Page 20 of 24

Re: Quick Diachronics Challenge

Posted: Sat 09 Sep 2017, 15:55
by Creyeditor
Based on your hints I tried to figure out which languages are possibly related by comparing the number of possible routes from one language to the other.
Spoiler:
Image
This lead me to the following grouping, which is really not based on similarity, only on the number of possible routes. One possibly route is a major subfamily border, two possible routes is a minor boundary. I then went on to reconstruct the forms bottom up and this lead me to my guess. I feel it's worse than the first one.
Spoiler:
*kˤaːdin
Image

Re: Quick Diachronics Challenge

Posted: Sat 09 Sep 2017, 20:45
by sangi39
Spoiler:
Now I'm thinking either *ŋ̊aðel
Spoiler:
Image

I still think the bottom-level groupings (1 through 8) are correct, but I've attempted to throw the two words I was stumped by into the scheme as well, one going to group 8 and the other to 4. I've also taken another look at the higher level groupings and come to this:

Code: Select all

PROTO -+-> 123  -+-> 12 -+-> 1
       |         |       |
       |         |       +-> 2
       |         |
       |         +->  3
       |
       +-> 478   +-> 78  -+-> 7
       |         |        |
       |         |        +-> 8
       |         |
       |         +-> 4
       |
       +->  56  -+-> 5
                 |
                 +-> 6
I'm still a little unsure of where group 4 goes, just because of its geographical location and its phonetics.

Proto-word at bottom-level groupings:
Group 1: *kaheŋ
Group 2: *kaziŋ
Group 3: *xazen
Group 4: *hieðei (from earlier *heðei
Group 5: *ɣẽçel (from earlier *ɣẽj̊el)
Group 6: *ŋail
Group 7: *ʕe:ðo (from earlier *ʕe:ðew) (lengthening of the first vowel in open syllables)
Group 8: *hezel (from earlier *ɦezel)

Second-level groupings:
Group 1,2: *kazeŋ
Group 5,6: *ŋajel
Group 7,8: *ʕeðel

Third-level groupings:
Group 1,2,3: *kazen (from ealier *kaðen)
Group 4,7,8: *ɣẽðel (from earlier *gẽðel)

Fourth-level groupings (really hypothetical)
Group 4,7,8,5,6: ŋaðel

Proto-Word: *ŋ̊aðel

A medial fricative might explain why there are no affricates resulting from palatalisation anywhere on the map, and dental fricatives are capable of moving all over the place from to /z/, /d/ and /j/.

I still think the final consonant is *l, nasalising to *n in Group 1,2,3 and palatalising to in Group 4 (as a result of the preceding *e) before shifting to *j (the resulting *ei monophthongising to *e in two languages).
Spoiler:
I am vaguely considering that the proto-word is *kaðel and that the nasalisation in Groups 5 and 6 is the result of rhinoglottophilia, but I find that harder to justify the voiced initials in Group 7.

Re: Quick Diachronics Challenge

Posted: Sun 10 Sep 2017, 01:55
by qwed117
Kinda would want to know, where do y'all place the urheimat?

Re: Quick Diachronics Challenge

Posted: Sun 10 Sep 2017, 02:04
by sangi39
qwed117 wrote:Kinda would want to know, where do y'all place the urheimat?
Spoiler:
I've been working under the assumption that it's somewhere around about the east-south-east of the area, roughly on the coast to the east of Group 4 with one population group moving south (5,6), another moving north (1,2,3) and another moving west through the central gap between the two mountain ranges (4,7,8).

Re: Quick Diachronics Challenge

Posted: Sun 10 Sep 2017, 14:00
by Creyeditor
qwed117 wrote:Kinda would want to know, where do y'all place the urheimat?
Spoiler:
Definitely in the western bay.

