third declension accusative plural

[Aenesidemus]

At Suet. Calig. 9 we have

Apud quos quantum praeterea per hanc nutrimentorum consuetudinem amore et gratia valuerit, maxime cognitum est, cum post excessum Augusti tumultuantis et in furorem usque praecipites solus haud dubie ex conspectu suo flexit


"tumultuantis ...praecipites " bothers me. Are they both accusative plural? if they are, why would Suetonius use different endings in the same sentence? Does the -is ending seem more appropriate to him with a present participle?

 

[cinefactus]

Interesting. Just looking on phi, one of the first hits is
transiit in Asiam auxiliisque contractis et praefecto regis prouincia expulso nutantis ac dubias ciuitates retinuit in fide.

Most of the hits are on participles, although there are some nouns as well:
Studium et fides erga clientis ne iuueni quidem defuerunt.
Prisco rege inter minores gentis adlecta in senatum

 

[Aenesidemus]

Interesting. Just looking on phi, one of the first hits is
transiit in Asiam auxiliisque contractis et praefecto regis prouincia expulso nutantis ac dubias ciuitates retinuit in fide.

Most of the hits are on participles, although there are some nouns as well:
Studium et fides erga clientis ne iuueni quidem defuerunt.
Prisco rege inter minores gentis adlecta in senatum

Thanks for your response. in the third quote you offer, we have the same problem, don't we? why two different endings in two adjacent words, minores and gentis?

 

[Avunculus H]

Historically, -īs is the ending of the Indo-European i-stem class, -ēs of the IE consonant stem class, which to a big degree were merged in Latin; -ēs is the ending that generally won out, while - īs is an archaism that stuck in specific words, comparable to how some English words have "irregular" plurals like oxen or mice. gens is one of these old i-stems (the nom. sg. was *g'n.h1tis in Proto.Indo-European, the i was lost due to apocope in the prehistory of Latin.) Sometimes - īs was also wrongly applied to words that aren't old i-stems (comparable to how people wrongly construct forms like I *doth or he *doest when trying to write archaizing English), like the participles. These things are irregular and so depend on what forms a specific writer saw as the correct form.

 

[Iacobinus]

Historically, -īs is the ending of the Indo-European i-stem class, -ēs of the IE consonant stem class, which to a big degree were merged in Latin; -ēs is the ending that generally won out, while - īs is an archaism that stuck in specific words, comparable to how some English words have "irregular" plurals like oxen or mice. gens is one of these old i-stems (the nom. sg. was *g'n.h1tis in Proto.Indo-European, the i was lost due to apocope in the prehistory of Latin.) Sometimes - īs was also wrongly applied to words that aren't old i-stems (comparable to how people wrongly construct forms like I *doth or he *doest when trying to write archaizing English), like the participles. These things are irregular and so depend on what forms a specific writer saw as the correct form.

Thinking about it: i-stem and u-stem aren't exactly vocalic stems, but rather sonorant stem. Couldn't, in fact, the consonant stem, the i-stem and the u-stem had been part of one unique consonant-sonorant stem declension, at a moment of old latin or of proto-latin? (I mean... I believe that there was most probably just one declension in the earliest steps of proto-indo-european, but the whole declensions still appears quite close in some archaic words of classical latin.)

neutral NOM., VOC. & ACC.	T-stem (dental)	N-stem (nasal)	S-stem (sibilant)	R-stem (liquid)	I-stem (semi-vowel)	short V-stem (semi-vowel)
singular: -	T- (caput < capit)	N- (flumen < flumin)	S- (corpus < corpos)	R- (fulgur)	I- (mare < mari)	V- (cornu)
plural: -A	T-A (capit·a)	N-A (flumin·a)	S-A (corpor·a < corpos·a)	R-A (fulgur·a)	I-A (mari·a)	V-A (cornu·a)

common singular cases	T-stem (dental)	B-stem (labial)	N-stem (nasal)	G-stem (velar)	S-stem (sibilant)	R-stem (liquid)	I-stem (semi-vowel)	short V-stem (semi-vowel)	long V-stem (semi-vowel)	non vocalised V-stem
NOM.: -s	T-S (civita·s < civitat·s)	B-S (pleb·s)	N-S (sermo < sermon·s)	G-S (rex < reg·s)	S-S (mos)	R-S (soror < soror·s)	I-S (puppi·s)	V-S (lacu·s)	V-S (su·s)	V-S (bo·s < bov·s ?)
VOC.: -s	T-S (civita·s < civitat·s)	B-S (pleb·s)	N-S (sermo < sermon·s)	G-S (rex < reg·s)	S-S (mos)	R-S (soror < soror·s)	I-S (puppi·s)	V-S (lacu·s)	V-S (su·s)	V-S (bo·s < bov·s ?)
ACC.: -(e)m	T-EM (civitat·em)	B-EM (pleb·em)	N-EM (sermon·em)	G-EM (reg·em)	S-EM (mor·em < mos·em)	R-EM (soror·em)	I-M (puppi·m)	V-M (lacu·m)	V-EM (su·em)	V-EM (bov·em)
GEN.: -is	T-IS (civitat·is)	B-IS (pleb·is)	N-IS (sermon·is)	G-IS (reg·is)	S-IS (mor·is < mos·is)	R-IS (soror·is)	I-IS (puppi·s)	V-IS (lacu·s < lacu·is ?)	V-IS (su·is)	V-IS (bov·is)
DAT.: -i < -ei	T-I (civitat·i < civitat·ei)	B-I (pleb·i)	N-I (sermon·i)	G-I (reg·i < reg·ei)	S-I (mor·i < mos·ei)	R-I (soror·i)	I-I (puppi < puppi·ei ?)	V-I (lacu·i < lacu·ei ?)	V-I (su·i)	V-I (bov·i)
ABL.: -(e) < (e)d	T-E (civitat·e < civitat·ed ?)	B-E (pleb·e)	N-E (sermon·e)	G-E (reg·e < reg·ed ?)	S-E (mor·e < mos·ed ?)	R-E (soror·e)	I- (puppi < puppi·d)	V- (lacu < lacu·d)	V-E (su·e)	V-E (bov·e)

