Reconstruction talk:Latin/montania

I don't think Vulgar Latin ever had all those declensions intact, as shown in the table. At one stage, there was probably in the singular just nom *montanea, gen *montanē, acc *montaneã, and the nom and acc certainly merged soon enough. I think maybe we should make Vulgar Latin declensional tables, more in the style of Romanian or Asturian tables, rather than pretending that the whole suite of Classical cases still existed. —Μετάknowledge^{discuss/deeds} 06:53, 15 December 2012 (UTC)

I concur. — Ungoliant ^(Falai) 07:11, 15 December 2012 (UTC)

I think Vulgar Latin actually had *-nia instead, so the genitive would have been *montaniē. e tended to become i in that position, and later merged with the preceding consonant in certain circumstances. Vulgar Latin certainly still used the ablative case, as it was used to form the common Romance adverbial suffix -mente. —CodeCat 19:39, 17 December 2012 (UTC)

w:Appendix Probi in the 3rd century has a long list of "corrections", which are presumably meant to give pairs of vulgar forms with classic literary forms. One such correction is "vinea non vinia", which is the same change as in this word. So I think we can assume the vulgar form was really *montānia. —CodeCat 19:42, 17 December 2012 (UTC)

We should spell it as most dictionaries’ etymology section spell it, per “likely to run across”. Mine spells it *montanea. — Ungoliant ^(Falai) 20:24, 17 December 2012 (UTC)

@CC: *montania is reasonable, yes, comparable to the step between, say, teneo and tengo, if I understand that correctly. I don't see how a frozen form like -mente is proof that the ablative was functioning outside of set phrases, however. I suspect that it wasn't, and the way to prove that it was would be to find evidence of the ablative plural, because the singular is so similar to the accusative. —Μετάknowledge^{discuss/deeds} 21:53, 17 December 2012 (UTC)

That is the thing though. That phrase wasn't actually a fossilized phrase because it isn't even found in literary Latin as far as I know. It is a vulgar innovation, which obviously proves that the ablative was still in use at the time. Presumably once the ablative fell out of use (it had fallen together in form with the accusative in many declensions), the phrase was reanalysed or perhaps it had already become grammaticalised to some extent. But I would say the ablative is still securely reconstructable for the vulgar Latin of the late Roman empire, albeit diminishing much as the instrumental was in Germanic at that time.

Considering *montanea or *montania... I'm not sure why we would reconstruct *montanea. That seems like "fixing" vulgar form to make it look classical, which is really an anachronism. That other dictionaries do this is not terribly surprising, considering that classical Latin is the standard everyone uses. But it is certainly not correct Vulgar Latin. As for teneo > tengo, I'm not sure. Presuming that *tenio was the earlier stage, I don't think that could have developed into tengo, but it would be teño. However if you look at Catalan you see that a lot more verbs have added this -g-, so it is likely the result of morphological analogy and not regular sound change. —CodeCat 22:43, 17 December 2012 (UTC)

Re: tengo: I thought [ŋ] and [ɲ] were in flux in some lects? Maybe hypercorrection?

Re: ablative: You still haven't proved anything. Of course classical writers wouldn't use it much — they already had Latin's rich adverbial system to work with! That doesn't mean that there was necessarily still a functioning ablative. Because we know it disappeared, I think you need to demonstrate that it still existed instead of assuming it on evidence that scanty. (Proto-Polynesian is infinitely less attested, and so much easier to reconstruct...) —Μετάknowledge^{discuss/deeds} 23:04, 17 December 2012 (UTC)

Well, we can be sure at least that when the phrase came into use, the ablative was used to indicate circumstance. It's likely that it was on its way out, but it can't have quite disappeared yet, unless the accusative had just taken over its functions (like it did for prepositions). But when cases merge and take over the uses of both, you can't really say that there is an accusative or ablative case, it's really an accusative-ablative case then. And about tengo, can you give examples of this alternation? I know that the sequence [ŋn] <gn> became [ɲ] but that's quite different and doesn't apply here. —CodeCat 23:45, 17 December 2012 (UTC)

I'd be fine with calling it "accusative-ablative" as long as you admit that it was the same case :) As for the alternation, I have trouble thinking of great examples but pénher < pingere and frañer < frangere come to mind. —Μετάknowledge^{discuss/deeds} 23:52, 17 December 2012 (UTC)

And of course, not to mention nearly identical examples like vengo < veniō. —Μετάknowledge^{discuss/deeds} 00:15, 18 December 2012 (UTC)

OK, I moved the page and added my idea of a declensional table. Please correct it as necessary, because I'm just looking at it from the classical point of view, but you seem to know more about it. —Μετάknowledge^{discuss/deeds} 00:09, 18 December 2012 (UTC)

I'm not saying that it did merge, though.Purely by sound change, you would get the following pronunciations of the 12 forms of *montānea in 4rd century Western and Italo-Romance (i.e. all except Eastern Romance/Romanian and Sardinian/Corsican): (pasted down below)

So some cases had merged phonetically, but differently in singular and plural. The Wikipedia article also says that the genitive went out of use in favour of de + ablative before the dative went out of use. Also keep in mind that Romanian has a distinct dative-genitive case. So before we begin to consider Vulgar Latin we really have to first consider which Vulgar Latin. Which century, which groups do we include or exclude? According to the table about the vowel development on Wikipedia, the vowel system ancestral to Sardinian had already diverged by the 3rd century (since it didn't have the merger of ē and i), and Romanian diverged only a century later (since it lacks the merger of ō and u). —CodeCat 00:15, 18 December 2012 (UTC)