Re: Quick Diachronics Challenge

Posted: Mon 11 Sep 2017, 05:18
by qwed117
Sangi is almost 100% on with the grouping. Just switch one word in group 8 with one in group 4, and you should be there.
Proto-word wise, no one has really gotten any much better. So I'll give reconstructed phonology:
/m n ŋ/
/p t k q b d g ʔ/
/r s j w/
/a ɛ e i u ɔ o/
/aː uː/
/aːɪ̯ aːʊ̯ uːɪ̯/

Hint 3:
Spoiler:
ʕ reconstructs to a phoneme that was likely more similar to ʔˁ

Re: Quick Diachronics Challenge

Posted: Mon 11 Sep 2017, 12:01
by Davush
I had a *q or in mind for the initial consonant, so since it's confirmed in the proto-lang phonology, I guess the word was:
Spoiler:
*qa:ɪdɛl

*q > k/χ/ʔ seems reasonable, with ʔ then becoming pharyngealised and χ > h in the relevant languages. Perhaps also something like *q > ɢ > ŋ

a:ɪ > a:/æ: seems plausible, and a:ɪ > i: is not too uncommon.
the /ie/ in hieðǝ etc. seems like some sort of breaking before a voiced consonant, so perhaps an early branch had a:ɪ > i:.

*d seems to be the most plausible medial consonant, as d > ð/z/j/Ø are common changes. I don't think I've seen d > h, but I'm sure it's happened.

*l seems to be the final consonant as coda *l > n makes more sense to me than the opposite way round. It also allows for ɛl > ɛu > o/œ type stuff. Possibly also ɛl > ɛʎ > ɛj > eɪ/e: etc.

The final vowel is a bit tricky, ranging from ǝ, e, u, o, œ, e:, eɪ etc. As above, I think *ɛl can explain most of these pretty reasonably.
Feel free to disregard my guess when announcing the winner as I realise I have joined quite late and after some hints have been given!

Re: Quick Diachronics Challenge

Posted: Mon 11 Sep 2017, 15:14
by jimydog000
Yeah, i had /q/ in mind as well... [:'(]

Re: Quick Diachronics Challenge

Posted: Mon 11 Sep 2017, 15:29
by Creyeditor
So, just for the sake of completeness, my third guess is *ʔaːden

Re: Quick Diachronics Challenge

Posted: Mon 11 Sep 2017, 20:49
by sangi39
Spoiler:
Given hint 3 and the given proto-phoneme inventory I'd probably take a guess at something like *qaden

Re: Quick Diachronics Challenge

Posted: Tue 12 Sep 2017, 14:13
by jimydog000
I'll post something more expansive tomorrow, but right now I'm stuck at connecting the /ŋail/ branch with anything else.

Guess:
Spoiler:
Is it /qaːɪdew/? A coda /w/ would explain coda /l, ŋ,/ and the rounded second vowel of group 8 and 7.

Re: Quick Diachronics Challenge

Posted: Tue 12 Sep 2017, 22:33
by qwed117
We've got all the right phonemes, just not the right combinations of them. [:P]

Re: Quick Diachronics Challenge

Posted: Tue 12 Sep 2017, 23:20
by sangi39
Spoiler:
Hmmm, I guess either of *qa:ɪden or *qa:den could explain the mid and high vowels outside of groups 1, 2 and 3.

*qa:ɪden might be slightly more likely. It explains why, when not a low vowel, the first vowel is front, but might also explain the pharyngealisation in group 7 (since /ʕ̞/ as the semivowel equivalent to /ɑ̯/)

Re: Quick Diachronics Challenge

Posted: Wed 13 Sep 2017, 05:41
by qwed117
I'm willing to reveal the answer, since the major ideas have all been given. Jimydog was the closest at first, with *kayzen. Sangi's grouping was pretty accurate. Davush pieced together the bits that nobody else had, but didn't do well on the parts that were already correct.
The word was *qa:isel (for the purposes of this, I considered both s and z the same; they're intervocalic allophones). I'm going to give this to Jimydog for now...

Re: Quick Diachronics Challenge

Posted: Wed 13 Sep 2017, 09:25
by Creyeditor
So intervocalic fortition was the key change most of us missed, right?

Re: Quick Diachronics Challenge

Posted: Wed 13 Sep 2017, 09:35
by jimydog000
qwed117 wrote:Sangi is almost 100% on with the grouping. Just switch one word in group 8 with one in group 4, and you should be there.
Proto-word wise, no one has really gotten any much better. So I'll give reconstructed phonology:
/m n ŋ/
/p t k q b d g ʔ/
/r s j w/
/a ɛ e i u ɔ o/
/aː uː/
/aːɪ̯ aːʊ̯ uːɪ̯/
qwed117 wrote:The word was *qa:isel
You mention no /l/ in the phonology.

