LINGUIST List 2.450

Mon 02 Sep 1991

Disc: Sound-Change and Teleology

Editor for this issue: <>

Directory

1. Joe Stemberger, Re: Why 'wh-'
2. , Teleology of function/purpose

Bruce Samuelson asks about why there are phonological-semantic congruences
at all.
First, we have to keep in mind that there are different historical pathways
to such congruences. The wh-words are presumably just a synchronic residue
of standard word-formation processes at some ancient stage of the language.
The "gl" words that someone else reported (GLITTER, GLEAM, etc.) are
another example.
The "th" congruence that Bruce brought up is quite different:
 "th" is voiced in function words
 "th" is voiceless in content words
This was never part of word formation, and is the result of sound change.
All words originally had voiceless "th", but it voiced in word-initial
position in just the function words. (I seem to remember that this happened
in the early Middle English period, but I wouldn't swear on it.) The sound
change was identical in Swedish, Norwegian, and Danish (which, because of
a later change, now show stops, with /d/ in function words and /t/ in
content words). The change in German was more extensive; "th" became voiced
in ALL words, and then word-initial /s/ also voiced (and in some obscure
dialects, even word-initial "sh" has voiced). I don't think that anyone
knows why the change went so much further in German.
I question whether English really shows a congruence between phonology and
SEMANTICS, though. There is little in common semantically between the
affected function words. It correlates with whatever makes function words
function words, but that is probably not semantic. (Some people would claim
it is more syntactic in nature. You might be able to get away with arguing
that it is pragmatic in nature.)
It is possible that the change didn't really even have anything to do with
the function/content distinction at all. Sound changes often spread through
the lexicon one word at a time, with two common patterns: frequent words
change first (with infrequent words unaffected), or infrequent words change
first (with frequent words unaffected). (Labov had a paper in LANGUAGE a
few years back on what determines whether sound change will be
word-by-word. There was a paper in LINGUISTICS ca. 1984 that tried to sort
out what determines whether it is infrequent or frequent words that change
first; I don't recall the author's name.) It is quite possible that this
change affected frequent words first, which would largely be function
words, since they are the most frequent words in the language. The change
may have ended (and who knows why?) before spreading to even relatively
frequent content words like THINK. (It may not have spread to words like
THREE and THROUGH because of a phonological constraint in English barring
voiced fricatives in syllable-initial consonant clusters. English doesn't
allow *VRY or *ZLIP, either.)
Anyway, this congruence may be entirely accidental. Anyone know more of the
details? I'm not an expert on such long-ago sound changes.
I also agree with the comments that we have no idea why sounds change,
except in the most general sense. Why did original Germanic /sk/ change to
"sh" in West Germanic, but survive as [sk] in most North Germanic languages
until comparatively recently? In Swedish, before front vowels and /j/, it
is now "sh" in many dialects (a standard palatoalveolar fricative), but is
a highly unusual labiopalatal or labiovelar fricative (with frication at
both the palate and lips) in many other dialects. But in Danish, I think,
it is still [sk]. Anyone who can explain that sort of variability should
come forward.
---joe stemberger

In Linguist List: Vol-2-444, Bert Peeters says, re sound change,
that he doesn't believe in teleology; instead, he opts for
(French) functionalism. But is it *really* possible to maintain
the distinction between teleology of purpose and teleology of function
in the explanation of language change?
 Suppose the Latin change "-nct- > -nt-" (quinctus > quintus;
sanctus > santus; &c) took place because the sequence -nct- was
"not easy enough". This is scarcely substantially different from
saying that this change took place for the purpose of ease
of pronunciation. That is: the gist of Martinet's principle of
economy remains the same, whether economy is conceived as a telos,
or as a function.
 -- Martti Nyman
 Univ of Helsinki, Finland