Praudhamanorama p172

p172

Let us return to the matter at hand. "Operation of the technical name": meaning, smāy and the rest occasioned by the technical name 'sarvanāman'. "Operation within the group": meaning, "aḍ ḍatarādibhyaḥ", "tyadādīnām aḥ", "tadoḥ saḥ sau", and so forth.
"Meaning of an instrumental compound": here, the compound prescribed by "pūrvasadṛśa—" is taken because it is specifically stated, not [the one by] "kartṛkaraṇe kṛtā". Therefore, there is no prohibition in 'tvayakā kṛtam'.
"Of the aggregate": because the word 'dvandva' is primary only there, but there is no syntactical connection: 'those words beginning with sarva which are in a dvandva', because kārakas connect only with an action. Nor is the connection: 'those which exist in a dvandva', because supplying the action of 'existing' involves cumbersome assumption, and because it is only proper that there be connection with the action of 'becoming', which is to be negated, precisely on account of its prominence. Moreover, there is no dvandva of words beginning with tyad, because it is barred by single remainder; those preceding it are other than tvad, and there, there is no difference whether there is ka or akac. Nor should it be said: why should it not be for the sake of prohibiting the treatment as a masculine even when desired, as in 'dakṣiṇottarapūrvāṇām'? For the mention of 'mātra' in "sarvanāmno vṛttimātre" means that there is treatment as a masculine even in the absence of the technical name at present for words seen somewhere as pronouns. Therefore, on the strength of reasoning and on the strength of results, taking the aggregate alone is proper.
Although, on the view that the word tvad ends in 't', a result could easily be stated even if the prohibition applied to a part, nevertheless the treatise would have a restricted scope and reasoning would be contradicted; this is the purport.