zompist bboard

I recently got The Dialects of Italy by Martin Maiden and Mair Perry, and wow, I knew that Italy had quite a few regional languages straddling multiple branches of Romance, but after acquainting myself with the book, I have to say: Italy truly is a treasure-trove of linguistic diversity, at least relative to other Romance-speaking areas. So I'll be sharing some interesting things I've found in Maiden and Perry's work that may serve as inspiration for conlangers or simply shed light on interesting developments within Romance.

To start off with, some interesting or unusual sound changes:

Palatalization

Fortition of /j/ after labials: ᴘʟᴀɴᴜ(ᴍ) > [pʎaŋ] > [pjãŋ] > [ptʃãŋ], Germanic *ʙʟᴀɴᴋ > [bjaŋk] > [bdʒaŋk] (Highland Lombard)

Labials assimilate to /k/ before /j/ or /ʎ/: [pl] > [pʎ] > [pj] > [kj] (Neapolitan)

Postconsonantal /r/ > /j/, e.g. ғʀᴀᴛʀᴇ(ᴍ) > /’fjati/ (Outer Trapanese)

Nasals

/n/ > /ŋ/ not only word-finally, but also syllable-finally, including before non-velar consonants - e.g. [ʒɛŋt], and even intervocalically, passing through a stage of /ŋn/ (Piedmontese)

Nasalized offglides /j̃ w̃/ fortited to /ŋ/, e.g. ᴘᴏɴᴛᴇ(ᴍ) > /põw̃t/ > /paŋt/ and ᴜɪɴᴜ(ᴍ) > /vẽj̃/ > /veŋ/ (Emilian)

Insertion of a homorganic consonant after a final nasal, e.g. [sa’lamp] “salame”, [omp] “man” (Friulian)

Length

All consonants except /ŋ/ lengthened after stressed short vowels - e.g. /mel/ > [ melː] but /meːl/ > [meːl] - including the first element of consonant clusters: ʙᴜsᴛᴜ(ᴍ) > [bosːt] (Bolognese)

Lengthening of all cases of word-initial /r/ (Calabrian)

Word-initial /b/ > /bb/ vs. > /v/ everywhere else (Northeastern Sicilian)

Allegedly some dialects of Italian preserve -u in words derived from Latin final -um (Sardinian does this too, but Ive read on Wikipedia that some dialects of Italian itself also do)

https://www.bulgnais.com/ <--- Bulgnais is a language spoken in Italy that looks like it belongs to its own branch of the Romance language family. (edit: I see you mentioned it, but I think the language itself is distinct enough to deserve further exploration)

Sicilian might have more [z] than [s], which is cross-linguistically unusual.

Somewhere around Milan is a tiny little dialect that has changed intervocalic /l/ to a non-syllabic [e] glide. Sorry,i couldnt find the name of the dialect on a Wikipedia search, but i'll edit this post if I do.

Tuscan Italian has word-initial consonant mutation very similar to that of Gaelic : porta "door"; la forta "the door".

------
A few of those features I recognize from other Romance languages ... e.g. /r/ > /ř/ in Spanish, /pʲ/ > /kʲ/ in Romanian ... and may have been part of an areal change at the time. Post-consonantal /r/ > /j/ may or may not have had something to do with the similar change of post-consonantal /l/ to /j/ in Italian, Spanish, and Portuguese (and probably others).

Soap wrote:Somewhere around Milan is a tiny little dialect that has changed intervocalic /l/ to a non-syllabic [e] glide. Sorry,i couldnt find the name of the dialect on a Wikipedia search, but i'll edit this post if I do.

Isn't this a feature of some varieties of Venetian? See: https://en.wikipedia.org/wiki/%C5%81#Other_languages.

The article on Venetian claims the value is [ɰ], at least in dialects spoken near the city proper.

