The Correspondence Library

jburke · Post by **jburke** » Thu Sep 18, 2003 8:09 am

Here's an idea: since ZBB posters are, collectively, familiar with a wide variety of languages, why don't we start compiling sound change and correspondence lists for the languages we know? This way, when a conlanger is basing his work on a certain language family, and wants to know how sounds tend to change in that family, he can consult these lists for inspiration. This would be a great resource, and would save much time for those who don't have ready access to an academic library.

I'll start off with a comprehensive (_not_ exhaustive) list of changes from Proto Algonquian > Cheyenne, taken primarily from Bloomfield and Leman:

PA=Proto Algonquian; Ch = Cheyenne

Vowels:

PA /a/ > Ch /o/
PA /o/ > Ch /e/
PA /i/ > Ch /e/
PA /e/ > Ch /a/

Consonants:

PA /t/ > Ch /ht/
PA /l/ > Ch /t/
PA /th/ > /t/
PA /tl/ > Ch /t/
PA /p/ > Ch /hp/ in a few cases; mostly, it dropped out
PA /k/ > Ch /hk/ in a few cases; mostly, it dropped out
PA /ch/ > Ch /s/
PA /y/ > Ch /t/ or /e/
PA /w/ > Ch /v/ or /o/
PA /sh/ > Ch /sh/ or /x/
PA /s/ > Ch /h/

Consonant clusters:

All clusters involving /k/ > ?
In nasal + non-nasal clusters, the nasal survived; in nasal+nasal clusters, /m/ dropped out
In semivowel + nasal clusters, the nasal survived
In /p/ clusters, the /p/ either dropped out or changed to a /t/
In /t/ + semivowel clusters, the /t/ survived

Dudicon · Post by **Dudicon** » Thu Sep 18, 2003 9:17 am

I like this idea. I can list right now a set of comprehensive (also, possibly not exhaustive) set of changes from proto-Poylnesian to Hawai'ian--note that there don't seem to be any conditioned changes involved, and also no changes to the vowels, which is remarkable.

h

0
?

0
s

h
f

h
k

?
t

k
ng

n
r

l
v

w

Well, I guess Hawai'ian must just be very conservative; my source is Trask.

While we're at it, why don't I include the Greek vowel shift , also from Trask (my only readily available historical book right now), although I certainly can't speak for the rest of Greek.

u

y
u:

y:
o:

u:
e:

i:
E:

e:
ai

E
Oi

y:
e:

i:
y

i
y:

i:
V:

V
Eu

Ev
au

av

jburke · Post by **jburke** » Thu Sep 18, 2003 9:41 am

note that there don't seem to be any conditioned changes involved

Unless you count the cluster changes, there are no conditioned changed in PA > Cheyenne either.

I guess Hawai'ian must just be very conservative

Miami is one of the most conservative Algonquian languages. The PA preaspirates passed into Miami basically intact; and some morphemes are exactly the same as in PA (though the head marking system has changed, and so the prefixes are different). Miami retains most of the PA consonant clusters.

Sapir theorized that the further you get away from the original homeland of a proto language, the more conservative the daughters become. This would mean that Midwestern languages like Miami are far from the original homeland, and that the PA homeland should be properly located in western Idaho. This is the running theory at the moment (though some still believe tha PA homeland was in southern Canada, around Georgian Bay.)

Dudicon · Post by **Dudicon** » Thu Sep 18, 2003 10:48 am

jburke wrote:Sapir theorized that the further you get away from the original homeland of a proto language, the more conservative the daughters become. This would mean that Midwestern languages like Miami are far from the original homeland, and that the PA homeland should be properly located in western Idaho. This is the running theory at the moment (though some still believe tha PA homeland was in southern Canada, around Georgian Bay.)

That seems almost counterintuitive--it would seem more likely that languages closer to their place of origin would be more conservative. However, it does agree with the evidence from Hawai'ian as well--Hawai'ian is, I believe, the furthest eastward Polynesian language.

Post by **zompist** » Thu Sep 18, 2003 11:28 am

Dudicon wrote:
jburke wrote:Sapir theorized that the further you get away from the original homeland of a proto language, the more conservative the daughters become. This would mean that Midwestern languages like Miami are far from the original homeland, and that the PA homeland should be properly located in western Idaho. This is the running theory at the moment (though some still believe tha PA homeland was in southern Canada, around Georgian Bay.)
That seems almost counterintuitive--it would seem more likely that languages closer to their place of origin would be more conservative. However, it does agree with the evidence from Hawai'ian as well--Hawai'ian is, I believe, the furthest eastward Polynesian language.

