New CBB

I hope people are still interested in this kind of game After playing this a few times, I thought I'd like to make one too...



This time, researchers have collected cognates all over NW South America. All of the above forms are related. It should be fairly easy to figure out the major groupings (although there are a few chance resemblances), but I'm also interested in the many smaller, shallower groups and their relation to each other. For almost all words on the map, at least 3 or 4 successive generations of protoforms can be reconstructed. So, it's your task to:

- find as many valid genetic groups as possible, and reconstruct their protoforms
- find out which of these groups are related to each other in language families, subfamilies, and branches of a subfamily, and reconstruct the protoword for each of those branches and families
- reconstruct the protowords for the major families
- find the protoword of the whole stock

Have fun!

_________________
Blog: audmanh.wordpress.com
Conlangs: Buruya Nzaysa | Ndok Aisô | Doayâu | Tmaśareʔ

OK, I'll take a stab at this.

Here is the protoword:

I know it's not 100% correct, because there are still words I can't account for from this reconstruction (and plenty of other words that could derive from it, but I have no idea what group to classify them in). Since the isobars cross each other many times, it's hard to say which sound changes denote family boundaries, and which ones are just areal features that cut across genetic groupings, and that makes subcategories very difficult. But I made a few cautious classifications anyway. Here is a map:

I think the proto-word is *gʷakti

This is my current suggestion for major groups

_________________
КЕРВЕНСКИЙ

I had a similar problem with the western branch, Aszev. But I think an orange affinity makes more sense, since z > r seems like less of a leap than tt > r. I really like your categories overall. I think it will be tough getting all the sub-sub-sub-categories done right, but I'm pretty sure we're very close when it comes to the proto-word.

That's true, but once you look at the immediate surroundings the leap becomes much smaller. A hypothetical *baize could cover the nw group, as well as immediate neighbors from both red and orange :\ Cedh really succeeded in creating overlapping glosses...

I have some ideas for minor groups, but I'm waiting for a respond to our suggestions so far before diving further into it.

_________________
КЕРВЕНСКИЙ

My least favorite is oaha/oaka/wakka/waxa or whatever it is in the southeast. No matter how you slice it, the isobars overlap. Maddening!

Both of your proposals for the major divisions are pretty good. Brandrinn has even found the correct protoword already: it's indeed *gʷaikti . Well done!

As for the subgroups:

- Aszev's blue group is a valid genetic grouping (though not a first-level branch), but the protoword *ʐæːti is not correct . The largest lower node within this branch has been correctly identified by brandrinn as the *ʐeːti group ; Aszev's protoword is closer to the actual one for this lower node though.

- Aszev's pink grouping is a valid first-level branch , although there's one language incorrectly associated with this group . However, the protoword *gʷakkɹe is fairly different from the actual protoword . Brandrinn's attempt at subdividing this branch is quite good: the *weːkri group has the correct protoword but misses one language that belongs here , and both the *gʷaitʃi and *guttja groups are fairly close both in boundaries and in protoword shape .

- Aszev's red group is a valid genetic grouping (though not a first-level branch), if one leaves out the three northwestern languages, which have been correctly associated with the yellow group by brandrinn . Brandrinn has also correctly identified the two major subdivisions of Red . All the suggested protowords are wrong though, and not exactly close to the actual forms either .

- Aszev's yellow group in its maximal extent is a valid first-level branch , and the two major subgroups have been correctly identified by brandrinn . *weize is the correct protoword for the western subgroup , and *geisœ is not far off for the eastern subgroup. Aszev's suggestion *gʷɔize for the yellow group as a whole is not correct though .

_________________
Blog: audmanh.wordpress.com
Conlangs: Buruya Nzaysa | Ndok Aisô | Doayâu | Tmaśareʔ

I can't figure out what the proto word for red could be. Now that we know g_waikti is correct, we can use that to help us. I'm leaning toward *bVçti, but that first vowel... Many of the modern forms have a rounded vowel, so I'm wondering if it's something like *we or *œ, but then some forms also have j in them. It's a conundrum. Since we already guessed œ and y, I'll try a different vowel. How about *bwɛçti?

*bwɛçti is correct . The forms with j are a secondary development, (mostly) resulting from a change of stressed *ɛ → jɛ as in Western Romance.

_________________
Blog: audmanh.wordpress.com
Conlangs: Buruya Nzaysa | Ndok Aisô | Doayâu | Tmaśareʔ

I think we've gotten as close as we're going to get. Why not give us the answer in a spoiler box?

Just a minor quick thing, I'm guessing the proto-word for my blue group is *g(ʲ)aːti with the major of the two groups having *ʐaːti.

_________________
КЕРВЕНСКИЙ

The second word *ʐaːti is correct , and the first one is very close - it's actually *ɣaːti.


OK, here's a summary of the first and second level divisions. I'm still interested in seeing your guesses about minor divisions though, especially in the *ʐaːti, *bwɛçti and *weːkri subgroups.

_________________
Blog: audmanh.wordpress.com
Conlangs: Buruya Nzaysa | Ndok Aisô | Doayâu | Tmaśareʔ

Oh... I've missed a lot of fun :(


Did you mean something like the following?

Well, there are still unsolved ones left here in the forum ;p

_________________
КЕРВЕНСКИЙ

Yes, although not necessarily this detailed. Your subdivision of the *ɡʷaise group is almost completely correct, except that jɛsə belongs to the same group as ɡɛise and ɡɛsɛ, the first separate proto-form for the *ɡaiso group is actually *ɡʷaiso, and the last common proto-form for the *ɡəizə/*ɡoiza group is *ɡoizə.

_________________
Blog: audmanh.wordpress.com
Conlangs: Buruya Nzaysa | Ndok Aisô | Doayâu | Tmaśareʔ

Yeah, I'm going to try that game, too. But it appears difficult, and I'm new to this business =)


Well, I thought I'd be able to understand the whole picture better if I had as many intermediary protoforms as possible :)

But now it seems that I had no chances; I mean, unless there are some hidden cues that I failed to notice.


I see. I hesitated a bit. But the development ɡ -> j was unique anyway, and e -> ə appeared just a little bit more special than fronting the nucleus of -ai-, so I decided that the former must be given a priority in classifying the dialects.


This one was particularly difficult. Everything seemed to break so nicely into just g-group and w-group... I'd never imagine an insular reserve of ɡʷ deep inside the g-zone :(


Here I might notice that øː in ɡøːɾa requires an extra step if əi is taken as the starting point; but there are other examples of quite uneven pace of changes in the sample, so this one didn't seem too special, and I opted for the sequence ai -> əi -> oi, without supposing the rounding to appear-then-disappear.

I'm indeed curious if there are other things I could notice but didn't.

It's actually quite simple: the initial split within the g-zone is *ɡʷaisə vs. *ɡaise, with the latter then producing an offshoot with fronted vowels in the first syllable (*ɡɛise etc.), and the former splitting into *ɡʷaiso > *ɡaiso with a backed final vowel on the one hand, and *ɡoisə > *ɡoizə with a rounded diphthong (by fusion of [ʷai]) and later a voiced sibilant on the other hand. The full tree of this branch of the family is as follows:

_________________
Blog: audmanh.wordpress.com
Conlangs: Buruya Nzaysa | Ndok Aisô | Doayâu | Tmaśareʔ