I'm surprised no Ligurian features made your list. There you have such dramatic changes as intervocalic sonorants > /j/ or /0/ and labials palatised all the way to alveopalatals (e.g. PLUVERE > ceuve ['ʧø:ve]), FLORE(M) > sciô ['ʃu:]).

Soap wrote:Tuscan Italian has word-initial consonant mutation very similar to that of Gaelic : porta "door"; la forta "the door".

This is an interesting development, but it's actually strictly a sandhi phenomenon, i.e. only allophonic and not phonemic like in the Celtic languages. There are other dialects with phonemic consonant mutation, though. Well, Tuscan itself does have a different phenomenon (raddoppiamento sintattico) which is phonemic consonant mutation. And that's actually quite widespread throughout the dialects of Italy. But I do mean that there are other varieties that have phonemic lenition-based consonant mutation.

linguoboy wrote:I'm surprised no Ligurian features made your list. There you have such dramatic changes as intervocalic sonorants > /j/ or /0/ and labials palatised all the way to alveopalatals (e.g. PLUVERE > ceuve ['ʧø:ve]), FLORE(M) > sciô ['ʃu:]).

Oh, yeah, I think I did notice that, though I'll have to go back and take a closer look at Ligurian. The thing is, though it's particularly thoroughgoing in Ligurian, the lenition and complete loss of intervocalic stops and sonorants was actually so common among the various dialects that it didn't actually seem that notable.

And the palatalized labials too. I first heard about that kind of change in Tsakonian, and then I found out it happened in French too, and it presumably happened in Portuguese during /pl/ > /tʃ/. I guess it is definitely an interesting change, but it happened in a ton of different Italian dialects, not just Ligurian.

Here are a few more interesting sound changes. Next up is some stuff about dialectal zones with interesting outcomes of the Proto-Romance vowel system.

Epenthesis

Epenthesis of /v/ or /g/ into an empty onset position: ʙʟᴀᴅᴀ(ᴍ) > /ˈbjeva/, ᴜᴜᴀ(ᴍ) > /ˈyga/ (Piacentino of Travo)

Addition of [jjə] to final stressed vowels and monosyllables: [saˈpere] > [saˈpe] > [saˈpejjə] “to know”, [me] > [ˈmejjə] (Abruzzese and Molisan)

Fortition

Initial /j/ > /k/, e.g. [je] > [ke] as in [ˈkeu] “I” (Melendugno Salentino)

Voiced stops devoice and merge with voiceless stops in all positions, including intervocalically and word-initially, except for initial geminates: [maˈtɔnna] “Madonna”, [ˈtetʃi ˈliri] “ten lire” but [ku dˈdetʃi ˈliri] “with ten lire” (Salentino)

Okay, so you know how the Romance languages are often divided into primary branches based on how they resolved the Proto-Romance vowels? E.g. as shown here:



Well, it turns out Sicilian and some related Calabrian dialects have done something slightly different. They've merged Proto-Romance /i ɪ e/ into /i/ and, likewise, /u ʊ o/ into /u/. The authors state that this was potentially due to influence from Byzantine Greek, which underwent an analogous vowel development. This vowel system is shown in the chart below:



The authors also describe a particular Corsican vowel system, in an intermediate area between the north and central areas (with Western Romance vowel outcomes) and the far south (with Southern Romance vowel outcomes). They call it the "Taravo Vowel System". Uniquely, it preserves the distinction between short and long /i u/ and merged short and long /e o/.

If we treat vowel outcomes as the gold standard for dividing Romance into subfamilies, Calabro-Sicilian could still be Western Romance, since its vowel system contains no contrasts that Western Romance did not preserve, i.e. it could have evolved from Western Romance. But the Taravo system couldn't have evolved from either Western Romance or Southern Romance. So technically speaking, by this standard, it could be its own indepdent branch of the Romance languages! Central Romance, anyone?

