Zompist Bboard Again

The source here is almost entirely consistent with IPA which makes it a lot easier to work with than the Russian source, but it also doesn't really discuss ordering (not that it matters for the majority of the changes). Consensus is good, although consensus regarding terminology is weird. "Skou" or "Sko" refers to the Skou language, as well as the Skou family presented here, which is also part of the Skou family that includes I'saka/Krisa, Rawo, Puari and Warapu/Barupu. The second of these Skous is also referred to as Sko, or Vanimo, or Inner Skou; but "Vanimo" is also the name of the branch consisting of Dumo and Dusur, which may be dialects of one language (this latter Vanimo is sometimes called "Vanimo Proper"). Only "Inner Skou" is described here. I think "Macro-Skou" and "Inner Skou" are preferable terms for the proto-languages, and "Skou" for the modern day language. Exhaustivity is relatively poor given that Donohue also mentions that there are several different correspondence sets which he can't explain especially for vowels, and he also doesn't go into the developments of verb classes which appear to have derived from a distinct set of cluster reductions. However, there is usually at least a full set of singleton consonant and vowel correspondences so it's not completely shit. Detail is generally pretty good, although some conditioning factors can't be notated easily because they are restricted to certain word classes (probably due to certain phonetic conditions but he doesn't go into those).
I've put the proto-inventory in tabular form first ɔ I think this is a nice feature to have – then I list the conventions he uses where they differ from IPA; notably "ɵ" can refer to "the ɵ correspondence set" which can be basically any non-low and non-back vowel. Also there are several cases where I've added environments of exception (like "*b → *w !_l") which aren't explicitly noted in the source but are implied by different changes involving clusters. The intermediate stages have considerably shorter lists of changes which may not include some changes shared by all members of a given branch; I've just followed Donohue's list of "taxa-defining changes" which are not necessarily exhaustive.

---

Reference: Donohue, M. "Which Sounds Change: Descent and Borrowing in the Skou Family". In, Oceanic Linguistics, Volume 41, no. 1 (June 2002).
DOI: 10.1353/ol.2002.0018
Transcribed by: Darren
Reviewed by: (none)
Ordering ☆☆☆
Exhaustivity ★☆☆
Detail ★★☆
Consensus ★★★
Transcription ★★★


Reconstructed phonemes for Proto-Inner Skou:

Consonants:

*p	*t		*k	*kʷ
*b	*d	*j	*g	*gʷ
*m	*n		*ŋ
*f	*s				*h
	*l	*y		(*w)

"The *w cannot be reconstructed with great certainty and is probably not part of the system."

Vowels:

*i	*ʉ	*u
*e	*ɵ	*o
*ɛ		*ɔ
	*a

Suprasegmentals:

• Tone: "A three- or four-way contrast in pitch is present on monosyllables in all languages—High (H), Low (L) and Fall (HL)— so it seems reasonable to posit at least these tonal contrasts in the protolanguage." However, Donohue does "not refer to the tones of lexical items" in his reconstruction.
• Nasalisation: "It seems more likely [...] that there was a nasality contrast in Proto-Skou on all syllables regardless of their onset".

Non-IPA correspondences:

Symbol	Likely IPA	Notes
*j	[ɟ]	Listed amongst the voiced stops and later described as a "palatal stop"
*y	[j]	Listed alongside other sonorants
*ɵ	[ɵ~ɨ~ø~ʉ~ʏ]	"The sound represented by *ɵ is, in most of the languages that reflect it, variable between [ɵ], [ɨ], [ø], and sometimes even [ʉ] ... In Skou the reflex of this sound ... varies between [ø] and [ʏ]."

