Re: To Matt Pearson

--- Kou wrote:
> Well, human language is rule-governed, pure and simple, so there's got to
be a set of rules in our heads somewhere.

Says who? I'm no scientist, but statements like "X is true, so Y *must* be
true" (a *hypothesis*) fail the scientific method litmus test, unless one
can come up with proof that Y is true ("Apples fall, so oranges must fall."
is a hypothesis where we can test if oranges fall.) (As opposed to: Apples
fall, I've seen an orange fall, apples and oranges are round, so all oranges
*must* fall.)) When *we* were in school :) linguistics profs seemed to want
to make linguistics legit as a science (as opposed to a discipline) by using
scientific language, but I frequently found the Y part lacking.
--- end of quote --

As I understand it, the chain of reasoning is:

[a] Language is rule-governed: utterances systematically obey discoverable rules.
(This is a hypothesis, of course, but the evidence seems to be strongly in its
favour.)
[b] Language is in our heads: we have linguistic competence. (I take this as a
given; the only alternative is that supernatural creatures are speaking through
us. :-))
[c] Given (a) and (b), it follows that there are rules in our heads.

I know that this is not a proper syllogism or anything, and it may not strictly
follow the scientific method (I know far too little about the history or
philosophy of science). But I just can't see how [c] could be a controversial
claim.

Notice that I've said nothing about *what form* the rules of grammar should take
(generative grammar offers one of many ways for stating such rules formally).
I've also said nothing about *how* those rules are instantiated in our brains
neurologically (generative grammar has little to say about this, but
neurolinguists and psychologists have started making some guesses). But that
our grammatical competence includes a set of rules, and that these rules are in
our brains--somewhere, somehow--strikes me as an entirely straightforward
assumption.

In Natural Science, no hypothesis that is non-trivial is absolutely provable (and
Natural Science of the human mind is even worse, considering how complex and
abstract the human mind is). But in practice, most natural scientists seem to
implicitly accept a 'threshold of plausibility' in evaluating hypotheses: Once
a sufficient amount of evidence has built up which supports a particular
hypothesis, that hypothesis is treated *as if it were absolutely proven* and
becomes a "fact". Natural selection is an example of a hypothesis which
evolutionary biologists treat as a "fact", because the evidence for it is so
overwhelming and because no better explanation for that evidence is
forthcoming. That human beings have some system of rules of grammar in their
heads seems to me to be a "fact" in this same sense.

--- beginning of quote ---
> Generative grammar seeks to find out what those rules are. What we
actually *do* with those rules when we speak and understand sentences in
real time is, in principle (and in practice), a different issue.

While I appreciate that generative (transformational?) grammar seeks to
provide us with paradigms to explain grammar rules that we in the here and
now can understand cognitively, that is no proof that that is how the mind
works. (Processing research sounds intriguing.) (Freudian models are/were
useful to explain dreamlife, but trains into tunnels as sexual intercourse,
all the rage in their day, seem not so applicable anymore [though I don't
think one throws Freud out with the bath water.]) (And was it scientific?).
-- end of quote ---

As I tried to explain in my earlier response to David, generative grammar is
explicitly a theory of knowledge, not a theory of process/behaviour. This is
where Chomsky broke ranks with the social science tradition of his day: The
behaviourists argued that you couldn't have a theory of knowledge which was
scientifically rigorous. Chomsky disagreed, and brought up some pretty
convincing counterarguments to the prevailing 'stimulus-and-response' theories
of language. But the debate he started is still going on. There are plenty of
people around today (mostly biologists and psychologists) who would argue that
generative linguistics is no more "scientific" than Freudianism.

My hope is that this rift between the materialists and the science-of-mind people
will begin to close up as we learn more about the physiology of the brain and
the neurological foundations of cognition. But we'll have to see...

--- beginning of quote ---
(Here again [and I took those notorious Ling101 classes back in the 80s
which you referred to]: the premise was that, like a computer, the human
mind must use the simplest rules available for language [a specious claim,
and far from the scientific method]) if for no other reason than efficiency.
I balked, 'cause the "there" transformation made no sense (I often hear the
speech 'error' by native speakers, "There's books on the table." but not in
the original "core" sentence, "Books are on the table." [the transformation
would involve inversion etc.] Then there are sentences like, "Let's go,
shall we?" and "Stay, won't you?" which require untold mental gymnastics to
find a sequence of transformations that work. (Linguistics prof offered a
free "A" to anyone who solved that dilemma). If there's a gaping hole in
your theory...
--- end of quote ---

The absence of a complete account of inversion or eixstential constructions is
hardly grounds for abandoning generative grammar (*no* theory has an entirely
satisfactory account of these phenomena). And anyway, so what? An incorrect (or
inadequate, or unverifiable) hypothesis does not falsify the theory within
which that hypothesis is couched. To hold generative syntax to a higher
standard of explanatory adequacy than other theories or disciplines--especially
considering how young the science of linguistics is--is hardly reasonable.
We're working as hard as we can on a very intractable problem! :-)

--- beginning of quote ---
The field may have come a long way in twenty years, and I would genuinely
love to read the next wave of research, but if it takes on the same ol'
hubris of "this is the way it is", then I fear "we haven't come a long way,
baby."

> This is a very common-sense attitude,
as is: "the mind must work economically, like a computer" (I hope you
realize I'm not taking you personally to task for what my prof said 20 years
ago)
--- end of quote ---

Yes, I realise that.

The hubris you speak of is--or should be--purely rhetorical. I was taught that the
only way to proceed is to make the strongest possible claim you can on the
basis of the evidence you have. The stronger your claim is, the easier it is to
falsify--and the best theory is the one that's most easily falsified.

As a scientist, one is obliged to heed Occam's Razor (and Sherlock Holmes) and
assume that the simplest theory which fits all the facts is the correct one.
You start with the most straightforward theory possible (e.g., the mind is
maximally economical) and work from there. If you start from the assumption
that things are complicated--i.e., that a system is more complex than it needs
to be, or that it does not operate according to rules, or that it operates
according to rules which are ultimately undiscoverable--then there's no way to
make scientific progress.

Gotta go prepare for class, so I'll end here.

Cheers,
Matt.

Matt Pearson
Department of Linguistics
Reed College
3203 SE Woodstock Blvd
Portland, OR 97202 USA
ph:  503-771-1112 (x 7618)

Reply