I also just finished reading 'Words and Rules'. I read it for a very
specific purpose: I plan to write a book on morphology (how words work)
and Pinker's book is a convenient summary of a recent major debate. So
I'm somewhat on the inside of this issue in linguistics.

The point of his book is to prove, not that affixation of irregular
verbs and nouns is NOT computed by rule, but that affixation of REGULAR
verbs and nouns IS computed by rule. Pinker also argues that the parts
of the brain that do each kind of word -- regular vs. irregular -- are
at least to some extent physiologically separate. To skip straight to a
short summary of his point, skip down to the paragraph that begins 'To
sum it up broadly...'

Pinker is responding to criticisms of the general line of linguistics
founded by Noam Chomsky and his followers. Chomskyan linguistics and its
descendants like the idea that the various 'components' of language --
syntax, phonology, semantics, etc. -- are at least to some extent
independent of one another, and that each has to be explained on its own
terms (that is, that you can't explain one [for instance, syntax] in
terms of another [for instance, semantics]). ('Interfaces' between the
components are recognized as necessary, though.) Components of language
that are independent of semantics use some kind of formal computation (a
la logic, mathematics) to accomplish parts of language production (e.g.,
adding suffixes to regular verbs).

A line of theorizing that is sometimes called 'alternative linguistics',
and falls generally under labels such as 'functional', 'cognitive', or
'discourse' linguistics challenges this idea, and suggests that perhaps
no special brain area or cognitive mechanism is needed for language --
language is done with the same parts of the mind that take care of
general higher cognition. Many linguists in this alternative view also
question the independence of language components from one another.

This is a major split in current linguistic theory. One's job
eligibility can be impacted by which 'ideology' one follows, for example
-- some ling. depts. will only hire a Chomsky-type, others only an
alternative type -- and there is often not a little animosity involved
in the debates, when the two camps actually meet in debate (they
function to some extent separately from one another).

Pinker was critiquing in particular the connectionist approach, which
uses a particular kind of computer modeling to simulate actions of the
brain. The model imitates the brain by being configured as a network of
processing units that are connected to one another, imitating the
structure of neurons connected by synapses in the brain. The computer is
programmed to 'learn' rules of language, and then the resulting
configuration of the network is studied to see if one can 'find the
rule' in the patterns of processing activity the network develops in the
process of learning. Connectionist networks are one kind of 'neural
network', a relatively new computer modelling technique. Connectionists
have not been able to 'find a rule', but instead find a very complex
pattern of activation strengths distributed across the network.

To sum it up broadly, Pinker is responding to connectionist claims that
both regular and irregular verbs are stored and assembled for speech in
the same brain apparatus and by the same means. A lot of cognitive
linguists model both regular and irregular affixation patterns in what
would be called the 'lexicon', the word storage area. He wants to prove
that these ideas are wrong, and that, even if we can posit a
network-like storehouse for irregular words and vocabulary in general,
we still need a computational section of the language faculty that is
responsible for regular affixation. Pinker is actually conceding a good
bit to functionalists/cognitive linguists by adopting their model of the
lexicon. But he insists that regular affixation is accomplished by a
distinct part of the language faculty. So he is trying to 'have the best
of both worlds'. This is a pretty welcome stance, given the acrimony
that has characterized the debate in the past.

I have so far read only one review of the book; so far as I have been
able to find out, only one linguistics journal has commissioned a review
to date. The other existing reviews are in cog. science journals or the
general media. If I were to get into the details of what might be wrong
with Pinker's argument, I'm sure I would lose whatever of the ATEG crowd
I haven't lost already. In general, some of his claims are a bit shaky
(esp. with regard to German plurals). I also think that he sets aside
data that doesn't agree with his theory (I need to check this, but I
believe that it's pretty common in British English to form compounds
such as 'antiques dealer' with regular plurals -- something which should
not be common, acc. to Pinker). Also, alternative linguistics has a way
of including regular rules in the lexicon that Pinker almost achieves,
but he doesn't quite go there. From my personal point of view, I also am
suspicious of the evidence he cites from psychological experiments.
Results are rarely as clean-cut and unambiguous as he presents them to
be. However, I have not read the original studies he cites, so cannot
critique them.

I (a cognitive linguist) do not stake my life on the idea that language
is NOT centered in various, sometimes independent parts of the brain,
but I'm not totally sold on the idea either. Pinker's book presents some
strong arguments, but for me the jury is still out, at least until the
psychological studies he cites can be fully critiqued, and more data can
be discussed. I'm more convinced than I was before I read the book, but
have several questions.

I find it hard to believe that this would be a popular book, simply
because of the arcane nature of the issue and the pace at which he
waltzes us through the argument. The details one needs to keep track of
are quite complex for even the average educated reader, if said reader
doesn't have a strong background in linguistics. Pinker and his
publishers seem to think there is a large audience of people out there
avid for this kind of exposition. I'm doubtful. Maybe I've just been out
of the major-research-institution ivory tower for too long. I can
imagine Mensa members (no slur intended, you Mensa members out there!)
and cog-sci freaks would enjoy the book, but I can't see it appealing to
a wide audience even of college-educated people. I wonder how the book
is selling.

It is, however, a nice model of how to construct a really complicated
argument!

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Johanna Rubba   Assistant Professor, Linguistics
English Department, California Polytechnic State University
One Grand Avenue  • San Luis Obispo, CA 93407
Tel. (805)-756-2184  •  Fax: (805)-756-6374 • Dept. Phone.  756-259
• E-mail: [log in to unmask] •  Home page: http://www.calpoly.edu/~jrubba
                                       **
"Understanding is a lot like sex; it's got a practical purpose,
but that's not why people do it normally"  -            Frank  Oppenheimer
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~