Praudhamanorama 60_283

60_283

In reality, however, the view that it is a locative of subsequent condition is also easily maintained. One should keep in mind that the metarule of non-completion of what is already arranged is an exception to the metarule of internal conditioning.
[Objection:] If that is so, then even outside a compound, when asuṅ has been applied, because the stem accent takes precedence over the affix accent as the rule taught subsequently, the asuṅ [segment] would have the acute accent; however, the acute accent is desired for the u [of puṃs]. And in passages such as pumān pumāṃsaṃ paripātu, it is recited precisely in that manner! If this is objected, [we reply:] True. By virtue of the marker ñ in ḍumsuñ, the stem accent does not apply—so Haradatta and others maintain. [Objection:] Now, this is incorrect, because the rule "phiṣaḥ" applies only to bases devoid of any division into base and affix! If this is argued, [we reply:] True. This is an inferential sign (liṅga) that it applies in certain cases even to bases that have parts. Although this is accomplished merely by [the metarule] 'asiddhaṃ bahiraṅgam', nevertheless the inclusion of ñ is meant to indicate that the bahiraṅga metarule is also not universally binding: this is the conclusion.
[On the phrase] 'for the purpose of pronunciation': As for what the older grammarian said, that u and ṅ are markers, and what his grandson explained, that the u is for the sake of operations applying to bases with an u-marker, the intention is that that is unfounded.
[On the phrase] 'he puman': Because of the statement 'asambuddhau' [in the rule], lengthening does not occur.