Nah-- it's true of most language families.

-- Central Quechua is more varied and has changed more than Peripheral Quechua
-- Mandarin has proceeded further than the southern Chinese dialects in losing syllabic distinctions and simplifying tone
-- American English has some significant retentions of older words and sounds compared with RP
-- in the Spanish-speaking world, coastal regions (historically, connected by fast sea travel) share some recent sound changes, such as loss of final -s and medial post-stress d
-- within Romance, the rather peripheral Sard is the only language to have escaped the palatalization of /k/; the longer retention of cases in French and Romanian is also interesting
-- within Germanic, Icelandic is one of the most conservative languages

Now, all this shouldn't be over-generalized. It's never true, for instance, that peripheral languages are unchanged. All languages are subject to sound change. And of course some peripheral languages do change a lot-- French within Romance, for instance.

(Sorry to get off-topic; I'll try to remember to post my Quechua sound changes.)

Glenn · Post by **Glenn** » Thu Sep 18, 2003 11:57 am

Dudicon wrote:That seems almost counterintuitive--it would seem more likely that languages closer to their place of origin would be more conservative. However, it does agree with the evidence from Hawai'ian as well--Hawai'ian is, I believe, the furthest eastward Polynesian language.

It may indeed seem counterintuitive, but one way to think about this phenomenon is that populations migrating outward often retain and carry with them "older" features of the original language, which are later lost in the place of origin. American English is one example of this. Another comes from my own background: my fathers' ancestors were Germans (specifically Swabians from the Duchy of Wurttemberg) who emigrated to the Russian Empire in the late 1700s and early 1800s, and who later came to the United States (c. 1900), settling in the region of the Dakotas. The German dialect spoken by those immigrants to the U.S. (Central Dakota German) apparently retained some archaic 18th-century features lost in standard German, as well as adding some innovations of its own.
(My father owns a book on Central Dakota German, which describes some of the features of that language. I don't speak Dakota German myself, but I had a great-aunt who did.)

In Guns, Germs, and Steel, Jared Diamond describes a related phenomenon: the languages spoken at the original core of a language family often show greater fundamental diversity than those resulting from far-flung expansion; there may be a large number of peripheral languages, but they all descend from a small number of "migrating" ones, often within the same subfamily, and thus share many of the same basic features. According to Diamond, this phenomenon has been used to show that the Polynesian languages spread south and east from an original homeland in Taiwan (and perhaps southeast China), and the Bantu languages, from what is now southern Nigeria and Cameroon.

On a somewhat different note, Mark also noted in the LCK that the dialect spoken in a capital city or other centers of culture often changes more quickly, while dialects of outlying areas remain more conservative.

I don't want to distract from the thread either, however, particularly as I have little to add to the main theme myself. Does anyone else have sound changes to offer?

p@,
Glenn

So Haleza Grise · Post by **So Haleza Grise** » Thu Sep 18, 2003 5:35 pm

I remember a list of sound changes affecting Proto-Tangkic in the transition to Lardil, Yukulta, and Kayartilt, but I'm not sure I'll be able to find that anytime soon. It was hard enough to track it down the first time.

Ran · Post by **Ran** » Thu Sep 18, 2003 5:50 pm

Glenn Kempf wrote:On a somewhat different note, Mark also noted in the LCK that the dialect spoken in a capital city or other centers of culture often changes more quickly, while dialects of outlying areas remain more conservative.

The more I read into this matter, the more it seems that all major cities are automatically linguistic islands.

I don't want to distract from the thread either, however, particularly as I have little to add to the main theme myself. Does anyone else have sound changes to offer?

p@,
Glenn

I agree. I'm following this conversation with great interest, but I think it should be moved out to another thread.

