The Correspondence Library

[Qwynegold]

Are there any Proto-Celtic > Cornish (any kind, but preferrably (Revived) Middle Cornish) SCs? I only found this, which is not an exhaustive list. Also, can anyone verify if this really is the phoneme inventory of (Reconstructed) Proto-Celtic?:

/i iː u uː/
/e eː o/
/a aː/
/oj ow/
/aj aːj aw aːw/
/b t d k kʷ ɡ ɡʷ/
/f s/ [z x]
/m n/
/l r/
/j w/

[Whimemsz]

My friends, I present to you the development of Proto-Bantu consonants in Tswana!

Initially this was going to be a private message to Radius to tie in with something else we were talking about, but it sort of expanded past what I expected and in any case I think it would probably interest a lot of people here so hooray I'm turning it into a full post!

The information here I'm taking from an article by Denis Creissels, "Remarks on the Sound Correspondences between Proto-Bantu and Tswana (S.31), with Particular Attention to Problems Involving *j (or *y), *į and Sequences *NC", in the collection Bantu Historical Linguistics: Theoretical and Empirical Perspectives, ed. Jean-Marie Hombert and Larry M. Hyman (1999).

As background, Proto-Bantu (henceforth PB) is reconstructed with the consonants: */p t ʧ k/, */b d ʤ ɡ/, */m n ɲ/, and the vowels: */i u ɪ ʊ e o a/ (Creissels writes the PB high vowels ⟨ į ⟩ and ⟨ ų ⟩ and the mid-high vowels ⟨ i ⟩ and ⟨ u ⟩, and I've also seen the vowels reconstructed as */i u e o ɛ ɔ a/ -- but the most important point is there were four degrees of aperture). Tswana (henceforth Ts., also Setswana) has the consonant inventory: /p t tɬ ts ʧ k/, /pʰ tʰ tɬʰ tsʰ ʧʰ kʰ qʰ/, /b d ʤ/, /s ʃ χ h/, /m n ɲ ŋ/, /r/ -- plus (though Creissels doesn't view them as separate phonemes) the semivowels /w j/. /d/ has an allophone [l] and some dialects have an additional phoneme /f/. Ts. vowels are /i u ɪ ʊ e o ɛ ɔ a/ according to Creissels. There were also tones in PB and are tones in Ts. but I don't really know the details so whatever.

The normal, unconditioned reflexes of the PB consonants in Ts. are:

• *p → h
• *t → r
• *ʧ → tɬʰ
• *k → χ
• *b → b
• *d → d
• *ʤ → Ø
• *ɡ → Ø
• *m → m
• *n → n
• *ɲ → n

In the case of N+C clusters (or prenasalized stops?) the normal, unconditioned reflexes were:

• *mp → pʰ
• *nt → tʰ
• *ɲʧ → tɬʰ
• *nk → qʰ
• *mb → p
• *nd → t
• *ɲʤ → tɬ
• *nɡ → k

However--and here's where it gets interesting--in a number of cases the following vowel or vowel sequence conditions different changes in the consonants. I'll provide examples of most of these changes since some of them are pretty bizarre (in the examples, recall the Bantu noun class prefixes, which Creissels provides for the Ts. forms but not for the PB forms--in some cases where they are particularly relevant I've provided what I *think* the PB prefix should be...).

For consonants preceding single vowels, there are the following changes:

• *ʧ, *k → s /__(i,ɪ,e) (e.g.: *-kɪ́mbà "dung" → lɪ̀-sɪ́pá)
• *ɲʧ, *nk → tsʰ /__(i,ɪ,e) (e.g.: *n-kíì "eyebrow" → ǹ-tsʰí "eyelash")
• *ɲʤ, *nɡ → ts /__(i,ɪ,e) (e.g.: *n-ɡìɡè "locust" → tsìè)

• *k → h /__u (e.g.: *-kútà "fat" → mà-húrá)
• *nk → kʰ /__u (e.g.: *n-kúdù "tortoise" → kʰúdú)

There are contradictory reflexes of PB *t before a front vowel, but Creissels proposes that r is the regular reflex.

