LINGUIST List 2.490

Tue 10 Sep 1991

Disc: Sound-Change

Editor for this issue: <>

Directory

1. Susan Ervin-Tripp, Re: Sound-Change/471
2. bert peeters, "Natural" = "easy"
3. Geoffrey Russom, Re: Sound-Change
4. , Sound-Change/471
5. bert peeters, Teleology and sound change

Another indicator of naturalness--which Fromkin could speak to--is the
direction of production errors. Possibly these are so different cross-
linguistically that no generalizations can be made. For example, the
proposal that 'nuclear' and 'nucular' differ either in measurable
complexity measures, or just in being made of cv syllables or not, may account
for the very common replacement of the first by the second. But are such
replacement errors common in other languages which permit consonant clusters?
S. Ervin-Tripp

> Date: Tue, 03 Sep 91 11:46:19 EDT
> From: Geoffrey Russom <EL403015brownvm.brown.edu>
> Subject: Re: Sound-Change
>
> On "natural" sound change: Any discussion of this topic would I think
> have to account for the fact that there are isolated languages like
> Hawaiian in which you have almost all open syllables (e.g. the name of
> a fish, humuhumunukunukuapua, if I remember correctly). Japanese comes
> pretty close, too. Aren't these much easier for a random foreigner to
> pronounce properly than languages like English and Russian? If so,
> they would appear to be more "natural" in a universal sense.
It's not a matter of open vs closed syllables (why do most people say
[hawaj] rather than [hawa-i]?), but of strongly contrasting sounds in
succession (usually, of course, CVCVCVCVCV...).
Dr Bert Peeters Tel: +61 02 202344
Department of Modern Languages 002 202344
University of Tasmania at Hobart Fax: 002 202186
GPO Box 252C Bert.Peetersmodlang.utas.edu.au
Hobart TAS 7001
Australia

On "natural" sound change: Any discussion of this topic would I think
have to account for the fact that there are isolated languages like
Hawaiian in which you have almost all open syllables (e.g. the name of
a fish, humuhumunukunukuapua, if I remember correctly). Japanese comes
pretty close, too. Aren't these much easier for a random foreigner to
pronounce properly than languages like English and Russian? If so,
they would appear to be more "natural" in a universal sense. It is also
worth noting that the more cosmopolitan languages (the ones with armies
and navies) tend to borrow foreign words and blend adjacent dialects
to a considerable extent (radicalism of the center), so the "standards"
we study are always being rendered more complex, creating numerous
counterexamples to "ease of articulation" that may be no more than
apparent counterexamples.
 -- Rick Russom

This isn't to Rick Russom's main point, but rather a reaction to his
comment that "there are isolated languages like Hawaiian in which you
have almost all open syllables ...". In the first place, it isn't
"almost": all syllables are open. Second, I'm not sure what is meant
by an "isolated language". If he means Hawaii is an isolated place,
that's true enough, though the Hawaiian Department of Tourism is
spending millions to make it less true. But if there is some
implication of a "language isolate", then it certainly doesn't apply
to Hawaiian. There are some 30 Polynesian languages, all clearly and
closely related, and they all have open syllables.
Paul Chapin

