LINGUIST List 6.629

Sun 30 Apr 1995

Disc: Affricates

Editor for this issue: <>

Directory

1. Alexis Manaster Ramer, Re: 6.610, Affricates
2. "Sarah G. Thomason", Re: 6.610, Affricates
3. Hala'sz Sa'ndor, Re: 6.610, Affricates
4. Jack, Re: 6.610, Affricates

Message 1: Re: 6.610, Affricates

I am glad to have Don Churma come out and confirm that I am not
hallucinating when I say there is such a thing a a [t+S] vs. [tS]
contrast in some languages, incl. my native Polish (where it has
been well-known for ages; it is even mentioned in Bloomfield's
Lg). I am puzzled by Marc Picard's statement that there is
no difference between French *tch* and English *ch*, since the
former is not aspirated in environments where the latter is,
for example. On the other hand, I agree with him that (in
the speech of my informant at the time anyway), French did have
a one-segment [tS], not the [t-S] mentioned by Churma.
Maybe different speakers differ (such segments being restricted
to foreign words, if I am not mistaken, whre variation is expected).

But whatever the situation in French, the contrast itself is alive
and well, e.g., in Polish, and in the speech of some English speakers
too, as I keep pointing out.

Alexis MR

Message 2: Re: 6.610, Affricates

Marc Picard says that he is "extremely suspicious of any purported
difference between [t+S]/[tS] in any language". Montana Salish
(a.k.a. Flathead) has a phonemic distinction between a cluster
/t.S/ and an affricate /tS/; in the cluster, the /t/ is released
before the fricative. So, for instance, the word for `bull elk'
is /t.Sec'/; the diminutive form is l*-tt.Sec', with initial-consonant
reduplication. (l* = voiceless lateral fricative.) And the root
for `soft' is /tSep/; the full word, with initial-consonant
reduplication (and also with the diminutive prefix) is l*-tStSep.
The initial-C reduplication process shows the difference between
cluster and affricate clearly, but the phonetic difference is quite
striking in unreduplicated forms. (Stops in clusters are regularly
released in Montana Salish.)

 Similar cluster-vs.-affricate distinctions can be found in
other Salishan languages, I believe.

 Sally Thomason
 sallyisp.pitt.edu

Message 3: Re: 6.610, Affricates

 Marc Picard:
 .... I'm also extremely suspicious of any purported
 difference between [t+S]/[tS] in any language.

Well, in Hungarish "t" and "s" (Hung. spelling) combine to form not "cs"
(Hung. spelling) but "ccs"/"tcs", the long form of it. But combining "tt"
and "s" would also yield "ccs"/"tcs" (combining "t" and "ss" is out of the
question; it would mean that a word or ending begins with "ss").

Message 4: Re: 6.610, Affricates

Leonard Bloomfield approached this problem long ago in a "cute" footnote
stating (I quote from aging memory). "Anyone who hears no difference
between the (affricate) in 'catch it' and the (sequence) in 'that shirt'
is apt to find himself in difficulties." In those more delicate days one
would not have cited in print the "minimal pair" 'catch it' vs 'cat shit'.
(a fuller utterance ending with these in an interrogative intonation will
work best to avoid reduction to schwa of the vowel in 'it'.)

 /paks/

Ki semenat ispinaza, non andet iskultsu!

J. A. Rea jareaukcc.uky.edu