not to say "Give me unpoisoned food." 
And so paryáyokta is simply a figure of speech according to the opinion 
of the ancients. This is what it all comes down to. 

In apahnuti and dipaka: These are figures which he has already 
discussed; so he says, as is well known. He means the matter has 
already been proved with valid means of proof. Previously he brought 
up these figures to serve as illustration of how* a figure does not take 
its name from the [subordinate] simile, etc., [which it might contain] 
whereas now he mentions them in a different way to show that they 
are not dhvans because their suggestions are not prominent. This men- 
tioning of them a second time is to keep the order of his original list* 
so that he may make his text all of one piece.5 But the matter on both 
occasions is basically the same, for one might suspect them of being 
dhvans from the fact that they suggest a simile. 

As for the statement of the author of the Vivarana,’ based on an 
examination of many instances, that a (suggestion of] simile does not 
always accompany a dipaka, it is unhelpful, without merit, and can 
easily be refuted. In the stanza: 

Infatuation creates desire; 
and that, love with its loss of pride; 
that, a yearning to gain the beloved; 
and that, unbearable pain of heart. 
(Bhámaha 2.27)* 

one can easily imegine a relation of subject and simile between the 
terms even though the terms are produced successively. One cannot 
say tbat such a relation of similarity is impossible among successive 
objects, for we have the verse: 

Dasaratha was like Rima, 

Raghu like Dasaratha and Aja like Raghu 
and the whole race of Diipa like Aja: 
marvellous is the glory of Rama,? 

where one cannot say that it does not exist. So why worry whether 
sequentiality or contemporaneity hinders the (suggestion of] simile? 
Enough of trying to milk a mule. 

1. In defining parydyokta Bh&maha merely says: “Parydyokta is when 
something is said in a diferent way.” (parydyoktam yad anyena prakárenábhi- 
dhiyate, 3.8.) Udbhata here repeats the words of Bhamaha but adds a second 

§1.13h L] 153 

half verse. which for the first time introduces into the figure the notion of sug- 
gestion. His definition is followed by Mammata (10.175, p. 680) and Ruyyaka 
(p. 141). Only late in the history of alarikárasástra does the restriction appear 
by which most present-day students recognize the figure. "In a different way” 
is then taken to mean “by stating the result." So SD 10.61. Thus parydyokta 
in the modern view is the sugestion of a cause by stating the result. 2. The 
quotation is presumably from some Purana. It is not found in MBh. or Ram. 
One is supposed, of course, to bow down and touch the teacher's feet before 
hearing his words; and the punishments of südras who overhear Vedic teaching 
are famous. But the quotation is here intended humorously. — 3. The English 
reader may need more explanation than Abhinava furnishes. The words are 
supposed to be spoken by Krishna on his visit to Sisupala, where he stands in 
danger of assassination. Without the hint of that possibility the words, taken 
literally, express the highly pious protocol that Krishna might be expected 
to follow. Out of respect he always bas his priests fed before himself. The 
expression is enlivened, is given a twist of wry humor, by the suggestion, but 
it is the expression itself that delights us. 4. amuyd cchdyayd: the same 
as amuná prokdrena, "of this sort, (namely, that a &gure, etc]." — S. Viz., 
in 1.13¢ A. See also 1.13g L, note 1. — 6. granthasayyd: a smooth text, 
an orderly presentation. Kaumudi: jayyá náma ekarüpah sannivesavisesah, 
“Sayyd is a particular way of arrangiog matters so that they are all of one 
and the same form.” See PW's third definition under sayyd, which gives as 
synonyms gumphana, Sabdagumphana, granthasya nirmitih. The term bas an 
interesting development in later writers, where padasayyd, like pāka, comes to 
mean the perfect choice of words; see Pratáparudriya 2.34 (p. 49) and Tarola 
on Ékàvalrl.l3 (p.22). T. Kane (HSP p. 126) supposed that this was a ref- 
erence to Udbhata as the author of the Bhámahavivarana, a work specifically 
referred to by Abhinava at 3.16 m L. But Kane was forced to rely on the incor- 
rect text of the KM edition, where the negative is omitted from upamánvayo 
ndstiti, The supposition seems to me impossible. In Udbhata's Kàvyálari- 
kórasorigroha there is no mention of such a theory as is here attributed to 
the Vivaranakm. In fact he there defines dipaka as containing a simile (1.14 
Indurája = 1.28 Vivrti). Furthermore, I cannot imagine Abhinava referring 
to Udbhata as a mule (see end of section) even if he disagreed with bim. 
8. The verse is given by Bhámaha as an example of ddidipaka, as the common 
verb ("creates") is given in tbe first quarter. It is really a sorites (káranamald) 
Sharing the pecularity of dipaka that the common verb is bracketed after its 
first occurrence. Presumably in order to make it more clearly illustrate di- 
paka, the Jayamarigala commentary on Bhattikávya 10.22 alters it drastically: 
mado janayati pritim ánandam mánabhariguram / yat priyásarigamotkanthám 
aschydm mánasah $ucam. — 9. The figure here is raosanopamá (chain simile); 
see KP 10.413, p. $80, SD 10.25. 


A In sarikara (fusion, see Locana, note 1) also, when one figure 
assists another with its color or charm, the suggested figure cannot be 
intended to be predominant and so falls outside the area of dhvani. 
Where either of two figures is possible, there is equal prominence of 
the literal and the suggested. But if the suggested figure subordinates 
to itself the literal figure, then we may assign it to the area of dhvani. 
We cannot say, however, that dhvani is just that, because of the same 
reasons that we gave in speaking of paryáyokta (cf. 1.13h A). Further- 
more, in all cases! of sarkara the very name "sarikara" will prevent us 
from thinking of dhvani.? 

1, We have followed Abhinava's interpretation of api ca sarikardlankdre 
‘pica kvacit... nirdkaroti to mean sarikarólarikóre ca kvacid apt... nirākaroti 
The natural meaning of Ánanda's sentence would be, "And sometimes in 
sarkaro the very name 'sarikaru' will prevent us from thinking of dhvani." 
But the natural interpretation gives a less logical train of thought. — 2. The 
name "sarikaro" denotes a figure of speech. A figure of speech cannot be an 
instance of dhvani. 

L In the figure sarikara.! One type of sarikara is defined thus: 

When contradictory figures appear, 
when they cannot function together, 
and when there is neither right nor wrong 
in accepting just the one, we have sarikara. 
(Udbbata, 5.11 Indurája = 5.20 Vivrti)? 

An example is a verse of mine: 

Moon-faced she is, 

dark water-lily-eyed 

and with teeth of white jasmine: 
God has given her forms of beauty 
from sky, river and earth. 

Either metaphor or simile may appear here, depending on whether 
we analyse “moon-faced” as “having a moon for her face” (by Pani 

$ 1131 L) 155 

2.1.72], or as "having a face like the moon" [by Panini 2.1.56]. As 
both figures cannot be entertained at the same time and as there is no 
compelling reason to accept or reject one of the other, we have here 
the figure sankara. As there is no evidence that one of the components 
is denoted and the other suggested, what possibility is there here of 

And what chance is there in the second type, where a figure of sound 
and a figure of sense occur in one [sentence]? As i 

Remember as K&ma your lover, 
in whose embrace you have found delight,? 

where we have the figure of sound yamaka‘ and the figure of sense 

In the third type too, where more than one figure of sense is found in 
a single portion of a sentence, inasmuch as both are equally presented 
literally], how can either be suggested? As in: 

They rise and sink together. 
so when the bright sun has set, 
the weary day for rest 
enters as it were the cave of night. 
(Bbámaha 3.48)* 

For in this stanza the partial metaphor (viz., “the cave of night"] im- 
plies a full metaphor of a well-bred man eager to perform the appro- 
„priate duty [of self-immolation] on the fall of his master. (The partial 
metaphor is directly expressed] and poetic fancy is also directly ex- 
pressed, viz., by the term, "as it were." These two types are embodied 
in the following definition: "When figures of sound and sense occur in 
one sentence or one portion of a sentence, we have the figure sarikara" 
(Udbhata, 5.12 Indurája = 5.22 Vivrti). 
The fourth type is where there is a relation of assistance between the 
figures. As in this: 

These glances of the long-eyed maid 
that tremble like water-lilies in the wind: 
did she borrow them from the does of the wood. 
or did the does borrow them from her? 
(Kumárasambhava 1.46) 

Although a likening of Párvati's glances to the glances of female deer 
is here suggested, the simile becomes subordinate from its role as as- 
sistant in giving rise to the figure sandeha (poetic doubt)," for by its 

(§ 213i L 

assistance we end up With the figure sandeha. This [fourth type of 
sarikara) is defined: "Where figures are placed in mutual assistance 
and lack independent being, that too is the figure sarikara" (Udbhata, 
5.13 Induraja = 5.25 Vivrti). 

This is what our author refers to in the words when a figure, etc. 
And so the existence of dhvani in the fourth type is ruled out. We 
have already stated that there is no possibility of dhvani in the second 
and third types. But in regard to the first type, exemplified in “Moon- 
faced she is,” etc., one might suspect the possibility. This he now rules 
out: where either of two figures. Equal: because our mind sways 
between the two. 

But now, when the suggested sense appears to be predominant, what 
are we to do? Take the example: 

The masses have no care of quality 

but easily fall for reputation. 

The moonstone sweats at sight of the moon 
but not at my true love's face." 

Here the figure arthántaranyása (substantiation) appears, literally ex- 
pressed, but vyatireka (contrast) and apahnuti (feigned denial) appear 
by suggestion and are predominant. Having this in mind, he says, but 
if, etc. He answers the problem with then we may assign it. That is 
to say, this is not sarikara at all, but is the second type of dhvani called 
alarikáradhvan: (sugestion of a figure of speech).? What was said under 
the subject of parydyokte {viz., that dhvani is of wider range than the 
figure and is always predominant} may be equally applied here. 

Now he gives a general means of denying the possibility of a [predomi- 
nant] suggested sense in any form of sarikara: furthermore. One must 
construe by changing the position of the particle apt. The sense is, “ev- 
erywhere in the figure sarikara," as distinguished from any particular 
variety. For fusion means a mingling, a complete commixture.'? How 
would one element predominate any more than in the mixture of milk 
and water? 

1. Frequently more than one figure of speech is found in a sentence or 
stanza. Such cases are assigned to samsrsti (association) or to sarikara (fu- 
sion). In samsrsti the figures are associated "like sesamum grains and rice 
grains"; that is, they can be distinguished and separated. [n sarkara the mix- 
ture, "like milk and water," is irresolvable. By Abhinava's time sarikara was 
divided into four types, which he will here define and illustrate. (a) alarikéra- 
sandigdhatva, where there is doubt to which of two alarikáras the case should 

$143iL] 157 

be assigned; (b) olarikáraikavákyánupravesa (also called alarikdratkaudcakdnu- 
pravesa), where two or more figures, with no doubt as to their identification, 
are combined in a single sentence; (c) alarikératkavdkyémsdnupravesa, the 
same as type two except that here the figures are combined in a single portion 
of a sentence; (d) alarikárárigárigitva (or alarikáránugráhyánugróhokabháva), 
where the two figures assist each other, where the charm of each depends 
on its involvement with the other. The whole scheme is often simplified by 
combining types two and three. Ananda shows a clear awareness of types 
one and four. Abhinava, in interpreting Ananda’s remarks to cover all four 
types. quotes ancient authors whose verses, as the footnotes will show, do 
not always fit exactly the later scheme which be bas in mind. 2. The text 
of Udbbata reads anekálarikriyó, “more than one figure," rather than "con- 
tradictory figures." 3. The better attested reading is priyam sma ramase 
(Kaumudi). If one takes priyam ramayase, the sense will be “whom you de- 
light with your embrace." The line is in prthui meter. — 4. Yamaka is the 
repetition of two or more syllables of the same sound but in different mean- 
ing, e.g.. smaram and sma ram(ase). — 5. Abbinava is taking this verse as an 
example of the third type of sarikaro called ekavdkydmsdnuprovesa. Bhimaha 
furnished it as an example of a figure which he called utpreksánvaya, a figure 
of far more restricted application than the type of sarikara which Abhinava 
wishes to exemplify. Bh&maha requires for his figure ao ambiguous expres- 
sion (slista), here instanced in udaydvasdna, which may mean the rising and 
setting of heavenly bodies, or the success and failure of humans. [n addi- 
tion there must be utprekyd (poetic fancy), bere evidenced by the particle 
iva, "as it were," and also "a sense of metaphor" (ripakértha). By the last 
stipulation I suppose he refers to the partial metaphor (ekadesavivartirüpaka) 
*that is directly expressed in the stanza. It is partial because it encompasses 
only one element of the sentence, the darkness to which the day goes. The 
much larger, suggested metaphor, which involves all elements of the sentence, 
is another matter, which Abhinava will speak of and which we shall remark 
on in the following note. 6. We are left with the difficult problem of the 
implied metaphor. It is difficult to see how it can be regarded as other than 
suggested. Presumably Abhinava takes the suggestion as subordinate to, and 
merely embellishing, the literally expressed partial metaphor. By such a view 
be could claim that as none of the figures is predominant, there is no dhvani 
in the verse. But surely this is a perverse reading of the poem. The suggested 
metaphor, or aprastutaprasamsd, forms its beart and its whole beauty. What 
we relish and remember is the suggestion of a faithful servant who begins to 
die a little when his beloved master dies. Abhinava's rejection of dhvani in this 
case can be explained only by his wish to exclude all cases of suggestion that 
were involved in the old system of olarikdras. This is an old verse by an au- 
thor who did not recognize dħvani. 7. Our reading sandeha here and in the 
next line is preferable to the Kaumudi's reading sasandeha. Udbhata, whose 

158 [$113i L 

remarks on this figure are referred to by Ananda later on (2.26 A), makes a 
distinction between sasandeha (embodiment of doubt) and sandeha poetic 
doubt). The first figure he finds in those verses where the doubt is expressly 
corrected, e.g., “People wonder on seeing the conch in Visnu's hand, ‘Is this 
a wild goose that has come to the lotus growing from bis navel?’ But no, it 
does not move.” In sandeha, on the other hand, the doubt is not removed. 
Rather, it gives rise to a suggestion of some otber figure of speech. The ex- 
ample here quoted from Kalidasa is clearly of the latter type. Whether we 
tegard the sandeha as giving rise to upamd or the suggested upamá as giving 
tise to sandeha is unimportant. The doubt is not resolved and there are two 
figures bere assisting each other. In later authors the terminology is changed. 
Both of Udbbata's figures are known by the same name (by Mammata as sa- 
sandeha, by most others as sandeha), but the former type is distinguished as 
"containing a resolution? (niscayagarbha or niscayánta) while the latter type 
is called "pure" (suddha). 8. We bave taken the readings of tbe Kaumudi, 
pohnavai (for the senseless pahinusai) and na before piámuhe. The point of 
the verse is that the poet would portray the face of his beloved as more beauti- 
ful than the moon. He arranges the stanza in such fashion that the suggestion 
of her beauty is expressed by an arthdntaranydsa (substantiation) apparently 
intended for quite a different purpose. For definitions of substantiation see 
Dandiu 2.169, Mammata 10.109. A general statement may be substantiated 
by a particular (as here), or a particular by a general. Moon-stones give off 
moisture (they are said to sweat or to weep) when exposed to the light of 
tbe moon. But, ignorant creatures that they are, they fail to sweat at the 
lady's face, which is more beautiful. This instance substantiates the general 
rule that common people respect reputation rather than true quality. The 
figure substantiation is explicit. The vyatireka (see 1.1 Intro. L, note 12), in 
tbe form “My true love's face is more beautiful than tbe moon,” is merely 
suggested, as is also the apahnuti in the form “This is not the moon; the true 
moon is my beloved's face” 9. Ananda will treat alarikáradhvani under 
10. We take this sense for lolibháva from the comment 

not found in PW. 

A So also in aprastutaprasamsá (reference by means of the ex- 
traneous, allegory).! When, by a relation of general and particular, or 
cause and effect, there is a connection of a literally stated extraneous 

$13j A] 159 

matter with the suggested subject in hand, then the literally stated 
and the suggested meanings are equally important. To begin with the 
case where there is a connection of a literally stated extraneous gen- 
erality with a suggested germane particular: although xe apprehend 
the particular as important (for that is the final intention of the sen- 
tence], we must admit that the general statement is equally important 
because the particular cannot exist without the general. Again, when 
a [literally stated) particular ends up in a general suggestion, [although 
the general suggestion is important as being the final intention} the 
particular is also important because all particulars are included in the 
general. The same principle holds where the relation is one of cause and 
effect [viz., cause cannot exist without effect, nor effect without cause; 
so both are equally important]. But in an oprastutaprosomsá where 
the connection of extraneous and germane is based solely on similarity, 
there, if the literally stated extraneous (member of the] similar [pairs] 
is not intended to be predominant, the case falls in the area of dhuani. 
Othervise, it will just be one of the figures. 

1. The literal meaning of the term aprastutaprasamsá is "praise by means 
of the extraneous.” The extraneous is the matter not in hand. Bhàmaha 
, (3.29-30) and Dandin (2.340) take the name quite literally. To them aprastu- 
taprasamsá was simply a special type of samásokti where the matter in hand 
(the prostuta or real subject intended) is praised by a statement of something 
extraneous (aprastuta). Thus in Bhàmaha's example we find praise of the 
magnanimity of trees, by which one understands the intended praise of good 
and generous men. Dandin's example praises the simple Life of deer in the 
forest, from which one is to understand his praise of a life away from court 
where one need not fawn upon kings. It is Udbbata (5.8 Induraja = 5.14 
Vivrti) who dispenses with the notion of praise. Any mention of an extraneous 
matter, if it is so connected with the subject in hand as to suggest it, is called 
oprastutaprasamsd. Ananda accepted Udbhata's definition and then divided 
the figure into three or five types. The five types are: where something 
general suggests the particular, where a particular suggests the general. where 
a cause suggests the result, where a result suggests the cause. and where like 
suggests like. The first and second, as also the third and fourth. may be taken 
together, giving only three types. The last type of all (also called anyokti) is 
essentially the Graeco-Latin allegory. Later writers went on to subdivide the 
last type also. 

[5143j L 

L The praise of something different, 
that is, other than the matter in hand, 
is called aprostutaprasarnsá. 
It is threefold." 

What is meant is a description of extraneous matter which hints at the 
matter in hand. This hint is threefold, as it is based on a relation of 
general and particular, on a relation of cause and effect, or on similar- 
ity. In the sentence beginning When by a relation and ending with 
equally important our author sets forth the thesis that in the first 
two of these types the extraneous matter and the matter in hand are 
equally important. 

In the type based on a relation of genera] and particular there are 
two methods of procedure. One method is where a general statement 
which is extraneous is literally expressed and the particular, which is 
the matter in hand, is suggested; as in the following: 

Ah, the cruelty of worldly life, 
the maliguity of misfortune! 
Ab, the tragic ways of fate 
deceptive in its very nature. 

Here the power of fate in its general form, which is extraneous to the 
poet's real intention, and which is stated throughout the verse, ends up 
in [a suggestion of] the matter in hand, which is a particular disaster 
that has befallen someone. Here the general statement, which is lit- 
erally expressed, is as important as the particular, which is suggested, 
because the species is logically included in its genus; for there is no con- 
tradiction in the simultaneous importance of general and particular. 

When a particular which is extraneous suggests a general statement 
which is germane, we have the second type; as in the following: 

It is not so much that at first the fool imagined 
a drop of water upon a lotus leaf 
to be a pearl; but hear what happened next: 
as he tried to take it slowly on his finger tip 
with gentle motion, it melted at his touch. 
At this be cried, "Alas, it has fown away!" 
and now he cannot sleep from inner grief. 
(Bhallatasataka 94)? 

§ 1.135 £] 

In this stanza the matter intended by the poet is the general princi- 
ple that people imagine greatness in what is really nothing, while the 
extraneous subject is particular, that of imagining a pearl in a drop 
of water. Here too there is no contradiction in the simultaneous im- 
portance of the general and the particular, as has been said. In this 
way our author has dealt with the first type in its two varieties in the 
passage from “to begin with” to “all particulars are included in 
the general.” 

Extending the same principle to the type based on a relation of cause 
and effect, our author shows that it too has two varieties: where the 
relation is one of cause and effect. Sometimes a cause, which is 
extraneous, is presented literally in order to suggest an effect, which is 
the matter in hand. For example: 

They who take joy in your success 
and stay with you in adversity 

are your true relatives and friends; 
other seek only their own benefit. 

Here the speaker states explicitly a cause which is extraneous to his real 
intent. The cause is that good men by their faithful attachment are 
friends and relatives to us. He states this in order to suggest what he 
really intends, namely that his own words should be trusted.? Although 
we apprehend the effect here, our apprehension of the cause becomes 
important from its giving life to the effect; so both cause and effect are 

Sometimes an effect, which is extraneous, being literally presented 
suggests a cause which is the matter in hand. An example will be 
found in the Setubandha: 

I remember before the churning of the sea 
heaven without its párijáta trees, 
Visnu without his Laksmi by his side 
or the kaustubha jewel upon bis breast, and Siva 
without the lovely moon within his locks. 
(Setubandha 4.20)* 

Here Jámbavàn describes his memory of Visnu’s breast without the 
kaustubha jewel or Laksmi, and so on, which are matters extraneous 
to, and at the same time a cause of, what is his real intention. He does 
this in order to suggest the result, namely that his service of the elder 
gods, his longevity, and his skill in negotiations fit him to be accepted 
as an advisor. Here the suggested and the literal senses are equally 

162 [$113jL 

important, for although we apprehend the cause to be important, the 
effect, which is literally presented, raises itself (into importance] by its 
giving life to that cause. 

After dealing with these two types, each of which is twofold, he ex- 
amines the third type, which is based on similarity. Here too there 
are two varieties. Sometimes the charm (and therefore the importance] 
comes from the extraneous matter which is literally presented, while 
the suggested matter is subservient. An example is this verse by my 
teacher Bhattendurdja: 

He who brought you back to life 
and by his strength supported, 

who carried you upon his back 

and even gave you worship: 

that man you kill with but a smile 
O brother zombie, you show yourself 
to be the prince of gratitude.* 

Here, although some other ingrate is suggested by the power of simi- 
larity, the charm and interest of the stanza lie in the anecdote of the 
zombie (vetála), which is extraneous. The sense is not impossible as 
would be a reproach against an insentient being, and the anecdote is 
not without attraction. So the predominance here lies in the literal 

But if the matter in hand is charming and is suggested by a literal 
description of an extraneous subject possessing properties such as in- 
sentiency which render it impossible for the described purpose, then 
we have a case of vastudhvani. An example is a verse of my own: 

Troop of delights, who storm the hearts of men 
and make them dance in many an antic step; 
concealing your own intention as you play; 

men call you brute and stupid, in their ignorance 
thinking themselves intelligent thereby. 

That title of stupidity, 1 think, 

if given to them would be honorific, 

for it -would seem to liken them to you. 

Here the germane matter, which is revealed by suggestion, is the ex- 
traordinary way of life of a man while he is being despised by the 
world as a fool. What is meant is a great man, who has rolled back 
the curtains of darkness by an eye of deep penetration and who lives 
a worldly life on the principle of “unimpassioned yet as though with 

$1413j L} 163 

passion," concealing his real self and accepting the reproaches of men 
as he causes their tongues to wag. 

[The literal meaning.) A delight, such as a garden or moonrise, is 
despised by men as being a brute thing, but it can make the heart 
of an absent lover grieve with yearning, or make the heart of another 
overflow with joy. Of what sort its own heart may be no one knows. 
In fact it is vastly deep and intelligent, utterly void of pride and skilful 
at play. Now if people for that reason call it brute and stupid when 
it stands rather in honor because of its implied intelligence, and for 
the same reason honor themselves as intelligent when they are, rather, 
worthy of being considered stupid, then the expression “you are brute 
and stupid,” being established as an epithet of the troop of delights, 
which we have seen are really intelligent, will, rather, be an honorific. 
What is hinted at is that men are worse than stupid. He indicates all 
this in the sentence but if. 

Otherwise: otherwise it will just be another figure of speech, that 
is, the particular figure of speech [aprastutaprasamsd], but never when 
the suggested element is predominant. 

1. The definition is, basically, the one given by Bhamaha (3.29), but 
Abbinava has changed the last páda. The original says nothing of the figure's 
being threefold. The second half reads aprastutaprasamseti sd caiva kathyate 
yothd. The division into types is first found in Ananda. 2. 1 have fol- 
lowed the interpretation of both Kaumudi and BP. The point of the stanza 
is that it is not so egregious a folly to mistake a drop of water for a pearl; 
after all, they look alike. But to be so convinced of one's error that one wall 
attribute volitional flight to a pearl shows the overwhelming power of vain 
hope in humanity. Abhinava's readings are superior to those of the printed 
text of Bhallata, which has srnvann akasmád api in b and tatas in place of 
Sonais in c. The alteration in b seems to have arisen from a reader who 
misunderstood tasya mukhdt to mean “from his mouth,” presumably from 
the mouth of someone playing a joke on the fool. The verse is quoted again 
by Mammata 10.441, p. 621. Bhallata's Sataka comprises the work of vari- 
ous authors; see 1.14 A, note 6. 3. One could scarcely elicit this meaning 
without knowing the context in which the words were spoken. But Abhinava 
doubtless knew the context, as did Hemacandra who also quotes the verse 
(AC 559, p. 365), where he says that the words are spoken by Jarésandha 
(the enemy of Krishna. soon to be slain by Bhima). Presumably the quota- 
tion is taken from some lost play. 4. The stanza forms the opening lines 
in Jàmbaván's speech of advice to the leaders of the army about to attempt 
an assualt on Lanka. Jümbavàn is the Methuselah of Indian legend. The 
trees, Laksmi, the kaustubha jewel, and the moon were all acquired from the 

164 (§ 1.135 L 

gods’ churning of the sea many thousand years ago. 5. The verse is ironic 
and is used allegorically. In the Vetdlaparicaviméati the belief is found that a 
corpse can be brought to life as a vetála (zombie). As first revived, the vetdla 
cannot walk but must be carried. lf worshipped with Tantric rites he may 
give the practitioner magic powers. Perhaps one may infer from this stanza 
that if the vetála smiles the practitioner will die. The translation of vetdla by 
“vampire,” though sanctioned by long usage, is utisleading. The vetdla does 
not suck blood; he is a revived corpse, a zombie. 6. What is one to make of 
Abhinava's account of his own verse? The literal meaning of the stanza is not 
dif&cult. "Men who decry, as do the noo-Tantric philosophers, the delights of 
love and of the senses, calling them brute pleasures, are really stupider than 
the pleasure they run down. So I will not copy them by calling names. To call 
them stupid would be to compliment them." Now it is true that the literal 
sense is impossible from the realistic point of view in which the words “pos- 
sible" (sambhávya) and “impossible” are used by the Álahkárikas. Neither 
garden nor moonrise, being insentient, actuallly makes the heart dance, nor 
do they conceal their own beart, for they have none. So one is forced to look 
for a second meaning. To pass to that second meaning is more difficult. Abhi- 
nava has thrown what seems to me a aeedless stumbling block in our way by 
the discrepancy between the plurality of delights (or stimulants, bhdvavrata) 
and the singularity of the great man (mahdpurusa). But the great man does 
conceal his thoughts. His causing the tongues of men to wag, in the case of the 
Pasupatas and | dare say of many Tantrics, was a premeditated instigation of 
reproach; see D. H. H. Ingalls, “Cynics and P&supatas: The Seeking of Dis- 
honor," Harvard Theological Review 55.4 (1962). There is also an underlying 
compatibility of the Tantric adept, seeking moksa by the path of bhoga, and 
the worldly stimulants amidst which he lives. 

A Here then is a summary of the matter: 

Where the suggested meaning does not predominate 
but merely accompanies the literal sense, 
there we clearly have ornaments of the literal 
[i.e., figures of speech] 
such as samásokti and the rest. 
Where the suggested appears only faintly, 
or merely follows along with the literal, 
or is not felt to be the more important: 
in such places there is no dhvani. 


Onlv where word and sense are subordi 
and directed toward the suggestion, 
and where there is no fusion (sarikara), 
are we in the area of dhvani. 

Accordingly, dhvani cannot be included in any other category. And for 
this reason too it cannot be included: because it is a particular poetic 
whole (arigin) that has been called dhvani. Its parts will be shown in 
the sequel to be the figures of speech, tbe qualities (gunas) and the 
alliterations (vrttis).' A part, if taken by itself, is never known as a 
whole, while if taken together with the whole, it is recognized as a part 
of it, not as the whole itself. Even where one of the elements [which 
are normally parts] does constitute a case of dhvani,? dhvani because 
of its vast range is not li ited to it. 

1. For the gunas and vrttis see 1.1a A, notes 4 and 5. — 2. As in certain 
cases of paryáyokta (see 1.13h A) or sarikara (see 1.13i A). 

L In the list (of seven figures of speech] with which our author 
began [bis discussion (1.13c A)], the words "and the like" refer to any 
other figure where a suggestion may be imagined, that is, to vydjastuti 
(trick praise), etc. Our author proceeds to give a general answer to 
all cases of that sort: Here then, etc. His feeling is that it is useless 
to write on each particular figure. 

Among such figures [we may give] an example of vyajastuti: 

What good is done by telling on other wives? 
And yet, being a chatterbox by nature 
and a southerner as well, 
I can't keep still. 
She's in everybody's house, 
in the market, at the crossroads and at drinking bouts; 
she runs about like a drunkard, does your mistress. 
Obo, but her name 
is Fame. 

Here it is the literal meaning that is embellished by the suggestion 
in the form of praise. Another critic has offered the following as an 
example of the figure: 

[$ 1.13k L 

The great earth, lord, engirdled by the sea, 
was wedded to your grandsire; 

she next became your mother, while today 
to raise a family you keep her as your wife. 
After a full century, without reproach, 

she will be married to your son. 

Say is this decent in a line of kings 

who know the rules of proper conduct? 

This stanza strikes me as obscene, because it causes one to think of 
higbly indecent things. And what does the praise amount to? That 
you are king by hereditary succession. What is so great about that? 
Trick praise of this sort will be reproved in any company of senstive 
critics and deserves to be ignored. 

[To explain the term “etc.” in bis remark "vyájastuti, etc.," Abhinava 
takes up another figure of speech, bháva (expression of inner feeling):] 

If an alteration [of a given state of mind], arising from an [apparently] un- 
connected cause, gives us to understand the intention and its coanection with 
that cause, we have the figure bhàvo. (Rudrata 7.38)? 

Here too if the literal sense is predominant, the case is one of a figure 
of speech, bhàvo. That is to say, if an alteration of a state of mind 
appears, such as the speaking of certain words not normally connected 
[with their apparent cause}, and gives us to understand for what reason 
the intention embodied in that state of mind [has arisen|—here the 
reason may be such as the aim of enjoying the pleasures of love without 
stint—we have the figure of speech, bhava. For example: 

As I am a weak woman, 
young and left in the house 
while my husband has gone abroad, 
with no one here but my blind and deaf mother-i 
how can you be so foolish, traveler, 
as to ask to spend the night? 
(Rudrata 7.41)* 

Here a suggestion embellishes the literal sense of each word and so the 
literal sense is predominant. On the other hand, if the suggested sense 
were predominant, we should not have a figure of speech at all, as we 
have shown before.* So enough of many words. 

Where: i.e., in a poem. Ornaments: it is because they are orna- 
ments that they act only to embellish the literal sense. 

$113k L] 167 

Appears only faintly: that is, where there is a vague impression. as 
in the simile, etc. (suggested by the figure dipako, etc.]. Follows along 
with the literal: the meaning of following along with the literal is 
the having of equal importance with it, as in aprastutaprasamsá. Not 
felt to be: where its predominance does not appear clearly, but must 
be forced on it and so does not really enter the heart, as in the verse, 
"Turn back, I beg you," as explained by other commentators.5 By these 
two verses he shows that in four cases we are not justified in speaking 
of dhvani: where the suggested sense, although it is present, is not pre- 
dominant [being a mere ornament], where it is faintly perceived, where 
it is equally important with the literal, and where its predominance is 
not clear. Where then are we justified? He tells us, only where word 
and sense are subordinate. He adds, where there is no fusion. 
By "fusion" he means the possibility of inserting any figure of speech. It 
is wrong to interpret as "without the figure called sarikara (fusion)," for 
then it would be difficult to take the proviso as prohibiting other figures. 

And for this reason too: not only by reference to the contradiction 
of denoter-denoted to suggestor-suggested can it be shown that the 
figures of speech and dhvani are not identical, but because there is a 
contradiction between the nature of a whole and of a part, as there is 
between that of master and servant. Its parts: that is, taken singly, 
as he goes on to say: if taken by itself. Very well, then, let us not 
take [a figure of speech] by itself but regard it in the context of the 
whole. In defense against this proposal he says, if taken together with 
the whole. It is then not the whole itself because other constituents 
afe included in the whole; and among its constituents is the suggested 
sense which is not a figure of speech because of its predominance. He 
makes this point with the words: not the whole itself. 

Now it may be objected against our author that he has consecrated 
an occasional instance [of what appears to be a figure of speech} with 
predominance and has recognized it as the soul of poetry, dhvani." 
With this in mind. he now says, even where one of these, etc. He 
has not consecrated any one of the figures, such as samásokti, as dhvani 
itself, because each figure may exist separately from dhvani and because 
dhvani is found in the absence of all figures of speech, samásokti and 
the rest, as in the stanzas “Mother-in-law sleeps here" (1.4c A) and 
“Who would not be angry" (1.4f A). He makes this point with the 
words, is not limited to it. 

1. Vyájastuti = vyájena stut',, "praise by means of a trick.” The immedi- 
ate impression is one of reproach, but as one thinks of the implications one sees 

168 [$ 113k L 

that praise has been expressed. This is the old sense of the term, as defined and 
illustrated by Bhamaha (3.31), Dandin (2.343), Vamana (4.3.24) and Udbhata 
(Indurája 5.9 = Vivrti 5.16-17). There is no evidence that Abhinava recog- 
nized the extended definition which Mammata (10.112, p. 670) picked up from 
Rudrata's vyàjaslesa (Rudrata 10.11) and which bas been followed by all later 
Alaükárikas. In the later view, vydjostuti may be either trick praise or false 
praise (vydjanipd stutih), i.e., a sentence apparently offering praise, but as one 
thinks of the implications, expressing reproach. — 2. The stanza is ascribed to 
Vidy& by the oldest of our preserved anthologies, SRK 996. The ascription has 
this io its favor, that Vidyá was a southerner and an admirable poetess. But 
other anthologies ascribe it to Mátabga-div&kara (Sdrvig. 1227, SubhA. 2544). 
The verse has given rise to much discussion; see Ruyyaka, p. 144, and Rasa- 
gorgádare, p. 418. 3. Abhinava here quotes Rudrata's definition of the 
first type of bhóva. He goes on to interpret the definition so that it may fit 
Rudrata's illustration of the second type of bháva, which he quotes, omitting 
the illustration of the first type and the definition of the second type. I can- 
not say whether Abhinava does this advisedly, with the intention of reducing 
Rudrata's two types to one, or by mistake. He may bave read from a defec- 
tive copy of Rudrata, where the verses were omitted, or his memory may have 
played him false. At any rate, the original, as it stands in Rudrata, runs as 
follows: 7.38 (naturally interpreted): When an emotional alteration (vikdra) 
of a person, arising from a cause which is not [normally] connected with it 
[i.e.. productive of it}, gives us to understand what that [cause] means to that 
person and tbat [in this case| it really is so connected, we bave the figure 
bháva. 7.39 (illustration): On seeing the village youth / with a varicula flower 
in his band / the face of the young girl / changes color. The commentator 
Namis&dhu explains. The girl would not normally be affected at sight of a vañ- 
cula flower, 3o we seek for an explanation. It lies in the suggestion that she has 
made a rendezvous with the youth in a varicula grove, whicb she was prevented 
from keeping. When she sees the youth with a varicula flower in his hand, she 
realizes tbat he kept the tryst and that she has missed the opportunity of love- 
making. 7.40: When a sentence, in denoting just this, gives us to understand 
something that differs from this in regard to good and bad, we have another 
type of bhdva 7.41 (illustration): As [ am a weak woman, etc., as quoted by 
Abhinava. 4. The verse has been a favorite and is quoted in virtually all the 
great anthologies: Särg. 3773, SubhA. 2234, SüktM. 87.11, Saduktı. $47. It 
portrays by innuendo an unchaste wife (asati) and has just as good a claim to 
be considered a case of vastudhvani as the verse “Go your rounds freely, pious 
monk" (1.46 A). But see the following footnote. — 5. Viz., 1.13h L, com- 
menting on the remark of A: "if the suggestion is predominant." It is of course 
a matter of taste whether one regards the literal sense or the suggested sense 
as more important. But Abhinava takes the fact that the verse was quoted by 
Rudrata as illustrating a figure of speech to show that the literal sense must 

$1131L] 169 

predominate. On the other band, the verse "Go your rounds freely, pious 
monk" was quoted by Ananda as an illustration of dhvani. A Sanskrit com- 
mentator, without strong provocation, will not argue with ancient authorities. 
The irreverent Westerner of course is free to do so. — 6. See the interpreta- 
tion given in 1.4e L of this verse, where it is shown that some critics took the 
verse to exemplify preyo'larikdra or rasavadalankáre. 7. See 1.13j A, end 
of passage, and the illustrative stanza “Troop of delights,” etc., in 1.13} L. 

A When [Kariké 1.13] says "which the wise call dhvani," this 
means that the term was invented by men of knowledge and that it 
bas not been put into use inadvisedly. The preeminent men of knowl- 
edge are the grammarians, for all the sciences rest upon grammar; and 
they gave the name dhvan: to the sounds of speech that are heard. In 
the same manner other wise men, who knew the true essence of po- 
etry, have followed the example of the grammarians by giving the title 
dhvani to that verbal entity which contains a mixture of denotative and 
denoted elements and which is designated as “a poem.” They did so 
because of the similarity [to acoustical dhvani] in its being a manifestor 
[of suggested meanings just as the heard sounds manifest words]. Now 
this being the nature of dhvani, the range of which is immense when 
one counts up all the types and subtypes which we shall soon describe, 
its ilumination bears no comparison to a report on some mere individ- 
ual figure of speech that has hitherto remained unknown (cf. 1.1c A]. 
So the excitement of those whose minds are saturated with dhvani is 
quite within reason, nor should others exhibit toward them an intel- 
lect stained by jealousy. And so by this, those at least who deny the 
existence of dhvani stand refuted. 

L Was invented by men of knowledge. The compound 
vidvad-upojfià is a bahuvrihi modifying uktih, literally, “a term of which 
the upajnd or first use was by men of knowledge." Accordingly, the 
neuter gender demanded by Pan. 24.21 when upajñā is used in a 
tatpurusa is here inapplicable. 

The sounds of speech which are heard: According to the process 
[described in the Ny&ya-Vaisesika] it is the last sound of a chain of 
Sounds that enters the orifice of the ear. so the heard sounds are sounds 
born of sounds, (not the original sounds produced by the organs of 

170 (§ 1.131 L 

speech]. These sounds in form are like the reverberations of a bell,? 
and it is these sounds that are called dhvani. As the master Bhartrhari 
has said: 

Others have expressed tbe view that the sphota is boru from conjunction and 
disjunction with the organs of articulation; the dhvanayoh (plural) are the 
sound-born sounds. ( Vakyapediya 1.102)? 

In the same way, the suggested meaning has been called dhvani as it 
too is often characterized? by a reverberation analogous to the pulsa- 
tions of a bell. Again, the phonemes as heard, technically called náda- 
Sabdas, manifest the semantic unit, which we comprehend as soon as 
we cognize the final phoneme. These phoneme-manifestors are called 
dhvanis. As the same master says: 

The true form |i.e., the semantic content] in the word that is manifested by 
the dhvani is determined by a series of cognitions (viz., the cognitions of the 
successive phonemes], which are unnameable [that is to say, each phoneme- 
cognition in itself is unassignable to this word or that], but favorable to the 
final [word-identifying] cognition. ( Vakyapadiya 1.83) 

So we too use the term dhvani for the word and the [literal] sense which 
manifest [the suggested meaning]." 

Furthermore, it is in the varnas (the phonemes produced by the 
conjunction and disjunction of the vocal organs) that the differences of 
prosodical length (e.g., a and à, i and i) reside, as has been said: 
Either the mind does not perceive a sound if it is pronounced too softly, or it 
perceives clearly the whole phoneme. (e.g., we never hear half an a]. (Kumā- 
rila, Slokavdrttika, Sphotavdda, vs. 10) 

As these same differences are heard in the secondary sound, [or dhvani, 
that reaches the ear], it is only such other elements of the speaker’s 
speech, such as its slow or rapid delivery, elements over and above 
the well-defined operations of articulation [e.g., prosodical length, as- 
piration, closure, etc.) that are more particularly called dhvani As 
Bhartrhari® has said: 

The derivative dhvanis (ie., the sounds which reach the ear) after they have 
manifested the word continue to carry in themselves the variations in speed of 
utterance, but the nature of the phonemes? is not altered by them. ( Vakya- 
padiya 1.77) 

In like manner we [poeticians| apply the word dhvani to an opera- 
tion over and above the well known operations of denotation, sentence- 
meaning and secondary usage.'? Thus we have four senses of dhvani; 

§ 1.131 L] im 

and by combining them we may speak of a whole poem as dhvani'! It 
is for this reason that statements both of its difference (from poetry, as 
in “dhvani is the soul of poetry," 1.1 K] and of its identity (as in 1.13 
K, which Abhinava interprets as speaking of "poetry, as a distinctive 
type of literature, which the wise call dhvani"| are not improper. 

[Abhinava now gives a highly artificial analysis of the compounds 
vdcya-vdcaka-sammisra and $abdátmá, which we translated as a mix- 
ture of denotative and denoted entities and as verbal entity 
respectively.| Vàcya-vàcaka-sammisra is a compound from which the 
penultimate member has been dropped.!? It stands for vdcya-vdcaka- 
sahita-sammisro, "possessing mixtures as well as denotative and de- 
noted elements." The sense of addition is given without the use of 
"and," as in "ox, horse, man, beast." By this analysis one can see that 
the denoted sense can be called dhvani and the denoted word can be 
called dhvani, for both of them are suggestive inasmuch as both hint 
at (dhvanati) the suggested sense. As they form a mixture when they 
are combined with the vibhávas and anubhávas, the suggested sense 
also, (which consists of such a mixture] can be called dhvani for it 
is what is hinted at (dhvanyate). The word sabda in the compound 
$abdátma means sabdana, "a putting into words,” or verbal operation, 
and that not in the form of the denotative operation but in the form of 
the dtman, "the soul,” of poetry. This [suggestive operation] can also 
be called dhvani, for it is a hinting. Finally the object “which is des- 
ignated as a ‘poem’” can also be called dhvani because it is composed 
of the four other types of dhvani in the manner just described. 

And so he states the reason common (to both the grammatical and 
the poetical traditions for the use of the same term dhvani]: because of 
the similarity in its being a manifestor. The relation of manifestor 
and manifested,’ present in general in all four senses [of poetic dhvani], 
is common [to the term as used by the grammarians|. This is the 

It was objected (in 1.1 c A] that because the possibilities of speech are 
endless, [there may well be some small variety which has hitherto re- 
mained unknown and which might be called dAvani]. He answers that 
Objection with now this being the nature of dhvani, etc. The 
types which we shall soon describe: there are two main types. 
The subtypes: for example, the subtype "where the literal is shifted 
to another sense" (arthántarasarikrumitavácya) and that "where the 
literal is entirely set aside" (atyantatiraskrtavácya), both of these be- 
longing to the main type "where the literal meaning is not intended" 

172 [$1131 Z 

(avivaksitavácya);'* and the subtypes “where the suggested meaning is 
produced without apparent sequence” (asamlaksyakramavyarigya) and 
that “where the sequence is apparent" (samlaksyakramavyarigya), these 
belonging to the type “where the literal is intended but is subordinated 
to a second meaning” (vivoksitànyaparavácya).!* And even among these 
subtypes there are further divisions. The range of which is im- 
mense: the sense is that it covers the whole of poetry. By the word 
individual he indicates the restricted nature [of a figure of speech]; 
by the word mere, its being subordinate. Saturated (bhávita): that 
is, whose minds are intent on the nature of dhvani, or, it may mean, 
whose minds, by being perfumed with its charm, cause them to show 
such symptoms of emotions as closing their eyes (cf. 1.1c A). Those 
who deny the existence of dhvani: that is, those divided into all 
three categories [described in 1.1a, b, c A]. 

1. The Nyàya-Vaisesikas regard sound as a quality set up in the ether 
by the conjunction and disjunction of matter. This quality spreads from 
its place of origin through ether in all directions, as waves spread out from 
& stone dropped in a pond. Just as the wave which reaches the shore is 
wave-produced, not stone-produced, so the sounds of speech which reach the 
ear are sound-produced. 2. The basis for comparison is the succession of 
replicas through a span of time 3. The view, although here ascribed to 
others, is close to Bhartrhari's own. Note that the word sphota as used here 
is close to its etymological origin, "explosion." But Bhartrbari regards the 
sphota not simply as an inferrable physical fact, viz., the sound-explosion 
that is the ultimate source of the derivative sound that one hears, but as 
the accompanying metaphysical explosion of a phonemic pattern, a pattern 
that is devoid of the speed or slurring or variations in pitch and volume of 
the heard sound (dħvanı). Bhartrhari never took the further step, taken 
by his commentators, of identifying the sphota with the semantic content of 
the sounds. See S. D. Joshi, The Sphotanirnaya, pp. 33 and 54. — 4. Here 
Abhinava justifies the use of “dhvani” in sense d of 1.1 K, note 1, viz., as 
the vyarigya. 5. Note that Abhinava is careful to use the word upaloksita, 
not loksito. As will be shown in Chapter 2 (see 2.2 and 220), it is only 
the varieties of divans other than rosadhvan: that give this impression of 
bell-like reverberation, for in them a sequence is perceived by the auditor 
between the literal and the suggested sense. In rosadvani no such sequence 
appears. Accordingly, it would have been wrong, by using the word laksita, 
to characterize dhvan: as a whole by an impression of reverberation. The 
word upalaksita is more modest and may be used of a partial or temporary 
characteristic. A house may be upalaksita by a crow perched on its roof. 
6. Sphota is here used i its final, post-Bhartrhari sense as the sememe or 

§1.13m A] 

meaningful unit of speech. 7. Here Abhinava justifies the use of “dhvani” 
in senses a and 6 of 1.1 K, note 1, viz., as fabdárthau vyañjakau. 8. Sa eva 
does not here refer to the author of the previous verse. It is used in the manner 
of Irish dialect: “Himself told me,” referring to the master of the house, or 
the most important man in the speaker's frame of reference. 9. Sphota is 
here used in Bbartrbari's sense, not the later semantic sense. 

10. Abhinava here justifies the use of "dhvani" in sense c of 1.1 K, note 1, 
as the vydpdra, vyarijana. 11. Here, finally, sense e. One can only marvel at 
Abhinava's skill in justifying all the poetic uses of "dhvani" by grammatical 
precedent. — 12. Such madhyamapadalopin compounds are legitimized by 
P&n. 2.1.69, Vårt. 8. The anointed example is séka-parthiva “the vegetable- 
(eating) king." 13. Abhinava is forcing the text here; vyarijakatua does not 
mean vyafgyo-vyarjako-bhóva. 14, See 2.1 K. 15. See 2.2 K and note 1 


A There is such as thing as dħvani. And it is in general of two 
sorts: where the literal sense is not intended (avivaksitavácya) and 
where the literal is intended but is subordinated to a second meaning 
(viveksitényaparavdcya). An example of the first is this: 

Three men reap the earth 
of its flower of gold: 
the warrior, the man of learning, 
and he who knows how to serve. 
(Mahabhdrata 5.35.64)! 

And of the second, this: 

On what mountain, 
for how long, 
and what was the name of his austerity? 
I mean this little parrot's 
that he should bite into a cherry 
as pink as your lip? 

174 [$113m A 

1. The verse is also found in late versions of the Paricatantra: Bom- 
bay 1.45, Kosegarten 1.51. One may argue over the syntax of suvarnapuspóm. 
My strong feeling is that it is a karmadháraya, used karmany akathite (Pan. 
1.4.51; the verb ci is listed in the pariganana). The learned Kuppusvámi 
Sastri was of the same opinion ( Upolocana, p. 254). The gender of puspd may 
have been influenced by that of /atd or by its proximity to prthivi. Those who 
would make the Mahdbhdrata agree with Classical syntax will of course take 
the compound as a bahuvrihi. Abhinava has a more improbable explanation. 
The "fower of gold" here does not mean what it says. [t means success, 
worldly advancement. 2. Quoted also in SRK 439. Poetical use is often 
made of the belief that sensual pleasures are a reward for merit gained by 
ascetecism in past lives. See SRK, translation, note on vs. 408. According to 
Abhinava’s first explanation, the stanza is an instance of rasadhvani. 

L He states the result of refuting [the position of those who 
deny the existence of dhvani]: there is such a thing. Now, that 
dhvani should be no more than secondary usage (laksaná) [which was 
the fourth objection to the dhvaniváda set forth under 1.1d A] can be 
easily explained and answered only after giving examples. So despite 
the fact that one would expect straightaway a refutation of the view 
that dhvani is secondary usage [the fourth objection], or that it is 
indescribable [the fifth objection, cf. 1.1e A], the author of the Vrtti 
here sets forth the two main varieties of dhvani in order to be able 
to give examples. In doing so be follows [the system set forth by the 
Kánikós in| the next chapter:! and it is 

To fit the five meanings of “dhvani,” the grammatical agreement [of 
the compound avivaksite-vdcya with dhvani| can be achieved by taking 
it as a bahuvrihi which has simultaneous instrumental, locative, abla- 
tive, dative, and genitive relations (to its exocentric member].? When 
“dhvani” has the sense of vácya (the literal meaning), dhvani itself will 
be referred to by the element vdcya in the compound. The avivaksita- 
vácyo [dhvaniA] will be that type of dhvani (i.e., vācya, literal meaning) 
by which that literal meaning itself is not intended, that is, is not in- 
tended to predominate [in the final meaning]. Such a dhvani ill be a 
suggestor.? The same sort of analysis can be made of vivaksitényapara- 
vácya. Or, in the one option [of the five] where "dhvani" means artha 
(the literal meaning), we can analyse the compounds as karmadhérayas 
(simple adjective compounds). Avivaksitavácyah will then be an unin- 
tended literal meaning. Vivaksitányaparavácyah will be "a meaning 
which is literal and which is subordinated to something else which is 

§1.13mL] 175 

intended (to predominatel" In these two types, sometimes [viz., in 
avivaksitavácya] the literal sense is not intended for such reasons as 
that it makes no sense in the context; sometimes [viz.. in vivaksitànya- 
paravácya| it is intended insofar as it does make sense in the context, 
but by the power of its beauty it extends our apprehension to a sug- 
gested sense. It is on this account that in the second type a meaning 
is primarily the suggestor, in the first type a word. But is it not a 
contradiction to say that a meaning is intended and then say that it is 
subordinated to something else? No, because what is meant is that it 
is intended only insofar as it is subordinated to a second meaning. 

In general of two sorts: His view is that although there are three 
kinds of dhvani— vastudhvami, alarikáradhvani, and rasadhvani—they 
are included in these two sorts. But we may ask what benefit accrues, 
after giving the name dhvani, from adding these [particular] names [for 
its two main varieties]. The benefit is this. The first name indicates the 
cooperation in the operation called dhvanana (suggestion) of the ap- 
prehension on the part of the auditor of meanings implied by the three 
other operations of denotation, sentence-meaning, and secondary us- 
age, while the second name indicates the cooperation of what is wished 
to be said, that is, of the intention on the part of the poet. In this way 
the true nature of dhvani is rendered more apparent. 

Flower of gold: the compound suvarnapuspám |is an upapada com- 
pound agreeing with prthivi; it]* means “which flowers forth in gold 
pieces." As the sentence thus embodies an impossible meaning, the 
lieral sense must be unintended. Accordingly, after setting forth the 
literal sense of the words by denotative power (abhidhd) and giving 
us the syntax by the power of sentence-meaning (tátparyasakti), the 
stanza. abandoning this sense because of the obstacle (of impossibil- 
ity), gives us by the power of secondary usage (laksand) a meaning 
which is related by similarity, namely that the three men easily par- 
take of great wealth. The purpose of this secondary usage is that the 
praise of the warrior, man of knowledge and servant, which hides be- 
cause of its not being expressed literally, should rise to the highest value 
by being suggested, just as the breasts of a beautiful woman [are the 
more beautiful from being half hidden].? The primary manifestor of the 
suggestion here is the word [suvarnapuspám], but the literal meaning 
cooperates. All four semantic operations are in use. 

On what mountain: for not even such mountains as Sriparvata, 
which give unobstructed success of the highest order, furnish such suc- 
cess as this. The time spent there must have been measured in millenia 

176 (§1.13m L 

of divine aeons. And no form of austerity, such as sitting amid five 
fires,” has been recorded as being productive of such reward. The word 
tava has been given separately lest it be understood with weakened 
force, as it would be if placed in compound. His intention is to bring the 
action of biting into close relation with “you.” Accordingly, those who 
Say that the author failed to use the expression tvad-adhara-pátalar* 
simply because it would not fit into the meter are mistaken. 

Bite: (the suggestion is that) he tastes it in an unbroken continuity; 
[that] he does not eat it like a glutton, but rather acts like a connoisseur. 
Hence [it is suggested that] he has gained his refined taste, just like the 
other reward, from his austerity. Little parrot: From this indication 
of his youth, we see that he has gained his reward at the proper time. 
This also must derive from his austerity. We have in this stanza a 
suggestion made by a lover, which stimulates the (dlambana-|vibhdva 
li-e., the lady who is the object of his love}. It takes the form of a 
clever compliment which transmits his own hidden desire [viz., to kiss 
the lady's lip]. 

In this stanza there are only three semantic operations, viz., de- 
notation, sentence meaning, and suggestion, [as opposed to the four 
operations contained in the preceding stanza], for here the third, or 
middle-stage, operation of metonymy (laksaná) is missing as there is 
no blocking of the primary meaning, etc. Or, by a different interpreta- 
tion, we may say that the third operation, laksand, does intervene. We 
may suppose tbat the primary meaning is blocked by the impossibil- 
ity of taking literally these improbable questions [directed to a parrot] 
and that a secondary meaning [of the lover as obtaining the chance 
to taste his lady's lip], based on similarity, ensues. But the purpose 
of the laksaná is still that which is being suggested, which enters in 
the fourth stage. There is this difference, however, that in the former 
verse ("Three men reap the earth") laksand was the chief semantic 
power to cooperate with suggestion, while here the chief powers are 
denotation and the power of sentence-meaning, for it is because of the 
beauty of the literal sense that we apprehend the suggestion. This 
shows that the operation of loksaná is helpful only to a very small ex- 
tent. [n that variety of suggestion where the sequence is not noticed 
at all (asamlaksyakramavyarigya) [= rasadhvani] there is no apprehen- 
sion of laksaná at all, for the sequence [from the literal to suggested 
meaning] is not apparent, as we shall show. So we may take it that in 
the second type of dhvani also (at least as here exemplified] all forms 
of semantic operation are present. 

§1.13m ZL] 177 

1. It has been argued that the present passage, taken in connection with 
2.1 K and 2.2 K, shows that the author of the Vrtti and of the Kārikās was 
one person. The argument takes the form of a question: “Why would the 
author of the Kárikás define the subtypes of dhvani under 2.1 and 2.2 if he 
had not already described the major types to which they belong?" But the 
argument is not conclusive, and I would base my belief in a single authorship 
on other grounds (see Introduction, pp. 25-27). — 2. An English example 
may put the reader who is innocent of Sanskrit grammar on the right track. 
The phrase "cut-rate competition" can be glossed as "competition by which 
rates are cut." Sanskrit grammar would regard "cut-rate" as a baAuvrihi com- 
pound. It agrees grammatically with the word “competitiou.” Its exocentric 
(unexpressed) member is "which." Its relation to that member is instrumen- 
tal. 3. Vyarijoka here = vyarijako 'rthah. Abhinava does not trouble to 
give analyses of the other relations. The locative will apply when “dhvani” 
has the sense of kdvya; the ablative ("because of which"), when “dhvani” bas 
the sense of vyarigya (the suggested meaning); the dative ("for the purpose 
of"), when "dÀvani" bas the sense of vyarijana (the suggestive operation); 
the genitive, when "dAvoni" bas the sense of vdcaka (the denotative word). 
4. Apparently Abhinava takes suvarnapuspdm as suvarna plus the root pusp 
(DAP 4.15, puspyati) plus suffix on (Pán. 3.2.1). — 5. The same simile, spelled 
out in full, is used by the Pala poet Vallana (9th or 10th century), SRK 1705. 
6. A peak of the Western Ghats. For an interesting description of it some cen- 
turies later, see the account of the 15th-century Russian traveler Athanasius 
Nikitin, India in the 15th Century, Hakluyt Society, 1857. — 7. Sitting with 
a fire on each side and the sun above. 8. It is normally considered incorrect 
to use an expression in which a member outside a compound (here tava) must 
be construed with the subordinate member of the compound (bere adhara in 
6dhara-pátalam). Abhinava's remarks are designed to clear the author of the 
verse from the charge of incorrect usage. The meaning that Abhinava reads 
into the sentence would be rendered in English by emphasis: “that he bites a 
cherry as pink as your lip." Accordingly, the commentaries Kaumudi and BP 
say that if the word for you were here placed in compound, the author would 
be guilty of avimrstavidheyo, "not giving sufficient prominence to a predica- 
tive element." . Abhinava's reason for offering the second interpretation 
Ues in his general view of the pertinence of both stanzas quoted in this section 
by Ananda. He views them as introduced before the discussion of bhakti (= 
laksaná) in 1.14 in order to give examples where laksaná occurs but where it 
can be shown to be distinct from dhvani. It is true that avivaksitavdcya will 
always furnish such examples. But Abhinava wishes to show that there are 
cases even of vivaksitányaparavácya where a small degree of laksaná is possi- 
ble. Otherwise there would be no point in Ananda’s mentioning or illustrating 
that type of dhvani at this point. 

[5114 K 

K This dhvani is not indentical with bhakti (secondary opera- 
tion), because it differs from it in form; nor is it defined by that, because 
the definition would be both too wide and too narrow. 

A Here he refutes the objection that dhvani is nothing more than 
secondary operation (cf. 1.1 K, 1.1d A). This dhvani, that is, dhvani of 
the sorts just mentioned, is not identical with bhakti, because it differs 
from it in form. Dhvani is where a meaning other than the literal is 
revealed by the literal word and meaning to be the final sense and in 
that sense a suggested meaning is predominant. A secondary operation 
is merely a subordinate one. The second balf of the couplet shows that 
bhakti cannot define dhvani. Why cannot dhvani be defined by bhakti? 
Because the definition would be both too wide and too narrow. Too 
wide, because bhakti occurs in areas outside those of dhvani. Poets are 
often found to use words in an associated sense, being prompted by 
idiom or conformity (to convention], without any great beauty being 
generated by the suggestion. An example is: 

Wilting at either end 

from touch of heavy breasts and loins, 
green in between 

from a waist that bears no weight, 
and here all disordered 
from tossing about of loosened arms: 
this lotus-petal couch 
speaks the fever of a slender maid. 
(Harga, Ratndvali 2.12)? 
Or agai 

You kiss a hundred times, 

embrace each other a thousand 

and rest only to unite again. 

But when this happens with a lover, 
it is not tautology. 

$1.14 A] 
And again, 

Whether angry or pleased, 

in tears or in smiles. 

however you catch them, wanton girls 
carry away your heart.‘ 

With a newly flowered vine 

the husband gives his young wife a tap 
on her breast. 

Such a gentle tap; 

but the pain went straight to the heart 
of her fellow wives.* 

It suffers pressure for others' sake, is sweet when broken, 

even in altered state it is prized by all. 

What if it fails to grow if cast on barren land: 

is this the fault of the sugar cane or of the hostile desert? 
(Bhallatasataka 56)* 

In this last stanza the word "suffers" [is to be taken in a secondary 
sense]. Instances like those quoted above are never the domain of 
dhvani, because ... [sentence completed by 1.15 K}. 

1. Idiom and conformity to convention: this distinction between prasid- 
dhi and anurodho is made clear by Abhinava in 3.33i L (Text, p. 426), where 
Ananda uses these words in the dual. 2. Here it is the verb "speaks" that is 
used in a secondary or associated sense for "makes it clear that." Couches do 
not actually speak. [n the play, the lotus-couch furnishes the king with evi- 
dence of the heroine's love fever. The stanza is quoted in SRK 709 and in most 
of the great anthologies. — 3. "Tautology" is a trope for “tedious repitition." 
The action here is repetitive but not tedious. The verse is badly misprinted in 
the Kashi ed.; see Corrections of Text. The verbs cumbijjai, avarundhijjai, and 
ramijjai are passives used statically (bhàve yak): lit., "there is kissing,” etc. 
4. Here “carry away” is a trope for to charm or fascinate. I am not sure of the 
sense of ucchinta. Prakrit ucchitta means “crazy” according to Pdia-sadde- 
mahannovo. Abhinava renders it together with mahildo as svairinyaA, lit., 
women who disregard in the pursuit of their own inclinations the constraints 
of husbands or parents. The term is used as a synonym for abhisdrikah; see 
SRK 233. 5. According to Abhinava, "gives to" is here used in the sense of 
"rewards with." — 6. The verse is an aprastuteprosamsd (see 1.13j A, note 1), 
an allegory of a virtuous courtier who cannot live under a vicious king. The 

180 (81.14 A 

whole verse makes this suggestion, which is embodied in the figure of speech, 
as it is the literal meaning that is charming. But the verse is here quoted sim- 
ply for the secondary usage of "suffer." Sugar cane cannot suffer, so "suffers 
pressure" means no more than "is squeezed." The stanza is certeinly not by 
Bhallata, who lived under Sankaravarman (A.D. 883-902), that is, after the 
time of Ananda; see RájTar. 5.204 and Jacobi, ZDMG $6, p. 405. Bhallata's 
Sataka is an anthology drawn from many authors This stanza is ascribed to 
Indur3ja in Sárrig. 1052 and SuktiM. 35.5, to Yasas in SubhÁ. 947. 

L Andsoitisonly after giving examples of both (major types of 
dhvani| that he comments on and refutes the view expressed {in 1.1 K] 
by "some say it is bhákta (secondary operation)." What he has in mind 
is this. Are bhakti and dhvani identical, the two words being merely 
synonymns? Or does bhakti define dhvani, marking it off from what is 
other than dhvani as prthivitva (earthness) marks off from earth what 
is other than earth?! Or is it an upalaksana (occasional characteris- 
tic), as a crow may be of Devadatta's house because a crow may be 
perched on it? (The Küriká) begins by denying the first possibility: 
not identical with bhakti. The (Vrtti's] phrase, of the sort just 
mentioned, should be referred to the five meanings of dhvani: word, 
litera] meaning, operation, suggested meaning, and poem.? To show 
the difference (of the two concepts] he begins by stating the nature 
of dhvani: (where a meaning other than] the literal, etc. To 
be the final sense: i.e., the sense that permits our apprehension to 
come to rest, the sense which is the purpose [of the poet's employing 
just those particular words]. Is revealed: he means, is suggested. 
Merely a subordinate operation (upacára): that is, secondary us- 
age (gunavrtti), metonomy (laksaná). To use an associated sense is to 
speak in transgression (of the literal sense]. By the word "merely" he 
is referring to cases where a fourth semantic operation, of the nature 
of suggestion, may be inferred as possible from the mere fact that the 
third operation has been employed,’ but where this fourth operation, 
being neither useful nor important, is as good as absent. For the defi- 
nition of a [primary] purpose is "that with a view to [the acquiring or 
avoiding of] which one starts about an action."* Inasmuch as secondary _ 
usage is found even in such cases, how can one say tbat the suggestive 
operation and the operation of dhvani are one and the same thing? 

[The Karika} then rules out the second possibility: because the 
definition would be too wide, etc. Nor is it, viz., dhvani, defined 
by that, viz., by secondary operation. It may be objected that the 

$114 L] 

suggestive operation must take place [wherever a secondary operation is 
employed], so how can one distinguish the area [of dhvani from that of 
laksand)? To provide against this, [the Vrtti| says: without any great 
beauty. He means, for the reason that the purpose in question is of no 
importance and nothing is therefore accomplished by the suggestion of 
it. By the use of the word "great" [he admits that there is some beauty 
in such cases but that] there is only enough to qualify as a guna (poetic 
quality), as shown by the definition [of the quality, aptness (samádhi)]: 
“the imposition of the property [of one thing] on some [other] thing 
should be called aptness."* 

But if there is no real purpose in using such expressions, how is it that 
an author uses them? He answers this objection: "because of prasiddhy- 
anurodha," for such has been the usage through a long succession [of 
writers]. But I would say that prosiddhi implies [not that there is no 
purpose but] that the purpose is too obvious. Although the purpose 
appears in clear form, it stands in need of something hidden,” like 
buried treasure [in order to be in the domain of dhvani]. For in the case 
of the secondary usage (upacdra) illustrated by "speaks" [in the first 
of Ánanda's illustrative verses] the purpose is the apprehension by the 
auditor of “making something quite clear." If the author had used the 
obvious literal expression, would the verse lose an element of beauty? 
Or by his using this covert expression is any element of beauty added? 
With this in mind our author will say [that dhvani reveals] “a beauty 
which cannot be conveyed by any other form of expression" (1.15 K). 

[In the second verse quoted by Ananda] the Prakrit word avarun- 
dhijjai means dlirigate (there is embracing). "Tautology" is used in its 
secondary sense of something to be avoided, because the literal meaning 
[viz., repetition of words} is impossible in the context. 

[For the third of the quoted verses Abhinava furnishes a Sanskrit 
translation of the Prakrit and then adds:] In this verse "catch" is used 
in the secondary sense of receive, "carry away" in the sense of subjecting 
to their power. 

[The fourth verse:] Here a husband in the course of normal love- play 
gives a light stroke of a fresh vine to the breast of bis youngest wife 
As this indicates the favor she has attained with him, it proves from 
its very softness bard to bear in the heart of her fellow wives, who no 
longer share in his love-play. That a light stroke given to one person 
should have its effect on another and be hard to bear “even although so 
soft,” is startling. In this stanza giving is used in the secondary sense 
of rewarding 

182 [$1.14 L 

[The fifth verse:} Although the word "suffers" is used in its primary 
sense with respect to the virtuous man, who is the subject really in- 
tended, still. with respect to the extraneous subject, sugar cane, which 
is being praised, as the “suffering” of pressure is impossible, the word 
is used in a secondary sense of undergoing pressure, so that the whole 
expression ends up meaning nothing more than "is squeezed." 

Now in such cases of course there is some purpose, so how can we 
avoid saying that there is dhvani? It is in response to such an objection 
that our author says: Instances like those, etc. 

1. prthivi prthivituasambandhat, Prasastapáda, Benares ed., p. 41. This 
so-called definition of earth is justified by Udayana in his Kiranávali. He 
points out that it is not intended to be definitive. We shall still require 
à definition of prihivitvo, which he gives. But such definitions, he insists, do 
not lead to an infinite regress, in logic any more than in medicine or grammar. 
One goes on defining until the uneducated man is educated. Then one stops. 
2. This interpretation is historically impossible. Ánanda never distinguishes 
the five seoses in this way. Ananda’s phrase actually refers to the two major 
types of dAvoni that he has just mentioned: avivaksitavácya and vivaksitónya- 
porovácya. 3. Every use of laksaná implies some purpose, for otherwise the 
speaker would not have departed from literal usage. Thus even in the first 
of Ànanda's quotations (the verse from the Ratnávali) one may say that the 
word vadati is used instead of gamayati in order to give the notion that the 
couch shows very clearly the lady's suffering. But this is a secondary sort 
of purpose, not the main purpose of the sentence. It is not really useful to 
that purpose nor poetically important. —4. Nyáyo S. 1.1.24. Harsa did not 
set about writing his verse with tbe view to expressing only a little more 
vividly the action of the couch on the king's inferential judgment. 5. The 
source of this definition of samadhi appears to be lost. For other, and better, 
definitions of the term, see BANS 16.102 and Dandin 1.93. Dandin's example 
is: kumudéni nimilanti kamolány unmisanti ca, "the water-lilies close their 
eyes and the lotuses open theirs." The imposition must be within the range 
of everyday speech, or we would have a figure such as hyperbole or fancy. It 
must be apt and should impart vividness. 6. Abhinava seems to take the 
compound prasiddhyanurodha here as a tatpurusa. Kuppuswami Sastri in his 
Upalocana on the Kaumudi remarks that Abhinava's explanation here does 
not fit with the use of prasiddhyanurodha in 3.33i A. 7. Kaumudi and BP 
explain the hidden something as the transmission of an element of beauty that 
does not appear in cases of non-suggestive use. 