> I agree that sound change must be explained "in terms of origin" (though
> this isn't enough).
I don't understand. What else is there to be explained? Once one knows why
a change originated, one has an explanation - although the explanatory
process may have (will have) to be renewed, as due consideration is to be
given to social factors and the like.
> However, that doesn't preclude teleological explanation.
Indeed, I would say it makes teleological explanation unnecessary.
> In 1949, Martinet wrote: "Puisque ... on parle pour se faire comprendre,
> les de'viations accidentelles, ine'vitables, auront des changes d'e^tre
> e'limine'es si elles tendent a` empe^cher la compre'hension mutuelle,
> puisque le locuteur devra se corriger s'il veut atteindre son but"
> (quoted from B.Peeters, La Linguistique 19, 1983, 114). This may be
> a rare bird, but scarcely a lapsus. In Folia Linguistica (20, 1983,
> p.540), Peeters quotes Martinet's dictum "les langues changent parce
> qu'elles fonctionnent", adding by way of a comment: "c.a`.d. servent
> a` la communication". Isn't Martinet represented as a crypto-
> teleologist here? He didn't write "les langues changent pour
> fonctionner". Why?
The first passage may indeed have a teleological ring. It should not be
forgotten, however, that in 1949 Martinet had not yet reached the final
formulation of his principle of economy, and that he was trying hard to
get his own point (against teleology) across. It is not surprising that
he was not successful in his earliest attempts.
More importantly, neither passage addresses a particular instance of
sound change. The first one tells us why certain changes DO NOT take
place, the second one is meant to explain why LANGUAGES as such change.
The argument is that as soon as languages do not function anymore - do
not serve communicative purposes anymore - they will stop changing and
become extinct.
Some people may feel Martinet is to be accused of inconsistency: he
looks very much like a teleologist as far as synchrony is concerned
(stressing as he does that languages serve communicative purposes)
whereas he speaks out against teleology as soon as language change is
at the order of the day. I don't feel there is an inconsistency at all.
An analogy with eating may be in place. At the synchronical level, one
may say that humans eat in order to stay alive (teleology? yes, in some
sense). But once one studies why dishes do not always remain the same,
the explanation is obviously not teleological at all: they change
because our culinary tastes are changing - because we need a change.
> Enough cavilling! Peeters (Folia Linguistica 1986, p.539) objects to
> Josef Vachek's statement that Martinet endorses the view that 'function'
> presupposes a teleological approach. Whereas Vachek may be wrong with
> respect to Martinet, he is obviously right in metatheory. Functional
> explanation consists in relating a set of functions to some goal(s).
That is one way of defining what "functional explanation" is: it all
boils down to "explaining functions" (in terms of the goals related to
them). However, "functional explanation" can mean something entirely
different - and Martinet has that other meaning in mind: he thinks of
functional explanation as of "explanations with respect to function".
That is to say, he tries to explain sound change while taking into
account the function of the elements involved, and their place in
the phonological system.
> So, teleology
> is supposed to bring forth phonetic anarchy. This fallacy is
> probably due to Martinet's autonomist conception of phonological
> systems. If phonology is looked upon as a sub-system interacting
> with other grammatical sub-systems (morphology, syntax, ..), nothing
> of this ilk will happen. That language is neither anarchically nor
> ideally organized is due to the interplay between sub-systems.
Nyman's argument seems to be that teleology will only bring forth phonetic
anarchy if phonological systems are looked upon as autonomous systems
which don't interact with the morphology, the syntax, the lexicon of a
language. I must disagree. As in language there is no goal, defined once
and for ever, and agreed upon by all the speakers of a language, even at
a subconscious level, teleology must lead to anarchy - in the area of
language change, but not for instance in theology, where teleology (cf.
the works of Pierre Teilhard de Chardin) has probably far more reality.
(Don't quote me on this one: I'm not a theologian). In the case of language
change, simplicity may seem to be a common goal. But what is simple for
one speaker is not necessarily simple for the next; we may have different
simplicities in mind and set off different changes, but the fact that our
languages are not too bad after all generally stops us from doing so (ex-
cept when communicative needs are so pressing that we cannot but give in).
I also believe that it is unfair to accuse Martinet of blindness as far
as interrelatedness of language subsystems is concerned. One example that
comes immediately to mind is where he talks about the strategies used
by speakers when the opposition between the two a's in French (anterior
and posterior) started breaking up: there was lexical substitution (ta^che
being more and more often replaced with travail, devoir, boulot; las being
replaced with fatigue', etc. - in order to prevent misunderstanding as
there are words that used to differ only with respect to the nature of the
vowel: tache = spot (without a circumflex); la` = there)
Have I convinced anyone?
Dr Bert Peeters Tel: +61 02 202344
Department of Modern Languages 002 202344
University of Tasmania at Hobart Fax: 002 202186
GPO Box 252C Bert.Peetersmodlang.utas.edu.au
Hobart TAS 7001
Australia