How to select the right verbal person: A change of perspective between Pāṇini and his commentators

In many languages verbs and subjects share the same person (as well as other relevant features such as number and gender; see Mereu 1995; Chomsky 2000: 122-124; Moro 1997: 60-70). The ancient Indian grammarian Pāṇini (IV c. BCE) formulated three rules (namely Aṣṭādhyāyī 1.4.105, 107, 108) to explicitly account for such a phenomenon in Sanskrit. In the present study we analyse 1.4.105, 107, 108, as well as Kātyāyana’s and Patañjali’s (III and II c. BCE) discussion of these rules, and achieve two main results. The first result is that Kātyāyana and Patañjali supplemented A 1.4.105, 107, and 108 with special devices – such as the utsarga/ apavāda opposition and the conflict resolution rule A 1.4.2 – in order to enable these three rules to yield the right verbal person in the syntactic environments involving a coordinative phrase of the form [yuṣmad- + X], [asmad- + X], or [yuṣmad- + asmad-] (where X stands for any nominal or pronominal base). The second result is that a proper understanding of the rule segment sthāniny api featuring in A 1.4.105 and 107 suffices to enable these rules to yield the right verbal person in those syntactic environments, so that the aforementioned special devices deployed by Kātyāyana and Patañjali become unnecessary and unwarranted. By combining these two results, we conclude that an innovation occurred in the history of the interpretation of A 1.4.105, 107, 108, and that such an innovation was the result of a misunderstanding, on the part of Kātyāyana and Patañjali, of the relevant Aṣṭādhyāyī rules (A 1.4.105, 107, 108).