$145 L] 

K A word can justly be termed dhvani only if in its being sug- 
gestive it lights up a beauty which could be achieved by no other [i.e.. 
non-suggestive manner of] expression.! 

1. [n the bracketed words of our translation we are following Abhinava's 
interpretation. Without them the sentence might seem to be limiting dAvani 
to the Flaubertian mot juste, which is surely not the case. 

A In the area of the examples just given there was no word 
that caused the manifestation of a beauty which could not have been 
achieved by some other [manner of] expression. 

L By no other expression: that is, by the use of literal word 
and meaning, and so other than dhvant. "Word" is to be understood 
in all its five senses. Can be justly termed dhvani: i.e., can be de- 
noted by the word "dhvani" The examples just given: e.g., "speaks" 
[in the first verse quoted under 1.14 A]. 

1. Such is certainly not Ánanda's intention, but the interpretation makes 
for a more systematic presentation. The Kaumudi, followed by BP. shows that 
Sabda may be analysed so as to give the same five senses given by Abhinava in 
his analysis of dhvani (1.13 L). Thus: óabdyata iti sabdah = vdcyah; fabdyate 
‘neneti Jabdoh = vàcakoh; sabdyate vyajyata iti = vyarigyam (here I suppose 
one must employ the principle anekarthd dhátavah); fabdanam = vydparah; 
while the combination of these four senses will give the samudáyah = kávyam. 

(§ 1.16 K 

A Furthermore, 

K Words such as lávanya, which are used idiomatically in a sense 
other than their proper (etymological) sense, are never instances of 

A In these words there is indeed a secondary semantic operation. 
And in the context where they occur we may speak of dhuani, if it bap- 
pens to arise, but that will be only because of some other consideration. 
Dhvoni never occurs primarily because of such words. 

L Our author has already said that there is no operation of 
dhvani where the purpose [of choosing to employ a secondary sense| 
is unimportant. Where there is secondary usage without any basic 
purpose at all, there too there will naturally be no operation of dhvani. 
He states this by saying, furthermore, etc. 

Words like lávanya, which means properly “possessing a salty taste," 
but which is idiomatically used in such other senses as "charm, beauty," 
by the very fact that they are used idiomatically lack that separation 
(from their proper meaning] which is occasioned by the presence of the 
triad! [and so cannot give rise to dhuani]. As has been said, “Some cases 
of secondary usage, being idiomatic, so far as their force is concerned 
are just like direct denotation” ( Tantravárttika, p. 683). Such words, 
although used in a sense different from their etymological sense, do not 
carry any dhvani and we cannot speak of dhvani in such cases. By 
speaking of secondary semantic operation he refers to secondary usage 
of both the gauni and làksaniki varieties." By saying "such as" ldvanya, 
[the Karika] would include words like dnulomya, prátikülya, sabrahma- 
cérin. Anulomya (smoothness, orderliness) means literally rubbing in 
the direction in which the hair grows. Pratikila (antagonistic) is used 
properly of a current that fights against the bank of the stream. The 

$117 K | 185 

primary meaning of sabrahmacdrin (companion) is one who bas had 
the same teacher. In each case the other meaning is secondary. As one 
does not adopt the secondary use of these words with any purpose in 
mind, we cannot speak of a suggestive operation in their case. But now 
in such verses as: 

de vaditi lundhi paludisi gasittha 

làvannujjala-gugharidhollavapattà (7)? 
a suggestion is indeed apprehended in the presence of the word lávanya. 
True, but it does not derive from the word làvanya. It derives rather 
from the operation of dhvani that follows after we have understood 
the meaning of the whole sentence, for in the sentence it is suggested 
that his beloved's face has illuminated the whole sky. But enough of 
many words. Our author says the same thing: because of some 
other consideration. He means, because of the suggestive power [of 
the sentence], not because of the use of such a word as lávanya in à 
secondary sense. 

1, The triad is: blocking of the literal sense, connection of the literal 
object with the secondary object, and a purpose (prayojana) for shifting from 
the use of one sense to the other. Actually it is only the third of these 
conditions that the word lávanyo, as used idiomatically, lacks. 2. Govni is 
based on the simil ity of the primary and secondary object, láksaniki is based 
on some other relation subsisting between them; see 1.1 K, note 1. 3. No 
one to our knowledge has been able to make sense of any of the versions of the 

,Prakrit verse. Each manuscript shows a different reading, all of them being 
mostly jibberish. After the first two syllables | have transcribed the readings 
of Kaumudi’s MS ca, in which at least the words lévannuyjala, “resplendent 
with beauty,” make sense. 

A And agai 

K When a word abandons its primary operation and reveals an 

Object by secondary usage, the purpose for which this is done is one to 
which the word moves without stumbling. 


A Because when a purpose is to be achieved of revealing a mean- 
ing of special beauty, if the word accomplished that purpose only 
through a non-primary sense, the author would be at fault in using 
it. But that is not the case. 

1. That is, he should have used some other word. What is meant be- 
comes clearer by the help of an example. “A village on the Ganges" suggests 
the beauty, peacefulness, and holiness of the village. These suggestions spring 
from the primary sense of the word "Ganges," not from the secondary, or 
shifted, sense of "bank," which we need in order to make sense of the expres- 
sion. It is logic that demands the secondary sense (see 1 4b L, note 6). The 
suggestion, the poetry, springs directly from the primary. 

L The foregoing argument has shown it wrong to say that wher- 
ever there is bhakti (associated usage) there is dhvani. And so, if we 
use bhakti to define dhvani, the definition will be too wide. But even if 
we were to grant, for the sake of argument, that bhakti occurs wherever 
there is dhvani, the object on which the bhakti operates will be different 
from the object on which dhvani operates. There can be no relation of 
substance and attribute between entities tbat occupy different areas; 
and a definition must be an attribute.’ Laksana (bhakti) operates on 
a secondary object. The operation of dhvani takes place in the area 
of the purpose. The second semantic operation, laksand, does not take 
place in that area, because that area lacks the set of conditions (block- 
age, similarity, etc.) for secondary usage. lt is with this in mind that 
our author says, and again, etc 

[Comment on the Karikà:] When a word abandons its pri- 
mary operation, that is, after completing its denotative operation, 
and reveals, that is, brings to our apprehension, a secondary ob- 
ject by secondary use (gunavrtti)—here gunavrtti is a synonym of 
laksanà—the purpose, or goal, at which—accusative case—this il- 
lumination is aimed, is one where another operation comes into play. 
And this operation is not laksaná, because the laksanà of a word is an 
operation which moves baltingly, that is, the word's power of giving 
information is disturbed by the working of some hindrance, whereas 
that word in giving us to understand the purpose does not meet with 
any hindrance. For if it did, we should have to discover a reason for 
it, [which could only be) some further purpose, which we should have 

$117 L} 187 

to discover; and so we should be forced into an infinite regress. Ac- 
cordingly, there is no place here for laksita-laksand (secondary usage 
growing out of secondary usage). To reveal: the word darsanam is a 
form that includes the causative suffix [i.e., the meaning is “revelation,” 
not "sight"]. 

[Comment on the Vrtti] Accomplish here means “suggest.” 

Through a non-primary sense: i.e, as disturbed by some bin- 
drance. In using it, viz., the word. Would be at fault: It is so 
that the purpose may be easily apprehended that a word is used in 
its secondary sense. In the expression “the boy is a lion,” where the 
notion of the boy’s unusual bravery is to be conveyed, if the word were 
to operate haltingly, it would not convey to us this notion; so why 
should it have been used? If you reply that it will convey that notion 
by a [further] secondary operation, then we shall have to discover a 
subsequent purpose and still another secondary operation to go with 
it. So we shall be led to an infinite regress. On the other hand, if you 
admit tbat there is no halting motion here. then there can be no sec- 
ondary operation prompted by the purpose to be conveyed. for the set 
of conditions for a secondary operation (hindrance, etc.,] will be absent. 
You cannot say that there is no operation [for conveying the purpose]; 
and that operation cannot be denotation, for the conventional agree- 
ment [between denoter and denoted] is absent. This operation, over 
and above the operations of denotation and secondary usage, can only 
be the operation of dhvani. 
* But that is not the case: The author is not at fault, because the 
purpose is readily apprehended. So we see that the denotative power, 
wishing to pass to its primary meaning but being blocked by some 
hindrance, continues on because it has not fulfilled its aim. Tbat is why, 
in speaking (of the associated sense which it does reach], we say, “This 
is the non-primary sense of the word.” As there is an acceptance of 
convention even in the non-primary sense, {one may say that] secondary 
usage (lekgand) is simply an appendage to denotation (abhidha).? 

1. To define a substance is to furnish its peculiar attribute. that is, the 
attribute which rules out all other substances. If the attribute occurs in 
a different area from that of the substance, it obviously cannot serve as a 
definition. — 2. Such is the correct reading furnished by the Kaumudi. The 
other printed texts read laksana- for laksita-. See 1.4b L and note 18. BP gives 
an example to which the term laksito-laksaná may be properly applied: "How 
is this, slender maid! The month of Sravana is in your eyes, autumn in your 
cheeks, summer in your limbs and winter in your lotus face." Here the month 

188 [$1.17 L 

of $t&vana' is used in the sense of the rainy season, which in turn gives rise to 
the second secondary sense of tears. See also 2.9 L and note 2. — 3. These 
tather odd remarks are occasioned by Abhinava's effort to clarify the following 
Kürikà. Kārikā 1.18 will say that secondary usage (gunavrtti) depends on no 
more than vácokatva (the denotative power of words), a statement that seems 
at variance with the semantic system that has been described, where laksand 
(as likewise gunavrtti) is an independent semantic power which depends on a 
triad of conditions. Abhinava furnishes us here with two considerations which 
may explain the apparent change of viewpoint. Loksaná arrives at its object, 
the secondary sense, only at the end of a journey which set out toward the 
denoted object. In that sense it occurs as an appendage to abhidhà. Then 
also, it shares in the peculiar property of abhidhà, the convention (samaya, 
sariketa) between word and meaning. Laksand does not veer aside from the 
primary sense to any meaning, but only to a secondary meaning that also 
attaches to the word by convention. One may find many of the secondary 
senses of a word in a dictionary. In this sense too laksaná is closely related 
to abhidhd. Dhvani, on the other hand, although it requires denotative word 
and meaning as its trigger, is not dependent on, or closely related to, abhidAa 
in the two respects here brought to our notice. 

K Secondary usage (gunavrtti) depends on no more than the 
denotative power of words. How can it be used to define dhvani, which 
is based wholly on suggestive power?! 

1. The distinction is expressed here unclearly. Both gunavrtti and dhvani 
depend to a greater or lesser extent on vàácakatva. The distinction lies in 
the manner in which they depend on it. Abhinava does much to clarify the 
passage. See 1.17 L, last paragraph and note 3, and his comment below. 

A Therefore dhvani is one thing and secondary usage another. 
As a definition it would be too narrow too, because the type of dhvani 
where tbe literal is intended but is subordinated to a second meaning, 

$1181] 189 

and many other varieties, do not f | in the same area with associated 
meaning (bhakti). Therefore associated meaning furnishes no defini. 

L He sums up: therefore. Since secondary usage (laksang) 
forms simply an appendage to denotation, for that reason; that is to 
Say, Since secondary usage (gunaurtti)—he means both its varieties, 
gauni and láksaniki—depends on denotative operation inasmuch as it 
arises from an obstruction to that operation and forms an appendage 
to it, how can secondary usage form a definition of dhvani, which is a 
suggestive operation? The two processes occur in different areas. 

[The Vrtti] sums up the matter: therefore. The author means, 
because he has shown its forming too broad a definition and apropos of 
that discussion (has noted] that it occupies a different area. So, havi 
explained the overextension referred to in the Káriká which stated "it is 
not defined by that, because the definition would be both too wide and 
too narrow (1.14 K)," he now explains the underextension: it would 
be too narrow. 

It would be: that is, secondary usage would be. The definition 
would be of sufficient extension only if bhakti (associated usage) oc. 
curred wherever dhvani occurs. And that is not the case, for while 
bhakts occurs in the presence of that type of dhvani where the litera) 
sense is not intended, as in such verses as “Three men reap the earth” 
(1.13m A), how shall we find it in such verses as "On what mountain” 

[Objection from the Mimámsá point of view:] But now, [let us exam. 
ine the concept of] /akgand (secondary usage); it extends throughout 
the qualitative (gauna) as well [as the relational (làksanika)]. The only 
difference! between the qualitative and relational varieties is that in 
the qualitative the word which indicates an object by laksaná enjoys 
grammatical agreement with the word [which denotes that object], as 
in "The boy is a lion."? Or, we may say that one object (e.g. a lion], 
by indicating a second object [e.g., a boy who shares in the qualities 
of a lion], makes the word that denotes the second object agree with 
the word that denotes itself (i.e., makes the word “boy” agree gram- 
matically with the word "lion"]. Or, we may say that the word “lion” 
and the object lion cooperate in indicating; and so both of them fuse 
with the word denoting that thing and with the thing itself. It has 

190 (§ 1.18 L 

been said that “the [denotative] word is used in the qualitative vari- 
ety, but is not used in the relational variety (laksana). But secondary 
usage (laksaná)? is present there too and so extends to all varieties. 
And this relational variety is of five sorts,‘ as it is based on (a) a 
conjunction (of the secondary object] with the direct object: for exam- 
ple, a direct object of the word “dvirepha” (“possessing two r's") is the 
word “bhramara”—now the word "bhramara" is connected [by the con- 
ventions of language] with the six-legged creature, a bee; that object 
may be indicated by the word "dvirepha" because of the bee's connec- 
tion with a direct object [of the word "dvirepha"]; (b) the proxi ity 
(sámipya) of the secondary object [to the direct object], as in "a village 
on the Ganges" [where the bank of the river is close to the direct object, 
river]; (c) samaváyo, that is, a connection;? for example, "Bring in the 
spears," meaning bring in the spearmen, (who are connected with the 
spears|; (d) opposition, as when one says with reference to an enemy, 
“In what has he not benefitted me?";* (e) its being connected with 
the same activity, that is, when it is based on a relation of cause and 
effect: for example, *he takes away my life," when the expression is 
used of one who takes away one's food [here the direct object, life, is 
the effect of the secondary object, food]. In this way, laksaná occurs in 
all varieties. 

And thus, in your verse "On what mountain," since an obstruction to 
the literal sense is introduced by the improbable questions (addressed 
to a parrot], laksana, based on similarity, comes into play. 

(Abhinava:] But we admitted this, saying that laksaná occurred i 
the middle stage [between denotation and suggestion]. 

[MImámsaka:] Then how can your author say that “the type of dhvani 
where the literal is intended but is subordinated to a second meaning, 
[and many other varieties, do not fall in the area of laksana]"? 

[Abhinava:] He was referring to the chief subtype (of this type of 
dhvani], viz., where a succession from literal to suggested meaning is 
not noticed. And by its many varieties he meant the suggestion of rasa, 
of bhava, of improper rasa, of improper bháva, or of the incipience or 
cessation of any of these;" all this, as well as the subvarieties of each.* 
And in every one of these, laksana is impossible.” The reason is that 
in a poem that sets forth the vibhávas and anubhávas there can be no 
hindrance to the primary meaning; so what chance is there for laksaná 
to arise? 

[M:] Let us forget about hindrance. The true definition of laksaná 
is this: "That is called laksaná where we have an apprehension of an 

§1.18 L] 191 

object invariably connected with the primary meaning.”!° In the type 
of poetry here of which you are speaking, when the vibhávas, the anu- 
bhávas, etc., ate the primary meanings (of the words], the rasas, etc., 
being invariably connected with them, will be indicated by laksaná, for 
the vibhávas and anubhdvas stand in a cause-effect relation to the rasas, 
while the vyabhicdribhdvas stand in a relation of accompaniment.!! 

A:] By no means. At this rate, whenever we apprehend the object 
smoke from the word "smoke," we would be reminded by laksaná of 
fire, and from that we would be reminded of the repelling of cold, and 
so on, until there would be no end to a word's meanings. 

M:] The word "smoke" finds a point of resti its own literal 

so it will not go on operating any further. 

A:] Here you have come around to our view, that it is a hindrance 
to the literal meaning that brings laksanà to life, for when such a 
hindrance occurs, the word does not rest in its literal meaning. And in 
the setting forth of the vibhàvas, etc., there is no hindrance. 

M:] Very well then, we may say that just as the memory of fire arises 
after one has understood the presence of smoke, so the apprehension 
of such mental states as sexual desire arises after the apprehension of 
the vibhávas, etc. There is no verbal operation here at all.!? 

A:] This Mimàmsaka who knows so much about our apprehension'? 
should be asked this question. Do you suppose that the apprehension 
of rasa is nothing more than the apprehension of another person's men- 
tal state? You should not make such an error. What aesthetic relish 
(rasatà) would there be in the mere inference of emotional states that 
are found in the everyday world? The relishing of rasa is a super- 
normal (alaukika) delight. It consists in savoring the vibhàvas, etc., 
which are found in poetry, and it must not be degraded to the level 
of memory and inference, or the like. Rather, the connoisseur, whose 
mind has been trained by everyday inferences from effect to cause, 
when he apprehends the vibhávas, etc., does so not in an uninvolved 
manner (tétasthya) [in which everyday inferences are achieved], but by 
bringing them into the power of his sensibility, or in other words, con- 
fronting them with sympathy, a process that forms the seedling for the 
full relishing of rasa about to ensue. The nature of his apprehension 
consists in a savoring of the vibhàvas that springs from his identifying 
with them, a process that is entirely removed from the path of mem- 
ory and inference. This savoring has not been produced by some other 
means of knowledge in the past so that it might qualify as memory.'* 
Nor is it produced now from any other means of knowledge, because 

192 [5118 L 

perception and the other means do not operate in a super-normal ex- 
perience. That is why the terms vibhávo, [anubháva,| etc., are used, 
which are terms for the super-normal.'5 As [Bharata] says: "It is called 
a vibhdva because its purpose is to give understanding (vijria)."!5 In 
dealing with normal experience we speak of a cause, not a vibháva.'" 
Anubhdva also is a term for the super-normal. "The dramatic repre- 
sentation of voice, body, and emotional expression, since it per its us 
to experience (anubhávayati) [the represented mental states], is called 
anubháva."'* By experiencing (anubhavana) is meant one's identifica- 
tion with the mental states [so represented]. In dealing with normal 
experience we speak of an effect, not an anubháva.!? And so, lest one 
suppose that [in the enjoyment of rasa] the mental state of another 
person is being inferred, Bharata omitted the word “basic emotion” 
(sthéytbhdva) from his definition: "Rasa is born from the combination 
of vibhávas, anubhdvas, and vyabhicáribhávas."?? Had it been included 
it would have proved a stumbling block. 

The statement that a basic emotion turns into a rasa?! is based on 
suitability, that is, because the relishing of beauty arises in us from 
our memory bank (samskdra) of mental states which are suitable to 
the vibhdvas and anubhdvas of those basic emotions [that are being 
portrayed in the characters of a literary work]; and because, while we 
are in the stage of understanding the mental state of another person, 
which may prove helpful to our sympathetic response, we understand 
such basic emotions as sexual desire from such (vibhávas] as a garden 
and such [anubhávas] as horripilation of the actor's skin. The tran- 
sient emotion (vyabhicáribháva) is also a mental state [resident in the 
portrayed character], but since it is relished only as dependent on a 
principal mental state, [there is no danger in its being mentioned in 
the sūtra and so] it is listed along with the vibhávas and anubhávas. 

So what is born here is a rosyamánatà (a being tasted, a gustation, 
of beauty),”? that is, a savoring that eclipses such worldly mental states 
as the joy that might be produced by reunion with a constant stream 
of old friends. And for this reason, [viz., because of its super-normal 
character,] the savoring serves to manifest something, not to inform one 
of something, as might be done by an established means of knowledge 
(pramána). It is not a production such as results from the working of 
a cause. 

(M:] But if it is not a cognition and is not produced, what is it? 
[A:] We have told you. It is the super-normal rasa. 

§ 1.18 L] 193 

[M:] But the vibhdvas, etc.—are they what inform us of it, or are 
they productive of it? 

[A:] Neitber informative nor productive; 

(M:] Where else is such a thing found? 

[A:] The fact that there exists nothing else of this sort is why we have 
called it super-normal. 

[M:] But then rosa will not be a valid means of knowledge. 

[A:] So be it. Since we receive pleasure and instruction from savori 
it, what else do you want? 

[M:] But still, there will be no way of knowing that it exists. 

[A:] Wrong. It is proved by our own self-awareness, because savoring 
is a form of knowledge. 

Enough arguing now. But one more reason why rosa is super-normal: 
namely, that soft and harsh alliterations, while they have no effect on 
meaning, can be suggestive of rasa.?? What possibility of laksand can 
there be in such a case? One finds too that there can be relishing of 
the words of a poem by mulling them over, for we see a connoisseur 
reading the same poem over and over and savoring it. That the words 
of such a poem are unlike other words, which become useless after we 
have understood them according to the dictum, "One takes them up [as 
means] only to abandon them [after they have served their purpose],"?* 
shows that such words must have a suggestive operation. And that is 
why the succession [of the suggestion to the primary meaning] passes 

It has been objected by a certain person that this theory of sugges- 
tion would result in vàkyabheda (giving two different meanings to one 
sentence),?* but his objection is based on a misconception because [of 
the following consideration]. 

A ritual text (sdstra),?® uttered once, conveys a meaning by means 
of semantic convention. As it is impossible for us to remember simul- 
taneously numerous contradictory conventions, the text cannot convey 
two meanings, while if the conventions are not contradictory, then ob- 
viously there will be just one sentence meaning. The sentence cannot 
convey contradictory meanings successively, because a semantic oper- 
ation cannot begin again after it has once ceased. If the sentence is 
uttered a second time, the meaning will be the same, because there has 
been no change in the semantic conventions or the context. If it could 
Convey a second meaning, obliterating the meaning obtained by con- 
text and convention, there would be no semantic rules at all. At this 

194 [5118 Z 

rate vv hen we hear the words, "A man desirous of heaven must perform 
the fire oblation,” what assurance should we have that the meaning is 
not that a man must eat dog meat?" Nor would matters stop there. 
Language would become wholly unreliable. Accordingly, vakyabheda is 
there counted as a fault. 

But in poetry, the vibhávas, etc., as they are being conveyed to us, 
tend to become objects of our relish. There is no need tben (or semantic 
convention." There is nothing here comparable to the apprehensions 
we receive from a ritual text: “I am enjoined to do this; I will do it; I 
have accomplished my purpose," since such a text, being on a normal 
plane,?? tends toward what one should do in the future. But in poetry, 
the savoring of the vibhàvas, etc., is essentially a matter of the present, 
arising like a magic flower, without reference to past or future. In this 
sense the relishing of rasa is different from normal relishing, as it is 
also different from a yogi's meditation. 

And so it is that in the stanza “On what mountain.” connoisseurs will 
sense the clever compliment which is the speaker's intention without 
noticing the succession to a blocking of the primary meaning. That is 
the [true] reason why our author has stated as a general principle that 
there is no bhakti in that type of dhvani where the literal meaning is 
intended but is subordinated to a second meaning. It was we, seeking 
to persuade you when you were being obstinate,” who granted that 
laksoná might be found in the stanza. for it was our thought that no 
matter how angry you might be, you could say nothing against cases 
where no succession could be noticed. But if you will not be angry, 
we will point out that even in the stanza "Three men reap the earth,” 
where the literal sense is not intended, regardless of the fact that all 
the conditions for laksané, such as blocking of the primary meaning, 
are present, the sentence comes to rest in the suggested meaning. And 
so enough of tbis long discussion 

Our author sums up: therefore associated meaning [furnishes no 

1. In the text, kevalam construes with the words ity evam laksanikdd 
gunasya bhedoh four lines below. 2. "Lion" is here used by secondary usage 
to mean a boy with the qualities of a lion. It is accompanied by the denotative 
word "boy," with which it stands in grammatical agreement. [n the relational 
variety, on the other hand, this accompanying use of the denotative word is 
not found. "A village on the Ganges" indicates a village on the river bank, 
but the denotative word "bank" is not used. 3. It is confusing to have the 
same word (loksand) used in two different senses in the same sentence. To 

8118 L] 195 

make matters worse for the reader, the pronoun (sa) with which the next 
sentence begins must refer to the first, not the second sense. 4. Abhinava 
now specifies the five sorts, following the verse of Bharirmitra which he quoted 
under 1.1d L. But he here follows a different reading from that whicb he there 
gave: abhidheyena samyogàt sámipyót samaváyatoh in place of abhidheyena 
sdmipydt sárüpyàt samaváyatah. See l.1d L, note 4. In effect the version 
given under 1.1d L includes the gauni variety, that being the variety based 
on sórüpyo, whereas the present version excludes it. The sense bere given 
to samyoga is peculiar; in Nyàya a vácyavácakasambandha is never referred 
to as samyoga. Furthermore, the illustration of the type based on samaváya 
implies a non-Nyàya sense of sarnaváya. But these facts cannot be used as 
arguments against the present version, for Bhartrmitra may have defined the 
words in other than their Nyáya-Vai$ ika senses. $. Note that semaudya 
and semyoga are here glossed by the same general word. In its normal sense 
samaváya cannot be used of the relation between spears and the men who hold 
them. 6. This is a case of irony. The direct sense of the word “benefitted” 
is the opposite of the secondary sense which the speaker intends. 7. For 
these varieties see 2.3 K and for examples, 2.3 L. 8. Rasadhvani may be 
divided into the eight, or according to some, nine, rusas. Even single rosas 
have their subvarieties; thus, sambAoga-srrigáro and vipralambha-srrigára. So 
also with the bhdvas and rasábhásas, etc. 9. The text from this point, Kashi 
ed., p. 154, line 4. to p. 158, Line 7, has been translated by Gnoli. Aesthetic 
Ezperience, pp. 102-106. 

10. The quotation. given inexactly, is from the Tantravdrtttka, p. 354 
The original reads abhidheyavinábhüte praurttir laksonesyate, "the operation 
of a word on an object invariably connected with its direct meaning sbould 
be called latsand” It is not a good definition, because the connection need 
not be invariable. — 11. The vibhávas are the determinants of the rasas; 
they cause them to arise. The onubhávos result from tbe rasas (at least 
as rasa is understood by the Mimámsaka here speaking, viz., as a sharp- 
ened emotion). The vyabhicdribhdvas accompany the rasas. 12. Here the 
Mimàmsaka would reduce the achievement of rasa to an inference, the third 
valid means of knowledge. The fourth means, verbal apprehension, would play 
no part. 13. The words are ironic. Among poeticians the Vedic ritualists 
are stigmatized as the most insensitive of all men to poetic beauty and to the 
understanding of literature. 14. This is another distinction between rasa- 
pratiti and inference (anumdna). In inference the sight of smoke gives rise 
to a memory of fire which has been perceived in the past as accompanying 
smoke. 1$. alaukika eva vyavahdra: that is, they refer to a super-normal 
means of apprehension. Abhinava returns to this idea in commenting on 2.4 
The apprehension (pratiti) of rasa, he says there. is lokottararüpá; it tran- 
scends the experience of the workday world. The term alaukika had already 
been used bv Nyàya and Buddhist epistemologists for types of perception thac 

196 {51.18 L 

could not be explained by normal physical causes. 16. BANS 7.3 (Vol. I, 
p. 346): atha vibháva iti kasmát / ucyate, vibhdvo vijñānārthah. 17. Here 
Abhinava is adding to the words of Bharata, who does not make this distinc- 
tion. He merely says, uibhdvah kdranam nimittam hetur iti parydydh (ibid.). 
18. BANS 7.4 (Vol. I, p. 347). Abhinava is following a Northern version of the 
text, closer to that given by MS ba of the Gaekwad edition than to the text 
as it has been printed. 19. One may say that the determinant (vibháva) is 
the cause of the mental state and the dramatic portrayal (anubhávo) is the 
result, so long as one remembers that these words are used not of an everyday 
mental state produced in the actor, but of the super-normal state produced 
in the audience. 

20. BANS 6.31 (Vol. I, p.272). 21. stháyino bħävä rasalvam ápnuvanti, 
BANS Vol. I, p. 288. 22. That is to say, the birth spoken of by Bharata 
is not of a physical thing called rasa, produced by everyday causes, but the 
birth of a super-normal enjoyment engendered by poetic means. Abhinava 
insists on the same distinction in his comment on 2.4. 23. Abhinava's view 
is more complex than this sentence seems to imply. See 3.33b L, note 3. 
24. Vàkyapadiya 2.38. The dictum of course applies only to words that are 
denotative. That the words of poetry are different shows that they must 
have a different sort of power and operation 25. The anonymous oppo- 
nent (Bhattanàyaka?) must have held Mimamsé views, for it is only in the 
Mimamsé that much is made of this fault. To the ritualists it was impor- 
tant to barmonize the many apparently discrepant statements in the Vedic 
texts. They did this by subordinating certain passages to others and so pro- 
ducing ekavdkyatd, "the state of a single sentence," or, if many grammatically 
distinct sentences were involved, by forming a single mahávákya, "great sen- 
tence,” i.e., a single consistent logical presentation. The grammarians, on 
the other hand, eager to make the text of Panini apply to as many linguis- 
tic situations as possible, often instruct us to make two statements out of 
his one (yogavibhdga or vákyabheda). One feels that Abhinava would have 
answered the present charge more simply and more truthfully by pointing 
out that in cases of dhvani the primary sense is always subordinate to the 
suggested sense, whereas in the alañkāras, if there is a suggested sense, it is 
subordinated to the literal. In either case there is ekavdkyatd. Instead of this, 
he makes an elaborate distinction between $àstra and poetry, which leads up 
to the claim that vdkyabheda is a fault in the former but not in the latter. 
26. The Kaumudi advises us to take Sdstra in a broad sense, as referring 
to any text that is not literary. But the argument that follows, it seems to 
me, implies a very narrow sense. It is only a ritual or Mimamsé text that 
is "directed to what one should do in the future,” or that gives rise to such 
notions as “I am enjoined to do this," ete. 27. The words tendgnihotram 

kā promá form a sloko. Gnoli (Udbhata's Vivarana, p. cod, note 1) has 
pointed out that it is a quotation from Dharmakirti's Pramánavarttika, 1.318 

$149 L] 197 

(Manoratha ed. 3.318) 28. Kaumudi and BP make clear that the need 
for semantic convention occurs in the stage of conveying the ribhávas, etc. In 
the next stage, where they suggest rasa, semantic convention plays no part 
29. As opposed to the supernormal plane of rasásváda. 

30. Durdurütam: the word appears also in other forms in the MSS— 
dvondvarüdhom, dadurutam, durdurabham. As I see no satisfactory etymol- 
ogy, I translate simply from context. 

K lt might, however, be an adventitious mark (upalaksana) of a 
certain type of dhvani. And if dhvani has been defined by others, our 
view would stand confirmed. 

A While bhakti might be considered an adventitious mark of one 
type out of the various types of dÀvani that we shall describe, if one 
were to say that dhvani is fully characterized by secondary usage, one 
might say that all the figures of speech which differ from dhvani are 
characterized by the operation of denotation, a statement that would 
imply the futility of constructing definitions of the individual figures of 

Even if dhvani had been defined by others previously, this would 
simply confirm our view, for our view is that there is such a thing 
as dhvani. If this has already been proved, we have gained our wish 
without effort. 

Those too who have said that the nature of dhvani is something 
sensed in the hearts of connoisseurs but incapable of being expressed, 
have spoken without reflection. For when the general definition has 
been given in the manner stated and the particular definitions in the 
manner about to be stated, if it were still held to be inexpressible, the 
same charge would apply to all things. On the other hand, if they mean 
this as a hyperbolical way of saying that its nature is superior to all 
other types of poetry, then they are speaking the truth. 

L Now our opponents may admit that dhvani and bhakti are 
not identical and also that bhakti does not fully define dhvani. But it 

198 [$119 L 

does sometimes characterize it. As bhakti occurs in [some] areas where 
dhvani occurs, it serves as an upalaksana (adventitious characteristic) 
of dhvani. True enough, but it does not occur everywhere that dhvani 
occurs, so what do our opponents gain by this fact, or what do we lose? 
And so [the Kàriká] says, it might be of a certain type. 

And bhakti has been described by the ancient authors. Using it 
simply as an adventitious characteristic, we might be able to mark 
out dhvani and know it in all its varieties. What use is a [specific] 
definition? To dispel such a view, [the Vrtti| argues, if one were to 
Say, etc. The relation of denoter and denoted is found in all areas 
where figures of speech are found. So, since the working of denotation 
has been described by the grammarians and the ritualists, where is 
there any need of work by definers of figures of speech? In the same 
way, by adopting the dictum of the logicians that the effect is born of 
the cause, we might ask what useful new accomplishment has ever been 
made by any creator or discoverer, beginning with God.' At this rate 
00 one would ever undertake anything. Our author states this, saying 
[it] would imply the futility of constructing definitions. 

Let us suppose that we have not revealed anything new, that [the 
nature of dhvani| has already been revealed and we have merely de- 
scribed it correctly [by following precedent]. What harm would this 
do? With this in mind, our author says, even if, etc. Previously, 
that is, previous to our work. 

By thus refuting the three varieties of the view that dhvani does not 
exist, as also the view that it is included in bhakti, it follows that among 
false views the view that it is inexpressible has also been refuted.? 
Accordingly, there is no Káriká directly aimed at refuting this view. 
The author of the Vrtti, however, in order to make a neat presentation 
of the whole subject, brings up this implicitly refuted view and gives 
an explicit refutation: those too who have said, etc. 

Its general definition has been given in the manner stated, viz., "That 
[type of poetry] in which sense or word," etc. (1.13 K). The particular 
definitions will be given in the manner about to be stated, viz., "It may 
be shifted to a different meaning," etc. (2.1 K). In regard to these [we 
may note that] in Chapter One the general definition of dhvani has been 
given by the author of the Karikds. In Chapter Two, the author of the 
Karikás gives definitions of the subtypes of dhvani and refers to the two 
main types as though they had already been given. In confor ity with 
this, the author of the Vrtti had stated the basic divisions already in 
this chapter. by saying, "and it is in general of two sorts” (1.13m A)? 

§1.19 L] 199 

To all things: both everyday matters and scientific matters. A 
hyperbolical way of saying: by this he shows that the inexpress- 
ibility may have been used as a hyperbolical expression, as in the verse 
"those syllables keep on sounding their ineffable message in my heart,”* 
in order to show its supreme excellence. 

(May my words prove] auspicious. 

(There follows a colophon verse made up of puns, so that one is forced 
to render it by two separate translations.) 

Will the world be clear even by moonlight. 
if there is no eye to see? 
So Abhinavagupta has opened an eye. 
I praise the goddess Siva, 

- God's blessed Sakti of understanding, 
who resides within her own self and who. 
by the power which awakens within her 
wakes instantly the universe.* 

Will the (Schrday]- Aloka be clear, even with the 
help of the Candrikà,? if it lacks the Locana? 
So I have made a beginning of the Locana 

I praise the blessed inspiration [of poetry], 
which resides within (the poet] himself 

and which by its power of revelation 

reveals instantly the universe. 

Herewith the First Chapter of the Sohrdayáloka-locona, an exposition 
of dhvoni, revcaled by the great Saiva master, the revered teacher, Abhi- 

1. The argument of the Karikd, explained in the preceding paragraph, was 
that bhakti cannot serve to define dhvani, that is, to rule out the application of 
the term “dhvani” from all instances of non-dhvoni, because bhakti is present 
only in some instances of dhvani. The argument of the Vrtti, explained in 
this paragraph, is more general. Even if-a characteristic occurs throughout 
the area of the thing to be defined, it is of no use as a definition if it is too 
general. 2. The references are to the five false views set forth or implied in 
1.1 K. 3. I suppose the reason for Abhinava's again (see 1.13 m L) bringing 
up this distinction between the exposition of the Karikds and the Vrtti at 
this point is to justify the curious discrepancy between the future tense of 

200 (§ 1.19 L 

vaksyamédnayd ("in the manner about to be stated") and the past tense of 
pratipádite (“has been given”) in the present passage of the Vriti. — 4. The 
quotation forms the last line of the following stanza: nidrárdhanimihtadro 
(or in some versions nidrdnimilitadrso) madamantharáyá nápy arthavanti na 
ca yáni nirarthakáni / adyápi me varatanor madhurdni tasyás tàny aksaráni 
hrdaye kim api dhvanonti 

With half closed eyes, lazy with wine, 

my lady spoke sweet syllables 

They were neither meaningful 

nor yet unmeaning, but were such 

that even now they keep on sounding 

their inefable message in my heart. 
Abhinava may well have taken the quotation from Kuntaka, who uses it in 
his Vakroktifivita (1.19, vs. $1) and says that the indefinite pronoun kim ap: 
conveys the fact that the excitement of the heart cannot be expressed but can 
only be experienced. The stanza is found as number 36 in the Kashmiri re- 
cension of Bilhapa's Cauroporicásiká (ed. W. Solf, Kiel, 1886) and is variously 
ascribed in anthologies to Bilhana (Sárrig. 3468), Kalasa (SuktiM. 43.26), or 
Kalasaka (SubhÁ. 1280). As Bilhana was a contemporary of King Kalasa(ka) 
(regn. a.D. 1080-1088), the ascriptions must be wrong. Kuntaka lived a cen- 
tury or more earlier. — 5. [n the final slokas to the chapters of his Locana 
Abhinava renders homage to the powers or stages of Vdc: at the end of the 
First Chapter, to the pard Sakti; of the Second, to the pasyanti gakts; of the 
Third, to the madhyamd sakti; and of the Fourth, to the vatkhari sakti. That 
speech exists in four stages is a concept as old as Rigveda 1.164.45. We first 
meet with the traditional names of these stages in Bhartrhari; they are miss- 
ing in the Mohdbhdsya, although that work quotes the famous Rigvedic verse 
(Kielhorn ed. I, p. 3, lines 24-25). To Tantric authors like Abhinava the four 
Saktis represent a double process of evolution: on the metaphysical plane, 
from undifferentiated unity through two intermediate stages to the sensible 
everyday world of diversity; on a linguistic plane, from the undifferentiated 
Saddabrahman, again through two intermediate stages, to the physically man- 
ifested speech by which worldly communication is carried on. For Abhinava's 
description of these stages, see his Paryantaparicáfiká, verses 41-48, or Jaya- 
ratha’s commentary on Tantráloka 1.18. Pratibha (imagination, inspiration) 
is used, as well as cit (the self-conscious), as a name of Siva's highest Sakti 
in TA 1.2. It is a serviceable term in the present stanza, because the prati- 
bhà of a poet is the power which enables him to make use of dhvani for the 
achievement of rasa. Compare the verses with which Abhinava concludes the 
later chapters of the book. 6. Candrikd: name of a commentary on the 
Dhvanyáloka that preceded Abhinava's Locana. 


A Thus it has been shown that dhvani is of two kinds:' one where 
the literal meaning is not intended (avivaksitavácya), the other where 
the literal meaning is subordinated to a second meaning (vivaksitànya- 
paravdcya) In this regard, the Karika makes a statement in order to 
show the subdivisions of the first kind. 

1. The word *dhvani" may be taken here in any of the last three senses 
listed in 1.1 K, note 1, viz., as suggestion, suggested meaning, or a suggestive 
type of poetry. Note that it is the Vrtti chat bas shown the distinction between 
the two main types of dhvant, not the Kárikás. See 1.13m A and 1.13m L, 
note 1. 

LI praise the magic branch that bears 
, the wished-for fruit, 
Siva the generous; 
who, as we think of her, gives blessing 
and removes our grief. 

The author of the Vrtti, providing the logical connection between 
this [and the preceding] chapter, begins by saying thus, etc. Shown 
means: "by me, in my capacity as the author of the Vrtti. Nor in 
saying this! did I depart from the intention? of the basic text; rather, 
I followed the intention of the author of the Karikds.” That is why he 
says what follows.? 

In this regard: that is, in regard to the statement of two kinds of 
suggestion made by the author of the Vrtts, this (Káriká which now 
follows) is the seed [from which that statement sprang]. Or we may 
take tatra to mean "at that previous time." The sense will then be: 
the author of the Vriti showed previously, in the First Chapter, the 


202 [8 2.1 Introduction L 

kind of suggestion where the literal sense is not intended.‘ To show 
a difference (prabheda), that is, a sub-type within that, the Karka 
now makes a statement. And by showing these sub-types it will be 
possible, by reference to them, to show [in the second Kdrikd] the 
difference (prabheda) of (the first kind of suggestion] avivaksitavdcya, 
that is, the difference of its nature (prabhinnatva) from (the second 
type] vivaksitányaparavácya.* Basically, the sense is that the author of 
the Kàrikás is in full agreement with the view that there are two kinds 
of suggestion. 

1. "This" refers to the naming of the two varieties of dhvani in 1.13 m A. 
2. Utsütram means “beyond the scope or intention of the basic text,” viz., the 
Künikás. The word is used in Sisupdla 2.112. — 3. The meaning of Ánanda's 
second sentence, on which Abhinava now comments, is perfectly clear. But 
Abhinava is not satisfied with the apparent sense, because the distinction of 
the two main types of suggestion has nowhere been stated in the Kàárikás, 
but only in the Vrtti Accordingly, Abhinava tortures the sense in order to 
turn it into a statement by which the Vrttikára may justify himself. — 4. We 
have emended the word prakdsitah to pra-kdsitasya. As the text stancs, one 
can make sense of it only by the expedient of taking avivaksitavdcyasya yah 
prabhedah to mean avivaksitavdcyanipo yah prabhedah, on the analogy of such 
phrases as rdhoh sirah. 5. Tbus Abhinava takes probheda in two different 
senses. He thereby increases the implications of Ánanda's statement. 

K In that type of suggestion where the literal meaning is not in- 
tended, it may be shifted (sankramita) to a different [associated] sense, 
or it may be entirely set aside (tiraskrta) 

A The suggested meaning likewise is distinguished (i.e., differ- 
entiated into two kinds) by virtue of the two mentioned varieties.! 

§21L]) 203 

1. The point of this sentence is that one might question the propriety 
of the term vdcya (literal sense) in a definition that should deal with the 
suggested sense (dhvani). The response is that the suggestion is also distin- 
guished by the very distinction that has been made in the vdcyo. The KM 
edition adds after the word visesa the following: iti vyarigyaprakásanaporasya 
dhvaner prakároh, apparently only in order to make clear what we have stated 
in this note. The extra words are not found in the recorded MSS and seem 
to have been missing from Abhinava's text. 

L  Thecausative suffix in the past passive participle sarikramita! 
denotes a group of conditions (sahakárivarga)," and it is by their power 
that the sense is caused to shift in the suggestive operation. The word 
tiraskrta refers to the same agents. The connection is this: that literal 
meaning which, being unintended, gives to the form of suggestion called 
"where the literal meaning is unintended” (avivoksitavácya) its name, 
is twofold. (In the first variety,] that meaning which. although possible, 
is not as such of any use; which seems to have become something else 
because of its involvement with various properties; and which remains 
as an unnoticed? property-possessor like the thread of a necklace, is said 
to be (shifted, i.e..] developed into a different shape. That meaning, on 
the other hand, which is not possible in the context and which serves 

. merely as a means to perceiving some other [suggested] sense, after 
which it runs away as it were, is said to be “set aside.” 

Now it may be objected that we are supposed to be describing sub- 
varieties of dhvani, that is to say, the suggested sense. So surely it is 
inappropriate to speak of varieties of the literal sense. With this doubt 
in mind, our author says: the suggested meaning likewise. The 
word ca (likewise) is used in the sense of "because." He means: Be- 
cause of variety in the suggestor [i.e., the vdcydrtha. the literal sense], 
we can appropriately speak of variety in the suggested sense [i.e., the 
distinctions of the suggested meaning depend on distinctions of the lit- 
eral meaning). Or if we take the word dhvani to be used in the sense of 
suggestor |i.e., the literal sense, as explained in 1.1 K, note 1], there 
will be no difficulty. 

Feeling that there is no need to give a definition since the defini- 
tion of this variety is furnished by its very name, the author lof the 
commentary) proceeds to give an example. 

204 [$21 L 

1. sarikramita is the past passive particle of the causative verb: sam + 
kram + nic + it + kta. The suffix nic drops by Pan. 6.4.52, but not witb- 
out denoting the causative agent. Abhinava identifies the agent, i.e., that 
which causes the shifting of the sense, as "the group of conditions" (saAa- 
kárivarga). — 2. The sohakárivorga is composed of the three conditions of 
laksand, namely mukhyérthabddha, nimitta (i.e., sombandha), and prayojana- 
vattva. Both forms of avivoksitavácya depend on laksaná and thus on the 
conditions that bring lcksand into play. Another name for avivaksitavácya 
is loksanámuladhvani. — 3. Against the interpretation of BP, I have taken 
loksyamónaA to include a negative, i.e., as laksyamdnah Only thus does the 
passage make sense. Both the sarikramitavdcya and the thread of a necklace 
are unnoticed property-possessors. What are noticed are the meanings sug- 
gested by the vácya and the flowers or gems strung on the necklace. — 4. Abhi- 
nava bas already explained several times in the first chapter, e.g., Text p. 31 
(Translation 66), 99 (Tr. 125), 105 (Tr. 132), 135 (Tr. 171), that one of the 
meanings of dhvani is vyarijaka, i.e., the vácyártho. The alternate interpreta- 
tion strikes me as impossible, for what is obviously at issue is the vyarigyártha. 
Moreover, the full name is avivoksitavácyadhvani. 

A An example where the literal meaning is shifted (arthántora- 
sarikramitavácya) is the following: 

White herons circle against dark clouds 

that paint the sky with their wet lustre. 

Winds carry the small rain. 

The peacocks, friends of the clouds, cry out with joy. 
Let all this be: my heart is hard; 

lam Rama and can bear it all. 

But Vaideh!, how will she live? 

Alas, my queen, alas, be brave!! 

In this verse the [suggestive] word [whose sense is shifted] is “Rama.” 
By this word we understand Rama as developed into various suggested 
qualities, not simply as the possessor of the name. 

$2.1a L] 205 

1. The stanza is quoted anonymously by Mammata 4, vs. 112 (p. 188), 
Hemacandra AC, vs. 68, and SuktiM. 90.6. Jacobi pointed out (JRAS, 1898, 
p. 296) that it occurs as Mahànátaka 5.7 (the Eastern version of the Hanu- 
manndtaka). The Western version lacks it. The quotation begins with a 
description of the rainy season, a time when lovers long to be together. The 
point in quoting the verse here is that the word Rima qualified by “hard- 
hearted" reminds one of, or rather suggests, many other qualities that are 
associated with Ráma, namely his having undergone so many other hardships, 
such as the loss of his kingdom, his exile in the forest, etc. The beauties of 
the verse are explained by L below. One may point in addition, however, to 
the skill with which the stimulants of sight, touch, and hearing are combined 
in the first two lines (the first four lines of the translation). 

L Taking it for granted that the definitions are sufficiently indi- 
cated by the names of the types aad their subtypes, our author proceeds 
to give an example. An example where the literal meaning is 
shifted: this phrase is syntactically joined with “As in this verse, the 
suggestive word is Rama.” Snigdha means "moist" because of its con- 
nection with water. Syàmala is the dark color commonly found among 
Southern women. Kanti means brilliance. Lipta means covered; that 
is, the sky is covered by clouds of just such a brilliance. "Clouds that 
are vellad-balàkah," that is, in which the herons, a species of white 
bird, are vellat: the word means "conspicuous" (vijrmbhamdna), sc., 
because of the contrast! (of their white bodies with the black clouds], 

- and "flying about" (calat), sc., because of their joy (at being with 
their friends, the clouds]. And so the sky is painful to look at [since 
it reminds one of days of love]. All the directions are also hard to 
bear. The use of the plural in “winds” shows that they blow from 
all directions; and by their releasing small drops of water it is sug- 
gested that they are blowing very gently [and thus linger over one's 
body and make one all the more love-sick]. Well then, perhaps Rama 
should enter a cave somewhere and stay there for the duration of the 
rainy season. With this in mind, the poet says that the clouds have 
friends (or helpers) among whom are the peacocks? who produce out of 

7 joy sweet sounds that resemble the sadja? note and become reminders 
of that whole unbearable scene of the clouds. On their own as well, 
these sounds are quite unbearable. This is what is meant. In this way 
Rama. whose feeling of love in separation has been aroused by stim- 
ulating factors (uddipanavibhàávas), knowing that these determinants 
of emotion (vibhdvas)* will be shared (by Sita]. since love is based on 

208 (§ 21a Z 

mutual feeling, from here on in the poem conjures up his beloved in 
his heart. First he reports on himself: "Let this be." Drdham means 
"exceedingly." The word "hard-hearted" (kathora-hrdaya) gives scope 
to the particular suggestion that is achieved through the word Ráma 
and its literal sense,? just as the word natabhitti in the verse that 
begins tad geham natabhitti ("That house with crumbling walls," cf. 
3.161 A). Otherwise how would the word Rama not suggest other mean- 
ings connected with other qualities, e.g., the fact that he was born in 
the family of Dasaratha, that he was the object of Kausalya’s love, 
the deeds of his childhood, and the acquisition of Sità? Asmi means: 
“I am the self-same person (who has undergone all these sorrows]." 
Bhavisyati expresses action in general, so the meaning is: What will 
she do? It can also be taken in the sense that “her very being is im- 
possible" (i.e., she will kill herself. or die of a broken heart]. In this 
way by a succession of memory, name (sc. "Vaidehi"], and specula- 
tion (sc., “what will happen to her?"], he has conjured up his beloved 
from his heart into being present before him. To her, as her heart 
is about to break. he says with agitation, “Alas, my queen, alas. be 
brave!" The word "queen" suggests that fortitude will be the proper re- 

By this: that is, by the word Ráma, the literal sense of which is 
not strictly useful here [to the idea intended|. The suggestions of other 
properties, which suggestions form the purpose (for shifting from the 
literal to the secondary meaning] are endless; for example, his banish- 
ment from the kingdom, etc. And since these suggestions are countless, 
they cannot be conveyed [simultaneously] by means of the denotative 
function of words. Even if these innumerable suggested properties were 
to be conveyed [by denotation] one by one. since they will not be had 
in one single act of cognition, they will not be the source of a wondrous 
aesthetic experience and hence they will not give rise to great beauty. 
But if these properties are suggested, they will assume countless forms 
(kim kim rüpam na sahate) because in the suggestion their separate- 
ness will not be clearly perceived. In this way they will become the 
source of a strikingly beautiful aesthetic pleasure that is analogous to 
the flavor of a wonderful drink, or cake, or sweet confection [where the 
individual ingredients cannot be separately tasted but yet add to the 
flavor of the final product]. For it has been said already that a word 
which is suggestive reveals a beauty "which cannot be conveyed by an- 
other form of expresion" (1.15 K). In all cases where the purpose [in a 
secondary usage] is to achieve a suggestion. this (viz., the simultaneity 

§2.1b L] 207 

of a number of suggestions] should be considered the cause of excel- 
lence. The word simply shows that the literal sense of the denoted 
object (sarijriin) is not wholly set aside. 

1. porabhága: from the sense of emphasis, work done in relief, etc., para- 
bhága often comes to mean contrast, though that sense is not given by PW; 
cf. Ragh. 5.70, KirArj 8.42. — 2. Sobhanahrdaya is merely a literal gloss 
of suhrd. — 3. This is a poetic convention (cf. Ragh. 1.39 with Mallinatha 
thereon), just as the cuckoo is said to imitate the pancama. — 4. Note that 
Abhinava mentions elements from rosadhvani, whereas Ananda has given the 
verse as an example of avivoksitavdcya. Abhinava mentions the terms vibháva 
and vipralambAa(srrigára). This is not, however, a departure from the view 
of the basic text. There is textual justification, but it will not come until 
Chapter Three. There, under 3.43, mention is made of all the possible com- 
binations of dAvani. gunibhütavyangya, and vdcyélankdras. In fact Ananda 
there mentions this verse as an example of the combination of different types 
of dhvani. — 5. The idea is that “Rama” by itself might conjure up sugges- 
tions that are not meant. But the fact that the adjective kathorahrdaya is 
used lets us know that it is Ráma's character in the face of sufferings that is 

A Another example is found in my Visamabanalilá: 
Virtues blossom 
when admired by men of taste. 
When graced by the sun's rays 
a lotus becomes a lotus. 

is example the suggestive word is "lotus" in its second occurrence.’ 

1. By "virtues" the verse no doubt refers to poetic beauties. They shine 
forth only under the eye of the connoisseur, just as the lotus opens its petals 
only in the sunlight. We have here a case of pure vastudhvani as opposed to 
the combination of vestudhvani and rasadhvami in the previous example. 

L Another example: tālā in Prakrit means "then," and jalé 
"when."! Gheppanti means "are taken [i.e., admired]."? He gives a cor- 
roborative statement in the line beginning with ravikirana. 

208 [$21bL 

The suggestive word is "lotus": he is speaking of the bearer 
of the name "lotus" as transformed into a hundred varieties by such 
properties as being the abode of the goddess of beauty, etc. [In both 
this and the previous example] there is blocking of the pure. literal 
sense [of the words “Rama” and "lotus"]. The cause [of adopting the 
secondary sense] is the inherence of those [secondary] qualities in the 
literal or primary sense. It is through this cause (viz., this connection 
or inherence) that the word “Rama” conveys by means of secondary 
usage (laksayati) a meaning that is transformed by other qualities. 
The sense that is suggested [and that forms the purpose of employ- 
ing the secondary usage] consists of still other qualities which are ex- 
traordinary and which are beyond the scope of words [i.e., beyond the 
scope of denotation]. The same holds true of the word "lotus."? On 
the other hand, the word “virtues” (guna) denotes merely the bearer 
of its primary sense. The forced view taken by some that what we 
have here (sc., in “Rama” and "lotus"] is [simply] a metaphorical (or 
secondary) sense (dropite) is unconviocing, because wherever a pri- 
mary sense is blocked by its uselessness we are in the area of sug- 
gestion and the secondary sense is only that from which the sugges- 
tion arises.* 

As for what the Hrdayadarpana says, that the aesthetic delight of the 
verse [of 2.1a] is occasioned by the distress (samrambha) that comes 
from the particles “alas, alas,” we must point out that even in this 
interpretation the suggestion of an aesthetic experience (rasadhvani) 
is admitted, because distress or agitation (dvega) is a transient emo- 
tion (vyabhicdribhdva) of the aesthetic experience of love in separation 
(vipralambhosrrigàra). Furthermore, without the help of the meaning 
suggested by the word “Rama,” this distress would not blossom forth. 
For the emotion takes the form of "I can bear it, but what will happen 
to her?" Moreover, in the case of the word "lotus" what agitation can 
there be? Enough arguing then.> 

As there is failure of the literal sense, that sense being useless, we 
have in these verses examples of something that is based on the sec- 
ondary meaning (laksanámüla) and that may rightly be called sugges- 
tion of the type where the literal sense is not intended (avivaksita- 
vdcya), for in it there is no intention to express its direct meaning. 
But the literal sense in the form of property-possessor is not totally set 
aside, since it is carried along [in the sense of Rama, who is qualified 
by the loss of his kingdom, etc.]. Hence I have used the expression 
parinata (developed or transformed). 

§21cA] 209 

1. See Hemacandra, Prákrtavyókarana 8.3.65. 2. Toid., 8.4.256 
3. According to BP the laksydrtha of "RAma" is Rama transformed by exile 
from his kingdom, etc. (rdjyabhramsddiparinatardma) and the vyarigyárthas 
are discouragement, weariness, etc. (nirvedaglánimohádi); of "lotus" the laks- 
yartha is loksmipatratuddiparinatakamala, the vyangyárthas are manoratha- 
tvàdi. Abhinava is not so precise. — 4. What Abhinava is objecting to is an 
interpretation that would rest satisfied with finding some single property of 
Ràra or of the lotus as the final intention of the verses, this property be- 
ing arrived at by laksand or óropa. Such an interpretation is far from being 
unreasonable and can be dismissed only by appeal to the general theory of 
suggestion. The theory demands that everywhere that we find mukhydrtha- 
bädha together with a laksya sense that does not rest merely on rüdAi, there 
must be a suggestion. In the present examples the suggested sense, except 
that it is said to be multiple, does not differ very noticeably from the laksya 
sense. 5. Abhinava's point is that the samrambha does not actually consist 
of the words ha hā Ad. It consists in the fact that Rama thinks, “I can bear 
it, but she cannot.” Now this knowledge we cannot have without the use of 
tbe word “Rama” as a vyanjekasabda. He is not fair, however, in saying that 
Bhattanáyaka's interpretation must admit rasadhvani in the stanza. Bhatta- 
náyaka admits that there is rasa in the stanza, but not dħvani. As though 
aware that he has pressed bis criticism too far, Abhinava turns to an example 
where there is no rasa. And yet the process of our understanding “lotus” is 
the same as the process of our understanding "Ráma." 6. [n the case of the 
first example “Rama” in its literal sense is useless to the idea of the capacity 
to endure those stimulating emotions (uddipanavibhávas) that are described 
in the stadza. In the case of the second example the second occurrence of the 
word "lotus" is useless in its literal sense, which would merely be a repetition 
6f the prior occurrence. 

A An example of the second variety, where the primary meaning 
is entirely set aside (atyantatiraskrta) is a verse by the first of poets, 

The sun has stolen our affection for the moon. 
whose circle now is dull with frost 

and like a mirror blinded by breath 

shines no more.' 

is example the suggestive word is “blinded.” 

210 [§ 21c A 

1. The verse is Rámáyana 3.15.13, from Laksmana's description of win- 
terime. Lis incorrect in assigning the words to Rama. Saubhdgya means 
literally success in love. The moon. which is loved for its coolness in sum- 
mer. loses its hold on our affection in winter, when we turn to the sun for 
warmth. Compare Rdémdyana 3.15.5 from the same description: subhago 
havyaváhonah, "and Gre has won our affection.” The point of quoting the 
verse is for its phrase “a blinded mirror,” which L will discuss in detail. 

L By the first of poets shows that [this kind of] suggestion is 
well known in literature.! 

The sun has stolen: these are the words of Rama describing the 
winter when he was at Paücavati. "Blind" (andha) means one whose 
sight is destroyed, for even a person born blind has had his sight de- 
stroyed in the womb. [When we say of someone:| "He is blind; he 
cannot see in front of him," the meaning of the word "blind" is only 
partly set aside, not entirely. But to a mirror, such as we have here, 
blindness cannot be applied even by an imaginary superimposition of 
this literal sense [for a mirror, being insentient, has no sight which 
could be destroyed]. The word "blind" can apply to a mirror only in 
the secondary sense of "being incapable of making a clear representa- 
tion," a sense occasioned by the presence of that incapability in a man 
who is literally blind. The purpose of using this word here is that it 
suggests numberless properties such as an exceptional loss of beauty, 
uselessness, etc. 

As for what Bhattanayaka has said, that because of the use of the 
word "like" (iva) there is no secondary usage at all in this stanza,” he 
said this without really thinking about the meaning of the stanza. For 
the word iva conveys the similarity between the moon and the mirror. 
The words "blinded by breath" qualify "mirror." If, however, the word 
iva is connected with the literal sense of the word "blind" (andha), 
then we shall be left with the presentation [of an identity| "the moon 
a mirror." This constuction of the word iva is harsh. To say that 
the word iva should be connected both with "blinded by breath" and 
with "mirror"* is not proper. This procedure might be valid in the 
Mimamsa system; it has no place in poetry. So enough 

1. The line is mispunctuated io both the Kashi and the Vidyabhavan edi- 
tions. In Kashi, p. 172. Locana. line 2, delete both dandas and put dandas in 
place of the dashes. 2. Bhattandyaka would join the word iva with andha: 
"a mirror blinded as it were." Here "blinded" is used in its literal sense, the 

§ 21d L] 211 

fancy being expressed by the word iva and therefore not being suggested. 
3. Abhinava's argument is that we need the word iva to connect moon and 
mirror: candramá ádarsa ivo. If you borrow the word iva to express the fancy 
andha iva. you will be left with an awkward rüpaka: ádar$as candrumá. The 
only solution would be to take the word iva twice (durttyd), as the grammar- 
ians and the Mimàmsakas sometimes do; the exegetical technique is known 
as tantra; see 2.4 L, note 12. Abhinava then repels such a solution. 4. In 
nih$vásándha iva, iva would serve as utpreksóvácaka (a word expressing the 
fancy). In ddersa iva it would serve as áupamyavácaka (a word expressing 
the simile). 

A [In the following verse the suggestive words are "drunken" 
(matta) and “pride” (aharikára). 

Though the sky is filled with drunken clouds 

and the woods with arjunos thrashing in the downpour, 
these black nights too when the moon has lost its pride 
carry off my beart.' 

1. This Prakrit verse is Caudavaho 406. It describes a monsoon night, 
dear because of its associations of marital intimacy. The orjuna is a grey- 
barked tree with leaves of immense size. It is probably on their account that 
it is singled out for thrashing in the monsoon storm. 

L "Though the sky,” etc. [L follows this by a Sanskrit trans- 
lation of the Prakrit verse; then continues:] The word ca should be 
understood in the sense of api (even).! Even when the sky is full of 
drunken clouds [it carries off my heart], not only when it is full of stars. 
So also the woods when their arjuna trees are thrashing in torrents of 
rain, not only when they are filled with mango blossoms shaken by the 
breeze from the Malaya moutains. Even these black nights when the 
moon has lost all pride carry off [my heart], not only nights that are 
whitened by the rays of the moon. "Carry off" means that they pro- 
duce longing. The word "drunken" in its primary sense is impossible in 

212 [$214 L 

the context. As its sense of "a person who has become intoxicated by 
the use of wine" is blocked, it applies to the clouds through a metaphor- 
ical extension to other properties common [to an intoxicated man] and 
so suggests thousands of properties, such as irrational conduct, unruly 
behavior, etc. The phrase "without pride" is also applied by secondary 
usage. In applying to the moon it suggests the subservience, the lack 
of luster, the lack of will to rise up and overcome, etc., of one who has 
literally lost his pride.? 

1. The literal rendering of the verse is: "And it be filled with drunken 
clouds,” etc. This use of “and” for "although" is found in Latin and Shake- 
spearean English as well as Sanskrit. — 2. The literal sense of “pride” can 
apply only to a sentient belng. In the compound tatpáratantryo, tat refers to 

K The soul of dhvani! where the literal meaning is intended 
(vivaksitábhidheya), has two varieties: one where the suggested mean- 
ing is produced without apparent sequence (alaksyakrama) [i.e., im- 
mediately, together with the primary meaning], the other where the 
sequence is apparent. 

1. Dhvaner átmá is a somewhat ambiguous phrase. One might translate 
the present Kárikà as “The nature of dhvant has two varieties," and this 
is the way Abhinava takes it. But in 23 K, A and in 2.11 A, 217 K, A, 
and 3.16 A, the term seems to mean dhvani par excellence, the very soul of 
dhvani. In all these passages the term refers to the suggested sense in vivakst- 
tànyaporavácyadhvani. For this eleven-syllable title, which will not fit into 
sloka meter, the Káriká substitutes the abbreviation vivaksitábhidheyo. As L 
will point out, the portion onyapara (subordinate to another sense) can be 
supplied from the context. 

A The nature (or very self, dtman) of dhvani is a suggested sense 
which takes precedence over the literal sense.' It has been divided into 
two varieties: one where the suggestion appears without a perceived 
sequence between the literal and the suggested meanings, the other 
where a sequence is apparent.” 

§ 2.3 Introduction A] 213 

1. That is to say, when the literal sense is subordinate to the suggested 
sense (anyapara = vyarigyaparo). 2. First we perceive the literal sense; 
then, after a momentary interval, the suggested sense dawns on us. There is 
of course a krama even in asamlaksyakrama, but it is so swift that we no more 
notice it than we notice the succession of punctures when a oeedle pierces a 
pile of lotus petals. 

L When the "different nature" (prabhinnatva; see 2.1 Intro. L) 
was mentioned of that type of suggestion where the literal meaning is 
unintended, from what was it meant to be different? For a thing cannot 
be different from itself. With this point in mind, he says that that type 
is different from the type where the literal sense is intended; for "in- 
tended” and "not intended" contradict one another. So the Kérikd says 
“imperceptible, etc.” The word asamlaksyakramoddyota is a bahuvrihi 
compound and means "that of which the revelation. i.e., the process of 
revelation, is such that its sequence in time cannot be well perceived." 
The fact that the literal sense is subordinated to something else (anya- 
para), although not expressly stated [in the Kànriká], is implied by the 
expression vivaksitábhidheya (“where the literal meaning is intended") 
because of the proximity of this expression to the word dhvani." 

Of dhvani: i.e., of the suggested sense. The nature (átman): In 
the previous Karika the varieties of suggested sense were distinguished 
on the basis of the literal sense. Now the present varieties are distin- 
guished solely within the suggested sense itself, the distinction being 
based on the process by which the suggestion operates.? But what se- 
quence can there be within the process of suggestion itself? He tells 
us: a sequence with respect to the literal sense. The literal sense will 
here be the vibhávas, etc. 

1. The word dhvani used in the Kàrikà implies that the vdcydrtha leads 
up to the suggested sense; in other words, that it is anyaparu. — 2. The divi- 
sion in vivaksitànyaparavácyadhvani is based on the vyarijanavyápára, whereas 
in avivaksitavácyadhvani the basis of division was the vácyártho. 

[$2.3 K 

K A rasa, bháva, rosübhása, bhdvdbhdsa, bhávaprasánti, etc., 
appearing as a predominant element and [so] constituting the soul of 
dhvani, are assigned to the non-sequential type.? 

1. For these technical terms see Introduction, pp. 17-20, 37, and Abbi- 
nava's remarks on this section and on 2.4 and its various subdivisions. For 
bhávadhvani Abhinava invents a wholly new meaning. 2. This is the nat- 
ural interpretation, with átman having its full value and vyavosthita having 
its normal meaning of "assigned, distributed." L's interpretation is slightly 

A For the suggested sense, such as a rasa, etc., is apprehended 
nearly at the same time as we apprehend the literal meaning. When it 
predominates, that is the soul of suggestion. 

L Of these: among the two. Only that meaning which is of 
the form of rosa, etc., constitutes the variety of suggestion that is non- 
sequential. But this is not to say that such a meaning is always non- 
sequential. For sometimes we find that rasa does involve a sequence. 
When it does, we have a variety of arthasaktyudbhavánusvánarüpa- 
dhvani (suggestion similar to the resonance of a bell, a suggestion based 
on the power of meaning), as will be stated later in the text. The word 
&tmà, which means literally one's own nature (svabhdvavacana), here 
conveys the idea of "variety" (prakdra). Hence a [suggested] sense 
such as rasa, etc., is called the nonsequential variety of dhvani. What 
is meant is that i it the sequence (from the literal to the suggested 
sense] is not perceived. 

Are rasa, etc., invariably a variety of dhvani? He says no, only when 
they are revealed as the predominant element, that is, as of major 
imporance [in the poem). Although at the time of giving the general 
definition (of dhvani in Chapter One, Karika 13], this was explained 
by using the phrase "to which all other elements are subordinated" 
(gunikrtasvárthau), it is here repeated in order to give occasion for the 

$2.3 L] 215 

discussion of rosavat (that which contains rasa in a subordinated po 
sition) and other figures of speech [in Karikd 2.5]. And these elements 
such as rosa are distributed (vyavasthita, i.e., are everywhere present 
in poetry), for without them there is no such thing as poetry.' 

Although it is through rosa that all poetry lives and despite the fact 
that rasa is essentially an undifferentiated mass of aesthetic delight, it 
may still give rise to an extra degree of aesthetic pleasure through the 
agency of some particular element which acts as its cause. In sucb cases, 
when some particular transient state of mind (vyabhicdrin) reaches a 
high pitch and gives rise to exceptional aesthetic delight, we have what 
is called bhávadhvani.? An example is | Vikramorvesiya 4.9]: 

Can she be angry 
and using her magic to remain invisi 
But she was never angry for long. 
Has she flown off to heaven? 
But she loves me deeply in her heart. 
Even the demoas could not steal her from me 
when I was with her. 
Yet now she has utterly disappeared. 
What turn of fate is this? 
Although tbe rasa of love in separation is present throughout this ex- 
ample, the exceptional pleasure is occasioned by the striking effect of 
the transient state of mind known as "speculation" (vitarka). 
Transient emotions have three states: inception, stasis, and cessation. 
, As has been said,” " Vyabhicárins are so called because in bringing [the 
rosas| before us they act in diverse ways." 
Sometimes the emotion is presented in the stage of inception. For 
The slender damsel heard him when in bed 
address her by another woman's name 
She thought of turning away— 
decided to try to do so— 
had almost done it, loosening one graceful arm— 
but could not lift her bosom 
from her lover's breast.* 
In this stanza the words "could not" show that the state of jealous 
anger is arrested in incipience, since they deny its full emergence; and 
on this depends the aesthetic enjoyment of the verse. 
The stasis [of a vyabhicáribhava, in this case speculation or doubt] 
has been illustrated just above in the verse “Can she be angry?” 

