Re: Back to mutations again was Re: Mutations in General

At 1:22 PM -0500 10/22/02, Peter Clark wrote:
>Quoting Dirk Elzinga <Dirk_Elzinga@...>:
>
>> The fact that many English speakers routinely delete the nasal in such
>> clusters gives support to the existence of such a tendency. Of course there
>> will be speakers who pronounce a full nasal; this doesn't mean that the
>> tendency isn't real (or "universal"), just that it's not active in that
>> variety of English.
>    Thanks for clearing up what you meant by "universal tendency"; I got bogged
>down on the "universal" part which tends to get equated with
>Greenburg's "universals," which aren't really universal at all (for the most
>part) and so now I tend to have a alergic reaction against "universal." My
>doctor has prescribed some pretty yellow pills for it... :)
I think of these kinds of statements as statements about surface forms which may
in principle be violated. However, such violation must be compelled by higher
priority considerations. In many varieties of English it is more important that
the surface form look like the underlying form than that the prohibition on
surface nasal/voiceless stop clusters (*NC) be observed. Thus, the effects of
*NC will be eclipsed in these varieties (the technical term is 'occultation'),
and /nt/ will be pronounced [nt]. For statements like *NC to be "universal"
simply means that they exist in all languages; their scope of influence in the
grammar of a language will determined by the kinds of constraints which take
precedence. (This idea has been formalized as Optimality Theory. There are
boatloads of papers and articles which have been written in this framework; see
the Rutgers Optimality Archive at http://roa.rutgers.edu for a large sample of
them; my dissertation is in there as well.)

> > I figured the discussion of mutation would eventually get bogged down in the
>> particulars of English pronunciation; it seems to be one of the strange
>> attractors for threads on this list.
>    Perhaps I can steer this thread back to its original topic. BTW Dirk,
>thanks for the definitions of lenition, gradation, and mutation. I knew
>lenition, but I was not familiar with gradation.
>    Now my question regarding mutation is how closely bound is it to the
>phonology of the language. You wrote that mutation is not phonologically
>predictable, but is there any specific tendencies *g* within mutations? For
>instance, Welsh has three, soft, nasal, and spiration. Soft can be generalized
>as follows: unvoiced stops are voiced, voiced stops are fricatized (is that a
>word) except for /g/ (did Welsh once have /X/ that disappeared?), and so on and
>so forth. In other words, there is a definite pattern that can be perceived. It
>is not the case that /p/ -> /b/ but /t/ -> /D/. This is what interests me.
When I wrote about the unpredictability of mutation, I was referring to the
conditioning environment and not to the consistent quality of the alternating
segments. In Welsh, the triggers for mutation are not phonological in nature;
they're morpho-syntactic. As such, mutation is phonologically unpredictable,
even if the effects of mutation are perfectly consistent.

>    Now, here's the heart of the matter: is the tendency shown by both Welsh
>and Fula frequent enough to call it a "universal tendency"? Or are there some
>really wacky mutation systems out there (granted, mutation is wacky enough as
>it is...)? Where, for instance, /p/ does not mutate to /b/, /f/, /m/, or
>something similar, but to /t/, /S/, or /l/? If there is such a system, would it
>be proper to actually call it "mutation," since the underlying change is
>probably not phonetic? If not, what would it be called? What would be some ways
>in which it could develop?
*I* wouldn't call it mutation. The overwhelming majority of cases of mutation
manipulate features of nasality, continuancy, and/or voicing (you could add
sonorancy and laryngeal features like aspiration and glottalization as well).
That is, the patterns that you see in Welsh and Fula are repeated in mutation
languages around the globe. So /p/ won't alternate with /t/, /S/, or /l/, since
that involves features other than nasality, continuancy, and voicing. More
typically, alternants of /p/ under mutation will be segments like /b/, /f/,
/v/, /w/, /m/, /mp/, /mb/, etc. The place of articulation won't change, but
continuancy, voicing, and nasality will.

The particular mutations that a language exhibits will be bound by the segmental
inventory, since mutation typically involves alternations among distinctive
elements (lenition, on the other hand, involves alternations among
non-distinctive segments; think again of AmEng flapping or Spanish
spirantization). So if a language has /p/ and /w/ but no /f/ or /v/, then a
spirantizing mutation would likely alternate /p/ and /w/ rather than some
non-distinctive segment like /f/ or /v/.

Dirk
--
Dirk Elzinga                                               Dirk_Elzinga@byu.edu

"It is important not to let one's aesthetics interfere with the appreciation of
fact." - Stephen Anderson