Thread:User talk:CodeCat/Proto-Finnic declension

In reply to your comments at Wikipedia:
 * 1) Yes, *s : *h is not attested in radical gradation at all. It only applies as suffixal gradation, i.e. regardless of syllable closure (hence essive forms such as *hambas : *hambahe-na > Fi. ). I believe this had at the PF level already been analogically extended to trisyllabic forms like *kuniŋgas. OTOH *mees-nä > miessä has been attested for "man", so this root seems to have remained non-gradating for some time.
 * 2) Chronologically *-t- has been proposed to be more original, but per Lauri Hakulinen in Suomen kielen rakenne ja kehitys, also *-tt- must've existed in Proto-Finnic already. He mentions a derivation based on forms where *-tt- is segmentable, with the 1st *-t- being from the stem. (The causatives with *-tt- are not related — these are either from PU *-pta, or from stacking two instances of PU *-ta, as in Fi.  ('rotten') →  ('to rot', itr.) →  ('to rot', tr.))
 * 3) An interesting idea. Hakulinen reports some similar effects from the plural genitives of *k-stems, such as *estek-ten > estetten (modern Fi has *estek-i-ten >, and the analogical ). This is well into OR territory though, and it's a problem that stem-medial *-kt- > *-tt- in South Estonian does gradate (*vakto-n > vatu, vs. Fi. ). (I think this would even fit well into an article I am working on. It's not likely to come out in a few years, but I'll take a mental note to credit you if this turns out to not have been noticed before!)
 * 4) It's possible that we might need to still reconstruct *-sna for (Middle) Proto-Finnic, yes — especially since the variant *-hna is also attested from the Southern Ostrobothnian dialect of Finnish (alongside *-ssa there). I'd want to see some sources for this though, since I don't know what South Estonian does for the potentials and essives of *s-stems (as in Fi.  : *pes-ne-(pi) > ; *tois-na > ). It's possible that rather than a cluster simplification, this involved the influence of the gradation *s : *h, perhaps via a contamination with the illative (*-zen, *-sna > *-zen, *-zna > *-hen, *-hna).
 * 5) The conditional is a relatively complicated issue, as one needs to account for the Samic languages having the supposedly innovative form with *-ńćə- (and not *-k-ńćə-). I don't think though that an imperfect conditional has been attested anywhere in Finnic. — OTOH yes, Hakulinen reports that 1PP -meˣ and 2PP -eˣ (and even -maˣ and -aˣ with an open vowel!) in the imperfect and conditional are attested from Finnish dialects.

Other comments that come to my mind:
 * The present passive originally ended in *-k-sen (according to Hakulinen, per Estonian, Votic and Veps), with *-(t)ta-hen in Finnish being a later innovation formed after the imperfect and the participles.
 * Marking half-long consonants with an apostrophe seems confusing. I suggest either sticking with traditional t̆t or IPA /tˑ/.
 * You'll probably need to add consonant-stem partitives for words like *hooneh 'room' (> *hoonehta). The Finnish forms like are remodelled after the *-k-stems, as are many other parts of the *-h-stems' inflection.
 * If we're working with Middle Proto-Finnic, the abessive ought to still be *-ktA.

Cheers!