216 (§ 2.32 

Sometimes it is the cessation of a transient state that occasions the 
aesthetic delight. This was illustrated earlier in the verse “They lay 
upon the bed"* and is called bhávaprasama (the cessation of an emo- 
tion). In the verse in question we might also speak of the cessation of 
a rosa, viz., of love in separation due to jealous anger. 

Sometimes, again, the cause of aesthetic relish is the coming together 
of two transient emotions." For example: 

He who bas kissed a face 
beautified by jealousy 
has known the bliss 

of drinking nectar.’ 

In this example, in which anger is directly expressed, the man who 
kisses the face of his beloved as she weeps softly, sobbing with anger, 
is said to bave known the satisfaction of swallowing successive drafts 
of nectar.? The cause of the aesthetic delight is thus a mixture [of the 
directly expressed] anger and (the suggested] reconciliation. 

Sometimes it is the mixture of different vyabhicdrins that is the object 
of [aesthetic] satisfaction. For example: 

How can a king of the Lunar Dynasty do a forbidden deed? 
May I see her once again! 

I have learned the scriptures in order to abstain (rom sin. 
Even in anger ber face was lovely. 

What will the wise and sinless say? 
One could not find her even in a dream. 

My heart, come to your senses! 
But what blessed man will drink her lower lip?'? 

In this stanza the states of mind arranged in pairs are mutually contra- 
dictory: compunction and longing, intellectual thought and memory, 
doubt and despair, firmness and anxiety. Yet since the last state is 
anxiety, the preceding states confer importance on anxiety alone and 
thus [by being mixed together] they give rise to the highest aesthetic 
pleasure. In like manner other [examples] may be supplied. All of 
these: incipience, conjuncture, mixture, etc., are intended by the use 
of the term etcetera (àdi) in the Kartkd. E^ 

It might be objected that in like manner great aesthetic delight is 
conveyed througb the vibhávas and anubhávas, and so we should speak 
also of vibhdvadhvani and anubhávadhvani. But no. For both vibhavas 
and anubhávas are conveyed directly by denotation [and not by sug- 
gestion. And the aesthetic delight arising from them terminates in 

82.3 L] 217 

[certain] states of mind (cittavrtti, namely the stháyibhávas) alone; so 
there is nothing to be aesthetically enjoyed in them apart from rosa and 
bhava [to whose aesthetic enjoyment they lead]. But there is nothing 
wrong with saying that when the vibhdva and anubhdva are suggested, 
they are cases of vastudhvani.! 

Now when false love (ratyábhàsa, literally “the appearance of love") 
arises out of a false vibhávo, the enjoyment is false because of the falsity 
of the vibháva and hence is known as rasdbhdsa, “false or improper aes- 
thetic enjoyment." An example is the srngàáràábhása (the false aesthetic 
feeling of love) that arises when we listen to the poem of Ravana.!? Al- 
though Bharata has said that "the imitation of the erotic (srrigára) is 
comic (hásya),"!? the feeling of its being comic arises only at a later 
time. [n the verse 

I merely heard her name 

and it acted as a magnet or a maddening charm. 
Since that day my heart has known 

no moment's rest without ber. 

there is no occasion for relishing comedy. Now someone will object, "In 
this verse love is not the basic emotion (stháyibháva), for it lacks the 
bond of mutual affection." But who ever said that there was love here? 
Rather, we have here a case of false love (ratydbhdsa, the appearance 
of love). And it is false precisely because it never occurs to the heart 
lof Ravana] that Sita might be indifferent to him or even hostile. For if 
this were to occur to him, bis desire would disappear. And even if he 
thinks that she is in love with bim, that thought gives no assurance, 
for he is infatuated with passion. Accordingly, tbe love is established 
to be in reality spurious, just as the silver which one cognizes in a 
piece of mother-of-pearl is spurious. But Bharata himself indicated as 
much when he used the term £rrigáránukrti (imitation of the erotic), 
for anukrti, amukhyatd, and dbhdsa are all synonyms. Therefore when 
writers use the word Sriigdra in situations where the love is one-sided, 
it should be understood in the sense of an imitation (àbhàsa) of real 
$rrigára. The word srrigára [in Bharata's phrase srrigáránukrtir yos tu 
so hdsyah| implies further the possibility of imitation (or falsity, ébhdse) 
of the heroic and other rasas. 

Thus, bhávadhvani, etc., are the outflow of rosadhvani. [In setting 
up these categories] we merely single out one major cause of aesthetic 
delight and consider it separately, just as connoisseurs of scent, even 
when they enjoy the unified flavor of a perfume, are able to say that the 

218 [$2.3 L 

sweetness comes from the nard,’* or some other element, alone. But 
rasadhvani is par ezcellence the intense relish occasioned by the audi- 
ence’s (pratipattuh) tasting of the basic emotional element when their 
understanding of this basic emotion has arisen from the combination 
of the vibhávas, anubhávas, and vyabhicdribhdvas. For example: 

My eyes with difficulty pass her thighs 

to wander long in the land about her hips: 

then at her waist, uneven with the triple fold. 
become quite powerless to move. 

But now at last, like travelers parched by thirst 
they've climbed the mountains of her breasts 

and see at last what they had hoped, 

their counterparts, her eyes, that flow witb tears.!* 

In this example, from the King of Vatsa's looking at the painting [of two 
figures] which is honored by bis own portrait and which he is describing 
[to bis friend] because of the portrait of the heroine, [we know that] the 
stháyibhàva of love is in that state where it is mutually shared. This 
sthdyibhdva, by means of the combination of vibhávas and anubhdvas, 
has reached a point where it can be aesthetically enjoyed. So enough of 
this long discussion. It is now established that a meaning in the form of 
rasa, etc., when it appears as the major element (in a work of literature] 
is a variety of that kind of dhvani known as asamlaksyakramavyarigya. 

Nearly at the same time: the word "nearly? (iva) shows that 
although there is a sequence, it is not perceptible. As we apprehend 
the literal sense: viz., the vibhávas, anubhávas, etc. 

1. Abhinava's literal interpretation of the Káriká is: rasa, bháva, etc., 
when appearing as predominant, constitute the variety of suggestion that 
can appear without perceived sequence, a variety that is widely distributed 
{in poetry]. 2. Presumably Abhinava understands this term as an elliptical 
compound meaning "suggestion of [a rasa brought about by] a striking vyabhi- 
cáribháva." 3. The quotation is from BANS Book 7 (Vol. 1, p. 355), where 
the MSS disagree on the text. The editors have chosen the reading vivi- 
dham dbhimukhyena rasesu carantiti vyabhicárinah. Abhinava's commentary 
on Book 7 is lost. Presumably he interpreted vividham (in various ways) to 
mean "by their inception, stasis, and cessation." 4. This verse is found 
also in Pratiharendurdja’s commentary on Udbhata. p. 88, and as No. 151 in 
the parigista to the Amarusataka. 5. Quoted above in 1.4g L. 6. Love 
in separation (viprolambhasrigdra) is a rasa, not a bháva. In the verse in 
question the separation is caused by jealousy. The rusa disappears into its 
contrary rasa, love in union. 7. Read vyabhicdrinoh. 8. This verse is 

§ 2.4 Introduction A ) 219 

found also in Hernacandra's Desínámamalà, 1.142. We do not know its origin 
For sumthidim read sumhide. By osuru Abhinava presumably understood 
irgyá, or possibly irgyásru (BP), as he says that the anger (i.e., irsyd) is 
directly expressed. 9. Literally, "the satisfaction of successive pleasures of 
swallowing nectar." 

10. This stanza is found also in the Kávyaprakása, 4, quotation $3; and 7, 
quotation 331. It is not from the Vikramorvasiya as many commentators on 
Mammata have said. BP says that it represents the words of King Yayáti, who 
has fallen in love with the brahmin girl Devayáni despite the restrictions of 
the laws of caste. The verse is later quoted by Ananda (3.20a A below) as an 
example of the introduction of elements from an obstructive rasa (here santa) 
in order by stopping short to magnify the rasa intended (here srvigéra). Com- 
punction, intelligence, doubt, and firmness are vyabhicáribhàvas of sántarasa. 
Longing, memory, despair, and anxiety are vyabhicdrins of love in separa- 
tion. 11. Vostudhvoni (the suggestion of an object or situation) belongs 
to the sequential variety of suggestion and will be described later. — 12. It 
is not clear whether the reference is simply to a poem about Ravana or to a 
work called the Rávanakávya. See 1.4g L, note 4, where one may add that 
the complete stanza is also quoted in AbABh., Vol. 1, p. 295. — 13. BANS 1, 
p.295. 14. "Gives no assurance” (na niscayena krtam): One MS reads 
krtyam. The parallel passage in AbABA., Vol. 1, 295, has niscayo hy anupa- 
yogi, "Such an assurance proves nothing.” 15. Māmsi is a kind of nard, 
Nardostachys jatamamsi (PW). — 16. Ratndvali 2.10. The king is looking at 
a portrait of Ságariká and himself. 

A Now it will be shown that this type of dhvani where the sug- 
gestion is without apparent sequence (from literal meaning to suggested 
meaning) is different from the figure of speech known as rasavat.’ 

1. The figure of speech rosavat was known to the older poeticians. Dandin 
says merely that it was a figure charming with rasa (2.275). Bhamaha says 
little more: “The figure rosavat is such that the rasas srrigára, etc., are clearly 
exhibited therein” (3.6). Udbhata brings into his definition the formative 
factors of rasa as given in BANS: “The figure rasavat is where the rise of a 
rasa such as jrrigáro is clearly exhibited. It is a locus of rasa, stháyibhàva, 
saricárin (= vyabhicdrin), vibháva, and dramatic portrayal” (4.3 Indur&ja = 

220 [8 2.4 Introduction A 

4.4 Vivrti). For Ananda to establish his system, in which the aim of poetry 
is rasa achieved by means of dhvoni, it was necessary to distinguish this rasa- 
dhvani from the old figure of speech. See Introduction, pp. 23-24. He proceeds 
to do so here by assigning rasadhvani to cases where the rasa is predominant 
and limiting the Ggure of speech to cases where the rosa is subordinate or 

K Wherever the varied word, meaning, and their causes of beauty 
are subordinated to rosa, etc., this is considered the domain of dhvani. 

A Wherever the words together with the alarikáras (ornaments, 
figures) of sound, the meanings together with the figures of meaniog, 
and the qualities (gunas), all variously arranged so as to be kept distinct 
from suggestion, subordinate themselves to the main suggested sense 
which consists of rasa, bhàva, rasébhdsa, bhávábhása, or bhdvapresanti, 
one may apply the term dhvani to that poem. 

L  Kérikd 2.3 spoke of rasa, etc, "appearing as predominant 
elements.” It may therefore be asked if rasa, etc., can ever appear as 
subordinate elements so that it should be necessary to qualify them [as 
predominant] in order to rule out (their subordination in dhvani]. By 
way of answer the [the Vrtti] proceeds to say, Now, etc. His thought 
is that there is subordination of rasa, etc., when they assume the form 
of the figures of speech known as rasavat, preyas,’ ürjasvin? and samá- 
hita. His turn of phrase indicates that rasadhvani, bhávadhvani, etc., 
are not included within the figures of speech rosavat, etc. [i.e., the 
provinces of the two are different], for (in the same way,) it was earlier 
shown that vastudhvani is not subsumed under the figures of speech 
samásokti, etc. (cf. 1:13 c]: 

The expression used in Káriká 2.4, vdcyavdcakacdrutuahetu (words, 
meanings, and their causes of beauty), is a dvandva compound mean- 
ing "the literal sense, the denoting word, and their causes of beauty." 
In the Vrtti [the expression sabdárthálankárá| is also a dvandva com- 
pound meaning "words together with figures of speech based on sound; 

$24 L] 221 

and meanings together with figures of speech based on meanings." Is 
considered: i.e., has already been said [by us in 1.13]. 

But now,‘ Bhattan&yaka has said: “If rasa were perceived (pratiyate) 
as belonging to someone else, the spectator would remain indifferent. 
Nor can rosa, which stems from a poem dealing with a subject like 
the life and deeds of Rama, be perceived as belonging to oneself. For 
if it were perceived as belonging to oneself, we should have to admit 
that there was a physical production (utpatti) of rasa within oneself. 
And such a physical production would be inappropriate coming from 
Sttà, for she cannot serve as an [objective] determinant (dlambena- 
vibháva) to the spectator. Should it be argued that a certain universal 
"belovedness" (kántátva) causes her to become such a determinant [to 
the spectator] in the sense of causing a flowering of his latent impres- 
sions (vàsaná), we may ask how such a process could be possible in 
the case of the description of gods, etc. Nor can it be said that dur- 
ing a dramatic performance there is (on the part of the spectator] a 
recollection of his own beloved. And how can a {stimulative} determi- 
nant (uddipanavibháva) such as building a bridge over the ocean by an 
extraordinary hero like Rama ever become generalized [since nobody 
else could ever do it]? Nor can it be said that Rama, as full of heroic 
energy (utsdha), is remembered,” for be has never formed part of our 
past experience. Again, to perceive Rama's energy through a verbal 
source of knowledge is not to experience rasa,* just as when we watch 
a couple making love there is no experience of rasa.? And if we accept 
that rosas bave a physical origin (utpotti), the spectator would be so 

“pained by his [physical] sorrow (karuna, i.e., Soka) that he would never 
return to watch a tragic (karuna) performance. Therefore there is no 
physical production. Neither is there a manifestation, !? for if the erotic 
rusa were a power [located within him| that is manifested, the specta- 
tor would make ever greater efforts to obtain those objects (which bring 
about the manifestation].!! And if you hold that rosa is manifested (we 
must ask the same question as before]: Is rosa in the spectator himself, 
or in someone else? The same difficulties arise now as arose before. 
Therefore rasa is not perceived (pratiyate), nor physically produced 
(utpadyate), nor manifested (abhivyajyate) by a poem. Rather, poetic 
words are of an altogether different nature from ordinary words, thanks 
to their threefold operation. Their denotative power (abhidháyakatva) 
operates within the limits of the literal meaning; their aesthetic efficacy 
(bhávakatva) operates in the area of the rosas, etc. [i.e., it transforms 
the vibhàvos, etc., into rasa]; and their efficacy of aesthetic enjoyment 

222 [824L 

(bhogakrttva) operates within the sensitive audience. The working of 
a poem consists of these three operations. If one were to claim that 
in poetry denotation alone held sway, then what would differentiate 
Slesa (artistic double meaning) and other figures of speech from such 
devices as the forcible taking of a word in two senses (tantra),'? etc., 
in scientific works? Moreover, the varieties of alliteration would be 
virtually useless. And what purpose would be served by the avoidance 
of such faults as indelicacy of sound (srutidusta)?'? Therefore there 
is a second operation known as the efficacy (bhávaná) of rasa (i.e., 
the ability to create rasa), thanks to which denotation assumes a new 
dimension. A poem's having the efficacy (bhdvakatua) to create rasas 
is nothing more than a poem's power of making the vibhávas, etc., 
universal. Once a rosa has been thus realized,’ its enjoyment (bhoga) 
[is possible], an enjoyment which is different from the apprehensions 
derived from memory or direct experience and which takes the form of 
melting, expansion, and radiance.!* This enjoyment is like the bliss that 
comes from realizing [one's identity] with the highest Brahman,! for it 
consists of repose in the bliss which is the true nature of one's own self, a 
nature which is basically sattva but is intermingled with the diversity 
of rajas and tamas." It is this aesthetic pleasure (bhoga) alone that 
is the major element [i.e., the purpose of poetry] and it is something 
already [eternally] accomplished (siddharüpa).'* Any instruction that 
poetry may furnish is incidental.” !? 

On this subject we may make the following remarks. To begin with, 
there are different opinions among the critics on the very nature of 
rasa. Some, for example (Lollata], say that what is a stable emotion 
(sthàyibhàva) in a former state, being nourished by the addition of 
the transient states of mind (vyabhicdrins), etc., just that much, as 
belonging to the character portrayed (anukárya), is rasa. The rasas 
(are not located in the actor or audience of a drama. However, they] are 
called "dramatic" rasas (nátyarasáh) because they are used in drama. 

[Against this view it may be said? that] a state of mind (cittavrtti) is 
something that underlies a series of properties [i.e., is really a trend of 
mind],?' so what can it mean to say that one state of mind is nourished 
by another state of mind? Wonder, grief, anger, etc., are not gradu- 
ally augmented; [on the contrary, they diminish with timie].?? Therefore 
there is no rasa in the character being portrayed. [On the other hand] 
if rasa were to lie in the actor, he would be unable to follow the tempo 
(laya),?9 etc. Again, if one were to say that rasa lies in the specta- 
tor, how could there be delight? On the contrary, in tragic (karuna) 

§ 2.4L] 223 

performances the spectator would experience only pain. Therefore the 
above theory will not do. What will do? As states of mind are endless 
[chains], it is impossible to imitate them exactly. Moreover it would be 
useless to do so because if we did perceive the exact state, we should 
derive nothing from it, because we should be indifferent (as the state 
would belong to someone else]. 

Therefore?* rosa is an apprehension (pratipatti) of a stable emotion 
whose nature is not exactly fixed. It results from the addition of the 
vibhdvas, anubhávas, and vyabhicárins and takes the form a of relishing 
(&sváda) different from memory because there is a direct object (9o- 
cara) of its perception of the stháyin, e.g., “This Rama [standing before 
me as represented by the actor] is happy."?5 This apprehension of rosa 
depends on the actor and is found only in plays. It requites no other 
basis. But the actor must be thought to be? the character portrayed 
in order for the audience to enjoy the experience. Only this much?" 
and nothing more is required for the aesthetic experience of the rasa. 
Therefore, rosa exists only in the drama and not in the characters to 
be portrayed, etc. This is the view of some. 

Others?? say: The appearance (or semblance, avabhdsa) of a stable 
emotion in the actor, which has been brought about by a set of causes 
such as dramatic representation, etc., is like the semblance of a horse 
drawn on a wall by means of yellow and other pigments. When it 
is relished-by an act of perception, known otherwise as a relishing 
(dsvdda) because it is beyond ordinary experience, it is called rosa. 
And so the expression ndtyarasdh is to be explained as ndtydd rasáA, 
i.e., rasas arising from drama. 

Others, however, say: The vibhdvas and anubhávas are presented to 
the spectator with the help of particular dramatic equipment (visista- 
sámagri — acting, music, dialogue) so as to engage the latent impres- 
sions of the spectator that underlie that mental state which forms the 
stable emotion that is sought to be produced by these vibhávas and 
that is brought within view by these anubhávas.?? The vibhavas and 
anubhàvas when accompanied by this relishing of bliss within the self 
are the rasa. (That is to say,] these vibhdvas and anubhdvas themselves 
are the rasa and the term nátyarosdh means the rasas which are the 

Still others say?? that rasa is the vibháva alone, others that it is the 
anubháva alone, and some that it is the sthayibhdva alone, some that 
it is the vyabhicdribhdva, still others that the combination of these four 
is rasa. Some say that rosa is the character being portrayed. Others 

224 [524 L 

say that rosa is the conglomeration of these five elements. But enough 
on this score. 

Another point is that the occurrence of rasa in poetry is wholly 
analogous to its occurrence in drama.?! Where drama makes use of re- 
alistic style (lokadharmi) and theatrical style (nátyadharmi),? poetry 
uses the styles of direct expression (svabhávokti) and artificial expres- 
sion (vakrokti).?? In both cases rasa is produced in these styles by the 
combination of extraordinary?* vibhdvas, anubhdvas, and vyabhicdrins 
presented in language that is clear, sweet, and forceful.?* Granted that 
the perception of the rosas in poetry is distinct in nature from that 
experienced in drama, because the means differ whereby it is brought 
about, still, the same scheme (sarani) holds for both forms of art.?¢ 

Having arrived at this point, [we can see that Bhattanayaka's] criti- 
cism, pointing to the impossibility of a rasa’s residing in the spectator 
or in someone else, applies only to the first view [that we described 
after Bhattanáyaka's|.?" But in all the views [that followed Bhatta- 
n&yaka's] the unavoidable fact remains that rasa is perceived. For if 
it were not perceived, we could have no dealings with it, just as we 
can have no dealings with a goblin. However, just as we have sen- 
sory, inferential, verbal, intuitional, and yogic perception, all of which 
are undifferentiated so far as being perceptions, but each of which is 
distinct because of differences in its means of production, just so may 
we have this other type of perception that is called tasting (carvaná), 
relish (dsvdda), enjoyment (bhoga), [which is distinct from other types 
of perception,) because its basic components, namely the vibhávas, etc., 
helped by sympathetic response (hrdayasarnváda), etc., transcend [the 
experience of] the workaday world. To say that "rosas are perceived” 
is a turn of phrase as when we say “he is cooking the rice pudding,” 
for the rasa consists in the being perceived [of the vibhàvas, etc.].? 
Relishing (rasand) is a special kind of perception. This perception [of 
rosa] in drama is distinct from every-day cases of inference, although 
it depends on inference in the initial stages [since one first infers, from 
the vibhávas, etc., the stable emotion that is being portrayed]. Simi- 
larly, in poetry the perception of rasa is different from other kinds of — 
verbal cognition (abhidhà, laksaná, tátparya), but in the initial stages 
it depends on direct denotation (abhidhd) as a means [of reaching the 
suggested sense]. 

Accordingly, the case (of Bhattanadyaka] against us is destroyed be- 
cause it never really had occasion to arise.‘ Again, it is a rash state- 
ment indeed to say that the extraordinary deeds of Rama do not win a 

§24L] 225 

sympathetic response from everybody. For minds are characterized by 
a great variety of latent impressions (vdsand). As has been said: “La- 
tent impressions are endless because desire is eternal.” and “Though 
separated by birth, place and time, the latent impressions are unin- 
terrupted because of the correspondence of impressions and memory" 
[i-e., though several lives intervene, impressions still give rise to the old 

Therefore it is now established that there is perception of rasa. More- 
over, tbis perception in the form of aesthetic relishing is physically 
produced (utpadyate). And the verbal operation in bringing about this 
perception is the hinting (dhvanana), the suggesting (vyarjana), of the 
literal sense and denotative words, which is an operation different fom 
abhidhá and loksand. [What Bhattanayaka calls] the poem's operation 
of causing aesthetic enjoyment (bhogikarana) of the rasas is nothing 
other than thé operation of suggestiveness. As for aesthetic efficacy 
(bhávakatva), this too is nothing more than what is included in the 
use of appropriate qualities (gunas) and figures of speech, a subject of 
which we shall speak in some detail [later in this chapter]. What is new 
about all this? And when you say that poetry is effective (bhdvaka) 
of rasas, you have revived through your aesthetic efficacy the theory 
of physical production (utpatti) [which you had hoped to destroy].*? 
Again, one cannot say that in poetry the words alone are effective of 
rasa, for if their meaning is unknown, no rasa can arise. Nor can 
one say that it is the meaning alone, for if the same meaning is ex- 
ptessed in other words, rasa does not arise.* We explained that both 
word and meaning were effective when we said, "In which a sense or 
word suggest that suggested meaning, etc." [1.13 K]. Accordingly, with 
the operation known as suggestiveness serving as means and with the 
qualities, figures of speech, and propriety, etc., serving as procedure 
(itikortavyatà), poetry, which is effective (bhávaka) [of rosas], effects 
(bhavayati) the rosas; and in this three-termed scheme of efficacy (bhà- 
vand as understood by the Mimamsakas) suggestiveness fits in as the 

Again, aesthetic enjoyment (bhoga) is not produced by the words 
of poetry [i.e., the power of aesthetic enjoyment (bhogakrttva) is not 
a third function of poetic words, as BN would have it]. Aesthetic 
enjoyment, which is a melting, expansion and radiance,' otherwise 
known as relishing (dsvdda), comes about rather from the cessation 
of that obscuration [of the true nature of the self| which is caused 
by the thick darkness of ignorance. In bringing this cessation to its 

226 [524 L 

superlative degree it is suggestiveness that should be given the place of 
honor. When rasa has been achieved by means of suggestion, this power 
of aesthetic enjovment inevitably follows. For enjoyment is nothing 
other than the incomparable thrill of delight that arises from tasting 
the rasa. But it is wrong to think that the varieties of relishing are 
fully enumerated by melting, expansion, and radiance, because there 
are innumerable possible variations on account of the endless variety 
[of human character] created by the varying degrees of predominance 
among the components of character, sattva, rajas, and tamas. We 
admit [with Bhattanayaka] that the relishing of rasa bears a family 
resemblance to the relishing of the ultimate brahman. [We further 
admit that] the educative effect (vyutpádano) [of poetry] is different 
from that which comes from scripture through its mandates and fom 
history through its narrations.‘” For in addition to the analogy which 
it furnishes that we should behave like Rama [and not like Ravana], 
it produces in the final result an expansion of one’s imagination which 
serves as the means of tasting the rasas. With this view we find no 

Accordingly, it is established that rasas are suggested and that they 
are enjoyed by their very perception. Now this suggestion can be either 
primary [i.e., the rosa that is suggested can be the primary sense of 
the sentence or stanza) or secondary. If it is primary, it is a case of 
dhvani. 1f not [i.e., if the suggested rosa is only secondary], it is a case 
of a figure of speech such as rusavat, etc. This is what he now says 
in: the main suggested sense, etc. Kept distinct: viz., because of 
their having been determined to be distinct by the reasoning already 
employed [in 1.13a and following passages]. 

1. Preyas: name of a figure of speech also known as preyasvin (3.34 L) and 
preyolorikàro (14a A, et passim). Dandin and Bhàmaha give no definition, 
but their comon example (Dandin 2.276, Bhamaha 3.5) seems to show that 
they understood the figure to be a form of complimentary address. L on 
2.5 ascribes to Bhamaha the view that "preyolarikára is a loving description 
addressed to a god, a king, or a son." See 2.5a L, and note 2. 2. ürjasvin: 
an expression of pride or egoism, Dandin 2.275, Bhàmaha 3.7. Udbhata: "the 
description of bhdvas and rasas that exceed the bounds of propriety owing to 
love, anger, etc.” (4.5 Indurja = 4.9 Vivrt). 3. In Dandin (2.298) and 
Bhàmaha (3.10) samdhsta is the description of a happy coincidence. Udbhata 
completely changed the definition and brought it into connection with rosa: 
“A passage concerned with the cessation of rasa, bhava. or their improper 

824 L] 227 

varieties, in such a way that there is no trace of a new anubháva is called 
samáhita" (4.7 Induraja = 4.14 Vivrti). — 4. Abhinava now takes up the 
question of the nature and genesis of rosa. The question involves him in a 
long criticism of the views of his predecessors. For the views of Bhattanayaka 
and for the meaning of his technical terms bhávanó, bhdvakatva, and bávito, 
see the remarks in the Introduction, pp. 35-36. The passage 2.4 L from this 
point on has been translated in an appendix by Raniero Gnoli, The Aesthetic 
Experience according to Abhinavagupta, 2nd. ed., Chowkhamba, 1968. In the 
same book Gnoli also translates the longer version of the same argument found 
in the Abh., Vol. 1, 2nd. ed., pp. 272. — 5. BN means that Sità is a vib- 
hàva only with respect to Rama, not to the spectator. This holds, of course. 
only so long as rasa is regarded as a physically produced, perceptible, emo- 
tion. If rasa is admitted to be imaginatively achieved (bhdvita rather than 
utpadita), BN presumably would admit what Abhinava and all later critics 
hold, that Sità is a vibhdva with respect to the spectator's rasa. 6. vasand: 
literally, the perfuming of the self by its former experiences, including those 
of preious lives; hence, the latent impressions of the mind, which give it a 
proclivity to particular tastes and sympathies. — 7. Energy (utsaha) is the 
stháyibháva that underlies the rasa of heroism (virya). But how can Rama's 
energy produce any such resa in us? We cannot remember it, for the defini- 
tion of memory in Indian philosophy involves direct experience (anubhava). 
8. The expanded version of this passage in ABh., Vol. 1, p. 278 (Gnoli p. 10) 
reads: na ca Jabdaénumanddibhyas tatpratitau (where tat refers to utsahadi- 
sthdyibhdva) lokasya sarasatà yuktà pratyaksád iva. The point is that these 
various forms of perception lead only to information (jridna), not to rasa, 
which must be experienced through some other pramána, viz., through the 
bhavana of poetry. 9. Abhinava in ABh., Vol. 1, p. 278 (Gnoli p. 10) 
expands the analogy of watching a couple making love by adding: pratyuta 
lajjájugupsásprhádisvocitacittavrttyantarodayavyagratayà kd sarasatvakathápi 
syát, "On the contrary, because one becomes preoccupied with one's own re 
spective emotional reactions such as embarrassment, disgust, or even sexual 
desire, we cannot say that this is an aesthetic experience at all." Abhinava 
makes the same point in ABh., Vol. 1, p. 35. See also DR 4.39 and Avaloka 

10. BN's distinction of utpattt and abhivyakti is presumably the common 
distinction in Indian philosophy. Utpatti is the origin of an entity that was pre- 
viously non-existent. Thus the Naiyàyika says that when a jar is made, there is 
utpatti of the jar. Abhivyakti is the transformation into sensible form of what 
was formerly imperceptible although existent. When a jar is in a dark room, it 
is manifested by the light of a lamp. It seems unlikely in view of what follows 
that BN is using abhivyokti in Abhinava's more restricted sense. namely, the 
manifestation of a suggestion by verbal means. 11. The basic argument 
against abhivyakti is omitted, presumably as being obvious. BP supplies it: 

228 [52.4 L 

there can be no manif tation of rasa in the spectator because he had no rasa 
in the first place, that is, prior to his experience of the play or the poem. The 
argument against abhivyokti that is expressed is an argument against someone 
who supposes that the rasa might indeed be present in the form of the specta- 
tor's latent emotions. These might be roused into manifested, sensible, form 
by the wibhdvas. But then, the hypothesiser should push his analogy further. 
When we are looking for a treasure in a dark room, we want more than to be 
told about a lamp. We want to take the lamp in hand. Spectators would rise 
from their seats and try to carry Sitá away with them. See DR 4.39 and Ava- 
loko thereon: itaresém ostydnurdgdpahdrecchddayah prasajyeron. 12. This 
meaning of tantro is not found in PW. It occurs in Vàmana, KAS 4.3.7, in 
Padmapéda's Paricapádiká, Madras ed., p. 37, line 4, and elsewhere. BP ex- 
emplifies the technique of tantra by an old interpretation of Pán. 13.3. The 
sūtra reads hal antyam, apparently meaning "the final consonant (of a tech- 
nical term of grammar is an exponential marker].” But we cannot know that 
the expression “hal” means a consonant until we are told that its "I" is an 
exponential marker. So Katydyana ( Vart. 5) recommended taking the word 
hal in two ways, that is, by tantra, as (1) the l of “hal” [bere ħal is a tatpurusa 
compound] and as (2) ^al (= all the consonants from h to the exponential 
marker l). The sutra then means: "The | of hal is an exponential marker and 
all the consonants from A to the exponential marker | are used as exponential 
markers when final in a technical term of grammar." This is how to make a 
computer bank, not poetry. — 13. rutidus(a: see 2.11 below. The term goes 
back to Bhàmaha 1.47. — 14. bháuite ca rose: the notion of universalization 
is not present in the term bhdvita itself but derives from BN's explanation 
of how the realization or creation (bhávaná) comes about. — 15. Druti, vi- 
stara, and vikdse. We can point to no technical use of these terms earlier than 
BN, but they occur frequently in later texts. Dhanika on DR 4.43 assigns 
expansion (vistara) to the enjoyment of the heroic, and radiance (vikdsa) to 
the enjoyment of the erotic. Melting (druti) would naturally be connected 
with the rosa of compassion and with love in separation: see 2.8 K below. 
But BP, on the authority of the Kdvyaprakdsasariketa of RAj&naka Ruyyaka, 
identifies the three forms of enjoyment with the three respective components 
of the enjoyer: rajas, tamas, and sattva. On those terms as connected with 
the rasas, see Gnoli, p. 46, and Raghavan, Bhoja, pp. 467ff. 16. BN seems 
to have thought that aesthetic bliss was actually superior to yogic bliss; cf. 
the quotation of BN in 1.6 L. Abhinava in 3.43b L sems to reverse this value 
judgment. 17. Abhinava's summary is so condensed that it almost misses 
the point. From Mammata 4.28, prose following verse quotation 26 (page 90), 
it appears that BN conceived of the bliss of poetic enjoyment as bringing 
about a predominance (udreka) of the basic goodness (sattva) of the soul over 
its adscititious elements of passion (rajas) and brutishness (tamas). For an 
account of the relation of BN's views to the Simkbya. see M. Hiriyanna's 

824 L] 229 

article "Indian Aesthetics" in Proceedings and Transactions of the First Ori- 
ental Conference, Poona, 1922, pp. 246-247 and Raghavan. Bhoja, pp. 4668. 
18. Siddharüpa: The term and its implications are taken from Vedanta phi- 
losophy. Sankara argues that knowledge of Brabman cannot be learned or 
produced or manifested or developed in any way, for knowledge of Brahman 
is Brahman, which is not a sddhyo (that which is to be produced) but a 
siddho. Brahman and Brahman-knowledge are eternally present, the appear- 
ances to the contrary being due to illusion. Just so Bhattanàyaka would take 
aesthetic enjoyment to be an eternal mode of being, which is not produced or 
manifested. In the last analysis he puts the relation between the aesthetic ef- 
ficacy of the poem and the self-realization of the audience outside the relation 
of cause-effect which applies to the phenomenal world. It would be the same 
relation as that between the Upanishads and knowledge of Brahman. Abhi- 
nava in what follows will obj tto this high-Bying mysticism. — 19. Abhinava 
himself adopts this view at l.le L. But cf. 3.10-14f L. The view is not held 
by earber authors of known date, but may be found in the Visnudharmottara 
Puróna 15.2. 

20. BP, followed by Jagannath Pathak in the Chowkhamba edition, as- 
signs the following criticism to Srigankuka — 21. The point seems to be that 
a pure stháyibháva such as grief or anger is never found so that we could 
speak of its being nourished by the transient emotions (vyabhicárins). What 
we find is a complex mental state (cittavrtti), composed at any one time of a 
sthdyibhdva and vyabhicdrins, this complex flowing forth like a river until it 
loses itself with the passage of time. — 22. This passage has been expanded in 
ABh, Vol. 1; p. 272: Sokasya prathamam tivratvam kdlét tu mándyadarsanam. 
A mental state becomes itensified or weakened because of external objective 
stitnulants and not because of other mental states like the vyabhicáribhóvos. 
23. BP takes laya differently: layo ndma nrttagitavádyánàm ekatánarüpam 
sámyam. But we feel that the normal sense which the word laya has in music 
fits the context better. — 24. BP identifies the following view as svamata: 
"his own view,” meaning thereby the view of Saikuka, whose objections have 
just been recorded. Jagannath Pathak also supposes that Sritaükuka's re- 
marks are here continued. The identification seems to me probable but not 
quite certain. The seeing of the actor as “ayam RámoA sukhi” occurs in the 
account of Sa&kuka's view given in Abh., Vol. I, p. 273, and if nothing is here 
said about inference (see note 26 below), the same is the case with the account 
in Abh. The only occasion of doubt is that the analogy of the painted horse, 
which Mammata attributes to Srigafikuka. occurs not in this view but in the 
view which follows. 25. ayam rmah sukhi is explained by BP, p. 185, as 
rémo ‘yam sitévigayakaratimén. 26. Nothing is here said about inference. 
Apparently Srigankuka claimed that the spectator infers the identity of char- 
acter and actor. [t is this hypothesis of inference that brings the sharpest 
criticism on Sankuka in other texts. 27. adah means idam, as often in L. 

230 [$2.4 L 

See Text, p. 160, line 5; p. 239, line 3; p. 258, line 9. — 28. It is difficult to 
identify the person who held this opi ion. According to Mammata, line 8 of 
prose following 4.28 (page 88), the painted horse analogy (citraturaganyáya) 
belongs to Sabkuka. — 29. This is a difficult passage. Tadwbhávaniya means 
vibhávavibhávaniyo: "the sthdyibhdva that is sought to be produced by those 
vibhávas." Tadanubhdvaniya means anubhdvanubhdvaniya: “the stháyibháva 
which is intended to be brought within the purview of the spectators by 
means of those anubhávas." Visistasámagryá samarpyamánam means abhi- 
nayádisámagryá sámájikánám purostád upasthápyamánam. 

30. For the theories mentioned in this paragraph see Raghavan, Bhoja, 
p. 437. As he points out, Jagannátha Pandita also refers to such theories 
31. The preceding views, except for Bhattanàyaka's and perhaps those men- 
tioned in the last paragraph, were developed by commentators on the Natya- 
$ástro and apply strictly only to the theater. Abhinava now makes it clear 
that a theory of rosa must apply to poetry as well as drama. — 32. On loka- 
dharmi and nátyadharmi see the long article by V. Raghavan, JOR 7 (1933), 
pp. 359-375 and JOR 8 (1934), pp. 57-74. Lokadharmi refers to everything 
in the drama that is realistic, ndtyadharmi to those artificial conventions that 
are peculiar to the theater, such as asides that nobody else can hear, mono- 
logues, talking animals, gods on the stage. [o the thirteenth chapter of the 
Nétyasdstra (Vol. 2, beginning p. 214) Bharata gives a long list of the charac- 
teristics of each type. One may note an interesting verse that Abbinava quotes 
from his teacher on this passage. The verse reads as follows: yad atrásti na 
tatrásya kaver varnanam arhati / yan ndsambhavi latro syát sambhavy atra tu 
dharmatoh // "Not everything that is in the world deserves to be described 
by the poet in his plays. And what is not possible in the world may oc- 
cur in plays quite properly.” In ABA., Vol. 1, p. 269, Abhinava again voices 
his opposition to strict realism. 33. These are parallel terms, svabhàávokti 
corresponding to lokadharmi and vakrokti to nátyadharmi. Thus Abbinava 
is using the terms in their widest sense. The basic distinction is made by 
Bhamaha 1.30 and by Dandin 2.363; see also Udbhata 3.8-9 (Vivrti). The 
most valuable discussion of svabhávokti is found in the first chapter of Kun- 
taka's Vokroktijivito, but his use of terms is peculiar. For Kuntaka vakrokti is 
almost what dhvani is for Ananda. See the valuable article of V. Raghavan, 
"History of Svabhavokti” in Some Concepts. See also 4.7 A, note 1. 34. BP 
takes the word-stem alaukika to modify sabda: "presented in a language that 
is more than normally clear, sweet, and forceful." But the vibhávas are called 
lokottararüpa on the very next page (Text, p. 187, line 5), so they are probably 
the recipients of the synonymous epithet here. The intention, I suppose, is 
that the characters and emotional situations of a work of art are in themselves 
more striking than those that we know from the workaday world. 35. pra- 
sanna, madhura, ojasvin: these are the fabdagunas mentioned by Bhamaha, 
Dandin, and Vàmana. Anandavardhana completely altered the older teaching 

$2.4 L] 231 

by bringing them under the system of rasas. For him the gunas are the 
properties of the rasas; see 2.7 below. Instead of the ten gunos mentioned by 
the older writers, Ananda accepts only the three mentioned here. — 36. The 
difference in perception (pratiti) results from the difference in presentation 
(upasthdpana), the play being presented to the sight and the poem to the ear. 
But the constituents of rasa in the form of the vibhávas, etc., are the same 
in both. Furthermore, the same scheme of natural and artificial is found in 
both. 37. le., Lollata's view, because Lollata belongs to the utpattipaksa. 
He held that the rasas are physically produced. Others held that they were 
manifested or relished. 38. The insistance that rasa is perceived is directed 
against Bhattandyaka, who said, just above, that it was not. But Bhatta- 
náyaka would not on that account have admitted that rosa was nonexistent 
like a goblin. Obviously he meant that perception (pratiti) is too physical 
à concept to express the manner in which rasa is experienced. So Abhinava 
goes on to give "perception" a wider area of application. — 39. Although 
we say odonam pacati: “he is cooking the rice pudding,” a more accurate 
description would be tanduldn pacati, "he is cooking the rice grains,” because 
the pudding is the result of the activity while the grains are the object on which 
the activity works. Similarly, instead of the phrase rosáh pratiyante "the rasas 
are perceived,” we would be more accurate to say vibhávádi pratiyonte “the 
vibhdvas, etc., are perceived,” for the vibhávas, etc., form the object of our 
perception; the rasa is the result. Or one may even say that the raso is the 
process of perceiving itself. This qualification of his term perception leaves 
very little distance between Abhinava's view and that of BhattanSyaka. 

40. There is a similar passage in ABh., Vol. 1, p. 277, which Gnoli 
misunderstood: tatra purvapakso ‘yam bhattalollatapaksánabhyupagamád eva 
nábhyupagata iti taddüsonam anutthánopoatom eva, "This purvapakya [viz., 
the view of Bhattandyaka that has just been set forth] is unacceptable because 
we do not accept the view of Bbattalollata. (lf we did accept Lollata’s view 
of the physical production of rasa, the view called utpattipaksa, then Bhatta- 
náyaka would have a good argument against us; but we do not| So the 
criticisms made by the pürvapaksa are destroyed, never having had a chance 
to arise. 41. Yogosutra 4.10 and 4.9. In ABh., Vol 1, p 282, Abhinava has 
an interesting passage where he claims that the nine stháyibhávos are present 
in all human beings, although some predominate in certain people and oth- 
ers in others. He ends by saying: na Ay etac cittavrttwuásanásünyah práni 
bhavati. In the course of our beginningless journey through this universe we 
have experienced al! emotions. Thus nobody fully aware of his own humanity 
can fail to be moved by another person's experiences. Again, on p. 283 Abhi- 
nava quotes a line from the Yogabhdsya of Vyasa (2.4): na hi cattra ekasyám 
striyám rakta ity anydsu viroktah, "the fact that Caitra is in love with one 
woman does not mean that he is out of love with others." This is not meant 
hurnorously. but is intended to show that beneath the particular emotions 

232 [$24L 

which we manifest there lies a latent capability of many others. 42. This is 
a practical man's argument against a mystic. What else than cause and effect 
can Bhattanáyaka be speaking of when he says bhávite ca rose "once a rosa 
has been realized”? The identical argument is used against Sankara by the 
ritualists and by Bhaskara: we are not in mokga now; if we get there, some- 
thing must produce the change. See note 18 above. — 43. Abhinava's point 
is that BN said that there are three functions of words; he made no mention 
of the part played by meanings in producing rasa. This seems to be unfair of 
Abhinava, as he too says over and over that vyarijana is a fabdavyápáro. Why 
criticise Bhattandyaka for an imprecision of terminology of which Abhinava 
is equally guilty? 44. Cf. what Ananda says on 3.16m. — 45. Abbinava 
here reduces Bhattaniyaka's term bhdvand to the technical sense given it by 
the Mimāmsā. [n the Mimàmsà bhdvand is an efficacy residing within the 
verb of a Vedic sentence which explains how that verb can bring an actor to 
pursue a given aim by certain means and procedures. For example, we are 
given the sentence jyotistomena svargakámo yayeta "one who is desirous of 
heaven should sacrifice with a jyotistoma sacrifice." In the verb yajeta “one 
should sacrifice” there is said to reside bAdvand (we may here overlook the 
refinement of Kumarila who speaks of two sorts of bhdvand). This bhdvand 
is connected with three terms or factors: (1) a sádhya, an objective aimed at 
by the action, (2) a sédhana or karana, the means leading to that objective, 
and (3) an itikartavyatà, a procedure to be followed in reaching the objec- 
tive. These terms answer the questions: “What does it effect (kim bhdvayet)? 
With what does it effect it (kena bhávayet)? How does it effect it (katham 
bhduayet)?” In the case of the sentence “one who is desirous of heaven should 
sacrifice with a jyotistoma sacrifice," heaven is the sddhya, the jyotistoma is 
the sádhana, and the performance of the minor sacrifices praydja, anuydja, 
etc., is the ittkartavyaté. Abbinava now argues that in poetry or drama rosa 
(or rasásváda) is the sádhyo, dhvanana or vyarijanavyápára is the sádhano, 
and gunálankároucityádi is the itikartavyatà. By reducing Bhattanáyaka's 
term bAdvand to its position in the Mimamsa paradigm, Abhinava claims 
that it implies nothing more than is already furnished by his own theory. It is 
Abhinava's "suggestion" (dhvanana) which is the sádhana by which rasasváda 
is achieved. — 46. A few lines farther on Abhinava will refuse to limit aes- 
thetic enjoyment to this threefold description (melting, expansion, radiance). 
He is here simply using Bhattandyaka's description to show that even under 
Bhattanáyaka's definition aesthetic enjoyment should not be regarded as a 
power of words. It is rather a state of the audience that comes about after 
the words and meanings have produced rasa. 47. This admission falls a 
good bit short of BN's view. BN had said. “Any instruction that poetry may 
furnish is incidental.” Abhinava admits that enjoyment is the main goal and 
that the instruction of poetry is different from the instruction of other types 
of literature. But the instruction given by poetry is not without importance 

$2.5 L] 233 

to him. It bas a moral value. Furthermore, inasmuch as it trains us to 
experience aesthetic bliss, it may even be said to be spiritually instructive. In 
this way Abhinava is able to make enjoyment and instruction come to much 
the same thing. In commenting on 3.14 he will say: na caite pritivyutpatti 
bhinnarüpe eva, dvayor apy ekavisayatvót. 

K But where a rasa, or the like, is subordinate and the main 
purport of the sentence lies elsewhere, then it is my opinion that in 
that poem the rosa, or the like, is only a figure of speech. 

A Although others have defined the domain of the figure of 
speech known as rosavat, nonetheless my own position is this: if in 
a poem some other sense is principally conveyed and is the main pur- 
port of the sentence, to which purport a rosa, or the like [e.g., bhàva, 
rasábhása, etc.) is subordinated, then the rasa, or the like, comes under 
the domain of a figure of speech. 

L {Commentary on K:| Lies elsewhere: that is, in a rosa or 
something like a rosa, or in a mere fact, or in something peculiarly fit 
to form a figure of speech. My opinion: This shows that the author 
holds back other views as being faulty and is first giving his own view 
as being worthy of acceptance. 

[Commentary on A:] Nonetheless: for the province [of the figures 
of rasavat, etc.) as propounded by others is not reasonable, as will be 
shown later (in Ánanda's comment on this Káriká|. If in a poem: 
to clarify this loosely constructed sentence, it should be interpreted as 
follows: In that poem in which a rasa, or the like, as already men- 
tioned, is subordinated, and some other sense is the main purport of 
the sentence— understanding the word "and" in the sense of “but”— 
that rasa, or the like, which though subordinated is yet connected with 
the poem,' should be considered as falling within the province of the 
rasddi figures of speech, called by the terms rasavat, etc." The upshot 
of this is that a rasa, or the like, deserves the name of a figure of speech 
only when it is subordinate, not when it is otherwise. 

234 [$2.5L 

1. Thus Abhinava takes the word tasya in tasya córigabhüta (Text p. 191, 
line 4) as construing with yasmin kávye rather than with vákyàrthibhutosya. 
2. Tbe rusádi figures of speech are preyas, rosavat, ürjasvin, and samáhito. 

A Asan example of this, in passages of flattery, even though a 
figure of complimentary address (preyolarikdra) is the main purport of 
the sentence, a rasa, or the like, may be found as subordinate.! 

1. This sentence presents a difficulty. [t seems to speak of an alañ- 
kāra as forming the vàkyórtha (the main purport of a sentence), which is 
an impossibility, for an ornament (olarikára) must ornament the purport; it 
cannot be that purport. Abhinava will offer three solutions of the difficulty. 

L He gives an example for this statement: As an example of 
this. “This” refers to subordination. His meaning is: as in the example 
about to be given, so in other places also. 

If we follow the doctrine of Bhamaha,' we will take the passage cátugu 

drSyate all as one sentence, meaning: "even though preyolarikara 
is the main purport, (as it is] in passages of flattery, a rasa or the 
like may be found to form a part”; for Bhamaha has said, “preyolan- 
kāra is a loving description addressed to an elder, a god, a king, or 
a son."? In this interpretation, the word preyolankdra [in A will be a 
bahuvrihi compound and so) will mean: "that [passage] in which a very 
dear person (preydn) is an ornament."? Such a preyolarikára can be 
ornamented. It will not do to take the main purport itself to be an 
oranment (alarikàra). 

Or, we may take "the main purport" to mean simply the most im- 
portant element of the sentence, that is to say, the element that is most 
strikingly beautiful. 

Or, we may follow the doctrine of Udbhata and split the sentence in 
two. [First we will read] cátusu vakyarthatve preyolankdrasyapi visaych, 
taking cdtugu vàkyárthatve as locative absolute (visayasaptami), (trans- 
posing the word api] and supplying the word visaya from the previ- 
ous sentence. This will mean: "where flattery is the main purport of 

$2.5b A] 235 

the sentence, there is scope also for preyolarikóra." For according to 
Udbhata any bháva* can be called a preyos, for affection (preman) is 
used to cover all the bhdvas. The sense of the word api is that not only 
is there scope for the figure rasavadalarikdra, but for the figures preyos, 
etc., too. By the word rasavat and the word preyas all the (four) alar- 
kéras beginning with rasavat are included. This is just what he says in 
that [portion of the split sentence] which now follows: rosádayo 'riga- 
bhütà drsyante, to which we must supply uktavigaye: "the rasas, etc., 
when subordinate, are found to have the same scope.” 

1. So also, just below, "or if we follow the doctrine of Udbhata." One 
must not translate, “According to the opinion of Bhàmaha ... the sentence is 
to be taken as one,” for it is out of the question that such early authors could 
have seen and commented on the words of Ananda. 2. The text of Bhimaha 
as we now have it says no such thing, although the example of preyolarikáro. 
that Bhámaha furnishes at 3.5 does not disagree with such a definition. The 
absence of the quotation in Bhámaha has already been noticed by Kane, 
HSP, p. 82. But we cannot explain the absence, as Kane was inclined to do, 
by impugning our present MS of Bh&maba's verses, for the quotation is not in 
verse but prose. Possibly Bhamaba wrote a urtti on his verses which is now 
lost, but if so one would expect other quotations from it to survive. Nearly 
the same words as in the present quotation occur in R&janaka Tilaka's Vi „ti 
on Udbhata 4.2: rotir iha devagurunrpadiviyayé grhyote, kántávigayáyàs ta 
rateh sücone rasavadalarikdre vaksyate. — 3. Just such a meaning is given to 
preyas by Indurája in his comment on Udbhata 4.2 (p. 51, line 27): preyah- 
,fabdavácyena priyatarena ratydlambanena vibhdvena. 4. For bhdvdlarikdro, 
we must read bhdvud and omit alarikéra. See Udbbata 4.2. 

A This use of rasa, or the like, as a figure of speech can be either 
pure or mixed. An example of the first is: 

Why do you laugh? You will not get away again 
now that I have finally caught you. 

Pitiless man, what is this strange love of travel 
that drove you from me? 

So speak your enemies' wiv 

[52.5b A 

clinging to the necks of their husbands in their dreams, 
only to weep aloud 
when they wake to the empty circle of their arms.’ 

Clearly this is an example of the figure of speech known as rasavat, for 
pure karunarosa (the aesthetic experience of compassion) is subordi- 
nated [to the predominating eulogy]. And thus in si ilar cases it is 
clear that other sentiments can be subordinated. 

1. The verse occurs also in SubAA., 2570. Its author is unknown. I have 
followed Abhinava in the interpretation of riktabdhuvalaya. Another possible 
interpretation would be to take valaya not as a metaphor but simply in the 
sense of bracelet. The women's arms would be devoid of bracelets, as they are 
now widows. A and L speak of only two instances of dhvani in the verse: the 
suggestion of karunarasa, which is subordinate and which is therefore a case 
of the rasavat figure, and a vostudhvani, viz., the glory of the victorious kiog 
(nrpatiprabhávo), which is predominant. In commenting on similar stanzas 
commentators sometimes speak of the vastudhvani as leading to a third type 
of suggestion, namely the poet's love of the glorious king (nrpativisayarati- 
bhdva); cf. 2.5d L, note 1. 

L Pure means that there is no mixture with another subordi- 
nated rasa, or with any other figure of speech. Mixed means that 
there is a mixture. 

As a dream comes into being in the likeness of what one has experi- 
enced, a wife [here] sees her husband laughing in her dream. "You will 
not get away again,” that is, now that ] know your unfaithful nature, 1 
will not free you from the noose of my arms. This explains the "empty 
circle of their arms" later in the verse. [t is only natural to scold a 
lover who is received back; so she says, "Pitiless man, what is this ... 
that drove you." She means that she never nagged him even when he 
called her by another woman's name. "In their dreams" refers to their 
talking in their sleep.? "Dreams": the plural shows that this happens 
again and again. This is what the wives of your enemies say, that is, the 
wives of the kings hostile to you,? while their arms encircle the necks, 
that is, are tightly entwined about the necks of their beloveds. Just as 
they are in that condition, they awake to find the noose of their arms 
turned into a vacant circle, at which they cry aloud, with full throat. 
In the stanza the greatness of the king [which is the main purport of 
the sentence] is beautified by our relishing the flavor of compassion 

$2.5b L] 237 

(or, tragedy, karunarasa), whose sthdyibhdva is sorrow that is stimu- 
lated by the vision of the dream. Here compassion unmixed [with any 
other rosa or alarikáro] forms the figure of speech. The main purport 
of this sentence is not baldly stated, without any embellishment, as 
for example, "You have killed your enemies.” Rather, the purport of 
the sentence is beautified and the beauty is due to the experience of 
compassion [used as a subordinate element].* Just as* an object, such 
as a face, can be embellished by another object, such as the moon, for 
it appears with greater beauty by having the moon as its simile, so also 
a fact [such as the king's greatness], or another rasa, if embellished by 
a rosa used as a subordinate element, will appear with added beauty. 
So what objection can there be to our using a rasa, like any other ob- 
ject, as an ornament (alarikára)?* But tell us, an objector may say, in 
precisely what manner the matter in hand (prakrtdrtha)’ is beautified 
by rasa. Well; (we might ask him the same sort of question:] just how 
would it be beautified by a simile? If he says that the matter in hand 
would be compared to something else by the simile, we can say the 
same of rasa: the matter is made more tasteful (sarasikriyate) by the 
rasa. This is obvious. Therefore, what some have asked in the form 
of an objection, namely, “Just what among the vibhdvas, etc., can be 
beautified by a rasa?” is answered by our not accepting the premise.? 
For it has already been stated that it is the matter in hand that is 

He now shows that this [subordination of rasa to the main purport] is 
often found in literature: and thus. In similar cases means in such 
cases as those where the greatness of a king, etc., is being described, 

1. Perhaps she remembers a scene such a9 that depicted in the verse 
quoted at 2.18-19e A. 2. svapnáyita: talking in one's sleep; the more usual 
verb is utsvapnáyate, cf. Mélavikégnimstra 4.15.30. 3. Abhinava is explain- 
ing the use of te in the verse; it is sasthi sambandhe. — 4. It is a pity that 
Abhinava does not really explain here, or at 3.20c, d L, which deals with the 
Same subject, how the relish of compassion makes the main purport of such 
verses more beautiful. To judge from what follows, the problem troubled him. 
Most readers will agree tbat the stanza is beautiful. In HOS 44, pp. 372-373, 
1 have spoken of the cruelty of such verses, but also of "their strong poetic 
effect." But it is hard to explain the effect without admiting that the compas- 
sion is really more important in such verses than the glory of their nameless 
kings. 5. Read yathd vastrantaram for tathā vastvantaram and remove the 
danda aker cárutayávabhásanát. — 6. In this respect a rasa is not different 
from an upamá or other figure of speech. 7. prakrtdrthah, like prostuto 

238 [$2.5b L 

'rthah just below, is used as a synonym of the pradhdno vákyárthah of 2.5 K. 
8. We do not accept the fact that the vibhávos, etc., are beautified by rasa. 
The vibhávas, etc., are the means for achieving rasa. What is beautified by 
the rosa in the figure rasavat is the matter in hand or the main purport of 
the sentence 

A An example of mixed rosa, or the like, when it is subordi 

The women of the Triple City wept from lotus eyes 

as Sambhu's arrow-Bame embraced them; 

but still, though shaken off, the fire caught their hands, 
though struck, did pluck their garments’ hem, 

denied, it seized their hair, and, scorned 

like lover who has lately loved another, lay before their feet 
May this same fire burn away your sins.’ 

Here the [favor of love in] separation due to jealous anger (irsyd- 
vipralambhasrigdra]) together with the ambiguities (slesas) is sub- 
ordinate to the main purport of the verse, which is the extraordinary 
power of the enemy of the Triple City [i.e., Siva]. It is in such an area 
that the figures of speech like rasavat, etc. properly belong. That is why 
there is nothing wrong in including two [usually contradictory] rasas, 
namely, compassion (karuna) and love in separation, in one stanza, pre- 
cisely because they are both subordinate. For when rasa is the main 
purport of the sentence, how can it be a figure of speech? Everyone 
knows that a figure of speech is meant to add beauty to a poem. Surely 
a thing's own self cannot act as its own beautifier. 

1. This famous verse is usually attributed to Amaru, in whose collection 
it occurs as the second benedictory stanza. But Amaru is likely to have 
borrowed it, perhaps from Bana (so SRK 49). It describes the destruction of 
the demon city Tripura by the god Siva. an act that the demons themselves 
had long before prayed for. The act, although painful in immediate effect, is 
regarded as one of purification and mercy. It is here likened by sugestion to 
à lover's insistent embrace of a jealous mistress. The verse is quoted again 
and again in the critical literature; cf. Kosambi's apparatus, SRK ad loc. I 

82.5c L] 239 

have discussed the meaning of the verse, HOS, Vol. 44, pp. 21-22. Ananda 
will have more to say of it under 3.20c. — 2. The $lesas in the verse are not 
Sabdaslesas (puns), but arthaslesas, that is, adj tives (participles) that can 
apply to two sorts of object, fire or a lover. 

L (Commentary on the verse:] “Shaken off” (ksipta) in the case 
of the lover means rejected; in the case of fire, physically shaken. When 
it is said that the lover was scorned (avadhüta), it means that he was 
not greeted with an embrace in return. In the case of the fire the same 
word means that the fire was fragmented (visarárükrta) through the 
shaking of the entire body. Their eyes were filled with tears in the one 
case because of’ jealousy and in the other because of dispair. 

(Commentary on A's comment:] The love in separation because of 
jealousy that is suggested by the simile “like a lover," which simile is 
in turn supported by ambiguities (viz., ky:pta and avadhüta), is sub- 
ordinated together with tbe simile and ambiguities and not by itself 
alone. It is subordinated to the main purport, namely the greatness 
of Siva. Although the relish of compassion (karunarusa) is in fact also 
present in the stanza, inasmuch as it is not involved in our perception 
of the beauty of this [love in separation], Ananda has said that this 
[relish of love in separation] is combined only with ambiguities and has 
nót said that it is combined with karunarasa.' As this matter [namely 
the subordination of the figures rasavat, etc.) was not thought of by 
previous writers, he emphasizes it, saying: only such. That is why: 
sc., because [the relish of love in] separation is only an ornament in this 
verse and not the main purport of the stanza. Nothing wrong: If 
either of the two rasas [viprolambhasrrigára or karuna| had been pre- 
dominant in the verse, the second rasa could not have been combined 
with the first. For love in separation, whose primary emotion is love 
(rati) and which consists in a hope for reunion, is in contradiction to 
tragedy (karuna), whose primary emotion is sorrow and which consists 
in the despair of reunion. 

Having thus demonstrated the [possible] co-presence (of two opposed 
rasas] in the course of treating the expression alarikdra [used in 2.5 K 
“alarikdro rasádir iti me matih”), he now explains the intention behind 
his use of the emphatic word] “only” (eva) [in his commenti by saying: 
for when, etc. 

240 [$2.5cL 

1. The word tat in the compound tat-córutva refers, as BP correctly 
states, to the irgyávipralambha mentioned in the preceding sentence. Abhi- 
nave is not denying tragedy or the aesthetic relish of compassion (karunarosa) 
in the stanza. How could he, when Ananda admits it in his next sentence? He 
is simply making it clear that the karunarasa does not beautify the srrigdra- 
rasa. The two are held distinct and both are in subordinate positions. Hence 
they can both be contained in one stanza; cf. 3.20 K together with A's com- 
ment. The reason that they cannot be combined in rasadAvani is that in 
rasadhvani one would have to be subordinated to the other, in which case 
their contradictory natures would clash. 

A Here is a summary of the matter: 

What makes any figure of speech a figure of speech is the fact that it is 
introduced in dependence on a ruso, bhóva, or the like, which serves as the 
purport of the sentence. 

Therefore, wherever a rasa or the like forms the main purport of a 
sentence, that is not to be included under the domain of the figure of 
speech rasavat or any other, but must be considered to fall under the 
domain of suggestion (dhvani) itself. Of this [suggestion], simile and 
the like act as the ornaments. On the other hand, wherever anything 
else predominates as the meaning of the sentence, and where rosa or 
the like contributes to its beauty, that is where rasa or the like acts as 
a figure of speech. Thus the domains of suggestion and of such figures 
as simile and rosavat are distinct. 

. L Any figure of speech: e.g., simile, etc. The sense is this: 
the function [viz., ornamentation] that makes a simile a figure of speech 
works equally with the rasas, etc., [making them figures of speech when 
they serve in the same function). There must always be something else 
to be ornamented. Now this something else, even if it be only a fact 
(vastumátro), if it develops into a vibháva, etc., may furnish a rasa or 

$2.5d L] 241 

the like as the purport of the sentence.! And so everywhere it is rasa- 
dhvani that is the soul or essence. This is what has been said in the 
words a rasa, bhava, or the like, which serve as the purport of 
the sentence. 

Of this: viz., of that which is in a dominant position and is the soul 
lof the poem]. That is to say, although it is the literal meaning [of a 
sentence] that is ornamented by simile, nonetheless, the oranmenting 
of this literal sense consists in its being endowed with the capacity to 
convey a suggested sense, and so in reality it is the soul of dhvani? 
that is ornamented and not the literal sense. For it is the soul that 
is ornamented by bracelets, armbands, etc., which are worn on the 
body, for these ornaments indicate what is likely to be in the thoughts 
(of the wearers|.? Thus, an insentient corpse, even if provided with 
earrings, etc., is not beautiful, for there is nothing (sc., no soul] to be 
ornamented. And the body of an ascetic adorned with bangles, etc., 
would look ridiculous because of their inappropriateness to that which 
is to be ornamented. Nor is anything inappropriate to a body; so it 
must really be the soul that is ornamented. This agrees with what 
people feel. They say “I am adorned” (and not “My body is adorned.”| 
Where rasa or the like acts as a figure of speech: [the Sanskrit 
phrase] exhibits two genitives that are not in apposition. The meaning 
is: that alone is the province of the functioning as a figure of speech 
of rosa and the like. [n consonance with this, one should explain in 
similar manner the earlier passage in the Vrtti (Text, p. 198, line 1, na 
rasáder alarikárasya visayah], understanding it to mean: “that is not 

“the area of an act of beautifying, of which act rosa or the like is the 

Thus: that is, by distinguishing their spheres as we have done. Such 
figures of speech as simile: [n cases where rasa is that which is 
ornamented [i.e., where rosa is predominant], and where there is no 
other rasa subordinated, i.e., no rasavadalarikára, there such figures of 
speech as simile are pure [i.e., are not mixed with any second figure of 
speech]. Therefore the domain of simile, etc., is not eliminated by the 
figure somsrsti* 

And rasavat: the term here includes the figures based on bháva and 
the like [viz., rasábhása, bhdudbhdsa, and bhávaprasama when used as 
an ornament], that is, the figures called preyasvin. ürjasvin, and samá- 
hita [as well as rasavat]. 

Here is an example of a bhdva used as an ornament without any other 
figure of speech or rasa mixed therewith: 

[525d L 

Your foot, O Mother, 

with its sole as soft and red as a lotus petal 

sings with the jingling of its anklets 

like the gabbling of walking geese. 

How did this foot, 

stamping on the head of the buffalo demon, 

grow as ponderous as the Golden Mountain? 

n this example the main purport of the verse is praise of the Goddess, 
jut the emotions (vyabhicáribhávas) wonder and speculation act as 
sources of beauty. As they are subordinated to the main purport, this 
s a case of bháválarikára." 
An example of rasábhàása used as an ornament is a prayer of my own 


If all poetic qualities 

and every ornament of speech 

were to embellish you, my Muse, 

you would not show so fair 

as by taking your words whichever way they come, 

if thus they may delight your heart's love, Siva. 

So only will you be beyond compare. 

n this example the main purport of the sentence is that the highest task 
»f speech is to praise God. This fact is made to appear more beautiful 
>y an appearance of the relish of love (srrigárábhása) together with the 
igure of ambiguity (slega). It is not the normal relish of love because 
‘be beloved is without qualities and without ornaments. For Bharata 
1as said that jrrigára "deals with beautiful clothes and with young and 
1oble people” (BANS 1, p. 300). 
An example of bhdvdbhdsa as subordinated to the main sense of a 

voem is the following: 

May he protect you, 

upon whose killing of their kinsmen 

the surviving demons tremble 

to see the dark color 

even in the collyrium used by their women 

to paint their lovely waterlily eyes.” 

n this verse we have an example of bhávábhása because the accom- 
vanying emotion (vyabhicdribhdva) of dread (trdsa) [which is subor- 

$2.5d L] 243 

are terrifying by nature.'? Similarly, an example of bhdvaprasama [as 
subordinated to the main purport] may be supplied. 

1. Abhinava here has in mind the third sort of suggestion that we men- 
tioned in the footnote to 2.5b A. The BP gives an example. In the verse 
"Why do you laugh?" the fact that is ornamented is the power of the vic- 
torious king. But this fact itself may assume the form of a stimulant to the 
emotion (bhdva) of love for that king in the heart of the poet. One might 
better choose as an example "The women of the Triple City." The power of 
God is a vastumdtra, but one will readily grant that for many persons this 
fact stimulates their love of God. 2. dhvanyátmá: for the ambiguity of 
the term see 2.2 K, note 1. BP glosses the present occurrence as dhvanirüpa 
átmá& What is meant is the suggested sense in the form of a raso, etc.; but 
by calling it dhuanydtma Abbinava is able to give the analogy which follows. 
3. BP. Thus, the necklace and bracelets worn on the body of a young persoa 
indicate the likeliness of affairs of tbe heart, while the staff and orange robe of 
the ascetic indicate his distaste for worldly pleasures. — 4. Abhinava wishes 
to make the two passages parallel. But surely in rasdder alarikdrasya Ananda 
intended to use the words in apposition: "of the figures of speech called rasddi 
(i.e.. of rasavadolarikáro, etc.)." 5. The reason for this comment of Abhi- 
nava's on upamddindm is that, strictly speaking, tbere is no igure upama in 
the verse "The women of tbe Triple City" that has been under discussion. 
The upamá in that verse is associated with rasavadalarikdra, so tbe figure 
should be known technically as samsrgti (association); see 1.13i L, note 1. 
At this rate it might be thought that samsrsti will preempt tbe whole habi- 
tat of upamá and of rasavadalarikdra, and that Avanda should have written 
dhvaneh samsrstes ca vibhaktavisayotà bhavoti. But no, says Abhinava. There 
are other instances wbere upamá in a pure form may ornament tbe predomi- 
nant meaning. So Ananda is justified. 6. Abhinava bere constructs a more 
logical system of tbe rasddy clarikdras than we find in the older critics. Orig- 
inally the four figures rasavat, preyasvin, ürjasvin, and samáhita formed a 
very disparate collection. In Bhámaha and Dandin tbey represent, respec- 
tively, an emotional passage, a complimentary address, a passage of boasting 
or pride, and a happy coincidence. Udbhata was the first to bring some sort of 
logical order to the group, defining rosavat as a passage with strong $rrigára- 
rasa, preyas or preyasvin as a passage where some other rasa was involved, 
ürjasvin as a passage showing rasdbhdsa or bhávábhása, and samóhita as a 
description of bhdvapragama. Abhinava now takes the five concepts listed in 
Kárikà 2.3, viz., rasa, bháva, rasábhása. bhávábhàsa, and bAdvaprasanti, and 
in effect sets up five alarikdrus to cover tbe five cases where one or the other of 
these concepts plays a subordinate role in a sentence. — 7. According to the 
older scheme of alarikdras the figure here will be called preyasvin. 8. Abhi- 
nava's point is that srrigáro, suggested through slesa, here serves an ulterior 

244 (§25d L 

purpose. Hrdayavallabha means "one's heart's love” and Siva will suggest 
the ndyaka because of the masculine gender, whereas vänt (speech, which we 
have translated as “Muse”) will suggest the ndyikd because of the feminine 
gender. Rorijayeh in the case of the lover will mean not to delight but to 
make love to. The $rrigára that is here suggested by ambiguities and that in 
turn ornaments the religious statement of the verse is not $rrigára precisely as 
defined by Bharata. Accordingly, Abhinava calls it $rrigárábhása. The trans- 
lation “false love” here would misrepresent Abhinava's intent and one should 
not press his remarks on 1.4g L (anaucityena taddbhdsah) too far. There is 
nothing inappropriate here in the relation of the poet's muse to God. The 
Gbhdsatva comes simply from the fact that the relish of this love falls outside 
the strict definition of srrigdrarasa. God, as conceived in Abhinava's Saiva- 
Vedanta theology, is essentially without qualities. — 9. The black collyrium 
reminds them of the black-complexioned Krishna. The metaphor in “waterlily 
eyes" (nayanotpala) derives from the color of the dark blue waterlily, to which 
the color of the pupil is likened. The verse is quoted also in ABA. I, p. 297 
It appears in SubhA. as No. 32, ascribed to Candaka. 