1.1 Proto-Inner Skou to Skou
*t → r
*kʷ → k, w
*s → r, j
*d → t
*g → h, Ø
*gʷ → p
*ŋ → Ø, k, n /#_ (in bound prefixes)
*ŋ → n (in pronouns)
*ŋ → Ø (in nouns)
*ɛ, ɔ → a /h_
*ɛ, ɔ → a /_[+falling]
*e → ɛ (merger with *ɛ)
*o → ɔ (merger with *ɔ)
*ʉ̃ → ɵ̃
*ʉ → ʉ (otherwise)
*ɵ̃ → ø̃
*ɵ → ø /_N
*ɵ → u
*pl → p, l
*tl → r, t
*kl → l
*bl → p
*ml → p
*fl → p
*hy → h
*VnV → V[+nasal]


1.2 Proto-Inner Skou to Proto-Eastern Skou
*V[+nasal]d → *V[–nasal]n
*ʉ → *i /S_
*ʉ̃ → *ũ
*ʉ → *u /{k,g,kʷ,gʷ}_

1.2.1 Proto-Eastern Skou to Proto-West Coast
*gʷ → *g → *j (ongoing change)

1.2.1.1 Proto-West Coast to Proto-Border

*j → *tʃ
*l → *n /_V[+nasal]
*tl → *hl
*b → *p !_l

1.2.1.1.1 Proto-Border to Sangke
*d → t
*tʃ → t
*g → h, x, Ø
*gʷ → tʃ (in bound prefixes, free pronouns and nouns)
*gʷ → f (in free pronouns)
*ŋ → Ø /#_ (in bound prefixes)
*ŋ → n (in free pronouns)
*ŋ → Ø (in nouns)
*ʉ̃ → ũ
*ɵ → e (listed in tables, but later noted that "*ʉ merg[es] with *ɵ, which splits depending on nasality")
*pl → f, pr
*hl → hr
*kl → jr
*bl → l
*fl → fr, hr, hn
*hy → h, s ("probably reflect[ing] the endpoint of a series of changes following the path *hy > **hẙ > **hç > ç > ʃ > s")
*VnV → V[+nasal]
*l → r


1.2.1.1.2 Proto-Border to Wutung
*k → ʔ ("innovatio[n] borrowed from Vanimo")
*kw → ʔw
*b → w, p (for reflex w only, "innovatio[n] borrowed from Vanimo")
*b → m /_V[+nasal] ("innovatio[n] borrowed from Vanimo")
*d → d (in the 3pl. pronoun only)
*d → t (elsewhere)
*j → t
*g → h, Ø
*gʷ → tʃ
*ŋ → Ø /#_ (in bound prefixes)
*ŋ → n (in free pronouns)
*ŋ → Ø (in nouns)
*o → u
*ɵ̃ → ũ
*ɵ → i (otherwise)
*pl → pl, hl
*tl → hl, hn
*kl → ʔl
*ml → hl
*hy → hɲ, hɲdʒ, s ("probably reflect[ing] the endpoint of a series of changes following the path *hy > **hẙ > **hç > ç > ʃ > s")
*VnV → V[+nasal]


1.2.1.2 Proto-Eastern Skou to Proto-Vanimo
*b → *w !_l
*b → *m /_V[+nasal]
*p → *b !_l
*k kl kʷ → *ʔ ʔl ʔw
*tl → t

1.2.1.2.1 Proto-Vanimo to Dumo
*V[+nasal]d → V[–nasal]n
*g → Ø
*gʷ → b /#_ (in bound prefixes)
*gʷ → b (in free pronouns)
*gʷ → d (in nouns)
*ŋ → Ø /#_ (in bound prefixes)
*ŋ → n (in free pronouns)
*ŋ → Ø (in nouns)
*f → p /V[+nasal]_
*f → p /_i
*f → ɸ, p (in free allophonic variation, elsewhere)
*h → Ø
*ɵ̃ → ũ
*ɵ → ɨ (otherwise)
*ʉ → i /S_
*ʉ̃ → ũ
*ʉ → u /{k,g,kʷ,gʷ}_
*fl → pl
*hy → y
*VnV → V[+nasal]


1.2.1.2.2 Proto-Vanimo to Dusur
*h → Ø
*ʔ ʔw ʔl → h hw hl (occurs following loss of *h)
*V[+nasal]d → V[–nasal]n
*g → g /#_ (in the 1sg verbal prefix)
*g → Ø (in nouns)
*gʷ → w /#_ (in bound prefixes)
*gʷ → b (in free pronouns)
*gʷ → d (in nouns)
*ŋ → ŋ /#_ (in bound prefixes)
*ŋ → n (in free pronouns)
*f → p
*ɵ̃ → ũ
*ɵ → ɵ (elsewhere; "ɵ has a more rounded range of allophones than it does in Wutung or Dumo")
*ʉ → i /S_
*ʉ̃ → ũ
*ʉ → u /{k,g,kʷ,gʷ}_
*fl → pl
*hy → y
*VnV → V[+nasal]


