LINGUIST List 2.171

Saturday, 27 Apr 1991

Disc: Phonology (part 1)

Editor for this issue: <>

Directory

1. , Re: Morphophonology (Part 1)
2. Mike Hammond, English Stress
3. , Phonology

Frank McLellan brings up the following interesting example:

>...I discount the evidence for
>the phonemic status of [y] offered by a phonetics teachre in Leningrad who
>insisted that the name of the state museum housed in the Winter Palace was
>called the [yrmit'ash]. That word is a French borrowing which still has not
>been assimilated into the phonology of the language.

I disagree about this word not having been assimilated into native phonology. 
It happens that all instances of unstressed initial /e/ are in loan words,
but it is doubtful than many Russians are aware that such words are nonnative.
The problem here is that such sounds do not always alternate with stressed 
[e], and it becomes controversial to argue that [y] is a normal reduced variant
of /e/. So there is the potential for claiming a contrast with words 
beginning in /i/->[i]. My argument for /e/ would be that vowel reduction is
variable here--i.e. sometimes unreduced, unstressed [e] is heard--and the 
orthography (again) may be a factor influencing phonological sensibilities.
But you can see why the i/y controversy doesn't get put to rest very easily.

The fact that /i/ and /e/ do not neutralize totally under vowel reduction
is reminiscent of those arguments, often made by experimental phoneticians,
that so-called "phonemic neutralization" doesn't really exist. We have all
seen studies arguing that /t/ and /d/ do not really "neutralize" under flapping
in American or that final devoicing actually produces phonetic traces that 
are distinct from voiceless phonemes. I welcome such objective verification of
the Baudouin's physiophonetics (=phonology). Psychophonetic alternations
never leave phonetic traces of underlying abstract representations (because,
I would argue, speakers are not trying to pronounce those representations). 
Phonetic evidence against neutralization almost never resolves the question
of whether hearers actually perceive the minute phonetic traces. Evidence from
rhyme suggests they don't. For example, voiceless /t/ and final devoiced /d/
in 'vot' [vot] and 'rod' [rod] produce perfect rhyme. So the perceptual
'L-phonemic' level is attested to by rhyme, and the production-based 
'M-phonemic' level by the lack of pure phonetic neutralization.

 -Rick Wojcik (rwojcikatc.boeing.com)

p.s. David Stampe once remarked that there is no solid evidence that one 
language ever really 'borrows' the phonology of another. This is not to say
that there is no Sprachbund effect in language contact, but that words per se,
to be acceptable to native speakers, must be pronounceable within the 
constraints of the borrowing language. People who persist in using foreign
pronunciations are not always appreciated by their fellow speakers. On the
other hand, intimacy between different languages may drive the phonology of
one into a position where it can accept previously banned structures. Bringing
one language into resonance with another is more complex than just borrowing
idiosyncratic pronunciations.

With respect to the spate of responses to the debate on English
stress:

Alexis Manaster-Ramer notes that there are words like Admiralty and
baronetage with preantepenultimate stress. I would maintain that this
is only possible in suffixed forms. In nonsuffixed forms, there must
be a stress, primary or secondary, within the last three syllables.

With respect to the difference between words like orchestra and words
like industry, Alexis maintains they should be treated the same way. I
think this is reasonable. Words like galaxy require something more
though. The generalization at issue is whether a stressless penult
with the shape [laek] can occur before syllables ending in other than
[-i]. If not, then this is an argument for the SPE solution.

With respect to Polish, yow! Did I manage to make my remarks to a
native speaker??? (If so, I'd have stepped more lightly!) Nonetheless,
I stand by my original position: intuitions of (ir)regularity aside,
stress in Polish is lexical in the final three syllables of a word. I
think Alexis and I agree there's more to say, but I predict there's more
to say in Greek too! [Incidently, I have a paper on the Polish stuff
that appeared in _Phonology_6_.]

Harry Bochner maintains that the positions AMR and I have debated
exclude a viable alternative.

"The disagreement comes, I think, from the fact that Mike (and SPE-style
phonology generally) seems to think that there is a simple dichotomy: either
a) stress is 'rule-governed' in the sense that every case can be ground out by
 the rule system (using exception features if necessary), or 
b) it is completely arbitrary lexically, not governed by any significant
 generalizations.
While I'm not sure what Alexis has in mind, I doubt that it's (b). In any
case, I claim that this dichotomy is a false one.
What we need is a theory that allows us to express generalizations about
listed facts. This is what word-based theories of Morphology are about:
when we say that 'sanity' has a lexical entry in such a theory, we are
_not_ denying the validity of the generalization that -ity often forms
nouns from adjectives. Similarly, saying that stress is lexical in English
does not deny that it is often predictable."

I think Harry's analysis of the two possibilities in (a) and (b) is
correct, but I maintain that that's not a false dichotomy. Rules are
rules. What else is there? If something isn't rule-governed, then it's
not rule-governed. Why should sanity be listed if it's behavior is
rule-governed?


mike hammond

In response to Wojcik, I don't know what I can say save to repeat
that Russians do indeed use 'y' as the name of the letter in question.
If people don't want to consider 'y' : 'i' to be a minimal pair,
that is their business. I don't know what a minimal pair is if
this is not one, but I am just a country computer scientist. In
any case, Baudouin would clearly have recognized that it is not
the case that the choice of [i] vs. [y] is physiologically conditioned
if he saw this example, and that is what counts for the argument
we are having. 

Likewise, I still do not see how one can invoke orthography as
the factor that determines what people hear as distinct sounds
when I keep pointing out that Russians do NOT hear [i] in those
places where 'i' is written and [y] in those places where 'y'
is written, and everybody agrees with this, but still maintains
that orthography IS the crucial factor. Also, I DID say that
I do not believe that distribution is what determines whether
people hear sounds as same or as different, but that (a sophisticated
theory of) phonetic similarity (at least as sophisticated as that
of Stampe and Donegan but probably more so, as I also pointed out)
is. And I cited such other, and classic, examples as [h] and [ng]
in English. Or is that also due to the orthography.

[End Linguist List, Vol. 2, No. 171]