A little reconstruction game

[Salmoneus]

/sighs with relief. For a moment there I thought there were two other typos, but no, in both cases I did know what I was doing all along!


It seems fair to confirm that, as you both suspect, the front rounded vowels in B are a relatively superficial development adjacent to labials, with some rightward spreading. However, while it may seem superficial, there's something else you can learn from this.

I think it's premature for me to comment on specific vowel reconstructions just yet. However, I will say that Hallow is on the right track in worrying what on earth is going on with vowel length, and I think it's fair for me to confirm that vowel length differences in these languages shouldn't necessarily be taken to indicate vowel quality differences in the parent.

[KathTheDragon]

The only problem form for the front rounded vowels is tæːʏz. I have no idea where that rounding comes from.

The ɪ could be epenthetic, yeah - B seems to not tolerate clusters other than CR and RC. That said, I'd prefer to call it allomorphy until we have a better idea of the morphonology in general.

Regarding length, from looking over my correspondences, the only case I've identified of a short vowel in A corresponding to a long vowel in B is vægjuːvɛ ~ βɪːdʒɨβɤ. Conceivably that could be related to the palatalisation gj > dʒ in B. It looks like a lot of the cases of a long vowel in A corresponding to a short vowel in B are due to the loss of word-final voiced fricatives: contrast ðɛ ~ ðɛh, ɸɛː ~ pøɣ, plɐdɛ ~ blɔtɛg. But then there are a lot of cases where B has no consonant either.

[KathTheDragon]

On the other hand, perhaps it was actually open syllable lengthening. A lot of what looks like compensatory lengthening involves what could easily be more inflectional suffixes, which weren't present in A...

Except it can't be that simple either. Gosh darn it.

[dhok]

Re: tæ:ʏz, perhaps *tæ:wɪz with a dropped intervocalic *w triggering rounding?

[mèþru]

That seems pretty likely

[evmdbm]

My sense is that A and B didn't separate very long ago - perhaps the same sort of relation as German and Dutch?

To be fair, German and Dutch separated not much less than 2,000 years ago, so 'very long ago' is clearly relative! However, it's likely that the close contact of German and Dutch kept them closer than they otherwise 'ought' to be...

OK Dutch and Afrikaans then. I know that wasn't 2000 year ago!

[mèþru]

Afrikaans gradually diverged throughout the 17th to 20th centuries. In any case, Dutch and Afrikaans are mostly mutually intelligible.

[Salmoneus]

One, hopefully last, correction: I was careless when I added that 9th sentence; A should have /ɾɛɸɐɪ/ rather than, as I originally wrote, ɾɛbɐɪ.

I'm going to try to triple-check the next language before I put it up...

[evmdbm]

Afrikaans gradually diverged throughout the 17th to 20th centuries. In any case, Dutch and Afrikaans are mostly mutually intelligible.

The day job is law - not comparative philology - so there is not much I, as an amateur, can contribute, but hopefully (despite the picking holes in my examples) you can see what I'm getting at. They're pretty closely related (closer than Dutch and German, not as close as Afrikaans and Dutch - anyone get a better suggestion?).

[Moose-tache]

As far as the low vowels go (using red for Language A and blue for Language B), æ and ä correspond well enough that they can be treated as one phoneme. Meanwhile, a ɑ and æ ɑ correspond to each other pretty well in no particular order (Language B a only appears once and it's not obvious that it has a cognate, so I'm leaving that out). So I've been working with the assumption that there is an open back or open central vowel *A, which was conditioned by different features in the two daughter languages (thus yielding open-front or open-back vowels differently), and an open front vowel *Æ, which was pulled to the center in one of the languages after the *A division became more pronounced. I may be missing the secret key that allows all three low vowels to remain distinct back to the proto-language, but as far as I can tell, there isn't one.

[mèþru]

I agree that there probably isn't one.
Also not a philologist, either amateur or professional. I'm flattered that I can give the impression of being so knowledgeable, but I'm really at sea here.

[Salmoneus]

/sighs, goes to check the low vowel correspondences.

please let me not have screwed up again, please let me not have screwed up again...



[Since progress seems to have slowed, I'm thinking I'll put up a third language tonight, but only if I can make sure I'm not going to screw it up... copyediting, it turns out, is not my strong suit...]

[Salmoneus]

OK, on the first point (and somehow I had to read through three times to even find it), 5B shouldn't have /a/, that should be /ä/. Sorry!

EDIT: however, on the broader point, other than that missing umlaut, I think I'm happy with the low vowels. If you're particularly perturbed by the fact that both languages have /sɑ/ (etc), that's meant to be there (and, FWIW, is not a loanword). If there's any other specific instance you're puzzled by let me know and I'll double-check.

[akam chinjir]

Sorry if you've answered this already---will you be adding all corrections and updates to the top post? (I'm trying to avoid looking at the thread till I've been able to spend some time on it myself, so I hope so, and also may have missed it if this has already come up.)

[Hallow XIII]

Sal, it might be sufficient if you just post more examples for the time being. If you can wait until tonight I'll gladly post some requests for context...

[Salmoneus]

Sorry if you've answered this already---will you be adding all corrections and updates to the top post? (I'm trying to avoid looking at the thread till I've been able to spend some time on it myself, so I hope so, and also may have missed it if this has already come up.)

Yes, I've been updating the original post, so that should be as correct as possible. [assuming I didn't make a mistake in the process of correcting mistakes]. Also, the ninth sentence that I added later is in the first post too.

[akam chinjir]

Thanks!

[Moose-tache]

These are the issues that are plaguing me. Do they bother anyone else?
1. The ɑ ~ æ correspondence: I can't think of any phoneme that could yield this kind of spread that's not already taken. And we don't have enough data to see if it's an existing phoneme conditioned by environment.
2. The æ ~ ɪ: and ɑ: ~ ɪ correspondences: These look like they were both diphthongs at some point (the first one could easily be a diphthong that merged with e: in language B), but I can think of no plausible reason why the vowel length would end up in this opposite configuration.

[Salmoneus]

I haven't put up a third language yet, but I have put up three further sample sentences that I thought might be interesting. I've tried to avoid errors, obviously, but I'll have another look at them tomorrow with cleaner eyes!

[I've also replaced the pernicious little 'labiodental approximants' with the appropriate vowels, I hope]

[Salmoneus]

Well, some things got in the way on my end, and interest seems to have died down anyway, but if anyone's still engaged with this, I've put up sample sentences for a third language now.