1.2.2 Proto-Eastern Skou to Leitre
*p → b !_l
*kʷ → kw
*b → w
*b → m /_V[+nasal]
*V[+nasal]d → V[–nasal]n
*a → i /j_ ("prior to *j merging with *s as s")
*j → s
*gʷ → gʷ /#_ (in bound prefixes)
*gʷ → b (in free pronouns and nouns)
*ŋ → Ø, ŋ /#_ (in bound prefixes)
*ŋ → ɲ (in free pronouns and perhaps in nouns)
*l → l, n
*f → p
*h → Ø
*o → u
*ʉ → i /S_
*ʉ̃ → ɔ̃
*ʉ → u /{k,g,kʷ,gʷ}_
*ɵ → ɔ
*l → Ø /C_
*hy → y

This is really great. I think I’ll transcribe all this to a webpage once I’m finished with the latest updates to Polynesian.

Darren wrote: ↑Sat Sep 30, 2023 8:56 pm I've put the proto-inventory in tabular form first ɔ I think this is a nice feature to have

I like this a lot. I considered doing so for Polynesian, but never bothered; this convinces me that it’s easy enough to do. (Note that we got an explicit request for this feature.)

That being said, I think it’s better to do this as a per-language thing, rather than per-source.

Suprasegmentals:

• Tone: "A three- or four-way contrast in pitch is present on monosyllables in all languages—High (H), Low (L) and Fall (HL)— so it seems reasonable to posit at least these tonal contrasts in the protolanguage." However, Donohue does "not refer to the tones of lexical items" in his reconstruction.
• Nasalisation: "It seems more likely [...] that there was a nasality contrast in Proto-Skou on all syllables regardless of their onset".

I’m not sure how to handle this. In your view, is it worth giving an IPA mapping for these, or not?

Thanks!

bradrn wrote: ↑Sat Sep 30, 2023 9:17 pm

Suprasegmentals:

• Tone: "A three- or four-way contrast in pitch is present on monosyllables in all languages—High (H), Low (L) and Fall (HL)— so it seems reasonable to posit at least these tonal contrasts in the protolanguage." However, Donohue does "not refer to the tones of lexical items" in his reconstruction.
• Nasalisation: "It seems more likely [...] that there was a nasality contrast in Proto-Skou on all syllables regardless of their onset".

I’m not sure how to handle this. In your view, is it worth giving an IPA mapping for these, or not?

Tone only crops up once in the changes, so I just notate it as [+falling]. Nasal vowels he writes as in IPA so they don't require notes, or as conditions I use [+nasal]. I think generally suprasegmentals can be written as [+X] when they don't crop up super often (although this might pose some problems for the search feature).

Darren wrote: ↑Sat Sep 30, 2023 9:52 pm

bradrn wrote: ↑Sat Sep 30, 2023 9:17 pm

Suprasegmentals:

• Tone: "A three- or four-way contrast in pitch is present on monosyllables in all languages—High (H), Low (L) and Fall (HL)— so it seems reasonable to posit at least these tonal contrasts in the protolanguage." However, Donohue does "not refer to the tones of lexical items" in his reconstruction.
• Nasalisation: "It seems more likely [...] that there was a nasality contrast in Proto-Skou on all syllables regardless of their onset".

I’m not sure how to handle this. In your view, is it worth giving an IPA mapping for these, or not?

Tone only crops up once in the changes, so I just notate it as [+falling]. Nasal vowels he writes as in IPA so they don't require notes, or as conditions I use [+nasal]. I think generally suprasegmentals can be written as [+X] when they don't crop up super often (although this might pose some problems for the search feature).

Sounds reasonable. I consider these less of a problem for search, since they’re whole words — there’s little ambiguity around what ‘falling’ is.