10. The figure will therefore be ürjasvin according to the definition of 
Udbhata; see note 6 above. 

A Should one urge that figures like rasavat occur only when 
the purport of the sentence deals with sentient creatures, this would 
amount to saying that such figures of speech as si ile bave a very 
small domain, or have no domain at all. The reason is that when the 
situation of something that is not a sentient creature forms the purport 
of the sentence, we shall find that in some way or other there is some 
connection with the activity of a sentient creature. If you argue that, 
regardless of this connection, wherever the literal sense of the sentence 
concerns that which is not a sentient creature, this is not an instance 
of rosavat, your argument will amount to.saying that great works of 
poetry, the very mainstay of rasa, are devoid of rasa.' 

1. The Sanskrit passage rendered in this paragraph is puzzling enough 
to have made Jacobi suggest that a portion of it (from tarhi, Text, p. 200, 
line 1, through satydm api tasydm, line 3) is an insertion. But it is clear that 
Abhinava read the passage as we have it and the passage as a whole becomes 

$2.5e L] 245 

clear if one bears in mind that the objector belongs to an old-fashioned school 
of critics whose view of rasa is radically different from Ananda's. To the 
objector there is no such thing as dhvani. Accordingly, he makes no distinction 
between rasa and rasavadalankàra. [f the sentence meaning contains rasa, 
bhàva, etc., we have a case of rasavadalarikdra. If it does not, we have the 
domain of the other olarikáros such as simile. Now as the criterion by which 
to distinguish these domains, the objector first proposes cetanaurttdnta. If the 
purport of the sentence concerns a sentient creature, we have rasavadalankara; 
if it concerns anything else, we have simile, etc. Ananda replies that this will 
leave no domain for simile, etc., for we can always show some connection 
between the purport of the sentence and sentient creatures. The objector 
then refines his criterion. He will disregard these possible connections: if 
the sentence meaning is prima facie concerned with what is not a sentient 
creature, the domain belongs to simile, etc., not to rasavat, The answer to 
this argument is that great passages of poetry, which everyone recognizes as 
the paradigms af rasa—for example, the description of the oncoming season 
of rain in the Ramayana, or Purüravas' apostrophes to nature in the mad 
scene of the Vikramorvasiya—will lack rasa by the objector's criterion. This 
is because the objector's theory can admit rasa into a poem only in the form 
of rasavadalankara and by his last refinement he is excluding the descriptive 
passages of the Ramayana, etc., from that figure. 

L Should one: with these words he introduces the refutation 
of the opponent's view that was hinted at by the words “my opinion" 
lin Kárika 2.5). The opponent's view is this: as a rasa is a mental state, 
it cannot belong to non-sentient objects.' So there is no possibility of 
rusavadalarikára or the like in descriptions of such objects. Accordingly, 
the sphere of simile, etc., is different from the sphere of rasavad, etc. 
Our author refutes this. This would amount: that is, because of the 
preceding statement. But the opponent may reply, "It has [already| 
been said [by us] that the descriptions of non-sentient objects is the 
sphere [of simile and the like." Anticipating such a reply, he gives 
the reason for his statement [that simile and the like would have little 
or no sphere at all: the reason is. In some way or other: that 
is, by taking the form of a vibháva. etc. Regardless of this: i.e., 
even though there may be some connection with a sentient creature. 
Devoid of rasa: the position of the opponent is that wherever there 
is rasa there must necessarily be rosavadalarikára. So if there is no 
rasavadalarikára, there will surely be no rasa. Hence, according to this 
view of the opponent, there would be no rosa in cases where a living 
creature is not described. But (our position is different]: we do not say 

246 [§25eL 

that there will be no rasa if there is no rasavadalankdra, but only if 
there is no rasa in its form of dhvani. And this kind of rasa is present 
in such cases [as the description of the rainy season in the Rámáyana]. 

1. As the objector belongs to an old-fashioned school (see note on 2.5 e A), 
he presumably follows Bhámaha and Dandin in supposing that rosa is simply 
an intense form of bháva that occurs in the character portrayed. It was the 
Dhvanydloka that changed Indian views on this matter. 

A For example: 

Its waves are her frowns, 
its startled birds the strings of ber jeweled belt. 
In her anger she trails an opening robe of foam 
and twists and turns as her heart strikes upon 

my many faults. 
Surely her grievance bas transformed my love 
into this river.' 

[Vikramorvasiya 4.52] 

This slender vine with its rain-wet leav 
shows me her tear-washed lip; 
its flowers fallen with the passing of spr’ 
show her without her jewels. 
The loss of its bees 
is the silence of anxiety, 
as though my angry lady, having spurned me 

at her feet, 
now feels remorse. 

(Vikramorvasiya 4.66] 

Say, happy friend, if all is well still with the bowers 
that grow upon the Jumna bank, 

companions to the dalliance of cowherd girls 

and witnesses of Radha's love. 


Now that there is no use to cut their fronds 
to make them into beds of love, 
I know their greenness will have faded 
and they grown old and hard. 
(SRK 808, attributed to Vidys| 

1. This and the following quotation are ftom the mad scene, where Purü- 
ravas, distracted by the disappearance of his beloved Urvasi, fancies that he 
sees her transformed into the various natural beauties of the forest. The com- 
mentators differ as to whether yathá-viddham in the third line of the first 
stanza is to be taken as two words or one. In the first alternative: “since [the 
river exhibits these properties, it must be Urvasi].” [n the second, yathdvid- 
dham will mean "struck from side to side, with irregular motion" In the 
second stanza, line 3, read cinté-maunam together as a compound. 

L Its waves, etc. Tarorigabhrübharigà is a bahuvrihi compound, 

literally, "whose frowns are |or, have become] waves." "She trails”: 
literally, “forcibly pulling away the trailing (foam|." “Robe”: garment. 
The idea is (that she pulls away her dress] to prevent her lover from 
holding it. Collecting together in her heart my “many faults,” that is, 
offenses, she is unable to bear them.! What is meant is a proud, angry 
woman (mánini). And then, unable to bear the remorse caused by my 
absence, in order to calm her suffering, she has transformed herself into 
a river, 
, This slender vine, etc. A woman who has grown thin from separa- 
tion and is stricken with remorse abandons her ornaments. "(With the 
passing of] its season (of flowering]”: this is usually spring or summer. 
She is silent because she is considering a means (to effect reconciliation] 
and also because she is thinking, “Why did I reject my beloved who had 
fallen at my feet (in supplication|?" "Angry": wrathful. Although the 
sentence-purport in these two stanzas is the description of a river and 
a vine, in fact they are the utterances of Purüravas who is overcome 
by madness (and thus they are clearly connected with an emotional 

Say, happy friend, etc. “Those bowers” means those that are firmly 
entrenched in my heart. They are the friends in pleasure, the gay com- 
panions of the gopis, the wives of the cowherds. For none else [but 
bowers and vines] can really be the friends of women who are engaged 
in secret love affairs (since only they can be trusted to keep a secret 
and to provide shelter for love-making|. He shows that these bowers 

248 [$2.5fL 

were much loved by R&dhà: they were direct witnesses of RAdhà's love- 
making. Are those arbors on the bank of the daughter of the Kalinda 
mountain, i.e., the Jumna, well? The question is conveyed by inter- 
rogative intonation. The blessed Krishna, residing now in Dvárakà, 
his memories awakened at seeing the cowherd, asks him this question, 
and then soliloquizes in a manner full of longing and with feelings of 
love awakend by memory of the dlambanavibhdvas |viz., Radha and the 
gopis| and uddipanavibhávas [viz., the bowers|.? The purpose of these 
bowers lay in their being cut, because they vere soft, that is, delicate, 
for making beds of love, that is, love-couches; and that purpose is now 
fulfilled. Or, we can interpret the line to mean that the preparation, 
that is, the finished arrangement, into a love-couch, was gentle, that is, 
delicate; that this was the result of the cutting, which result has now 
come to an end. The idea is: now that I no longer recline there, what 
use is there in making love-couches? And so, with assurance that bis 
love for the gopis is mutual [i.e., that they will never make love-couches 
for anyone else}, he says, "I know, etc." The grammatical object of the 
verb "I know" is the sense of the whole sentence which follows. "Will 
have grown old and hard": If I were present, the leaves would be con- 
stantly used, as described above, and would never [have time to] reach 
such an old, faded, left-over condition. The remark that the leaves 
have lost their dark lustre suggests that Krishna has been away [from 
Gokula| for some time, but is still filed with constant longing for it. 
Thus, the second half of the stanza can be taken to be what Krishna 
says to himself. Or, we can take it as addressed to the cowherd to 
ascertain [the fate of the bowers]. The many examples here cited bear 
out what he has said, [that this applies to] “great works,” that is, to 
many works, of poetry. 

1. Abhinava brings out only half the meaning, the half that applies to a 
sentient being. In reference to the physical river, the line will mean: “aiming 
frequently at rough or stony ground, its course is irregular.” 2. We have 
followed BP in taking the three compounds beginning with prabuddharati- 
bhdvam as adverbs, but it is not impossible to take the second and third as 
adj tives qualifying ratibhdvam, the object of dha. 

§ 2.5g L] 

A In these and other similar examples, although the main pur- 
port of the sentences is a portrayal of inanimate objects, we find that à 
portrayal of human beings is connected therewith. Now, if you should 
say that whenever such portrayals are given, we may regard them 85 
examples of rasa, etc., used as figures of speech, you would leave simile, 
etc., with a reduced domain or with no domain at all. Because there 
is really no case of a portrayal of something insentient which is not 
connected with a portrayal of living creatures, at least in so far as it 
furnishes an emotional determinant (vibhàvatva). Therefore rasa and 
the like are ornaments (figures of speech) when they are subordinate. 
But when a rasa or a bhàva predominates, then, by all means, it is that 
which is ornamented, that is, the soul of suggestion (dhvani). 

L Now: one may here supply "with a view to prevent a lack 
of rasa from obtaining in such cases." But it might be objected that 
simile and the like will have their domain in those cases where there iS 
absolutely no question of that which concerns a living being, With this 
m mind, he says: Because, etc. At least: When insentient reactions 
are described, for example paralysis, borripilation, etc., because of the 
fact that such reactions are anubhávas (the physical consequences of 
an emotional situation), they necessarily bring into consideration living 
beings [as their basis]. Why say more?! Even a totally insentient object 
such as the moon, a garden, etc., though it is described lapparently] 
as an end in itself, will necessarily have no part in poetry at all, oF 
even in historical or learned literature, except in so far as it may be 2 
determinant (vibháva) of some state of mind (cittavrtti)?. Having thus 
refuted the opponent's position, he ends the discussion by repeating his 
own view: Therefore. He means, since the distinction in the domains 
proposed by the opponent is not correct. Or a bhava: The word “ot” 
shows that bhavdbhdsa, bhávaprasama, etc., are to be included. By a! 
means (sarvákáram): to be taken adverbially, “in any way" That 
which is ornamented: and not an ornament (figure of speech). 

250 [$25gL 

1. Replace the comma after tàvat with a danda and replace the danda 
after kim atrocyate with a comma. 2. BP: What is implied is that it could 
not otherwise be written about. There is no such thing as does not provoke 
some sort of mental state or thought. 

A Furthermore, 

K Whatever depends on the predominant sense should be re- 
garded as qualities (gunas). On the other hand, whatever resides in 
the non-predominant sense should be considered as ornaments (figures 
of speech), just like bracelets, etc. 

A Whatever matters depend on the predominant sense [of a 
poem], which [sense] will be a rasa or the like, are called qualities, 
just as courage, etc. [are called qualities of the human soul]. On the 
other band, whatever matters reside in the constituent elements (of the 
poem], namely, the words and their literal meanings, are to be con- 
sidered ornaments (i.e., figures of speech). just as bracelets, etc. (are 
considered ornaments of the body]. 

L An ornament must be admitted to be different from that 
which is ornamented, for this is what we find in the ordinary world, j 
as a quality is different from [the substance] which possesses the qual- 
ity. Furthermore, the use of the words quality and ornament is possible 
only where there is something that may be qualified and something that 
may be adorned. Now it is only by our theory that this usage can be 
seen to be logical. With these two propositions in mind,' he says: Fur- 
thermore. The word has the sense of addition, i.e., it is not only the 

§2.7 K] 251 

arguments already given? that show that rasa must be predominant, 
for there are others as well. The Karika too can be brought into accord 
with these two propositions. However, in regard to the first, the first 
half of the Karika should be explained as an illustration. The text of 
the Vrtti should also be explained in tbis way. 

1. The two propositions are: (o) An olarkáro (figure of speech) must be 
different from the olarikárya (the body of poetry); (b) One can only speak 
of an alarikára and a guna (poetic quality) where an olorikárya and a gunin 
(viz., rasa or the like) are present. — 2. It is not clear just what arguments 
(yuktijátam) Abhinava is referring to. But yukti refers to something less than 
formal proof (sádhana), e.g., to circumstantial proof and analogy, and one 
can find arguments of this sort for the predominance of rasa in the comments 
on 25. Even when the main purport of a sentence is a mere thing, it will lead 
eventually to raso if the sentence deserves the name of poetry. The two new 
propositions are more in tbe nature of formal proof. — 3. Tbe first half of tbe 
Kárikà does not state the proposition that an alarikára must be different from 
an alarikárya, but illustrates it by the analogy: a guna is that whicb depends 
on à gunin. 

A Soalso 

K It is just srrigára (the flavor of love) that is the sweetest and 
most delightful flavor (rasa). Sweetness (mádhurya) has its seat 1D 
poetry that is full of this flavor." 

252 [5277 K 

1. The old vocabulary of poetic criticism is here preserved, but it is 
made to express new relations. From the time of Bharata the critics spoke 
of the qualities or virtues (guna) of poetry. The early critics emphasized 
the qualities of sound. Later, qualities of meaning came in for their share 
of attention. But in all these critics the qualities were considered virtues 
in themselves. Now the Dhvonyóloka reduces the ten qualities of Bharata 
to three: sweetness (mádhuryo), force (ojas), and clarity (prasáda); and it 
regards them as virtues only insofar as they lead the audience to rasa. This is 
Ánanda's position: the qualities reside in the sound and sense but depend on, 
exist only for the production of, the rasa. Abhinava goes even further, saying 
that the qualities are ultimately qualities of the rasa, of the experience itself. 
Their names are then applied metonymously to the sound and the sense that 
bring about such an experience. 

A The rasa of love is sweet in comparison with the other rosas, 
because it gives delight. The quality known as sweetness is attributed 
to a poem if the poem consists of words and meanings which reveal 
this Bavor. As for being pleasing to the ear, this characteristic belongs 
to (the poetic quality] force (ojas) also. 

L Now it might be claimed that the qualities sweetness and the 
like belong to word and meaning. One may therefore ask how these 
qualities were said to depend on the predominant element, namely a 
rasa or the like (see 2.6]. Anticipating this question, he says: so also. 
He means that this [dependency] is perfectly reasonable because of the 
particular refutation (of the old-fashioned view] that our author has in 
mind and that he is about to announce. 

(Commentary on K) "It is just srngára, etc." He gives the reason 
why it is called sweet with the words *most delightful." For in all crea- 
tures: gods, anirnals, men, and the like, there is an unbroken proclivity 
(vásoná) towards the emotion of love (rati).! Acccordingly, there is no 
one who is not inclined to respond sympathetically to it. Even an as- 
cetic can be struck by its charm. And so it is said to be sweet. For 
a sweet substance, such as sugar or the like, when it comes in contact 
with the tongue, will immediately prove desirable to everyone, wise or 
foolish, healthy or sick. “Full of this”: viz., that poem in which srngára 
is the main concern of the poet and is conveyed through suggestion as 
the essence of the poem.? “Poetry,” i.e., both word and sense. “Has its 
seat": takes its seat. This is as much as to say that in reality what is 

§ 2.8 K] 253 

called sweetness is a quality only of the rasa, such as srngára, ete. This 
[quality] has been metaphorically transferred to the word and meaning 
which are suggestive of this sweet relish. For the definition of sweet- 
ness is the capacity of words and meanings to suggest the sweet relish 
of love [etc.]. Therefore it was correct to say, “whatever depends on,” 
etc. (2.6 K}. 

[Commentary on A] The Vrtti gives the meaning of the Karikà:: The 
rasa of love, [etc.]. Now sweetness has been defined [by Bhamaha 2.3, 
where he says|: "That [poem| which is pleasing to the ear and which 
does not contain many compound words and compound ideas is what 
is called sweet." Our author rejects this with the words pleasing to 
the ear, etc. He uses "pleasing to the ear" as an abbreviation of 
the whole definition of Bhámaha. Belongs to force also: for in the 
verse "Whatever man proud of his strong arm," [which is an example 
of force] there is both pleasingness to the ear and absence of long 

1. All creatures have had some previous experience, if not in this life at 
least in some previous life, of roti. This has left them with a latent proclivity 
toward this emotion, a predisposition to be affected by the stimulants of this 
emotion. 2. To speak in simple terms, Abbinava is rejecting the sense 
“poetry which is full of this relish” and substituting the sense “poetry in 
which this relish forms the essence." To speak in technical terms, Abhinava is 
rejecting the Kasikd's interpretation of P&a. 5.4.21: tatprakrtavacane mayat- 
Here the Küsiká says that prakrta means prdcuryena prastutem Thus the 
sjitro means, according to the Káfikd, "the suffix mayat (may be added to 
a word designating a substance when we wish] to express that there is an 
abundance [of that substance." Abhinava, on the other hand, wishes to 
take prokrta to mean dtmatvena prakrta and would understand the sūtra to 

... When we wish to express that the substance forms the essence {of 
|.” 3. The verse will be given in full in 2.9 A 

K In the rasa of love in separation and that of compassion (ka- 
runa) sweetness is intense. This is because in these cases the heart is 
Softened to a greater degree. 


A In both these rasas, sweetness alone [of the qualities] is at its 
most intense, because the heart of the sensitive audience is overcome 
[by these rasas] to a greater degree. 

L As the relish of love in separation is sweeter than that of love 
in union, and the relish of compassion is sweeter still, it follows that 
the higher degrees of sweetness of word and meaning are in reality their 
ability to suggest these [two forms of aesthetic experience]. It is with 
this in mind that he says, In the rasa of love, etc. And in that of 
compassion: the word "and" shows the sequence. Intense: what is 
meant is an intensity that progresses [in step with the sequence]. Soft- 
ened: the meaning is that the heart of the sensitive audience hereby 
abandons its natural hardness, its imperviousness, its liability to the 
flame of anger and its passion (rága) for the marvelous and for laugh- 
ter. To a greater degree: to a progressively greater degree. This is 
as much as to say that in the relish of compassion the heart completely 
melts. Now an objector will question us: "If there is sweetness in the 
relish of compassion as well, how do you explain the fact.that in the 
previous Kariké the word "just" (eva) was used |"It is just srngára," 
etc.]? Our reply is that this was not meant to exclude other rasas, but 
only to show that in strict fact poetic qualities such as sweetness and 
the like belong only to a rasa, the very soul [of poetry], and that only 
by extension are they attributed to words and meanings. This is what 
the use of "just" was meant to indicate.’ 

The Vrtti gives the sense of the Káriká: in both these rasas, etc. 

1. Abhinava is quick to notice discrepancies of statement. But he does 
not rest content with-noticing them, as a Western critic might do. He is a 
partisan of his author, so he hastens to defend him. He does so in the present 
case by suggesting that we understand the first half of 2.7 K to say “It is just 
the flavor (rasa) in srngára that is sweet. not the words, etc." 

$2.9 A] 

K The rosas of fury (raudro) and the like in poetry are charac 
terized by excitement (dipti). Strength (ojes) has its proper place ii 
words and meanings that manifest this excitement. 

A The rasas of fury and the like give rise to a high state o 
excitement. of inner flaring up, and so by metonymy (laksaná) it i 
said that these rasas themselves are “excitement” (dipti).' A word 
group capable of producing this excitement is a sentence adorned b 
the use of lengthy compounds, as for example: 

The brutal war-club whirling in my arm 
will crush both thighs of this Suyodhana. 
so that he whose name is truly Bhima 
may deck your hair, my lady, with his hands 
new-reddened in that fresh-congealing blood. 
[VenisamAáro 1.23] 

Equally capable of producing this excitement, (however,] is a meanin, 
which is expressed in lucid words,? without recourse to a style of length; 
compounds, as for example: 

Whatever man proud of his strong arm 
bears weapons in the Pàáücála clan: dotards. 
children, down to babes in the very womb; 
and whoever saw that deed: 
I will slay them all when 1 come upon the field, 
every man who fights me, though he be Death himself: 
in blind fury will I be the death of Death! 
(Venisamhéra 3.32] 

In both these and other similar examples there is strength. 

1. All printed editions except Krishnamoorthy's read ucyate (singular; 
Krishnamoorthy takes the reading ucyante from his Moodabidre MS. Eithe 
teading may be justified, the former by taking the whole iti clause as th 
subject, the latter by taking ta eva as the subject. — 2. The translation doe 
not attempt to imitate the long compounds of the Sanskrit verse. For th 

256 [82.9 L 

mise-en-scène see Editor's Introduction, pp. 34-35, where the effect of heavy 
compounding in Sanskrit is also discussed. 3. That is, words which have the 
poetical quality prasdda (clarity). — 4. In this verse ASvatthaman is vowing 
vengeance for the treacherous killing of his father, Drona. 

L Fury and the like: the word ádi is used in the sense of 
“si ilar cases.” Thus the rosa of the heroic and that of the marvelous 
are included as well.' Excitement is an apprehension characterized by 
radiance, expansion, and a blazing forth in the heart of the sensitive 
audience. It is what is primarily denoted by the word strength (ojas). 
Raudra and other similar rosos consist in the relishing of this excite- 
ment. They are characterized as effects, tbat is, they are distinguished 
from other rosas, by this excitement, this particular kind of relishing. 
Hence, by a metonymy which applies the name of the effect to its cause, 
raudra and si ilar rasas are called by the name “strength” (ojas). 
Then by resorting to a second metonymy,? even a word-group which 
reveals excitement, namely a sentence with long compounds, is called 
“an excitement” (dipti)? for example, the verse "The brutal war-club 
whirling.” In addition to this, a meaning which produces excitement, 
being expressed by lucid, quickly intelligible words‘ without recourse 
to compounds, is also called “an excitement” (dipti), for example, the 
verse “Whatever man proud of his strong arm.” 

(The brutal war club, etc.:] With this “brutal,” that is, cruel, war- 
club which will be wielded by my “whirling” arms, that is, arms circling 
with great speed; with the full attack of this war-club the “two thighs” 
will be simultaneously "somcürnita, that is, completely crushed so 
that the man cannot stand up again. I shall so despise? Suyodhana 
(= Duryodhana] of whom the thighs are in this condition that I will 
become one of whom the hands are “red,” crimson, with the blood 
that is “thick,” that is, does not run off because it comes from a deep 
[wound] and is not the thin liquid (of a superficial cut], and which 
will stick to my hands because it is “congealing,” that is, in the pro- 
cess of coagulating, there not having been time enough for it fully to 
dry. It is for just this reason that I shall be “Bhima,” that is, one 
who strikes terror into the hearts of cowards. “Your” hair: of you, to 
whom so many insults were given— "my lady" shows how little deserv- 
ing she was—] shall deck the hair, that is, I shall change it from its 
braided state and give it as it were a coronet of red flowers with the 
bits of blood falling from my hands. Thus we have an implied figure 

$2.9 L] 25 

of utpreksà (poetic fancy). By using the vocative “my lady," whic 
reminds us that a noble woman was humiliated [by the Kauravas], tt 
author has applied a stimulus (uddipanavibháva) to the relish of ange 
Accordingly, [in spite of the reference to a normally amorous act, tt 
decking of a beloved's hair] one cannot suspect any suggestion of tk 
relish of love. There is both despite of Suyodhana and an absence « 
any effort to strike him again with the war-club because his thighs wi 
have been completely crushed [with the one blow]. The use of the wot 
"congealing" suggests Bhima's impatience to wash away the grievant 
of Draupadi [before the blood even dries]. From the long compoun: 
flowing in an uninterrupted stream and allowing [the hearer] no pau: 
in all its course, there results an apprehension of the whole scene as 
unity up to the scorn of the broken-thighed Suvodbana. This serves ! 
intensify the impression of [Bhima’s] violence. Other commentators, * 
may note, take suyodhanasya as a possessive genitive and explain th: 
Bhima's hands will be reddened by the thick blood “of” Suyodhat 
that will cling, congealing, to them. 

Whatever man proud of the strength of his own arm and [fightin 
among the [Pándava] regiments: this refers to such warriors as Arjun 
Because Drona was killed by Drstadyumna, the son of the Páücá 
king, A$vattháman is especially incensed against that clan. "Wboev 
saw that deed" refers to such as Karpa. [Now we may understar 
carati mayi rane in either of two ways:| (1) when battle is to be e 
gaged, whoever comes against me, that is, interferes with my fightin 

„or (2) when I am moving in battle, whoever opposes or resists me. Suc 
a man (will I kill] even if he were the death of all creatures, how Mur 
more if he were a mere mortal or a god. Here the anger [of the speake 
rises to the highest pitch [by a progession] from word to word, throug 
meanings which, being presented separately, are reflected upon by t! 
hearer in succession. And so the very absence of compounds acts as 
cause of dipti. 

By showing that sweetness (mádhurya) and excitement (dipti = ojt 
strength) are opposed to each other in love and fury, our author wou 
indicate that in comedy (hàsya), the fearsome (bhaydnaka), the loat 
some (bibhatsa), and peace (sànta), these qualities exist together 
varying proportions. So far as comedy is subordinate to love, sweetne 
will predominate; and so far as it partakes of the nature of expansi 
(vikdsa), strength will predominate; so the two qualities come to 
equal in that genre. In the fearsome. although that rasa consists in 
broken (bhagna)' state of mind. the stimulant (vibhdva) is exciteme 

258 (52.9 L 

and so strength is greater and sweetness less (in that rasa]. The same 
applies to the loathsome. But for the rasa of peace there is a great 
variety of determinants, so that sometimes strength will predominate 
and sometimes sweetness.® Such are the distinctions (among the rasas}. 

1. BP and Mahadevasastri (see Text, p. 208, footnote) are worried by 
this statement because the relish of the marvelous should produce råga rather 
than dipti according to what Abhinava has just said (2.8 L; Text, p. 208, line 1 
of Locana). But we need not suppose the characteristics of the separate rosas 
to be so rigidly departmentalized. — 2. A laksitoloksanó is a second loksoná 
which comes into effect after a first laksand has already operated. 3. Ob- 
serve how elaborately Abhinava bas complicated the natural sense of these 
words. To take the words naturally, ojas (strength) is a property of words and 
meaniogs wbich enables them to produce a raso of fury or of tbe heroic. Tbis 
rasa is characterized by dipti (excitement, or, literally, inflammation"). But 
io Abhinava's explanation, dipti and ojas become synonyms. Dipti is what 
ojas primarily means and ojas denotes primarily the result of an experience. 
The word ojas (or dipti) is then transferred by metonymy to the experience, 
the relisbing, itself. Then by a second metonymy dipti (or ojos) is transferred 
to the words and meanings that produce the experience. — 4. Abhinava uses 
the word gamako to paraphrase the word prasanna used in the Vrtti (Text, 
p. 211, line 1]; so it means jhatityarthabodhaka (BP) “instantaneously con- 
veying the intended sense." Abhidhiyaména paraphrases the abhidheya of the 
Vrti 5. As will appear from his further comments, Abhinava is taking 
the genitive suyodhanasya as anddare sasthi (Pén 2.3.38) and eliciting some 
such meaning as "I shall so despise Suyodhana in his weakness that I will 
wash my hands in his blood." This is of course an unwarrantable extension 
of what Pánini means by the gentive absolute of despite. Furthermore, this 
genitive in actual use almost always contains a participle, e.g., rudato 'vràjit. 
"despite his (father's) weeping he became a monk." The Kdsiké actually in- 
sists on the presence of a verb form. The interpretation probably appealed 
to Abhinava on two grounds. It avoids construing suyodhanasya with the 
subordinate member of a compound, foni; and it intensifies the raudrarasa. 
Abhinava uses a similar interpretation for the locative, 3.10-14¢ L and note 1, 
3.24 L and note 2. — 6. No iva is used, so it is not a direct utpreksd. But 
the red blood reminds the hearer of red flowers, so that a fancy is suggested. 
7. The reading bhagna of P&thak's edition seems better than the magna of 
our text or the dhaya of Tripathi's. The point would be that in relishing fear, 
our mind, it is true, is reduced to weakness: but as it is strong stimulation 
that has so weakend it, strength plays a predominant part in literature of 
that genre. 8. Under 4.5 Ananda will assign the Mahdbhdrata to the genre 

al maara haranea ite variad erenes lead anr minde hv a enrt af catharsis ta 

§ 2.10 L] 25¢ 

an attitude of peace. Of the Mahābhārata one may well say that strengtt 
sometimes predominates and sometimes sweetness. 

K A poem's ability to communicate [to the reader] any rosa, at 
ability which is found operative in all rasas and styles, is called clarity 

A Clarity is clearness both of word and meaning. And while it i: 
a quality common to all rasas and common to all structures (racaná) 
we should restrict it primarily to its connection with the suggeste 

1. Here again the vocabulary is traditional but the idea is new; cf. 2.7 F 
note. Clarity is generally taken by the older criti to be a quality of mean 
ing; only Vámana added a "clarity" of sound. This clarity was regarded as : 
virtue in itself without regard to what was being clarified. The Dhvanyólok 
now defines it primarily in terms of ruso. Clarity belongs only to that worc 
and meaning which clearly reveal a rosa. It is common to all rosas, that is, i 
is found in the words and meanings which produce all rasas. [n this respec 
it differs from sweetness and force, which have more restricted domains. Ir 
saying tbat it is common to all structures. Ananda may be making a criti 
cism of Vamana (3.1.6), who believed the sabdaguna prasáda to consist in ai 
uncompounded structure. 

L Ability to communicate, literally an ability to communi 
cate (arpakatva) that is complete (samyak). is the ability to fill (vyápa 
katva)’ the hearer quickly with a sympathetic response by entering int: 
him, as fire quickly pervades dry wood, or water pervades a clean cloth. 
This lack of impurity, this lucidity, is a quality of all the rosas. B: 
metaphorical extension the word clarity is also used of the power, pos 
sessed by both words and meaning, to communicate suggested mean 
ings of that sort (viz., rasas) 

260 [52.10 L 

The Vriti explains: Clarity, etc. Anticipating an objector who 
might ask how this purity that belongs to the rasas can be attributed 
to words and meanings, he says, And it is, etc. The word “and” is 
here restrictive [i.e., is used in the sense of eva|.? (In its primary sense] 
clarity is a quality only of all the rosas and only this quality is such (sc., 
common to all the resas].‘ (In a secondary sense] clarity is common to 
al! structures both of word and sense, whether they make use of many 
compounds or none. Primarily: This ability to communicate can be 
intended only with reference to the suggested meaning, not others, for 
what is so wonderful in a word's communicating its literal sense that 
one should use the word "quality" or *virtue" of such a word? This is 
what our author has in mind. 

Thus according to Bhámaba (2.1-3] only three qualities were ac- 
cepted: sweetness, strength, and clarity. And these three qualities refer 
primarily to the relishing of the audience. By metaphor their names 
are extended to that which is enjoyed, namely the rasa (in that word's 
sense of object ratber than process], and even further to the words and 
meanings which suggest this rasa. This is the essence of what he means 
to say. 

1. We follow BP in reading vyápakatvam in place of vyápóárakatvam; cf. 
the use of vyópyate in the verse from which the analogy of fire and dry wood 
is taken (BANS 7.7, quoted Dhv. 1.1e L, Text, p. 39, lines 2-3). 2. Remove 
the danda after drstántena and place it after co. For the first analogy see 
previous footnote. The second is used by Mammata 8.70 (Jhalkikar p. 476), 
who has doubtless taken it from Abhinava, as is clear from the commentary 
of Sridhara, p. 277. The sense is that if a clean cloth is put into water, it will 
quickly absorb the water, whereas if the cloth is greasy the water will not easily 
be absorbed. Cf. Jhalkikar, p. 476: svacchajalavat—jolam yathà svaccham 
patam sohasata vydpnott malinapote jalasydprasarat tadvat, — 3. This sense 
of ca has the authority of Patanjali, Mahábhásya on Pao. 2.1.16. Kaiyata 
justifes it by the fact that "particles have many meanings" (anekárthatván 
nipátànàm). Abhinava not only takes the word here in this unnatural sense, 
but he reads it twice (by àvrtti or tantra) so that it may restrict both rasa- 
sddhdrano and sa gunoh. 4. BP: the first restriction rules out its being a 
property of word and meaning; the second restriction rules out sweetness and 

§ 2.11 L} 

K And? those faults such as indelicacy of sound,? which have 
been shown [by earlier writers] to be relative, are found in (their] illus- 
trations to be shunned only in the essence of dhvani (and there only] 
in the rasa of love? 

1. None of the commentators or translators remarks on this little word 
“and.” It would seem to connect the present Kànikà with the immediately 
preceding Vrtti in a way that can be shown by the following paraphrase. “The 
designation of clarity should be restricted primarily to the suggested meaning 
and the relative faults also are faults only where the meaning is suggested.” 
2. Four faults are intended, as first listed by Bh&maha 1.44: srutidusta, artho- 
duşta, kalpanddusta, and srutikasto. They are explained below by L. 3. That 
is to say, indelicacy of sound and the like are faults only in vivaksitányapara- 
vácyadAvani where the vyarigya is $rngórarosa. 

A  Faults such as indelicacy of sound, which have been indicated 
to be relative, also need not be shunned when the litera] meaning is the 
sole object, nor when a meaning other than the rosa of love is intended, 
nor when the rasa of love is not suggested by that type called the soul of 
dhvani. It is seen from illustrations that these faults are to be avoided 
only in the soul of suggestion, when love is suggested as the primary 
element [of the poem]. If this were not the case, these faults would not 
be relative. 

So now the essence of dhvani, where suggestion is indicated with- 
out apparent sequence (asamlaksyakramavyarigya), has been shown in 
general terms. 

L Having previously shown that the distinction in usage between 
the words "quality" (guna) and “ornament” (alarikára) is a logical one 
only if one accepts our position (see 2.6 L), our author now sets forth 
to show that the distinction of absolute and relative faults also makes 
sense in our position only. Such as indelicacy of sound (sruti- 
dusta): Words like vànta (vomited) are indelicate of sound! because 

262 [$2311 

they bring to mind objects that one does not speak of in polite com- 
pany. Indelicacies of meaning (arthadusta) are such as cause us to 
understand something indecent because of the purport of the sentence 
as a whole. An example is (the ambiguous verse]: "Seeking his chance, 
the great proud [man] advanced to the attack,” [which can also mean 
"Seeking for the hole, the great stiff (penis) advanced to the attack"]. 
An indelicate arrangement (kalpanádusta) occurs by the arranging of 
two words: for example kuru rucim (do what you like) if arranged 
in reverse.? Harshness of sound (srutikasta) is seen in such words as 
adhàksit (has burned), aksautsit (has stamped), trnedhi (pierces). The 
relish of love is meant to include other appropriate rosas as well, be- 
cause these faults are avoided in other rasas such as the heroic, the 
peaceful, the marvelous, etc. 

Have been indicated: The early authors have not shown the rel- 
ativity of these faults by showing any distinction in the areas (where 
they may and where they may not be used|. Neither have they shown 
how they differ from such (absolute] faults as error in meter, nor have 
they really shown how they differ from virtue? Our position is that 
they are allowable in the relish of disgust, of comedy, and of fury; and 
that they are to be shunned in the relish of love and the like. This 
explains why they are called relative. Such is what our author has in 

1. Place a danda after $rutidusta instead of hetovoh. — 2. Le., rucim 
kuru. BP says that ciriku means the clitoris. The word is not listed in PW. 
Bhámaha's example involves no reversal but merely an infeicitous juxtapo- 
sition: sa sauryábharunah "this ornament of courage," which brings together 
the syllables ydbha, the vulgar Sanskrit word for sexual intercourse. 3. An 
indecent ambiguity might be a virtue in a comic verse; one can easily find 
examples in the anthologies. As for Abhinava's negations, they are all strictly 
true of Bhimaha. Bhàmaha admits that some of these faults may appear as 
virtues in just the right context (1.54-55) and he gives us two or three exam- 
ples. But he states no general principle such as is stated here by Ananda. 

$2.12 A] 

K The varieties of the elements subordinate to this [rasa or the 
like] and the varieties within itself, when one imagines all their possible 
combinations with one another, are infinite. 

A It has been stated that a rasa or the like when suggested as 
the main element of a poem constitutes the very soul of that type of 
dhvani in which the literal meaning, though intended. is subordinate 
to a second meaning. Of its subordinate elements, namely the figures 
of speech which are based on word or meaning, there are endless vari- 
eties. And of the main element itself, the varieties. namely the rasas, 
bhávas, rasdbhdsas, bhdvdbhdsas, bhávaprasamas, taken together with 
the vibhàvas, anubhdvas and vyabhicárins which produce these rosas, 
etc., are endless, having no limit in respect to the bases [in which they 
may be portrayed]. Now if one were to consider all the possible com- 
binations of these [two sets of] varieties, one could not count up the 
varieties of a single raso, much less of all. Thus, of the rasa $rrigàra 
there are two main categories to begin with: love in union (sambhoga- 
$rrigára) and love in separation (vipralambhasrrigàra).? Of love in union 
there are the varieties represented by the lovers' looking at each other 
lovingly, their sexual enjoyment, their recreation, and so on. Of love 
in separation we have yearning in separation, the separation caused by 
jealousy, by love quarrels, by exile, and so on. And all of these may 
be divided according to the vibhávas, anubhdvas, and vyabhicárins in 
each case. And as there are differences of these according to the base in 
some location and the stage in some time-span [in which and at which 
they occur], even this one single rasa will prove to be endless in respect 
to the varieties which exist within itself. How much more so if we take 
account of the varieties of its subordinate elements! For if we consider 
all the varieties of subordinate elements [i.e., all the figures of speech] 
in their possible combinations with the varieties of each main element 
li-e., with each rasa, bháva. etc.), we shall surely arrive at infinity. 

264 (§2.12 A 

1. svásraydpeksayá (Text, p. 215, line 6) cannot mean “when considered 
in themselves," as Jacobi seems to have taken it, for that much is already 
implied by svágotás in line 4. Ananda expands the phrase on page 217, line 2 
into desakdlddydsraydvasthdbheda iti sudgatabheddpeksayd. In both cases sud- 
$ruya must refer to the dsraya, the physical base, in which the vibhávas, etc., 
reside. In the expansion Ananda has added a reference also to their residence 
in time, that is, to the various stages of the emotion. There are notably nine 
stages of love in separation before the subject dies of it. Now the spacially 
qualified residences of the emotions can scarcely be anything else than the 
characters in whom the emotions are displayed, “the appropriate male and fe- 
male characters,” as Abbinava puts it. But Abhinava strangely misinterprets 
the expanded phrase. See below. 2. [t may be well to give literal transla- 
tions of these technical terms. Sambhogasrrigára is literally love in enjoyment; 
vipralambhasrrigáro love in frustration. for viprolambha means literally vañ- 
cand, deception, frustration. The subtypes of love in frustration are abhildga- 
vipralambha: frustration that takes the form of yearning; irsyévipralambha: 
frustration caused by the jealously of the woman; virchaviprolambha: frus- 
tration caused by the woman's intentional separation (viroha) of herself from 
ber lover; and provdsaviprolambha: frustration caused by the man’s leaving 
bome (because of tour of duty, exile, etc.). 

L Subordinate elements: that is, figures of speech. Within 
itself: belonging to itself (àtmagata), such as love in union and love in 
separation, and belonging to what belongs to itself (atmiyagata), that 
is, whatever belongs to the vibhàvas, etc. One could not compute all 
the relations of principal and subordinate even by a marker and table 
of elements.' 

The bases: what is meant are such bases as the appropriate male 
and female characters. Looking at each other lovingly implies other 
varieties as well, such as talking together, etc. Sexual enjoyment is 
[divided into] sixty-four items, beginning with the [eight] embraces.” 
Recreation: that is, walking in the park. "And so on" will include 
such activities as water sports, drinking parties, watching the moonrise, 
games, etc. Yearning in separation is when a mutual love (rati) has 
arisen in a man and woman such that each values the other as his own 
life and when for some reason they are unable to meet, as in the case 
of the King of Vatsa and Ratnàvali in the Ratnàvali from the passage 
“Why do you ask whether she pleases me?" onward. But not before 
that passage in the play, because when mutual love (roti) is absent, 
one can speak only of the stage of physical desire (kama). Separation 
caused by jealousy is to be taken as the estrangement of a woman 

§ 2.12 L) 265 

who has been injured by her lover's infidelity, etc. Separation caused 
by a love-quarrel refers to an experience with just such a woman. 
who has not accepted her lover’s attempts to placate her and who 
is subsequently filled with remorse and longing because of his absence- 
Separation caused by exile refers to a husband's love for his wife whom 
he has left at home. “And so on” will include separation caused by 2 
curse,? etc. The word viprolambha is used ina Secondary sense to mean 
that which is "like a deception,” * for in a deception one fails to gi? 
one's object; and such is the case here. Of these means “of love-iD- 
union, etc." on the one hand and “of the vibhàvas, etc." on the othe! 
If we were to take the base (àsaya) to mean a geographical location 
such as the Malaya Mountain which is the location of a uibhava SU 
as the Malaya breeze, its sense would already have been anticipated bY 
the word "location" (desa). Therefore it is best to take “base” in tbe 
sense of "causé." An example [of such variety of cause is found iD s 
verse of mine: 

I have kept this garland woven by my love 
close to my beart. 

Even as it fades, it drips ambrosia to dispel 
my pain of loneliness. 

This one: viz., srrigáro. Main element: what is meant are the 12595 
if we consider all the possible combinations with the various rosas- 

, X. For lostoprostáranyáya see Jacob, Vol. 1, p. 44. The losta (piece of 
clay or pebble) is the marker that was moved from one position to the ne 
to count up the number of possibilities. Abbinava uses the maxim 865^ 
3.161L. 2. This is the dictum of the Babbraviyas, Kamasütra 2.2.4 (K95bl 
ed. 2.2.5). The editor of our Text, Pandit Pattabhirama Sastri, lists the 82°” 
four items in his footnote p. 216. 3. As in the Sékuntala. 4. See 8DOV® 
2.12 A, note2. S. It is difficult to see why Abhinava chooses this most Im- 
probable interpretation. We have explained our interpretation above (2-12 A e 
note 1). Even the interpretation to which Abhinava objects, presumably PY 
some earlier commentator, is preferable to his, for Abhinava must interpret 
tegàm desakáládyásrayávasthábheda to mean “difference of these [aspects ° 
love] brought about by place, time, etc., and by causes and stages.” ^$ °? 
example of variety of cause, he cites a verse in which a garland act$ s 
peculiar cause (uddipanavibháva) of viprolambhasrrigárg in that by reraindibÉ 
the lover of his beloved it pains him, but through his knowledge that Sbe ba 
made it it assuages his pain. 


[$2.13 1 

K A mere indication will be given here, by which the minds : 
persons of education and taste may be furnished with a criterion fc 
all other cases. 

A Because persons of education and taste, if by a mere indicatio 
they recognize the [correct] relation of subordinate and predomina: 
between the figures of speech and one particular rasa [viz., srrigdra 
will bave a criterion for their judgment in all other cases. 

L By which: sc., by which indication. Persons of taste: t 
has in mind those who wish to become great poets and connoisseurs 
For all cases: this should be construed [with the bahuvrihi compoun 
to mean: a judgment by which a criterion, an understanding, a correc 
educated view, has been obtained witb regard to all rasas and the lik 

1. That is, the instruction is addressed both to writers and readers. 

A This being the case, [we will state that] 

K Alliteration used continuously in the same form, because < 
its laboredness, is not helpful in revealing love in any of its varietit 
when this love is [intended to be| predominant. 

§2.15 A] 

A The varieties of love when it is predominant have been given. 
In all these varieties, alliteration used continuously and always in the 
same form is not suggestive. By the use of the word “predominant” it 
is implied that the continuous use of alliteration in the same form is 
permissible when love is subordinate. 

L This being the case: i.e., since we are about to give an 
indication. Because of its laboredness: because it can be accom- 
Plished only by labor; he means this as a reason [why such alliteration 
is ineffective]. He uses the phrase "the same form" to indicate that if 
one abandons this monotony and composes in varied alliterations, this 
will not constitute a fault. 

K In love, when it is the soul of suggestive poetry, the use of 
echoing alliteration (yamokas)! and the like, even if the author is well 
able to compose them, is (what can only be excused as) carelessness on 
his part, especially in love-in-separation. 

71. Yamaka is the repetition of a set of phonemes in the same order. If the 
sets bear meaning (i.e., if they form complete words), their meanings must 

A In love, the very soul of suggestive poetry when it is suggested 
as the primary sense by words and meanings [which are subordinate}, 
the use of yamakas and similar figures, such as difficult [arrangements], 
or puns involving the breaking up of words in two different ways, even 
if the author is well able to compose them, is carelessness on his part. 
The word “carelessness” implies that while a solitary instance of ya- 
maka may occur by accident, this figure should not be used in abun- 
dance as subordinate to a rasa in the way that other figures of speech 
may be used. By saying “especially in love-in-separation" it is shown 
that love-in-separation is exceptionally delicate. When it is being sug- 
gested, a subordinate yamaka or the like should not be employed 

268 [$2.15 A 

1. We have translated duskara as a noun, rather than as an adjective, in 
accordance with Abhinava's interpretation and with Ananda’s use of the word 
in 2.16 (yomakaduskaramárgesu). 2. A simple pun (sabdaslesa) would be, 
for example, the word guna used so as to mean both bowatring and virtue. A 
sabdabharigaslesa is a more complicated pun, a word that can be divided in 
different ways, e.g., asoka (= the asoka tree, or = a-soka “not grief"). 

L  Yamakas and the like: the word adi refers to similar fig- 
ures. Difficult (arrangements): e.g., verses shaped like drums or 
wheels! Puns involving the breaking up of words: note that 
a simple ambiguity (arthaslesa)? is not a fault, as in the example "You 
are rakta.”? Even in puns involving the breaking up of words in two 
ways it is only the difficult ones that make for a fault, not (an easy pun] 
like asoka.* 

1. Most of the handbooks of alarikdras give descriptions of these citra- 
bandhas. The drum-shape can be read zigzag as well as linearly. See the 
illustration in the Kashi ed. of Sisupdla facing p. 716. See also 3.41-42 A and 
note 1. 2. An arthaslega is not a pun but an ambiguous word, usually an 
adjective which can apply to two different substaptives. It is distinguished 
from a pun (Sabdaslesa) by the fact that the effect remains the same even 
when we substitute a synonym. 3. The example will be quoted in context 
2.18-19c A. 4. See the example quoted in 2.18-19c A. 

A The reason for this is: 

K Only a figure which can be composed in the course of one's 
preoccupation with rasa and that requires no separate effort in itself is 
acceptable as an ornament in suggestive poetry. 

$2.16 A] 

A Although the emergence of a given figure of speech may seem 
wonderful, it may still be acceptable as an ornament in the type of sug- 
gestive poetry in which the passage from literal to suggested meaning 
is unnoticed, if it can be produced through one's very preoccupation 
with reso. For only this sort of figure is, in the real sense of the term, 
a subordinate element of rasa. For example: 

Your palm erases from your cheek the painted ornament 

and sighs have drunk the ambrosia! flavor of your lip; 

the tears that choke you agitate your breast: 

anger has become your lover, stubborn one, in place of me. 
[Amarusatoka 81; SRK 664, etc.]' 

Inasmuch as a figure that is subordinate to a rasa is characterized by 
the fact that no separate effort is required on the part of the poet 
to create it, it follows that if a poet who is concentrating on putting 
together a rasa should leave? that trend of thought and apply himself to 
some other effort, the figure that might result would not be subordinate 
to the raso. When one intentionally and repeatedly makes yamakas, 
there invariably is involved the undertaking of a separate effort, which 
takes the form of searching for the particular words that will fit. To 
the objection that the same will hold true of other figures of speech, 
we say not at all; other figures, even those which are difficult when 
described, will rush to present themselves to a poet of imaginative 
genius precisely while he is concentrating his mind on the rasa. An 
example is the passage in the Kádambari where K&dambarT is [first] seen 
[by the hero]; again, in the Setubandha where Queen Sita is shocked by 
the illustion of the severed head of Rama. And this stands to reason, for 
rusas are suggested by particular meanings and by words that convey 
these meanings.’ Now it is figures of speech such as rūpaka (metaphor) 
and the like that are the particular meanings which are able to reveal 
rasas. Therefore they are not extraneous devices in helping to suggest 
these rosas. On the other hand, this [character of being extraneous| 
does attach to yamokas and difficult arrangements of words. As for 
those few yamakas and the like which are found to possess rasa, those 
are cases of the rasa's being subordinate and where the yamaka or 
the like is the predominant element. Jn rasábhása, of course, it does 
Dot contradict [our theory} for a yamaka to be used as a subordinate 
element. But when rasa is to be suggested and is to be of primary 

270 [$2.16 A 

importance, yamakas and the like cannot be subordinated to the rasa 
because they require of the poet a separate effort. 
The following is a verse-summary of the matter. 

A great poet can produce with a single effort some matters that contain 
rasa together with figures of speech. But for composing yamakas and the 
like, he must make a separate effort even if he is well able to compose them. 
Therefore these figures cannot play a part subordinate to rasa. There is no 
objection to using yamakas and the like as elements subordinate to rasábhása. 
But this subordination is impossible in the case of love. the soul of suggestive 


1. The point of citing the verse is that it contains several figures of 
speech which detract in no way from the overall rasa. which is love in sep- 
aration caused by jealousy (irsyávipralambha). The figures are ambiguity 
(arthaslesa), metaphor (rupaka: the anger acting like a lover), and contrast 
(vyatireka: anger is your lover. not I). 2. atyuAya: according to Pan. 7.4.23 
the ü should beshortened: atyuhya. 3. Remove the danda after ákseptavyáh 
and place it after fabdaih. The tat in tatpratipádakaih refers to vácyavisesa. 
The tat in tatprakdsinah refers to rosas. 

L The reason: what is meant is a general principle. Only, 
etc.: on the road to rusa it is only what one happens on directly, as 
one is fitting the vibhávas and the like into the combined! form of rasa, 
that can serve as ornament. Accordingly, yamakas and the like are 
always an obstruction to the rosa (aesthetic experience) both of the 
poet and of the reader, [not only in $rrigára but) also in the heroic, 
the marvelous, and other rosas. That our author spoke (in 2.15] of 
avoiding those figures especially in love-in-separation was doubtless for 
the purpose of drawing the attention of persons who are in the rut of 
tradition and who have not attained the height of good taste.? And so 
in what follows he will speak in very general terms, saying, "therefore 
these figures cannot play a part subordinate to rasa." 

The emergence: what he means is its self-generation through the 
favor of genius without any apparent making of it. May seem won- 
derful: that is, one wonders how it could have been formed. 

(In the verse quoted from Amaru] a lady is described as resting her 
face on her tender hand, her lower lip pale because of her sighs, her 
throat choked by a welling stream of tears. and her breast shaken by 
her steady sobbing. The lady, unwilling to renounce her anger, is being 

$246 L] 271 

appeased by the speaker with flattering compliments. And while he 
does so his mind concentrates on relishing the symptoms (anubhávas) 
[which appear in the lady] of this love in separation caused by jeal- 
ousy. To this speaker (who is here the poet] such figures of speech 
as ambiguity, metaphor and contrast emerge effortlesssly and cause nc 
interruption of his relish, any more than they do of the relish of the 

Characterized: he means invariably characterized. Repeatedly 
(prabandhena): the word construes with the word “being made" (kri- 
yamáne); and since that which is made repeatedly must be made with 
intent, he uses the word "intentionally."? A separate effort: an effort 
other than that of joining one component with another into a rasa. 
These spontaneous figures are described as being difficult (nirüpya- 
mánáàni durghatanani), that is to say, even if one wishes to make them 
one could scarcely do so. And they are difficult when described (nirüp- 
yamáne durghatanüni), that is to say, they arouse the reader's wonder 
at their having been made. Will rush to present themselves (aham. 
Pürvikayó parápatanti): the Sanskrit term derives from the phrase “me 
first" meaning "I will go first." The abstraction of this is ahampürviká. 
meaning a situation where each one seeks to go first. The element 
aham is a particle of the same phonetic structure as an inflected form 
of “I” and with the meaning of “I."* This: viz., this rush to present 
themselves. Those few refers to those used by Kalidasa and others.? 

[In the verse summary] even if he is well able construes with “he 
must make a separate effort,” not with with what follows. These 
refers to yamakas and the like. What was said above (in 2.15-16| is 
here summarized in its essentials in the half verse: in the case of 
love, the soul of suggestion. 

1. samavadhána is a technical term of Nyàya meaning co-presence, e.g. 
the presence of z together (with y) in z. See Upaskdra on VaisS. 3.1.12 and 
3.2.1 (p. 103, line 11); or Nydyasiddhdntamuktdvali p. 167. line 2. Abhinava's 
use of the term here is very precise. He means when one is forming a vibhóvo 
or the like so as to fit with the co-presence (the presence of all the other 
elements: anubhdvas, vyabhicárins) required in a given rasa. He uses the 
same word again eleven lines farther on 2. Abhinava imagines the author 
to be reasoning as follows. Yamaka is so well-loved a figure that it will be 
impossible to convince a conservative that it should be abandoned in all good 
poetry. 1 may be able, however, to make him see that its most glaring misuse: 
are in bad taste. 3. Abhinava's intention here is to guard against taking 
prabandhena directly with buddhipürvokam. It is not that the poet constantly 

272 [$246 L 

has the intention of making yamakas, but that he makes them constantly 
and we therefore infer that he makes them intentionally. — 4. The doctrine 
of vibhaktiprotirüpakanipdtas goes back to a Ganasütra on Pin. 1.4.57. It is 
useful in justifying irregular And idiomatic expressions. To take the present 
case: if the aham in ahampurvikayd were the real pronoun aham, it would 
have to shift to its stem-form mad in the the compound (P&n. 7.2.98). S. As 
in Raghuvamsa 9.1-54 

A Now it is shown that a whole group of ornaments (i.e., figures 
of speech) can suggest love, the soul of dhvani.' 

1. Le., can suggest srrigdrorase as the predominant meaning of a sentence. 
See 22 K, note 1. 

K The whole group of ornaments such as metaphor and the like, 
when they are used with circumspection in srigérarasa, the soul of 
dhvani, will merit their name. 

A An ornament is said to be a factor that beautifies the element 
of primary importance (Le. the rasa] just as an external ornament 
(e-g., an earring or bracelet beautifies the person]. The whole group of 
expressed figures! such as metaphor and the like, both tbose that have 
been described (in the past] and those that will be described by anyone 
in the future—for figures of speech are endless—this whole group is 
such that if any of them is introduced [in a poem| with circumspection, 
it may serve as a source of beauty for all the varieties of suggestion 
where the passage from the literal meaning to the suggested meaning 
is unperceived (alaksyakramavyarigya). 

§ 2.18-19 A] 273 

1. vdcydlarikdra, expressed figures, as opposed to suggested figures ol 
speech, which will be treated in 2.25-27. 

L Now: he has in mind that the things to be avoided have 
been stated and that it is now time to state the things which should be 
sought. Can suggest: supply "which" and “how,” [so as to understanc 
the passage as “it is shown which group of ornaments can suggest love 
and how they can suggest it^]. Merit their name means to merit 
their name as causes of beauty. That have been described: sc., by 
Bhàmaha and others who have given definitions of the figures of speech. 
And those that will be described: he gives the reason for this in 
the words “for figures of speech are endless.” He means that they will 
be described by other critics, for (critical| genius (also] is endless 

A Now this is the circumspection [that is needed] in introducing 
a figure: 

K The intention must be to keep them subordinate [or helpfull 
and never acting as the chief element; they should be taken up and 
dropped at the proper time and should never be oversustained; when 
sustained [throughout a verse] special care should be taken to insure 
their subordination [or helpfulness|: these are the means of insuring 
the subordination of the figures metaphor and the like. 

A A figure of speech which a poet intends as subordinate will be 
able to manifest a rasa.! as in this example where the poet gives great 
care to the forming of the rasa: 

(§ 218-19 A 

Many times you touch 
the trembling corner of her eye, 
as if telling secrets 
buzzing in her ear, 
or, while she shakes her hands, 
drinking Love's treasure in her lip. 
Here 1 am cursed with asking questions, 
while you, O bee, have entered heaven. 
(Kalidasa, S&kuntala 1.20] 

for here the figure svabhávokti (naturalistic description) of the bee is 
entirely harmonious with the rasa. 

1. "Will be able to manifest a rasa” is the conclusion of the sentence, 
found nine pages later (Text, p. 233, line 5). For the analysis of this sentence, 
which we bave broken up, see Abhinava's remarks below. 

L By the word circumspection' he means what is stated in 
the Kdrika. In the first sloka and a half [of the two Kárikàs| we 
have the means stated of achieving subordination. Of the figures 
metaphor and the like: this should be construed with each [preced- 
ing half sloka].? 

[In the Vrtti on these two Kdrikds] we have a long complex sentence 
[of which the framework is as follows]. A figure of speech which a poet 
desires to present as subservient and not as the principal element; a 
figure which he takes up at the proper time and which he drops at the 
proper time; a figure which he does not push too far; a figure which 
[even if carried throughout the verse] he is careful to keep subsidiary: a 
figure so composed will bring about the manifestation of a rasa.? Within 
this complex sentence other matter has been included in order to give: 
(a) the point which forms the occasion for furnishing an illustration; 
(5) the illustration itself; (c) the application of the illustration to the 
point; and (d) the justification [i.e., showing how the observance of 
a given point is congenial to the rasa|. Such is the structure of the 
passage in the Vrtti. 

Many times you touch the trembling corner: O bee, although 
we are eager for just such flirtatious action and words (as you are ex- 
hibiting), we are cursed, that is, we are burdened, with the trouble of 
seeking the truth of a problem that must be solved (sc., the nature of 

§ 2.18-19 L] 275 

Sakuntalà's parentage, whether it is noble or brahmin). “While you": 
the particle khalu (“of course,” “as is well known”) indicates that the 
bee has gained his object effortlessly. These are the words of Dusyan:a, 
who desires Sakuntalà [and speaks to the bee as if to say]: "How can 
I become the recipient of her sidelong glance? How can [ get her to 
listen to secret words that will suggest my intentions? How can I steal 
a kiss from her even if she is unwilling? All these aims which lie in the 
land of my daydreams have been accomplished by you with no effort." 
For the bee keeps touching the corner of her eye, thinking it a dark 
waterlily, which it does indeed resemble. And as his illusion contin- 
ues, for her eye extends [like a lily ornament] to the opening of her 
ear, he continues to buzz there. And while she is frightened because 
of the timidity oí her natural delicacy, he drinks her lower lip, which 
is as sweet as the fragrance of lotuses in full bloom and is the very 
treaasury of love (rati). Thus the figure of speech, which is a svabha- 
vokti (a naturalistic description) of the bee, has become subordinate 
to that rasa [viz., abhilásavipralambhasrrigára] which is the poet's chief 
concern. Other [commentators], however, interpret the Vrtti's phrase 
bhramarasvabhavoktir alarikàrah to mean "the figure of which there is 
an expression in the course of describing the bee's behavior" and they 
identify this figure as rüpakavyatireka.* 

1. Clearly one must read samiksó in place of samiksya in order to get 
a feminine noun for ukté to agree with. In the next line one must correct 
rupakádir to rupakáder as in the Kàriká. — 2. [n other words, one is to un- 
derstand angitvena rüpokóder, grahanatydgau ripakdder, etc. The same effect 
will be obtained in the translation by substituting “metaphor and the like” 
for "them" and “they.” 3. Thus the relative pronoun yam in yam alañ- 
kdram. Text, p. 224 line 2, is repeated p. 226 line 1, p. 227 line 2, p. 232 
lines 1 and 7, and is finally resumed by the apodeictic sa alankàro on 
p. 233 line 5, where the sentence ends. A Sanskrit sentence (vákya) of course 
admits of more parenthetical material than a sentence of Latin or English 
4. Abhinava is reading into the close succession of the first two images of the 
verse the suggestion that Sakuntalà's eye "stretches to her ear," this being 
the convention by which the erotic poets refer to a beautiful eye that keeps 
casting sidelong glances. 5. The rūpaka (the superimposition of the charac- 
ter of lover on the bee) would be suggested, the vyatireka (contrast) would be 
directly expressed. While this identification of figures is at least possible, it is 
surely not what Anandavardhana meant when he used the phrase bhramara- 
svabhávoktir alarikérah. Whether Abhinava accepts the second interpretation 
is unclear. Usually he refers to other commentators only to disagree with 

276 {$ 2.18-19 L 

them. Accordingly, BP claims that he disagrees here, that is, that he prefers 
the first interpretation. But in 3.43b Abhinava seems to favor the second. 

A “Never acting as the chief element" means never predomi- 
nant. For it sometimes happens that an ornament which the poet has 
intended to be subservient to a rasa or the like actually appears to be 
intended as the chief factor. For example: 

By the imperious command of his discus stroke 
he rendered love's festival for the wives of Rabu 
empty of passionate embraces 

and left only with a kiss.’ 

In this stanza [there appears to be] an intention to make the figure 
paryáyokta (periphrasis)? the chief element although the overall mean- 
ing of the sentence is a rasa or the like.? 

1. The author of the verse is unknown. It refers to Vispu's cutting off 
the head of the demon Rahu with his discus. As Ráhu's disembodied head 
continued to live on—to be the cause of eclipses by its occasional swallowing of 
the moon or sun—it still remains possible for Ráhu's wives to enjoy his kisses. 
But they have been robbed of the full enjoyment of love. 2. Parydyokta: 
see 1.13h L, note 1. The more restricted definition (“an implication of the 
cause by statement of the effect") will apply to our verse. 3. Ánanda's text 
is not as logical as one could desire and we have been forced to add the words 
"there appears to be” in brackets. The overall meaning of the stanza is the 
courage of Visnu ( Vósudevasya pratàpaA), which suggests the relish of heroism 
(virarasa). But the striking effect of the figure of speech, its cleverness and 
brilliance, casts this overall meaning into the shade and what appears as the 
predominant element in the stanza is the gure of speech. The difficulty with 
Ananda's sentence lies in vivakyd. How does he know what the poet intended? 
All we know is the result, from which we may say that the poet seems to have 
intended the figure to be the chief factor. We need the word drsyate after 
vivoksà just as we had it after vivaksito two lines before the verse. But I fear 
to change the reading lest I be charged with emending the author rather than 


§ 2.18-19a LJ 

L He who, by means of his imperious command, that is, his 
untransgressible order (delivered] in the form of his discus stroke, made 
the festival of love to possess a mere remnant in the form of kisses; for 
this festival of love was rendered barren, that is, empty, of its amorous 
sportings among which embraces are the chief element.’ But someone 
may object that what the poet here intended to be predominant was 
precisely the figure paryáyokta and not rosa or the like. So how can 
one claim that the overall meaning of the sentence is rosa or the like? 
But no. For what is intended to be expressed here is the courage of 
Visnu. And this does not appear as a cause of beauty,? whereas the 
paryáyokta does. Although the poem is not faulty on this account, it 
may Still serve as an example of how a subordinate figure can obscure 
the nature of the matter in hand which it was supposed to support. 
And from this there results a certain impropriety. Such is our author's 
view. And so when he states [a few lines below] that "the criticism of 
a great poet is simply a criticism of oneself," it must be clear that he 
did not give this example as an example of a fault.? 

1. Abhinava takes dliriganoddámavilása as a karmodháraya containing a 
bahuvrihi. ln doing so he takes uddáma as a noun in the unexampled sense 
of chief element. We have preferred to give udddma its normal meaning and 
to take the compound as a tatpurusa, literally, “ a wild passion of embraces.” 
2. To explain this puzzling passage I suppose that Abhinava is interpreting the 
Stanza as an instance of rasavadalarikdra, in which the poet intends the initial 
suggestion of the heroic rasa to subserve an ultimate suggestion of love of 
Visnu; see 2.5 b and c A. But in the working out of his verse the poet has given 
greater brilliance to the parydyokta than to the virarasa. This interpretation 
would allow Abhinava to avoid condemning the poem, for the ultimate rasa 
would still be an arigin to the ariga of parydyokta. Only the intermediate rasa 
would be downgraded. 3. | doubt that Ánanda's intention was as Abhinava 
says. Ananda’s opinion here, it seems to me, is that the overall meaning of 
the stanza, which is the heroic rasa. has been spoiled by the cleverness of the 
figure of periphrasis. 

[$2.18-19b , 

A Even when a figure is intended as subordinate (or helpful t 
the rasa], it must be taken up at the proper time and not at the wron 
time. An example of a figure taken up at the proper time is the upamé 
$lesa (simile with ambiguity and puns) of the following: 

[The translation of the second meanings is given in small print belo 
the translation of the first meaning.| 
It is bursting with new buds and pale of hue; 
She longing 
It has just begun to blossom 
sbe stretch witb languor 
and exhibits a reaction to the constant advent of the breezes. 
her sighs 
. this garden vine today on its madana tree 
in her passion 
is like a rival woman. and by my gazing on i 
I shall doubtless make my queen's face Bush with anger. A 
[Harsa-deva, Ratnávali 2.4] 

1. The use of the simile helped out by puns and ambiguities is indee 
felicitous, as Abhinava will point out. The king here describes a vine i: 
the palace garden in terms that will apply to Ratnavali, who has recentl 
arrived in his barem as a servant girl. The verse occurs in the play befor 
the queen knows of the king's new infatuation with Ratnával. The puns an. 
ambiguities, given in small type in the translation, not only confirm the simil 
but prepare us for the scenes of jealousy to come. If the verse had been use 
after the queen's discovery, the figure would not be apt. The word prárabdho 
jrmbhà has been rendered by one translator with "has commenced to yawn. 
The translation carries a sadly inappropriate image to the English readei 
What is meant is the impatient and anticipatory stretching of the body of 
girl who is daydreaming of her lover and forced to bide her time. The sam 
expression is used in SRK 370. The same gesture arising from the same caus 
is described by the Latin poets also; see Juvenal 6.64-65. 

L Itis bursting with new buds, etc.: uddámotkalikám mean 
"whose buds (kalikàh) have arisen (udgatdh)”; also "whose longin, 

§ 2.18-19¢ 4] 279 

(utkalikd) has arisen." Prérabdhajrmbham ksanát means "at a moment. 
i.e., at that very instant, it has begun to blossom"; but jrmbà is also 
a stretching of the limbs caused by love. "Exhibits a reaction.” namely 
the motion of swaying, a reaction on the part of the vine caused by the 
advent of the wind, namely the gentle approach of the spring breeze. 
The phrase also means "exhibits a reaction," that is, manifests the 
fever in her heart, by the succession of her sighs. Samadanam means 
"with a madana, a species of tree" and also "with passion (kàma)." 
Here the figure upamásleso, being placed as trail-blazer to the relish of 
jealous love that is to follow, calls our attention to the enjoyment of 
that relish. So the figure is taken up at just the right moment inasmuch 
as it comes forth just as this rosa is about to begin. That is what our 
author bas in mind. 

