LINGUIST List 9.704

Tue May 12 1998

Sum: Sound Change

Editor for this issue: Martin Jacobsen <>


  1. monica prieto, SUM: sound change

Message 1: SUM: sound change

Date: Tue, 12 May 1998 15:38:54 -0500 (CDT)
From: monica prieto <>
Subject: SUM: sound change

 I am glad to see that so much interest arised about my
question of the change from [tS] to [ts].in Spanish. First of all,
what has been called the "unnamed dialect" is the Spanish of the
Basque Country (Spain). It seems that both sounds are in free
variation for some speakers (I include myself in that group).
I want to thank all the people that answered the question, they have
been of a great help; Benjamin W., Joaquim Brandao de Carvalho,
Aleksas Girdenis, Rick MacAllister, Chris Miller, keith Goeringer,
Daniel E. Collins, Joey Laubach II, Sean Jensen, Tim Beasly, Jose Luis
Estevez, Miguel Carrasquer, Paul Boersma, Curt F. Woolh, Johnny
Thomsen, C. Whiteley.

This is what I found out (some of the responses were not directly
related but I have included them because they were very interesting);

1. A change like tS > ts is commonly assumed as the last stage of the
drift of Latin /k/+j,i,e in the Western Romance languages : compare
Italian cielo,vicino, braccio, which preserve the first stage (tS),
with the corresponding forms in Spanish, Portuguese, Gallo-Romance and
North Italian, which all suppose /ts, dz/. Indeed, Old Spanish had
/ts/ and /dz/ for what was written c/c+cedilla and z respectively. So,
the modern change
 is the 'repetition' of an older one, much like ly > Z in both medieval
Castilian and modern Rioplatense.

2. We have such phenomenon in some Southern and Eastern Lithuanian
(so-called Dzukish) dialects. E.g. 'Common' Lithuanian [tSe] 'here' =
"Dzukish" [tse], C.Lith. [patS'o:s] (' - the signe of palatalization)
'of the wife' = Dz.[pats'o:s] or [patso:s]. The evolution of the
voiced [dZ] is the same: C.Lith. [dZ'au.ksmas] 'joy', [gaidZ'u:] 'of
the roosters, of the cocks' =Dz. [dz'ou.ksmas] or [dzou.ksmas],
[gaidz'u:] or [gaidzu:]. But our dialectologists and language
historians are inclined to interpret these differences as results of
the divergent evolution of the Common East Baltic clusters *tj, *dj
(in most dialects *tj, *dj > tS', dZ', in the Dzukish ones *tj, *dj
>ts', dz' (>ts,dz).
 I suspect that the change [tS] > [ts] also can be found in some
Northern Italian dialects.

3. This kind of sound change, where [ts] in some varieties coexists
synchronically with [tS] in others, is found in modern-day Cree
(Northern Canada) and Occitan (the indigenous "langue d'oc" of the
southern third of France). In Occitan, the [ts] variant (along with a
[dz] for [d3]) is found mainly in certain geographically central
varieties, north and east of Toulouse/Tolosa. Certain Bantu languages
(including northern coastal varieties of Swahili) show a similar
evolution, but to an interdental affricate. It would be worthwhile
investigating the historical evolution of latin velar stops before
front vowels in the modern langges. Whereas Italian and Romanian
changed these stops into [tS] and [d3] and, for the most part, haven't
proceeded any further, others have extended the change further. So in
Occitan as in most western Romance languages, the surmised series of
changes (for k/[-back] __ ) is k>tS>ts>s. Castilian Spanish differed
from the others in changing [ts] to [TH] (the voiceless interdental
fricative). Slavic languages show differential palatalizations of
underlying /k/ to [tS] or [ts] (with a similar split for /g/) at
different levels of their lexical phonology. Much work has been done
on these issues with respect to Russian and Polish, for example.