Before sequences of two different vowels, the changes to consonants are often different. In these cases, the initial vowel of the sequence drops following the consonant change. A large number of mergers result (unfortunately Creissels doesn't provide many examples...):

• *p, *t, *ʧ, *k → s /__i+V (e.g.: **-díki- "lead to pasture" → -dís-á)
• *mp, *nt, *ɲʧ, *nk → tsʰ /__i+V
• *(m)b, *(n)d, *(ɲ)ʤ, *(n)ɡ → ts /__i+V (e.g.: -bíad- "give birth" → -tsál-á)
• *m, *n, *ɲ → ɲ /__i+V

• *ʧ, *k → s /__(ɪ,e)+V (e.g.: *-kɪ́á "stump of tree" → sɪ̀-sáná)
• *ɲʧ, *nk → tsʰ /__(ɪ,e)+V
• *ɲʤ, *nɡ → ts /__(ɪ,e)+V

• *p → ʃ(w) /__(ɪ,e)+V (e.g.: *-pɪ́à "new" → -ʃ(w)á)
• *mp → ʧʰ(w) /__(ɪ,e)+V
• *b → ʤ(w) /__(ɪ,e)+V (e.g.: *-béad- "sow" → ʤ(w)ál-á)
• *mb → ʧ(w) /__(ɪ,e)+V
• *d → ʤ /__(ɪ,e)+V (e.g.: *-dɪ̀ànɡò "doorway" → mʊ̀-ʤàkɔ̀)

• *m → ŋw /__(ɪ,e)+V (the following four changes seem to be mostly in diminutive formation, and I'm not entirely sure about them)
• *n → ɲ /__(ɪ,e)+V
• *r → tsʰ /__(ɪ,e)+V (*r from earlier *t)
• *l → ʤ /__(ɪ,e)+V (*l from *d)

• *p, *t, *ʧ, *k → sw /__u+V (e.g.: *-kúá "heritage" → bʊ̀-swá)
• *mp, *nt, *ɲʧ, *nk → tsʰw /__u+V
• *(m)b, *(n)d, *(ɲ)ʤ, *(n)ɡ → tsw /__u+V (e.g.: *-ɡúɪ̀ "arrow" → mʊ̀-tswí)
• *m, *n, *ɲ → ɲw /__u+V (e.g.: *-jámu- "suck the breast" → -áɲ(w)-a)

• *p → ʃ(w) /__(ʊ,o)+V
• *mp → ʧʰ(w) /__(ʊ,o)+V (e.g.: *m-pòí "ostrich" → ɲ̀-ʧʰ(w)é)
• *b → ʤ(w) /__(ʊ,o)+V
• *mb → ʧ(w) /__(ʊ,o)+V (e.g.: *m-bʊ́à "dog" → ɲ̀-ʧ(w)á)
• *m → ŋw /__(ʊ,o)+V (e.g.: *-mòì "one" → -ŋwɪ̀)

[Whimemsz]

My interpretation of some of the Austronesian changes given in Robert Blust's work (cf. and pace Rorschach's earlier post)

Proto-Malayo-Polynesian > Chamorro
(From Robert Blust (2000). "Chamorro Historical Phonology." Oceanic Linguistics 39(1): 83-122)

ə → u
ə → Ø /VC__CV
V → Ø /VC__CV (sporadic)
i, u → e, o /__CC, __C#
i, u → e, o /CC__# (sporadic)
a → æ (in some forms; environment and conditioning unclear)
ui̯ → i (one form) ~ u (one other form)
iu̯ → u (only one example)
p → f
c → s
q → ʔ
k → Ø /__# (sporadically)
k → k /__# (otherwise)
k → h (in other environments)
h → Ø
V₁V₁ → V₁
b, dz → p, ts
d → Ø /__#
d → h
ɟ → ʔ
l → d /__C, __#
ʀ → ɡ
[+obs] → [-voice] /__C, __#
Ø → j /i__a
Ø → w /u__a, a__u, #__V
j → dz
w → ɡʷ
ɡʷ → ɡ /__V[+round]