[One may] also [note that] the acting out of the primarily intended 
meaning, [name]y that pertaining to the vine,] should be at every word. 
while the acting out of the secondary meaning (which pertains to the 
woman| should be only of the general meaning of the stanza and should 
be effected by updrigas (facial gestures).! On the other hand. it would 
be wrong to give no gesture at all [to the secondary meaning|. But 
enough on this incidental matter. The word “doubtless” (dhruvam) is 
the very heart of this preparation for the coming jealousy. 

1. We wish we knew more about the abAinayo, the acting out a descrip- 
tion by means of gesticulation and expression. that Abhinava has in mind. 
How would one act out the description of the vine? At just what point 
would one use an upáriga to indicate that the description applies also to a 
woman? The argas and updrigas are listed in BANS 8.14 (Vol. 2, p. 3): 
tasya Sirohastorahparsvakotipádatah sad arigáni / netrabhründsádharakapola- 
cibukdny upérigént. From this it will be seen that an upàriga is a facial ex- 
pression, a subtler indication of meaning than an ariga, which is a gesture of 
the body. 

A A figure that has been taken up, if it is abandoned at the right 
time for another figure better adapted to the rasa. will likewise [bring 
about the manifestation of a rosa], as in: 

($2.18-19c A 

You are rakta with your new blossoms 
and I am too with my beloved's virtues; 
the silimukhas come to you, 
so do those shot by Love to me; 
the stroke of a damsel's foot brings joy to you, 
so would it me and both of us should be, 
asoka tree, the same, if fate 
had not made me sasoka. 
[Yasovarman; also Hanumannátaka 5.24]! 

For here the puns and ambiguities (slesa) with which the stanza begins, 
by their being abandoned for the expression of a contrast (vyatireka), 
give strength to the particular rasa [sc., love-in-separation]. It may 
be objected that there are not two figures of speech here, but one 
entirely different figure of speech, of composite nature like a man-lion, 
a figure consisting of ambiguity and contrast together.? But our answer 
would be no, because that composite type of figure has a different 
sort of distribution. The domain of the composite figure is where we 
apprehend the contrast? in the very same word in which we apprehend 
the ambiguity, as in the line: "The god [Indra] is Hari by name (sa 
harih), [whereas] your majesty is saharih (one who bas horses) because 
you have a host of steeds." But in our stanza one word is the domain 
of ambiguity and another is the domain of contrast.‘ If we were to 
imagine the presence of the "entirely different figure" (i.e., sarkara] in 
an area such as this [i.e., so wide as this], no area at all would be left 
for samsrstis 

1. Rakta: “red” and “in love”; silimukhas: "bees" and “arrows”; afoka: 
name of a kind of tree and "without grief"; sasoko: “with grief." The 
Honumannétaka, in which the verse occurs, is a cento made up of verses 
taken from many poets and adapted, sometimes by slightly altering the text, 
to the story of RAma; see Kosambi, HOS Vol. 42, p. civ. The Subhasitàvali 
(1364) ascribes the present verse to Yasovarman, which is not improbable. To 
judge from line c the original context would have been one of love in separa- 
tion caused by jealousy (irsydvipralembha). There is a superstition that the 
a$oka tree will blossom only at the touch of a woman's foot; see Bloomfield 
JAOS 40, 1-24 and Ingalls HOS Vol. 44, p. 111. As regards the lover, the 
stroke of his mistress' foot might serve as his penance and mark the end of her 
anger. But with the adaptation to the Rama story the context becomes dif- 
ferent. Sita is not jealous, but has been abducted. Hence the commentators 
on the Hanumannátaka and on the Rosagarigádhara (KM edition, p. 354), 

§ 2.18-19¢ A] 281 

with invincible logic but questionable taste, explain the kick desired by Rima 
as a reference to an acrobatic position of intercourse. 

The stanza has furnished matter for disagreement to Sanskrit critics for ten 
centuries (see Kosambi's references under SRK 770), so it is not surprising 
that even the present editors bold different opinions concerning the figure 
or figures of speech which it contains. The problem briefly is this. Ananda 
quotes the verse as an example of a felicitous shift from one figure of speech to 
another. The verse begins with similes strengthened by puns and ambiguities. 
The lover in three respects is like the asoka tree. But the lover is separated 
from his beloved so that he is in grief (sosoka), whereas the asoka tree is, as 
its name proclaims, griefless (a-soka). The shift to contrast (vyatireka) in the 
last line is certainly effective in emphasizing the relish of love in separation 

The chief objection to Ananda's analysis is that the stanza does not really 
drop one figure and take up another; the two figures are interlocked. The 
contrast cannot arise without the pun in “asoka” to support it. Ananda 
himself raises this objection (2.18-19d) and answers it, not very happily, by 
pointing to other cases where contrast is effective without the use of puns. 
On this matter J. Masson remarks, “It is perfectly true, as Ananda says, that 
vyatireka can arise apart from slesa. But surely the point is not whether it 
can or cannot theoretically, but whether it does in the case of the verse 'raktos 
tvam,' etc.” Accordingly, Masson finds himself on the side of the obj tor. 
But Abhinava supports Ananda (see below), saying that the opponent's view 
is incorrect because it goes against one's inner feeling for the poem, a feeling 
tbat even the objector must share. Ingalls finds himself basically on the side 
of Ananda and Abhinava. Granted that the figures are interlocked, the charm 
of the stanza derives from its shift, its turpabout. The logic of Ananda's 
argument is not convincing, although Abhinava does much to improve it, as 
we shall point out in the notes which follow. But the poetic sensitivity of these 
two critics, it seems to Ingalls, was correct. After examining what Ananda and 
Abhinava have to say, the reader may come to his own conclusion. 2. This 
“entirely different figure of speech” would be a form of fusion (sarikara) as 
opposed to samsrsti (the association of distinct figures). The critics divide all 
cases of the presence of two or more figures in a single sentence or stanza into 
these two categories. Sarikaro (as opposed to samsrstt) is a figure distinct from 
the interlocked or interdependent members of which it consists. The objector 
is claiming that the asoka-stanza embodies that figure. Ananda will argue 
that it embodies a samsrsti. 3. One should drop the prakdrantarena before 
vyatirekapratitir. It probably arose by dittography from the prakdrantarena 
in the preceding line. — 4. This is strictly true: raktas, silimukhàh, mude and 
asoka are the domain of slesa; asoka and sasoka are the domain of vyatireka. 
The fact that the figures overlap in asoka does not make Ananda’s statement 
false. 5. For samsrgti see note 2 above. The argument from nirvisayatva is 
a favorite with the grammarians. Ananda has already used it before (2.5e A). 

[$2.18-19c L 

L  "Rakta" means red. “I am too": here rakta means "my pas- 
sion has been aroused." One should understand the red color of the 
asoka bloom to be the stimulant (vibháva) which arouses this passion 
And hence the ideas expressed first in each quarter of the stanza are 
to be explained as stimulants [of what is expressed in the second half 
of the quarters].! So this is a hetuslesa. For slesa is very commonly an 
aid to the figures sahokti, upamá, and hetu. This is all that Bhàmaha 
meant in describing $lesa as being "of three varieties, viz., sahokti, upa- 
mā, and hetu.” [n making that remark he did not intend to deny that 
Slesa may be an aid to other figures also.? 

The particular rasa: viz., love-in-separation. The word sasoka, 
which brings in the contrast, also gives scope to such transient states 
of mind (vyabhicáribhaávas) as depression, anxiety, and the like, which 
strengthen the relish of love-in-separation. But: this [different figure of 
speech] is the single figure fusion sarkara, so there can be no question 
of discarding one figure (slesa) and adopting another (vyatireka). That 
is the objector's point. That: i.e., fusion. For fusion is the flashing 
into view of two figures of speech in one area. Thus, the word saharih 
is a single area. (It has two meanings] "he is Hari" (sa harih) and 
possessing horses (saha haribhih). But in our verse: the particle hi 
is here used in the sense of “but” (tu).* The reference is to the stanza 
raktas tvam. One word: the words rakta, etc. [ie. rakta, silimukha, 
mude, which are the area of double meaning]. And another: viz., 
asoka, etc. [i.e. asoka and sasoka, which are the area of contrast|.? But 
it might be urged that if we take the area to be the whole sentence, we 
can still have fusion here residing in a "single area." Anticipating such 
a suggestion, he says if. In an area such as this: that is, a [whole] 
sentence. The singular inflection of visaye is intentional. The sense is, 
if one will call something a single area from its being a single sentence, 
then there can be no association (samsrsti) anywhere because it will 
be logically included in fusion. 

1. In the second line, it is the sight of the bees fying toward the asoka 
flowers that excites the lover's feeling of separation from his beloved. In the 
third line, it is the touch of a woman's foot which gives delight to the asoka 
and may cause the lover to think of his own lady's touching him in the same 
way. He is also excited at seeing the joy of the asoka. Abhinava's point is 
that each element in the description of the osoka tree furthers the lover's 
emotion and is therefore the cause of the second meaning of the ambiguous 

§ 2.18-19d A] 283 

expression. 2. Bhàmaba divides slesa into three varieties (3.17): sahokti- 
Slesa, upamé-slesa, and Aetu-lega. Sahokti-slega is where a word having the 
meaning “and” or "together with" (saha) is used in connection with the am- 
biguity, e.g.. "Being easy to ascend (or approach) and generous of fruit (or 
reward), wayside trees and good men serve the benefits of others.” Upama- 
lesa is where a word meaning "like" is used, e.g., the same example with 
the substitution of "like" (iva or tulya) for *and" (ca). Hetu-slesa is where 
a cause is expressed, e.g., "Because you are unfathomable and because you 
never transgress the proper bounds, you are like the ocean." In this example 
of hetu-slesa (taken from Bhamaha) both the word "like" and an expression of 
cause (the ablative case) are used. [n the asoka-verse a word for “like” is used 
and. according to Abhinava, cause is suggested. Bhàmaha would doubtless 
have identified the figure in the asoka-verse as upamd-slesa, but Abhinava's 
extension of the term Aetu-slesa allows a more suggestive interpretation of the 
verse. Whether Bhàmaha would have allowed $lesa to be an aid to still other 
figures, as Abhinava claims, is uncertain. But the opinion is ancient. Dandin 
says that slesa can increase the beauty of any figure (2.363). — 3. And so this 
one area furnishes both pun and contrast. — 4. And so we have translated 
it. However, 1] think that the literal meaning of atra Ai ia Text 229, line 1 is 
the same as of the atra hi with which 2.18-19c A begins. The sentence here 
is furnishing another reason to explain Ananda's disagreement with the ob- 
jector (iti cen na Text 228, line 3). But Abbiaava is fond of this explanation 
of the particle Ai (e.g. 3.6a L) and uses Ai in this sense himself (3.33k L, 
Text 434, line 7). — 5. The correct reading is doubtless asokádih, to which 
some absent-minded scribe added safoka without thinking to remove the édi. 
As the text stands it is nonsense, for it is only the words asoka and sasoka 
which carry the vyatireka. 

A Should one object that our stanza is not an Instance of sam- 
srsti because the contrast comes into existence only by means of a 
pun [whereas samsrsti demands that the two figures be independent 
of one another], our reply is no, for we see that contrast can arise ina 
quite different manner also (i.e., without a si ile based on puns].' For 

[$2.18-19d A 

The wind of doomsday, whose fierce blast 
will tear down mountains, will not blow it out; 
its beauty blazes far by day, 
it is quite untouched by the soot of night. 
Patoriga gives birth to this unique wick 
nor ever serves to extinguish it: 
may this light for the lamps of every land, this sunlight, 
ever bring you joy. 
[(Mayüra, Süryasataka 23]? 

This is an example of contrast without the expression of any similarity 
at all? Again, it cannot be said that we perceive there (in the asoka- 
stanza] a cause of beauty to lie in the puns alone [without regard to 
similarity] and that the puos must therefore be intended to be subor- 
dinate to the contrast and not an independent figure on their own.* 
Because in such an area [viz., the area of contrast] we find that beauty 
can result also from a well-stated simile alone [without puns], as in the 
following and other examples: 

My groans are like your thunder; 
the water of my eyes, your ceaseless downpour; 
the Gre of grief born of her separation, 
is like your flashing lightning; 
I bear my loved one's face 
within my heart, you hide the moon in yours: 
in all this we are similar, friend Cloud; 
why then would you destroy me? 

1. Ánanda's sentence suffers from both ellipsis and bad logic. Abhinava 
fills in the ellipsis, as we have done in brackets. But the conclusion does not 
follow from the premise. One cannot deduce a particular from a particular. 
For Abhinava's attempt to improve the logic, see his commentary below and 
our note 1 thereon. 2. The stanza contaios a pun in patariga: the sun 
(which gives birth to sunlight), or a moth (which serves to extinguish the 
wick of a lamp). So Ananda will furnish another example below of contrast in 
the complete absence of puns. — 3. There is no expressed simil ity because 
the stanza contains no word "like" or “similar to (tulya)." There is of course a 
suggested simile as Abhinava will point out. But Ananda overlooks this point 
because it is unimportant. See end of note 1 on Abhinava's comment below. 
4. The disavowed statement, if true, would imply that the verse contains 
the figure of interlocked members sarikara and not the two figures slesa and 

§ 2.18-19d L] 285 

vyatireka beld distinct in samsrstt. — 5. The verse occurs in SRK 240. It is 
attributed to Yasodharman by Sodukti. 993; to Ya$ovarman by SuktM. 43.33. 

L Now an objection may be raised. Contrast always contains 
asi ile within itself; and in the case here [of the asoka-stapza| the 
Simile has been brought in by means of puns. So the puns in the 
verse are aD aid to the contrast. Thus the stanza is the domain of the 
figure fusion. On the other hand, where there is no relation of aider 
and aided between the two figures, that can be a case of association, 
even if the two figures reside in one sentence. This [is the objection 
which} our author now states. By means of a pun: what be means 
is "by means of a simile which is brought in by force of a pun." Our 
author counters the objection with no. What he has in mind is this: 
Does contrast everywhere arise only when the si ile (or similarity] is 
directly expressed, or does it arise wben it is implied?! He rejects the 
first alternative by saying: in a quite different manner. He means, 
"even thougb thesi ile [or similarity] is not directly expressed. 

"Will not blow out" means "cannot extinguish.” The flame of a lamp, 
on the other band, can be extinguished by a mere breeze. "The soot of 
night”: soot in the form of night. "Not not freed from" means “quite 
untouched by." The wick of a lamp, on the other hand, is accompanied 
by night (or darkness) because, as the wick is very slender, soot envelops 
its surface. “From patariga, that is, from the sun." The wick of a lamp, 
on the other hand, is extinguished by a patariga, that is, by a moth, 
and does not take its birth therefrom. 

Similarity: he means, with a series of similitudes or similes but 
without any word, e.g., "like" or "such as," proper to the conveying of 
this [relation]. This is as much as to say that as a simile by being merely 
suggested can be helpful to a contrast, the simile need not be expressed 
in so many words. Accordingly, the slesopama (simile occasioned by 
puns) was not taken into our verse for its being helpful to the contrast. 

But, [a new opponent may object,| granted that this may not be the 
reason in other cases, in the case of the asoka-stanza the slesopamá 
has been taken in because it was favorable to the contrast, because 
if it were not favorable, it would have no power to beautify. This is 
why the slesopamé cannot be a separate alarikdra (figure of speech = 
beautifier). This [is tbe objection which] our author now states with it 
cannot be said. Our author realizes in his heart that the opponent's 
view is wrone because it goes against one's inner feeling [for the versel. 

286 [ § 2.18-19d L 

Accordingly, he produces an example in which beauty arises without 
puns, simply through simile, which silences the opponent, who is really 
denying his own inner feeling. He does this with the sentence because, 

In the exemplar verse the word “like” is to be construed with all 
the words ending in the instrumental case. Everything else is to be 
understood as in the case of the stanza on the asoka? 

1. Ananda’s argument is actually different from that which Abhinava 
here supposes. Ananda in effect shows that (a) contrast can be produced 
by puns without simile, and (b) contrast can be produced by si ile without 
puns. From this he passes to the conclusion (c) that the contrast in the asoka- 
stanza is not in need of either puns or simile: the figures of that stanza are 
independent. This is arguing from a particular to a particular. Doubtless 
Abhinava saw the logical fault, for he makes an attempt to remove it. He 
supposes an objector to base an argument on a universal, thus: 

Contrast always involves simile, 
the simile in the asoka-verse is slesopamó, 
therefore the contrast and the flesa are i 

This argument can be logically disproved by showing a single case where con- 
trast does not contain a simile. The verse from Mayüra is such an example, 
for there is no expressly denoted simile in it. Nothing is said of its suggested 
simile, for that is beside the point. A suggested simile falls in the category 
of dhvani, not of figures of speech. Technically, the interlocking of 5lesa and 
suggested simile does not produce sarikara. 2. BP explains that the thun- 
dering of the cloud is to be taken as the cause of the speaker's groans, and so 
on for the other elements of the verse; see 2.18-19c L and note 1 thereon. 

A Again, when the poet's mind is concentrated on carrying out 
the rasa, a figure which he will not wish to press too far [will bring 
about a manifestation of the rasa].! For example: 

§ 2.18-19e L| 

In anger she has bound him 
tightly in the noose of her soft arms 
and in the evening leads him to the bedroom, 
where before her attendant friends 
she points to the signs of his deceit and conjures hi 
never, never to do such a thing again. 
O lucky lover: 
as he is hiding his transgression with a laugh, 
she weeps and strikes him. 
[Amarusataka 9|? 

for here a metaphor (rupaka), being introduced but not fully carried 
out. greatly strengthens the rasa. 

1. The bracketed words follow in Text 233, line 5; see 2.18-19 L and 
note 3. 2. The stanza belongs to a type of Sanskrit verse which depicts 
a man's delight in provoking an outburst of jealousy from his mistress; see 
SRK 682. The essence of the situation is furnished by the collocation of the 
two contradictory words (almost an oxymoron) at the very end of the stanza: 
rudatyd hasan “he laughing (is struck) by her weeping.” 

L Having thus shown the application of (the advice for] taking 
up and dropping a figure of speech, he now explains the portion [of 
tbe advice] that says “they should never be oversustained." (Again, 
when carrying out] the rasa: the word "again" (ca) serves to add 
another variety of circumspection [see 2.18-19 A, note 1]. Were one 
to continue the metaphor furnished by the woman's creeper-like arms" 
acting as a noose for binding, the woman would become a huntress, 
the bedroom would become a prison or a cage, and so on, all of which 
would be most inappropriate. 

"Before her friends": the implication is that they have all along been 
telling her that her lover is faithful; well, let them just look now. "[Con- 
jures him] with faltering voice," that is, with a voice that is indistinct 
because of her access of anger and that is also a sweet voice. And what 
does she say? "Never do such a thing again" What she means by 
"such a thing" is shown by the word "deceit," for example nail-marks 
and the like {imprinted by a rival], to which she points with her finger 
"She strikes him": she cannot be held back by the placations of her 
friends, because he is intent on hiding his fault under the pretext of 

288 [52.18-19e L 

laughing and because he is so dear to her. Who would be able to bear 
the infidelity of such a man? 

A Even if the poet decides to sustain the figure, if he takes care 
to keep it subordinate, [it will bring about a manifestation of the rasa], 
as in the following and other examples: 

I see your body in the syàmá vines, 
your glance in the startled eyes of deer, 
your cheek in the moon, your bair in tbe peacock's tail, 
the play of your eyebrow in the rippling stream. 
Alas, my timid darling, I can nowhere 
find your complete likeness in one place. 
[Kálidása, Meghadüta 2.41 = Pathak ed. 109}! 

1. Bhiru (timid), on which L expatiates, is a characteristically Kashmiri 
reading, found also in Vallabha's commentary and in the Subhdsitdveli, as 
opposed to the vulgate candi (cruel). 

L [Decides] to sustai is means, to carry all the way 

In the $yámà vines: that is, in the fragrant priyarigu (Aglaia 
oderata), because it is pale, slender, and kantakita ("thorny" or “ex- 
hibiting horripilation"). In the moon: because of its paleness. "I see 
(utpasyámi)," that is, by effort I fancy. The sense is that he indulges 
this fancy in order to preserve his life. Alas: what a misfortune. The 
implication is that since I cannot find: your whole likeness in one place, 
I wander about, but wherever I go I lack the satisfaction of a single 
likeness of you in anything. Timid: the sense is that one of timid 
heart will not put all her wealth in one place. 

In this stanza the similarity [between the Yaksa's wife and various 
objects of nature], which animates the figure utpreksd (poetic fancy), a 

§ 2.18-19g L ) 289 

sustained fashion just as it was begun and yet gives strength to the 
[dominant] relish of love-in-separation 


A À figure of speech which a poet forms under these precau- 
tions will bring about a manifestation of rasa. On the other hand, if he 
departs from these principles, a loss of rasa will certainly ensue. Exam- 
ples of such loss are to be found in abundance even in the works of great 
poets. But we have not demonstrated these lapses because publishing 
the faults of great men who have shown their greatness in thousands of 
fine verses would be simply a criticism of oneself. However, the general 
direction has been given of how the whole collection of figures of speech 
such as metaphor and the like can be useful in suggesting rasa. If a 
good poet, with concentrated mind, will follow this lead, discovering 
still other principles on his own account, and will thereby construct 
the soul of dhvani, of which we have just spoken, (that is, a suggested 
meaning] which appears without a perceived interval, then this (type 
of dhvani] will arise in all its glory.’ 

1. Tasya (Text. p. 234, line 1) refers back to dhvaner à: 

L But these, sc., examples, should be construed with the word 
"demonstrated." With however, etc., he shows that although no coun- 
terexamples have been given, one will achieve what is necessary by 
studying the positive examples. Other principles: he means other 
types of precaution; for example the taking up again at the right time 
a figure that has been dropped, as in a verse of my own composition: 

If the rays of the moon are pencils of ambrosia 
how come they to burn my heart? 

Or were they long ago infected 

by their dwelling with the kdlakuta poison? 
Then how is it they have not destroyed my life? 
Perhaps it is saved by the magic syllables 

in telling over my beloved's name 

Then why do I faint? 

Mic) sica. Tuo eue 

290 [§ 2.18-19¢ L 

In this verse the figures ripeka,? sandeha,’ and nidarsand,‘ being suc- 
cessively abandoned and then again taken up,” lead to a strengthening 
of the rasa. And so enough. 

1. The moon is traditionally supposed to be the reservoir of ambrosia 
drunk by the gods, but it once dwelled in the sea where one of its companions 
was the kdlakita poison. Pencils of ambrosia (amrtacchatah): the dictionary 
definitions of chatd are misleading. It is regularly used as a noun adjunct 
for things which are long and slender, e.g., drsticchaté (Text 309, line 6 of 
Loc.), sasidyotacchatà (Udbhata, 2.°15 Indurája), katáksacchatà (SRK 465). 
2. There is rüpaka (metaphor) in the phrase "if the rays are pencils of am- 
brosia." The figure is abandoned in the next clause for sandeha. — 3. There 
is sandeha (poetic doubt) in "or were they long ago infected," etc. [t is aban- 
doned in the next sentence for nidarsand. 4. The phrase "saved by magic 
syllables in telling over my beloved's name" (priyatamásanjalpamontráksaroi 
raksyante) contains the figure nidarsand. This is the later conception of 
nidarsand (from Udbhata onward), more technically called asambhavadvastu- 
sambhavanibandhaná nidarsand. This type of nidarsand occurs where a rela- 
tion (in the present example the relation between roksyante and priyatamá- 
sanijalpa) is logically impossible unless we envisage a simile (unless we take 
the whole phrase to mean priyatamásanjalpena mantréksarair iva raksyante), 
in which the qualifiers (visesanáni) of the upamána and the upameya (in the 
present instance priyatamé and mantra) appear as image and reBection (bimba 
and pratibimba) of one another. In general, whenever a rüpaka is extended 
by visesanas qualifying the upameya and upoméne such that these visesanas 
appear as bimba and protibimba, the figure is called nidarsand. not rüpaka. In 
the stanza under discussion the nidarsanó is given up in the question which 
follows. 5. Apparently the last sentence of the stanza "I know not what to 
think" (no vedmi keyam gatih) is regarded as reestablishing all three of the 
preceding possibilities and therefore reviving the three abandoned figures. 

K That form oí suggestion which appears after an interval and 
which is similar to the reverberation [of a bell] is itself distributed into 
two varieties: it can be dependent on the [suggestive] power of words 
(sebdasaktimüla), or it can be dependent on the [suggestive] power of 

§ 2.21 Introduction A | 

A Of suggestion where the literal meaning although intended 
leads to a further sense, the variety which is similar to a reverberation 
because the suggested meaning appears at an interval (from the literal 
meaning] has itself two varieties, one based on the suggestive] power 
of words and the other on the [suggestive] power of meaning. 

L Having thus considered the first variety of that suggestion 
where the literal meaning although intended leads to a further sense, 
that is. the variety where the sequence is not perceived. he now proceeds 
to analyse the second variety with the words (that form of suggestion 
which appears] after an interval. These words of the first quarter 
of the Karka, resuming a subject from the earlier discussion, are used 
to furnish a reason [for the relative clause that follows}. The resonance 
of a bell always appears at an interval from the sound produced by 
the striking of the bell. Is itself: not only is suggestion basically 
divided into two varieties [viz., avivoksitavdcya and vivaksitényapara- 
vácya]; and not only is vivaksitányaparavácya divided into two varieties 
[viz., asarnloksyakramavyarigya and sarnlaksyakramavyarigya|; but even 
samlaksyakramavyarigya is divided into two varieties (viz., sabdosakti- 
‘mula and arthasaktimüla).! This is the force of api (“also,” "itself"). 

1. Samioksyakramavyorigya really has a third variety also, based on both 
the power of words and the power of meaning. See 2.23. 

A Now it may be objected that if this domain, where a second 
meaning appears by the power of a word, is given to a variety of sugges- 
tion (dhvani), there will be no domain left for the pun ([sabda-]slesa). 
Rut ane can show that this is not the case: 

[5221 K 

K because it is where a figure of speech appears by the power 
of words, being only implied (àksipta) and without being directly ex- 
pressed by a word, that we have this [variety of] suggestion tbat arises 
by the power of words. 

A Since what we mean is, that a figure of speech, not a mere 
fact, appearing in a poem by the power of words, constitutes this type 
of dhvani arising by the power of words. But when two facts appear 
by the power of a single word. we have the [directly expressed] figure 
of the pun (slesa), as in the following: 

(In the punning verses which follow, the translation of the second 
meanings will be given in small print below the translation of the first 

He who destroyed the cart and is unborn 

He who destroyed the miad-born god 

and who once, the conqueror of Bali, made his body into a woman; 

and who once made the body of the conqueror of Bali into his weapon; 

who slew the upraised serpent; whose ultimate state is sound; 

whose necklace and bracelets are raised serpents; 

who upheld the mountain and the earth; 

who upheld the Ganges; 

to whom the immortals give the praiseworthy name 

to whom the immortals give the praiseworthy name 

of the Seizer of the head of the moon-destroyer; 

of the Seizer (hara) who bears on his head the moon; 

who brought about a habitation for the Andhakas; 

who brought about the destruction of Andhaka; 

who is all-giving and named Madhava: 

who is always the husband of Um: 

may he protect you.’ 

may he protect you. 

§ 2.21 L] 293 

1. The mythological references are explained below by L. The stanza is 
ascribed by SuktiM. 2.104 to Candraka, by Sadukti. 163 to Bháravi. It is left 
anonymous by SubhA., Mammata, and Hemacandra. 

L By since he explains the word "because" of the Karikà. By 
not a mere fact he shows what the Kàriká means to exclude when it 
says “a figure of speech.” But when two facts: the word ca (“and”) 
is used in che sense of “but.” 

[Interpretation of the verse as referring to Visnu:| He who destroyed 
the cart while he was playing as a child; the unborn, that is, who is 
without birth; balijit [Abhinava takes the word otherwise than as we 
have taken it]: "he who conquers the strong ones (balinoA), viz., the 
demons; who made his body into a woman long ago at the time when 
the ambrosia was churned from the sea; who slew the upraised, that 
is, proud, serpent named Kaliya; whose dissolution is into sound, for 
it is said that Visnu is the phoneme a;' who [as Krishna] held up the 
mountain Govardhana and [in bis boar incarnation] raised up the earth 
from Pátàla; to whom the sages give a praiseworthy name. What is 
this name? He who destroys (manth + null-suffix kvip) the moon, 
Ráhu; [Visnu's name] is "remover of the head of Rahu.” May this 
M&dhava, that is, Visnu, who is all-giving, protect you. [Again] how is 
he'quali&ed? As he who gave ksaya, that is, habitation, to the tribe of 
Andhakas at Dvárakà; or one can take the term to mean he who, at 
at the time of the Mausala battle, made ksaya, that is, destruction, of 
the Andhakas by means of the iron reeds. 

The second interpretation (referring to Siva}: Who, having destroyed 
Kama, transformed the body of the Destroyer of the Strong Ones, 
that is, of Visnu, into his weapon, that is, his arrow, at the time of 
the burning of the Triple Citadel; whose necklace and bracelets are 
upraised serpents and who bore the Ganges; to whom the sages give 
the praiseworthy name "he whose head carries the moon" and who is 
called Hara; may he, the blessed one, who himself was the cause of the 
demon Andhaka's destruction and who is always, that is, at all times, 
the husband, the lover, of Urná, protect you. 

In this stanza the second meanings which we perceive are simply facts 
and not a figure of speech; and so it belongs wholly to the domain of 
the pun (slesa). 

294 (§ 2.212 

1, Short a is the source of all the phonemes, from which comes the Veda, 
from which comes everything. God (Visnu) in his ultimate form is sabda- 
brahman, the verbal source of the universe. 

A But now there is a difficulty. Udbhata has shown! that even 
when a separate figure of speech appears [together with slesa (an am- 
biguity or pun)], that [combination] is to be designated an instance of 
the figure slesa. So now it appears that there is no domain left for 
dhvani based on the power of words. In anticipation of this difficulty, 
the Kdriké has used the word "implied." Here is what is meant. Wher- 
ever by the power of words a figure of speech appears in addition [to 
Slesa|—this figure being directly denoted—all that is the domain of 
Slesa. But where by the power of words a figure of speech appears in 
addition [to slesa], this figure being implied by the inherent capability 
of the situation (sámarthyóksipta)? and not directly denoted—in other 
words, being suggested (vyarigya)—all that is the domain of dhvani. 

The direct appearance by the power of words of a second figure [to- 
gether with $lesa| may be seen in the following: 

As even without a necklace 
they had a natural charm, 

in whom did this maiden's breasts 
not arouse wonder?? 

Here a transient state of mind (vyabhicáribháva) of love, named “won- 
der," and the directly expressed figure of speech virodha (contradiction) 
both appear [together with a pun]. So this is in the domain of slesa 
favoring the semblance of virodha.* It is not in the domain of that type 
of dhvani which is like a reverberation.? However, it is in the domain 
of the type of dhvani where the suggestion appears without an interval, 
the suggestion being here suggested either by the pun or the virodha.* 

sla ky 295 

1. This refers to Udbhata, Induràja 4.10 ( Vivrti 4.24), where he speaks 
of slesa as "generating the appearance of other figures of speech" (slistam ... 
alankdrdntaragatdm pratibhám janayat padaih). Both his commentators inter- 
pret his words to mean that where we have the impression both of slesa and of 
some other figure such as upamá or rüpaka one is to identify the figure as Jlesa 
only. The reason given is anavokásatvál, i.e., that $lesa has no other scope. It 
never appears without without the appearance of some other figure, whereas 
those other figures do have their own spheres where slesa is not present. If 
we do not ‘ve the overlapping instances to slesa. it will have no scope at all. 
Another solution, of course, would be to recognize the symbiotic nature of 
Slega and give it no independent domain. The overlapping instances would 
then be instances of fusion. In cases where slesa and a second figure are mutu- 
ally dependent, Abhinava (in his comment on this section; see below) claims 
that Ananda follows this anti-Ubdhata explanation. But that seems unlikely. 
2. The translation is clumsy but I cannot find a simpler English phrase that 
does not falsify the meaning. Sámarthyáksipta. taken most literally, means 
"implied by the capability (inherent in the situation or in the sentence." Com- 
pare 3.30c L arthasámarthyàd iti vákyárthasámarthyád iti yávat (Text 411, 
Loc. lines 2-3). One may render it less literally by "implied by the available 
possibilities.” or “implied by the principle of compatability." Compatability 
is one of the glosses given to sémarthya by the grammarians on Pan. 2.1.1. 
3. The verse occurs without ascription of author in SubhA 1534. The 
word virodhaccháyá, which we have translated “a semblance of virodho," is not 
a synonym of the later term virodhábhása. Ananda does not mean that the 
contradition itself is false or merely apparent, but that it forus a false figure. 
The Ggure is not really virodha because the dictum of Udbhata (see note 1 
above) requires us to identify the bgure as slesa.— S. Naturally not, for there 
is no suggested Bgure of speech in it. — 6. Ananda means that the stanza 
tasyà vindpi Adrena suggests srrigárarosa. It would thus fall under the type 
illustrated at 3.4d A, where rasadhvani is helped out by figures of speech. 
It is disconcerting, however, to find the actual name of a vyadhicdrin (viz., 
vismaya) given in a stanza which is said to suggest rasa. as this goes against 
the principle laid down in 1.4g A. Presumably Ananda regarded the word as 
merely incidental. Certainly it does not add to the rasa Finally, one wonders 
why Ananda wrote vá ("or") instead of ca (“and”). Abhinava's explanation 
seems unnatural. 

L  Toshow what the Kai means to exclude by the word “im- 
plied,” our author begins with a possible objection: But even 
when a separate figure of speech. 

As even without, etc.: Here the word “even” (api) directly de- 
notes a contradiction! and so forces the denotative power [of Adrineu] 

296 {§ 2.21a L 

into a double meaning. Hdrinau may mean "they must? captivate 
one's heart,” or it may mean “possessing a necklace (hara).”* And it 
is because of this [presence of api) that the word vismaya (wonder) 
strengthens this very sense. For if the word api were not present, there 
would be no direct denotation of a double meaning, since the natural 
charm of the maiden's breasts could be taken as the cause of wonder.‘ 

A state of mind named wonder: he uses this phrase to furnish 
an illustration. Just as wonder appears directly through the very word 
“wonder” [and not as suggested by the description of an anubháva or 
the like], just so does the figure of contradiction appear directly through 
the word “even” (api) [and is not a suggested figure]. Now one might 
doubt that there is any suggestion at all in this stanza. To allay that 
doubt he says, however, etc. Or by the virodha: by the use of 
"or" he shows that this is a case of the figure fusion composed of an 
interlocked slesa and virodho, for the word "or" indicates that due to 
the presence of mutual aid? there is no basis for rejecting the one and 
accepting the other. 

1. When the word opi is used, a virodha is said to be directly expresed 
(vácya). When it is omitted, the virodAa is said to be suggested (vyarigya). 
Thus Rangacbarya Raddi Shastri, commenting on Dandin 2.333, quotes Vā- 
mana's exemplar verse (KASV and says esa ca apisabdaprayoge vā- 
cyo, anyathd vyarigyah. 2. For the interpretation of the -in suffix (nini) 
as indicating necessity, see Pan. 3.3170. 3. Without the word "even" we 
would take the word Adrinau only in the first, its common, sense. The word 
“even” makes us look for a contradiction, which we find in the second, un- 
usual, meaning. 4. As the verse stands, the reader understands it to mean 
tbat men wondered at the contradiction, viz., that the maiden's breasts could 
have a necklace without having a necklace. Lf the word “even" were omit- 
ted, the reader might understand the verse to mean that men wondered at 
her breasts, which happend to be without a necklace, simply because the 
breasts were beautiful. 5. The presence of mutual aid (anugrahayogat): 
What Abhinava means is that the relation of aider and aided (anugrahydnu- 
grahakabhdva) between slesa and virodha is mutual. If there were no pun 
(if, for example, we substituted the the word manoharau for Adrinau), there 
would be no contradiction. Again, if the contradiction expressed in api did 
not make us look for a second meaning, there would be no pun. According to 
Abhinava, Ananda would go against Udbhata's dictum (see 2.21a A, note 1) 
in such instances and identify the figure not as slesa but as sankara. This is 
putting words into Ánanda's mouth that he might reject. 

$221b L] 

A A similar instance occurs in a verse of my own: 
As he holds in his hand the discus "Beautiful" 
As he is one whose band is beautiful 
while her whole body merits praise; 

as he had stepped across the universe 
surpassed the universe 

with the graceful motion of his lotus feet 

while she had conquered the universe 

with the grace of every limb: 

as he bears an eye that is the moon 

while her whole face is of lunar beauty: 

it was with reason that Hari regarded Rukmini 
as more precious than his very self. 

as superior to his own body. 

I pray that now she give you aid." 

In this stanza slesa appears, favoring a vyatireka (contrast) through its 
direct expression. 

1. The name of Hari's (= Visnu's) discus is "Sudar$ana" (the name means 
"Beautiful"). In three steps Visau covered earth, sky, and heaven. In his 
cosmic form his two eyes are the sun and the moon. Rukmini was his favorite 
wife when he lived as Krishna in Dvaraka. 

L Normally sudarsanakara will mean "he who carries the discus 
Sudarsana in bis hand." But in the alternative demanded by the con- 
trast it will have to mean "he whose hands only, (not the other parts of 
his body,] are beautiful." "The graceful motion of his lotus feet," that 
is, the playful action of stepping across the three worlds. "Bearing aD 
eye": possessing an eye that consists of the moon. 

Through its direct expression: because the contrast is directly 
expressed by the words "superior to his own body." 

[$221c A 

A Aad another example is: 

The cloud serpents pour forth water 
pour forth venom 
which brings to ladies whose husbands are away 
à sudden dizziness, a listlessness and weariness of heart, 
then fainting, darkness, emaciation, death 

Or again: 

Whose war elephants are his mighty arms 
whose fame has spread through their crushing 
who have scattered pollen by their crusbing 

of the golden lotuses of the hearts of his foes 
of the golden lotuses of Lake Mánasa. 

and whose flow of gifts is ceaseless 

and who flow constantly with ichor.” 

In these two stanzas slesa appears as aiding the semblance of metaphor 
(rüpaka)? through its direct expression. 

1. The ascription to Sakavrddhi is furnished by Subhd. 1538, which 
Quotes some twelve verses by this poet. Ananda quotes the stanza again 
at 320b A. 2. The verb camadha is given by Páiasaddamahannavo as 
Meaning to crush; hence camadhia (here camohia), crushed. “Of his foes” is 
Supplied by Abhinava. The words mánasa (heart; also the lake of that name) 
and dana (gift, or ichor) are puns. In the last quarter of the Prakrit one must 
read ccia for vvia (see Hemacandra, Prák 2.184). This should be translated 
by eva, not iva. The word iva would produce upamó. not rüpaka. It may also 
be remarked that Sanskrit has trouble in rendering the grammatical number 
of the Prakrit nouns. As the ending -à represents equally the dual or plural 
in Praksit, the war elephants (plural) of the king can be likened to his arms 
(dual) without any grammatical difficulty. But in Sanskrit with its distinct 
inflections this would constitute a fault, which the Sanskrit translator has 
avoided at the expense of giving the king more than two arms. 3. In the 
first Stanza the compound joladabhujaga must be analyzed as a rüpakasamása 
into the elements jaladé eva bhujagáh, where the eva shows that the metaphor 
is directly expressed. For the distinction of rüpakosarnásas, which are justi- 
fied by Pan. 2.1.57. from upamitasamásas. which are justified by Pan. 2.1.56, 

$221d A] 299 

see Nágoji Bhatta on Mammata 10, example 421. and Rangacharya Raddi 
Shastri on Dandin 2.66. 

L In view of the meaning of “serpent.” the word visa cannot 
stop at its normal denotation of water but is forced to denote its sec- 
ond meaning, "venom," for without that meaning the denotative power 
[of the sentence! cannot be completed.! Dizzinesss and the other [afflic- 
tions], on the other hand, are common [to both interpretations].? 

(The application of the second verse to the king is as follows.| The 
hearts, that is the hearts of his enemies, have been crushed by having 
been plunged in despair. These hearts are called lotuses made of gold 
because of their strength. Because of these [broken hearts|, his “fra- 
grance has been churned,” that is, the essence of his valor has been 
spread abroad. His war elephants are his mighty arms, which flow con- 
stantly with gifts. Because of the presence of the word "war elephants" 
the words camahio, parimala, and dàna transmit the respective mean- 
ings crush, perfume, and ichor; but they have not completed their 
denotative function in so doing and they go on to denote the other 
meanings which I have enumerated above. 

1. Water does not normally bring dizziness and death. To make sense 
of the literal meaning of the sentence, we have to understand the second 
literal sense of visa, viz., venom, as well as the first, water, which was needed 
for the connection with clouds. — 2. -Dizziness and the other afflictions can 
‘result equally from snake-bite or from the absence of one's husband during 
the monsoon. Accordingly, the words bhromim. etc., exemplify arthoslesa, 
not Sobdosleso. 3. The heart is often likened to a lotus. But BP notes the 
instance in KumSam. 5.19 where Parvati's heart is likened to a lotus made 
of gold because, while tender, it can endure the most severe exertions. The 
text of L is mispointed; one should place the danda after sasáratvát instead 
of before. Again, in the next line, one should place the danda after prasrta- 
protàpasara(h) instead of before 

A Where the figure of speech, although implied, is then again 
directly expressed by some other word [in the stanza], we likewise can- 
not speak of dhvani in the form of a resonance arising by the power 

300 [52214 A 

of words. In such cases we must speak of some directly denoted figure 
such as vakrokti or the like,' as in this example: 
O Keéava, my eyes were blind 
my judgment was blinded 
with the dust raised by your cattle; 
by passion for the cowherd: 

I could see nothing and so I stumbled; 
1 fell from virtue; 

why do you not help me up, my lord? 
why do you not take me as a husband [takes a wife]? 
You are the one refuge of the weak 

of women 

when their hearts fail them on rough roads. 
io their troubles. 

Thus did the Gopi once express a hint to Hari, 
who now, I pray, may grant his help to you.? 

Everything of this sort we would place in the domain of the expressed 
figure slesa. 

1. Vakrokti in its slesa variety is the figure exhibited in the verse which 
follows. Ananda uses the term here in the sense of Rudrata 2.14 and Mammata 
9.1 (= sūtra 103), of a specific figure (sabdálarikára) rather than in the general 
sense of “ornamented speech" employed by earlier writers. Mammata's defini- 
tion is " Vakrokt: (crooked speech) is when a sentence expresses one meaning 
taken in one way and another taken in another, the combination being effected 
by puns (4lesa) or by tone of voice (kdku).” Vakroktt of the first type is really 
nothing more than a complex or extended sleşa. — 2. The stanza is quoted 
in SuktiM. 2.93, where one MS ascribes it to Tribhuvanapala. 

L Having thus shown what was meant to be excluded by the 
word "implied" (dksipta) [in 2.21 K], he now proceeds to show what 
is excluded by the word "only" (eva): where the (figure although 
implied], etc. The meaning of this sentence is as follows. When words 
capable of denoting two senses are employed. if there is no reason for 
restricting the power of denotation to only one of these senses, as, for 
example, in the stanza "He who destroyed the cart" (2.21 A], or to 
take a different situation, when there is some reason [such as the use of 
words like "even"| which awakens us to the presence of a second power 
of denotation [which conveys a second sense]. as for example in the 
stanzas beginning from "As even without a necklace" [2.21 a 4] and 

§2.21d L] 301 

extending to “Whose war elephants are his mighty arms” (2.21¢ A], it 
is of course obvious that that [second] sense is denoted, [not suggested]. 
But where there is some reason such as the context, which delimits the 
power of denotation to only one of the two senses, so that the power 
of denotation cannot extend to the second sense: in such a case the 
second sense is said to be suggested. Now even in such a case, if a 
word is then used [in the poem], by which that restrictive factor, such 
as the context, etc., loses its force, then that power of denotation, 
although once inhibited from rendering the second sense, is revived so 
to speak and such cases do not fall in the domain of suggestion. The 
particle ca [the second word in the sentence of the Vrtti which has 
just been explained] has the sense of api and has been placed out of 
position; [it belongs after àksipta]. Thus what is referred to is a figure 
of speech which, although (api) implied. that is, it begins by giving 
the immediate impression of being implied, is not really implied, but 
rather is directly expressed because of the revival of the [second] power 
of denotation by some other word. The word again (punah) indicates 
this revival of the power of denotation as explained above. Thus the 
word "only" (eva) (in Karika 2.21] rules out a figure of speech that is 
apparently, but not really, suggested. This is the meaning. 

“O Kesava": here the words have their powers of denotation limited 
by the context in the following sense. O Kesava, as my eyesight was 
blinded by the dust raised by [the feet of] the cattle, I could see nothing 
and because of that I stumbled on the way. Why now, that is, for 

* what possible reason, do you not give me. who have fallen, your hand 
for support? For you alone, being of extraordinary strength, are the 
refuge, i.e., the means of support in rough places, for all those who are 
weak: children, the aged, and women, whose hearts are distressed and 
who are unable to proceed. 

But while the context inhibits the denotative powers of the words 
from the second sense, which will be explained presently, these deno- 
tative powers are revived by the word salesam “with a hint." Lesa 
means something small and. as to hint at something is to make a small 
[reference] to it, salesam means "with a hint." The meaning hinted at 
is this: O Kesava, O cowherd, O lord of my life! Because my eyes were 
blinded by passion—or the words may be construed as follows: ~Be- 
cause | was deprived of my judgment by my passion for Kesava"'—1 
stumbled, that is, I became guilty of a moral lapse. Why do you not 
assume patità, that is, the office of a husband, toward me? You are the 
one, that is, you alone possess perfect success in love, for all women, 

302 [$2214 L 

their hearts yearning with desire but without any stain of jealousy, 
worship you as their refuge, that is, for the preservation of their lives. 
This is the second sense. 

1. In the prior interpretation gopa is taken as a separate word in the 
vocative. In the latter, and far preferable interpretation, gopa is taken as the 
stem-form in compound with raga. 

A On the other hand, where by the power of words a figure of 
speech in addition [to lesa] appears, this figure being implied by the 
inherent capability of the situation: all such cases are in the domain of 
dhvani. As [in a prose passage]: 

Meanwhile the long period named Summer, 
Meanwhile the God of Destruction, 

when the market stalls are white with the laughter 
whose terrible laughter is white 

of their blossoming jasmine flowers, 

as jasmine flowers, 

expanded as it put an end to the two months of Spring." 
yawned as He put an end to the aeons of time. 

(Bana. Hargacanta 2, lines 19-20] 
And as [in a verse]: 

They are high. with flashing necklace 

They are high. with Rashing downpours 

and dark with aloe paste: 

and dark as aloe paste: — 

whom would the breasts of this slender maid 

whom would tbis wealth of clouds 

not fill with yearning? " 

$221e A] 

iving joy to all creatures 
to their progeny 
by their absorption and release of water, 
of milk, 
scattering to all directions in the morning 
and disappearing at the close of day: 
and gathering together at tbe close of day: 
they are a ship for crossing 
the sea of transmigration, the source of our long pain. 
May these rays of the blazing sun engender 
May these cows 
in your purified selves unmeasured bliss. 
[Mayüra, Süryasatoka 9)? 

In these examples. by the power of words a second, non-contextual 
(aprákaranika), meaning appears. In order that the sentence should 
not convey a [second] meaning that is unconnected [with the first], 
one imagines a relation of image and subject (upamdna and upameya) 
between the non-contextual meaning [e.g., cows] and the contextual 
meaning (e.g., the rays of the sun], this imagining being made possible 
by the inherent capability of the situation (sámarthyát). And so the 
Slesa here is implied by the sense and not furnished by words.* Thus 
the domain of that type of dhvani which is like a reverberation is indeed 
different from that of the [figure of speech] slesa. 

1. The figure is rüpako. We have translated according to Abhinava's 
interpretation. On the other hand, the natura] way of taking pAullamallikà- 
dhavoláttaháso is as a rüpokasamása: "whose white laughter was the blos- 
soming jasmine.” But taking that compound as a rüpakasamása would make 
the passage unfit as an example of a sugggested figure of speech. Hence Abhi- 
nava's interpretation. Note that the way in which such passages are explained 
by the Alankdrikas may be defended logically but does not satisfy the psy- 
chological process of our apprehension. Meeting with the passage from Bana, 
atréntare dhavaldttahdso mahákàlah, etc., the reader immediateely sees 
what Ananda regards as the suggested sense: "Then Siva with his terrible 
laughter,” ete. Only later and painfully does he absorb the other meaning 
To Ananda the direct meaning (sáksádvácya) is the contextual meaning. Af- 
ter all, the story is describing the shift of spring to summer. There is no 
reason of syntax, no word marking a figure of speech (like iva for upamà, tu 
for vyatireka. eva for rüpaka), that would make us choose the meaning that 
refers to Siva the destroyer. So that meaning, which the reader has absorbed 

304 {§2.21e A 

so readily, is to Ananda the suggested meaning. But now for the psychological 
difficulty. This suggested meaning is defined as samlaksyakramavyarigya, that 
is, a meaning which is apprehended at a moment recognizably later than our 
apprehension of the denoted meaning. In instances like that of the quotation 
under discussion, that simply is not true. — 2. Here the suggested figure is 
simile. The girl's breasts are like clouds. The stanza is quoted by Subhasitávali 
(1538), which quotes some twelve verses by this same poet. 3. The text of 
the stanza is uncertain, with the variants aklistasrstaih appearing in a (Dav. 
ed. Badari Nath Sarma) and pavands táh appearing in d (KávyasamgraAa, 
Vol. 2). For the image of the salvific ship, one may note that the soul of the 
dying man passes by way of the rays of the sun to release (Jā Up.) and that 
the cow, being sacred, is sometimes used as a psychopomp, the tail of a cow 
being placed in the hand of a dying man to lead him to heaven. 4. The 
flesa in the preceding three examples is furnished by the capability, inherent 
in the two senses of the stanza, of entering into a relation of upamdna and 
upameya. The two senses have an inherent similarity. If, for example, the two 
senses of Mayüra's stanza did not have this capability, we would not thiak of 
taking the words prajéndm, payobhih, and gàvah in two senses. On the other 
hand, in a verse like “As even without a necklace" (2.21a A), the slesa in 
hárinov is pointed out by the word api. 

L Having thus distinguished the domain of the figure of speech 
lesa, he now explains the domain of suggestion(dhvani): on the other 
hand, where, etc. 

(Comment on the first example from Bana.) The contextual meaning 
is as follows. The summer puts an end to the two months that con- 
stitute the season of spring; [the summer,] in which the laughter, that 
is, the blossoming, the whiteness, of the full-blown jasmine flowers is 
such as to whiten, to make beautiful, the atfan:, that is, the market 
stalls. If [phullamallikédhavaldttahdse is] explained as [a rüpakasamása, 
viz.,] "[summer] which possesses Siva's white laughter in the form of 
full-blown jasmine flowers," this example would be in no way differ- 
ent from "cloud serpents” (2.21 b A; it would be a case of a directly 
expressed metaphor]. It is a "long time" (mahákálaA), that is, a long 
season, because its days are-long and hard to endure. Here the de- 
notative powers of the words are restricted by the context, namely a 
description of the summer season. For that reason mahdkdla [which as 
a compound means the destructive form of Siva] and the other [word 
attahása, which as a compound means the wild laughter of Siva] do not 
follow the maxim that "the denotative power of a compound is stronger 

$221eL] 305 

than the denotative power of its components," but fulfill their deno- 
tative functions [by furnisbing the sense of their component members| 
in the way we have described.! The apprehension of the sense [of the 
compound] which takes place afterward is the result of the power of 
suggestion based on the (denotative] power of words. 

On this problem, some people hold the following view. Inasmuch as 
these words (mahakdla, etc.,] have been seen in former contexts to have 
a different power [from that of mahàn kālah, etc.], giving a different 
sense [from that of “a long season"], it is from that other power which 
has been seen to give that other sense that the hearer can now appre- 
hend that sense (rom these words even when their denotative power is 
restricted by context, this apprehension being due to the operation of 
suggestion. Accordingly, there is no contradiction in saying that this 
sense is a suggested sense based on the [formerly experienced] denota- 
tive power of the word.? 

Others say that since the second denotative power [e.g., the power 
in moAàkàla that furnishes the meaning of "Siva"] relies for aid on the 
inherent capability of the situation, namely the similar properties of 
summer to those of the terrible god, it is therefore said to take the 
form of a suggestive operation.? 

Some follow the view that just as sabdaslesa is possible only where 
two separate words are present, so also in arthaslesa there must be 
two words because there are two denotations.* Accordingly, in [both| 
these cases a second word is bought in. Sometimes this is done by a 
denotative operation, for example where the answer sveto dhavati is 
given to two separate questions,* or in riddles and the like. In these 
cases the figure [slesa| is a denoted figure. But where the second word 
is brought in by a suggestive operation, it is reasonable to regard the 
meaning understood from the second word as a suggested (pratiya- 
māna) meaning because it is based on that which is suggested. 

Others say that inasmuch as it is a second denotative power that 
is revived according to the explanation of the second view [put forth 
above], the second meaning must be denoted and not suggested. But 
there appears an identification of this second meaning, after it has 
been apprehended, with the first, contextual, meaning; and as this 
identification cannot come from a non-linguistic source, it must come 
from the suggestive power of the words; because one cannot suppose 
that any denotative power is responsible for it.’ And this (suggested 
identification] is based on the second denotative power, for without 

306 [$2.21e L 

that power it would not arise.* Accordingly, it is reasonable to speak 
of this as the suggestion of a figure of speech (alarikaradhvani). And 
the Vrttikàra is about to say, "in order that the sentence should not 
convey a [second] meaning that is unconnected [with the first].” Now in 
the previous (examples, where the figure was not suggested,| the lack of 
connection [in the verse “O Kesava, my eyes were blind"] was prevented 
by the word salesam “with a bint”; in the verse "He who destroyed the 
cart" a lack of connection simply did not appear;? in tbe verse "As 
even without a necklace" the lack of connection was prevented by the 
word api "even"; in the verse "As he holds in his band the discus," by 
the word adhike "superior"; in the verse "The cloud serpents,” by the 
metaphor-compound [the rüpakasamása "cloud-serpents"]. (They say 
that] this is the overall meaning.'? 

(Commentary on the verse of Mayüra.] The word payobhih means 
both "with water" and “with milk” Samhāra means "disappearance" 
and “gathering into one place." Gavo means "rays" and "cows." 

Conveying a meaning that is unconnected: that is, a meaning 
that is unintelligible. A relation of image and subject: By this rela- 
tion, which is in the form [that the figurative operation takes| in simile, 
one should judge that contrasting [one thing with another], denying 
{one thing in favor of another], etc., in fact, any form of [suggested] 
operation, furnishes us with the chief goal of aesthetic delight, ratber 
thao the base and simile, etc., [on which these operations work].!! This 
consideration applies to all suggested figures oí speech. 

By the inherent capability of the situation: that is, by the 
suggestive operation. 

1. Other things being equal, one would naturally take mahdkdlah to mean 
Siva. But the present context forces us to take the denoted sense as that of 
moAán káloh, a long season. If we also apprehend the sense of Siva, that 
sense must be suggested, for the denotative operation has been completed. 
2. This is the view accepted by Mammata 2.19 (Jhalkikar ed. p. 63) and by 
Viévanàtha (SD 2.14). — 3. This interpretation stays closer to the language 
of Ananda. The hearer already knows that mohókála can mean Siva. But 
hearing the word in a context where it must denote a long season, he would 
not think of the denotation Siva unless there was some similarity in the new 
context that suggested that other sense. The fact that summer puts an end 
to a period of time, namely the spring, and Siva puts an end to a period 
of time, namely the aeon, constitutes the similiarity that allows the sugges- 
tion to operate. — 4. This is the view of Udbhata as is also the peculiar 
distinction of Sabdaslesa and arthaslesa. Udbhata defines slesa, by which he 

$221eL] 307 

means a sticking together, an adhesion, of two meanings or of two similar 
sound groups, in 4.9-10 Indurdja (4.23-24 Vivrti): ekaprcyatnoccáryánám 
toccháyám cawa bibhratám / svaritádigunair bhinnair bandhaA sligtam iho- 
cyate // alarikdréntaragatém protibhám janayat padaih / dvividhair artha- 
sabdoktwisistam tat pratiyatam // "The use of expressions tbat bave identical 
phonetic shape, or of expressions that seem to have identical shape because 
their differences are in properties such as the Vedic accent, is called slista 
(= Slega). This figure produces the appearance of other figures by means 
of word-pairs of these two types and is to be understood as characterized 
by a [conjoined] expression of meanin or of sound-groups [i.e.. as artha- 
Slesa ot Sabdaslesa.” The commentators explain that in Udbhata's view there 
must be a separate word (padam) for every meaning. Thus, if kam is used 
to mean both “hand” and “ray,” we have two words of identical phonetic 
shape. This usage he calls arthaslesa, adhesion of meaning, for two mean- 
ings adhere together in one phonetic datum. On the other hand. when the 
expression asvápaphala is used to mean both "producing a result that is not 
easily obtained" (a-su-dpam phalam yasya) and "the result of lack of sleep” 
(a-svápa-phalam), we do not have two words of identical phonetic sbape. Ac- 
cording to Pán. 6.2.172 the former will be accented asvdpaphald (a negative 
bahuvrihi has terminal accent), while by Pan. 6.2.139 the latter will be ac- 
cented asvāpaphála (in a tatpurusa the final member takes its natural accent). 
This usage Udbbata calls fabdaslesa, adhesion of sound, for two sound-units 
adhere so as to appear identical. Udbhata's two types are almost equivalent to 
Dandin's abhinnapada and bhinnapada. Udbhata's terminology is infelicitous 
because both types depend on sound and this terminology was abandoned by 
the later Alaàkárikas whose works are preserved to us. But the commentator 
from whom Abhinava is here quoting apparently accepted it. — 5. The two 
questions are ka ito dhávati "what is running hither?” and kimvarno dhé- 
vati “What color is the thing that is running?” The answer to the first is 
fveto ($và ito) dhávati “A dog is running hither”; to the latter sveto dhdvatt 
"Something white is running." The second word (sveto "white") is brought in 
by denotative operation. The question demands that some color be denoted 
The example goes back to Patafjali, Mahábhásya, Intr. to 1.1.1 (Kielhorn 
Vol. 1, p. 14) and 8.23 (Kielhorn Vol. 3, p. 388). — 6. Context demands 
that mahdkdlah apply to summer. So the "first word” denotes a long season. 
But we have previously heard a "second word" of identical phonetic shape, 
which means Siva. If we now understand the meaning Siva also, it is because 
the second word has been brought in by suggestion. Vācya and pratiyamana 
are used, like srauta and drtha, of that which is expressed and that which is 
understood (suggested). 7. Because the denotative powers are used up in 
furnishing the frst and second meanings. 8. The suggested identification is 
not furnished (utpádita) by a denotative power because the denotative powers 
are now exhausted. But it is based on, that is, it presupposes, a denotative 

308 [§221eL 

power. Specifically it presupposes the second denotative power because one 
cannot have the relation of identity without having a second term to identify 
with the first. 9. Both meanings of the verse furnish praise of God and so 
both are contextual. There is no non-contextual meaning in the stanza which 
could exhibit a disconnection. 

10. According to BP this view (the fourth view of "others" given by 
Abbinava) is accepted by Appayya Diksita in his Kuvoloyánanda. — 11. In 
characterizing suggested figures of speech Ananda spoke merely of our being 
forced to imagine a similarity. Abhinava extends his statement, taking it as 
an upalaksona of other relations which we may be forced to imagine. He 
then points out that this imagining of various relations is what gives the chief 
aesthetic value to this species of dhvani. In denoted figures of speech our 
pleasure ends with the apprehension of the objects which are brought into 
relation by the figure, e.g., the subj t (upameya) and the image (upamóna). 
In suggested Ggures of speech we receive a special relish from imagining the 

A Other figures of speech as well [as simile) can occur in this 
type of suggestive poetry that is based on the power of words and 
where the suggested meaning is like the reverberation of a bell. Thus, 
contradiction (virodha) may appear in the form of a reverberation based 
on the power of words, as in Bhatt& Bána's description of the land called 

where the women have the [slow] gait of elephants 
have affairs with outcastes, 
and are virtuous, 
are of fair complexion and fond of wealth, 
are Gauris and are fond of places where Siva is absent, 
are youthful and wear rubies, 
are black and have the red color of lotuses, 
have mouths that are bright with white teeth 
have mouths as pure as those of pure brahming 
and breaths that are perfumed with wine. 
[Bána.Harsacarita, p. 98, lines 3-4] 
(Chapter 3, lines 228-229 out of 654) 

§2.21f A] 309 

For in this example one cannot say that the contradiction is directly 
expressed or that the puns favoring the semblance of contradiction are 
denoted; because the figure contradiction is not directly revealed by 
any word [such as "although"]. For where the figure of contradiction is 
directly expressed by a word, in such a poetic expression of ambiguity 
we have the domain of a denoted figure of speech, either contradiction 
(virodha) or pun (slesa). 
An example may be given from the same work of Bàna: 

She was the meeting place as it were of contradictions, 
for her figure was brilliant 
for there was the figure of the sun 
although accompanied by the blackness of her hair. 
even in the presence of the youog night. 
(Bana, Harsacarito, p. 27, line 15] 
(Chapter 1, line 403 out of 689) 

Or, [one may see an example of suggested contradiction] in a verse 
of my own own: 

Bow down to the sole refuge of men, the everlasting, 
to tbe soie house of men that is no house, 

tbe overlord, the lord of our thoughts, 

the non-lord of thoughts that is lord of our thoughts, 

Hari-Krishna, fourfold of nature,’ beyond all action, 

the golden, the black, of dexterous self who does not act, 

the destroyer of enemies who bears the wheel. 
the destroyer of the spoke-bolder who bears the wheel. 

For in this verse contradiction in the form of a [suggested] reverberation 
based on the power of words is clearly understood. 

Contrast (vyatireka) of the same sort also appears i 
of my own: 

May both sets of the sun god's feet lead you to welfare 
[those which are his rays and those on which he stands]: 
those which light up the sky, dispelling darkness, 

and those whose toenails are refulgent; 

and those which do oot illumiae the sky; 

those which nourish the beauty of the pond lotus 

and those whose beauty puts the lotuses to shame; 
those which shine on the tops of mountains 

those which shine on the heads of kings 

and those which tread on the heads of the immortals. 

310 [§ 2.21f A 

In the same way there are other varieties of dhvani based on the power 
of words where the suggested meaning is like the reverberation of a 
bell. They may be sought out by sensitive readers on their own. I have 
not dealt with them here in detail lest my book should become too 

A reference to the four emanations (vyühas) of Visnu. 

L (Commentary on the first quotation from Bana.| (Mátariga- 
güminyah| means they walk like elephants. The contrast (with what 
follows] lies in the [second] meaning "they visit outcastes." "They 
delight in wealth" also means they take delight in a place where Siva 
is absent. "They possess the gems called padmarága (rubies)" also 
(means) they have the red color (raga) of lotuses (padma). "Their 
mouths are pure," that is, bright "with white teeth (dvija)" also [means] 
their mouths are as pure as those of pure, that is, the most exalted, 
brahmins (dvija). 

For where: namely in a poetic expression of ambiguity [where the 
contradiction is directly expressed], there we have the domain of con- 
tradiction (virodha) or pun (slesa), that is to say, of the figure fusion 
(samkara).! He means that this (figure in such instances) becomes the 
domain—of what?—of a denoted figure of speech, of a denoted 
ornament. The meaning is, it becomes the domain of something which 
possesses the property of a denoted figure of speech.? This is as much 
as to say that it is only in such cases that one may rightly call the 
contradiction or the pun a denoted figure. 

(Commentary on the second quotation from Bana.] Night, that is, 
blackness, was in her hair (vàála), or the night, that is, darkness, was 
young (bàla), new. 

Now it might be objected that in the passage "Where the women 
have the gait of elephants," etc., the particle ca ("and") being used 
with the pairs of properties actually expresses the contradiction. For 
if a mere additive sense had been intended, ca would have been used 
with each property separately, or would have been used just once at 
the end, or would not have been used at all.? With this objection in 
mind, he furnishes another example: or. 

How can a Sarana, a house, be in the form of a-ksaya, a non-house? 
How can he who is not dhisa, lord of our thoughts, be lord of our 
thoughts? How can he who is golden (hari) be black (krsna)? How 

§ 2.22 A] 311 

can one whose self is dexterous, valiant, be actionless? How can he 
who is the destroyer of that which possesses spokes proudly bear the 

Contradiction: the word virodha here means contradiction in gen- 
eral [not the expressed figure of speech virodha|.* Is understood: what 
he has in mind is that it is clearly understood but is not expressed by 
any word. 

[Commentary on the final verse.] "Those whose toenails are resplen- 
dent” also means "which certainly do not shine in the sky." "Both" 
[sets of feet] means those which are his rays aod those which are limbs 
composed of toes, heel, etc. 

1. Le., such an instance falls into the category of the figure fusion. See 
above, 2.21a A, note l and 2.21a L, note 4. — 2. le., such a case of the 
figure fusion falls into the category of a denoted figure of speech. — 3. The 
same effect appears in English. If I say, "He is wise and young, handsome and 
not proud,” I am emphasizing the contradictions as I would not do if I used 
the word "and" three times, or just once (before "not proud"), or not at all. 
4. The reason for Abhinava's gloss is that in his view the suggested figure is 
sañkara, not virodha. 

K On the other hand, we have another type of [dAvani] that 
arises from the power of meaning (arthasaktyudbhava) when a meaning 
appears which by itself and without [the use of| words, manifests a 
second meaning as the tátparya (the chief meaning of the sentence). 

A Where a meaning by its own inherent capability manifests, 
without the operation of words, another meaning, we have that variety 
of dhvani, arising from the power of meaning, where the suggested 
meaning is similar to a reverberation. For example: 

While the heavenly visitor was speaking, Parvati, 
standing with lowered face beside her father, 
counted the petals of the lotus in her hand 
(Kalidasa, KumSam. 6.84] 

312 {§ 2.22 A 

For here the counting of the petals of the lotus subordinates itself and 
without the help of any verbal operation reveals another matter in the 
form of a transient state of mind (vyabhicdribhava) (of the emotion love, 
namely shyness]. This example does not fall under the sole heading of 
“suggestion without a perceived interval” (alaksyakramavyarigya), for 
it is only where we apprehend the rasa or the like directly (sáksát, i.e., 
immediately) through a verbal presentation of the vibhávas, anubhavas, 
and vyabhicáribhávas that we have a suggestion that is exclusively of 
that type,! as for example in the Kumárasambhava where, in the con- 
text of the advent of spring, we have the description of events beginning 
with the arrival of Parvati wearing spring flowers for jewelry, up to the 
point where the god of love places an arrow on his bow and takes aim 
at Siva. AU this, as well as the description of the particular actions 
of Siva as his calmness is stitred, is conveyed directly by words. But 
in the present example the rasa is apprehended through its transient 
state of mind, which in turn is implied by the inherent capability [of 
the described action of counting the lotus petals, etc.]. Therefore this 
is a different variety of dhvani. 

1, Note that Ananda will admit, in 3.43 A, that this stanza does contain 
rasadhvani, for the suggestion of the shyness leads on to an apprehension of 
the rasa, órrigóro. In the stanza the two varieties of dhvani, be says, are 
fused. What he says here is that the stanza does not contain rosadAvant 
only (alaksyokromavyarigyadhvani). [t also contains a suggestion of perceived 
interval, namely tbe suggestion of shyness. 

The phrase sáksácchabdanivedita used here and three lines below has long 
caused difficulty. Sridhara in commenting on the Kávyaprokása (Vol. 1, p. 128) 
attributed the phrase to Ananda’s temporary forgetfulness or inattention. For 
modern discussions of the passage, in addition to Jacobi's note, ZDMG 56 
(1902), p. 766, see K. Krishnamoorthy, The Dhvanydloka and its Critics, 
P. 266, and M. V. Patwardhan and J. L. Masson, "Solution to a Long-confused 
Issue in the Dhvanyáloka,^ JO/Baroda 22 (1972-73), pp. 48-56, to which 
Krishnamoorthy has replied in a long note to his text and translation of the 
Dhv., pp. 354-360. 

