LINGUIST List 2.156

Monday, 22 Apr 1991

Disc: Morphophonology

Editor for this issue: <>

Directory

1. Mike Hammond, Morphophonology
2. , Re: Russian i/y controversy
3. , Re: Phonology (reply to Manaster-Ramer)
4. "Michael Kac", Hutchinson's Argument Contra Halle

With regard to Alexis Manaster-Ramer's last response to my last
response to his....

English stuff:

I agree that there are a number of words like _Kentucky_ that would be
treated as exceptions if, as I am arguing, English stress is
rule-governed.

I agree that there are words like _salamander_ and _Abernathy_ with
preantepenultimate main stress. The generalization that I want to
maintain is that there is always a stress within the last three
syllables of any English noun. That stress is usually a primary, but
can be a secondary as well. Hence the two words above have a
penultimate secondary stress.

General stuff:

Alexis likens English to Greek where stress is restricted to the last
three syllables, but, he maintains, is lexical in that domain. The
status of Greek notwithstanding, I think a comparason with Polish is
more appropriate. Polish restricts main stress to the last three
syllables. Stress usually falls on the penult, but in exceptionally
marked forms, it falls on the antepenult or ultima. As discussed by a
number of people (Bernard Comrie, Steve Franks, Halle & Vergnaud, and
me), when suffixes are added, words surface with penultimate stress.
The moral is that even though stresss is "lexical" in the
three-syllable domain, it would be a mistake to say it's not
rule-governed.

Finally, with regard to the SPE analysis of _industry_ as containing a
final underlying /y/, Alexis maintains that this is objectionable for
two reasons: "I object to this because this is completely arbitrary:
you can do anything you want once you are allowed to take impossible
sound sequences and put them in your URs. If there are alternations,
that is one thing. But in this case there are not." I don't agree with
either objection. First, there are bizillions of analyses out there
with unsyllabified underlying representations, where such
representations are of course unpronounceable and contain impossible
sequences of sounds. Second, I don't understand why alternations are
criterial for setting up underlying representations. Why is
distributional evidence not sufficient?

mike hammond

This replies to Alexis Manaster-Ramer's 3 "easy refutations" of my speculation
about the i/y controversy. First, some background:

 - Russian i/y (high front/back unrounded) are in complementary distribution
 - Russian has a phonemic distinction between "hard" (velarized) and "soft"
 (palatalized) consonants
 - Russian vowels assimilate to preceding consonants allophonically. Thus:
 - [y] occurs only after hard consonants; [i] occurs initially and
 after soft consonants (i.e. elsewhere)
 - Cyrillic has distinct symbols for [i] and [y] allophones, which violates
 the principle that alphabetic writing is based on phonemes.
 - Russian speakers find [i] and [y] to be perceptually salient, leading
 some [Leningraders] to call them separate phonemes.

My speculation was that the salience might be attributable to the unusual
orthographic situation. Here are Alexis' 3 refutations (paraphrased) and my
replies:

(i) Russian have no difficulty perceiving [y], even when spelled with "i"
 after "zh", "sh", and "ts".
reply: This is a pure orthographic convention, as AMR knows. /zh/, /sh/, &
 /ts/ are inherently hard consonants with no soft counterparts. So the
 perceptual salience is not easily confused. In fact, the confusion 
 here is in spelling--since native speakers sometimes write "y" here
 to mark the "hardness" of the preceding consonants.

(ii) Spelling is not enough because Russians taught to spell fronted allo-
 phones of other vowels in phonetic transcription do not find them easier
 to perceive.
reply: Assuming this claim is true, it is still weak. Lifelong immersion in
 a childhood-learned standard orthography hardly compares with adult-
 learned training in phonetic transcription.