Proto-Malayo-Polynesian > Proto-North-Sarawak
(from Robert Blust (2002). "Kiput Historical Phonology." Oceanic Linguistics 41(2): 384-438 and Robert Blust (2007). "Òma Lóngh Historical Phonology." Oceanic Linguistics 46(1): 1-53)

q → ʔ
ʔ, h → Ø /#__, V₁[+high]__V₂
h → Ø /__#
h → ʔ /V₁__V₁, a__(i,u)
V₁V₁ → V₁
ə → Ø /V__, __V
a → ə /__%%(+)# (i.e., in prepenultimate syllables)
ə → Ø /__V
ə → Ø /VC__CV
N → [αPOA] /__S[αPOA] (in some words)
N → S₁ /__S₁ (in other words)
C₁[+obs, αPOA] → C₂ /V__C₂[+obs, βPOA]V (in reduplicated monosyllabic roots)
C₁[+obs, αPOA] → C₂ /V__C₂[+obs, βPOA]V (in non-reduplicated bases which had undergone the change of schwa syncope in medial syllables)
C(?) → [+long] /ə__V
ɟ(ː) → d(ː)
bː, dː, dzː, ɡː → bʰ, dʰ, dzʰ, ɡʰ (voiced stops with voiceless releases, i.e. [bb̥, dd̥, dz̥, ɡɡ̊] or [bp, dt, ds, ɡk], but treated as unit phonemes, not clusters)
ə → Ø /#__%%(+)# (i.e., in word-initial position in prepenultimate syllables)
pː, tː, cː, kː → [-long]
c → s


Proto-North-Sarawak > Kiput
(from Robert Blust (2002). "Kiput Historical Phonology." Oceanic Linguistics 41(2): 384-438)

Primary stress → Ultimate syllable
ʔ → Ø /V__V
k → Ø /V__V (in some forms)
ai, au → ai̯, au̯ /__#
ə → a /__ʔ#
ai, au → ɛː, ɔː /__...#
i, u → əi̯, əu̯ /__#
Ø → h /a__#
s → Ø /V__V (sporadic)
V₁V₁ → V₁
ə → Ø /V__, __V
V[+stress (!ə)] → [+long] /__C[!h]# (applies to diphthongal nuclei as well a monophthongs)
ʀ → l (non-finally, in some words)
ʀ → ɾ (non-finally elsewhere)
l → Ø (non-finally, irregularly)
u → əw /__V (but also cases of (C)u → w /__V)
i → əj /__V (but also cases of (C)i → j /__V)
w → v
j → ɟ
iu̯ → ui̯
s → Ø /__#
i, u → ɛ, ɔ /__C[-lab]# (sporadic)
ʀ → ʔ /__# (in a handful of forms)
ʀ → ɾ (elsewhere)
a → i /C[+obs +voice]...__(C)# (blocked if there was an intervening nasal, and sometimes if there was an intervening voiceless stop or liquid)
i, u → əi̯, əu̯ /__(ʔ)#
ə → a /__i̯, __u̯ (unless a voiced obstruent appears earlier in the word)
bʰ, dʰ, dzʰ, ɡʰ → f, s, s, k
f → s
v, ɟ, ɡ → f, c, k /V__V
v, ɟ → f, c /#__
i, u, ɛ, ɔ → iə̯, uə̯, iə̯, uə̯ /__k#, __ŋ# (and also sporadically before final *t and *n and some other consonants)
k, ŋ → ʔ, Ø /Və̯__#
ə → ə~a /__C# (free variants)
NS[-voice] → S[+long]
Numerous different possible reflexes of N--S[+voice] clusters
(C)V → Ø /#__C... (irregular)
n → l /#__
ə → Ø /#__
l, ɾ → n /__#
dz → d /#__ (but → s in a few forms and → ɟ in a few others)
s, c → t /__VsV, __VcV
b, d → p, t /__#


Proto-North-Sarawak > Proto-Kenyah
(from Robert Blust (2007). "Òma Lóngh Historical Phonology." Oceanic Linguistics 46(1): 1-53)

d → l /#__ (sporadic)
ʀ → h /V__V(C)#, __#
ʀ → Ø
s → h /__#
i, u → e, o /__h#
h → Ø /__#
S → [-voice] /__#
l → n /__#
s → t /__VsVC
CV → Ø /__NCVC (in reduplications)
N → [αPOA] /#__C[αPOA]


