Talk:-ellus

Not -ellus

• ager + -lus → agellus (compare agrellus, which really uses -ellus)
• circulus + -lus → circellus (having both circulus and circellus in CAT:Latin words suffixed with -lus is exquisitely counterintuitive)
• claustrum → clauster- + -lus → claustellum
• culter + -lus → cultellus
• femina + -lus → femella
• frater + -lus → fratellus
• geminus + -lus → gemellus
• lamina + -lus → lamella
• liber + -lus → libellus
• rastrum → raster- + -lus → rastellum
• sacer + -lus → sacellum
• signum → sigen- + -lus → sigillum

@JohnC5: Do you agree with this? We should simply mention at -lus that there are regressive assimilations at work.

Also, shouldn't we merge the cats for -ulus, -ula, -ulum, -olus with -lus, which should be the main entry instead of -ulus?

@GuitarDudeness, why did you make -ulus the main entry? --Per utramque cavernam (talk) 11:52, 22 January 2018 (UTC)

@Per utramque cavernam: Restored that link to References of -ulus. The origin seems to be Proto-Italic *-elos. I now do not remember whether this was already Proto-Indo-European or some later fabrication from simple *-los. By various interpretations and syncopes came all other forms. And I agree with the above index (but it seems to be always -ulus often with syncope of -u-) and we should be careful in the indication of the suffix. Sihler (p. 209) gives corolla < *koronla < koronela, agellus < *agerlo < *agro-lo, puella < *puerla < *puero-la, which do not seem consistent, and he precisely does not divide *koronela... -GuitarDudeness (talk) 16:25, 22 January 2018 (UTC)

@Per utramque cavernam: This list looks good. -ula, -ulum should definitely link back to -ulus. @GuitarDudeness: In the business of -lus vs. -ulus, are you suggesting everything be moved to -ulus? —*i̯óh₁nC^[5] 22:54, 22 January 2018 (UTC)

@JohnC5:: If truly -ulus Proto-Italic *-elos) is the ground form. -GuitarDudeness (talk) 23:35, 22 January 2018 (UTC)

@JohnC5, GuitarDudeness: The cases that bother me most are the diminutives of diminutives. It seems a bit fussy to posit as a (if and only if) condition that there must be no simple/"intermediate" diminutives in -lus for a word to count as a true instance of -ellus.

• albulus exists > albellus doesn't use -ellus
• circulus exists > circellus doesn't use -ellus
• loculus exists > locellus doesn't use -ellus
• lupulus exists > lupellus doesn't use -ellus
• (oculus exists > ocellus doesn't use -ellus)
• porculus exists > porcellus doesn't use -ellus

I'm always torn between the want to be diachronically accurate, and the temptation to be synchronically helpful. Oh well, being diachronically accurate is probably better, if what we're aiming at is a truly scientific work.

Isn't the u of -ulus nothing more than a thematic vowel? I don't really see why we should bother with it. --Per utramque cavernam (talk) 23:42, 22 January 2018 (UTC)

@Per utramque cavernam:: Depends on it coming from Proto-Italic *-elos, for which are cognates in other Italic languages. I also want to sort this. The connection with PIE -los seems obvious, but whence -e-?... I would say -lus be the ground form, but then what to make of that Faliscan cognate? And thus the kind of regulus and calculus (from reg- and calc-) would already be secondary (or insertion in *reglos and *calclos?). This needs more study and confirmation from sources. And I would suggest all other forms being derived from the ground form and not "Alternative forms" and the suffix must be better explained in every word page of some diminutive form. -GuitarDudeness (talk) 00:02, 23 January 2018 (UTC)

@GuitarDudeness: So pre-Latin had a highly productive rule *-Tl- > *-Tol- which produced -culum (< pre-Latin *-colom < *-clom < Proto-Italic *-tlom < Proto-Indo-European *-tlom), -bulum (< pre-Latin *-bolom < *-blom < *-βlom < Proto-Italic *-ðlom < Proto-Indo-European *-dʰlom or < Proto-Italic *-zlom < Proto-Indo-European *-s-lom), and similarly -bilis (but with "l exīlis" treatment, giving the -i- outcome). This issue is actually very interesting and is treated ad nauseam by Sen in Syllable and Segment in Latin. As such, there is no need to posit *-elos for consonant stems in Latin. —*i̯óh₁nC^[5] 01:30, 23 January 2018 (UTC)

@Per utramque cavernam: Weiss in Outline of the Historical and Comparative Grammar of Latin says -ulus was generalized from -lus using the thematic verbs (bibulus < *bibe-los, tremulus < *treme-los) and from the phonetic processes described above pertaining to *-Tl-. So, it would seem that -lus is definitely the base form. —*i̯óh₁nC^[5] 05:32, 23 January 2018 (UTC)

@JohnC5: Thanks for this. So, should we start moving things shamelessly?

By the way, I think there's yet another category to empty: CAT:Latin words suffixed with -iter > CAT:Latin words suffixed with -ter; do you agree?

And last thing: I've suggested the deletion of several entries and cats (-ificus, -ficatio) at Wiktionary:Requests_for_deletion/Non-English#-ificus. Could you weigh in?

Pinging @GuitarDudeness also. --Per utramque cavernam (talk) 19:50, 23 January 2018 (UTC)

@Per utramque cavernam:: Certainly. And -tas, not -itas... all this class needs improvement... -GuitarDudeness (talk) 20:26, 23 January 2018 (UTC)

@GuitarDudeness: CAT:Latin words suffixed with -itas has already been emptied, fortunately. I think the entry -itas in itself doesn't do too much harm (as long as no one starts using it again), but I wouldn't be too sad to see it deleted either. A note could as well be added to -tas. --Per utramque cavernam (talk) 20:32, 23 January 2018 (UTC)

@Per utramque cavernam:: Good. Sorting then needs to come when from civi- is derived civi-tas and from veloc- is derived veloc-itas with the apparent misinterpreted/free use of -itas for -tas. And so with all others supra said. -GuitarDudeness (talk) 20:39, 23 January 2018 (UTC)