Dear Rishi, Dear Madhav and Dear All,
Rishi Rajpopat has presented a very well reflected and researched work on Pānini's grammar which deserves close attention and analysis by specialists.
As I have pointed out a little while ago (Houben 2003), "Cardona's Pāṇini: A Survey of Research (1976), giving a bibliographic overview plus a systematic and critical discussion of all research done up to 1975" has a bibliographic part (p. 3-136) which "is nearly complete within the limits the author had set for himself, while the systematic part (p. 139-371, the Survey proper) can serve as an excellent introduction to the field, the achievements up to 1975, the major issues of scholarly discussion, the principal views on these issues as defended by various scholars, and the evidence on which they based themselves."The problem of "Rule order and the application of rules" is discussed by George Cardona (1976) on pages 189-191. On p. 190 he mentions Faddegon (1936), Scharfe (1961), Shefts (1961), Vidya Niwas Mishra (1966), and himself (1970a) as scholars who accept that "1.4.2 applies only in the section headed by 1.4.1".
In view of (1) this statement in 1976; (2) Cardona 1970a; (3) the absence of the formulation of any exceptions to the position taken in Cardona 1970a and 1976; (6) the organic suitability of Cardona 1970a and 1976 to his passionate and well-reflected and widely attested endeavour to give full value to the commentarial tradition, we can, I believe, safely infer that there is little chance that he would deviate from this in the case under discussion. Even after Cardona 1976, Joshi and Roodbergen in their Aṣṭādhyāyī of Pāṇini vol. 4, pp. 14, accept that "1.4.2 applies only in the section headed by 1.4.1" and note that Faddegon only "casually notes" that AA 1.4.2 is a paribhāṣā valid upto AA 2.2: to this we should add that the Faddegon does define, as the main topic of his study, the principles on the basis of which he accepts the validity of AA 1.4.2 up to AA 2.2 only.
With regard to the main thesis, it is attempted to show that it is possible to take AA 1.4.2 vipratiṣedhe paraṁ kāryam as applicable throughout the Aṣṭādhyāyī with an interpretation of the rule that differs considerably from those proposed by Pāṇinīyas.
What does not emerge in a convincing way is: why in that case the rule is placed immediately after AA 1.4.1 ā kaḍārād ekā saṁjñā, and the detailed answer to this question given by Barend Faddegon has not been refuted.
Moreover, by making the rule AA 1.4.2 vipratiṣedhe paraṁ kāryam apply in a very different domain than where it was till now thought to be applicable on account of its position in the sūtrapāṭha -- and, more specifically, in conflict with the evident adhikāra function of AA 1.4.1 -- creates a lacuna precisely in the formulation of rules in that section.
If we do not want to use this reference to authority to entirely dismiss our young colleague Rajpopat's well-researched thesis, we can see how far we can come if we join him, preliminarily and sketchily, in his argumentation.
Although unsurmountable from a strict traditional approach, a solution to the problem pointed out above -- since for this rule this has already been done by all modern scholars in the field -- could be the following:
if and only if a thorough review of all the cases and the argument presented by Rajpopat are accepted for his new application (where the post-Pānini tradition has worked with siddha/asidddha etc.), the rule vipratiṣedhe paraṁ kāryam should have been placed earlier in the sūtrapāṭha, before AA 1.4.1ff, because even if the search for the strict application of linearity in the post-Patañjali tradition shows that this principle was not there in the grammar as conceived by Pāṇini in the sūtrapāṭha and Annexes as he accepted them, and even if we accept Joshi and Kiparsky's argument that the explicit statements on siddha and asiddha imply the default acceptation of a sarvatra siddham principle, this cannot mean that the statement of the rules was in an entirely arbitrary manner: here we see that the entire sūtrapāṭha shows a Triple Dichotomy (see my "Pānini sans détour", conférences de l'année 2013-2014,
https://journals.openedition.org/ashp/1748), with a first division into a parsing part (Adhyāya 1-5) and a prakriyā part (Adhyāya 6-8), which provides a meaningful position to each section and to the rules within each section (even the Prakriyā works or "reordered grammars" which do not do away with Pānini's sūtrapāṭha but supplement it often take regarding any given topic entire continuous series of sūtras as they were given by Pānini).