Proto-Kenyah > Òma Lóngh
(from Robert Blust (2007). "Òma Lóngh Historical Phonology." Oceanic Linguistics 46(1): 1-53)

bʰ, dʰ, dzʰ, ɡʰ → p, t, c, k
i → e /__k#
i → iə /__ŋ#
p → k /u__#
u → o /__k#
u → o /__ŋ# (sporadically failed to occur)
u → ɯ /__(C)# (but not before final ʔ)
a → o /__# (not in all forms)
ʔ → Ø /__# (but aʔ → əʔ in some forms)
k → ʔ /__#
a → ɛ /{t,n}__#
p, t → c̚ /i__#, ɛ__#
m → ɲ /i__#
n → ɲ /i__#, ɛ__#
ai̯, au̯ → ɛ, ɔ
ui̯, iu̯ → e
i, u → e, o /__CV[+closemid](C)# (iə is treated as close mid for this change)
i, u → ɛ, ɔ /__CV[+openmid](C)#
i, u → i, u /__CV[+close (!ɯ)](C)#
i, u → ɛ, ɔ /__CVC#
i, u → e, o /__Cɯ#
ŋ → ŋ̊ /__#
p, t → k /ɯ__#
m, n → ŋ /ɯ__#
i → əj /__V(C)#
u → əw /__V(C)#
j, w → z, v
p → f /#__
ə → Ø /#__
p, k → p, k /ə[+stress]__V
p, k → f, ɣ /V__V (elsewhere)
d → r /V__V (irregular)
dz → ɟ /V__V
b, d, dz, ɡ → p, t, c, k /N__
N → Ø /__S (sporadic)
h → Ø /V__V

[Note that the above changes mean that Òma Lóngh has a system of vowel harmony! Any vowel can cooccur with itself in the same word, and either of the neutral vowels /a ə/ can cooccur with any of the other vowels. However, there is otherwise a constraint that restricts the vowels in a word to all be of the same height: /i u/ can cooccur, as can /e o/, as can /ɛ ɔ/, but none of the vowels from any of these groups can cooccur with any member of a different group (excepting, of course, the neutral vowels)]

[----]

Has anybody found anything for Proto-Mon-Khmer to Vietnamese?

[Pogostick Man]

Has anybody found anything for Proto-Mon-Khmer to Vietnamese?

I've been working (or at least meaning to) on some Vietnamese stuff for here for awhile; not from Proto-Mon-Khmer, but still it hopefully might be useful a little bit. Now I have a reason to finish it.

On a related note…if I were to create a PDF of the Library, under what license would its release be required? I'm aware of some of the licensing requirements due to KneeQuickie but would like some advice.

[Herr Dunkel]

Has anybody found anything for Proto-Mon-Khmer to Vietnamese?

I've been working (or at least meaning to) on some Vietnamese stuff for here for awhile; not from Proto-Mon-Khmer, but still it hopefully might be useful a little bit. Now I have a reason to finish it.

On a related note…if I were to create a PDF of the Library, under what license would its release be required? I'm aware of some of the licensing requirements due to KneeQuickie but would like some advice.

Open source sharealike, I think.

[Tropylium]

A bit old, but…

Old Indo-Aryan to Something Middle Indo-Aryan
* {β,ð,ɣ} → h / V_V
* aħ → o
* kʂ → tstʃ

You probably mean *bʰ *dʰ *gʰ (<bh dh gh>) for the first, *ah (<aḥ>) for the 2nd, and *tːʃ (<cch>) for the 3rd?

* j w → dʒ b / V_V
* C → C[+voiced] / V_V
* {j,v} → Ø / V_V
* b {d(ʒ),g} → v j / V_V

Where do those second /j v/ come from? And *j → *dʒ → j seems kind of unnecessary — is this out of order maybe? (*g → *j first, then *j → dʒ could make sense…)

[Nortaneous]

Here's an attempt at Late West Brythonic -> Middle Welsh, from here (pdf). These probably aren't in chronological order because this pdf is fucking shit about organization, and when chronology is given, it occasionally contradicts itself.

Symbols: (some of these are nonstandard because, if there is any standard notation for them, I don't know it)
N = nasal
L = liquid
P = plosive
C = consonant
Q = any cluster
+ = syllable break (used for umlaut)

VOWELS:

Late West Brythonic vowel system:

Code: Select all

iː i       ʉː u
eː e (ə ɵ)    o
ɛː         ɔː
      a

Late West Brythonic to 6th C.
i > ɨ
o > u / _N _NP _LP

New quantity system - ~6th C.
In stress-bearing monosyllables:
* V > Vː / _C# EXCEPT _m#
* Vː > V / _Q# _m#
* Word-final voiceless stops did not exist in inherited vocabulary, but when they emerged, vowels before them were short.
* Unclear whether vowels were short or long before /ɬ/.

