transl.:"Grammaticality"

[witzel at HUSC3.HARVARD.EDU]

RE: W. O`Flaherty/Doniger`s RV, JB and Manu translations.


As promised, a final answer to the various messages on translation 
received in the few past weeks.   

My challenge to show mistakes in my evaluation of W. O`Flaherty/Doniger`s 
translations of the Rgveda, Jaiminiya Brahmana and Manu have been taken 
up only by a few, and in my mind, they have not been able to refute my 
statements.

I briefly treat comments that deal with `cool` translation, `a decent 
read` which `has a lot of punch`, `an acceptable talk` which`scans well`. 

This is worse than the much maligned orientalist approach. The 
orientalists, `who romanticized certain aspects of the past` of India, to 
quote a recent Indian author, `treated [it] with a certain amount of 
respect` (K. Roy, The emergence of monarchy in North India, Delhi, OUP, 
1994, p. 5).  The `cool` approach, however, is decidedly (N. American) 
ethno-centric: it imposes *present N.American* concepts on the text under 
discussion and does not bother about its *original* intent, which a 
translation, after all, should attempt to present to its readers.

The GI and `cool` approach is, in fact, a definite step backwards, behind 
what the Orientalists tried to do. It remains ethno-centric and cannot 
*translate* others` texts, and even less their concepts so that we may 
attempt to understand them.
 
<<< A few details: 
Robin Kornman has not understood the points I made: (1) IN RV 10.95, the 
poet depicts Pururavas as `mad` (drpyaamy aham?) as the Brahmana texts 
say, and *therefore* lets him speak with anacoluths; the joke is, of 
course, that he says `sensibly` while he does not talk sensibly at all 
(which Urvasi does when pointing out repeatedly that she, being an 
Apsaras,  cannot stay on as a wife)...; thus, I precisely try to present 
`the voice of the text`; (2) the new born baby will discern/understand 
only *later* on, and then will let a tear roll (deploring the loss of his 
mother)... (3) The sakhya remark is not about sexism at all. It just 
states a truism: `monks cannot be members of a nunnery`. Similarly, in 
Vedic times only men belong to a vratya sodality; some women hang on a 
`groupies`. To bring in present ideas of sexism is ethno-centric again. 
etc. etc.  (4) A.D. Sood reminds me of Walter Benjamin`s article on the 
task of the translator, which,  *translated* into English (from: 
Illuminationen. Ausgewaehlte Schriften, hg. Siegfried Unseld, Frankfurt 
(Suhrkamp) 1955/61, seems to be the bible of North American Indological 
translators these days. But, apart from his usual great insights, there 
has been a long and specific discussion on translating from Vedic 
Sanskrit, see next message.  (5) Incidentally, if it is even `cool` that 
O`s RV  cover is incorrectly described as one of `the RV` and not one of 
an unknown commentary of the RV: I suggest that Aditya Deva Sood print 
his first paper/book under the name of his cousin  X. Y. Sood -- after 
all, she/he also is a Sood... What is in a name? >>>


`GRAMMATICALITY`

All of this, however, was not what I was really talking about. I was also 
not talking about `grammaticality` and `translating in the same tense and 
order as in the original.` Rather about translating (grammatically 
correctly) a Vedic text into English (etc.); at first, perhaps, not into 
modern idiomatic English but into an English utterance which brings out 
the original intent of the text in question. How to transpose that 
initial translation into idiomatic English (etc.) is quite another 
question, and as a non-native speaker I would not attempt to write one 
without proper countercheck by colleagues or, preferably, a poet, as O` 
Fl. has done herself.  

Now, many of not most Vedic sentences can be translated into English 
while following the patterns of Skt./Engl. grammar. Sanskrit is, after 
all, also an Indo-European language.  And many of the grammatical 
categories are in place in both languages with the same or very similar 
`meaning` (rather: function). 

Thus  a vocative remains a vocative in English as well (though in Engl. 
it has `no` special suffix, rather a 0-morpheme.) In  English, too, you 
do not address your friend by the nominative, saying `The/a John, come 
here!`  (The Victorian crutch is, of course, to translate `o John`, etc.).

All of this is quite another matter in languages which do not have 
<exactly> the same categories, say, the Old Japanese 
`nominative/genetive` suffixes  -ga, -wa, -no, -tu, not to speak of Whorf`s 
horror case of the non-existence of nouns in Nootka: `it houses` just 
like `it flames` (=burns). Only in such cases can we speak, with some 
degree of justification, of an  `untranslatability` of certain 
grammatical categories into other languages. Obviously, whatever the 
rather ad hoc designation (the `genetive` :: genere/genus etc.) of a 
grammatical category may be, we have to look at the grammatical function 
it has and the noematic category/ies it represents before translating 
literally, and then more idiomatically, into our own languages. 

