Praudhamanorama p9

p9

Regarding [the gloss] madhyagānām: an aggregate as a whole is implied by its two parts, the initial and the final; and because its simultaneous application to an example is impossible, the technical name descending onto its parts comes to rest on those situated in the middle, but not on the initial and final, because their subordinacy to another has been determined by their inclusion within the own form of the technical name—this is the idea. Regarding svasya ca: this is due to the continuation of "svaṃ rūpam" [Pāṇini 1.1.68]. And the own form is grasped only of the initial, not of the final, because of the mention in the non-principal instrumental case [antyena], [and] because pronouns generally refer to the principal element. Regarding ādyoccāraṇam: although in the Mahābhāṣya, the glosses, etc., upadeśa has been explained through an instrumental derivation as "that by which [grammar] is taught (upadiśyate 'nena)", namely the grammatical treatise, nevertheless that is a mere rhetorical assertion. For [the affix] ghañ is hard to obtain in the instrumental sense, since it is set aside by lyuṭ. Nor can it be [the affix] gha, because it is not a technical name, since the explanation "generally in a technical name" is far-fetched, and because it is rejected by the Bhāṣyakāra with the words "surely a condition cannot have a condition", and so forth. For this very reason, ghañ also does not occur here by "halaś ca" [Pāṇini 3.3.121], which is an exception to gha. Variational application, however, is a last resort. For this very reason, the instrumental derivation "those by which words are derived (prakriyante śabdā yābhiḥ)" is also refuted. And likewise there is the vārttika: "The feminine affixes and khal-arthas set aside ac and ap; the feminine affixes set aside khal-arthas by conflict of rules." Therefore, an affix in the action-noun sense alone is proper—this is the idea.
As for what some recite, beginning with "roots, sūtras…", and what they explain: "The status of being an original enunciation belongs to those mentioned in the first half of the verse through an instrumental derivation, but to augments and the like through an object derivation. Even though the Uṇādi and the Liṅgānuśāsana are also established [as upadeśa] because of being sūtras, their separate mention is by the maxim of the cow and the bull. Similarly, the verbal root also is established just by being part of a list (gaṇa)"—and so forth. On this, the following must be stated: Is this a technical name, 'upadeśa', for these, or is it merely etymological? Not the first: because the technical name is not stated by the authors of the authoritative texts, and because of the Mahābhāṣya's statement that the affix gha does not apply because it is not a technical name; and because if a technical name coined by people like us were to regulate the authoritative treatise, there would be an unwarranted over-extension.