6th C. to ??? - I can't tell whether this happened before or after new quantity system
ɔː > ɔ / pretonic syllables
eː ɛː > uɨ oɨ
ɔː > au / stressed syllables

A-affection and i-affection - common to Welsh, Cornish, and Breton, a-affection happened first
A-affection:
* i u > e o / stressed, penultimate syllables with a aː in next syllable

I-affection, first stage: (all changes occur in penultimate syllables)
* a > ei / _+j
* a > ei or ɨ / _+iː
* o > ei or ɨ / _+j
* e u > ɨ / _+{iː j}
* "With diphthongs, more complex results obtain."

I-affection, second stage -- internal i-affection: (affects vowels in all syllables, applies recursively)
* a o > e / _+{i iː}
* a e o > ei / _+j

ɔ > o

Early Welsh vowel system:

Code: Select all

i  ɨ ʉ  u
e (ə ɵ) o
    a

----------------------------

Changes dated to before Old Welsh accent shift: (~9th C.)
o > ɵ (> ə?) / unstressed prefixes
u ɨ > ɵ ə > ə / nonfinal syllables (except in some dialects? -- see IRC log below)
0 > ə / _sP (remains productive into Middle Welsh)
ei eʉ > ai aɨ / stressed syllables
nt > n / _# except stressed syllables

Accent shift:
Stress accent shifted to penultimate syllable in polysyllables; pitch accent disassociated and remained on final.

Changes dated to after the accent shift:
Nh > N / everywhere except onset of stressed syllable
au > o / posttonic syllables

----------------------------

Contrastive vowel length re-emerges:
nn rr > n r
ɣ > 0 / C_#

In northern varieties:
* V > V: / _sP _ɬt (in monosyllables)

Middle Welsh period:
ʉ > ɨw / _x in monosyllables
ʉ > ɨ

0 > ə / C_{r l n} {l r}_{m v} ð_v
* "In some cases the epenthetic vowel assimilates to the vowel in the previous syllable."

----------------------------

CONSONANTS:

Old Welsh consonant inventory:

Code: Select all

p b t d k g gʷ
m̥ m n̥ n ŋ ŋ̥ 
ɸ β θ ð x ɣ
  β̃ 
    s
    ɬ l j   w
    r̥ r

β > 0 / _u u_ and occasionally _ʉ ʉ_ (in Middle Welsh)
β̃ > β (~12th C.)
ɸ β > f v

initial consonant mutation, I honestly can't be bothered

Voiced fricative fuckery:
ɣ > 0 ("over an extended period from the sixth to the ninth century")
v > 0 / au_# in monosyllables, ʉ_#, _# sporadically (note: 10:37 < Yng> final v loss is pretty much universal)
ð > 0 / V[+high -round]_#, _# sporadically, also sporadically in other positions, "variability remains in many of these items to this day"
0 > ð / V[+high]_V[+high] (in southern varieties, "in the latter half of the Middle Welsh period")

Development of word-final glides:
g > x > j > ə̯ / {r l}_ (g > x dated to Brythonic -> Old Welsh, x > j dated to Middle Welsh)
ə̯ > u / u+_
ə̯ > either a or 0
w > u / C_# in monosyllables
w > 0 / C_# in polysyllables, sometimes with metathesis (e.g. gwarchadw > gwarchawd) (c. 15th C.)

s > ʃ / _i
d > dʒ / _i in Pembrokeshire?

Fortition in southeastern dialects: (from Yng)
P[+voice] > P[-voice] / onset of post-tonic syllable -- e.g. 'gwybod' is [gwɨbod] in standard but [gupɔd]~[gupɔt] in those dialects

x > h / #_w (Middle Welsh, in southern dialects)
h > 0 / #_w (modern, southern dialects -- Yng says this is a specific regional thing accompanied by voicing of n̥ m̥ ŋ̥ r̥ -- "the dialect losing h")