The diffculty is this. Ananda has already said (1.4g A) that rusa and the 
like (rasddi) are never sáksácchabdávyapáravisaya, that is, never the object of 
the direct (denotative) operation of words; they are always suggested. And 
within the term "and the like" the bhávos and vyabhicüribhóvas are included. 
If we take sáksácchabdanivedita in the present sentence to mean “where the 
vyabhicárins are conveyed by direct denotation, or actually named," we will 
have to charge Ananda with flatly contradicting one of his most basic theories. 
Furthermore, in the passages of the Kumdrasambhava to which he refers as 

§ 2.22 L] 313 

exemplifications of alaksyakrama, the vyabhicdribhdvas are nowhere directly 

Abhinava's solution, followed by Patwardhan and Masson in the article 
just referred to, is to connect the sense of the adverb sáksát more closely with 
pratiyate than with Sabdaniveditebhyo. It is when we apprehend the suggestion 
directly, that is, immediately, from the description of the vibhávas, anubhávas 
and vyabhicárins that we have the type of suggestion where the interval is not 
perceived (asamlaksyakramavyarigyadhvani). In Ananda's judgment the verse 
which describes Párvati's counting the petals of the lotus does not belong in 
that category. He feels that we must reflect for a moment before we realize 
that what is being described is really the shyness of young love. So he assigns 
the verse to the samlaksyakrama type. 

Abhinava's solution becomes somewhat more difficult in the passage three 
lines below, where sáksácchabdanivedita is used without the verb protiyate, 
the sense of which must be understood. That is, we must understand "con- 
veyed by direct words" to mean "conveyed by words which give an immediate 
understanding of the suggestion." Difficult or not, this interpretation fts with 
the whole tenor of Ánanda's aesthetics. We can now charge him perhaps with 
inexactitude of expression (Sridhara's inattention) but not with a lapse of 
theory (Sridhara's forgetfulness). 

L Having in this way dealt with that variety ofdhvani where the 
suggestion. rests on the power of words, he now proceeds to explain 
the variety that depends on the power of meaning: [On the other 
hand], etc. Another: sc., other than that which rests on the power 
of words. By itself (manifests) as the tátparya: This word (tét- 
parya), being followed as it is by a denial of the denotative function (in 
the words uktim vind], must refer to the suggestive process, not to the 
tátparyasakti (the power belonging to the syntax of the sentence), for 
we have already said [1.4b L] that the power belonging to the syntax 
of the sentence has exhausted itself once we have understood the literal 

With this same intention he now says in the Vrtti: Where a mean- 
ing by its own inherent capability. The term "by itself" (svatah) 
of the Kárikà has been rendered by "its own" (sva) in the Vrtti. He 
now explains the significance of "without words": without the oper- 
ation of words. He illustrates: For example "While [the heavenly 
visitor], etc." 

Another matter: shyness. Directly: the meaning here intended 
is that the vyabhicürins are “directly conveyed" where our perception 

314 {§ 2.22 L 

of them arises from their appropriate vibhávas and anubhdves with- 
out anything's being interposed, inasmuch as we do not perceive any 
interval [between the literal and suggested meanings]. Hence there is 
no contradiction with what was stated before [in 1.4g A]. For it was 
stated before at some length that the vyabhicáribhávas, being emotional 
manifestations (bhávas), cannot be conveyed by being directly named 
(svafabdatah). This is as much as to say that although rasa, bhàva, 
and the like are invariably suggested and never directly expressed, still 
they are not always in the domain of that type of suggestion where the 
interval [between the literal and suggested meanings] is not perceived. 
We have that type of suggestion only in those cases where the sugges- 
tion of rasa and the like occurs immediately from the full presentation 
of the vibhavas and anubhdvas belonging to the sthdyibhavas and the 
vyabhicáribhàvas. For example: 

Then came the daughter of the mountain king 
with her attendant nymphs. Her loveliness 
seemed to rekindle in the god of love 
his near extinguished courage 
[KumSam. 3.52] 

In this and the following stanzas there is a full description of [Parvati's] 
nature which is well suited to act both as an objective and a stimulative 
determinant (dlambanavibhdva and uddipanavibhava.)! 

We are told how these vibhévas take effect in the stanza: 

As Siva, favoring his devotee, 
moved to accept the offering, 
the love-god fitted to his fowery bow 
the deadly shaft Infatuation. 
[KumSam. 3.66] 

But Siva. stirring slightly from his cal 
as the ocean stirs at moonrise, 
busied his eyes on Umi's face 
and her fruit-like lower lip 
(KumSam. 3.67] 

First we have been told of Parvati’s inclination toward Siva and now 
we are told of Siva's turning his attention toward Parvati. His prej- 
udice in her favor has been indicated as a “favoring of his devotee.” 

§ 2.22 L) 315 

From a strengthening of this prejudice comes the basic emotion (sthayi- 
bhava) love (rati). The author has revealed all the symptoms (anu- 
bhdvas) common to this basic emotion and to its transient states of 
mind (vyabhicárins) eagerness, agitation (dvega), instability (capalya), 
and the like. And so our relish of the determinants (vibhávas) and 
the symptoms ends up as (paryavasyati) a relish of the transient states 
of mind. And as the transient states of mind are dependent [on the 
basic emotion], there is no perceived interval between our relishing of 
them and our aesthetic goal of relishing the basic emotion, which is like 
the string of a garland [of which the transient states are the separate 

But in the stanza which our author is discussing, a young maiden's 
counting of lotus petals and lowering her face can be imagined as due 
to other causes [such as ipattention or naivete] and so do not cause 
our heart to fix immediately on shyness (as her state of mind]. Rather, 
these acts suggest the idea of shyness in love only after the reader calls 
to mind the earlier incidents [of the poem], such as Párvati's asceticism 
(undertaken in order to win Siva as her husband]. So the suggestion of 
[the accompanying emotion) shyness comes after a perceived interval. 

The rasa in the stanza is also removed, but as it appears the moment 
that the true nature of [Párvati's| state of mind bas been realized, it 
is not at a perceived interval from that. The suggestion of perceived 
interval in this stanza is with respect to the shyness. This is what our 
author has meant to indicate by bis use of the words "sole" (eva) and 
“exclusive” (kevala).? 

1. Parvati herself acts as the àlambanavibhàva; her physical qualities act 
as the uddipanavibhóvas. 2. There is no strict or exclusive asamlaksya- 
kromadhvani in the lotus petal stanza because while the rasa is asamloksya- 
krama from the vyabhicárin, the vyabhicárin is samlaksyakroma from the 
presentation of the anubhdva, To put the matter in simpler form: there is 
a perceived interval between the presentation of the anubhdva (counting the 
letus petals) and the suggestion of the vyabhicárin (shyness); there is no 
perceived interval between the suggestion of shyness and the suggestion. of 
the rosa of love. This is a subtle distinction, which permits Abhinava to say 
that the rosa itself is always suggested without interval even when it "stands 
far off” (düratoh) by reason of its transient emotion’s being suggested at an 

[$2222 A 

A But a meaning which is aided by the (denotative] operation 
of words in suggesting a second meaning does not fall in this category 
[of suggestion by the power of meaning]. For example: 

Knowing that her gallant had set his heart 

on a rendezvous, the subtle lass 

smiled and to show her meaning folded 

the petals of the lotus in her hand.' 
Here we are expressly told of the suggestiveness of the girl's folding up 
the lotus blossom with which she is playing. 

1. As lotuses close their petals at sundown, she means that he is to meet 
her at that time. Ananda quotes the stanza again at 3.34 A as an example 
of alaksyakramagunibhutaryarigya. The suggestion is the fact that her lover 
should come at night. We arrive at the suggested meaning iramediately. The 
stanza is also found annonymously in Subh À. 2043. Dandin has a very similar 
stanza, KA 2.261. 

L He proceeds to show wbat is sought to be excluded by the 
Káriká's phrase “without (the use of] words": But a meaning, etc. 
The particle ca ("and") bere has the sense of "but." In this [cate- 
gory}: he bas in mind that it may, however, fall in [a different category, 
namely] that of suggestion where the interval is not noticed.! He illus- 
trates: "Knowing that her gallant, etc.” 

The suggestiveness: sc., that evening will be the right time [for the 
rendezvous}. Expressly told: sc., by the first three lines. It is true that 
no one word in these three lines, even in connection with its neighbors, 
has the power of denoting the sense of “evening,” and to that extent the 
suggestiveness of the stanza is not undone. However, we are expressly 
told that the sense is suggestive of some other sense and thereby the 
very life of suggestion, which consists of the charm of something's being 
said in a hidden manner, is destroyed. It is as if some one should say: 

Tam a deep man. 

No one knows what I will do. 

I do not say anything 

that my facial expression has i icated 

$2.23 A] 317 

Here the speaker, who should indicate his depth of character by his 
facial expression, instead speaks of it outright. Hence our author says, 
"we have been told of the suggestiveness."? 

1. Compare 2.21a A, end; also 3.38 A. 2. I am not sure that I have 
understood the verse fragment. Jagannath Pathak translates it into Hindi: 
maim gambhir hüm, bind batde marà kám koi bhi nahim jāntā, (is lie) kuch 
kahtá him, "| am a deep man. No one knows of my doings unless I give a 
hint, (therefore) I say something.” BP punctuates so as to get a very different 
sense: “I am not a deep man. Therefore no one knows my intention although 
it is indicated (by my expression]. So I will say something.” Pathak may be 
right. BP, [ am sure. is wrong 

K When a meaning, even though it has been implied by the 
power of words or meanings, is then again revealed by the poet in so 
many words, it is [to be considered as] a figure of speech and different 
from dhvani. 

A When a meaning, even though it has been implied by the 
power of words, by the power of meaning, or by the power of words 
and meaning, is then again proclaimed in so many words by the poet, 
it is [to be considered as] a figure of speech and different from this type 
of dhvani that is like a reverberation. Or, if there is a possibility in 
the verse of a suggestion of unperceived interval, this (implied and then 
revealed meaning] will be a figure of speech and different from that sort 
of suggestion.! 

318 [$2.23 A 

An example of a meaning implied by the power of words (and then 
revealed] is: 

“My child, come hither and be not distressed; 
come to this one (Visnu| and go not to the poison-eater [Siva]; 
leave off this rapid upward surging breath; 
abandon the rapid Wind-god and bim of vertical motion [Fire]; 
and why this heavy trembling? Have done with 
what use is tbe god of Water or Brahm’? Have done with 
this exhausting stretching of your limbs." 
the proud destroyer of Bala [[ndra]." 
Thus Ocean, under guise of calming Laksmi's fear, 
for she was dazed by the churning of the sea, 
caused her to reject the other gods and gave her 
to him who now, I pray, may burn away your sins.? 

An example of a meaning implied by the power of meaning (and then 
revealed] is: 

"That's where my aged mother sleeps, and there 
sleeps daddy, the oldest man you've ever met. 

Here sleeps the slave-girl worn out by her chores, 
and here sleep I, who rnust be guilty 

to deserve these few days absence of my lord." 

By these statements the youthful wife suggested 
to the the traveler his opportunity.* 


An example of a meaning implied by both powers is “O Kesava, my 
eyes were blind" [2.21d A]. 

1. The genitive aloksyokromavyafigyasya goes naturally with sambhave 
and tádrs refers most naturally to oloksyakramovyarigya. There is no need for 
the grammatical gymnastics employed by Abhinava (see below). — 2. Vertical 
motion is a basic characteristic of fire in the Vaisesika system ( VS 5.2.13) and 
is used as an epithet of fire in poetry (e.g. Sisupálavadha 1.2). — 3. The 
stanza is quoted anonymously in SüktiM. 3.65 and under Kuv. 155. Laksmi 
was among the precious objects churned up by the gods from the sea. The 
motion might well have left her out of breath and trembling. For jrmbhita, 
context as well as the opinion of Abhinava favors the meaning of stretching the 
limbs (arigasammardana) rather than yawning; see also 2.18-19b A, note 1. 
The ocean is pictured as Laksmi's father, comforting her and by puns directing 
her away from all suitors but her future husband, Visnu. While the sequential 
suggestion of second meanings is destroyed by our being told of it in so many 
words, there is the possibility of the other type of suggestion, with no perceived 

§ 2.23 L] 319 

interval (asamlaksyakramavyarigyadhvani) in the verse. We may take the main 
purport of the stanza to be a relish (srrigdrarasdbhdsa) of the poet's love 
of Vispu ( Visnuvisayakaratibhàva) of which the stimulant (uddipanavibháva) 
would be Visnu's qualities which are here suggested to be superior to those 
of all the other gods by the fact that Laksmi chose him for her husband. 
4. The reading avasara in line d is found in ancient quotations of the stanza 
only here and in RG. With this reading a literal translation of the line will 
be: "Thus was the traveler addressed by the young woman with a hint of 
his opportunity in her statements.” Most ancient quotations (SRK 812, Sad- 
ukti. 2.15.3, Aufrecht ZDMG 36, 539) read abhimatam: the wife informed 
the traveler of her intention. SudhA. 2247 reads avasatha (resting place): 
the young wife spoke to the traveler under the pretext of stating the resting 
places (of the members of her family). The verse is variously ascribed in the 
anthologies to Rudrata or Bhatta. The two authors: Rudrata, author of the 
Kávyálarikáro, and Rudra Bhatta, author of the Srrigdratilaka are constantly 
confused in the anthologies. As Rudra Bhatta was apparently of later date 
than Anandavardbana (see Kane HSP, pp. 149-151), it is Rudrata that has 
the best claim to the verse. 

L The author of the Vrtti wishes in a single construction to 
resume the foregoing two types of suggestion and to indicate a third 
type. Accordingly, he introduces the Karika with a phrase that serves 
both purposes: and in the same way.! The meaning is that together 
with the aforementioned two types a third type must be considered. 

' [In the Karika] sabdártha is an ekasesa compound.’ Different: it is 
not a suggestion but an [expressed] figure of speech such as slesa. Or, 
taking "suggestion" (dhvani) [in the Kdrika| to mean suggestion with 
an unperceived interval, this suggested meaning can be considered an 
ornament of that suggestion to which it is subordinated, an ornament 
that is different from merely expressed figures of speech and one that 
forms a second, far superior (lokottara) type of figure of speech. He 
will explain the passage in these two ways in the Vrtti? 

[Comment on the first verse.) Visdda (distress) [also] means eater of 
poison. "Of vertical motion" refers to Fire. (To obtain a satisfactory 
pun} one müst understand the word "and" [as connecting svasanam 
and ürdhvapravritam]. Kampah (trembling) [also] means lord (pah) 
of waters (kam). Or [why] choose kah, that is, Brahma, your ancestor 
(guru)?* Have done with balabhid, that is, Indra, who is jrmbhita, drunk 
with the pride of his sovereignty; that is [one] meaning. Jrmbhita also 
means a stretching of the limbs, which is balabhid, that is. destructive 

320 (§ 2.23 L 

of one's strength because it causes exertion. By the word "rejection" 
(pratyókhyána) we are informed that a second set of meanings is de- 
noted by the words. “Having caused her": now Laksmi had arisen 
from the sea with a desire for Pundarikaksa (= Visnu) in her heart 
and so she naturally (would have] rejected the other gods; but because 
of the delicacy of her constitution she was in a state of shock from the 
crashing waves churned up by Mt. Mandara and was brought to do 
what was natural to her only by her father’s reminder in the form of 
his revealing the faults of the other gods and his saying, “Come hither 
(come to this one, [Visnu])” with a gesture indicating his respect for 
all the virtues [of her future spouse]. That is why the poet says “dazed 
by the churning." The structure of the sentence is this. May he burn 
away your sins, to whom Ocean gave Laksmi after causing, under the 
guise of stilling her fears in the manner described, her, who was dazed 
from the churning, to rejct the other gods. 

[That's where my aged] mother [sleeps]: the suggestiveness of 
the individual words in this stanza can easily be imagined by a sensitive 
reader, so we refrain from explaining them in so many words. The 
phrase “with a hint of” constitutes the direct expression which is the 
statement of the poet himself. 

Under the guise of summing up’ {what has gone before], our author 
has described and illustrated two varieties of suggestion. He now pro- 
ceeds to mention a third variety: implied by both powers. (The 
stanza "O KeSava, my eyes are blind" contains sequential suggestion 
by} the power of words because of the puns in gopardga |“the dust 
raised by your cattle” or “passion for the cowherd"], etc.; the power of 
meaning because of the context.* For so long as it is not known that 
Krishna is the object of the intense, secret love of all the young women 
[of Gokula], the second sense [i.e., the suggested sense] cannot be per- 
ceived. The word salesam ("with a hint") is the statement of the poet 

1. “In the same way” (tathd) refers to the foregoing two types “based on 
the power of words” and “based on the power of meaning.” “And” (ca) indi- 
cates that there exists a further, third type “based on the power of words and 
meaning.” 2. Pan. 1.2.64. Just as urksdh can be analyzed into vrkso-vrksa- 
vrksáh = vrksas ca vrksas ca urksas ca, so sabdárthah can be analyzed into 
abdaf ca arthas ca śabdārthaś ca. — 3. Ananda explains the Küriká phrase 
sdnyaivdlankrtir dhvaneh in two ways, but not in the two ways described by 
Abhinava. There is no need to take dhvaneh in the Kárikà as a genitive as well 
as an ablative. [t appears as a genitive in the second explanation of the Vrtti 

§ 2.24 A] 321 

simply because it is in construction with sambhave. Nor is there any reason to 
suppose that Ananda made a value distinction between the alarikara involved 
in the first interpretation and that involved in the second. 4. Abhinava 
here discovers still another pun, which we have not rendered in the transla- 
tion. The ioterrogative kasmo: in RV 10.121 "To whom shall we offer the 
oblation?” was understood as a name of Brahma. Brahma would be an im- 
possible choice as husband for Laksmi ae he was her grandfather. $. BP: 
Abhinava is taking the causative suffix nic in the sense of causing a person 
to do what he is naturally about to do, not in the sense of setting bim on à 
course which he has not begun on; because this interpretation is harmonious 
with the rasa of the verse. 6. The statement "proclaimed in so many words 
by the poet,” as A has put it, which gives away the suggestion. 7. The 
phrase upasarnháravydjena may mean little more than "by way of summing 
up.” But we bave taken it at full value. Ananda has actually done more than 
sum up what had been said before. His examples of sequential suggestion 
that fails or is spoiled by open revelation give us a fuller knowledge of this 
type of suggestion. 8. A knowledge of context, of course, is necessary in 
sabdasaktyudbhavadhvani also. But it is a particular kind of context that Ab- 
hinava has in mind, as bis next sentence shows. The knowledge of Krishna's 
character permits the arthasaktyudbhavadAvani here just as the knowledge of 
the gay housewife's character permits the same type of suggestion in "That's 
where my aged mother sleeps.” 

K A meaning that reveals a second fact [or situation, vastu) is 
also of two kinds. It may be given body simply by an imaginative 
expression (praudhokti),! or it may be inherently possible (svatah sam- 

1. For an explanation of proudhokti based on the etymology of the word, 
see Abbinava below. The term is used of a sophisticated, striking (camat- 
küránuguna) expression arising from the poet's imagination rather than from 
the data directly-presented by the everyday world. We translate it by "ima£- 
inative expression,” but the term always connotes boldness, vividness, fancy, 
and the irreality of what is expressed. 

A In the type of suggestive poetry that is like a reverberation, 
ithin the variety that is based on the power of meaning, the suggestive 

322 [$2.24 A 

meaning is itself of two types: the first, which is given body simply by 
an imaginative expression of the poet or of a character created by the 
poet; and the second, which is inberently possible. Àn example of the 
type which is given body simply by an imaginative expression of the 
poet is: 

The fragrant month prepares, 

but gives not yet for his use against young maids, 

the arrows, pointed with mango bud 

and feathered with new leaves, to the god of love." 

An example of the type which is given body by an imaginative ex- 
pression of a speaker created by the poet is the verse already quoted 
"On what mountain, for bow long" (see 1.13m A]. Or, 

Attentive youth 

has lent a hand to your breasts 

that they might rise as it were to greet 
the visitor love.? 

The type that is inherently possible occurs when the fact can be 
imagined as appropriate in the world of reality, a fact the substance of 
which is not produced only by a turn of phrase. An example is the verse 
we have quoted "While the heavenly visitor was speaking" [2.22 A]. Or, 

The hunter's wife strolls proudly 
with peacock feather behind her ear. 
She strolls amid fellow wives 
who are decked with pearls.? 
[Sattasas 2.73] 

1. The stanza describes the earliest days of spring before its full effect 
is seen. For similar descriptions see SRK 164 and 166. The syntax follows 
the metrical pattern. That is, there is a syntactic break after surahimdso; the 
words na ddva are to be construed with what follows. By failing to observe 
this, Jacobi's translation misses the point of the stanza. In line b the alternate 
reading lakkhasohe given by BP is better than lakkhamuhe., as it avoids the 
repetition of the word muhe. V. V. Mirashi in "Some Royal Poets of the 
Vàkàtaka Age,” IHQ 21, pp. 196ff, ascribes this verse to the Harivijaya of the 
royal poet Sarvasena, for whom see below 3.10-14e A. There is no compelling 
reason for the ascription, but the verse is written in the literary dialect and 
has the simple charm of other verses of that author. 2. The author is 
not known. Literally, "An abhyutthdna (rising to meet a guest) has been 
given to Love by your swelling breasts, an abhyutthdna that has the support, 
respectfully bestowed (vitirna), of the hand of Youth." 3. The readings of 

82.24 L] 323 

the Sattasai are of a purer Máharástri than Ananda’s: (a) sihipehundvaamsd 
bahud; (c) gaamottiaracia-. Abhinava will explain the suggestions of this little 
stanza below. 

L In this way the author has so far given a general definition of 
suggestion that arises from the power of meaning. He has also explained 
that its province is separate from that of figures of speech such as slesa 
and the like. He now proceeds to explain its subdivisions: It may be 
given body, etc. The meaning which [in the Karikd] is said to reveal, 
that is, to suggest, another meaning is also of two kinds. Not only 
is reverberatory suggestion [i.e., samlaksyakramadhvani] of two kinds 
[viz., based on the power of words and based on the power of meaning], 
but even its second variety is of two kinds through the dichotomy of 
the suggestive meaning. That is the force of the word also. 

[In the Vreti} he states that imaginative expression also has subvari- 
eties: of the poet [or of a character]. Hence there are three varieties 
of the suggestive meaning. Praudha is formed of pra in the sense of 
highly, extremely (prakarsena) and üdhah “carried out,” that is, [of a 
meaning, when it is| fully competent for the matter to be conveyed; so 
praudha means effective. An expression also is called praudha when it 
is appropriate to the matter to be conveyed. 

[Abhinava gives a Sanskrit translation of the Maharastri verse "The 
sweet month." He then comments.) Here spring, figured as a sentient 
being and the friend of the god of love, only prepares but does not give 
over [the arrows to his friend]. By this expression, which is effective in 
conveying the meaning which should be conveyed, that stage of spring 
is referred to when the mango is just coming into bud. It is hereby 
suggested that the depredations of love are just beginning and that 
they will gradually grow stronger and stronger. 1f, on the other hand, 
the poet had written, "In the spring the mango begins to bud and 
leaf," he would have stated a fact that suggests nothing. This is an 
imaginative expression of the poet speaking in his own person.? 

“On what mountain": here if the poet had written “the parrot bites 
a red bimba fruit," there would be no suggestiveness at all. But when 
there is an imaginative expression of this sort contained in the stanza, 
spoken by a young man of the poet's invention, who is filled with desire, 
there is suggestiveness. 

324 [52.24 L 

[Abhinava gives a Sanskrit translation of the Māhārāstrī stanza "At- 
tentive youth" and then comments.] Here the girl's breasts have be- 
come great persons, but love is worthy of still greater respect, so they 
are figuratively said to rise up to greet him; and youth acts as their 
attendant servant. By this striking expression the speaker's intention 
is indirectly suggested: that "every man will fall deeply in love with 
you as soon as he sees your breasts." [f he had said, "Your breasts are 
high because of your youth," there would be no suggestion. 

Not [produced] only by: this phrase implies that strikingness of 
expression is always useful. 

(Abbinava translates the Mabarastri stanza “The hunter's wife” and 
then comments.) In his attachment to her the most that he can do is kill 
a peacock. When attached to his other wives, he killed even elephants.” 
Thus the success in love (or sexual attractiveness, saubhágya) [of the 
new wife] is suggested [ukta, i.e., vyañjita] by the wording of the stanza. 
By saying that the other wives are decked with pearl ornaments, that is, 
possess pearl ornaments which are variously arranged, it is suggested 
that because these wives are not preoccupied with sexual enjoyment 
their major occupation is a display of skill in the arrangement of these 
ornaments, which in turn suggests that now (after the arrival of the new 
wife] they suffer the greatest neglect. One need not fear that the poet's 
statement that she is “proud” gives away the suggestion, because it is 
possible for the wife's pride to come from the ignorance of youth or the 
like [and not from a realization of her success in love]. This matter, as 
it is described, or put aside the description and suppose that one were 
to see it directly in the external world, suggests the great success in 
love of the hunter's wife. [Hence it is an inberently possible situation.]* 

1. Abhinava soon overlooks the subvarieties of praudhokti and will speak 
of “the two varieties of suggestion arising from the power of meaning,” 2.25 L, 
first sentence. Of later authors Hemacandra (AC, p. 73) denies explicitly, and 
Jagannatha (RC, p. 136) implicitly, that there are three types. — 2. Spring 
is not really a sentient being and does not prepare and hand over or refrain 
from handing over arrows to the god of love. 3. Later commentators note 
that the elephants would be farther off in the forest, with the result that 
killing them would require of the hunter a longer renunciation of his amorous 
sports. 4. I should prefer a simpler explanation of the stanza. The young 
bride might well be proud of her peacock feathers, for they are the very 
hallmark of a hunter, worn next to his body, around his waist (see KumSam. 
1.15). But Abhinava's explanation is essentially the same as that given by 
Ananda under 3.1i A and it remains the explanation invariably given by 

§ 2.25 L] 325 

later commentators. Also in its favor is the large number of similar verses in Prakrit. 
See the whole of Section 22, on hunters, in the Vazjdlagga, together with Patwar- 
dhan's translations aad notes. 

K Also where a new figure of speech is understood by the power 
of meaning, we have another variety of suggestive poetry in which! the 
suggested sense is like the reverberation [of bell]. 

1, In the Kàriká the relative clause modifies prakérsah (variety). In the 
Vrtti it modifies dhvanih (suggestion). The difference in meaning is mio- 
imal. 2. In the previous four examples of sequential suggestion arising 
by the power of meaning (arthasaktyuduhavasamlaksyakramavyarigyadhuani) 
a meaning suggested a fact or situation (an artha suggested a vastu). The 
present Kárikó shows that a meaning may also suggest a figure of speech (an 
artha may suggest an alarikára). 

A Where a new figure of speech, that is, a figure other than 
an expressed figure of speech, appears to our understanding from the 
inherent capability of a meaning, that is another [type of] suggestive 
poetry, arising from the power of meaning, in which the suggested sense 
takes the form of a reverberation. 

L The two varieties of suggestion! arising from the power of 
meaning have heretofore been described as a form of vastudhvani (the 
suggestion of a fact or situation) because that which was suggested in 
both types was a mere vastu.? He now states that this [same type of 
suggestion arising from the power of meaning] can be an alarikáradhvani 
if that which is suggested is in the form of a figure of speech. Thus he 
says Also where a new figure of speech, etc. The purport of the 
word "also" is that a figure of speech is not necessarily suggested by 
the power of words, as has been described, but may also be suggested 
bythe power of meaning. Or, we may explain the word "also" as 
meaning that it is not only a vostu that may be understood there (viz., 

326 (§ 2.25 L 

in suggestion by the power of meaning], since a figure of speech also 
may be suggested. 

The Vrtti explains the word “new” (anya): other than an ex- 
pressed figure of speech. 

1. See 224 L. note 1. 2. Abhinava fails to mention here that in the 
svateh sambhavin variety the suggestion may be of a rasádi rather than a 
vastumátra. An example is the stanza quoted in 2.22 A. 

A Lest it be thought that the scope of this [type of suggestive 
poetry] must be very small, the following is said: 

K It has been shown that the whole collection of figures of speech 
such as metaphor and the like, which use direct expression, are often 
met with in suggested form. 

A It has been shown by the venerable Bhattodbhata and others 
that figures of speech such as metaphor and the like, which are well 
known to be directly expressed in some occurrences, in other occur- 
rences are understood (i.e., suggested), and this quite frequently. For 
example, it has been shown that in such a figure as doubt (sasandeha)! 
there may appear the figures simile. metaphor, hyperbole, and the like 
And so it is not difficult to prove that one figure of speech may be 
found suggested in another figure of speech.? 

$2.26 L] 321 

1. Sasandeha: In the light of Udbhata's only surviving work one would 
say that Ananda, and following him Abhinava, are here using the wrong word. 
Udbhata distinguishes sasandeha from sandeha (see 1.13i L, note 7) and says 
specifically tbat it is the latter figure which gives rise to the appearance of 
other figures. His example of sandeha (6.3* Induràja, 6.6 Vivrti) is a stanza 
which describes Visnu ‘of whom, black of color and riding on the sun-bright 
king of birds, one doubts whether he is a black cloud on Mt. Meru or smoke 
over the fire of doomsday.” Induràja identifies the suggested figure in this 
stanza as poetic fancy (utpreksd), Udbhata's two examples of sasandeha do 
not suggest a second figure in this fashion. If the reading of our text is wrong, 
the error must be ancient, for Abhinava (see below) clearly read sasandeha. 
Krishnamoorthy in his var. lect. notices sandeha, which he attributes to BP. 
But this is an error; the word does not occur in that text. 2. This is the 
sense that Abhinava finally assigns to the passage. He begias, however, by 
taking alarikdrdntare as a locative of cause, which would give the passage 
the meaning, "it is not difficult to prove that one figure of speech may be 
suggested because another [directly expressed| figure is present." 

L Lest it be thought: the source of his concern is that it is 
easily understandable that by the power of words such figures as slesa 
and the like may appear, but it is hard to see what figure of speech 
could appear through the power of meaning. By the words "the whole 
[collection]? and “has been shown" he makes it clear that this difficulty 
is specious 

'[Bhàmaha 3.43 and Udbhata 6.2 have defined sasandeha as follows.) 
"A statement containing doubt made for the sake of praise by a person 
who states the identity [of the upameya] with the upamdna and then 
again its difference. is known as sasandeha."! For example, 

Is this her hand. or might it be 

a frond whose finger-leaves are moving in the breeze?? 
In such lines there is a suggestion of either si ile or metaphor. And 
as for hyperbole (atigayokti) it is suggested in almost all expressed 

Because another figure of speech is present: [Abhinava begins 
by understanding the Vrtti to say that one figure may be suggested 
because another figure is present, i.e., by means of another figure; cf. 
2.26 A, note 2.] If a figure of speech suggests another. it is not impossi- 
ble that a figure may be suggested by a vastu (fact or situation). If this 
is the meaning, the author of the Vrtti has used the word alarikdrantare 
[to exclude that possibility here], but it does not fit the context. The 

328 {§ 2.26 L 

context here is not that a figure can be suggested by a figure. The 
context is that in dhvani based upon meaning, a figure, just as well as 
a vastu, can be suggested. Thus in the passage where the subject is 
summed up, the Karikd (2.28) will state that "these figures attain the 
highest beauty when they form a part of dhvani^; on which the author 
of the Vrtti begins by saying that a figure can belong to suggestion in 
either of two ways [viz., by suggesting or by being suggested} and con- 
cludes that “here, because of the subject at issue we must understand 
‘a figure that is suggested’ to be meant, (not one that suggests." As 
an alternative (which will avoid this lapse from the subject at issue] 
we shall take the word antara in both occurrences to be a synonym of 
sesa“ and take the locative as a locative of the sphere rather than a 
locative of cause.* Thus the meaning will be as follows. “In an area of 
expressed figures various suggested figures [may also] appear." This has 
been stated by Udbhata and others and they have thereby admitted 
that a figure can be suggested by the power of meaning. The only qual- 
ification to be made is that as they were definers of figures of speech, 
they spoke of these [suggested] figures in an area of expressed figures. 
Such is the real meaning of the passage. 

1. Abhinava does not quote the verse which follows this in Udbhata (6.3 
Indurdja, 6.5 Vivrti), where Udbhata specifically states that in sandeha (not 
sasandeha) one figure may suggest another. 2. The source of this example 
is unknown to us. 3. See below, 3.36. — 4. One of the thirteen meanings 
of antara given by the Amarokosa is bheda = usesa = difference, variety. 
5. See 226 A, note 2. 

is much deserves to be stated: 

K Even where a second figure of speech is apprehended [without 
being directly expressed], if the [first,] expressed figure does not appear 
as subordinate to it, we are not on the road of dhvani. 

§ 2.27 L] 

A Even where there is an apprehension, like a reverberation, of 
other figures of speech in [expressed] figures of speech, if the beauty 
of the expressed figure does not appear chiefly in its conveying of the 
suggested figure, we are not on the road of dhvani For example, al- 
though simile is regularly understood in a figure like zeugma (dipaka), 
if the beauty of the zeugma does not lie in its suggesting the simile, 
one should not use the designation of dhvani [i.e., one should not label 
thesi ile as upamádhvani). For example: 

Night is ennobled by moonbght, 

the pond by its lilies, a vine 

by its clusters of Bowers, 

the beauty of autumn by wild geese, 
and the very name of poetry 

by good listeners." 

In passages of this sort, although a simile is contained [within the di- 
paka), the beauty of the poem lies chiefly in the expressed figure and 
not in any subservience of it to the suggested figure. Therefore it is 
reasonable for the poem to be given its designation by the expressed 

1. Author unknown. For kàvyokathá, literally “all talk of poetry," see 
Abhinava's remarks below. Soyyanaih means literally “by good persons.” But 
what is meant is persons of taste and fair judgment, the opposite of those 
asajjanéh (SRK, Section 38) who seek out the faults of a poem "as a camel 
looks for thorns” (SRK 1255). The expressed figure of the verse is dipaka 
because the same action, ennobling, is predicated of several subjects, one of 
which is prakrta (truly the subject matter, here “poetry”) and others of which 
are not. 

L Now it might be objected that if all this has been said by 
former authorities, there is no need for our author to exert himself. 
Sensing this objection, he says: but this much [deserves to be stated]; 
supply "by us.” The word "but" indicates that there will be some 
difference [in what he will say] from what those [authorities] have said. 

"By moonlight": moonlight achieves its greatest glory only at night; 
and what would good listeners be good for without poetry?! The en- 
nobling of night by moonlight consists in its rendering the night brilliant. 

330 [82.27 L 

and enjoyable; the ennobling of the pond by its lilies in their giving it 
beauty, fragrance and wealth; the ennobling of the vine by flowers, in 
giving it attractiveness and charm; of the autumn by wild geese, in 
their making it a delight to the ear and charming. All these [effects of 
ennoblement| are granted to poetry by good listeners. All these mean- 
ings are furnished by the words "is ennobled" by force of the zeugma 
(dipaka).? The "name"? of poetry implies this: put aside the subtle dis- 
tinctions of poetry; the very word "poetry" disappears without good 
listeners. But when they are present, a mere collection of words be- 
comes blessed and possessed of the appelation "poetry," for the effect 
of these (good listeners] is to bring it to a position of honor. Thus it is 
the figure dipaka that predominates here, not thesi ile. 

1. It is odd that Abhinava begins his comment on the stanza by eliciting 
this incidental suggestion from a reversal of its terms. This reverse implication 
is not important and he does not refer to it again 2. That is, the sense 
"is ennobled by" is supplied to each pair of subject and agent by the figure of 
speech. 3. Abhinava is here explaining why the word kávyakathá (literally, 
"all talk of poetry") is used rather than the simple word kávya (poetry). 

A But where the expressed figure of speech is placed in subor- 
dination to a suggested figure, it is reasonable to designate the poem 
by the suggested figure. As in 

"Why should he, who has attained to royal glory, 
who has won Srl for wife, 
burden me again with the pain of churning? 
I cannot believe that one so active 
Should seek his former sleep. 
Why, when he is attended by lords of all the islands, 
should he build a bridge once more?" 
Such are the doubts, it seems, which make the ocean tremble 
when your Majesty marches to its shore." 

§2.27a L] 
Or? as in a verse of my own: 

Truly insensate is the ocean 

that it is not now stirred by this your smiling face, 
tremulous-eyed beauty, 

which fills the horizon with the splendour of its loveliness.? 

In instances sucb as these, inasmuch as tbe beauty of the poem lies 
in a metaphor that appears like a reverberation. it is reasonable to 
designate the poem as poetic suggestion of metaphor (rüpakadhvani). 

1. Author unknown. The stapza flatters the royal patron of the poet 
by suggesting, in a fused figure (samkara) of doubt (sandeha) and poetic 
fancy (utpreksd, indicated by "it seems”), that the patron is Visnu whom the 
ocean knew on three previous occasions: before the obtaining of Sri from the 
churning of the sea, in the time of pralaya when Visnu slept on the sea, and 
when, as Rama. he built a bridge to Lanka to defeat Ravana. — 2. At this 
point we should probably read the passage rejected by Abhinava; see below. 
We omit it in the translation only in order to avoid repetition. 3. The 
verse contains a pun. The woman's face, like the moon. should stir the ocean 
(alarási) if it were not insensate (jadarási). In puns l and d are regarded as 
identical (ladayor abhedoh). The verse has been picked up by several of the 
anthologies; see Kosambi's apparatus on SRK 421. 

L In this way he has shown the [negative] meaning of the Karika 
by'a [negative] example. He now explains the positive intention implied 
by what the Kàrikà has excluded. This positive intention is that where 
[the expressed figure shows] subordination to the suggested figure, we 
are on the road of dhvani: But where, etc. Actually there are three 
possibilities here [viz., in the area of figures of speech suggested by 
the power of meaning|: (a) a second figure may be suggested by an 
expressed figure; (5) there may be an expressed figure but one that 
does not suggest; and (c) there may be no expressed figure at all. One 
sbould attach these possibilities, each as it may fit, to the examples 
[that follow].* 

He illustrates: "Why should he, who has attained to royal glory," 
etc. A certain king has come to the shore of the sea with an immense 
collection of troops, whereupon the sea begins to tremble (or surge) 
eitber because of moonrise or because of the plunging of these troops 
into it.? As this trembling is fancied to be due to doubts (as described 
io the stanza], we have the fusion of sasandeha and utpreksd, so the 

332 (§227aL 

expressed figure is fusion (samkara). And by this there is suggested 
the superimposition of the character of Vasudeva (= Visnu] on that 
king [ie., the figure rüpaka is suggested]. Granted that a contrasted 
superiority (vyatireka) of the king (to Vasudeva] also appears, that is 
because of a superiority to Vasudeva in his former, not in his present, 
condition. Because the blessed Vasudeva has now obtained Sri and 
lives as an active monarch who is victorious over the kings of all the 
continents. The rüpaka is not forced upon us by any impossibility of the 
fancied doubt so that we might say that it served simply as a support 
(upaskára) to the expressed figure, for we can imagine (the ocean's} 
reasoning to take the form that anyone who has not obtained wealth 
and is inspired by an unconcealed desire oí conquest might be likely to 
churn its waters.‘ Nor is this sense [of the identification of the king with 
Vasudeva] forced on us by the words “again” (punar api), “former” 
(pürvam), and “once more” (bhüyah). Because the words "again" and 
"once more" can reasonably be used even if the agents should differ, 
since the ocean remains the same. As one might say, "The earth was 
formerly conquered by Kartavirya and then again by Jamadagnya."* 
And "his former sleep" can be explained as the king's habit of sleep 
when he was a prince [before he had assumed the responsibilities of a 
ruler]. Thus it is established that the stanza is an example of suggested 
metaphor (nipekadhvani)® [and not a case of a metaphor subordinate 
to an expressed figure]; because we apprehend the superimposing [of the 
character of Vasudeva on the king] in the absence of direct denotation 
simply from the beauty of the expressed sense." 
Here some [authorities] read another example: "And as in 

On this sand bank of the Sarayü 

whitened by flooding streams of moonlight 
two angels once held long debate. 

One claimed that Kesip was the first to die; 
the other, Kamsa. Tell us truly now 

which did you slay first?" 

This passage is spurious,® because the sense that "you are Vasudeva” 
is here made clear by the direct expression of the word "you." 
"Loveliness" (lávanya), that is, a charm of configuration; "splen- 
dour,” i.e., brilliance. The horizon is "filled," that is, furnished with, 
made enchanting by, these two qualities of your face.? "Now^: now that 
your face inclines toward graciousness after its recent disturbance by 
[the] anger [of jealousy]. "Smiling": with lips slightly parted in a smile. 

§2.27a L] 333 

“Tremulous-eyed beauty”: the word is a vocative, designating a per- 
son of whom the eyes are tremulous, that is, beautiful because of their 
motion and dilation prompted by their owner's graciousness. And yet 
the ocean is not stirred “now,” although it was stirred a moment ago 
[when the moon rose]. Inasmuch as your face, flushed by its [recent] 
disturbance of anger and [now] smiling is the very disc of the full moon 
rising at sunset,!? any sensitive being must be stirred, must experience 
a motion of the heart. As the ocean "is not stirred," it clearly exhibits 
an accumulation of insentiency (jádyasancaya) in conformity with its 
name "the accumulation of waters” (jalarási). We have said before [see 
1.1 Intro. L and note 10 thereon] that adjectives like jada can be nom- 
inalized [e.g., jada can be used in the sense of jádya]. In the stanza the 
denotative function of the words comes to a halt after furnishing the 
sense "a sensitive person must be Stirred,” must experience an alter- 
nation brought about by love. “on looking at your face." Accordingly, 
the metaphor (rupeka, i.e., tbe identifying your face with the moon) 
must be the work of suggestion. The pun (slega) is an expressed figure 
of speech in the stanza, but this pun does not suggest anything. The 
metaphor that appears like a reverberation is suggested by the power 
of meaning; and as the beauty of the poem depends on this metaphor, 
the poem should be designated by it as an instance of rüpokadhvani. 
This is the connection [between the verse and the thesis that stood i 
need of illustration}. 

1. The passage from “actually there are three possibilities here" (tatra 
ca. etc.) to “one should attach to the [following] examples" (udáAaranesu 
yogyam] becomes clear after one reads further in the corumentary. Abbinava 
was apparently struck by the fact that several of the examples given under 
this Kàrikà by the Vrtti do not exhibit a figure suggested by another figure. 
Although that seems to be the ana marked out by the Kariké, the Vrtti cov- 
ers the wider area of figures suggested in any way by the power of meaning 
Abhinava divides this wider aren into three categories. The first, which is 
clearly referred to by the Karikd in where the suggested figure is suggested by 
an expressed figure. This is exemplitied by the following example, praptasrir 
esa “Why should he.” The secon, where the verse exhibits an expressed fig- 
ure but where the suggested figure is Suggested by something other than the 
expressed figure, is exemplified hy the stanza lávanyakánti “Truly insensible 
is the ocean.” The third, where ther is no expressed figure but where there 
is a suggested figure, would be exemplified by the verse jyotsndptraprasara 
"On this sand bank of the Sarayu, which Abhinava rejects as an insertion, 
if the suggestion were not spoil! by the direct expression "you." — 2. These 

334 [§227aL 

are the real causes of the surging of the sea, as opposed to the fancied cause. 
In supposing that the trampling of an army could disturb the sea, Abhinava 
was doubtless picturing in his mind a Kashmirian lake rather than the Indian 
Ocean. 3. If the nipaka arose in this way, it would be subordinate to the 
samkara and we could not categorize the poem as rüpakadhvani. The point 
of view that Abhinava here denies might be put as follows. So long as the 
character of Visnu is not superimposed on the king, the speculations (vitarka) 
regarding the purpose of the king's visit to the ocean, which constitute the 
Sondehdlarikdra, are not reasonable. Nor is the fancy (utpreksá) that these 
speculations are the cause of the ocean's agitation reasonable. The superim- 
position is needed in order to justify or rationalize the expressed figures. But 
Abhinava denies this view. The speculations regarding the purpose of the 
king's visit to the ocean do not force the rüpoko upon us, he says, because 
they can be differently explained. 4. As the ocean was supposed to be a 

ine of jewels, anyone desiring wealth would be a potential miner of these 
jewels. Again, a king bent on conquest would be likely to churn the sea by 
tranporting his army to distant lands. It does not take a Vispu to think of 
troubling the sea. 5. The argument which Abhinava rejects might be put 
thus. We grant that any ambitious king might churn the sea. But only Visnu 
could churn it "again" and build a bridge across it "once more." Abhinava's 
rejection points out that the adverbs can refer to the relation of verb and 
Object rather than to the relation of verb and subject. The sea, churned 
by Visnu, cap be agitated at the thought of being churned by someone else. 
6. This categorizing of the verse by Ananda and Abhinava is criticized in RG, 
p.247. 7. These words are added to make it clear that the present example 
is a case of arthosaktyudbhavadhvani 8. Abhinava's criterion of textual 
authenticity is that what the text says must be worthy of his author. If, in his 
opinion, it is not, the text must be spurious. By modern principles of textual 
criticism the rejected passage has a good claim to be genuine. [t is included 
in the Nepali manuscript of the Dhv. (see Krishnamoorty's ed.) and is cited 
in the Vyaktiviveka (p. 430) where it appears between the verses práptasrir 
esa and lávanyakónti, precisely its position in the Nepali MS. 9. Abhinava 
takes ldvanyakanti as a dvandva. In the translation we have taken it as a 

10. The full moon also as it rises turns from red to white. 

§ 2.27b L] 

A The following is an example of the poetic suggestion of si 
The eyes of warriors take not such joy 
in their ladies’ saffron painted breasts 
as they take in the cranial lobes, painted with red minium. 
of their enemies' elephants." 

Another example is from my Visamabànalila. speaking of the conquest 
of the demons by the god of love: 

Their hearts once bent on theft of those gems 
born of the same womb as Sri 

were transferred by the god of flower arrows 
to tbe bimba-like lips of their women. 

1. The comparison of a woman's breasts to the two frontal lobes of an 
elephant's cranium is a stock simile of Sanskrit poetry, but the notion that a 
warrior might take more joy in handling and crushing the upamdna than the 
upameya of this simile is original. 

L Our author gives two examples of suggested simile (upamá- 
dhvani), but does not state expressly how the definition [of a suggested 
figure] applies to them because the application is the same [as in the 
preceding examples of suggested metaphor]. 

[Abhinava gives a Sanskrit translation of the Prakrit verse "The eyes 
of warriors," on which be comments as follows.| Here, while there is 
a weighing of the inclination [of these warriors] to fondle their orna- 
mented ladies against their emotional eagerness for an approaching bat- 
tle, the eagerness for battle.is [shown to be] greater. Thus the directly 
expressed figure of speech is contrast (vyatireka). But a similitude is 
suggested between the cranial lobes of enemy elephants, ready for bat- 
Ule, objects of terror to all men, and the swelling breasts of their ladies. 
By this si ile we see the esteem [of these men for battle], as if they 
were deriving sexual pleasure from it, and this si ile effects a striking 

336 [$ 2.27b L 

portrayal of their heroism. Hence the [suggested] simile is of primary 
importance [in the verse]. 

The conquest of the demons: for in that poem his conquest of 
all three worlds is described. “Their hearts,” that is, the hearts of the 
demons who dwell in Pátála and who had engaged in every sort of 
outrage, such as injuring the city of Indra and the like; hearts whose 
resolution was unshaken by the most arduous adventures.! "Born of 
the same womb as Sri": the sense is "and therefore of inexpressible 
value." The hearts of these demons which had been bent on, that is, 
wholly intent on, the theft, that is, the abduction from every [hiding 
place], of these gems: these hearts were transferred by the god of flower 
artows—he was supplied with the gentlest artillery imaginable—to the 
lips of their women; that is to say, Kama brought their hearts to regard 
the gazing at, ki ing, and biting of these lips as the highest purpose 
of these demons’ lives, whereas their hearts had just now been blazing 
with the fire of martial ambition. Here the expressed figure of speech is 
hyperbole (atisayokti).? The suggested figure is simile, for the bimba- 
like lips are similar to the best of al! gems and hence the high value 
placed upon these lips is real. That is why the suggested figure is 
not rüpaka, because a rüpaka is a superimposition [of the character 
of one object on another, actually different, object] and so is unreal? 
Tbe similarity of the lips of their women to the best of gems appears 
to these demons as a matter of actual fact and this similarity by its 
predominance is the source of the striking effect of the verse. 

1. The words tam Aiaam of the verse are glossed by Abhinava by 
Ardayam tac ca: the heart, and such a heart! 2. Mammata quotes 
this verse (Book 10, vs. 515) as an example of the trope parydya (where 
one obj t, here "their hearts,” is described as existi in different places). 
Paryáya is apparently an invention of Mammata's, as it is not found in such 
older authors as are preserved to us. His reason for the assignment of it to this 
verse is that atíayokti had ceased to mean hyperbole in his time and the verse 
will not fit easily into any of his four types of atiayokti. But the verse fits 
naturally into the older definitions, e.g., that of Bhamaha 2.81 "A statement 
the sense of which exceeds reality, when made for some [poetic] purpose should 
be considered atisayokti (hyperbole)^; or of Udbhata (Vivrt 2.23). BP by 
ingenuity manages to fit the present verse even into a late definition of ati- 
Sayokti. "Heart," it says, really refers to nothing more than cittavrtti, a mental 
or emotional state. The demons' emotional drive toward robbery really had 
no connection with their later amorous state, but the two states are here 
pictured as connected. This fits the definition asambandhe sambandhah of 

§2.27¢ A] 337 

Alarikárasarvasva, p. 83 and of SdhDarp. 10.47. 3. If I say “My lady's face 
is like the moon,” my statement is true or real because the face does have, 
at least to me, qualities of brightness and charm which are Like those of the 
moon. But if I say her face is the moon, the literal sense of my words is false. 
This distinction between the reality of upamá and the irreality of rüpaka goes 
back to the beginning of the Alankara tradition; cf. BANS 16.56, savikalpena 
racitam rüpakam, "rüpoka is a figure formed by one's imagination." 

A An example of poetic suggestion of dksepa (feigned or preg- 
nant denial) is as follows: 

He can express all Hayagriva's virtues 
who can measure by jars the water of the sea.' 

Here by means of the [expressed] figure hyperbole there appears a [sug- 
gested] óksepa, which takes the form of proclaiming that Hayagriva's 
virtues are indescribable and which has the purpose of showing that in 
their excellence these virtues are unique.” 

1. The verse is later quoted by Hemacandra AC 1.77 and AlSary, p. 151. 
Possibly it is taken from Bhartrmentha's lost poem, the Hayagrivavadha. 
2. The directly expressed figure here, in the older system of poetics, is hy- 
perbole (atisayokti) because measuring the sea in jars is a purely imaginary 
action that exceeds the bounds of reality; see 2.27 b L, note 2. In the later 
system the figure would be identified as nidarsand, as it is identified by our 
late commentators on Sak. 1.16, which contains a similar turn of phrase. The 
suggested figure is áksepa, a denial which hints at something unexpressed; cf. 
1.13e A, note 3. Bhamaha (2.68) defines the figure as “a feigned denial of 
what is or was one's intention, made with the purpose of emphasis (or ex- 
aggeration).": pratisedha ivestasya yo visesdbhidhitsayd. What he means can 
best be seen from examples. “It is wonderful that you have no pride although 
you have conquered the earth. But what can produce an alteration of the 
sea?” (Bhàmaha 2.70). "As he thought about her, it was wonderful that 
his thought knew no end. But where is there any end to thoughts of love, 
or time?" (Udbhata, 22* IndurSja, 2.5 Viurts). In both examples there is 
implied a denial (the denial is not directly expressed in these examples) of 
the predication “wonderful.” but in both cases the denial is not because the 

338 (§2.27¢ A 

denied fact is not wonderful (it is wonderful) but because the word wonderful 
is not strong enough. The graciousness of the conquering king and the never 
ceasing thoughts of the lover are more than wonderful; the one is as steady 
as the ocean, the other as ceaseless as time. In the present example cited by 
Anandvardhana, there is an implied denial that the virtues of Hayagriva can 
be enumerated or described (both Ananda and Abhinava speak of avarnani- 
yatd rather than aparisarikhyeyatá). This is not because the poet really means 
that they are indescribable (his poem doubtless went on to describe them), 
but because he wishes to emphasize or exaggerate. He means that Hayagriva's 
virtues were unique. 

L By means of hyperbole: he means, as the expressed figure. 
The implication that his virtues are indescribable is a form of áksepa 
(feigned or pregnant denial), because it is a denial of what is actually 
intended.! He shows us that this àksepa forms the predominant sense 
of the verse by the adjectival compound "[which has the purpose of 
showing that these virtues are] unique.” 

A Poetic suggestion of the figure substantiation (arthántaranyá- 
$a) is possible both as a suggestion arising from the power of words 
(sabdasaktyudbhavavyarigyadhvani) and as a suggestion arising from the 
power of meaning (arthasaktyudbhavavyarigyadhvani). An illustration 
of the former is: 

Since fruit depends on fate, what can be done? 
This much, however. we can say: 
the flowers of the red asoka tree 
are unlike those of others. 
[Sattasai 3.79]' 

As the suggestion is here revealed by a word (viz., phale "fruit"], it does 
not conflict with the overall sense of the sentence. which is different. 

$2.27d A] 
An example of the second variety is as follows: 

O clever lover, to apologize 

even when I have disguised my face 
and hidden the grievance in my heart 
Though you have been unfaithful, 
one cannot be angry.? 

By the fact that one cannot be angry at a particular individual denoted 
[by the primary force of the words], who is clever though unfaithful, 
there is suggested as the primary sense a substantiating general state- 
ment connected [with that particular statement]. 

1. The Sattasoí reads devdattammi, the regular reflex of Sanskrit devá- 
yatte, and Pischel knows only the form datta. But dya might change to e on 
the analogy of aya, so our reading may be correct. For karilillapallaváh read 
karikellipallavà. Karikelli is the Maharastri name of the asoka; see Hema- 
candra, Abhidhánacintámani 1135. The word occurs again in Sattasai $.4. 
The asoka tree bears brilliant red flowers, which appear on its fronds (pallavdh) 
before the leaves appear. Its fruit is inconspicuous. The direct meaning and 
suggested figures of the stanza are as follows. Direct meaning: "The asoka 
tree bears no fruit; that cannot be helped. [t bears beautiful lowers." There 
is no arthdntaranydsa (substantiation, see 1.13i L, note 8) in the direct mean- 
ing because a particular cannot be substantiated by a particular. The first 
suggestion arises from the double meaning of the word phale (fruit, reward) 
and so is a suggestion arising from the power of a word. "The asoka tree bears 
novfruit. The reward of all living things depends on fate." This is arthdntara- 
nyàsa of Mammata’s second type (10.109), where a particular statement is 
substantiated by a general statement. The second suggestion depends on the 
power of meaning and derives from the stanza as a whole, read in the context 
of advice to a king. "Even a hero may fail of success, for success depends on 
fate. But his virtues shine above those of other men." This is aprastutapra- 
Samsd, a figure which consists in a statement (prasamsd) of a non-contextual 
or allegorical (aprastuta) meaning and which suggests a contextual (prostuta) 
meaning, that is, a sense that concerns the matter that is really in one's mind 
The prastuta meaning in an aprostutaprasamsá is always suggested; it cannot 
be expressed. A problem now arises. Our text has said that we can speak of a 
suggested figure of speech only when that figure forms the predominant sense 
of the verse (2.22 K). How can there be two predominant suggested senses 
in one stanza? Ánanda does not answer the question very satisfactorily. He 
merely says that there is no contradiction because one figure is suggested by 
a single word, the other by the sentence as a whole. 2. Viewing this little 
verse out of context and without reference to the commentators, one may be 

340 (§2.27d A 

puzzled how to take it, whether as the remark of a woman who is naive, forgiv- 
ing, cynical, or clever. Ananda’s implied and Abhinava's expressed opinion 
is that it is the remark of a clever woman, who has been hurt and who is 
letting her husband or lover know that she sees through his excuses. This 
is why they assign the verse to "suggested substantiation” (arthdntaranydsa- 
dhvani), a figure that demands that a particular statement be substantiated 
by a general statement or vice versa. The particular statement of the woman 
concerning her lover suggests the general principle that false lovers always act 
in this way. After reading Abhinava's comment on the verse one will agree, I 
think, that this interpretation gives the verse a greater charm than it would 
bave under other interpretations. 

L Is possible: by this word he shows that the consideration 
at this point of suggestion based on the power of words is incidental 
[the matter properly under consideration being suggestion based on tbe 
power of meaning]. 

[After translating the Prakrit stanza "Since fruit depends on fate," 
etc., into Sanskrit, Abhinava continues:| Unlike other trees such as the 
mango and the like, the asoka tree does not bear fruit. What can be 
done about it? But its flowers are extremely beautiful. By express- 
ing this mucb the denotation of the sentence is complete. Before [we 
reach this point], we apprehend a substantiation of this [particular] 
fact through the [double] power of the [ambiguous] word pole (fruit or 
reward). The general proposition takes the form of the following obser- 
vation: "Sometimes, because of fate, the result ip tbe form of success 
may not be achieved even by a man of uncommon ambition who sets 
forth with the correct means." But now a difficulty arises because an 
allegory (aprastutaprasamsá) is suggested! as the predominant mean- 
ing of the sentence as a whole. How then can the figure substantiation 
be suggested. since two predominant meanings cannot coexist? In view 
of this difficulty he says, As the suggestion is here revealed by a 
word. For it will be stated (in 3.1-2} that all the forms of suggestion are 
revealed either by single words or by complete sentences. In the verse 
we are discussing substantiation (arthántaranyása) is the predominant 
suggestion in a single word, whereas allegory (aprastutaprasamsá) is 
the predominant suggestion in the sentence. But of these two it is 
the relation of substantiating and substantiated statements (i.e., the 
arthantaranydsa) furnished by the word phale that appears in greater 
predominance? and so one will categorize the verse as an example of 
suggested substantiation (arthántaranyásadhvani). 

§2.27e A] 341 

“When I have hidden the grievance in my heart”: literally, “I being 
one by whom the grievance is placed within the heart and not revealed.” 
Accordingly, you clever one who apologize to me even when I show no 
anger in my face, there can be no cause of anger at you although you 
have been unfaithful. The vocative singular “O clever one” determines 
[the denoted sense to refer to] a particular. But then as one considers 
the sense, a substantiation in the form of a general statement is appre- 
hended and this is strikingly poetic. For we have here a woman who 
bas been wronged [by her husband or lover], who is clever and who, 
when he apologizes, shows her annoyance by saying (in effect|: “Every 
clever rascal although he has been unfaithful covers up the occasion 
of his transgression in this way; do not pride yourself falsely on that 
account.” Connected: sc., because the general statement is connected 
with the particular. 

1. aprastutaprasamsa prádhanyena vyarigyd: Abhinava is speaking inex- 
actly, as BP points out. What he means is that the contextual (prastuta) 
meaning of the aprostutaprasamsd is suggested (aprastutaprasamsdsthale pro- 
stutártho vyarigyah). The distinction should be kept in mind because in 2.27h 
Abhinava will give us an example where the aprustuta sense as well as the pra- 
Stuta sense is suggested. That is not the case here. — 2. Abhinava gives no 
reason for this greater predominance. BP says simply that it appears that 
way to men of taste. 

A Poetic suggestion of contrast (vyatirekadhvani) is also possible 
in both forms [sc., by the power of words and by the power of meaning]. 
The first of these types has been illustrated above (2.21 f A]. An example 
of the second is this: 

1 would rather be born somewhere i 
as a crippled, leafless tree 
than be born in the world of men, a man 
yearning to give and poor. 
[Sattasai 3.30] 

In this example it is directly stated that to be born as a poor man who 
wishes to be generous is not a matter for rejoicing and that to be born 

342 L5 2.27e A 

as a stunted, leafless tree is a matter for rejoicing. This statement, after 
we first apprehend the comparability (upamdnopameyatua) of such a 
man with such a tree, shows as the predominant meaning of the verse 
that the man's degree of misery is greater. 

L A suggestion of contrast is also possible: By using the 
word “also” he shows that just like the suggestion of substantiation 
(arthantaranyásadhvani), the suggestion of contrast vyatirekadhvani) 
is also of two kinds. Above refers to the verse “May both sets of the 
sun god's feet” [2.21 f A].' 

“I would rather be born somewhere in a forest": in a secluded part 
of a forest, in a thicket where, being surrounded by the luxuriance 
of numerous conspicuous trees, I shall not even be noticed by anyone. 
“Crippled”: unfit for being made into any shape. "Leafless": the mean- 
ing is that it affords no shade, how much less should it have flowers or 
fruit. What the poet wishes to convey is that even such a tree might 
be useful for making charcoal or might serve as a roost for owls. “Of 
men”: that is, where suppliants are plentiful. "In this world”: sc., 
where he is seen by suppliants and where suppliants are seen by him 
and yet he can do nothing [to help them]. The poet means that this is 
an agonizing situation. There is no directly expressed figure of speech 
in the verse. 

Comparability: thus our author prepares the ground for contrast 
(vyatireka) (as contrast is always based on similarily. Shows that 
the degree [of misery is) greater: i.e., shows the contrast.” 

1, In our text ‘kham ye 'tyujjvalayanti' iti is followed by ‘raktas tvam 
navapallavaih' iti. As BP notes, the second reference is in error, for that 
stanza was not an example of vyatirekadhvani but of vácyavyatireka adopted 
after giving up the slesopamá. It is easier to suppose that a careless reader 
added the extra reference than to suppose that na tu has fallen out before 
raktas. 2. Cf. Mammata 10.105 upameyasya vyatireka ádhikyam ""Contrast" 
is where the subject of comparison (upameya) possesses [some property in] a 
greater degree [than the image (upàmiána)]." 

§ 2.27f A] 

A’ An example of the poetic suggestion of fancy (utpreksadhvani) 
is this: 

In spring the Malabar wind, 
swollen by the breathing of snakes 
that encircle sandalwood trees, 
makes travelers swoon.’ 

For in this example the capacity of the Malabar wind to cause travelers 
to swoon in’spring is [actually] due to its stimulation of love. But this 
capacity is fancied (utpreksita) to be due to the swelling of the wind 
by the poisonous breath of snakes wrapped around the sandalwood 
trees [of Malabar]. This fancy, although it is not directly stated, is 
noticed like a reverberation because of the inherent capability of the 
sentence meaning ? One may not object that because there is no use of 
such words as iva [e.g., “as if," “like,” "as it were^] the fancy in such 
cases is disconnected (asambaddha, also “nonsensical”), for it is readily 
understandable, as we find in other cases too that fancy arises through 
suggestion even in the absence of such words. For example in 

Although your face is afflicted by anger, 

the full moon, having achieved this once 

a similarity thereto, 

cannot, it seems, contain himself. 

Or, asi 

A timid deer ran about among the tents. 
No men with bows pursued him; yet he rested not, 
his lovely eyes being struck with shame 
by the ear-reaching arrow-glances of the women. 
(Magha, SisupalavadAa 5.26] 

When it comes to the relation between words and sense [i.e., to the 
question whether particular words suggest a given sense or not], the 
generally accepted view is authoritative? 

344 [$ 2.27f A 

1. Sandalwood trees grow in Malabar and are said to be a favorite haunt 
of snakes (SRK 801, 1078). The verb murcchati means both to swell and to 
swoon. Swooning may be caused either by snake bite or unfulfilled love. The 
spring breeze brings back to travelers memories of their wives and mistresses 
waiting at home. With this verse compare Dandin 2.238. — 2. Or, because 
of the natural compatibility of the sentence meaning with such a fancy. On 
sóámorthya see 2.21a A, note 2. 3. BP: As persons of taste will understand 
the sense of fancy in such passages even in the absence of a word like ivo, the 
rejection of such passages as disconnected or nonsense is wrong. P 

L Is fancied: because [it is fancied that] the wind, swollen 
by, that is magnified or increased by, poisonous exhalations, causes 
swooning. Furthermore, it is fancied that one individual among the 
travelers |viz., the wind], being poisoned (mürcchita), causes rnürcchá, 
that is, loss of health, to other individuals. So we bave a double fancy 
(utpreksà). Now it might be objected that we have no fancy here at all, 
because the adjectival phrase (“swollen by the breathing of snakes”), 
being otherwise a useless addition, makes sense only if taken as giving 
the cause [of the increase of the wind and the swooning of the travelers]. 
But what if the phrase is taken as giving the cause. This is not the 
teal cause [of those effects]; rather, it is fancied as the cause. So the 
objection amounts to nothing.’ Of such words: because we see that 
this sense, viz., that of something fancied, may be understood even 
when words like 1va are not used. He now illustrates this point: As 
in, etc. 

"Afflicted by anger": tinged with redness. Should the moon attain 
similarity to your face when your face is pleased, or tbat it should 
always be similar, then what would it not do (in its joy|? That it could 
really become your face—why, this would be beyond the power even of 
its daydreams. This is what the word "although" (api) conveys. He 
“cannot contain himself" in his own body, as be is filling the borizon 
[with moonlight]. "Just this once" means after such a long time and 
then too for one night only (sc., for the night of the full moon]. In this 
stanza the filling of the horizon by the light of the full moon, which is 
a natural event, is fancied in this manner [sc., as being prompted by 
tbe moon's desire to equal the face of the woman]. 

But now our author senses that the word nanu ("it seems”) might 
be regarded as a expression of reflection, that is, of fancy (utpreks4), 
and tbat by this direct expression the disconnection (asambaddhatà) 
[to remove which, as we bave claimed, the reader supplies a suggested 

$227g A] 345 

fancy) might be removed. Accordingly, he gives us another example: 
Or, as in, etc. 

That the deer, running among, that is, all about, the tents, did not 
halt although it was not chased by any men with bows and arrows, 
was due, in actual fact, to its natural timidity and restlessness. But a 
poetic fancy is here suggested: that the deer did not balt because the 
beauty of its eyes, which was the deer's most treasured possession, had 
been hit [i.e., surpassed] by the women with arrows in the form of their 
large eyes (also: arrows pulled back on the bow string as far as the 
ears]. Our author anticipates a criticism, namely that this verse too? 
is disconnected (or nonsense) as there is no word like iva to denote the 
fancy; and so he says:when it comes to the relation, etc. 

1. In actual fact the wind is not appr iably increased by the breath of 
snakes, nor in actual fact does it poison travelers so that they swoon. — 2. As 
well as the verse, "In spring the Malabar wind.” 

A The following contains a poetic suggestion of puns (slesa). 

Where young men with their wives enjoyed 
covered terraces with sloping eaves, 
with curving folds at the waist, 

decked with banners to give them beauty 
winning fame because they were beautiful 

and stirring passion by their privacy. 
by their adornment.! 

(M&gha, Sisupalavadha 3.53] 

In this stanza after we have perceived the direct sense of the sentence, 
viz., that with their wives they enjoyed the teraces, we perceive puns 
(slesa) to the effect that the terraces are like the wives. This $lesa, 
[which arises} from the non-verbal inherent compatability of the (ex- 
Pressed] meanings [with the second meanings], acts as the predominant 
figure of the stanza 

346 [5 2.27g A 

1. This stanza, which forms part of Māgha's desccription of the city of 
Dv&raké, has been analyzed by Vallabhadeva and Mallinütha as well as by 
Ananda and Abhinava. There is no disagreement about the double meanings 
of the adjectival phrases. But the commentators disagree on how to interpret 
the little word samam ("together with”) and on this question hangs the un- 
derstanding of the expressed meaning of che stanza, the identification of its 
figures, and the problem of what, if anything, the stanza suggests. Samam 
is an adverb, but does it modify the verb's relation to its subject or to its 
Object? In the first case, the direct meaning will be: "together with their 
wives the young men enjoyed the terraces.” In the second case, the direct 
meaning will be: "the young men enjoyed the terraces together with, that 
is, as well as enjoying, their wives." Only Mallinatha takes the latter course. 
Ananda and Abhinava take the former course and explain the appearance of 
the second meaning as a suggestion. Ánanda's explanation, which is ellipti- 
cal, and Abhinava's, which is clearer, may be put as follows. After we have 
understood the direct meaning (the first of those just given), there occurs to 
the mind the second meanings of the adjectival phrases. These senses are 
suggested to the mind by the natural compatability of the primary sense for 
these second meaniogs. What our authors mean is that the primary sense 
gives us an erotic context in which it is entirely appropriate that the women 
be beautiful, endowed with the graceful triple fold at the waist, and elegantly 
adorned. Once we perceive these puns, we realize that the terraces are like the 
wives. That is to say, we arrive at the second meaning by way of a simile that 
grows out of the suggested puns. As the charm of the verse lies in the sug- 
gested puns, it should be counted as an example of slesadhvani. Mallinitha, 
setting off from a different direct meaning of the verse, sees in it the expressed 
figure tulyayogitd (paired objects). The Dhvonyóloka analysis seems to me a 
better description of the psychological process of understanding the stanza; 
Mallinátha's of course is simpler. 