4rd century, lol. OK, to business: all the sound changes look good to me except the [r] in the genpl, which is obviously beside the point. What's the evidence for Wikipedia's statement there about the genitive's disappearance? And why can't we just choose a dialect by fiat, like 5th century Gallo-Italian or something neat like that? Or would you rather a kind of Ur-Romance, more like the slang of the Barracks Emperors' days? —Μετάknowledge^{discuss/deeds} 00:23, 18 December 2012 (UTC)

Here are the first 4 declensions in a different order than usual, grouped for comparison:

nom	mɔnˈtanja	mɔnˈtanjɛ	ˈmuros	ˈmuri	ˈhɔmo	ˈhɔm(e)nes	ˈmanos	ˈmanus
acc		mɔnˈtanjas	ˈmuro	ˈmuros	ˈhɔm(e)nɛ		ˈmano
abl		mɔnˈtanjis		ˈmuris		hɔˈmen(e)bos	ˈmanu	ˈman(e)bos
dat	mɔnˈtanjɛ				ˈhɔm(e)ni		ˈmanwi
gen		mɔntanˈjaro	ˈmuri	muˈroro	ˈhɔm(e)nes	ˈhɔm(e)no	ˈmanus	ˈmanwo

So dative and ablative merge completely in the plural, as do nominative and accusative of the 3rd and 4th declensions (as they already had in classical Latin). Accusative and ablative merge in all but the 4th declension, and even there the difference is marginal (so presumably they merged there, too). However, in Romanian, short u merges with long ū, so already there is a difference as the 2nd declension accusative remains distinct as ˈmuru, but instead the 4th declension accusative merges with the ablative there, as do the 4th declension nominative and genitive. In the 1st declension, nominative, ablative and accusative merge, as do dative and genitive. So it seems that the genitive remained very much distinct as a case at least through sound change. But if it was the first to disappear (as Wikipedia suggests at w:Vulgar Latin) then that would have left the nominative as the most distinct case (except in the 1st), and the ablative would have been alternately like the accusative (in singular) or the dative (in plural), and sometimes both (2nd declension). Assuming that the dative was also falling out of use for other reasons, and because the plural was probably used more than the singular, I believe that the ablative plural was eventually eventually replaced by the accusative plural, completing the merger of the cases. That would have then left the genitive and dative out of the picture, and the nominative remaining alongside the accusative, which would have become a general prepositional case. I don't know at what point the genitive and dative fell out of use, though; they probably lingered on in fixed phrases even after they fell out of general use, like in family names (like still in Italian surnames in -i!). —CodeCat 01:11, 18 December 2012 (UTC)

Now, about your question of choosing a dialect... we can do that in principle. But the point of having a common Vulgar Latin reconstruction is to provide a view into Proto-Romance, the common ancestor of all (or at least most) Romance languages. But considering that the dialects were already diverging noticeably by the 4th century, it's hard to say what to choose. We'd have to start with the "true" common ancestor, which was basically the Latin of the 1st and 2nd centuries... and then as we go progressively further, first Sardinian splits off, then Eastern Romance/Romanian, and then presumably Italian branches off next. What we choose would have to be arbitrary of course, but we would still have to choose if we want to be consistent. —CodeCat 01:11, 18 December 2012 (UTC)

OK, now I'm disagreeing a bit more. There's no way that initial /h/ of homō was preserved! And doesn't the diphthong of uomo point to /o/, not /ɔ/? I also find it unlikely that the Classical ū of manūs and manū would be preserved, except for word stress, so I would expect <manos> and <mano> there too.

Perhaps a couple tables are in order. We could have one on each side of attestability. One could have the full (or near-full) suite of cases, and reflect Probus' stuff and include phonology parental to Sardinian and Romanian too, and another could have two (or possibly three) cases, and reflect Old French and Italian archaisms more with specific Gallo-Italian phonology. —Μετάknowledge^{discuss/deeds} 04:20, 18 December 2012 (UTC)

Yes you're right, the h was probably not preserved. I don't know about the effects of stress either. The table was an attempt to give an idea of the mergers, not to be correct in all the details. I don't know where the uo in Italian came from, either. I do know that the diphthongisation occurred before the change of au > o, because that sound wasn't affected (there is no au > uo). So Vulgar Latin did preserve au, at least initially. —CodeCat 04:45, 18 December 2012 (UTC)

The Italian <uo> (= Spanish <ue>) is generally from stressed ō. Vide: duomo < *dōmũ, uovo < *ōvũ. Oh, and did you see my ŋ~ɲ thing above? What do you think? Anyway, how do you like the two-table solution? I'll try it and you can see what you think if you have no objections. —Μετάknowledge^{discuss/deeds} 05:57, 18 December 2012 (UTC)

There are other things to consider too. Vulgar Latin developed a 7-vowel system (except in Sardinia, which had 5), but the Latin alphabet wasn't equipped to write this. And of course when we write -us, that's only really correct for Romanian and Sardinian since Italo-Western had -os by then. So we would need to decide whether we follow Latin orthography even when it no longer fits the vulgar pronunciation, or make our own that reflects the 7 vowels, but no longer looks like Latin. —CodeCat 14:24, 18 December 2012 (UTC)

Yeah, I prefer a pronunciation-based Gallo-Italian approach. As for vowels, I think a narrow transcription of early G-I would recognize more like 9 vowels, although some are just (un)stressed allophones of each other. —Μετάknowledge^{discuss/deeds} 18:11, 18 December 2012 (UTC)