---- edit: some information on Modern Welsh ----
10:26 < Yng> chwarae > hwarae > warae
10:26 < Yng> what else happens in southern dialects
10:26 < Yng> ɨ and i merge
10:27 < Yng> that thing that happens with schwas in northern dialects has various different results across the rest of Wales
10:27 < Yng> some southern dialects develop an [æ~ɛ]
10:27 < Yng> where others have a long a
10:27 < Yng> pretty much all dialects collapse diphthongs in various ways
10:27 < Yng> in fact
10:28 < Yng> I have a feeling that the schwa thing of u ɨ does not in fact happen across Wales as a whole, only in northern dialects
10:28 < Yng> because in most southern dialects you have a realisation somewhere around /i/
10:29 < Nortaneous> what schwa thing?
10:33 < Yng> u ɨ > ɵ ə > ə / nonfinal syllables
10:33 < Yng> because that produces words like, uhh
10:33 < Yng> <tyfu> /təvɨ/
10:33 < Yng> but!
10:34 < Yng> in a lot of dialects that's realised as like
10:34 < Yng> /tɨvɨ/ or /tivɨ/
10:34 < Yng> although this might be a later shift because original isn't reflected
10:35 < Yng> there's also
10:35 < Nortaneous> do u 1 still merge in that position?
10:35 < Yng> let me see if I can find some examples of caleta
10:35 < Yng> yeah

10:46 < Yng> there's also a ton of examples of voiced fricatives disappearing before consonants, often with compensatory lengthening
10:46 < Yng> <colofn> /kɔlo:n/
10:47 < Yng> that or consonant clusters being reduced or having epenthetic vowels inserted
10:47 < Yng> <ffenestr> is usually /fɛnɛst/ but I guess could be /fɛnɛstɨr/ with varying qualities of epenthetic vowels
10:48 < Yng> Another aspect, peculiar to this region, is the tendency to not use the dark y sound; in its place we frequently see i e.g. cifan = cyfan (all), finy = i fyny (up), i mochyn = y mochyn (the pig) , and sometimes, w e.g. cwmeryd = cymryd (to take), and cwmwdog = cymydog (neighbour). (source)
10:49 < Yng> the dark y sound is the [ə] in non-final syllables
10:49 < Yng> so it looks like [u ɨ] might NOT have merged after all

10:51 < Yng> oh yeah and some dialects have [dʒ] in what I think must be native words
10:52 < Yng> pembrokeshire for 'escape' is [dʒɛn.ɡid]
10:52 < Yng> which is dianc in standard Welsh, but also has an archaic alternative verbnoun in diengyd

11:04 < Yng> how does my dialect handle diphthongs
11:04 < Yng> I think it's something like
11:04 < Yng> [ai ae] > [ai] unconditionally, I don't think anyone retains those as separate
11:06 < Yng> [ai aɨ] > [ɛ] _# in polysyllables, [eː] in the tonic syllable in polysyllables, [ɑː] in monosyllables
11:06 < Yng> so blaen [blɑːn], blaenau [bleːnɛ]
11:07 < Yng> oh also [ɛ] in final syllables generally, I think: cerddoriaeth [kɛrðɔriɛθ]
11:09 < Yng> except
11:10 < Yng> final stressed syllables, which result from hiatus in old Welsh, are often immune to this change
11:10 < Yng> so casáu and Cymraeg [ka.ˈsaɨ kəm.ˈraiɡ]
11:10 < Yng> not always immune though, you get all sorts of realisations of Cymraeg:
11:11 < Yng> the seminal work on Welsh dialects is called Cymraeg Cymrâg Cymrêg

[Herr Dunkel]

Voiced fricative fuckery:
ɣ > 0 ("over an extended period from the sixth to the ninth century")
v > 0 / au_# in monosyllables, ʉ_#, _# sporadically
ð > 0 / V[+high -round]_#, _# sporadically, also sporadically in other positions, "variability remains in many of these items to this day"
0 > ð / V[+high]_V[+high] (in southern varieties, "in the latter half of the Middle Welsh period")

By the Seven Thanes of Phoneticks, what fuckery is this?

[Nortaneous]

/G/ disappears in all positions, other voiced fricatives drop occasionally word-finally, /v/ drops after some high rounded vowels, /D/ drops after some high unrounded vowels, epenthetic /D/ inserted between two high vowels

duh

[Herr Dunkel]

/G/ disappears in all positions, other voiced fricatives drop occasionally word-finally, /v/ drops after some high rounded vowels, /D/ drops after some high unrounded vowels, epenthetic /D/ inserted between two high vowels

duh

Duh

But what fuckery is this?

[Jetboy]

Anyone have something to the tune of the vowel changes between PIE and Shakespeare?

[jmcd]

Wikipedia has this. There are other things out there though considering how well studied the language is.