Old Mandarin (~1300) to modern Pekingese

(From Hsueh's Phonology of Old Mandarin)

ng > 0 /#_
e, o > a /_w
m > n /V_
o > e /_ng
e, o > a /_n
i > e /C_?
? > 0
y > 0 /y + ?_ (y = /j/)
#r + ?# > #? + r#
v > 0
y > 0 /Pr_ (Pr = retroflex consonant IIRC, nothing to do with bilabials)
o > e /_#
? > e /_E (E = ending consonant) except _r
y > 0 /C_w + e + ng
k, kh, h > c, ch, s /_y (c, ch, s are /t?, t?h, ?/)

It goes without saying that this reflects just one interpretation of Old Mandarin phonology. It is also far from exhaustive, since the author intended the list to be an "afterthought" to his book.

Mecislau · Post by **Mecislau** » Thu Sep 18, 2003 6:14 pm

Does anyone know about sound changes in Russian? I need to know for a conlang I'm starting on

Post by **zompist** » Thu Sep 18, 2003 10:15 pm

Quechua sound changes

Not final by any means...

Ayacucho

?

ch
sh

s
q

[X] (uvular fricative)

(phonetic interpretation of proto-Quechua ? is uncertain; there was a distinction between ch and ? though)

Cuzco

ch

ch' (word-initially, 50% of cases... otherwise

ch)
?

ch
sh

s
m

n / _[+dental | +velar]

p

[F] / _[+stop] F = bilabial fricative
t, ch, ?

s / _C
k

x / _C
q

X / _C
ll

l / _q

0

h / #_V

Rare or incipient changes:

?

n (4 of 36 cases)
k

h / V_V (3 of 47 cases)
q

h / V_V (2 of 42 cases)
w

y / _$ (3 of 18 cases; $ = syllable boundary)

Cajamarca

?

tr
h

0 / #_
q

k
ll

Z / V_V

Jun?n

?

tr
r

l
q

0 / #_
q

? (glottal stop) elsewhere
s

h / #_ (about 35% of the time; but 58% of cases before a)

Ancash (Huaraz)

?

ch
ch

ts (word-initially, 2/3 of the time; otherwise

ch)
s

h / #_ (about 45% of the time)
s

0 / [+labial]_V
s

y / V_i, i_V
s

w / u_V
s

h / other V_V
?

n / #_
ll

l / #_ (about 18% of the time)
ay

ee
uy

ii
aw

oo
q

X / V_[-voiced]
q

[+voice] / *_[+voiced] (* = any phoneme)

Ran · Post by **Ran** » Sat Sep 20, 2003 12:24 am

Late Middle Chinese (~1000) to Old Mandarin (~1300)
(Hsueh's Phonology of Old Mandarin)

The author uses V1, V2, V3, V4, Vn, Vch, and Vta to represent Late Middle Chinese vowels. I am going to very tentatively assign the values of o, a, ia, e, ?, a(ch) and a(ta) to these vowels. These should be taken as orthographical convenience rather than actual speculation.

I am also going to number tones according to their traditional order, i.e. F1, F2, F3, F4, F5. Since tones change rapidly, it is impossible to accurately reconstruct their values; we can only know how many there were, and agree on an order to renumerate them. Middle Chinese starts out with no F2.

A few more things:

H = a voiced /h/ and a breathy element across the syllable
y = palatalization
w = labialization
j = palatal articulation of previous consonant
r = retroflex articulation of previous consonant

These (H, y, w, j, r) will be treated as separate phonemes in the sound change list; note that they are not.

Middle Chinese phonology

vowels
V1 (*o)
V2 (*a)
V3 (*ia)
V4 (*(i)e)
Vn (*?) can be palatalized (*y?)
Vta and Vch - ???

initial consonants
p ph pH m
t th tH
n l
ts tsh tsH s sH
tj tjh tjH
tsr tsrh tsrH sr srH
tsj tsjh tsjH sj sjH nj(nr)
k kh kH x xH ng
? 0

final consonants
y w m n ng p t k

tones
F1 F3 F4 F5
(F5 exists if and only if in syllables ending with stop)

List:

0 > w / bilabial_V
xH > 0 / _y(w){? or ia or a(ta)}
y > 0 / retro(af)fricat_
j > r
n > 0 /_r (retroflex n loses its nasality; what remains is a retroflex approximant)
F1 > F2 / syllables with H or liquid(incl. nasals)
F3 > F4 / syllables with H
? > 0
F5 > F2 / syllables with H
F5 > F4 / syllables with liquid(incl. nasals)
F5 > F3 / all other syllables
H > h / stop _ toneF2
H > 0 / all other cases
y > 0 / m_w + ?{w or k or ng}
p(h) > f / _y+w
m > v / _y+w
y > 0 / {f or v}_
w > o / y_? + w
ia > a / C(w)_
ia > e / _
0 > y / velar/pharyn. _ {a or a(ch)}
o > a / C_C as long as ending C is not w, and if beginning C is a velar/pharyn. then the ending C cannot be p,t,k
ng > n / n_y + a(ta)
ng > ng / #_o(w)
ng > 0 / #_
k > y / {o or a or e}_
k > w / V_
o, a > e / _ng
a(ta), a(ch) > a / _ng; o otherwise
? > o / _ng
y > 0 / alveolar.(af)fric or retroflex _? + y
y > 0 / C + ?_
t > ts / _r
p, t > y / y + ?_
p, t > 0 / V_
o > a / y_w
0 > w / C_o# as long as C is not a velar/pharyn.
w > 0 / retroflex + y_? + w ---- optional
w > 0 / w + ?_
? > a / C + w_y
o, e > ? / _y
y > 0 / _w + ? + y
0 > w / retrof._a + ng
e > o / w_ng ---- optional
y > 0 / _w + a +ng
y > 0 / retro_w + o + ng
m > n / syllables beginning with bilabial

hwhatting · Post by **hwhatting** » Tue Nov 04, 2003 4:28 am

O.k., I re-submitted and edited my post to get rid of the smilies.
Maknas: yes, I know about writing in word first and submitting then. First, I wanted to post about the vowels only, but the post grew and grew...

Maknas wrote:Does anyone know about sound changes in Russian? I need to know for a conlang I'm starting on

Which ones? In the development towards Russian, the following stages can be distinguished:

PIE > Proto-Slavic > Old Russian > Modern (Standard) Russian.

I'll start with the vowels:

PIE (I'll use a notation without laryngeals; in the history of the Slavic languages, the influence of the laryngeals is mostly on accentuation, which I won't touch here) > PS:

a, o > Balto-Slavic /a/ > PS /o/; > e after /j/
@ : same as a,o or > 0
e > e
i > ь (reduced front vowel)
u > ъ (reduced back vowel)

a:, o: > a
e: > ě ("yat'")
i: > i
u: > y (X-Sampa /1/)

ei, e:i > i
ai, oi > ě (i after /j/)
a:i, o:i > ě, sometimes a

Syllabic /l/, /r/ > ьl / ъl, ьr / ъr
Syllabic /m/, /n/ > ьm / ъm, ьn / ъn in open syllables, e~ / o~ in closed syllables

Some conditional developments:
1. ě > a after palatal fricatives and and affricates
2. e( : ), i( : ) before nasals in closed syllables > e~
a( : ), o( : ), u( : ) before nasals in closed syllables > o~
3. o > ъ in some instances in final syllables
4. oi > i in some instances in final syllables

Consonants:

PIE Media and Media Aspirata merge, while the Tenues are kept:

p, t, k > p, t, k

b / bh, d/dh, g/gh > b, d, g

The labiovelars become plain velars:

kw, gw/gwh > k, g

The palatal stops become dental fricatives (probably via a stage of palatal fricatives in Balto-Slavic):

k', g'/g'h > s, z

All other consonants remain unchanged generally.

Some conditional developments:

1. Law of open syllables:

Proto-Slavic allows syllable-initital consonant clusters, but not syllable-final consonants except for liquids. So every syllable-closing stop and fricative (Late PIE had only /s/) vanishes without trace, while syllable-closing nasals vanish leaving nasalised vowels as a trace.

2. iurk-rule:
PIE /s/ > /x/ after (PIE) i, u, r, k

3. 1st palatalisation

k, g, x > /t_C/, /Z/, /S/ before PIE e( : ), i( : ), e( : )i

4. 2nd palatalisation

k, g, x > /t_s/, /d_z/, /s/ before PS ě, i from PIE ai, oi

5. 3rd palatalisation

k, g, x > /t_s/, /d_z/, /s/ after some syllables with front vowels

(The names 2nd and 3rd palatalisation are due to scientific history; nowadays it's mostly assumed that the 3rd came before the 2nd palatalisation.)