Also, fun fact: There are certain dialectal zones in Calabria that have Eastern Romance (i.e Romanian) vowel outcomes. There is also an area that has Southern Romance vowel outcomes (the Area Lausberg), but the authors say that some scholars argue that these supposed pockets of what appears to be "Southern Romance" are actually descended from regular Western Romance seven-vowel systems, as shown by metaphonic evidence.

Just out of curiosity, did you take that "Proto-Romance" vowel system from the book?

Serafín wrote:Just out of curiosity, did you take that "Proto-Romance" vowel system from the book?

I mean, it was what the book had, but I didn't really, since I've read about the Romance languages before and that's just the standard reconstruction of the Proto-Romance vowel system, e.g. as seen in this Wikipedia article.

Porphyrogenitos wrote:

Soap wrote:Tuscan Italian has word-initial consonant mutation very similar to that of Gaelic : porta "door"; la forta "the door".

This is an interesting development, but it's actually strictly a sandhi phenomenon, i.e. only allophonic and not phonemic like in the Celtic languages.

Ive seen it analyzed as phonemic. Vowel-final words that used to end in a consonant and are in common use do not trigger the mutation, e.g. tre porte, "three doors". Wikipedia analyzes it as https://en.wikipedia.org/wiki/Syntactic_gemination , which appears to be simply the English translation of raddoppiamento sintattico. So the two phenomena feed on each other and can even be considered to be two sides of the same coin.

Porphyrogenitos wrote:

Serafín wrote:Just out of curiosity, did you take that "Proto-Romance" vowel system from the book?

I mean, it was what the book had, but I didn't really, since I've read about the Romance languages before and that's just the standard reconstruction of the Proto-Romance vowel system, e.g. as seen in this Wikipedia article.

Even though Robert Hall did posit /ɪ/ and /ʊ/ in Proto-Romance Phonology (1976), for stressed syllables at any rate, it's very common for scholars to refer to your "Western" 7-vowel system as "Proto-Romance", "Common Romance" or even "Vulgar Latin", even when they do recognize a stage with a 9-vowel system in the same work, therefore treating Romanian and Sardinian as separate things. I am a bit surprised that you say this book refers to the 9-vowel system as "Proto-Romance".

Serafín wrote:

Porphyrogenitos wrote:

Serafín wrote:Just out of curiosity, did you take that "Proto-Romance" vowel system from the book?

I mean, it was what the book had, but I didn't really, since I've read about the Romance languages before and that's just the standard reconstruction of the Proto-Romance vowel system, e.g. as seen in this Wikipedia article.

Even though Robert Hall did posit /ɪ/ and /ʊ/ in Proto-Romance Phonology (1976), for stressed syllables at any rate, it's very common for scholars to refer to your "Western" 7-vowel system as "Proto-Romance", "Common Romance" or even "Vulgar Latin", even when they do recognize a stage with a 9-vowel system in the same work, therefore treating Romanian and Sardinian as separate things. I am a bit surprised that you say this book refers to the 9-vowel system as "Proto-Romance".

Clearly Proto-Romance must have had 9 vowels. Those who refer to the 7-vowel system as 'Proto-Romance' are either idiots or being lazy. It would be like saying that the labiovelars had already merged with the plain velars in PIE...

Serafín wrote: Even though Robert Hall did posit /ɪ/ and /ʊ/ in Proto-Romance Phonology (1976), for stressed syllables at any rate, it's very common for scholars to refer to your "Western" 7-vowel system as "Proto-Romance", "Common Romance" or even "Vulgar Latin", even when they do recognize a stage with a 9-vowel system in the same work, therefore treating Romanian and Sardinian as separate things. I am a bit surprised that you say this book refers to the 9-vowel system as "Proto-Romance".

I went back and looked at the relevant page and I realized I must have gotten it mixed up with some other table I was looking at. The chart in the book just showed the correspondences to the Latin stressed vowels.

But yeah, I will say I haven't seen any work refer to the seven-vowel system as "Proto-Romance". I've always seen that referred to as Western Romance.