I’ve been trying to transcribe these changes into a webpage (apologies for the delay in doing so), and I find myself unsure about what to do with changes like:

Darren wrote: ↑Sat Sep 30, 2023 8:56 pm *kʷ → k, w
*s → r, j
*ŋ → Ø, k, n /#_ (in bound prefixes)
*tl → r, t
*g → h, x, Ø
[et cetera, et cetera]

Does Donohue ever give conditioning factors for the resulting phonemes? I can’t find any in the paper, but you’ve looked through it more extensively than I have.

bradrn wrote: ↑Fri Oct 06, 2023 2:10 am I’ve been trying to transcribe these changes into a webpage (apologies for the delay in doing so), and I find myself unsure about what to do with changes like:

Darren wrote: ↑Sat Sep 30, 2023 8:56 pm *kʷ → k, w
*s → r, j
*ŋ → Ø, k, n /#_ (in bound prefixes)
*tl → r, t
*g → h, x, Ø
[et cetera, et cetera]

Does Donohue ever give conditioning factors for the resulting phonemes? I can’t find any in the paper, but you’ve looked through it more extensively than I have.

Where he gives conditioning features I've noted them (although there may be some I missed). Is there anything wrong with having something like:

*kʷ → k, w (unknown conditioning)

?

Darren wrote: ↑Fri Oct 06, 2023 2:36 am Is there anything wrong with having something like:

*kʷ → k, w (unknown conditioning)

?

I don’t see anything wrong with that if that’s the best we can do.

Darren wrote: ↑Sat Sep 30, 2023 8:56 pm *ʉ̃ → ɵ̃
*ʉ → ʉ (otherwise)

Now that I have more time, I’ve resumed going through these changes, and I don’t understand the second one quoted. By ‘otherwise’, do you simply mean ‘when not nasalised’?

bradrn wrote: ↑Sat Nov 04, 2023 8:35 am

Darren wrote: ↑Sat Sep 30, 2023 8:56 pm *ʉ̃ → ɵ̃
*ʉ → ʉ (otherwise)

Now that I have more time, I’ve resumed going through these changes, and I don’t understand the second one quoted. By ‘otherwise’, do you simply mean ‘when not nasalised’?

Yes.

Darren wrote: ↑Sat Sep 30, 2023 8:56 pm *ʉ → *i /S_

What is S here?

(We should really figure out some coherent convention for these uppercase symbols…)

bradrn wrote: ↑Sat Nov 04, 2023 8:26 pm

Darren wrote: ↑Sat Sep 30, 2023 8:56 pm *ʉ → *i /S_

What is S here?

(We should really figure out some coherent convention for these uppercase symbols…)

Sorry, that's supposed to be just /s/. Idk why I capitalised it.

As for uppercase symbols, I think there's no debate as to
C = consonant
V = vowel
N = nasal
R = approximant

If we wanted to we could just generalise coronals as the defaults, so in addition to N and R you'd have
T = voiceless plosive
D = voiced plosive
S = voiceless fricative
Z = voiced fricative

Maybe also G = semivowel (that seems fairly widespread). Probably best to explicitly give the contents of these classes at the beginning of any branch they're being used in.

Darren wrote: ↑Sat Nov 04, 2023 8:41 pm

bradrn wrote: ↑Sat Nov 04, 2023 8:26 pm

Darren wrote: ↑Sat Sep 30, 2023 8:56 pm *ʉ → *i /S_

What is S here?

(We should really figure out some coherent convention for these uppercase symbols…)

Sorry, that's supposed to be just /s/. Idk why I capitalised it.

Thanks!

As for uppercase symbols, I think there's no debate as to
C = consonant
V = vowel
N = nasal
R = approximant

I haven’t seen ⟨R⟩ before (or ⟨G⟩), but these seem reasonable.

If we wanted to we could just generalise coronals as the defaults, so in addition to N and R you'd have
T = voiceless plosive
D = voiced plosive
S = voiceless fricative
Z = voiced fricative

I feel this would just be inviting confusion — nothing wrong with [-voice+stop].

Probably best to explicitly give the contents of these classes at the beginning of any branch they're being used in.