L “Possessing patdkdh,” that is. [of terraces] possessing flags on 
flagstaffs. The phrase ramyé iti ("to give them beauty") expresses 
the cause.' [As applied to the women] "possessing patákáh" will mean 
possessing fame. The phrase ramya iti will then express the form that 
this fame takes [i.e., the fame of being beautiful]. Viviktah (private) [is 
used of the terraces] because they are not crowded with people; and for 
that reason they increase raga, that is. the desire for sexual enjoyment. 
Other [commentators], however, explain rága here as color, "having a 
profusion of beautiful color."? Also, (as applied to the women,| the sense 
will be "increasing (their husbands’) passion (anurága) since they are 
viviktàh, of ornamented limbs, that is, handsome. (The terraces] have 

§ 2.27g L] 347 

curved eaves, that is, curved edges of the thatch. (The women] have 
curved lines at their waist in the form of the triple fold. Samam means 
together with. But someone may say that from the word sama (equal, 
similar to) one understands [that the terraces are] similar to (the wives]. 
That is true; that too (is understood] from the puns. But the puns are 
not brought in by the denotative force of the words, but by the inherent 
compatibility? [of the primary sense for these second meanings]; and 
so the puns are most certainly suggested. That is why the author of 
the Vrtti, while be says tbat the terraces are like the women, does 
not say that we have here a case of suggested simile (upamádhvani), 
because the simile is based on the puns. If the poet bad said clearly 
[that the young men enjoyed terraces which were] samá (similar to) 
[their wives], then the simile would be directly expressed and the puns 
would be brought in by the simile. But the indeclinable samam has 
the direct meaning of saha ("together with"), which can furnish a pun 
only by suggestion through its modifying the verb.‘ If «e omit the puns, 
there is no incompletion of the denotative power of the words. So it is 
only after the denotative power of the words has exhausted itself that a 
second meaning is understood and this only by persons of refined taste, 
who understand it without any separate effort.* As bas already been 
said (1.7 K], “Not by a mere knowledge of grammar and dictionaries 
is the suggested sense known." And this [principle of analysis] is to 
be followed in all (these) examples. In the sentence -Fat Caitra does 
not eat by day," it is the denotative power that is incomplete and 
that demands a further meaning or a further word for its exhaustion 
Afid that demand furnishes occasion for the inference (anumdna) of a 
Naiyàyika or the verbal presumption (srutárthápatti) of a Mimamsaka. 
There is no occasion there for suggestion.’ But enough of this long 
lexplanation]. It is this [principle of analysis] that our author refers to 
when he says non-verbal. 

1. The terraces were decked witb flags because they (the flags) were beau- 
tiful; ramyé refers to patākāh, not to volabhih. — 2. Abbinava is here probably 
referring to Vallabbadeva. who in his commentary on the Sisupélavadha ex- 
plains raga in this verse as referring to the painting of the terraces with red 
realgar and minium. 3. The context. I think, demands crthasámarthya in 
place of arthasaundarya. — 4. In vadhübhih sama valabhih tke word for similar 
connects "terraces" and "wives" directly. In vadhubhih samam valabhir asev- 
yanta "terraces" and “wives” are connected only through the verb. — 5. If the 
denotative sense were incomplete without the puns (as in tke verse of 221c A 

348 [$2-27g L 

"The cloud serpents pour forth water"), the puns would be part of the deno- 
tative process and would not be suggested. Furthermore, the understanding 
of the two meanings of the puns would be by two separate efforts, for there 
would be two denotative powers to understand. Here the second meaning 
arises after the denotative process has ceased and its understanding is part of 
the same effort by which we understand the denoted meaning. 6. By "this 
principle" Abhinava evidently means the rule that once a logically complete 
meaning has been given by direct denotation, any further meaning that ap- 
pears in a poem does so by suggestion, not by anumána or arthápatti, of which 
he goes on to speak. 7. The Naiyàyika regards the sentence "Fat Caitra 
does not eat by day" as furnishing the liriga for the inference that Caitra eats 
at night. The Bhátta Mimàmsaka regards it as presuming a suppletion by the 
word rátrau (“at night"). 

A An example of poetic suggestion of yathdsarikhya (ordered 
sequence) is the following: 

The mango tree puts forth 
sprout, leaf, bud and flower. 
In our heart love puts forth 
sprout, leaf, bud and flower." 

In this stanza there is a beauty similar to a reverberation in the fact 
that the words "sprout," etc., used in the predicate of "love," occur 
in the sarne sequence in which they were used in the prior statement 
[about the mango]. This beauty is noticed as something distinct from 
the expressed figures of tulyayogità and samuccaya that apply to love 
and the mango tree.” 

In this way other figures of speech may be adduced [as examples of 
figures suggested by the power of meaning], according to the proper 
connection in each case.? 

1. BP notes that for the nominatives in the first verse-balf another read- 
ing gives locatives; and such is the reading of SRK 188 and Sadukti. 1232: 
arikunte pollavite karakite puspite ca sahakére. But the change to nominatives 
is doubtless intentional on Ánanda's part and is repeated in the Vyaktiviveka, 
P. 442; see the following note. None of the sources gives the author's name 

$2.27h L] 349 

2. The expressed Ggures are tulyayogità, the assignment of the same verbs 
to two different subjects (the mango tree and love) that are both proper to 
the context; and samuccaya, defined by Rudrata 7.27 as the occurrence of 
the same quality or action in two different objects at the same time. Yathà- 
sarikhya (ordered sequence) occurs when a number of things are referred to 
in a later passage in the exact order in which they were referred to in a prior 
passage. In an expressed yathdsarithya the sense of the sentence is not logi- 
cally complete or satisfactory unless we apply this principle of ordering. “You 
surpass the lotus, the moon, bees, elephants ... by your face, radiance, eyes, 
gait ..."(Bh&maha 2.90). If we mix up the references, we get nonsense. So 
yathdsarikhya is necessary for an understanding of the primary sense. The 
same necessity applies to our present verse if we read locatives in the first half 
line, for the verse will then mean that as the mango puts forth sprouts, love 
begins; as it leafs, love unfolds; as its buds appear, love develops into a sharp 
yearniog; and as it flowers, love reaches it full development. The yathdsankhya 
principle is needed ih order to assign the appropriate conditions to the appro- 
priate effects. But if we read nominatives, the verse sets two whole sentences 
against each other, as we have done in the translation of A. It is not necessary 
to correlate the components of the two sentences in order to avoid nonsense. 
Rather, the perception comes as a pleasant surprise, after we have understood 
the direct meaning, that the components of the two sentences can be sequen- 
tially paired off. Thus the yathásarikhya appears as a suggestion similar to a 
reverberation and not as an expressed figure. One may note that SimhabbG- 
pila (also spelled Si&gabhüp&la and Singabbüp&la) equated the terms of our 
little verse with various Sanskrit words for love: arikura with prema, pollava 
with māna, koraka (= kalikd) with pronaya, puspa with sneha, phala with rága 
(Rasárnavasudhàkaro, ed. T. Venkatacharya, 2.111ff.). Mallinátha refers to 
this passage at KumSam. 8.15. 3. [n some cases the connection will be with 
à suggestor (vyaijaka) which is itself a figure of speech; in other cases with a 
suggestor that is a vastu (fact or situation). 

L In this way other [figures of speech]: Every figure of 
sense (arthdlarikdra) can be found in suggested form. For example, a 
suggestion of dipaka (zeugma) is: 

Not fire or wind 

or maddened elephant! 
Not ax or lightoing! 

O tree, may you be ever 
safe with your vine. 

(The meaning of the verse is "May fire, wind, etc., not harm you,” but} 
the zeugma of the verb “harm” is hidden, from which we understand 

350 [$2.27h L 

that the tree is specially dear to the poet! and by which the verse is 
given beauty. 
A suggestion of aprastutaprasamsá is as follows:? 
Flying about the ketaka trees, 
you will kill yourself on their thorns, O bee; 
while in your wandering you will not find 
a flower sweet as the jasmine.? 
A certain lady, walking in a garden with her lover, sees a bee and 
addresses these words to it. Thus the direct statement about the bee 
is contextual.‘ You cannot say that we understand this meaning to be 
non-contextual from the use of the vocative, because the vocative can 
be contextual, being due to the lady’s simplicity. So the aprastuta- 
prosomsá cannot be reached simply by the power of denotation. It is 
only after the denotative process has been completed that a reference 
to some other matter is suggested by the force of the directly expressed 
sense.’ The suggestion is that a lady of good family, beautiful and 
proud, comparable to a delicate, sweet-smelling jasmine, prompted by 
her sincere love, berates her lover who has been roaming hither and yon 
among prostitutes of low birth, who have gained a reputation for their 
artificial skill, who are surrounded by thorns in the persons of their 
bawds, and who can well be represented by a grove of ketaka trees with 
its far-reaching perfume. 
A suggestion of apahnuti (denial accompanied by false affirmation) 
is found in a verse by my teacher Bhattenduraja: 
On this reservoir of nectar, whose bold and lovely curve 
is like the curve of a pale maiden’s breast, 
that which we see, precious as the application 
of ornamental markings in black aloe, 
is Love, my graceful lady, stretching out his limbs 
on that cold bed to assuage the heat engendered 
by his dwelling in the heart of many a damsel 
burning with the fire of separation.* 
In this verse a denial of the mark on the moon is suggested by the 
[false] affirmation that it is the figure of Love, turned black by the fire 
arising from the hearts of damsels who suffer the fire of separation. 
The same verse carries also a suggestion of the figure sasandeha 
(doubt), for the mark of the moon has not been mentioned by name but 
has been referred to as exhibiting the value, or excellence, of the spot 
painted with ornamental markings of black aloe on the moon as rep- 
resented by the curve of a pale maiden's breast. Accordingly, a doubt 

$2.27h L] 351 

is suggested to the effect that "I do not know what it is." There is 
also a suggestion of prativastupamá.? The prativastupamá that is sug- 
gested is this. A lady who has refused to accept her lover's apologies 
subsequently feels remorse and pain at the separation occasioned (by 
her anger], so she prepares to welcome her lover back and after com- 
pleting the decoration of her person is ready in the love chamber for 
his arrival. It is a full-moon night and her lover arrives, guided by the 
go-between. The lover flatters the lady, telling her, "the ornamental 
markings on your swelling breast stir my passion." to which the anal- 
ogous sentence is that "this beautiful appearance, dark as the petal of 
a blue waterlily, on the moon does the same." '? The term "reservoir of 
nectar," although it is used only as a synonym for the moon, still sug- 
gests the reason for the god of love's seeking to allay his heat there. So 
we have also a suggestion of hetvalarikára (the figure cause).!! There 
is also a suggestion of sahokti (pairing) in the form that the beauty 
of your breasts and the beauty of the moon together stir my passion. 
From the apprehension of the sense that "the moon is like your breast 
and the curve of your breast is like that of che moon, the figure upa- 
meyopamá" is also suggested. In this way, other varieties [of suggested 
figures of speech| can be imagined. For these words of a great poet are 
like a magic wishing-cow, (endless in the gifts they can furnish]. For it 
has been said: 

The playful gesture of one man produces 
results unending; another's effort 

cannot achieve a particle of gain. 

If an Elephant of the Quarters stirs a hair, 

he shakes the earth, while if a bee should fall 
from heaven itself, he would not shake a vine.'? 

The combination of these figures [in Bhattendurája's verse] into sam- 
srsti and sarikara (associated and fused combinations) can be worked 
out (by the reader for himself]. 

A suggestion of atisayokti (hyperbole) is in a verse of my own: 

Your eyes are the chief embodiment of beauty's springtime 
when it sends forth its first shoots of gaiety: 

the successive motions of your eyebrow are the bow 
bending in Love's hand to its varied curvature; 

the wine of your lotus mouth, fair lady. at first sip 

begets intoxication: truly God has centered 

in you alone the treasures of his universe. 

352 [§ 2.27h L 

There is a suggestion of atisayokti here, for while spring, love, and wine 
are blessed with success in the world by reinforcing one another, they 
are here [said to have} combined into an unique body in your person.” 
There is also a suggestion of the figure vibhdvand (the occurrence of 
a result in the absence of cause), for by saying that “the wine of your 
mouth causes intoxication at first sip” it is suggested that a result, 
namely intoxication, comes about without its proper cause, namely a 
frequent imbibing (of the inebriant]. There is also a suggestion of tulya- 
yogità (paired subjects) in saying that the lady's two eyes are the chief 
embodiment of beauty’s springtime. 

It should be understood that in this way every figure of speech can 
be suggested. It is not that only certain figures can be suggested within 
certain limits as some writers claim.'* 

According to the proper connection in each case: one should 
connect (with the suggestion] the (power of} meaning that is appropriate 
in each case. In some cases the suggestor will be a figure of speech; in 
some cases it will be a vastu (a fact or situation). 

1. It is so dear to him that be silences the unlucky word bédhista (“harm”). 
The dipoka, not being expressed in words, is suggested by the contextual 
meaning. One cannot argue that as the verb must be supplied in order to 
get a logically satisfactory sense, the dipaka should belong to the denotative 
operation; because we might get a satisfactory primary sense by supplying 
a different verb with each subject (may fire not burn you, wind not break 
you, etc.). The cboice of a single verb, which occasions the dipaka, is due to 
suggestion. Abbinava does not here observe the distinction between dipaka 
and tulyayogitd that we fnd in Udbbata and later authors. See 1.13f L, note 1. 
2. Here Abhinava is speaking precisely (contrast 2.27d L, aote 1). Not only 
the contextual (prastuta) meaning but the non-contextual meaning (aprastu- 
ta) itself will be suggested. — 3. The verse is included as stanza 985 in Weber's 
supplement to the Sattasai (Das Saptacatakam des Hala, p. 512). It is also 
quoted by Mammata. The ketaka is Pandanus oderatissimus, Linn., a tree 
with leaves "drooping, from three to five feet long, tapering to a very long, 
Bne, triangular point, very smooth and glossy, margins and back armed with 
very fine sharp spines” (Roxburgh, 707). Without being given some special 
explanation one would take the verse as a directly expressed aprastutapragam- 
så It is only by inventing a special context that the figure can assume the 
form of a suggestion. Abhinava proceeds to furnish such a context. 4. That 
is to say that the expressed meaning is not aprostutaprosamsó, where the 
direct meaning must be non-contextual. 5. The point of the objection is 
that a person would not normally address a bee in this personal fashion. In 
order to make sense of the vocative we have to suppose that the lady is really 

§ 2.27h L] 353 

referring to something else. 6. The lady is not a biologist or a logician. 
She might well address a personal remark to a bee. 7. The description of 
the bee's folly is singularly compatible with the folly of the lady's lover. So a 
reference to his fickleness suggests itself to us. — 8. The "reservoir of nectar” 
is the moon. That which appears on the moon is its mark (our “man in the 
moon"). Women painted designs (pattrabhariga) in red saffron or in black 
aloe on their breast or cheek; see Ingalls, An Anthology of Sanskrit Court 
Poetry, HOS Vol. 44, note on vs. 389, p. 498. This verse is discussed by 
J. L. Masson, “Abhinavagupta as a Poet," Journal of the Oriental Institute 
(Madras) Vol. 19 (1970), pp. 247-253. 9. A prativastipamd is a figure 
consisting of two sentences, the latter forming an analogy to the former. In 
the later system of figures (from Udbbata onward) a prativastüpamá must 
fulfil) three conditions: (/) there must be no word such as iva to denote the 
similarity; (2) the analogous objects must possess the same quality or action 
(not a similar one as in the figure udáharana (“example”); (3) the common 
property must be expressed by a different word or phrase in each of the two 
sentences. An example is Kálid&sa's description of King Dilipa (Rogh. 1.18) 
"To benefit bis people he took taxes from them. The sun draws up water to 
pour it forth a thousandfold." "Take" and "draw up” refer to the same action 
by different words. 

10. The literal sense of the stanza is tbat a substance which lies on the 
moon and which has a beauty like the beauty of aloe markings on a woman's 
breast, is really the god of love. This suggests two sentences related as in a 
protivostüpamó: (1) The aloe markings on your breast are really the god of 
love"; and (2) "The dark substance on the moon is really tbe god of love.” 
The phrase "dark as the petal of a blue waterlily” is introduced by Abhinava 
in order to fulfill the third condition of a protivastüpamá, mentioned in the 
previous footnote, that the phrase used to describe the common quality in the 
second sentence must be different from the phrase used in the first sentence, 
11. Hetu is Dandin's name for the term (2.235). Udbhata (Induraja 6.7) 
calls it kdvyahetu and latet writers know it as kdvyaliriga, 12. Upameyopa- 
md: a simile followed by a sentence in which the two terms of the simile are 
reversed. 13. Source unknown. In the cosmology that regards the earth as 
Bat and circular, four (or sometimes eight) elephants are said to support it at 
the chief compass points. Their stirring causes earthquakes. The impact of a 
bee, on the other hand, though the bee falls from a great height, is scarcely 
felt. 14. For once BP errs i. interpreting the sentence. The identification 
of the lady's eyes-with springtime, of ber eyebrows with love's bow, and of 
her mouth with wine are metaphors (rüpaka) and are directly expressed. The 
atisayokti consists in joining three objects (spring, love and wine) which in 
the real world are not joined but merely cooperative. This joining is only 
suggested. Spring, love and wine are not actually said to be joined in the 
Stanza, but as they are identified with the eyes, brow and mouth of a single 

354 [5227h L 

body, we see that they must be pictured as joined. 15. We do not know 
who these authors were. 

A Having thus analysed the province of suggested figures of 
speech, in order now to show their usefulness [the Kariká| says: 

K Those figures of speech which cannot be made into the body 
of a poem when they are directly expressed, attain the bighest beauty 
when they form a part of dhvani. 

A Forming a part of dhvani can occur in two ways, by being a 
suggestor or by being suggested. Here, in view of the context, we are 
to understand that being suggested is meant. But figures of speech, 
even when suggested (vyarigya), form a variety of dhvani only when 
they form the predominant sense of the passage. Otherwise they fall 
under “subordinated suggestion,” as will be explained [viz., in 3.34]. 

L Now figures of speech bave long ago been defined by the crit- 
ics, so it might be doubted that anything very wonderful has been 
accomplished by our author in sbowing that tbey may be suggested. 
To allay this doubt he says, Having thus analysed, etc. 

[Abhinava first interprets the Kàrik by taking vácyatvena as a single 
word, thus eliminating tbe negative.] "Those figures of speech which by 
being directly expressed are made into the body of a poem." Making 
into the body of a poem means taking a figure, which as it is something 
other than the subject matter in band is not naturally the body of the 
poem any more than a bracelet or the like is the body of the person 

$229 K | 355 

who wears it, and making this figure into the body of the poem, a 
transformation that can be effortlessly achieved by great poets. 

[Abhinava next gives the more natural interpretation of the Kérikd, 
dividing the word vácyatve from na so as to express a negative.) Or 
[we may interpret the Karikd to say that] figures of speech which are 
not made into even the body of a poem—the sense is that it is difficult 
so to make them—when they are directly expressed: these same figures 
of speech by taking a suggested form become an integral part of the 
operation of suggestion or of the poem itself and even attain to the 
highest and rarest beauty, which one may call the very soul [of the 
poem}. What is implied is this. A good poet, like an experienced 
woman, uses ornaments skilfully; and yet, it is difficult for him to make 
an oranment pass for the body of a poem, just as it is difficult to make 
rouge pass [for the true flesh]; how much less can an oranment become 
the soul itself. But such is the property of being suggested that it 
imparts to an ornament, even when the ornament is not predominant, 
a superiority over expressed oranments. [t is like the property of royalty 
which imparts a peculiar distinction even in tbe games of children.’ It 
is with the foregoing in mind that our author says, otherwise, etc. 

1. The "king" even in children's games acts differently from other chil- 
dren. Compare Horace, Epist. I, i, 59-60. 

A Now there are two ways in which a figure of speech can be 
suggested as the predominant element: it may be suggested by a mere 
fact or situation (vastu), or it may be suggested by [another] figure of 
speech. Of these [two ways], 

K when figures are suggested by a mere fact or situation, they 
invariably form a variety of dhvani. 

(§ 2.20 A 

A The reason for this is that 

K the poetic functioning is founded on them. 

A That is, because in such cases the whole poem comes into 
being in dependence on this sort of suggested figure. Otherwise the 
poem would be merely a statement (and no poem at all]. 

L Of these: there being these two ways. The reason for 
this: these words form part of the Vrtti, [not the Karikd]. Because 
the "poetic functioning,” that is, the functioning of the poet's activity 
is slanted toward the figure which depends on that (fact or situation]. 
Otherwise; sc., if it did not come into being in dependence on that. 
Accordingly, there is no room in such cases for a suspicion that the 
suggested figure might be subordinate.’ 

1. The reasoning that underlies K, A, and L in this section is the fol- 
lowing. In order to have poetry one must either have suggestion predominant 
in the poem (as in dhvani), or one must have an olarikára predominant in 
the poem (as in subordinated suggestion). When one olarikáro (directly ex- 
pressed) suggests another alarikdra, the question can arise whether the second 
alarikdra is predominant or the first. But where a mere fact or situation sug- 
gests an alarikaro, the suggested alarikdra must be predominant, for if it were 
not, the verse would not be poetry at all. 

A These same figures 

K when they are suggested by another figure, 

§ 2.30 L} 

K they will form a part of dhvani if the suggested sense is seen 
to be predominant by its greater degree of beauty, 

A for we have already said (1.13e A] that “the decision whether 
the literal or the suggested meaning is the more important depends on 
which is the more charming.” 

The province of figures of speech which are suggested by a mere 
fact or situation can be deduced from the examples just given. Ac- 
cordingly, arthasaktyudbhavánurananarüpavyarigyo dhvanih (that type 
of suggestive poetry where the suggested sense appears like a rever- 
beration arising from the power of meaning) is to be understood as 
occurring wherever a meaning of any sort, or a meaning in the form of 
a particular figure of speech, gives rise to a second meaning or a second 
figure of speech such that the second is predominant by its possessing 
a greater degree of beauty than the first. 

L These same figures: this furnishes the words that must be 
supplied in the Káriká which immediately follows. Then: this word 
ig supplied in the middle of the Karikd. The word dhvanyarigatà [used 
in the Kàrikà and meaning literally "a property of dhvani"| means "a 
variety of dhvant" If the suggested sense is predominant: the 
reason follows: by its greater degree of beauty. “If: what our au- 
thor has in mind is that if the expressed sense rather than the suggested 
sense is predominant, the suggested sense wil! belong to "subordinated 

Now a figure of speech is sometimes suggested by a fact or situation 
and sometimes it is suggested by another figure of speech. So it may 
be asked why our author did not give examples [of the (ormer variety 
as well as of the latter].! He replies [to this question by the sentence, 
"The province of figures of speech which are suggested] by a mere fact 
or situation, etc.” 

He summarizes the whole matter in conclusion: Accordingly, etc. 
The upshot is that there are four varieties of arthasaktyudbhavadhvani 

358 [$2.30 L 

(suggestion arising from the power of meaning), deriving from the two 
forms, vastu (fact or situation) and alarikára (figure of speech), in which 
either the suggestor or the suggested may appear. 

1. All the examples under 2.27 are of an alarkára suggested by an 
alerikéra. Under previous Karikds examples were gives of suggestion by a 
vastu, but not of an alankdra being suggested by a vastu. 

A Thus the varieties of dhvani have been given. The followi 
is now said in order to distinguish them from the false varieties 

K Where the suggested meaning appears indistinctly,’ or as sub- 
ordinate to the expressed meaning, that is not the province of this (type 
of] dhvani. 

1. Manuscript authority seems to be equally divided between promlista- 
tvena, “indistinctly” (see Pag. 7.2.18) and proklistatvena, “with difficulty." 

A A suggested sense is of two sorts: it may be clear or indistinct. 
Whether occasioned by the power of words or of meaning, it is only 
that suggestion that appears clearly that falls in the province of dhvani, 
not an indistinct suggestion. And even a clear suggestion is not in the 
province of the [type of] dhvani where the suggested sense is similar to 
a reverberation, if it appears as subordinate to the expressed meaning. 
For example: 

O auntie! Without touching the lotuses 
or scaring away the geese, 
someone has laid out a cloud 
upside down on the village tank." 
(Mianka, Sattasai 2.10} 
The suggestion here, that a naive girl bas seen the reflection of a cloud 
in the water, is subordinate to the expressed meaning. 

$2311] 359 

1. malià = Sanskrit mrdit&h (Pischel p. 171), not malitdÀ as given in 
the cháy. piucchd = Sanskrit pitrsvasar (Pischel p. 112); 1 know not what 
confusion has lead to the gloss sahasd. uttànoom: on its back, i.e., in the 
reverse of the position it held in the sky. phaliham (= Sanskrit sphatika, 
parigha, partkhd) makes no sense to me; surely it cannot mean proksiptam as 
Abhinava seems to take it. I have substituted the Sattosai reading vvüdham 
(= vyüdham). Both Ananda and Abhinava see nothing more in the verse 
than a naive expression of wonder. If that is all there is, they are correct in 
finding a greater charm in the naive expression itself than in the suggested 
explanation of optical illusion. But the commentators on the Sattasai see 
much more. A young wife, they say, has gone off in the early morning on the 
pretext of drawing water from the village tank before it is muddied by the 
advent of others, but really in order to keep a rendezvous with her lover. The 
lover, however, failed to appear. Later he comes by as the young woman is 
chatting with her aunt. The verse is spoken by the woman in order to inform 
him that she kept her promise as he did not, but at the same time to hide 
this information from her aunt. As so often with verses from the Sattasai, one 
does not really know how much the author himself intended. 

L Thus (the varieties of dhvani have been given, etc 
The two major divisions [of suggestion] are: avivaksitavácya (where 
the literal sense is not intended) and vivaksitaparavácya where it is 
intended, but is subordinated to a further, suggested, sense). The for- 
mer, is divided into atyantatiraskrtavdcya (where the literal sense is 
entirely set aside) and arthdntarasarikramitavdcya (where the literal 
sense is shifted). The latter is divided into alaksyakrama (where no 
interval is perceived between the literal and suggested meanings) and 
anurananaripa (where the suggestion is similar to the reverberation 
of a bell). Of this pair the former has endless varieties. The latter 
has two: sabdasaktimüla (where the suggetion arises by the power of 
words) and arthasaktimüla (where the suggestion arises by the power 
of meaning). This last has three subdivisions: kavipraudhoktikrtasarira 
(where the suggestion is embodied in an imaginative expression of the 
poet), kavinibaddhavaktrpraudhoktikrtaéarira (where the suggestion is 
embodied in an imaginative expression of a character invented by the 
poet), and svatahsambhavin (where the suggestion is inherently possi- 
ble in the real world). Now each of these three subdivisions is of four 
kinds depending on the distinction just given that the suggested sense 
and the suggestor may be either [a vastu (fact, situation) or an alarikára 

360 (§2.31 L 

(figure of speech)]. Adding to these the four major divisions of sugges- 
tion that were mentioned first (viz., atyantatiraskrtavácya, arthéntara- 
somkramitavacys, alakgyakrama, and sabdasaktimüla), we arrive at six- 
teen varieties of suggestion. Now each of these sixteen varieties will be 
stated [3.1 K] to be twofold according as the suggestion is revealed 
by a word or by a sentence. But as olaksyakramadhvani is revealed 
not only by words and sentences but by phonemes (varna), by style 
(samghataná), or by an entire work |cf. 3.2], we get in all thirty-five 

To distinguish, that is, to separate, [these thirty-five varieties| from 
the false varieties of suggestion 

That sort of poem is not the province, not the domain, of this 
dhvani which is the soul [of poetry]. 

[Abhinava translates the verse from the Sattasai into Sanskrit as 

The lotuses have not been sullied 
and the geese have not suddenly been driven off 
[but] someone bas thrown a cloud 
upside down into the village tank. 

But others say that piucchá [does not mean “suddenly,” but] is a voca- 
tive, meaning "aunt." "Someone": some one extremely dexterous. 

Subordinate to the expressed meaning: from the expressed 
meaning, it being in the form of a manifestation of wonder, we under- 
stand the extreme naivete [of the girl]; and so the charm of tbe verse lies 
just in the expressed meaning. On the other band, it is only because 
the expressed meaning wants support in order to become rationally 
intelligible that it suggests to us a second meaning.! 

1. To say that a cloud has been thrown (as Abbinava understands the 
verse) into the village tank is absurd and irrational. The absurdity is ratio- 
nalized by our accepting the suggestion that the gir! has seen a reflection of 
the cloud in the water. 

§2.31b A] 

As the young wife 
busy with her housework hears birds 
flying up from the cane thicket, 

her limbs fail her.* 

Such verses, as will be shown later, are generally to be adduced as 
examples of subordinated suggestion. 

1. The verse is included, as number 874, in Weber's edition of the Sat- 
tasai. The young woman has apparently promised to meet her lover in the 
cane thicket but is prevented by the dreary chores of her married life. The 
Aight of the birds tells her that her lover has kept his promise and is await- 
ing ber in vain. The suggestion (that her lover has entered the thicket) has 
no particular charm or beauty; hence Mammata categorizes the verse under 
Gsundoram vyarigyam (5.132). The description of the young woman, how- 
ever, physically broken by her disappointment, moves our emotion. Thus the 
expressed meaning is predominant over (more beautiful than) the suggested. 

L [After translating the Prakrit verse, Abhinava comments as 
follows.| What is here suggested, namely that the secret lover has 
arrived at the agreed upon rendezvous, serves merely to support (or 
rationalize) the expressed meaning. To explain: "busy with her house- 
work": although her mind is intent on other things. "The young wife": 
although she is constrained by great shyness and by subserviency (to 
her'elders]. “Her limbs”: there is not a single limb which can be pre- 
vented by the deepest dissimulation [from revealing her longing]. "Fail 
her”: far be it from finishing the housework, she cannot even support 
her own body. While she was engaged in housework her limbs were not 
seen to be in that state. From this expressed sense we understand that 
the young woman is utterly overcome by love and from this comes the 
charm of the verse. 

A But where the particular meaning of a direct expression has 
been determined by our understanding of the context or some like factor 

362 [$2.31b A 

and where this expressed meaning then appears as subordinate to the 
suggested meaning, we are indeed on the road of this type of dhvani 
that is similar to a reverberation. For example: 

O farmer's bride, 

gather the flowers on the ground 

and don't shake the sephalikà tree. 
Your bangles will end on an ugly note 
if your husband's father hears them.' 

Here we have a wife, who is engaged in sex with her paramour, being 
warned by a friend because of the noise, heard afar, of her jingling 
bangles. This [context] is necessary in order to understand the direct 
meaning. But after the expressed meaning bas been understood, inas- 
much as it has been expressed only in order to furnish the final meaning 
which is the hiding of the woman's adultery, it becomes subordinate to 
the suggestion. Accordingly, the verse should be included in the type 
of dhvani where the suggested sense is similar to a reverberation. 

1. This verse too is included in Weber's edition of the Sattasai (num- 
ber 959). The reading of our text in c can scarcely be right, as virávo repeats 
the sense of saddo. The Sattosai reading is esa avasánaviroso. Abhinava 
seems to have read esa avasánavisamo. For the sephálikà (or sinduvdra) tree, 
see M. Emeneau, University of California Publications in Classical Philol- 
ogy 12 (1933-34) pp. 333-346 and Ingalls, HOS 44, p. 490 (note on vs. 271a). 
The tree blooms only at night, with beautiful, scentless white lowers. Thus 
the time of the adultery as well as the place (a garden) is indicated. The 
father-in-law is presumably within bearing of the friend's advice. He might 
well be annoyed at his daughter-in-law's shaking flowers from his favorite tree, 
but the friend regards the rousing of such annoyance as a lesser evil than the 
fury which would seize the old man if he guessed the true cause of the jingling 

L But where: where the sense has its nature determined by 
our understanding of a factor in that set of factors beginning with con- 
text and including the proxi ity of another word, inherent capability, 
gender, etc.,! but where this expressed meaning then—then be 
cause, having been directly expressed, its own understanding has been 
completed—does not conclude the sentence meaning by just this sense 
but goes on to assume a subordinate role in a suggested meaning: such 
an instance is in the realm of dhvani. Our author's clearly stating the 
connection with a final meaning in the form of suggestion is tantamount 

§ 2.32 Introduction A ] 363 

to saying that we must consider such cases to be the very opposite of 
“subordinated suggestion.” 

-Your bangles will end on an ugly note if your husband's father hears 
them”: because the father-in-law has been preserving the sepháliká tree 
with particular care and will be angry at its being pulled or shaken. We 
are to suppose [the expressed meaning to be] that the ugly result will 
be on this account, for otherwise [that is, if the ugly result were taken 
to be on account of the daughter-in-law's adultery], the suggested sense 
of the verse would be given away by the direct expression. Beyond this, 
we are to interpret the stanza as we did the stanza, "Who wouldn't be 
angry" (kassa và na hoi roso).? 

This, i.e., this suggestion, is necessary for our understanding, our 
getting at, the direct meaning. That is to say, without it no direct 
meaning can be got because this meaning [without the suggested sense 
given by the context] would be something perfectly obvious [to the 
adulteress| and so not worth saying. But now it might be objected 
that at this rate the suggested meaning serves merely to support (or 
rationalize) the expressed meaning [and so is subordinate to it]. To 
prevent such an objection our author continues: but after the ex- 
pressed meaning has been understood; in other words, after it 
has been directly expressed. 

1. The reference is to the list of factors by means of which an inherently 
ambiguous expression can be narrowed down to the particular sense intended. 
The,list occurs in Bhartrhari, VP 2.315-316 and is repeated by Mammata 
under 2.19. 2. See 1.4f A, L, where Abhinava gave six different suggested 
senses, the suggestions differing according to the various persons supposed to 
be overbearing the verse. Tripathi (p. 649), folowing Abhinava's direction, 
does the same for the present stanza. 

A While on the subject of distinguishing in this way vivakgita- 
vàcya dhvani (sugestive poetry where the literal meaning is intended 
but leads to a further meaning) from the appearance thereof, one may 
distinguish avivaksitavácya (suggestive poetry where the literal mean- 
ing is not intended) in the same way. To do this, the K. árikà says: 

[52.32 K 

K fa word that shifts (from its primary meaning] (skhaladgati) 
is used out of lack of mature judgment (avyutpatti) or lack of skill, such 
[a word] too is not in the province of dhvani. 

A Ifa word that shifts (from its primary meaning], that is, a word 
of secondary sense (upacarita), is used out of lack of mature judgment 
or lack of skill, such (a word] too is not in the province of dhvani. 

L While on the subject of distinguishing: this is a loca- 
tive, used causally. The sense is, because of the cogency [to the subject 
about to be raised] of the (previouis| mention (prastáva) of the dis- 
tinction from the appearance—appearance of what?—the needed com- 
pletion of the sense is furnished by the word vivaksitavdcyasya On 
the other band, if we take prustuta in its simple (sposta) meaning (of 
“begun” or "under discussion"|, the passage makes no sense, for the 
distinguishing of vivoksitavácya from the appearance thereof has al- 
ready been completed; it is not now begun, nor is it to be continued in 
what follows. 

A word that shifts from its primary meaning: that is, a metaphor- 
ical (gauna) word or a word used in a relational secondary sense (là- 
ksanika).! Lack of mature judgment: as writing with a view merely 
to compose alliterations and the like; for example: 

Prerithatpremaprabandhapracuraparicaye praudhasimantininám 
cittákásávakáse viharati satatam yah sa saubhdgyabhumih. 
Happy is he who strolls within the rooms 

of women's bearts, tbe whicb are well acquainted 

with every subtle sort of swaying love. 

Here the word "swaying" (prerikhat), used in a relational secondary 
sense [sc., for unsteady, fickle], has been used because of the poet's 
passion for alliteration and the metaphor “rooms” which the poet has 
used [for hearts?] leads to no beautiful goal whatsoever in the form of 

§ 2.33 Introduction A ] 365 
Lack of skill: such as inability in filling out the meter; for example: 

O foremost of the numerous entourage of Love, 
your sinking to the ocean, this dish of rolling waves, 
has imparted undulation to your level self. 

Here the first word (the compound word “foremost of the numerous 
entourage of Love") is an indirect expression for the moon; “dish” is 
used for receptacle; and "level" (literally “wall-like”) for unmoving. 
None of this produces any beauty [or has any use at all} except to fill 
out the meter.‘ 

Such [a word] too: In tbe first Uddyota, apropos of "poets who 
use words in senses furnished merely by convention," our author gave 
an example of secondary usage (bhákta = upacanta), viz., “the lotus- 
petal couch speaks [tbe fever of a slender maid|.^ (1.14 A]. The force of 
the word "too" in the present passage is to say that not only was that 
example not in the domain of suggestive poetry, but so also a word 
such as is here referred to. 

1. See 1.1 K, note 2 2. BP seems to take ciltàkáía as a rüpaka- 
Samdso, i.e., cittaivdkdsa, heart and space having the same property of im- 
perceptability. Such an interpretation explains Abhinava's denomination of 
ākāía as gauna. But it would be more natural to take cittākāía as equiva- 
lent to Ardayakdsa, the physical interior of the heart. The real vice of the 
stanza, it seems to me, is not its misuse of laksand but its verbosity: avakdsa 
adds nothing (except alliteration) to cittákása, pracura is needless; saubhagya- 
bhdmih achieves a repetition of the phoneme bh but means nothing more than 
subhagah. 3. The only poetic purpose of using laksand is to produce dhvant 
Here it has been used only to produce alliteration. 4. The drutavilambita 
meter, incidentally, is mishandled despite all the padding, for the Grst caesura 
(between saficaya and pravara) falls within a word, forcing an unnatural pro- 
longation of the final syllable of saricaya. 

A This is because 

[82.33 K 

K the clear appearance of the suggested sense as the predom- 
inant sense of the passage is the essential mark of dhvani in all its 

A And this mark is found only in such examples as we have 
Here ends the Second Chapter of the Sahrdayáloka! composed by $n 
Rájánaka Anandavardhana. 

1. As in the colophon to the First Chapter, the MSS vary between 
Sahrdayóloka and Küvyáloka. None of them reads Dhvanydloka, the form 
printed in the Kashi text. 

L Taking the view that the Karikà now repeats the definition 
of dhvani as a cause for distinguishing it from the appearance thereof, 
the author of the Vrtti introduces the Kanka with because. 

The appearance: what is meant is "the suggested sense appearing 
clearly (as the predominant sense|"; this on the principle that when 
told to bring a verbal abstract, one brings a substance (qualified by 
that verbal activity! The essential mark of dhvani: tbat is, the full 
nature of dhvani. Or we may take "appearance" to mean the perception 
of dhvani; that is the mark, i.e., the proof (pramána) of dhvani and it 
is complete, for it informs one of the whole nature of dhvani. Or, the 
perception is the mark of dhvani because the mark of dhvani can only 
be ascertained by perception [of the suggested sense]. 

By the word only the author of the Vrtti indicates that what differs 
therefrom is a false appearance of dhvani and has thereby carried out 
the purpose which he undertook to distinguishdhvani from the appear- 
ance thereof. [May my words prove] auspicious. 

I, Abhinavagupta, praise God's perceptive force, 
which by its perceiving of the world, 

this vast and mere appearance, 

makes it to seem other than Himself? 

§ 2.33 L] 367 

Herewith the Second Chapter of the Sahrdaydlokalocana, an exposi- 
tion of dhvani, revealed by the great Saiva master, the revered teacher, 

1. The principle is enunciated by Pataüjali on Pan. 2.1.51 and again on 
5.1.59. For example, we cannot bring a “collection of five bushels,” if we 
understand “collection” (saméhdra), as Patañjali does in those passages, to 
be a verbal abstract meaning “the activity of collecting." We bring the five 
collected bushels. 2. In the Perceiving (Pasyanti) stage of metaphysical or 
linguistic evolution, arise time, the concept of tbe ego, and a differentiation 
of subject and obj t. See 1.19 L, note $ 


A [n this manner the nature of dhvani and of its varieties has 
been described through [an analysis of] that which is suggested. It will 
now be described again through (an analysis] of those factors which act 
as the suggestors.” 

L cal to mind the Goddess. 
who, after Siva had shown his skill 
in effortless annihilation of Love's body, 
stole half of Siva's body for herself. 

The author of the Vrtti proceeds to establish a logical connection 
with the preceding chapters: In this manner, etc.. While it is true 
that in the preceding chapters the types [of dAvani] such as avivaksi- 
tavácya were distinguished by means of tbe literal sense (vdcya) and 
thus, insofar as the literal sense may act as a suggestor—and we see 
from 1.13 "either sense or word, etc." that it can—these types have 
already been distinguished by the factors which act as suggestors, nev- 
ertheless these literal senses were there analysed according to wbat was 
suggested by them. Thus the unintended literal sense (avivaksitavdcya) 
was subordinated to the suggested sense and the intended literal sense 
(vivaksitányaparavácya) was said to “lead toward the suggested sense." 
Thus tbe meanings that act as suggestors of the basic types of sugges- 
tion and of the sub-types were distinguished only by recourse to, by 
reference to, the meanings they suggested. Therefore he says: through 
an analysis of those factors which act as the suggestors. 

But more than this, while it is true that a meaning that acts as 
a suggestor may also be capable of being suggested, a word can act 
only as a suggestor and can never be suggested. For this reason too 
he says: through an analysis of those factors which act as the 


370 [8 3.1 Introduction L 

suggestors.' [By the Vrttik&ra's use of the word punah ("again") he 
indicates the following.) It is not the case that in analyzing dhvani 
through the types avivaksita, etc. there was no analysis by reference 
to suggestor factors. That analysis was indeed made by reference to 
suggestor factors. But although made, it will now be made again 
by reference to purely suggestor factors.? Thus, without reference to 
the suggested sense, one can divide suggestors into words, sentences, 
phonemes, word-components, texture (sarighataná), and long sections 
of poetry. But none of these, like the literal meaning, is ever capable 
of being suggested. So the point of his statement is that the varieties 
lof dhvani in the Third Chapter] will be described exclusively on the 
basis of the [verbal] suggestors. 

A certain commentator? bas explained the Vrtti's phrase "through an 
analysis of what is suggested" as follows: "by reference to that which is 
suggested, namely vastu, alarikáro, and rasa.” He should be questioned 
in these terms: [t was not the author of the Karikás who made this 
threefold division; it was the author of the Vrtti who pointed it out 
[cf. 1.4a A]. And it is not the author of the Vrtti who will make the 
[new] divisions [in Chapter Three].* Therefore, what kind of logic would 
it be to say, “be did that and now he is doing this," when the agent 
is not one and the same person?* Moreover, this explanation hardly 
makes a contextual connection with the whole of the book that has 
gone before, because other varieties also, such as avivaksitavácya, have 
already been shown.® But enough argument with an elder member of 
my own family." 

1. Le., the discussion in Chapter Three will be distinguished by its treat- 
ing suggestor words (as opposed to suggestor meanings), which were not 
treated in the previous chapters. — 2. By purely suggestor factors ($uddha- 
vyarijaka) Abhinava means words, phonemes, etc., which are purely sugges- 
tive, never suggested. 3. The reference is to Abhinava's kinsman, the author 
of the Candriké. 4. The division into padaprokdsaté, vékyaprakdsaté, etc., 
is given by the author of the Karikas in 3.1-2 K. 5. Certainly Abhinava here 
speaks as if he supposed the author of the Kdrikds to be a different person 
from the author of the Vriti. But in introducing 3.2 he will take quite the 
opposite point of view. For the question whether Karikdkdra and Vrttikáro 
were the sarne man, see Introduction. pp. 25-27. In view of Abhinava's own 
ambiguous stance in regard to the question, his criticism of the Candrikákára. 
here seems a carping one. Nor is it greatly strengthened by the argument 
that Abhinava subjoins. 6. Abhinava's point is that the explanation given 
by the earlier commentator can, at the very most, demonstrate the logical 

83.14] 371 

connection of the third chapter with only a part of the earlier portion of the 
work, namely the portion in the first chapter where vastu, alarikarn, and rosa 
were mentioned as divisions of dhvani; it will not demonstrate the logical con- 
nection with chapter two, where avivaksitavácya and vivaksitanyaparavdcya 
have been mentioned as divisions of dhvani 7. Kane (HSP p. 198) is prob- 
ably correct in here emending pujyajanasagotrai^ to pirvajanasagotraih, for 
Abhinava in other passages regularly refers to the author of the Candriké as 
purvavamsa, and sogotro is merely a synonym of vamsa. The printed read- 
ing, if we take it to mean “with my own relatives who are persons I should 
respect," is awkward because of the separation of nija from sagotraiA. 

K Either a word or a sentence may serve as the suggestor in 
the type of dhvani where the literal sense is not intended and, of the 
other type, in that sub-type where the suggested sense resembles a 

,A One sub-type of dhvani where the literal sense is not intended 
is that where the literal meaning is entirely set aside. In this variety a 
word may serve as the suggestor, as in this line of the great sage Vyasa: 

these seven are the kindling sticks of royalty.' 
Another example is in a line of Kalidasa: 

When you put on your armor, who could be 
forgetful of his wife pining in his absense?? 

for what serves not as ornament 
to a sweet configuration?? 

In these examples the words “kindling sticks," “armor,” and “sweet” 
have been used for their suggestive qualities. 

372 [83.14 

1. From Vidura's advice to Dhrtatarastra, Mbh. 5.38.35. The verse is 
quoted in full by L below. The samidhoh were the sticks laid out at the 
base of the sacrificial fire, for the choice and arrangement of which elaborate 
directions are given in the ritual texts. The word is here used metaphorically 
of the virtues upon which tbe success of a king must be based. 2. Words 
of the Yaksa to the cloud in Meghodüta, vs. 8. As the monsoon season was 
normally a time for staying at home and enjoying domestic bliss, the sight 
of gathering clouds would naturally remind a traveler of his wife waiting for 
him at home. As a cloud does not actually don armor, one takes the word 
samnoddha in the secondary sense of udyata, prepared as for a campaign, here 
accompanied by wind, thunder and lightning. The suggestion, as L will point 
out, is of a cruel and irresistible opponent. — 3. Sákuntala 1.17. The whole 

The pond lily circled by moss is charming; 
the moon's mark though black, 

gives it a royal beauty; 

and this slender damsel, 

even in her rough dress, is lovely; 

for what serves not as ornament 

to a sweet configuration? 

L Inthe Káriká the word "and" (ca) serves to prevent a sequen- 
tial ordering of the two pairs of terms.' Thus avivaksitavácyo, in both 
its types, is twofold insofar as it can be suggested by either a word or by 
a sentence. The type, namely “suggested gradually" (kramadyotya), of 
vivaksitdnyaparavacya, which latter is other than avivakgitavécye, along 
with its sub-types [namely fabdasoktimüla and arthasoktimüla] is like- 
wise twofold. It is called anurananarüpavyarigya because its suggested 
meaning (vyarigya) is such that its appearance is like the appearance 
of a reverberation. 

The great sage: this harks back to what was said earlier [1.1e A], 
that suggestive poetry “is also found in such works as the Ramayana, 
Mahabharata, and the like." 

Firmness, forbearance, self-control 

purity, pity, kindliness of speech, 

and constant faithfulness to friends: 

these are the seven kindling sticks of royalty.” 

Here the literal meaning of the word kindling sticks is completely set 
aside because it is impossible. What the word “kindling sticks” suggests 

$3.1a A] 373 

is the intention of the speaker, the suggested sense, namely that the 
capacity to strengthen royalty depends on nothing other [than these 
seven factors].? Although the purpose of giving the present example 
has been served by such examples [already given] as “like a mirror 
blinded by breath (cf. 2.1 c A], nonetheless these further examples have 
been given here, where an occasion for them arises, in order to show 
how they pervade a great part of literature. As one may easily supply 
the discarding of the literal sense here from what our. author has just 
said, there is no need for his expressly repeating the fact. 

The word armor, because it is here impossible in the literal sense, 
conveys, by secondary usage, the sense of being prepared, and thus sug- 
gests what the writer intends, namely cruelty, irresistibility, rashness, 
etc. In the same way, the word sweet conveys, by secondary usage, the 
giving of pleasure, satisfaction and the like, to all people, and suggests 
the speaker's intention, namely that it is not surprising that such a 
Shape should be the object of intense desire.* 

1. Witboutca, we should follow the principle laid down by Panini (1.3.10) 
and connect the first member (avivaksitavdcya) of the pair of suggestions 
with the first member (poda) of the pair of suggestors; likewise the second 
member (anurananaripavyarigya) of the suggestions with tbe second member 
(vàkya) of the suggestors. This would be wrong, for either pada or vàkya 
can suggest either avivaksitovdcye or anurananorüpovyorgya — 2. See note 1 
to A above. In pada a our text reads dayá instead of dama, doubtless a 
misprint or a scribal error, for dayd would be nothing more than a synonym 
of kárunya in páda b. 3. [n a footnote to our text Pt. Sri Mabüdeva S&stri 
points out this distinction. The secondary meaning (laksydrtha) of the word 
samidh is capability of production; what is suggested (vyarigyártha) is the 
fact tbat royalty depends on nothing other than these seven factors. 4. In 
his commentary on Sak 1.17 Raghava Bbatta repeats Abbinave's remarks on 
madhura almost word for word. 

A Ofthe same type, [in the variety] where the literal meaning is 
shifted, [a word may serve as the suggestor], as in this: 

Rama, being overmuch in love witb Life, 
has failed, my beloved, to be worthy of his love. 

374 [$3.1a A 

Here the literal sense of the word "Ráma" has been shifted to the 
suggested sense of one who possesses the very quintessence of courage, 

Let others thus compare 

her cheek to the lunar orb; 

yet if they really compare, the moon 

is no more than the poor old moon. 

Here the sense of the second occurrence of the word moon is shifted 
from the literal. 

L Of the same type: viz., the type where tbe literal sense is 
not intended; of this type he refers to the second variety. (The complete 
stanza from which be quotes is as follows:|' 

The cruel demon treated you as one expects 

of such a being angered by rejection: 

and you too bore the blow as a lady of high bi 
should bear it, with your head held high. 

But the witness of your death, who bears 

his weapon now to no avail. 

Rama, being overmuch in love with life, 

has failed, my beloved, to be worthy of his love. 

When [Rávana] was forcibly rejected [by you], since he is cruel by 
his very nature as a demon and through evil pride feels that his or- 
ders are not to be disobeyed, he acted, blind with rage, in conformity 
with his spirit when he cut off your head, for his intention was that 
no one else should ever transgress his command. Treated “you”: the 
sense is, you, by whom even such [a despot] was not heeded. And you 
bore the blow without flinching, open-eyed and with a face as calm 
as on a holiday. "As": in the manner of. "A lady": even a woman 
of low birth who wishes to be called a lady will hold her head bigh 
to give the impression that she is a lady. But you held it up bravely 
at the moment of decapitation, as if to say, *let it be done quickly." 
You held your head high as rea! ladies do because you had always done 
so. Thus both Ravana and you behaved throughout with propriety. 
But the part I played turns out to be most improper. My banishment 
robbed me of opportunities to use my bow, but it might at least have 
proved useful in protecting my wife. Now that I have failed to protect 

$3.1a L] 375 

you in your need, it becomes wholly useless; yet here I still carry it. 
The only reason one can imagine is to protect my own life; and this is 
not proper. "By Ráma": the meaning of this word is developed into 
suggested qualities? [beside the literal meaning of Rama, the son of 
Dasaratha], such as being disposed to unequalled daring, trutbfulness, 
appropriate bebavior, etc. The explanation oí the term "et cetera," 
found in a (certain) commentary, that it stands for cowardice and other 
such qualities, is incorrect. For this behavior, far from being inap- 
propriate, would be proper in the case of a coward. "My beloved": 
this has become a mere word now, for that which justifies the use of 
the word "beloved" is love, and that has been sullied by impropriety. 
Thus, the tragedy of Ráma? is mede clear by the combination of [the 
stháyibháva] grief, the álambanavibháva [namely Sita's death], and the 
uddipanavibháva [namely her noble behavior]. 

Let others thus compare: (After translating the Prakrit stanza 
Abhinava continues:] By "thus" the poet implies that people are nat- 
urally blind to distinctions. "People": tbat is, those whose only course 
is the steps of those who have gone before. "Her cheek": that is, of her 
whose precious figure possesses unique beauties. “Give the moon as a 
comparison to her cheek": [n order to funish a rhetorical comparison 
to her cheek, which is the central and predominant element of a face 
which is the perfection of genuine beauty, one must find some object 
that is of greater beauty; whereas the orb of the moon, being spoiled 
by its spot, is grossly inferior. Thus, altbough ordinary people follow 
in a line like sheep, if discriminating people will consider the matter, 
[they will see tbat] this wretched thing deserving of pity that is called 
the moon is really a thing to which should attach the properties of 
waning, of being without real charm, and of being sullied. For the shift 
of the literal sense to various suggested properties, one may compare 
what I have said previously [viz., in 2.1 a L]. The same appears in what 

1. The verse is taken from some Rama play that has not been identified. 
Apparently it contained a scene in which Ravana produced before Rama an 
illusion of the severed head of Sit, an illusion that elicits from Rama the 
present verse. Such an element of plot would be an easy invention in view 
of the scene in the Ramayana where Ravana produces before Sità an illusion 
of the severed head of Rama. 2. See 2.la A and L. 3. Abhinava writes 
loosely here. The tragic experience (karunarasa) does not belong to Rima 

376 (§31aL 

but to the reader or auditor of the play. What belongs to Rama is the emotion 
grief (Soka). 

A (Coming back to] the sub-type where the literal meaning is 
entirely set aside, [we find that] in this sub-type a sentence (just as well 
as a single word] may serve as the suggestor. Thus: 

In what is night to all creatures 
the true ascetic wakes; 
where others wake, the sage who sees 
sees that it is night. 
[Bhagavadgita 2.69] 

For in this sentence the meanings night and waking are not intended. 
What is communicated is rather the attention of the saint to a knowl- 
edge of truth and his aversion to what is not truth. Thus the suggestive 
force is of the sub-type where the literal meaning is entirely set aside. 

L Having in this manner illustrated the two sub-types of the 
first variety as revealed by single words, he now illustrates them as 
revealed by sentences: "in what is night." Intended: if we take the 
words literally, they furnish no advice for those who are to be advised.! 
What would be the use of saying that one must remain awake during 
the night and that one must act as if it were night the rest of the time? 
Therefore this sentence, its primary sense being obstructed, suggests 
that the ascetic, because of his extraordinary nature, is attentive to the 
preception of truth and averse from false perception. As the word “all” 
is a relative term, the literal statement is logically possible. Accord- 
ingly, it is wrong to suppose.that.the true sense is [not suggested but 
is] implied (aksipta) by the fact that "all" cannot otherwise be logically 

[Now to interpret the verse:| That which is night, that which causes 
utter confusion, to all fourteen classes of living beings from Brahma 
down to plants is the vision of truth. The true ascetic wakes here, 

§ 31b L] 377 

seeking to attain it. This rather than mere avoidance of sensual plea- 
sure is what asceticism really consists in. Or, one may take the words 
differently. That which is night to all creatures is the deception (of 
mdya].? The ascetic wakes here, seeking to avoid it. On the other 
hand, all creatures wake in false perception, that is, they are wide 
awake to [its presentations], whereas to the ascetic it is night, a feld 
of non-awareness, for he is not awake to its activity. Thus he whose 
conduct is defined as extraordinary [i.e., the true ascetic] really sees 
and thinks. Of him alone are the outer and inner organs of knowledge 
profitably employed. Others do not see and do not think. The general 
sense of the verse is that one must be intent on the perception of truth 
In the same way, the words “sage” and “who sees" do not cease operat- 
ing on rendering the literal meanings, but only after giving a suggested 
meaning. As the pronouns yat and tat (represented in the translation 
by “what” and "where"| are dependent on the other words of the sen- 
tence, we may say that the whole sentence, verbs and all, is suggestive. 
Our author states as much in the words For in this sentence, etc. Is 
communicated means is suggested. 

1. It is assumed that every verse in the Bhagavadgité must constitute 
God's advice to those who are in need of it. A verse from the Gita must 
therefore be so interpreted as to furnish such advice. 2. An objector claims 
that the intended sense of the verse is furnished by arthápatti (that which 
one supplies in order to resolve a logical contradiction). Abhinava insists that 
the intended sense is furnished by suggestion. The objector's argument would 
sun as follows. “The literal sense of ‘night’ is ‘a time for sleeping.’ Now there 
is a logical contradiction in sayipg 'the ascetic wakes in the sleeping-time of 
all creatures,’ for ‘all creatures’ is inclusive of ascetics. Thus the so-called 
Suggested sense of waking at night, viz., the pursuit of truth, is a necessary 
implication, not a suggestion." Abhinava rejects the argument by allowing 
“all” to mean “most” or "all other.” Thus the literal sense becomes merely 
inappropriate, not logically impossible. 3. The second explanation is the 
one chosen in Abhinava's Citdbhdsya: yd sarvesám bhütánám niíà, mohini 
máyó, tasyam munir jégarti katham iyam heyeti. With this and with the 
statement just above that the true ascetic is more than one who simply avoids 
sensual pleasure, one may compare Abhinava's comment in his Citábhasya 
on this versé: “yogi ca sarvavyavahárán kurváno 'pi lokottara iti ninipayata 
parume$svarena samksipyásya svarüparn kathyate: “God (ie., Krishna) here 
Shows that the yogin, though he may take part in every sort of worldly activity, 
is of extraordinary nature; and he tells us briefly what this nature is." 


A Of the same type but of the sub-type where the literal sense 
is shifted a sentence may serve as suggestor, as in the following: 
The passing of time is poison to some, 
nectar to others; 
part poison part nect to some, 
neither poison nor nectar to others.' 

For in this sentence the information is conveyed by words which have 
been shifted from their literal sense of poison and nectar to the sense of 
pain and pleasure. Accordingly, the suggestive force is of the sub-type 
where the literal sense is shifted 

1. Source unknown. In our printed text pdda a is two mátrus short. 
Badarinath Sarma corrects it by inserting via after visamaio. In b he reads 
i 2 Sk. vyapacala(ya)ti (cf. Turner. Dict. 12167), which is better than 
valayati, though the latter is barely possible (time "rolls on"). 

L That which is made of poison (visamayitah) means that which 
has come to consist of poison (visamayatàm práptah). (The first group 
of] “some” are those who are wicked or those who have a keen judg- 
ment. For (the second group,] those who are virtuous or those who lack 
judgment, times passes as if it were made of nectar. For some, who are 
of mixed conduct or who are partly of sound, partly of weak judgment, 
[time is] part poison, part nectar. While for those who are complete 
fools or who have reached the final stage of yogic concentration, time 
passes as if it were neither poison or nectar.! This is the construction. 
The words poison and nectar, by a sort of dead metaphor such as one 
sees in the word /Idvanya,” are used in the sense of causes of pain and 

..pleasure, just as we say that a lemon is poison and a wood-apple is 
nectar; but they end up by referring to the pain itself and the pleasure 
that they cause. However, it is not the intention of the sentence that 
they should not at all refer to the cause, for without the cause pleasure 
and pain would not exist. This is what our author means by saying 
that they have been shifted from their literal sense.? (The sense 

$3.1d A] 379 

of the other words in the sentence has been shifted also:|‘ the sense 
of "some" is shifted from the indefnite to the definite [groups which 
the author has in mind|; the sense of "passes" is shifted to the general 
sense of an activity; the sense of "time" to all the elements of worldly 
life. The author of the Vrtti has used the shifting of the single words 
poison and nectar as an example (by which the other shiftings may be 
judged]. That is why he specifies "in this sentence." 

1. Time brings pain to the wicked by bringing them retribution for their 
bad deeds. It is likewise painful to the man of judgment, who sees the essenti 
misery of transmigration. Groups two and three follow naturally from the first. 
Group four is composed of those unaffected by time—those whose ahedony is 
due to stupid insensitivity and those who have overcome the emotions by yoga. 
2. For lávanya see 1.16 K and L. When used as a frozen metaphor to mean 
charm, beauty, it cannot give rise to dhvani. Tbe same holds for the words 
visa (visa) and-amrta (amia) in the present stanza. It is not these words but 
the stanza as a whole that gives rise to dhvani. — 3. So the primary intention 
of the author of the verse is to suggest the pain and pleasure of the world as 
it affects different types of persons. He does this by a secondary use of words, 
but the Lteral sense of the words is not wholly abandoned. 4. Abbinava 
here guards against an objection which might be made to Ánanda's example. 
We might ask how this example differs essentially from the example in 3.18, 
where a single word is shifted. Is the only difference that here two words 
are shifted? It will be seen that by Abhinava's interpretation the present 
example is essentially different from that of 3.1a. The suggestion here comes 
from a shift of meaning of the whole sentence. The way in which anything in 
the world affects persons depends on their karma, their judgment, and their 
practice of yoga. 

A Of that type of dhvani where the literal sense is intended but 
where there is (subordination to] a suggestion resembling a reverber- 
ation, in its sub-type arising from the power of words, a single word 
may act as suggestor, as in the following: 

1f fate will have it that I am not born 

to fill the wants of needy men for riches, 
why was I. being jado, not made to be a well 
or pond of limpid water by the wayside? 

380 [§31d 4 

Here the word jada (witless, insentient, cool), used by the discouraged 
speaker in grammatical agreement with himself, comes to have, as a 
reverberation brought about by its own power, grammatical agreement 
with the well.' 

1. A question may arise how to correlate the present passage with Anan- 
da's statement in 2.21. He said there that we have sabdasaktyudbhavadhvani 
only where an alarikáro is suggested. Where vastumátram (a mere fact or 
situation) is suggested, we have $lesa. In the light of that statement the 
present verse at first sight would seem to exhibit lesa and not any form of 
dhvant. The answer is that the present verse does carry the suggestion of an 
alarkáro. Although neither Ánanda nor Abhinava mentions it here, both of 
them refer to it in 3.33c. The verse suggests a simile (upamá), for the speaker 
is likened to the pond and well. One should note that the term sabdosakty- 
udbhavánuronanopamadÀvani applies here only to the suggestion of simile. 
There is also rasadhvont in the verse, namely, a suggestion of sdntarasa: "As 
my lot is so much worse than that of an insentient pond or well, all worldly 
objects must be regarded as useless and empty." This final suggestion of the 
verse is referred to in 3.33 c (see 3.33¢ A, note 6 and 3.33 c L). It is immediately 
perceived by the sensitive reader and does not arise like a reverberation. 

L Having in this manner given examples of the four kinds [of 
dhvani| referred to in the first half of the Karikd, he now proceeds to 
give, in order, examples of six other varietes covered by the second 
half [of the Karika] by saying: Of that type of dhvani where the 
literal sense is intended, etc. Prátum means "to fill" The plural 
in "riches" is meant to suggest tbe fulfilling of whatever need any par- 
ticular suppliant may have. This is why the word "needy" has been 
used. "Of men":! because generally people want money; they do not 
want help from virtues. "By fate," the decree of which cannot be ques- 
tioned. "I": that is to say, someone else certainly has been created (for 
this purpose], but not I, for which reason the speaker is dejected. A 
pond of limpid water is one that is useful to people. “Or a well”: the 
suggestion is: “even although it is not noticed by people.” Used in 
grammatical agreement with himself: the word jada has the sense” 
“unable to think of what to do.” In the same sense jada can apply to 
a well, for a well lacks knowledge of who needs what. And that is why 
the well is jada, “cool,” i.e., not fevered by distress at the situation 
At the same time the well is jada by its connection with cool water 
(sitalayala)? and so is able to help others. In this third sense the word 

$3.1e L] 381 

jada if applied to the pond would be tautologous.? So he says that it 
comes to have grammatical agreement with the well. By its 
own power: thus he assigns the suggestion to the category of those 
arising from the power of a word.* 

1. Abhina«a's point in this remark is that the word jana (men in general, 
ordinary people) has been used because the common run of men want just 
wealth. It is only a few unusual persons who may want moral or spiritual 
aid, which the speaker might be able to give. — 2. [n interpreting puns and 
suggestions, d and ! are regarded as interchangeable. So jada can be the same 
as jala (water). 3. Abhinava here seeks to explain why the Vrtti speaks of 
jada's attaining grammatical agreement with the well rather than the pond. 
1t is because this third meaning which he has discovered of jada (possessing 
cool water) would be tautologous with prasannámbudhara (containing limpid 
water) already used of the pond. — 4. More literally, thus he joins (to the 
suggestion] the property of arising from the power of a word. 

A In the same sub-type of dhvani a sentence may act as a sug- 
gestor, as in the speech of Simhanàda in the Harsacarita: 
In this great disaster you are now all that is left — — 
In this cosmic destruction you are now the world-serpent Sesa 
for the support of the earth.! 
[Harsacarita, p. 192 bottom 
(Chapter 6, lines 421-2 out of 628) 

for this sentence clearly suggests by the power of its words a second 
sense that comes like a reverberation. 

1. Simhanáda, the feld marshal, is addressing Harsavardhana after the 
deaths of the latter's father and elder brother. 

L The word maháprulaya in its first sense is to be analyzed as 
maha-à-pralaya, the complete cessation of happiness. When this cause 
of grief has occurred, vou are the only one remaining for the support, 

382 [$31eL 

the consolation, of the dharani. that is, the yoke-beam, of empire.’ 
When the sense of this sentence has been thus completed, a second 
sense ensues, namely that after the elephants of the quarters, etc., 
have perished, the king of serpents alone is able to support the weight 
of the earth.? 

1. Abhinava tries to elicit a pun from every word in the sentence. Surely 
no reader would understand this "direct meaning" at a moment noticeably 
previous to the “suggested meaning.” See 221e A, note l. 2. One might 
add that the two senses in conjunction suggest asi ile. vi2., that King Harsa 

il in qualities to the world serpent. 

A In the same type of dhvani [viz., where the suggested sense 
appears like a reverberation] but in the sub-type where the suggestion 
arises from the power of a meaning furnished by a poet's imaginative 
expression [cf. 2.24], a word may be the suggestor, as in the Harivijaya.' 

The face of early spring is decked with mango buds 
and scented with the rich sweet wine 

that soon will flow. 

The god of flower arrows 

snatches from her a kiss without consent.? 

Where it is here said that the god of love snatches a kiss from the face of 
spring without her consent, the word asamarpitam (lit., “unoffered”), 
which denotes the state [of early spring], suggests by the power of its 
meaning [since it implies that Love seized Spring without her consent] 
the violence of the god.? 

1. For this lost Prakrit poem by the royal poet Sarvasena, see V. Ragha- 
van, Bhoja, pp. 810-11. It described Krishna's victory over Indra and his car- 
rying off of the porijáta tree as a present for Satyabhama. A. K. Warder iden- 
tifies the author with the fourth-century Vakataka king of that name (Indian 
Kavya Literature, Vol. 3, p. 59). 2. The correct reading of the second pada 
is that given by Hernacandra: chanapasaramahagghamanaharasurdmoam. For 
Prakrit manahara = Sanskrit manohara, see Pischel, para. 347. Dhvanyáloka 

$3.1g A] 383 

KM edition (followed by Badarin&th Sarma) emends the verse to bring it into 
strict giti meter, but this is wrong. All the gáthàs that are preserved of the 
Harivijova appear to have twenty mátrás in the second pàda; see Sarasvati- 
kanthábharana 4.235; 5.287, 330, 350, 351. It will be seen that I interpret 
the verse diHerently from Abhinava. His interpretation seems to me impos- 
sible 3. Hemacandra (on AC 1.74) makes the suggestion more specific: 
“when spring bas come fully of age, what will he not do!” 

L The beginning of spring, or the face of spring, in which there is 
delight (Gmoda = camatkdra) on the part of the god (sura) Manohara 
(= the god of love Manmathadeva) because of the precious (mahdr- 
ghena) festival-influence (utsavaprasarena). Here the attributive “pre 
cious” is placed after its noun in the compound because there is no 
rule of order in Prakrit.! Chana means festival. Mukham means both 
beginning and face: and this is joined to surámodam.? The basic mean- 
ing is tbat in the beginning of spring love stirs our hearts. But this 
becomes suggestive of a further sense because of the striking expression 
of the poet. 

1. But this is not true. If it were, the Prakrit language would be unintel- 
ligible. Furthermore, in Abhinava's interpretation prasara makes very little 
sense. 2. Abhinava's interpretation amounts to this. At the very beginning 
of spring comes the Love-Festival (Manmathotsava). Under its influence the 
love god in our hearts takes delight in the vernal beauty and seizes it even 
in its childhood. This suggests how violent our love will become when spring 
is fully grown. Abhinava does not note the obvious meaning of surdmodam 
(scented with wine). 

A Within the same subtype a sentence may act as the sug- 
gestor, as in the verse already quoted “The fragrant month prepares,” 
etc., [see 2.24 A]. There the meaning of the sentence, viz., that the 
fragrant month prepares but does not yet give the arrows to the god of 
love, being embodied in an imaginative expression of the poet, suggests 
spring's destructive stage of stirring up love [which is about to come]. 

(§3.1g L 

L Our author has not given an example either of a word or 
a sentence as suggestor of the sub-type of suggestion arising from a 
meaning embodied in an imaginative expression of a character invented 
by the poet [cf. 2.24 A]. This is because what he has given is sufficient 
to illustrate the specific words of Káriká 2.24: "it may be given body 
simply by an imaginative expression." An example of a word acting as 
suggestor in that sub-type would be: 

Truly fair women are obj 
and truly wealth is fair; 
but life is unsteady and as quickly gone 
as the glance of a tipsy girl. 

