Re: your mail

On Tue, 22 Feb 2000, taliesin the storyteller wrote:

> "And Rosta" <a.rosta@...>
> Bcc:
> Subject: Re: Stress and syllables
> Reply-To:
> In-Reply-To: <Pine.GSO.4.00.10002221349140.17772-100000@...>
> X-Disclaimer: I speak for myself. Beware of humor.
>
> * Fredrik Ekman (ekman@lysator.liu.se) [000222 14:17]:
> > David Bell wrote:
> >
> > > While the "patterns" seem to make sense to me, I, unfortunately, have
> > > no understanding of the theory behind them.  I guess I'll have to
> > > study more phonology.
>
> I assume the framework And was using sorts under Autosegmental Theory,
> and any Phonology 101 class or book worth its name should at least
> describe that briefly and give pointers to more... we were treated to
> it already in the preliminary introduction to language and linguistics
> which everyone wanting to study any variant of linguistics -and/or- any
> language have to survive :)
>
> Hmm.. And? Maybe we should work together on a brief tutorial or
> something? Or maybe I can convince my phonology-prof to scribble
> something down...
>
> > I will have to sympathize with David here. While the subject matter is
> > extremely interesting, I feel at a loss to grasp the underlying concepts.
> > Could anyone recommend a good text (which would not require a doctorate in
> > linguistics to understand) on the subject of stress and syllable
> > construction from a general point of view? Preferably one available on the
> > Net or that could be found in most university libraries.
>
> I've spent quite a lot of time looking for such an online resource, no
> luck yet. If you are willing to have a look at texts heavily colored by
> the theoretic framework[*] they are based on, there's the Optimality
> archive at http://ruccs.rutgers.edu/roa.html.
I would not recommend this for the faint-of-heart. Most of the
papers (if not all) assume at least an upper division
understanding of phonology, and a thorough grounding in the
generative enterprise. Also, the papers have not been
peer-reviewed, and many are just plain bad.

Mike Hammond's writings have always been very clear; he's just
published a book on English Prosody, published by Oxford
University Press. I'm reading it at the moment myself, and he
spends a lot of time talking about the distributional
generalizations of English. The analysis is OT, but I think he
provides enough background to make his points clear even to the
OT innocent.

> If you're hunting for introductory books, I'd recommend those by Francis
> Katamba or Iggy Roca (Peter Ladefoged for the phonetic side of things,
> of course). Katamba's is more general, Roca's describe the various
> contemporary and competing theories better, IMHO.
A faculty member here has just reviewed Roca's introductory book
and was very critical of it. I read a draft of his review (and
Roca's comments on it) as well as sections of the text itself.
He makes some very idiosyncratic decisions about the phonology
of English, and I feel I should caution anyone who would like to
use this as an introduction to phonological theorizing (as well
as the phonology of English). I haven't read Katamba's book, but
if it's anything like his morphology text, I would recommend it.
Another text worth looking at is by Alan Kaye; the title is
_Cognitive Phonology_ (don't get too excited, Ed! Not that kind
of "cognitive" :-). The framework he adopts is also generative,
but it is a slightly different take on syllabic constituency and
the makeup of segments.

> [*] Well I guess using trees everywhere seem natural once you're used to
> it but optimality-theory is just plain weird :) unless written by Marc
> van Oostendorp (great guy tho' he's something of an Esperanto-fanatic).
:-) Nothing in OT precludes the use of trees; in fact, in my
dissertation, one of the crucial arguments about the segmental
status of a series of final consonants is based on their
syllabification preceding the accusative suffix. I have noticed
a trend in OT, though; current OT is moving away from
assumptions of phonological constituency that cannot be borne
out by surface patterns. A great example is the virtual
disappearance of Feature Geometry in OT work. I suspect that
given enough time, syllable structure may be pruned back
considerably (if not outright eliminated).

The really hot stuff right now is accounting for analogy and
statistical patterns (i.e., patterns which are largely, but not
universally true). It's stretching OT to the breaking point,
which is a Good Thing, IMO. Not because I dislike OT, but
because it points to a certain maturation of the field that it
is recognized that there can be regularities that aren't
universal. Phonologists are growing up.

Dirk

--
Dirk Elzinga
dirk.elzinga@m.cc.utah.edu