Praudhamanorama p154

p154

[On] "numgrahaṇam": the meaning is that it is an indirect indicator (upalakṣaṇa) of anusvāra alone, whether originating from that [num] or inherent. Thereby, in "bṛhi vṛddhau," there is num due to having 'i' as an indicatory letter; its anusvāra [appears]. In "tṛṃhū hiṃsāyām," the anusvāra here is inherent. [Thus] "bṛṃhaṇam", "tṛṃhaṇam"—in both, retroflexion of 'n' is established. In "ivi prīṇane", "prenvanam"—here, retroflexion of 'n' does not occur by "kṛtyacaḥ" (8.4.29) and "ijādeḥ sanumaḥ" (8.4.32). Just as in the injunction "having seen a star, one should release speech," because the seeing of a star is an indirect indicator of a specific time, speech is not released even if a star is seen during the day, because the indicated specific time is absent; and at night it is released even when a star is not seen, because that [time] is present—similarly it should be understood here as well. Objection: In that case, the mention of 'num' is useless, since anusvāra is included merely as being an 'aṭ'. For in "uraḥkeṇa" and the like, when there is intervention by visarga, jihvāmūlīya, etc., this is indeed the only way out. If this is said, he replies that it is true, by: "tacca".
[Objection:] Let this be so. By the stated reasoning, the mention of num in the rule "numvisarjanīya..." (Pā. Sū. 8.3.58) would also have to be abandoned, because there too it must indicate an anusvāra; for otherwise it is impossible. Indeed, at the time the retroflexion to ṣ applies, num is not obtained in examples like sarpīṃṣi, because since retroflexion is unaccomplished [with respect to the former], the operation of anusvāra takes place prior to it. Even if it is indeed possible in [the unlengthened base of] suhinsu, nevertheless, in order to avoid overapplication in that very case, an indicative meaning targeting an anusvāra that replaces num must certainly be resorted to in accordance with the target form, based on interpretation; and since that [anusvāra] is [within] śar, the retroflexion is already established across its intervention? [Reply:] Do not say so; although rejection of the mention of visarga there is possible by the stated reasoning, the mention of num is for the sake of a restriction. To explain: if retroflexion occurs across the intervention of an anusvāra, it is restricted [to occurring] only with an [anusvāra] that replaces num. Thus it does not occur in pumsu and the like. This is the guiding principle.
yasmāt pratyayavidhistadādi pratyaye'ṅgam (Pā. Sū. 1.4.13). On this, the older grammarian said: "from which verbal root or nominal stem." That is incorrect, because forms ending in tiṅ and the like are also the base (aṅga) with respect to suffixes like tarap and the like; because such a restriction is groundless; and because in āmuṣyāyaṇa and the like, the undesired consequence of the absence of vṛddhi would follow. For this taddhita suffix is not after a nominal stem, but after an inflected word ending in a case ending. This is the guiding principle.