In this verse there speaks a disenchanted (virdgin) character created by 
the poet, suggesting by force of the meaning embodied in the word “life” 
the following. All these sensual pleasures and riches are of use only to 
one's life. When one's life is gone, even though they still exist, they 
come to be as if non-existent. And life, which consists in preserving the 
vital breaths,? is nothing to rely on, because the functioning of the vital 
breaths is so tenuous. So what is the point in maliciously proclaiming 
the faults of the poor objects of sense? (Rather,| one's own life is to 
be blamed. But since life is by its very nature so unsteady, even it 
is not at fault. All this leads to a thorough disenchantment with the 

An example [of suggestion of this sub-type] produced by a sentence 
is the vese "On what mountai ,” etc. [see 1.13m A]. 

1. Anandavardhana will quote this verse under 3.30. It occurs in many 
C&nakya collections, as well as in SRK 1608, and is quoted by Hemacandra. 
Ksemendra, and others. For particulars see Sternbach. Cánakya Niti Tert 
Tradition, Vol. Il, p. 231, item H. 2. This is the deGnition of DAP: jiva 


A [In the same type of dhvani] in the sub-type where the sugges- 
tion springs from a meaning that is embodied in something inherently 
possible, a word may be the suggestor, as in: 

Ab merchant, how should we have ivory 

or tiger skins for sale, 

when daughter-in-law is strolling about the house 
with the curls dancing on her forehead?! 

Here the word lulitdlakamukhi ("with curls dancing on her forehead”), 
by the power of the situation, inherently possible, to which it refers, 
suggests the young wife's eagerness for sexual play and her husband's 
weakness from his constant enjoyment of that pleasure. 

1, The verse is again quoted by Mammata (10, vs. 528, p. 709) and is 
included in Weber's Sottosoi (951). Parisakkae: DhP lists the verb svoskate 
(DAP 1.100) in the sense of motion; it bas been found only in Prakrit. Lulità- 
lakamukAi: lit., whose face bears tossing curls; but as alaka regularly refers 
to the curls in what we call the bangs, I have translated "on her forehead." 
The picture is not of a slattern who hasn't time to fix her hair, but of a young 
woman who gives careful attention to arranging her hair so that it will be 
attractive to her husband. 

L |Abhinava translates the stanza, then adds:| Parisakkae: walks 
about flirtatiously. There is no difficulty here in applying the literal 
sense of the words.’ The adjective "[with curls] dancing [on her fore- 
head]” is a simple description of [the wife's] appearance, while the hus- 
band's failure to procure ivory might be due simply to his arrogance.” 

1. If the literal sense of the words were illogical, it might be argued that 
the suggested sense was forced on us (dksipta) as a necessary inference rather 
than being suggested (vyarigya). That is not the case. 2. BP: he might feel 
that he had made enough money already. 


A In the same [sub-type] a sentence may act as a suggestor, as 

The hunter's wife strolls proudly 
with peacock feather behind her ear. 
She strolls amid fellow wives 
who deck themselves with pearls. 
[Sattasai 2.73)! 

This verse suggests the good fortune in love of a certain hunter’s wife, 
newly wed, who wears a peacock feather behind her ear. For it is 
suggested from the meaning [i.e., from the given situation] that her 
husband, wholly intent on enjoying her charms, is now able to kill 
only peacocks; while the ill fortune of the other wives, who have been 
married a long time, is revealed in their decking themselves with pearls, 
for it is suggested from the situation that the same husband had time, 
when he was enjoying them, to slay elephants. 

L “The hunter's wife,” etc. The verse has been quoted i 

^A Now the following objection may be raised. “You have claimed 
that dhvani is a type of poetry [cf. 1.13 KJ. How then can there be a 
revelation of it in a single word? For a type of poetry is a particular 
collection of words that causes us to apprehend a particular meaning. 
Its nature is such that a single word cannot reveal it, for the individ- 
ual words remind us of objects; they do not denote."! To which we 

$31jL] 387 

reply: There might well be a fault here if it were the denotative power 
that occasioned our use of the term dhvani, but it is not so; it is by 
the suggestive power that we assign the term.? Furthermore, poems, 
like human bodies, are collections defined by a particular arrangement 
of parts. The idea that we form of their beauty can be assigned, by 
positive and negative agreement, to particular parts of the collection 
Accordingly, there is no contradition in our assigning of dhvani to indi- 
vidual words insofar as they are suggestive [of this kind of beauty in the 
whole poem]. The following slokas will give support | to our position]. 

"Just as the sound of what is unpleasant makes a verse faulty, as is 
manifested in the faults srutidusta, etc.; just so does the reminding us 
of what is pleasant constitute a virtue. 

"Therefore there is beauty in all those varieties of dhvani which ap- 
pear in single words, even though a single word serves only as a re- 

"Just as a woman appears beautiful by a single ornament which im- 
parts to her some special attraction, so the speech of a good poet is 
beautiful by means of the dhvani revealed by a single word." 

1. The objection must come from a Prabhakara Mimamsaka. In Prabhá- 
kara's theory vácakalvo, conceived as the power of transmitting valid infor- 
mation, must produce a new cognition. The individual words simply remind 
us of objects which we already know. — 2. Abhinava will point out that this 
answer is a chala, a trick in which the opponent's meaning is intentionally 
misunderstood (cf. Nydya S. 1.2.10-17). By vdcakatua the Prábbákara means 

"the power to convey a specific piece of new information. whether fact or sug- 
gestion. Ananda takes it in the more limited sense which it bears in his own 
system, viz., as the power to denote, as opposed to the power to suggest, a 

L Now, etc.: The objection arises from the view that dhvani isa 
collection.' Its nature: that is, the nature of a type of poetry. He first 
takes the phrase used by the objector "because words do not denote 
(avácakatvát)" and by intentional misunderstanding’ shows that this is 
an insufficient reason for the conclusion. Thus he says, There might 
well be a fault here, etc. After answering the objection by this trick, 
he then answers in all seriousness with Furthermore, etc. Suppose the 
objector were to reply, "I am not taking the word's lack of denotative 
power as a cause of its lack of dhvani. I am saying that dhvani is a 
poem. And a poem is a sentence which conveys a complete meaning; it 

388 [$331jL 

is not a single word." Our author might continue, "True; but neither 
have we said that dhvani is a word. Dhvani is a collection. That is 
why the word 'reveal' (prakása) may be used in speaking of it: 'dhvani 
is revealed by a single word.'" Now the objector may ask how it is, if 
a single word has such a capability, that the process of understanding 
[suggestive poetry] is an unbroken whole. It is with this in mind that 
our author says: poems [like human bodies], etc. For it bas been 
said before: "the teaching that (a sentence] has parts applies only at 
the time of analysis.”* 

But how can we shift the area in which we apprehend beauty to the 
parts (of the sentence, which are its words], for the individual words do 
not express the sentence meaning but merely bring objects to mind? 
Well, why not? Why should words not be causes of our apprehending 
beauty, since they remind us of beautiful suggestions? To take a con- 
trary instance: a word such as pelava “delicate” is not denotative of 
any obscene sense such as pela “testicle.” but merely brings that sense 
to mind; and on tbat account a poem, as being a thing of beauty, is 
“spoiled by its sound (srutidusta)." This fault of being $rutidusta is 
assigned by positive and negative agreement to the parts of a poem 
The same should be the case with [the causes of beauty] which are 
under discussion. He puts the matter thus: Just as the sound of 
what is unpleasant, etc. The sense is, the sound of what reminds 
us of that which is unpleasant. Makes a verse faulty: tbat is, un- 
beautiful. Virtue: as much as to say beauty. Having thus given a 
[counter-Jexample in the first three quarters of the sloka, he gives that 
which the example illustrates in the fourth. Then he sums up: There- 
fore there is beauty, etc. The syntax [of his argument) is this. As 
the memory of what is pleasant constitutes a beauty, therefore in all va- 
rieties of dhvani that have been described, even in that which appears 
in single words, even in that which is revealed by a single word, there 
is beauty, even though words function only as reminders. The word 
api ("even though") is construed in both directions on the principle 
of a crow's eye. Finally he shows by positive and negative agreement 
how our apprehension of beauty arises from a single word: Just as a 
woman, etc. 

1. That is, the view is based on the fifth of the five meanings assigned 
to dhvani by Abhinava (1.131). From the points of view represented by the 
other four definitions the objection could not arise. — 2. See 3.1j A note 2 
3. What follows is a pratyavasthána (see NyáyaSBhásya 1.2.12), a correction, 

$3.2L] 389 

of Ananda’s misinterpretation. 4. Cf. Vákyapadiya 1.90. — S. In the second 
parikarasloka the word api is to be taken both with smarakatve and with pada- 
mátrávabhásinah. A crow is popularly supposed to have only one eye. which 
he shakes from one side of his head to the other when he would change the 
direction of his gaze. Hence the term kdkdksinydya for the construction àsà 

K But [that variety of] dhuani where the passage from the literal 
to the suggested meaning is imperceptible shines forth in phonemes, 
wotds, etc., as well as in a sentence, in texture. and in a complete 

L Having commented thus on the [first] Káriká, he now proceeds 
to set forth in detail that variety [of dhvani,| which was not included 
there, namely, where the passage from the literal to the suggested mean- 
ing is imperceptible:' But [that variety], etc. The word “but” serves 
to contrást this variety with those treated above. A word is a collec- 
tion of phonemes. A sentence is a collection of words. Texture is a 
‘property both of words and sentences. A complete work is a collection 
of connected sentences. [t is with this in mind that the author of the 
Karikd has listed phonemes, etc., in the order here given. The term 
“etc.” refers to parts of a word, groups of two words, etc. The locative 
in varnapadddisu is the locative of cause.? The term “shines forth,” as 
it implies the illuminating of the entire poem, confirms the nature of 
[dhvani to be] a type of poetry, as [we have seen] before. 

1. Note that in this sentence Abhinava writes as if the author of the Vrtti 
were the same person as the author of the Kárikás. Contrast 3.1 Intr. L, 
note 5 2. Wherever possible Abhinava wants to understand -dhvant as 
"suggestive poetry," rather than as "suggestion." He is able to do so here by 
taking the locative according to Varttika 6 on P&n. 2.3.36 (carmant dvipinam 
hanti: one kills a leopard for its skin). Instead of meaning “dhvant (i.e., 
suggestion) shines forth in phonemes, words, etc." the Kdrikd will mean 
by Abhinava's interpretation, "dhvan: (ie., suggestive poetry) shines forth 

390 [8321 

because of (skilfully employed] phonemes, words, etc.” The locative in vákye, 
however, he will take differently; see 3.4¢ L. 

A Lest one should hesitate in accepting this statement, feeling 
that phonemes are meaningless and therefore cannot suggest anything, 
the following is stated. 

K The phonemes s, s, dh, and conjunct phonemes containing 
r, when used to excess, are hindrances in [the rasa of| love. These 
phonemes do not produce rasa [of that variety]. 
These same phonemes, when used in the rasa of loathing, [cru- 
elty,] etc., illumine their goal. Hence they do produce rasa [in those 

1. In Kárikà 3 one must read te na separately; in 4 one must read tena 
as a single word. Rasacyut can only mean "dripping rasa"; cf. the similar 
compounds madocyut, madhucyut, etc. In other words. Abhinava's first ex- 
planation (see below) is the only one that can be justified by idiom. The 
varieties of harshness remarked on in these Kárikás pass unnnoticed in so 
generally harsh a language as English, but they have often been noticed in 
more musical languages. For the effort to avoid sibilants in Latin, see Quintil- 
jan 9.4.37-38, for other harsh conjuncts ibid 11.3.35. The letter r was called 
by Persius (1.109) canina littera (the dog's letter) fram its suggestion of a 
dog's growl. Both Latin and Sanskrit had a sharply trilled r. 

A In this pair of slokas the suggestive power of phonemes has 
been shown by positive and negative [precept]. 

§ 3.3-4 L) 

L (Commentary on K:| When used to excess is to be con- 
strued with each [of the phonemes]. Thus one should explain as “where 
$ is used to excess, [or where 3 is used to excess," etc]. Conjunct 
phonemes in which r is predominant: e.g., kr, rhr, rdr! Are hin- 
derances: the harsh alliteration? is opposed to srrigára because these 
(te) phonemes when used to excess do not (na) emit, that is, do not 
let the rasa flow. Or, [we may interpret as follows:} therefore (tena), 
viz., because they are opposed to srrigára. the phonemes $, s. etc.. fall 
off from rasa (rasác cyavante), that is, do not suggest rasa. This is the 
negative precept. These same [phonemes]: viz., $, etc. Their goal, 
viz., the rasa of loathing (bibhatsá), etc. Illumine: reveal or suggest. 

[Commentary on A:] He explains the overall sense of the two Karikas: 
In this pair of, Slokas (slokadvyena). His avoidance of the dual form 
(slokábhyám) is to prevent our.taking the expression by the principle of 
sequential ordering,? for the negative precept is given by the first verse 
and the positive by the second. 

The net result of the teaching here given is that a man who seeks to 
be a good poet should not use the phonemes 5, 5. etc.. in that which 
is characterized as $rrigára; and it is because of this negative result of 
the teaching that the author of the Kdrikds places the negative precept 
first. The positive precept comes after in the form of a qualification 
that this usage is not always to be avoided but is permissible in such 
rasas as bibhatsd. The author of the Vrtti, on the other hand, places 
the positive first in order to observe the custom of placing the word 
anvaya before the word vyatireka.* 

The following is meant. Although the cause of aesthetic pleasure 
(rasa) is the combined apprehension of the vibhávas, anubhávas, and 
vyabhicáribhávas,? it is self-evident that the vibhávas, etc., are conveyed 
by words of a given phonetic shape. Therefore even the particular 
character of phonemes, as soft, harsh, etc., which is grasped by the ear 
regardless of whether the meaning has been noticed at the time when 
they are heard, is helpful to the relishing of rasa. It is on this account, 
namely in order to convey the fact that phonemes are helpful, that the 
locative of cause was used (in 3.2 K] in the expression varnapadádisu. It 
is not that rasa is suggested solely by phonemes, for we have said many 
times that aesthetic pleasure arises from a combination of the vibhávas, 
etc. But phonemes have a nature of their own, grasped only by the ear, 
which does take part in producing the flow of aesthetic pleasure. They 

392 [§3.3-4 L 

are similar in this respect to the sounds of a song without words, or to 
the various notes (játi), rhythms (karana) and ghra, etc.® of a drum, 
guitar, or the like. 

1. By Abhinava’s addition of the word pradhána, it would seem that he 
interprets the verse as a warning against conjuncts that contain a predomi- 
nance of r's, i.e., more than one r. [f so, we should expect rkr instead of kr in 
the first example. But the combination rkr is not likely to occur in Sanskrit. 
2. Parusá urttth “harsh alliteration” is defined by Udbhata, 1.4 [nduràja, 1.6 
Vivrü, as containing these phonemes. — 3. [f Ananda had written slokábhyám 
anvayavyatirekábhyám ... darsitam, we would naturally take his meaning to 
be "it has been shown by these two verses, by the positive statement (of 
the first) and by the negative statement (of the second)." — 4. The order 
anvaya-vyatireka is made obligatory by Pan. 2.2.33. — S. See above, 1.18 L 
and note 20; also Introduction, p. 16. 6. Our text reads ghrádi; others read 
ghádi. We are ignorant of the meaning. 

A In the variety of dhvani where the passage from the literal 
to the suggested meaning is imperceptible, suggestiveness may also be 
present in a word, as in the following: 

You were trembling; in your fear 
the robe was slipping from your shoulders; 
and you cast those eyes 
helpless in all directions. 
But the cruel fire, pitiless, 
burned on with swift att 
blinded by its smoke, 
it destroyed you without seeing. 
Mátrarája, Tápasavatsarájacanita 2.16' 

For in this verse it will be clear to sensitive readers that the word 
"those" (te) is full of rasa. 

$3.4a L] 393 

1. The verse gives the words of King Udayana, who wrongly believes his 
Queen Vásavadattá to have perished in a palace fire. The situation has been 
dramatized in many plays but this particular verse is from the play of M&tra- 
raja. [n commenting on this verse as quoted by Mammata (7, ex. 187) N&gesa 
Bhatta and others have erroneously ascribed it to the Ratndvali. Some MSS 
follow this verse by another verse exemplifying the same suggestive use of the 
word te. The extra verse, which is not commented on by L, begins jhagiti 
kanakacitre and is given by our text in a footnote to page 304. [ts translation 
is as follows. 

The moment the golden deer appeared, 

my beloved's eyes blossomed and sent forth those glances, 
like blue waterlily petals ruffled in a breeze, 

which as I remember them still burn my heart. 

The words will be those of Rama remembering Sita. 

L May also be present in a word: that is, may be present 
also: when a word acts as a cause! of the suggested sense. Thus the 
intention is as follows. Aesthetic pleasure comes [strictly speaking] only 
from the vibhdves, etc. But when these vibhàvas, etc., being conveyed 
by some particular word, bring about a specially delightful relish (rasa- 
camatkéra), we ascribe the power [of suggestion] to the word alone. 

For in this verse: this is the lament of Vatsaraja, in whom heavy 
sorrow is aroused by his hearing of the burning of Vasavadatta. And 
inasmuch as the sorrow has here arisen from the destruction of a beloved 
person, such gestures of that person as the motions of her eyebrows 
or her sidelong glances, which formerly made her an object of sexual 
desire—these very gestures as now recurring to his memory—give rise 
to tragedy (karuna), in which the sense of loss is absolute.” This much 
is clear. Now. in the phrase “those eyes,” the word “those” (te) serves 
as the special cause of the tragic rasa by suggesting various memory- 
pictures of the qualities that her eyes possessed. qualities which are 
indescribable and which can only be felt by the speaker. 

Thus, what a certain [commentator] has objected to and answered 
are both false. He objected that the word “those” cannot possess this 
power, since it must refer to something previously mentioned.) His 
answer was that the speaker was under the influence of rasa [when he 
spoke].‘ Neither the objection nor the answer should ever have been 
raised.5 Where the relative yad has shown that a thing possesses à 
property that mav occur together with some second property still to 

394 [§34aL 

be mentioned, there the word fad then shows the copresence [of that 
second property] with the first property that we still bear in mind.® So 
the rule that “yad and tad must go together” refers to the anaphoric 
usage of tad. On the other hand, where the word tad is used to suggest 
a particular memory-picture induced by some cause,” as in the phrase 
“that pot” [meaning the pot which I remember as having seen before], 
in such and similar expressions, the word tad has no reference whatever 
to anything mentioned previously. So enough of arguing with persons 
who think themselves learned but whose references are wrong. 

By the words “trembling,” etc., the speaker of the stanza imagines 
the symptoms (anubhávas) of the queen's fear. The thought that he has 
been unable to prevent [that fear] is a stimulus (vibháva) of the sorrow 
which fills him. (He speaks of] casting “those” eyes, that is, eyes which, 
although they were always the unique abode of beauty-in-motion, were 
now helpless, finding no goal of sight in their terror and as if asking 
"Who will save me? Where is my husband?" That "those" eyes of 
hers should have been reduced to such a condition acts as a stimulant 
(uddipana) of the speaker's sorrow in an exceptional degree. "Cruel": 
such is the very nature (of fire] and cannot be helped. And yet, the fire 
was blinded by smoke and so was unable to see (the queen], for it is 
inconceivable that an informed agent should do such an improper deed. 
Thus the memory of the beauty of her eyes now acts as a stimulus of the 
sorrow which overwhelms the king. All this development of meaning 
is achieved by the presence of the word "those." In this manner [the 
suggestive power of particular words] should be explained in the case 
of other examples 

1. Abhinava is taking the locative word pade (Text, p. 304, line 1) as 
ni ittasaptami rather than as expressing place where. See above, 3.2 L, 
note 1. 2. BANS distinguishes tragedy (karunarosa) from love in separa- 
tion (vipralambhasrrigára) by the fact that the emotion (bhdva) is absolute, 
unqualified (nirapeksa), whereas in vipralambhasrrigára the emotion is rela- 
tive; its object is merely removed, not destroyed. — 3. The commentator who 
is being criticized (presumably the Candrikakara) based his criticism on the 
anaphoric use of the pronoun tad (its use as "picking up" the relative yad). 
4.. And hence apparently unable to remember grammatical requirements. The 
remark is historically interesting, as it shows that the Candrik&-kàra still held 
to the old conception of rasa as simply a heightened form of bhava. See Intro- 
duction, p. 18 and footnote 29. — 5. For the expression anutthdnopahata see 
24 L., note 40. 6. The passage becomes more lucid if we read antiddeksya- 
máno, as BP suggests, in place of anuddisyaména. In the sentence yo vidván 

$34b A] 395 

sa püjyoh. "he who is learned is worthy of honor," the relative pronoun yo 
shows the man's possession of the property vidvattva to be combinable with 
some second property. This second property turns out to be püjyatva. The 
anaphoric pronoun sa shows püjyatva to be something copresent with the 
vidvattva that we still bear in mind. See also 3.161 L, note 3. 7. The 
appearance of recollections is always due to some cause, such as the experi- 
ence of something similar to or something in some way connected with the 

A The suggestion may arise through a part of a word (e.g. 
througb a single component of a compound word], as in the follow- 

Her face was bowed in shyness 
in the presence of our elders 
and she forced back the grief 
that gave motion to her breast. 
But did not the mere corner of her eye, 
lovelier than a startled deer's, 
somehow, as it dropped a tear, 
tell me not to go?! 

Here the component tribhdga ("corner")? (in the compound netratn- 
bhàga ("eye-corner") is suggestive]. 

1. The verse is ascribed by Sàrrig. 3464 to "Eye-corner" Brahmayasasvin, 
as though the poet had taken his sobriquet from this verse. In SuktiM. 43.21 
the verse is given as anonymous (kasyápi). One may supply a context in which 
a husband tells his friend of the difficulty in taking leave of his young bride 
to go on some journey. 2. The literal meaning of tribhága is “a third.” A 
"third of the eye," as Jacobi notes in his translation of the present passage;- 
implies the pupil of the eye. But what is precisely meant is the pupil in a 
position at the corner of the eye, as in a sidelong glance. It is this meaning 
that gives suggestiveness to the element in the present verse. If so small a 
fraction of a remembered trait could tell the speaker of his bride's love, how 
great must be his pain in separation from her. 


L The component "corner": The speaker remembers how she 
looked at him, despite the presence of their elders, with a sweet glance 
that contained yearning, grief, and despair.’ The recollection serves as 
a stimulant of the grief of separation, caused by a journey, of persons 
who cannot live without each other.? This stimulation is made clear by 
the presence of the word element “corner.” 

1. The reading of the Kashi edition -garvamantharam makes very lit- 
tle sense. We have preferred the KM reading (also accepted by Badarinath 
Sarma) -garbhamadhuram, and have so translated. — 2. Parasparahetuka- 
tvaprána seems to mean that each one is the cause of the sustenance of the 
other's life, i.e., if one should die, so would the other. 

A Where the passage to the suggested meaning is imperceptible, 
dhvani having the form of a sentence is of two sorts, being either pure, 
or mixed with a figure of speech. Of these, the pure type is exemplified 
in the verse from the Rémabhyudaya “though with feigned anger.”' For 
the sentence taken as a whole shows how the love [of Rama and Sita] 
for each other has reached full bloom and so reveals the perfect essence 
of rosa. 

1. The complete verse is given below by Abhinava. The lost play Ràmá- 
bhyudayo was writen by Yasovarman, the eighth-century king of Kanauj and 
patron of Vakpatiraja and Bhavabhüti. 

L Having the form of a sentence: the term is expressed in 
the nominative in order to show that the sentence and the dhvani are. 
coextensive.' For while the suggested meaning appears when phonemes, 
words, or components are present [as special causes], that meaning 
appears [over a greater area than theirs] as running throughout the 
whole verse, for it takes its life from the combination of vibhávas, [anu- 
bhávas, and vyabhicáribhávas|. Thus it is that phonemes, etc., are 

83.4c L) 397 

merely subsi iary causes of dhvani where the passage from literal to 
suggested is imperceptible, but the sentence is not a subsidiary cause, 
merely helpful like the phonemes, etc., but is engaged in conveying the 
whole complex of vibhávas, etc., and so appears as wholly made up of 
rasa or the like.? Accordingly, where 3.2 K says [that the suggested 
meanng may] "appear in a sentence," the word vákye ("in a sentence") 
is not to be interpreted as a mere locative of cause but rather as a 
locative [of place] having the sense that this type of dhvani can occur 
in no other area. 

The pure type, that is, unmixed with any figure of speech (is as 

Although with feigned anger, 
with tears and with despairing glances, 
my mother sought to hold you back, 
you followed me in exile out of love, 
who now, without you, gaze upon 
the horizon black with its new clouds: 
how hard this shows your lovet's heart. 
to be, my love, that he still lives! 
Yasovarman, Rámábhyudaya 

Her following him despite his mother's seeking to hold her back in these 
various ways shows that she disobeyed the command of a parent out of 
the depth of her love. The collocation of *your lover" and "my love" 
expresses the basic emotion (sthdyibhdva) of love where each of the 
lovers is the very life of the other. "New clouds" shows that Ráma is 
gazing at the clouds of the monsoon season which he has never before 
endured [in the absence of Sité] and so expresses a stimulant (uddi- 
panavibháva) of love-in-separation. In the phrase "still lives" (jivaty 
eva) the particle “still” (eva) by its expressing a qualification prevents 
the appearance of tragedy.? 

Taken as a whole: the sense is that no one word reveals tbe rosa 
more than another. Essence of rasa: the essence, that is, of love-in- 
separation (vipralambhasrrigàra). 

1. If the Vrtti had said vákye dhvanih "suggestion occ ring in a sen- 
tence," one might take the sentence to be merely a special cause of the sug- 
gestion, just as "in a word" and "in a component part" were taken in that 
sense (nimittasaptomi) by Abhinava in the foregoing comment on Káriká 3.2. 
2. *Or the like" refers to rasábháso, bháva, bhávábhása. — 3. If the verse 
Pictures Rama and Sità as never to meet again, its effect will be tragedy 

398 [§3.4eL 

(karuna). If it is felt that they will meet again, its effect will be the sad 
variety of the erotic (vipralambhasrrigóra); see 3.4a L, note 2. Abhinava here 
argues that the statement that Rama still lives implies that the lovers will 
meet again; for otherwise Rama surely would have died. One may of course 
challenge his interpretation. It is true that Rama will meet Sita again. But 
Rama, as pictured in this verse, does not know this; only the audience knows. 
And Abhinava has said (p. 107 = Text p. 79, line 1) that paurvdparyavimarsa 
is not relevant to the immediate aesthetic impact on the reader. That is why, 
as he explains. Rávana's love for Sita is a case of rrigàra. not of Adsa, although 
ultimately it becomes hása as the love is not shared. By analogy a case might 
well be made for assigning the present verse to tragedy. 

A The type that is mixed with a figure of speech may be ex- 
emplified by such a verse as "carried together by the flooding river of 
passion."' for in this verse the rosa, [viz., love-in-separation] is strongly 
manifested and is adorned with metaphor following the rules laid down 
above [viz., in 2.18| concerning the suggestor. 

1. The verse, from the Amarusataka. is given in full by Abhinava below. 

We have seen lovers carried together 

by the flooding river of passion, 

who find the flood to be blocked 

by a dam in the form of their parents. 

When forced. with desire unfilled, 

to stand frozen as in a painting. 

they still drink of each other's love 

through the lily stems of their eyes 

Narasimha (SubhA. 2057 = Amaru 104) 

With metaphor: passion is the flood of a new river, that is, a mon- 
soon freshet. because it has swollen up suddenly. “Carried together” by 
this, that is, brought face to face without having so planned it. There- 
upon their parents (guravah). mother-in-law and the rest, act as dams 

§ 3.5 Introduction A } 399 

by blocking the flood of their desire.! (There is also a pun here, for] the 
dams are heavy (guravah), that is, impassable. The will [of the lovers] 
is thereby blocked and so they stand "with desire unfilled." However, 
interchanging their persons as they face each other, with limbs as it 
were painted because devoid of all motion in their bodies, they pass 
their time in the strategem of mixing slender glances of mutual long- 
ipg, tasting the relish of each other's longing which is brought to them 
by the lily stems which are their eyes. 

Now it may be noticed that the metaphor is not made complete, for 
the lovers have not been identified with a pair of wild geese or cakravéka 
birds or the like, for sucb birds are accustomed to play at drinking water 
from a single lily stem." That is why our author speaks of the rules 
laid down above. For it was said above "the iptention must be keep 
them [viz., the figures of speech] subordinate and [they] should 
never be oversustained" (2.18 K). “Adorned”: by the ornamentation of 
the vibhávas the rasa is also adorned. 

1. In India it is considered unseemly for young married couples to kiss 
and fondle one another in the presence of their parents and in-laws. Thus 
the verse does not imply, as an English reader might at first take it, any 
opposition on the part of the parents to the relationship between the young 
people. The obstruction is merely to overt gestures. — 2. Sanskrit poetry is 
full of references to the monogamous affection of shelldrakes (cakraudkas) and 
wild geese (hamsas). A common picture is of a pair of such birds nibbling at 
the two ends of a nalini, the long stem of the water lily which descends below 
the surface of a pond. Now the eyes of the lovers have the shape and dark 
color of water lily buds. Their mutual glance is likened to a lily stem. They 
drink rosa just as the birds drink the water contained in the stem. Thus the 
poet could have completed, or elaborated, his metaphor by likening the lovers 
to such birds. But he did not. 

A It has been said [in 3.2 K) that the variety of suggestive po 
etry where the passage from the literal to the suggested meaning is 
imperceptible may shine forth in texture (sarighataná). So it is here 
necessary first to define the nature of texture. 

(535 K 

K Texture has been said to be of three sorts: lacking in com- 
pounds, having compounds of medium length, and having long com- 

A “Has been said," that is, by cetain (critics].! After simply 
inding the reader (of this definition]? the following is stated. 

1. Presumably, then, Káriká 3.5 is a quotation, but it is not known from 
whom. The term sarighataná is not used by the early critics, who use racaná 
or riti instead. On the evidence of Abhinava's comment on 3.6 it appears that 
Udbhata used the term. He may have been the first to use it and the present 
Káàniká is possibly a quotation from him. — 2. In other words the definition 
will be accepted without criticism or discussion. 

L In Texture: the word sarighatandyam is [an abstract noun] 
formed (from the verb sarighat- "to put together”} with an abstract 
suffix [viz., yuc = ana]. The form is locative of cause, like the forms 
of varna, etc. Has been said: viz., in 3.2 K. To define: that is, to 
determine how it differs from the qualities. 

K This, standing in dependence on the qualities (gundn ásri- 
tya tigthanti) such as sweetness, manifests the rosas. The principle by 
which it is regulated is that it must be appropriate to the speaker and 
to what is said. 

$3.6 A] 

A This texture, standing in dependence on the qualities, man- 
ifests the rasas, etc.' Now in this matter? one can imagine two main 
positions: (1) that texture and the qualities are one; or (2) that they 
are different. And if they are different, two further views are possible: 
(2a) that texture depends on the qualities; or (25) that the qualities 
depend on texture. Now if we accept the position of unity, or the view 
that the qualities depend on texture, the meaning of the verse will be 
that texture, depending on qualities that are its own self, or on quali- 
ties that reside in it, manifests the rasos. But if we accept the position 
of difference and within that position the view that texture depends 
on the qualities, then texture, standing in dependence on the qualities, 
will be by nature subordinated to the qualities but will not be identical 
with them. 

1. By adding "etc." the Vrtti shows that rasdbhdsa, bháva. and bAává- 
bhása are to be included. 2. The long and complicated commentary which 
follows is motivated by Ananda's desire to justify his very different view of 
texture from that of the older poeticians. Bhamaha (2.1-2) implies that the 
gunas of a poem depend on the degrees of word compounding in the texture. 
Vamana states explicitly (1.2.7-8) that the gunas are special properties of 
the texture (which he calls riti). Udbhata, according to Abhinava and oth- 
ers (see 3.6 L, note 2), states that the gunas are properties of the texture 
(sarighatand); and a property both resides in and is dependent on the sub- 
stance in which it resides. So the older view was that the gunas depend on 
texture. Ananda shows (in 3.6a A) that this older view fails to accord with 
the facts of literature. His own view of guna and texture is radically different. 
In Ananda’s view the gunas reside in the rosas. The srrigdrarasa is sweet (has 
sweetness), the raudrarasa is strong. The gunas are not related to the texture 
in this way. The texture, since its purpose is to manifest rasa. may rather 
be said to depend on the gunas. Abhinava points out that “depending on” 
(asritya) is here used in a different sense from the sense which it bears in the 
older view. Ananda does not mean that a given texture resides in a guna; he 
means that it follows the lead of, is subordinate to, operates for the sake of. 
a guna. The Vrtti is complicated by its examining other views. notably the 
gunasarighatanaikyapaksa or view that texture and gunas may be identical 
(3.6e A). The conclusion of the Vrtti is that on either view, Ananda's or that 
of identity, some regulation of the use of the various textures must be given. 
That rule is furnished by the second half of the present Karikd: the texture 
must be appropriate both to the speaker and to the content (rasa or other- 
wise) of what is said. Thus, Ananda accomplishes his underlying purpose of 

402 [83.6 A 

subordinating the old concept of texture (riti or sarighatand) as well as the 
old concept of the gunas to his new concept of rasas which must be suggested. 
See Introduction p. 21. 

L [Commentary on K:| ~The rasas” forms the initial word of the 
second half verse. The whole second half reads as follows: "the rasas. 
The principle by which it is regulated is that it must be appropriate to 
the speaker and to what is said." 

(Commentary on A:) The Vrttikára shows that the plural infection 
of the word rasdn is meant to include similar entities: the rasas, 
etc. Now in this matter, i.e., in regard to [what is said in] the first 
half of the verse, it is possible to imagine, or to explain, these various 
matters by means of alternatives, which he states: that texture and 
qualities are one, etc. He shows how three possible views can be 
explained: Now if we accept the position of unity, etc. 

That are its own self: To show the exact nature of a thing we 
often use an expression which refers to it as tbe basis of some entity 
hypothetically distinguished from it, as when we say that the property 
of tree-ness belongs to a simsupo.! Which reside in it: According 
to Bhattodbhata and others the qualities are properties of texture;? 
and the generally accepted view is that properties depend on their 

Will be subordinate to the qualities: in this case the expres- 
sion “depends on” will not refer to a physical relation of superstratum 
and substratum (ddhdrddheyabhdva).? For the texture does not reside 
physically in the qualities. So the sense is similar to what is intended 
when we say that the estates are based on the king, meaning that the 
ministers and the like are appropriate to that (kingdom) on which the 
king is based. Thus we arrive at the sense that texture is by nature 
subordinate to, is at the mercy of, looks up to, the qualities. 

1. A SimSupe is a tree. But in order to show this point clearly, we hy- 
pothesize as different from it a property tree-ness, which we then say belongs 
to the Simsupa. So in the case at issue. According to one theory, a given 
quality, strength, consists in (and so is really one with) the texture of long 
compounds. But we may bring out its nature with clarity by saying that the 
texture of long compounds belongs to the quality strength. 2. The view 
is attributed to Udbhata also by the Aatnaprabhd. which comments on the 
words of the Prataparudriya, sarighatandsrayd gunóh (p. 245, lines 2-3 = Sec- 
tion 7, lines 9-10) by stating Udbhatamatenoktam eva. Presumably Udbhata 

§3.6a A] 403 

expressed this view in his lost Bhémohavivarana. In his only preserved work 
he has no occasion to speak of gunas or texture, as he is concerned only with 
figures of speech. — 3. Adhárádheyabháva is the relation that obtains between 
property and substance, part and whole, etc. 

A Now wbat is the point is raising these various possibilities? It 
is this. If texture and qualities are one (1), or if the qualities depend 
on texture (25), we shall be forced into the untenable position that the 
qualities, like texture, have no fixed rules of usage. [It is an unten- 
able position,] because. of the qualities, we know that a high degree 
of sweetness’(mddhurya) and clarity (prosáda) is limited to the area 
of tragedy and love-in-separation and that strength (ojas) belongs to 
fury (roudra), wonder (adbhuta), and the like. Furthermore, sweetness 
and clarity are found only in the area of rosa, bhava, rusábhàso, and 
bhdvabhdsa. Thus the sphere of the qualities is regulated. But this 
breaks down in the case of texture. Thus we find the texture of long 
compounds in the area of love as well [as in the area of fury] and un- 
compounded texture in the area of fury [as well as in love]. Of these 
[irregularities] an example of the texture of long compounds in love is: 

with locks engoldened by the pollen 
of the flowers of Paradise; 
or such a verse as: 
anavarata-nayana-jola-lava-nipuna-parimusita-patralekham te, etc. 

Who would not grieve, fair lady, to see your face 

supported by your open hand 

as the ever dropping tears 

rob it of its painted ornament? 

Tn similar disagreement, the uncompounded texture is found in exam- 
ples of fury such as: 

Whatever man proud of his strong arm." 
Accordingly, the qualities are not one with the textures, nor are they 
dependent on the textures. 


L Like texture: because in the first view, since qualities and 
texture are held to be one, they will be equivalent in all respects, while 
in the other view (25), because the qualities are held to be properties 
of the texture. Suppose that there are indeed no fixed rules of usage. 
With this in mind, he says, because, of the qualities, etc. The word 
"because" here has the sense of "but."! On the one hand, this does not 
square with the facts (because the qualities are in fact regulated], while 
on the other hand, it is forced upon us by logic [if we accept either of 
these two views]. This: this regulation that has been laid down for 
the qualities. He now sets forth examples from the literature to prove 
that such is the case: Thus. 

By saying “we find,” he has let us know that there are places where 
this may be observed. This lays the ground for [his furnishing} an 
example: Of these. Lest some one object that there is no srrigára in 
this example,” he gives a second example: or such as. This is the 
speech of a lover for the purpose of placating his beloved who is angry 
over a love-quarrel. Accordingly: that is to say, these two views do 
not fit with the Karka. 

1. See 2.18-19c L, note 4. Abhinava's interpretation comes to the same 
result as ours, but is less literal. 2. Srigéra is produced by a combination 
of wibhávos, anubhávos, and vyabhicdribhdvas. [n the first example. as it is 
only the fragment of a verse, we are given only the àlambhonatwbhávo. 

A Now if the qualities do not depend on texture, on what do 
they depend? The answer has already been given [in the Second 
Chapter]: "Whatever depends on the predominant sense should be 
regarded as qualities. On the other hand, whatever depends on the 
non-predominant sense should be considered as ornaments (figures of 
speech), just like bracelets, etc" (2.6 K]. 

Or we can even let the qualities depend on words (rather than the 
sense of the words]. They will still not beon a par with alliteration and 

$3.6b L] 405 

the like. For alliteration and the like are properties of words regardless 
of the meaning of the words, whereas qualities are the properties of 
words capable of expressing a primary sense which gives rise to a certain 
suggested meaning. These qualities can be called properties of words 
although they really depend elsewhere, just as heroism and the like are 
said to depend on the body (as the body is the place where they are 

L On what do they depend: what he has in mind is that it 
has already been remarked on by previous [critics] that if they depend 
on words and meanings, they would differ in no way from figures of 
speech. Has already been given: viz., by the author of the Karikés. 

Or: This alternative is possible because it does not follow from the 
fact that two things depend on [i e., reside in] the same base that they 
are identical. If it did, the color [of an object] and its contact [with 
some other object] would be identical. If you object that the contact 
requires a second object (and so does not, strictly speaking, rest on the 
same base as the color!], the same may be said of the point at issue: 
the quality requires [in addition to its word base] a literal meaning [of 
that word] that may help it to a suggested sense. But this is not really 
my [i.e., the Vrttikdra’s] point of view. I would merely let the qualities 
be the property of words according to the opinion of those who do 
pot make clear distinctions, just as they take heroism and the like to 
be the properties of the body. For the man who does not make clear 
distinctions is unable to distinguish between primary and metaphorical 
usage. Still, there will be no fault. This is what I intended by my 

So he says they can be called properties of words. The sense of 
although they really depend elsewhere is: although they really 
belong to the soul.? 

1. Its base comprises two objects whereas the color base is one obj 
2. Le., to the rasa in the case of the poetic gunas. to the jivdtman in the c 
of beroism, etc. 

[§3.6c A 

A Now it may be objected that if the qualities depend on words, 
it will follow that they are identical with texture, or that they depend 
on texture. For words that are untextured (asarighatita, i.e., not struc- 
tured into a sentence) cannot be the basis of the qualities, because 
such words cannot express qualities, which depend on rasa and the 
like, since rosa and the like are conveyed by specific meanings (viz., the 
vibhávas, etc., not by the general meanings that belong to words taken 

But this objection does not hold, because it has been shown that rosa 
and the like can be suggested through phonemes and words. Or, if we 
admit that sentences suggest rosa and the like, we need not admit that 
these sentences depend by rule on any particular texture. One may thus 
say tbat the base of the qualities consists only of words, words that are 
untextured [i.e., free to belong to any one of the three textures] so long 
as they are accompanied by some particular suggestive meaning. 

L Depend on words: if, metaphorically speaking, the qualities 
reside in words, the conclusion will be as follows. The quality sweetness 
(mddhurya) is the capacity of words to convey a literal sense which [in 
its turn) suggests such rasas as love; and that capacity of words can 
be attained only by a specific texture. It follows that the texture is 
nothing separate [from the words]. Rather, it is textured words to 
which this capacity belongs. This amounts to saying that this capacity 
[to express the rasa of love, etc.) depends on texture. Such would be 
the conclusion. 

But let the gunas be properties of the words, or even identical with 
the words. What need is there to bring in texture? Anticipating this 
response, the objector continues: For words that are untextured 
cannot, etc. The rosas, bhdvas, and the false varieties and cessations 
of rasa and bhdva are suggested when they are conveyed by specific 
meanings [i.e., the vibhdvas, etc.], not by the general senses of individual 
words! unconnected with one another. Even metaphorically speaking, 
untextured words cannot be the basis of qualities dependent on, or 

§3.6c L] 407 

strictly speaking residing in, these rasas and the like. The reason for 
this is because such words cannot express, etc. For untextured 
words cannot express a literal meaning which is syntactically complete 
and therefore useful to the production of a suggested meaning. This is 
the sense. 

He now refutes (the foregoing objection]: But this objection does 
not hold. For just as it has been said that a phoneme can suggest 
a rosa, just so can a word, without expressing any meaning, suggest 
sweetness, which becomes the manifestor of a rasa, by the beauty that 
results from the mere hearing of it, as in the case of the phoneme. What 
need is there for texture? And just as it has been said that dhvani may 
be manifested by words, just so may a separate [unstructured] word, 
by its reminding us of its [general] sense, reveal a meaning capable of 
suggesting rosa. And this [meaning] in itself is sweetness. Here again 
what use is there.of texture? 

Now it may be objected that at least in that variety of dhvani which 
js manifested by the sentence it will be necessary to introduce texture, 
for without it how could the sentence or how could its literal meaning 
bave any beauty? With this objection in mind, he says: Or, if we 
admit. The word "or" is used in the sense of "also" and should be 
construed with the word vàkyavyarigyatve.? This is as much as to say: 
bring in texture; we make no objection to its presence. But a specific 
texture is not the base of, nor one with, sweetness, for sweetness and 
the like exist without it wherever rasa and the like are suggested by 
phonemes and single words. It follows that where rasa and the like are 
suggested by a sentence, it is the sentence, independently of any given 
texture, that suggests rasa and that the texture, although present, is 
needless for suggesting rasa. Hence, even if we speak metaphorically, 
the qualities depend only on words, [not on texture}. He states this 
conclusion by saying, only of words, etc. 

1. The objector is following the MimAms theory that the individual 
word denotes a genera! or class character. See 14b L, note 2. 2. Le., the 
meaning intended is abhyupagate vókyavyarigyatve ‘pi: "if it is also admitted 
that rasa and the like can be suggested by sentences 

(§3.6d A 

A Objection: "We can understand that this might be the case as 
far as sweetness is concerned, but we cannot understand that strength 
does not depend on words as set in a particular texture. Because a 
texture without compounds could never serve as the basis of strength.” 
This objection too we are not unwilling to answer, if your mind is not 
spoiled by habitual acceptance of what is commonly believed. Why 
should a texture without compounds not be a basis for strength? Af- 
ter all, it has already been shown (2.9 L] that strength is just another 
name for the excitement of a poem conveying the rasa of fury and the 
like. What fault is there if strength is expressed in a texture witb- 
out compounds and sensitive readers find no lack of beauty therein? 
Accordingly, there is nothing wrong with saying that qualities depend 
upon words that are not regulated by any particular texture. But these 
[qualities} will not stray from their own field any more than will the eye, 
etc., relate to a sense object which is not their own (e.g., the eye will 
not hear sounds]. Therefore the qualities are one thing and texture is 
another. Nor are the qualities dependent on texture. This is one view. 

L Objection: some [commentators] say that this objection is 
concerned only with dhvani suggested by sentences. But we would say 
that in strength, which is the special character of the rasas fury and the 
like, even when this strength is suggested by a phoneme or a word, its 
special beauty does not really blossom in the individual phonemes and 
words until they are given the mark of texture. And so we would take 
the objection as a general one. Conveying: the present participle (by 
Pan. 3.2.126) expresses characteristic or cause. Here the sense is that 
strength is characterized by the conveying of fury and the like.! And 
|sensitive: readers]: the word. "and" has the sense of "for." What 
he means is that since there is no lack of beauty in "Whatever man 
proud of bis strong arm” [see 2.9 A and 3.6a Al, it therefore follows, 
etc. By these: these qualities. Their own: the field [of love] bas 
been restricted [to sweetness] by the statement: "it is just srrigára that 
is the sweetest and most delightful flavor” (2.7 K). 

$3.60 A] 409 

1. Abhinave does not express bis meaning clearly. Prakdsayateh actually 
agrees with kdvyasya, not with ojas. But presumably he identifies the two: 
the strength of a poem which conveys rasa is a strength which conveys rasa. 
His meaning is that from the conveying of fury, etc., we can infer the strength 
of the poem. This is on the analogy of the stock example of P4n. 3.2.126: 
ayáná bhurijate yavandh “the Greeks eat lying down." From the knowledge 
that someone eats lying down we can infer that he is a Greek. 

A Or let us consider [another view, namely] that the qualities 
are one with texture. But it was said earlier that [if they are one,] the 
qualities, like texture, would have no fixed rules of usage, for in lit- 
erature we find irregularities [in the corrrelation of texture and rosa]. 
The reply to this is that when in literature we find an irregularity in 
a sphere that we have circumscribed, we should regard it as an aber- 
ration (virüpa). If you ask bow it is that sensitive readers nonetheless 
find beauty in such instances, our reply is: because the aberration is 
concealed by the poet’s skill (sakti). For a poetic fault is of two kinds: 
it may be due to the poet's lack of mature judgment (avyutpatti) or it 
may be due to bis lack of skill. A fault that is due to lack of mature 
judgment may be concealed by the poet's skill and so never be noticed. 
But a fault that is due to the poet's lack of skill will appear immedi- 
ately. The following sloka will give support to our position: "If a poet 
commits a fault out of lack of mature judgment, it may be concealed 
by his skill. But if the fault is due to lack of ski , it will immediately 

And so it is that the impropriety of a great poet, sucb as his well- 
known writing of the sexual enjoyment of the highest gods, does not 
appear as vulgarity because it is concealed by his skill. An exam- 
ple is the description of (Siva's| enjoyment of Parvati in the Kumára- 
sambhava.' That the charge of impropriety cannot be cancelled in such 
cases? has been shown in what follows (3.10-14b A]. But it will appear 
in conclusion by positive and negative examples that this fault can be 
concealed by poetic skill. That is why, if a poet devoid of this skill were 
to describe this type of love in the area of such actors, his work would 

410 (53.66 A 

clearly appear faulty. But if we adopt this view [that texture and qual- 
ities are one], what lack of beauty can we find in such such a stanza 
as "Whatever man proud of his strong arm"? The answer is that we 
may hypothesize a lack of beauty which is not perceived because it is 
concealed by skill.? 

1. The reference is to the Eighth Canto, of which there is no good rea- 
son to doubt Kàlidása's authorship. While the general meaning is clear, the 
exact wording and interpretation of Ánanda's text are in doubt. In the BP 
text, which we have adopted, it would also be possible to break the com- 
pound as uttama-devatá-avisaya; "(sexual enjoyment,) which is an improper 
area in dealing with the highest gods." Compare Abhinava's analogy nir- 
vydjapardkramasya purusasyávisaye ‘pr. Furthermore, Dr. Krishnamoorthy 
has reported from his Moodabidre MS the reading uttamadevatávisayaprati- 
siddhasambhogasrigdra: "such as his writing of sexual enjoyment, which is 
forbidden in the case of the highest gods." 2. aueityatydgas (MB MS, 
Krishnamoorthy) gives better sense than aucityátyágas (Kashied.). 3. The 
quoted stanza fails to use long compounds to express the rasa of fury. Ac- 
cordingly, since the quality stregnth (which by tradition is to be expressed 
by long compounds) belongs to fury, if quality and texture are the same, the 
stanza breaks the rules. ] have long puzzled over this passage before deciding 
to adopt the interpretation suggested by the punctuation of the Kashi text 
and the specific direction of BP, which states that nanu kim acdrutvam 
is an objection, to which apratiyamdnam evdropaydmah is the answer. By 
so interpreting we are forced to recognize the gunasarighatanaikya theory as 
an alternative acceptable to Ananda. My chief reason for accepting such a 
conclusion is that Ánanda uses the hypothesis of "fault hidden by skill" in 
3.10-14 b A as though he approved of it. The reader will do well, however, to 
consider carefully a very different interpretation proposed by Badarinüth Sar- 
mà in his Sanskrit commentary, p. 272. The whole gunasarighatanoikyapoksa, 
he says, is wrong and is not accepted by Ananda. He interprets the present 
passage as follows. "But if we adopt this view (that texture and the qualiti 
are one], will we [be willing to} superimpose on such a stanza as “What- 
ever man proud of his right arm” some wholly unperceived lack of beauty?” 
He goes on to say, “To hypothesize wilfully a lack of beauty in this stanza, 
when this lack is not perceived even by connoisseurs, is grossly unreasonable 
(mahiyasy anupapattih). Accordingly, this view (of the identity of quality and 
texture] is wrong." 

L Or [let us], etc. What he has in mind is that the power of 
words to manifest rasa consists in their being textured in some partic- 
ular fashion. Skill: the word sakti (lit., "power") means pratibhána 

§3.6e L] 4n 

(imagination or skill), the ability to make new presentations of every- 
thing one wishes to describe. Mature judgment (vyutpatti) is skill in 
the careful weighing (pürvaáparoparámarsa) of all that may be helpful 
to such [presentation]. His lack of skill: that is, the poet's. Impro- 
priety: the most important point in [producing] rasa is to avoid any 
disturbance of delight to those who are relishing it, because rasa is 
wholly tied to this relishing. Now treating the sexual enjoyment of the 
highest gods is like treating that of our parents. Shame and horror 
will leave us no room for delight. This is his meaning. Because it 
is concealed by his skill: for even sexual enjoyment is there so de- 
sctibed by the imaginative poet that our heart fixes on the description 
itself without any careful weighing of the context, just as when a man 
of unimpeachable valor is engaged in a battle, even if it be in a wrong 
cause, we give him our bravos at that moment, but not later when we 
weigh the matter carefully. Such is our author's meaning.' Has been 
Shown: he uses the past tense because the passage is by the author of 
the Karikas,? for it will be stated that "for the spoiling of rasa there 
is no cause other than impropriety” (3.10-14a A]. Is not perceived: 
that is, not even by those well-trained, who weigh matters carefully. 

1. We are not convinced of the justice of this interpretation, however 
artistically Abhinava has phrased it. Ananda says nothiag about subsequent 
compunction. He says only that the impropriety is tireskrta, concealed, or 
‘more literally, set aside, by the poet's skill. Nor do we find that those who 
have once loved the Eighth Canto of the Kumdrasambhava ever reverse their 
opinion of its beauty. Those critics who are shocked by its impropriety were 
doubtless shocked at their first reading. — 2. This is an extraordinary state- 
ment, for the quoted passage is not a Kànikà Nowhere else does Abhinava 
ascribe one of the sarigraha-slokas, or the parikara- or sariksepa-slokas, to the 
Karikakara. If one is to distinguish the Karik&kara from the Vrttikara, the 
sloka here indicated (3.10-14a A) must be ascribed either to the Vrttikára 
or to some extraneous author. If the sloka in question were a karikd, the 
Vrttikàra would have commented on it. I can only suppose that Abhinava's 
eagerness to justify the past tense of dorsitam has led him into confusion. 


A So, whether we suppose that texture and qualities are one or 
that they are different, we shall need some other determinant! of the 
correct use of texture. Accordingly, the Kdrikd goes on to say: "The 
principle by which it is regulated is that it must be appropriate to the 
speaker and to what is to be said (vdcya).” 

1. Other than the rule which says that heavily compounded texture, as 
identical with strength, is productive of the rosa of fury. 

L Or that they are different: if the two are different, there 
will be no principle for regulating the texture, while if they are one, the 
rusas cannot furnish the regulation.! So some othher principle must be 

The principle by which it is regulated, etc.: this forms the 
remaining portion of the Kārikā (3.6]. 

1. Because we see that the same type of texture can be used for quite 
different rosas. 

A Of the two factors, the speaker may be the poet or a character 
invented by the poet. If the latter, he may be devoid of rasa and bhava, 
or he may be possessed of rasa and bhdva.' The rasa may belong to the 
hero of the story, or to his rival. The hero of the story may be brave 
and noble (dhirodatta), or may belong to one of the other categories of 
heroes. Then too there are primary and secondary [heroes]. All these 
distinctions are possible. What is to be said (vécya) may be subsidiary 
to true rosa, which is the soul of dhvani, or it may be subsidiary to false 
or unconventional rasa (rasábhása). Its meaning may be dramatically 
representable or not. It may be concerned with upper class characters, 
or with others. Thus there are many varieties [of both speaker and 

§ 3.6g L) 413 

1. It is perfectly clear that Ananda is here using rasa in its old sense of 
a particularly vivid emotion (dhdva), especially of love, not in the new sense 
established by Abhinava, of aesthetic delight. See Introduction, pp. 18-19. 

L The hero of the story, called the ndyaka (leader) because he 
leads the story in the sense of subordinating it to his own activity, is 
he wbo enjoys the reward at the conclusion. Brave and noble, etc.: 
a brave and noble hero is most notably heroic in justice and righteous 
war; a brave and arrogant hero (dhiroddhata) is notable for herosim 
and fury; a brave and amorous hero (dhiralalita) is noted for beroism 
and love; a brave and spiritually calm hero (dhiraprasánta) is noted 
for heroic generosity and justice and for his spiritual calm. These four 
types of hero are for the most part represented by the sátvati, árabhati, 
kaifiki, and bharati modes of gesture and speech (vrtti)! respectively. 
Primary refers to the main hero, secondary to the secondary hero. 
Distinctions: differences of speaker 

What is to be said (vácyo): here "subsidiary to true rasa" means 
being a manifestor of this rasa which is the soul, that is, the very na- 
ture, of dhvani. [As vácyo has also the more techncial sense of “primary 
meaning," Abhinava now seeks to justify Ananda's statement if the 
word is taken in that sense.) A primary sense (vdcya) of dramatically 
representable meaning (abhineydrtha) is one where meaning in its sug- 
gested form, that is, the very nature of dhvani, can be brought (neya) 
into (dbhimukhyam) almost direct representation through speech, ges- 
ture, inner symptoms,” and costume. This is what is called (by Bharata] 
the goal of poetry (kdvydrtha). It alone is susceptible of enactment. 
As the sage [Bharata] has said in several places in such words as, 
“The bhdvas produce (bhávayanti) the goals (or meanings) of poetry 
(kdvyartha) with their accompaniment of speech, gesture, and inner 
symptorms."? But as the primary sense (vdcyo ‘rthah) in the form of 
the vibhávas, etc., is acted out in the course of enacting the rosas, it 
is quite proper to speak of the primary sense (vdcya) as having a (fur- 
ther, suggested) sense that is dramatically representable (abhineydr- 
tha).* We should not speak here of vyapadesivadbhava as others have 
done.5 Others: that is, it may be concerned with middle class or lower 
class characters. 

1. These four vritayah are described BANS 20.88. They are not there 
associated with the different types of hero, which are listed BANS 24.17. But 

414 [§ 3.6g L 

the first three associations are natural enough. The sátvati is the heroic 
mode par ezcellence, the drabhati is the mode of violence, and the katsikt 
the graceful, delicate mode. The bAdrati is more difficult to characterize and 
is usually limited to speech, not gesture. For the history of the modes in 
criticism see V. Raghavan, JOR 6.346-370 and 7.34-52, 91-112. In what fol- 
lows Kànkà 3.33 will refer briefly to the modes. Vrtti in this sense is to be 
distinguished from vrtti as a type of alliteration (see 1.La A, note 4) 2. In- 
ner symptoms (sattva): what is meant are the sáttvikabhávas, e.g., blushing, 
perspiration, etc. 3. BANS, prose preceding 7.1. We know of no other 
Statement to this effect in BANS. 4. Abhinava's point is that it is only the 
suggested meaning. the rasas and the like, that are really abhineya (to be 
dramatically represented). But it is justifiable to speak of the primary, lit- 
eral sense as abhineya because it forms an ever-associated part of the process. 
5. For vyapadesivodbháva, see Paribhdsendusekhara 30. It is a grammatical 
technique by which one treats a linguistic element that lacks some particular 
mark as if it were ao element which bears that mark. The following is a non- 
linguistic exarople. The demon Rabu consists only of a head, but we speak 
of rahoh siras "the head of Ráhu," placing raAoh in the genitive case as if 
Ráhu, like other beings, possessed a head. In the case at issue, some commen- 
tator(s) previous to Abhinava interpreted udcyam in vácyam abhineyártham 
by this principle. The vácyam (primary meaning), they must have said, is 
nothing other than the abhineyàárthah (the dramatically representable mean- 
ing). But in the grammatical analysis yasya artho ‘bhineyas tad vdcyam it 
appears in the genitive, as if distinct. Abhinava solves the difficulty by saying 
that the dramatically representable meaning is not the literal meaning but 
the suggested meaning to which the literal meaning leads. 

A Among these cases, when the poet as speaker is devoid of rasa 
and bháva, the type of structure (racaná) is optional. The same holds 
when the speaker is a character invented by the poet and is without 
rasa and bháva. But when the speaker, either the poet himself or a 
character invented by him, is possessed of rasa and bháva, and when the 
rasa, from its being the predominant element, forms the soul of dhvani, 
then, by necessity.! only the uncompounded texture or the texture 
employing compounds of medium length can be used. But in the rasas 
of tragedy and love-in-separation (the restriction is greater and| only 

$3.6h L] 415 

the uncompounded texture [is allowed]. Why is this? Our answer is, 
that when a rosa is set forth as primarily important, one should do 
one's very best to avoid anything that interferes with or opposes the 
perception of it. As compounds can be interpreted in many ways, a 
texture of long compounds sometimes interferes with our perception 
of the rasa. Accordingly, in passages of rasa, the frequent use of this 
texture spoils the effect, especially in drama and, in other forms of 
literature, especially when the rasa is tragedy or love-in-separation, for 
these are very delicate rasas where the slightest lack of clarity delays our 
understanding of the words and meaning. On the other hand, when 
other rosas are being presented, such as fury, a texture of medium 
length compounds and sometimes, in order to describe the action of 
a hero who is brave and arrogant, even a texture of long compounds, 
may not be at fault and need not be entirely avoided. in view of the 
needs of a literal meaning that becomes appropriate to the rasa only 
by recourse to this texture.” 

1. See note 1 on Abhinava's commentary below. — 2. Aksepa: literally, 
a drawing toward oneself, or introducing into one's work. Badarináth Sarmá 
explains the implication of its use here by the gloss, dksepo ‘nupapattimilako 
'dhyáháro^, “the supplying of an element because failure would otherwise 
ensue.” Note how Ananda justifies the use of texture here. It is “appropriate 
to the vácya, because it enables the vàcya to be appropriate to the rasa. 

L Having thus listed the varieties of speaker and the varieties of 
what is to be said (or primary sense), he now states the appropriateness 
to each of these which regulates [the texture: Among these cases. 

Of structure: that is, of texture. An ascetic devoid of, that is, 
unmoved by, rasa or bháva, may be helpful to the main rosa by his 
functioning in the plot, although he himself is indifferent. In his proper 
character, however, he is called devoid of rasa, The same: viz., option 
[of texture]. Having thus considered propriety as it relates solely to the 
speaker, he goes on to speak of it as combined with propriety to what 
is said: But when. Although the poet when he speaks should be filled 
with rasa, for otherwise the poem will be tasteless, as our author will 
state in the passage “If he himself becomes dispassionate,” etc. (3.41- 
42a A], still, when he gives his chief attention to exhibiting skill in 
complicated figures such as yamakas, he is said to be devoid of rasa. 
[When] the speaker is restricted (niyamena) to one filled with rasa and 
bhdva, and so is not indifferent; and [when] the rasa is limited (eva) 

416 [§3.6hL 

to that type which is the soul of poetry, not being of the type which 
exhibits rasa as a figure of speech (rasavadalarikára); then the type of 
texture must be only (eva) that which lacks compounds or which has 
compounds of medium length, whereas otherwise long compounds (are 
permissible]. This being the logical structure, one cannot complain of 
tautology in the use of the word niyama (restriction) and two restrictive 
particles eva [in one sentence].* 

Why is this: his attitude is, is this a pronouncement of some legal 
text? Our answer is: i.e., a reasonable explanation is. (Whatever 
interferes with] the perception of it: whatever interferes with the 
relishing of the rasa, that is, whatever is an obstacle to the relishing 
or opposes it by containing some contradictory relish. Can be inter- 
preted: the compounds can be interpreted [by the reader| in many 
ways; but the texture is [also] a causal agent in this interpretation. 
Hence there are two causal suffixes in sambhdvand.? Especially in 
drama: to begin with, one cannot act out the meaning of a compound 
without breaking up the suggested sense. The shifts of intonation and 
the like and the antara and prasdda songs? are difficult to perform in 
this case. Furthermore, in this case [i.e., in the use of long compounds} 
the understanding [of the audience] is subject to constant doubts, which 
is improper in a play, because in a play the understanding should be 
direct and immediate. And in other forms of literature: in non- 
dramatic forms. 

Delays our understanding: the sense is that our relishing of the 
rasa is hindered by the obstacles presented to it. (May not be at 
fault;| The reason why a texture of long compounds may cease to 
constitute a fault is that it may be needed by a literal meaning which 
has been chosen in order to suggest a rosa, which is appropriate to the 
rosa, but which is incapable of suggesting the rasa without the texture 
of long compounds. The explanation that has been given (by a previous 
commentator] of tadaksepa as “[only] by drawing in the action of this 
hero" [instead of "by recourse to this texture"] does not fit well; so 
enough of that. 

1. The problem is to explain the apparent presence of three expressions 
of restriction in the single sentence yadā tu kavih ... sarighatane. Abhinava 
does so by explaining that three different variables are restricted, three areas 
are excluded. This explanation plays havoc with the word order; ntyamenat- 
va must go together. — 2. Abhinava here avails himself of a grammatical 
fiction based on P&n. 6.4.51 ner aniți By this rule the causative suffix nic 

83.61 L) 417 

(= 1» € > ay) drops before any further suffix that is anit. As nic itself is 
anit, a causative suffix will drop before a second causative suffix and a double 
causative will be phonetically the same as a simple causative. Thus "Caitra 
causes Maitra to cause tbe boy to eat rice" becomes caitro maitrena bdlam 
odanam bhojayati (= bhuj + [nic] + mc + sap + ti). Lo the case at issue Abhi- 
nava is interpreting samdsdndm anekaprakárosambhávaná as “the causing by 
the texture of the reader to cause [i.e., bring about] a manifold analysis of 
compounds. 3. For antaragdna and prosádagána, see ABh. on BANS 6.29. 

A The quality called "clarity" is required throughout all types 
of texture, for it has already been said that it is “common to all the 
rasas and common to all the textures."! If one swerves from clarity, 
even’ a texture without compounds will not suggest tragedy or love- 
in-separation. If one holds to it, even a texture of medium length 
compounds will not fail to reveal them. So clarity is always to be 
sought. Accordingly, if you feel that tbe quality strength is missing in 
the verse “Whatever man proud of his strong arms," (we would point 
out that at least] it has the quality called clarity and it does not have 
sweetness.” Furthermore, it does not lack beauty since it reveals the 
rasa which its author intended. So whether texture is one with qualities 
or whether it is different, the sphere of [the various types of} texture 
is regulated by the proprieties set forth above. In this way texture 
too is a suggestor of rasa And the above mentioned principle of the 
regulation of texture as a cause of the manifestation of rasa [namely 
appropriateness to speaker and content| is precisely what regulates the 
qualities. So there is no contradiction in saying tbat the distribution 
(vyavasthána)? of texture is dependent upon the qualities. 

1. Ánanda is quoting his own remark from 2.10 A, substituting tbe syn- 
onym sarighataná for racand. — 2. Ananda's point is that strength may not 
be required for raudro rosa. Tbe quality of clarity without sweetness may 
suffice 3. Vyavasthdna implies tbe assignment of one type of texture to 
one type of speaker or content and of another type to another. 

L Throughout: this is as much as to say that every type of tex- 
ture is to be so constructed that there shall be immediate perception of 

418 (§3.6iL 

the primary sense. Has already been said: viz., under Káriká 2.10: 
“A poem's ability to communicate,” etc. Will not suggest: he means, 
because the suggestor will not transmit its own literal sense. To it: 
if one holds to clarity. As the matter [of clarity] is so important to 
our author, he explicitly gives a positive as well as a negative state 
ment. It does not have sweetness: he implies that as strength and 
sweetness have already been described as mutually contradictory,' a 
combination of the two is quite unheard of. Intended: the sense is 
that the roso has been revealed simply by clarity; it is wrong to say 
it has not been revealed. So: The meaning of the whole passage [3.6- 
3.6i] may be stated as follows. If the qualities are one with texture, 
the regulation of the texture will be identical with the regulation of the 
qualities. In the view that texture depends on the qualities, we reach 
the same conclusion. Even if the qualities depend on texture, the very 
appropriateness to speaker and content which bave been described as 
regulating, as being sources of, the texture, will likewise be sources for 
the regulation of the qualities. Thus there is no serious fault in any of 
the three views.” 

1. See 2.9 L, last paragraph. 2. This may be true, but Ananda never 
specifically defends the third view. 

K Also another sort of appropriateness, namely to tbe (particu- 
lar] visaya (genre) {in which one is writing], regulates the texture, for 
texture differs as it is found to occur in different varieties of literature. 

A  Inaddition to what is appropriate to speaker and content, that 
which is appropriate to the visaya (genre) also regulates the texture. 
For the varieties of literature are many, such as the muktaka (indepen- 
dent stanza), which is written in Sanskrit, Prakrit, or Apabhramsa; 
the sanddnitaka (couplet), visesaka (triplet), kalápaka (quatrain), and 
kulaka (connected group of more than four stanzas); the paryayabandha 
(poem on a fixed subject); parikathà (round of stories); khandakathà 

53.7 L] 419 

(short story) and sakalekathé (complete story); the poem in cantos 
(i.e., the Sanskrit mahákávya); the play; the dkhydyikd and kathd (two 
types of tale); and others. Texture assumes a particular form as it 
occurs in one or another of these. 

L He shows that there is another principle of regulation: to 
the vigaya. The word vigaya means a particular aggregate (sarighá- 
ta). And just as a man who enters a social aggregate such as the 
army, even if he should be individually a coward, adapts himself to the 
character appropriate to an army, just so a poetic sentence introduced 
into a particular aggregate such as a couplet, must become appropriate 
thereto. If the muktaka (independent stanza) has been listed under the 
word visaya, this is only in order to show that because of the absence 
of any aggregation of poetic units in the muktaka, it is entirely free 
[of regulation by aggregate] and rests on itself like the ether (of the 

By the word also he as much as says: when there also exists a need 
to be appropriate to speaker and content, the need to be appropriate 
to the vigaya extends only to differences of degree; the need to be 
appropriate to speaker and content is never set aside by the need of 
the visaya. 

The muktaka: the term is formed from mukta, "freed," not bound 
to anything else, plus.the suffix kan used in forming a conventional term 
(sorijiá).^ Because of [the conventional associations of] this (term], a 
verse occurring independently in a cohesive form of literature, even 
if that verse is in no need of syntactical completion, is not called a 
muktaka. The adjectival phrase