common plural cases	T-stem (dental)	B-stem (labial)	N-stem (nasal)	G-stem (velar)	S-stem (sibilant)	R-stem (liquid)	I-stem (semi-vowel)	short V-stem (semi-vowel)	long V-stem	non vocalised V-stem
NOM.: -es	T-ES (civitat·es)	B-ES (pleb·es)	N-ES (sermon·es)	G-ES (reg·es)	S-ES (mor·es < mos·es)	R-ES (soror·es)	I-ES (pupp·es < puppi·es ?)	V-ES (lacu·s < lacu·es ?)	V-ES (su·es)	V-ES (bov·es)
VOC.: -es	T-ES (civitat·es)	B-ES (pleb·es)	N-ES (sermon·es)	G-ES (reg·es)	S-ES (mor·es < mos·es)	R-ES (soror·es)	I-ES (pupp·es < puppi·es ?)	V-ES (lacu·s < lacu·es ?)	V-ES (su·es)	V-ES (bov·es)
ACC.: -(e)s < -(e)ns	T-ES (civitat·es < civitat·ens)	B-ES (pleb·es < pleb·ens)	N-ES (sermon·es < sermon·ens)	G-ES (reg·es < reg·ens)	S-ES (mor·es < mos·ens)	R-ES (soror·es < soror·ens)	I-S (puppi·s < puppi·ns)	V-S (lacu·s < lacu·ns)	V-ES (su·es)	V-ES (bov·es < bov·ens)
GEN.: -um < -om	T-VM (civitat·um)	B-VM (pleb·um)	N-VM (sermon·um)	G-VM (reg·um)	S-UM (mor·um < mos·om ?)	R-VM (soror·um)	I-VM (puppi·um)	V-VM (lacu·um)	V-VM (su·um)	V-VM (bou·m < bov·um ?)
DAT.: -(i)bus < -(e)bos	T-IBVS (civitat·ibus)	B-IBVS (pleb·ibus)	N-IBVS (sermon·ibus)	G-IBVS (reg·ibus)	S-IBVS (mor·ibus < mos·ebos ?)	R-IBVS (soror·ibus)	I-BVS (puppi·bus)	V-BVS (lacu·bus)	V-BVS (su·bus)	V-BVS (bo·bus < bov·ebos ?)
ABL.: -(i)bus < -(e)bos	T-IBVS (civitat·ibus)	B-IBVS (pleb·ibus)	N-IBVS (sermon·ibus)	G-IBVS (reg·ibus)	S-IBVS (mor·ibus < mos·ebos ?)	R-IBVS (soror·ibus)	I-BVS (puppi·bus)	V-BVS (lacu·bus)	V-BVS (su·bus)	V-BVS (bo·bus < bov·ebos ?)

(I have several doubts on some cases..., and I have not checked, so there might be many errors... sorry for that)

 

[Iáson]

This isn't a bad way to think about it, but in practice the details were not so neat as that. For example, in the -i stems the original genitive singular ending is thought to be *-oys (on the basis of Indo-European parallels), and Oscan has -eis, so we expect Latin *-īs; that we get -is seems to be a result of analogical influence from the consonant stems, or according to another view an original ending *-y-es > *-yes > -is. The original dative ending (judging by Vedic) seems to have been *-eyey, so that this ultimately ends up as -ī, the same as the consonant stem endings, via regular sound changes, is a coincidence. The original nominitive plural of the i-stems is *-eyes (Greek -εες), of the consonant stems *-es (Greek -ες); the i-stem ending *-eyes > -ēs, which was then generalised to the consonant stems, displacing the original ending. And so on.

This is not to deny that, e.g., an accusative sg. ending *-m has a common origin at some point in prehistory; just to add that over time different declensions can become closer to each other by analogy, obscuring original differences. Thus, if we rewind the clock, we won't necessarily end up at a point where all these classes constitute the same declension.

 

[Glabrigausapes]

@Iacobinus

Your table is more or less spot-on. I would just note that in Latin the abl. sg. of 3rd declension nouns, no matter their historical class, < IE locative (tonic/accented) short -i, which was leveled across the board except for the old long -i of true i-stems, which is presumably the stem vowel i + i. Thus G πατρί (dat. < loc.), V pitári, L patre (cons. stem), but ignī in some authors.

The vocative sg. traditionally is given no ending (-ø), not final -s, so you would reconstruct e.g. *sermon, which would in theory yield short sermŏ. However, in Latin I think the nominative served analogically as the voc. for all 3rd (although there may well be inscriptions or something with like o su "oh swine!"). In i- & u-stems, the so-called endingless "full grade" is found at least in Vedic, which would have yielded L *o manū, *o ignī.

 

[Avunculus H]

Adding to that, the i- and u- stems had two variants in PIE, one like normal consonant stems, with (e.g.) Gen. Sg. *-y-e/os for the i-stems and *-w-e/os for the u-stems, and ablauting endings with *-eis for the i-stems and *-eus for the u-stems. Both variants are still distinguished in Vedic, but most daughter languages, including Latin, mixed the variants into one paradigm.