[Whimemsz]

Proto-Malayo-Polynesian > Palauan
(From Robert Blust (2009), "Palauan Historical Phonology: Whence the Intrusive Velar Nasal?" Oceanic Linguistics 48(2): 307-336)

I've tried my best to arrange these in chronological order, but since the paper isn't organized chronologically I may have screwed up somewhere


{aj,aw,uj} → {e,o,i}
Ø → w /u__V
Ø → j /i__V
ə → Ø /#__
Ø → ə /C__C (for certain consonant combinations, which the paper doesn’t specify)
h → Ø
ə → o ~ e /[+stress] (result of /e/ or /o/ unpredictable)
p → w
wa[+stress] → wa /#__
wa[-stress] → o /#__
{aw,əw} → o /__#
V → ə /[-stress] (except /o/ arising from earlier #pa- sequences)
wə → u /#__
j → r
l → j
ə → Ø /__j
ji → i /C__
ə → Ø /__#
ə → Ø / (sporadic)
t → ð /except when adjacent to another stop
s → t
ʀ → r /__C[+dental]
ʀ → s
d → r
ɖ → r (only one example)
ɲ → n (but → ŋ initially?)
n → l
dz → r
rl → ll
ɟ → k /__C#
ɟ → s
ŋ → Ø /C__#
t → s ~ ð /unpredictably, to eliminate sV(C)t and tV(C)s sequences
Ø → ŋ /#__V
q → χ
χ → ʔ (within the historical period -- hence /ʔ/ is still spelled <ch> in the practical orthography)

The addition of a velar nasal to all vowel-initial words (the third from last change in the list) is definitely the most interesting and unusual. There is another example of the same change, incidentally, in the Northern Samoyedic languages (see this post). Blust also argues convincingly that the Palauan change cannot be due to anything like fusion with an old nominal proclitic (e.g. a demonstrative or other noun marker), but is instead a genuinely phonological change.

[Whimemsz]

Okay. Jeff posted a preliminary list of Proto-Algonquian to Cheyenne changes to kick this thread off many years ago. I've been trying to compile my best effort at a more complete and accurate list, from a few different sources, but principally Ives Goddard's chapter "Pre-Cheyenne *y" in In Honor of Mary Haas: From the Haas Festival Conference on Native American Linguistics (unfortunately I do not actually have the book and am going on what Googlebooks and Amazon's previews will show me, so...this is still rather tentative). I've done my best to put them in roughly chronological order (Algonquianists aren't big on framing these things chronologically).

PROTO-ALGONQUIAN > CHEYENNE

we → o
e → i /#__
w → Ø /{s,h}__i
w → j /C__
j → r /V__, #__
V[short] → [long] /#C₀V[short]C₀__NC, #C₀__NC(j)V# (but e → i: /{θ,r,ʃ}__) [that is, vowels are lengthened in the second syllable of a word before a N+C cluster if the vowel in the first syllable is short, and in the first syllable of disyllabic words]
Ø → j /k__e(:), {r,θ,ʃ}__e:
o(:) → i(:)
V[-long] → [low tone]
V[+long] → [high tone]
a → o
e → ɪ /j__
e → a
i → ɪ
n → h /__C
θ → r
hr → h
hʧ, hs, hʃ → ʧ, s, ʃ
ʔs, ʔr → ʔh
s, ʃ, r → ʔ /__S
s → h
ʃ, r, t → s /{V,#}__j
{p(j),k(j),h,m(j),n(j),w}V → Ø /V__# (in words of 3+ syllables) [i.e., “final syllables are lost if they begin with a PA *p, *k, *s, *h, *m, *n, or *w whose reflex in Cheyenne would have followed a vowel, or if they would have begun with Ch h reflecting a cluster”]
Ø → h /#__V
t → ht /V__
p → Ø /__(j)V (sometimes)
p → hp /V__V
k → Ø /C__(j)V, V__(j)V (sometimes)
k → hk /V__V
r → t
j → Ø /C[!h]__
j → n
h → Ø /ʔ__n
ʧ → s
ʃ → x
w → v
Ø → oʔ /V__ɪ#
Ø → V₁ʔ /V__V₁#
Ø → h /ɪ__tɪ#
h → s /ɪ__t
h → ʃ /ɪ__k
x → s /__ó
x → ʃ /ɪ(ʔ)__{C,#}, {C,#}__(ʔ)ɪ, ɪ(ʔ)__(ʔ)ɪ [that is, when “the only flanking vowel, or both flanking vowels (with or without an intervening ʔ) are” ɪ]
t → ts /__ɪ
A series of vowel devoicing and tone sandhi rules which conspire to create four (or five) surface tones [these are almost impossible to write in formal rules, and they definitely are impossible to write in legible formal rules, sorry]

[dhok]

Does anybody happen to have Proto-Germanic to Old Norse, and/or , for an encore, Old Norse to any of its five extant descendents?