Re: Quick Diachronics Challenge

Posted: Wed 13 Sep 2017, 11:22
by jimydog000
Welp, ok, I'll think up the next challenge

Just some closure:
Spoiler:
Image

Group 1: */kaheŋ/
Group 2: */kayzeŋ/
Group 3: *xajzeŋ
Group 4: *hiezeː
Group 5: *ɣɛ̃çel
Group 6: *ŋail
Group 7: *ʕiedœ
Group 8: *aːzol

Group 1 is a descendant of Group 2 with */kajeɲ/ a close cognate.

The Proto word of Group 2 and 3 (and also 1) is */xajzeŋ/

The proto word of Group 5 and 6 is */ɣɛ̃sel/ ~ */ɣɛ̃çel/

Using Sangi's groups kinda hurts my reconstruction, as we have [z] change into [d] in two branches 0.o
In other words, 4 and 7 seem to be in the same sub-branch.
Maybe it matters not, so I reconstruct the proto word of Groups 7, 8 and 4 as: */ʕiezel/

Proto 1, 2, 3 and 5, 6 is: */χˤaizel/ ~ */χˤaisel/. (the χ is uvular, this font makes it look velar)

*/ʕiezel/ is a direct descendant of */qaːisel/ : /ʕiezel/ < /ʕeezel/ < /ʕaːizel/ < */qaːisel/

Finally, Proto 1, 2, 3 and 5, 6 are directly descended from */qaːisel/ by a simple */χˤaisel/ < /qaːisel/
*/ɣɛ̃çəl/'s nasal vowel did come from a pharyngealized consonant right?

Re: Quick Diachronics Challenge

Posted: Wed 13 Sep 2017, 13:51
by sangi39
jimydog000 wrote:
qwed117 wrote:Sangi is almost 100% on with the grouping. Just switch one word in group 8 with one in group 4, and you should be there.
Proto-word wise, no one has really gotten any much better. So I'll give reconstructed phonology:
/m n ŋ/
/p t k q b d g ʔ/
/r s j w/
/a ɛ e i u ɔ o/
/aː uː/
/aːɪ̯ aːʊ̯ uːɪ̯/
qwed117 wrote:The word was *qa:isel
You mention no /l/ in the phonology.
That's the reason I switched to a final *n despite reasoning against it earlier, otherwise I would have gone with *qa:idel.

Re: Quick Diachronics Challenge

Posted: Wed 13 Sep 2017, 20:52
by qwed117
sangi39 wrote:
jimydog000 wrote:
qwed117 wrote:Sangi is almost 100% on with the grouping. Just switch one word in group 8 with one in group 4, and you should be there.
Proto-word wise, no one has really gotten any much better. So I'll give reconstructed phonology:
/m n ŋ/
/p t k q b d g ʔ/
/r s j w/
/a ɛ e i u ɔ o/
/aː uː/
/aːɪ̯ aːʊ̯ uːɪ̯/
qwed117 wrote:The word was *qa:isel
You mention no /l/ in the phonology.
That's the reason I switched to a final *n despite reasoning against it earlier, otherwise I would have gone with *qa:idel.
I can't believe I forgot to put *l [xP] I guess that does make jimydog spot on then, since he had implicated *qayzen.

Re: Quick Diachronics Challenge

Posted: Wed 13 Sep 2017, 21:22
by sangi39
qwed117 wrote:
sangi39 wrote:
jimydog000 wrote:
qwed117 wrote:Sangi is almost 100% on with the grouping. Just switch one word in group 8 with one in group 4, and you should be there.
Proto-word wise, no one has really gotten any much better. So I'll give reconstructed phonology:
/m n ŋ/
/p t k q b d g ʔ/
/r s j w/
/a ɛ e i u ɔ o/
/aː uː/
/aːɪ̯ aːʊ̯ uːɪ̯/
qwed117 wrote:The word was *qa:isel
You mention no /l/ in the phonology.
That's the reason I switched to a final *n despite reasoning against it earlier, otherwise I would have gone with *qa:idel.
I can't believe I forgot to put *l [xP]
Qwazy Qwed [:P]