However, if the rule vipratiṣedhe paraṁ kāryam is placed earlier in the sūtrapāṭha, before AA 1.4.1ff, the application in the more restricted domain of AA 1.4, would not have been evident -- that is, not sufficiently anchored in the teaching in Pāṇini's school -- and, moreover, the meaning of the terms used in the rule would have been different. It would, then, be necessary to have the rule twice: both at an earlier place in the sūtrapāṭha and where it is now. For the sake of lāghava the rule was nevertheless read at only one place, where its absence would have been most problematic, although it stands for both rules.
Whereas this is open to further discussion, Rajpopat's argument goes entirely wrong where he writes:
"Houben is not in favour of perceiving the Aṣṭādhyāyī as a derivational machine, thereby also implicitly dismissing both the concept of linearity and consistent conflict-resolution procedures."
However "Houben 2003" (i.e., "Three Myths in Modern Pāṇinian Studies" which is a critical review of and supplement to Cardona 1999) does not accept the Aṣṭādhyāyī as a "derivational machine" because it shows it is a "reconstitutive" machine -- it is more like a 4th century BCE spelling-and-grammar checker than a grammar in any modern term of the word -- and by this neither the linearity (to the extent this is inherent in the Aṣṭādhyāyī as sūtrapāṭha) nor the presence of conflict-resolution procedures (which are explicitly stated in the Aṣṭādhyāyī, even in the main rule under discussion in Rajpopat's thesis, whether interpreted according to the Pāṇinīyas or according to Rajpopat) are rejected in the "machine" part of the grammar, which is so impressive that it makes us overlook the importance of the unformalizable domains of the sentence and sentence meaning as the main units of language and the beginning and end point of the progressive re-creation (prakriyā) of words which are syntactically connected (another way of saying that they presuppose a sentence or part of a sentence).
See now also my “The Art of Grammar in Context: ‘Science’, Human Interest, and the Construction of Cultural and Political Worlds.” In: Science and Scientification in South Asia and Europe (ed. by Axel Michaels and Christoph Wulf): 13-41. London: Routledge, 2020.
(Houben 2020, p. 18: "On the rules, skilfully formulated by several generations of grammarians up to Pāṇini, a formalism was superimposed and finally brought to perfection by Pāṇini. Behind it, the skilful and even artful choices of description all but disappeared except to the discerning eye of a few critical thinkers, including the earliest great grammarian-philosopher in the Pāṇinian tradition, Bhartṛhari (fifth century CE), who at a few occasions emphasised the ‘arbitrariness’ of the descriptive choices").
That for Kātyāyana and Patañjali and many Pāṇinīyas Pāṇini’s grammar as a closed system of rules and metarules was not a given but a challenge is borne out not only by Patañjali’s discussions themselves and works such as Nāgeśa’s Paribhāṣenduśekhara, but also by (unduly neglected) Pāṇinīyas such as Nārāyaṇa Bhaṭṭa, author of the Prakriyāsarvasva (Houben 2015 “Pāṇinian grammar of living Sanskrit: features and principles of the Prakriyā-Sarvasva of Nārāyaṇa-Bhaṭṭa of Melputtūr,“ Bulletin d’Études Indiennes (BEI) vol. 32, pp. 149-170.) , for whom the metarules, often in the same wording as the paribhāṣas of Nāgeśa and numerous predecessors, are explicitly presented as nyāyas, ‘rules of thumb’. Pāṇinīyas such as Bhaṭṭojī Dīkṣita and Nāgeśa Bhaṭṭa have discovered regularity and occasionally constructed it where it was not there, just as regularity was apparently occasionally lost (na veti vibhā.sā) even at the time of Patañjali.
With best wishes for the festive season,