(iii) Latin-based "i" vs. "y" letters arose in the Middle Ages because of the
 need to represent perceptually salient sounds.
reply: I don't know the Latin-based facts, but the emergence of "y" in 
 Cyrillic was for purely orthographic reasons. Cyrillic originally had
 front and back "yers" to represent short high vowels. These vowels 
 disappeared (or tensed), helping to bring about a major rephonologization.
 The consonant system became bifurcated into "hard" and "soft" sounds.
 Now the spelling system had too many vowel symbols and too few consonants.
 It evolved that the vowel symbols and (silent) yers became markers of
 hard/soft quality on preceding consonants. The "y" letter (called yerih)
 consisted of a back yer (hard sign) followed by an iota. This was needed
 to fill a gap in the orthography. My suspicion is that Jan Hus was 
 addressing a similar need in Czech writing, not a perceptual autonomy of
 [i] and [y] sounds. The whole point is that modern Slavic languages mark
 the consonantal dichotomy with following vowel letters.

The i/y controversy is famous in Slavic circles, and I don't think we can be
totally conclusive here. But I am still attracted to the notion that ortho-
graphy is the wild card here--the cause, not the result, of the unusual
perceptual autonomy of these two allophones for Russian speakers.

Finally, a swipe at SPE again: Halle's analysis of Russian, which led to SPE
analysis, is essentially historical internal reconstruction called synchronic
linguistics. It arrived at a phonological level that recapitulated the 
ancient phonology, with no inherently palatalized phonemes. This is entirely
counterintuitive to modern speakers. A hundred years ago, linguistic theory
developed a principle for explaining alphabetic writing. Generative theory
has no principled basis for explaining the nature of such writing. Is
this progress?

 -Rick Wojcik (rwojcikatc.boeing.com)

Alexis and I are more in agreement, than disagreement, on phonological matters.
But I do want to set the record straight on a couple of matters:

>...all [structuralists] agreed in placing the phonemic level exactly
>where Halle found it--and found it wanting, that is, AFTER all
>neutralizing rules and before all allophonic ones...

In general, yes. Recall, however, that Bloomfield allowed phonemic neutrali-
zation to non-phonemic segments (e.g. flaps)--cf. McCawley, "Sapir's Phonologic
Representation", reprinted in Adverbs, Vowels, and Other Objects of Wonder.
And I would be wary of too sweeping a statement on structuralists.

I think I'm in agreement with Alexis on the Halle argument, although I reserve
my scholarly right to cavil and bicker on minor points. ;-) So I disagree
on the following statement:

> As noted earlier, what I call Moscow-phonemic is the level called
> phonemic by Baudouin, by the Moscow phonologists, and by Stampe.

Leaving out Moscow, which used 'phoneme' consistently in this way, Baudouin
and Stampe (and Sapir) used it to refer to two different levels: the speaker-
based neutralizing level (M-phonemic) *and* the hearer-based perceptual level
(L-phonemic). Stampe (in the classroom) has tended to use "morphophonemic" 
more often than not to describe the speaker-based level. Baudouin didn't have 
that term and applied "phonemic" to both. Ulaszyn, as far as I can understand
him from your descriptions, noticed the confusion in his teacher's terminology
and invented the term "morphophoneme" to resolve it. He did not join the
Leningrad camp, which confused phonology (physiophonetics) and morphonology
(psychophonetics), and Stampe's usage seems quite consistent with Ulaszyn's.
Indeed, SPE phonology is essentially the Leningrad confusion minus the 
L-phonemic level.
 -Rick wojcik (rwojcikatc.boeing.com)

Alexis Manaster-Ramer writes (about Larry Hutchinson's critique of Halle'
s argument):

That is, both he and the lexical phonologists seem to want to have 
their cake and eat it too by allowing a phonemic level (which seems 
to be close to the Leningrad-phonemic one) but allowing phonological 
rules to apply in complicated ways, so that the same rule can
apply in some cases before this level and in other cases after it.

It's been quite a while since I've looked at Larry's paper, but if I remember
correctly, he adopts a completely neutral stance on whether in fact there is
a phonemic level. His concern, as I recall, is only with showing that Halle's
argument against it is fallacious; it doesn't follow from that that there
aren't nonfallacious ones.

Michael Kac

[End Linguist List, Vol. 2, No. 156]