[clawgrip]

I can give some info on Japanese.
It's clear that Old Japanese had certain distinctions that Modern Japanese does not, but because Japanese writing has always been syllabic, it's still unclear if this variation was a greater number of vowels (8 vs. modern 5) or a greater number of consonants (palatalized vs. unpalatalized is a theory). Either way, these distinctions collapsed, resulting in plain consonants and 5 vowels.

Other changes:

Somewhere in Old Japanese (not sure if all of these are entirely proven)
ia → je
io → je
ai → e (origin of things like 目 me (me + *-i) vs. 瞼 mabuta
ui → wi (variations of this type of have been regularized)
oi → wi (origin of things like 木 ki (ko + *-i) vs. 木の実 konomi
oi → e (origin of things like 殿 tono + 入 iri vs. 舎人 toneri
uo → wo
ua → wo



Old Japanese to Middle Japanese
p → ɸ

Middle Japanese to Modern Japanese
phase 1
ɸ → w / V__V (there are only two exceptions (that I know of): haha and ahiru)
ɸ → h / #__a, #__i, #__e, #__o
ɸ → ɸ / #__ɯ

phase 2
j → Ø / __e
w → w / __a
w → Ø / __i, __e, __o (except in one word, wo, and only when spoken in a deliberately clear way)

au → oː
ou → oː
ao → oː
oo → oː
eu → joː
io → joː
iu → juː

[Aurora Rossa]

phase 2
h → w / V__V (there are only two exceptions (that I know of): haha and ahiru)

I would think this change happened before /ɸ/ > /h/ since it seems more likely for /ɸ/ to become /w/ through an intermediary like /v/ than for /h/ to labialize.

[clawgrip]

Yeah, that makes sense. I'll update it.

[Herr Dunkel]

w → w / __a

/w/ oft changes into itself, even outside enviroments.

[clawgrip]

You missed ɸ → ɸ / #__ɯ

I recognize the redundancy, and only included those two to point out the single place where each sound didn't change. Is there a better way to indicate "no change only in this environment" or is it better to just leave these two out and let people figure it out themselves?

[finlay]

phase 2
h → w / V__V (there are only two exceptions (that I know of): haha and ahiru)

I would think this change happened before /ɸ/ > /h/ since it seems more likely for /ɸ/ to become /w/ through an intermediary like /v/ than for /h/ to labialize.

Why do you always demand intermediaries? ɸ → w direct is just as likely. If you're desperate to rationalise it, think of ɸ as the voiceless counterpart of w. Turning to v first is less likely because you're changing the POA to labiodental and back again. Also, SQUARE FUCKING BRACKETS ALREADY.

As for the change itself, what about 'nihon'? Or did that only start getting used later?

[clawgrip]

These changes don't generally apply between morpheme boundaries. Since Nihon is actually ni + hon, the /h/ is still initial and so has not been changed.

Nevertheless, there are certain compounds that have undergone the change. The ending -hara often changes to -wara because of this rule, e.g. Fujiwara, Yoshiwara. Also 八幡 can be Yahata or Yawata.

There are also some archaic compounds that have undergone the change:

目方 [mape] → [maɸe] → [mawe] → [mae] 前
今日 [kepu] → [keɸu] → [keu] → [kjoː] 今日

I totally forgot to add the vowel mergers that result from loss of intervocalic /h/. I'll have to add them in.

[Herr Dunkel]

Nippon and Nihon are varieties of the same (that is, I presume originally *ni-ppon and *ni-pon), I believe.

[Pole, the]

Nevertheless, there are certain compounds that have undergone the change. The ending -hara often changes to -wara because of this rule, e.g. Fujiwara, Yoshiwara. Also 八幡 can be Yahata or Yawata.

Are there any cases when the ending happens after /ɴ/? What then?

Where did the geminates come from?

Also I had a personal theorem regarding the vowels, dunno whether to show it.