LINGUIST List 7.608

Wed Apr 24 1996

Disc: Lang & movies, Phonology, Citing

Editor for this issue: T. Daniel Seely <dseelyemunix.emich.edu>

Directory

1. Ernest Scatton, Re: 7.577, Disc: Lang & movies, Ungrammatical sentences
2. Deborah Schmidt, Substantive phonological constraints and UG
3. Peter GU Dept of Education HKU, Citing electronic information

Message 1: Re: 7.577, Disc: Lang & movies, Ungrammatical sentences

Regarding the discussion of accents in movies of the 30s and 40s: I've 
always been struck by them. The contrast between then and now is shown 
very vividly in the Hudsucker Proxy of the Coen Brothers. Jennifer Jason 
Lee is particularly good in this respect, reproducing the style of the 
fast-talking, hyperactive reporter. 

*****************************************************************************
Ernest Scatton Germanic & Slavic Hum254
518-442-4224 (w) UAlbany (SUNY)
518-482-4934 (h) Albany NY
518-442-4217 (fax) 12222
 cnsvax.albany.edu/~alin220/slav_dept (WWW)

**I am posting these thoughts in order to stimulate discussion, with the hope
that I might gain a deeper understanding of these issues and that I might be
set straight in any illogic.**

According to many proponents of Optimality Theory, one of the obvious
advantages of assuming that phonologies are made up wholly of UG
constraints, as opposed to idiosyncratic language-particular rules, is
that the existence of common cross-linguistic tendencies is thus
accounted for. For example, the fact that nasal place assimilation is
so common, while nasal place dissimilation is (I think) unheard of,
might be due to there being a UG constraint penalizing unassimilated
NC clusters. For another example, the reason that +lo,+ATR (surface)
vowels are often absent from surface vowel inventories might be due to
there being a UG constraint penalizing the cooccurence of +ATR with
+lo.

However, it seems to me wrong to want to account for the existence of
common cross-linguistic tendencies (substantive as opposed to formal
ones, at any rate) with appeals to UG constraints, if UG is understood
to be the presumably hard-sired design of the computational system.
Common cross- linguistic tendencies arise, I believe, not from any of
the inherent logic of UG's computational system, but instead from the
acoustic and articulatory demands of its phonetic interface.
Accordingly, the reason nasal place assimilation is so common
crosslinguistically has nothing to do with any putative UG constraint
against unassimilated NC clusters, and everything to do with the fact
that human beings are all born with essentially the same articulatory
physiology. Same goes for why +lo,+ATR vowels are cross-
linguistically rarer than +lo,-ATR vowels.

Now, some will rightly object to my straw-man characterization of
Optimality Theory. It has, after all, been proposed that the UG
constraints of OT should really be conceived of as purely schematic
(though much of the OT literature seems not to have taken this
proposal to heart). This leaves open the possibility that the
insertion of substantive feature content arguments into these purely
schematic UG constraints may indeed be entirely controlled by
considerations external to UG. I shall return to this line of
thinking below. In the meantime, I would like to suggest how the
existence of common cross-linguistic tendencies can be very nicely
accounted for in the old rule-based model.

Darwin noted how biological family trees were like linguistic
genealogies. We might now take inspiration from his work, and think
about how his theory of natural selection might apply to the Origin of
Dialects. Imagine that a language-acquiring child (or an adult
learning a second language) spontaneously posits a novel phonological
rule that has no precedent in the target language. This can be seen
as analogous to a random genetic mutation. The acoustic and
articulatory demands of the phonetic interface can be seen as
analogous to environmental pressures. Novel rules whose effect is to
increase the phonetic fitness of surface phonological structures (by
leading to phonetic interpretations that are more perceptually
salient, or easier to articulate) would be naturally favored for
survival (in the grammar under construction) and propagation (to the
grammars of subsequent generations of speakers) over most other
potential new rules. Randomness and the fact that acoustic pressures
generally push in a direction different from articulatory pressures
ensure that diachronic change does not proceed in one narrow directio
direction. Nevertheless, given the mechanism of natural selection, we
should not be surprised by the semblance of output goal orientation.

Given the mechanism of natural selection, the semblance of output goal
orientation is a product of the demands of the phonetic interface
guiding diachronic change. It is not a product of innate UG
constraints. Natural selection accounts for convergent evolution in
biology, and it can be recruited to account for common
cross-linguistic tendencies among different languages' phohologies,
too.

Now, I'm not saying that I don't believe in UG. It seems to me quite
reasonable to suppose that UG weighs in on such formal matters as, for
example, whether or not all distinctive features must be monovalent
binary (+F versus 0F, as opposed to +F versus -F or 1F versus 2F
versus 3F), whether all assimilation is by spreading rather than
copying, and whether a feature residing on organizing node X can
spread only to another node X, and not node Y, of a neighboring
segment.

To assume that languages change by adding new rules (and occasionally
reanalyzing URs, etc., when rule systems become too cumbersome) is
consistent with familiar views of diachronic change, and that is
partly the reason that I suggested above that the counterpart to a
random genetic mutation in biology is a spontaneously posited novel
rule. Can a comparable story about how natural selection naturally
leads to common cross-liniguistic tendencies be told if, in place of
random and idiosyncratic potential new rules, we imagined random and
idiosyncratic fixing of arguments within UG constraints and random and
idiosyncratic reranking of those constraints, with only those
innovations that confer increased fitness in some phonetic dimension
having any real chance of sticking? I suppose so. But if common
cross-linguistic tendencies result from phonetic prossures guiding
diachronic change, then Optimality Theory has no advantage over the
rule-based model with regard to accounting for them.

Your thoughts please?

David Weiss asked about reference styles for electronic information.
Here's a message I recently sent to the BILINGUAL list on bilingual
education and language planning. Hope it helps.

- ---------
We occasionally need to refer to articles or whatever info we see over
the net (including our discussions here) when we are writing our
conventional stuff. How exactly we should do this then within the
framework of APA (American Psychological Association) or MLA (Modern
Languages Association) publication formats remains a technical nuisance.
Some people are working on this, and I've posted some relevant articles
and URLs. I however found the following URLs much more comprehensive:

For a quick reference to APA style (4th ed., 1994) when you don't have
the manual at hand, try the following URL for the
APA Publication Manual Crib Sheet
http://www.gasou.edu/psychweb/tipsheet/apacrib.html

For APA style for electronic info, try
Web Extension to American Psychological Association Style
http://www.nyu.edu/pages/psychology/WEAPAS/

For a similar page on MLA style, try
MLA-Style Citations of Electronic Sources
http://www.cas.usf.edu/english/walker/mla.html

For the most comprehensive guide so far on both APA and MLA formats on
electronic reference, try
Bibliographic Formats for Citing Electronic Information
http://www.uvm.edu/~xli/reference/estyles.html


=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=
Peter Y. GU HW207 Robert Black Campus I, Department of English
Hong Kong Inst of Education, HK Email: h9290037hkuxa.hku.hk
Tel: 2601-1528 (H) 2361-7108 (O)