Fair enough. We’re already listing non-IPA symbols anyway.

bradrn wrote: ↑Sat Nov 04, 2023 9:02 pm I haven’t seen ⟨R⟩ before (or ⟨G⟩), but these seem reasonable.

Really?? R is practically everywhere. I'm not sure about G actually, I think only CRVN are the really intuitive ones. And also the only really useful ones.

I feel this would just be inviting confusion — nothing wrong with [-voice+stop].

Yeah you're right.

Darren wrote: ↑Sat Nov 04, 2023 9:18 pm

bradrn wrote: ↑Sat Nov 04, 2023 9:02 pm I haven’t seen ⟨R⟩ before (or ⟨G⟩), but these seem reasonable.

Really?? R is practically everywhere.

Well, I’m not particularly familiar with the diachronic literature.

CVRN is the Big 4 of the literature. Aside from that…P for labials, various forms of T or TS for your coronals, K for velars, Q for uvulars, H for laryngeals (as is tradition). Maybe L for laterals specifically. S for either voiceless stops or stops generally, Z if you need to specify voicing.

B for back vowels, E for front vowels.

Darren wrote: ↑Sat Sep 30, 2023 8:56 pm *gʷ → *g → *j (ongoing change)

I can’t find anything in the paper which says this is ongoing.

bradrn wrote: ↑Sun Nov 05, 2023 1:03 am

Darren wrote: ↑Sat Sep 30, 2023 8:56 pm *gʷ → *g → *j (ongoing change)

I can’t find anything in the paper which says this is ongoing.

"It is likely that the West Coast linkage actually represents a continuation of the Eastern Skou linkage after Leitre broke off from it, because all the characteristics that define this unit are found in Leitre, with the exception of *gw > **g > **j, which is a change that (a) spread differentially through the linkage, and did not apply uniformly in time to the whole linguistic community, and (b) was triggered through contact with Skou to its west, and the changes that diffused from that language."

Perhaps "incomplete" would be a better term.

Darren wrote: ↑Sun Nov 05, 2023 2:39 am

bradrn wrote: ↑Sun Nov 05, 2023 1:03 am

Darren wrote: ↑Sat Sep 30, 2023 8:56 pm *gʷ → *g → *j (ongoing change)

I can’t find anything in the paper which says this is ongoing.

"It is likely that the West Coast linkage actually represents a continuation of the Eastern Skou linkage after Leitre broke off from it, because all the characteristics that define this unit are found in Leitre, with the exception of *gw > **g > **j, which is a change that (a) spread differentially through the linkage, and did not apply uniformly in time to the whole linguistic community, and (b) was triggered through contact with Skou to its west, and the changes that diffused from that language."

Perhaps "incomplete" would be a better term.

I interpreted that to mean that it occurred at different times in different places, but that it did eventually occur everywhere. I can see no evidence in that paper to say that it’s incomplete.

And I might as well post what I have so far of the formatted document, since knowing myself, I may never get around to doing so otherwise: https://bradrn.com/files/skou-mockup.html

bradrn wrote: ↑Fri Nov 10, 2023 7:12 am

Darren wrote: ↑Sun Nov 05, 2023 2:39 am

bradrn wrote: ↑Sun Nov 05, 2023 1:03 am

I can’t find anything in the paper which says this is ongoing.

"It is likely that the West Coast linkage actually represents a continuation of the Eastern Skou linkage after Leitre broke off from it, because all the characteristics that define this unit are found in Leitre, with the exception of *gw > **g > **j, which is a change that (a) spread differentially through the linkage, and did not apply uniformly in time to the whole linguistic community, and (b) was triggered through contact with Skou to its west, and the changes that diffused from that language."

Perhaps "incomplete" would be a better term.

I interpreted that to mean that it occurred at different times in different places, but that it did eventually occur everywhere. I can see no evidence in that paper to say that it’s incomplete.

Yeah, that makes sense. I guess it is a characteristic change of West Coast, but since it occurred over a long period of time through diffusion, can it still count as a complete sound change from proto-Skou to proto-West Coast? I'm probably over-thinking this, let's just list it as a change and note Donohue's comment about "did not apply uniformly in time to the whole linguistic community".