Got it, thanks. Do you know what the verse is? And in Antoine's A Sanskrit Manual, he gives rules for determining whether the verb was seṭ, aniṭ, or veṭ that don't have to do with pitch accent. For example, his first rule is: "All the roots of the tenth conjugation and of the derivative conjugations, all the roots ending with consonants, and all the roots ending with long vocalic 'r' and long 'u' insert 'i' before the terminations of the periphrastic future." He then gives dozens of exceptions. Is this rule valid? What is its relationship to the rules relating to accent that you gave?
Just a little confused―what does "swara marking on a vowel" mean?
This doesn't directly answer my question. So what you're saying is that knowing the dhātu, class number, voice, and iṭ-ness is enough? Everything else can be derived from this with perfect regularity (true irregulars aside)? And I thought that iṭ-ness can also be determined with certain rules, although there would be numerous exceptions?
@jlawler Yet some sources still identify "principal parts." I'm just not sure whether these are actually necessary for conjugation, or they simply are given for convenience. And if they are necessary, how many are there? For example, although reduplication occurs in the Sanskrit perfect/future―again, as far as I know―in a fixed, predictable manner, the rules have been the source of many nightmares (reduplication occurs with 2nd and 3rd conj. Latin verbs in the perfect system, but I don't know whether one can always derive the exact perfect stem with the same regularity). (2/2)
@jlawler For Latin, almost all the 1st conj. verbs are "regular" in their principal parts, but the other conjugations (3rd in particular) are not so regular. As you said yourself, you need the principal parts for such verbs, because there are too many "irregularities" for there to be a rule (exact, as opposed to a "guideline"). The difference with Sanskrit is that, to my knowledge, there are "exact" rules, not generic guidelines (e.g., for 2nd conj. verbs, the first-person singular perfect active form will often end in -ui), to derive all the different forms from the root. (1/2)
added 137 characters in body
None of those are Indo-European languages. One of the tags on this question is "indo-european."
@Derfder: There is no genitive case in English. 's is a clitic.