Combinatoric changes with /j/:

/sj/, /zj/ > /S/, /Z/

/kj/, /gj/, /xj/ > /t_S/, /Z/, /S/

For /m/, labial and dental stops see below; in all other cases the combination resulted in strongly palatalised consonants.

Diverging developments in the Slavic branches:

Some developments which show similar tendencies in all Slavic branches but different results are normally assigned to a stage called Common Slavic. Those are:

1. The result of the 2nd palatalisation of /x/ is /s/ only in Eastern and Southern Slavic, in Western Slavic it's /S/.

2. In Western Slavic, the 2nd palatalisation does not aply to the sequence Kw': Polish kwiat Russian cvet "flower", Polish gwiazda Russian zvezda "star"

3. The sequence -VTlV- (T = t,d) gives /VdlV/ in Western Slavic, /l/ in Eastern and Southern Slavic: Polish mydło, Russian mylo "soap".

4. The sequence Vr, Vl, in closed syllables gives the following results (so-called "metathesis of liquids"):

or: Czech, Slovak, South Slavic /ra/, Polish /ro/; Russian /oro/ (in word-initial position Polish and Russian sometimes also have /ra/, depending on accentuation - the same is true for the other combinations also)

ol: Czech, Slovak, South Slavic /la/, Polish /ło/; Russian /olo/

er: Czech, Slovak, South Slavic /rě/, Polish /rze/; Russian /ere/

el: Czech, Slovak, South Slavic /lě/, Polish /le/; Russian /olo/

5. Outcomes of /mj/:

Western Slavic /m'/; Eastern and Southern Slavic /ml'/

Polish ziemia, Russian zeml'a "Earth, land"

EDIT: The same is valid for labial stops; cf. Russian kupl'u "I buy" with Polish kupię

6. Outcomes of /tj/, /dj/; /kt/ + PIE front vowel, /gd/ + PIE front vowel

Eastern Slavic /t_S/, /Z/; Western Slavic /t_s/, /d_z/; Bulgarian /St'/, /Zd'/; Serbo-Croatian /t_C/, /d_j\/; Makedonian /k'/, /g'/

That's it for today; I'll post something on the further developments to Russian later.
Best regards,
Hans-Werner

hwhatting · Post by **hwhatting** » Fri Nov 07, 2003 7:54 am

2nd instalment: Sound changes from Proto-Slavic (PS) to Old Russian (OR):

These are more limited; OTOH, some of the diverging developments enumerated in the previous post might also be put here.

1. Denasalisation of nasalised vowels:

PS /e~/ > /'a/ (' denotes palatalisation of the preceding consonant)
PS /o~/ > /u/

OCS je~zykъ Polish język Russian jazyk "tongue"
OCS ro~ka Polish ręka Russian ruka "hand"

(All Slavic Languages except Polish denasalised the PS nasalised vowels, although with differing results).

2. Treatment of word-initial (and some intervocalic) /j/

/jV/ > /V/ for V = /e/, /u/, /o~/; in this case, /e/ > /o/

OCS jezero Polish jezioro Russian ozero "lake"

OCS jutro Russian utro "morning" Polish jutro "tomorrow"

PS *pajo~kъ Polish pająk Russian pauk "spider"

/jь/ > /i/

PS *jьgra > Polish gra Russian igra "game"

3. Prothetic /j/

/#a/ > /ja/

PS *ablъko Polish jabłko Russian jabloko "apple"

Especially for 3) it is in many cases disputed to which extent the /j/ in those Slavic languages which have it is prothetical or original (with those languages that don't show it having lost the /j/ due to rules analogical to rule 2).

Last instalment (OR > Modern Russian) when I have time; I'll be away for a while, so don't expect anything for the next two weeks).

Best regards,
Hans-Werner

alice · Post by **alice** » Fri Nov 07, 2003 10:14 am

hwhatting wrote: ei, e:i > i
ai, oi > ? (i after /j/)
a:i, o:i > ?, sometimes a

You missed out the other diphthongs:

au, a:u, ou, o:u > u
eu, e:u > ju (not jy!)

I'm pretty sure about the last one, but I can't find an actual example. Does anyone know of any?

hwhatting · Post by **hwhatting** » Tue Nov 25, 2003 11:03 am

geoff wrote:
hwhatting wrote: ei, e:i > i
ai, oi > ? (i after /j/)
a:i, o:i > ?, sometimes a
You missed out the other diphthongs:

au, a:u, ou, o:u > u
eu, e:u > ju (not jy!)