4. I've only heard this among chilenos but what I've heard is more
like a /tsy/ e.g. muchacha becomes "mutsyatsya" they also palatalize
<ge, gi, j> as well gente > "giente"

Asturiano is said to have /ts/ someplaces where Spanish has borrowed
/ch/ from Portuguese, but it corresponds to Spanish <ll>;
e.g. chubasco vs. tsobasco

5. Within Slavic, some language have [tS] as the reflex of the
proto-sound, and others have [ts] -- in Russian, 'what' is *chto*
(though it's usual pronunciation is [Sto], it is still [tSto]
dialectally); in Polish, it is *co* [tso]. Also, certain (mainly
north, I believe) Russian dialects are distinguished by the
pronunciation of [tS] as [ts] -- this is a trait also noted in Old
Russian dialects.

6. Mandarin, or Putonghua, is the national language of China. It has
both /tS/ and /ts/ occuring with very great frequency. In the official
Pinyin orthography these are written <ch> and <c> respectively. We
encounter forms, for example, like <cai1> (where the <1> indicates the
"first (high-level) tone") meaning "to guess", and <chai1> meaning "to
demolish". However, in the Southern parts of China, including Taiwan,
<ch> and <c> are not distinguished, and only <c> is found. Hence for
speakers down South, we find only <cai1> meaning both "to guess" and
"to demolish". This "merger" is found with the unapirated counterparts
<zh> and <z> (/dZ/ and /dz/) and the fricatives <sh>, <s> (/S/, /s/).

Official Southern Gloss

chai1 cai1 demolish
cai1 cai1 guess 
zhai1 zai1 vegetarian diet
zai1 zai1 disaster
shai1 sai1 sieve
sai1 sai1 squeeze in

Frequently in Southern areas the <ch,zh,sh> forms are reintroduced
where they do not exist in "official" Putonghua. Thus we occasionally
find <chai1> used down South for <cai1> "to guess". This phenomenon is
said to be a "feminine nicety" and is felt to be somewhat "effeminate"
when used by men.

7. Standard Russian has /tS/ as a unitary phoneme, as well as /ts/.
Some dialects possess only one or the other. Both result from /k/
palatalizing before front vowels or /j/. /k/+/e/ > /tSe/ is pretty
much pan-Slavic.

8. analogous changes can be observed in a number of northern Slavic
dialects, including Northern Russian dialects, where the
palatoalveolar affricate [tS] has been replaced by a dental affricate
[ts] (or palatalized [ts'] in some dialects), e.g. standard Russian
[tSas] 'hour' = North Russian dialect [tsas], [ts'as]. In many Polish
dialects, all palatoalveolar affricates and fricatives have been
replaced by dental affricates and fricatives, e.g. standard Polish
[tSytaS] (orthographically czytasz) 'you (eg.) read' = Polish dialect
(Mazovia, Little Poland, Silesia) [tsytas]. Something similar can also
be observed in the dialects of Yiddish formerly spoken in eastern
Poland, Lithuania and Belarus.

9.You may notice a slight move forward in the direction of ts in the
colloquial speech of Madrid, maybe to imply a touch of irony, but it's
not so alveolar as in "cats".

10. in modern Greek, some speakers pronounce the equivalent of "ts" as
English "ch".

11. In Kartvilico, /tS/, /t'S'/ and /dZ/ in Zan and Svan correspond to
Georgian /ts/, /t's'/ and /dz/ ( traditional reconstruction *c1,
*c'1,*dz1), and probably Zan and Svan mantain the previous stage.

In proto-Semitic, the sound *c^ (*c^., *dz^) appears in Arabic as /T/
(/D./, /D/), probably through the same development than in Spanish (c^
> c > T -- > /s/ o > /t/ in modern Arabic dialects). Most of the other
semitic languages have /S/ (very common development c^ > s^,
i.e. francis <ch> /tS/ > /S/).

In general,the most common developments seem to be:

c^ > s^ and then: s^ > x > h, s^ > s
c^ > c and then: c > T, c > s; T > s, T > t
c^ > t [?]

Thank you very much
Monica Prieto 
Mail to author|Respond to list|Read more issues|LINGUIST home page|Top of issue