That is precisely why we can translate grammatical categories such as the
the Vedic "tenses" `imperfect` (ahan vrtram) or `perfect` (jaghaana) by
the Engl. past and perfect (he killed Vrtra <long ago> / has killed V.
<and now he is dead>) but we do not have a `tense` for the aorist
(ahaaniit) and have to substitute: `he killed <just now>`. Even worse, the
injunctive: `<as you know,> he killed` etc. etc. All of this `aporia
between languages` is well known and the daily bread of someone
specializing in Vedic;  we learn that in grad. school (see, e.g. again K.
Hoffmann`s book on the Injunctive). But I would advise to check just a few
of O./D.`s translations of these verbal categories. 

To make it absolutely clear:  When I was talking about wrong grammar, I 
critized O./D. had not paid attention to the grammatical *function* 
(`meaning`, the noematic aggregate) of the grammatical forms in question, 
therefore did not select the proper English *function* and chose the 
wrong tense, case  etc.  which resulted in an incorrect translation of 
the original intent of the Vedic text.

In short, S. Vidyasankar was perfectly correct when summing up the 
matter:  >> all the points about fracture and reconstitution are 
perfectly well-known to every serious academic. The point is that the 
reconstitutions should not be colored by preconceived ideas of what 
constitutes the original text.<<


The argument about `good translation` starts only when we talk about 
*idiomatic* translation that is based on a first, grammatically correct, 
but perhaps not so idiomatic one. There, we can easily disagree. But that 
was *not* my point at all.

However, when we do not translate a Skt. genetive as an Engl. genetive in
philosophical and other shaastra texts (Stephen Phillip`s case)-- then
this is nothing but a translators` convention adhered to in English and
other European languages as we abhor the extreme *nominal* style of
Sanskrit in our own languages. It may be possible to translate literally: 
naantaHkaraNasya vahnyaadidezagamanam by `There is no going of the inner
organ to the place of fire etc.` but more complex sentences make it
difficult and rather cumbersome to follow this practice throughout (and we
therefore turn the Skt. nominative into a verb and the genetive into a
nominative in our languages). 

This is a convention, and in such cases I would not stress translating
every Skt. case by an Engl. one (there aren`t so many in Engl.  anyhow!
Rather follow the established conventions of translating *shaastra*
Sanskrit). This has nothing to do with ungrammatical utterances which are
interpreted by a listener/reader corrrectly. This is a special case. If
you constantly and *inconsistently* (e.g.not mirroring your dialect/mother
tongue consistently) mix up grammatical categories you are not easily
understood.  -- Philos. Skt. is grammatically correct, and uses the Skt.
case system in the proper way, it is only that we do not want to imitate
this in Engl. 

Actually, we even have three levels of translation of the genetive if 
take the following case:  we have a sentence of Kalidasa`s, where the 
word in question is explained by a commentator  using a sutra of Panini`s 
grammar: Kalidasa`s gen. is to be translated  = Engl. gen., Panini`s gen. 
as meaning `in place of` and the commentator`s gen.  (by now, 
traditionally, for convenience sake) as nominative <and the nominative it 
depends on as verb>.  

This is due to three entirely different processes. Panini uses the gen. 
in his meta-language in an entirely different function, one not found in 
`normal` Vedic or Classical Sanskrit.  The grammatical function of the 
shaastra and Vedic/Classical genetive remains the same (possessive, etc. 
functions); but we conventionally choose to translate the shaastra gen.  
differently, thereby in fact  misrepresenting the grammatical structure 
of the sentence. Here,it is meaning and understandability that outweighs 
inter-grammaticality.

<<Actually we even do so, -- another convention --  very occasionally, in 
Vedic Sanskrit, e.g. when we translate: tat shivasya shivatvam not by 
`this is the shiva-ness of shiva` but by `This is the origin of the name 
Shiva`. >>

It should be clear that all of the above is quite different from 
*bending* the grammar and therefore the meaning of sentences, such as the 
few ones critized by me earlier (vocative viira, or the JB `future 
imperative` in -taad `and then do!` ... etc. ) These translations simply 
remain wrong.

We all make mistakes, -- and I will gladly correct mine, -- but we should 
not make a score of grammatical etc. mistakes per page when publishing 
something intended for an unsuspecting public!

(concluded in another message, on translatability and original intent )
***************************************