I'm pretty sure about the last one, but I can't find an actual example. Does anyone know of any?

Thanks for putting those in. I also forgot:
u > ь after /j/
ejV > ьjV
ewV > owV
w > v

and certainly some more, minor ones. I also did not include the word-final changes, which anyway are among the most debated ones.

On PIE /*eu/ : that's kinky. There seem to be examples for both the development PIE *eu to PS /ju/ (bljusti from *beudh- (?), ljub- from *leubh-, ljudi from *H1leudh-, etc.) and to PS /u/. The cases with *eu > u can, of course, always be explained away by assuming that the word in Pre-PS had the alternative ablaut stage /ou/.
Best regards,
Hans-Werner

Jaaaaaa · Post by **Jaaaaaa** » Tue Nov 25, 2003 4:15 pm

This kinda shows how low my knowledge is, but...

-The Great Vowel Shift-
(changes in the long vowels from Middle to Modern English)

a: > E: > eI
E: > e: > i:
e: > i:
i: > @i > ai
o: > u
O: > o
u: > @u > au
[y > ju] (bracketed because I'm not sure about this one)

...and that is all that I know. I'll try to get in the Great Consonant Shift when I make sure I got that one right.

pharazon · Post by **pharazon** » Tue Nov 25, 2003 5:25 pm

If it's English sound changes you want, there's some here.

Mecislau · Post by **Mecislau** » Sat Dec 06, 2003 5:29 pm

I am reviving this thread for three reasons:

(1) It is just too useful to let it die
(2) To ask if anyone else has additions
(3) To ask if anyone here knows about grammatical change.

By #3, I mean, what kinds of change can the grammar of a language go through? What causes them? How common are they? Can you show some natlang examples?

pharazon · Post by **pharazon** » Sat Dec 06, 2003 5:37 pm

A common grammar change is for particles to attach to words; that's often how case endings originate. Kind of the opposite can occur when endings wear down over time which makes the language more isolating, e.g. Latin to the modern Romance langs.

As for additions, I think I'll make a list of those English sound changes on that page.

Mecislau · Post by **Mecislau** » Sun Dec 07, 2003 8:46 am

How do languages gain/lose cases?

Jaaaaaa · Post by **Jaaaaaa** » Sun Dec 07, 2003 10:56 am

Maknas wrote:How do languages gain/lose cases?

Gain: postpositions and other junk merge with the noun they modify; and sometimes there's just inexplicable innovation.

Lose: Sound changes wear off word-final sounds; and sometimes cases' functions get taken over by other cases.

gach · Post by **gach** » Sun Dec 07, 2003 11:27 am

Jaaaaaa wrote:Lose: Sound changes wear off word-final sounds; and sometimes cases' functions get taken over by other cases.

Don't forget that also adpositions like to take over cases.

doctrellor · Post by **doctrellor** » Sun Dec 07, 2003 12:22 pm

Is there a "guide" on what exactly a language type does,,

for instance...

Agglutinating do...a,b,c,d, etc., etc.
Isolating languages do...
Fusional languages do...

something simple where we can see the flow, and see how the languages change from one with no adpositions, and 1 or 2 meansings per word (Isolating) to 2 or 3 affixes, 4 or 5 meanings per root(Agglutinating), to the affixes join together into a 'hard to seperate' mess (Fusional/Poly)

do you guys have something like that?

Ran · Post by **Ran** » Sun Dec 07, 2003 1:08 pm

btw Jhex: iirc the Great Vowel Shift went like this for the front vowels:

a: > E: > eI (mate)
E: > e: > i: (meat)
e: > i: (meet)
i: > @i > ai (mite)

Jaaaaaa · Post by **Jaaaaaa** » Sun Dec 07, 2003 2:52 pm

Ran wrote:btw Jhex: iirc the Great Vowel Shift went like this for the front vowels:

a: > E: > eI (mate)
E: > e: > i: (meat)
e: > i: (meet)
i: > @i > ai (mite)

Oh, yes, yerright. Sorry about the inconsistency; I listed the intermediate for the /i:/ > /ai/ but not the others... I shall correct!