Category talk:Proto-Italo-Western-Romance

Dalmatian and palatalization

According to The Oxford Guide to the Romance Languages, Dalmatian didn't take part (in the Ragusan dialect) in the palatalization of k before front vowels. In Vegliot, we see that k > c before i, but not before e. Still, according to them, it is a late palatalization. Kwékwlos (talk) 21:19, 29 March 2023 (UTC)

That's correct, and it also shows /kt/ > /pt/ and /ɡn/ > /mn/, while 'real' PIWR shows /kt/ > /tt/ and /ɡn/ > /ɲ(ɲ)/. The retention of /pt/ in Dalmatian is also 'archaic', as PIWR otherwise turned it to /tt/, with the exception of the word captīvus > *cactīvus in Gallo-Romance (generally explained as contamination with a Gaulish word).

On the other hand, Dalmatian /kt, ks/ > /pt, ps/ (as well as retention of Latin /ps/) are less of an issue, since PIWR still would have had /kt, ks, ps/, judging by the different outcomes in PItR (/tt, ss, ss/) vis-à-vis PWR (/jt, js, ss~js/).

In spite of this, Dalmatian does show the Italo-Western vowel mergers: both ĭ + ē and, crucially, ŭ + ō (for the latter, see Hadlich 1963, The Phonological History of Vegliote, pp. 83–86).

Dalmatian is clearly sui generis in many ways and I think it would ultimately be best to exclude it from a reconstruction of Italo-Western Romance, for the simple reason that all the other languages in that branch share further sound changes that the former does not.

Incidentally, the resulting PIWR reconstruction must then date to some time after approximately 610 CE, the point when the Slavs took over most of the Balkans, cutting off both the Dalmatians and the ancestors of the Romanians and Aromanians from the west.

Speaking of chronology, if we take the reduction of geminate plosives as a sine qua non for PWR, which it is, then the time frame would also have to be shifted forward from the one that you came up with.

Per Politzer 1951, On the chronology of the simplification of geminates in northern France: '...the simplification of geminates, which is practically absent in the seventh century, occurs in significant numbers in the eighth century documents... The conclusion which is inevitably suggested is that at least as far as Northern France is concerned the simplification of geminates is an eighth century phenomenon'.

Per Politer 1955, A note on the North Italian voicing of intervocalic stops: '...simplification of Latin double stops is practically absent in the documents' (dated ca. 720–774 CE).

On the whole, I think it best to do the following:

1) Treat Dalmatian as effectively non-PIWR.

2) Remove all of the chronological claims (dating various stages of Proto-Romance to specific centuries).

3) Only provide reconstructed pronunciations based on the descendants, not the date of attestation.

Nicodene (talk) 11:37, 30 March 2023 (UTC)

I would object to 3). The rest is fine. Kwékwlos (talk) 13:15, 30 March 2023 (UTC)

Cf. also Loporcaro's comment on Spain (in Syllable segment and prosody, Cambridge History of the Romance Languages, vol. I): ‘Degemination, even for obstruents, had not begun yet by the time of the Arab invasion (711), as shown by the fact that Arabic geminates are treated differently, in loanwords, from their singleton counterparts...’

I would not object to your putting a Proto-Romance pronunciation for something attested, say, in the fifth century or earlier. But clearly the later stages are complicated, and giving Isidore for example a 'PWR' pronunciation is anachronistic. Nicodene (talk) 13:30, 30 March 2023 (UTC)

Genitive plural

Old Spanish has juzgo and Old French has many, like Francor, ancienor, etc. Kwékwlos (talk) 15:03, 30 March 2023 (UTC)

As fossilized forms, yes. For OFr. it may simply be regarded as a derivational suffix. Nicodene (talk) 15:17, 30 March 2023 (UTC)

Does this imply that PIWR has a functioning gentive plural? Kwékwlos (talk) 15:42, 30 March 2023 (UTC)

A couple of fossils are not really enough evidence in my view. Better would be to analyze the chronology of the decline of -orum endings in early medieval texts. Nicodene (talk) 15:53, 30 March 2023 (UTC)