LINGUIST List 12.1820

Sat Jul 14 2001

Disc: Nonobjects of Syntactic Study

Editor for this issue: Lydia Grebenyova <lydialinguistlist.org>

Directory

1. Dan Everett, Re: Nonobjects of Syntactic Study
2. Dan Everett, Re: Nonobjects of Syntactic Study

Overall, I agree with your posting, John, in just about every point. But let
me say just how.

When I came into linguistics, in the 70s, my first teacher was Kenneth Pike.
Aside from the fact that I found it very hard to understand much of what he
said (the concept of 'introductory level' seems not to have been one he was
familiar with), I found his way of thinking about language extremely
liberating. I could look at data and I had the right to say anything I
wanted about the data (*except* what might be the *cause* of it. Nida in his
_Linguistic Interludes_ talks about the childishness of asking the question
'Why' instead of 'What' - but I wanted to ask 'Why'). I then took a course
in Generative Grammar in Brazil. This was even more liberating. Chomsky's
theory not only allowed me to ask and answer 'What', it explicitly took up
the question of 'Why', answering it in terms of innateness. That was
extremely appealing to me. I was bothered at the time by calling everything
that could be a subject an NP, even when they were PPs or Ss, and I wondered
why the 'Why' stopped with innateness, but, overall, it was the most
exciting intellectual experience of my life. For more than 20 years it
motivated my research and I was utterly convinced by the rightness of
Chomsky's review of Skinner, his intellectual histories, the linguistics,
etc. This was science.

I think that my own questioning of all this follows the progression you
mention for those of us who 'came of age' in the last half of the last
century. What was initially liberating began to raise more questions and the
answers it provided became ever less convincing. When one looks, for
example, at theories like Role and Reference Grammar, which are able to
account for the ECP, Subjacency, etc., without appeal to derivations,
innateness, UG, etc. then one asks what work UG is doing and, well, the itch
begins.

Your characterization of positivism as initially liberating sounds quite
right to me, resonating with my own reading, especially Russell's
discussions of it. But, as you say, it has led us to the position where, at
least among those of us who considered ourselves Generativists, the best
kind of research result is to name a new principle. So we have a
proliferating collection of so-called 'generalizations' named after this or
that Generativist, which, upon closer inspection are often little more than
spurious observations, far from generalizations. Especially as one does
fieldwork it becomes more and more difficult to keep the faith. 'Theoretical
research' on, say, an Amazonian language to some researchers is often just
adding a new node to make the data fit. Now that might be the result of a
lack of cleverness on the part of the researcher, but I have come to think
of it instead as a problem with the structuralist strains in Generative
research, whereby semantics and function may never be directly causally
implicated in the statement of a Generative solution (Chomsky eliminated
'indices' from the theory of reflexives, for example, because they led to
non-structural solutions, the only kind of solutions which Chomsky
understands to follow from 'virtual conceptual necessity').

Scientifically realist Cartesianism *and* positivism lead now to what I
consider to be unliberating intellectual positions. Even the term 'paradigm'
sounds jaded and less than useful. The American pragmatists, especially
James - whose brilliance I long underestimated, proposed that we study what
is useful, being concerned with usefulness more than the worn-out Popperian
notion of falsifiability. But will pragmatism outlast what you allude to as
the Hegelian march of ideas? No, because style and innovation will never
stay static. But pragmatism, in the variety I find most attractive, Rorty's,
simply says that 'Truth' is what will convince all future audiences. That
sounds like a goal we are unlikely to ever meet, so 'Truth' is not something
pragmatism commits itself to. UG is a form of 'truth-commitment' and, so,
inherently unattractive. UG will not go gentle into that good night, but nei
ther has the 'Oedipus complex' or the 'id'. But they are from the same
arsenal ultimately. Pragmatism does not project itself into the future as a
model of truth or rightness of theory. But where people feel liberated and
feel that the story is useful, there is Pragmatism, to a degree. UG cannot
be studied. I like theory too. But there is no particle accelerator to
support UG in the way that Einstein's view of the atom has been shown to be
more useful than Mach's. There cannot be.

Right. Not possible to resolve this on email, which is why I am writing on
it off-line so much these days and why, I suspect, you have so long been
concerned with these issues. They are profoundly important as we think of
the nature of our discipline in a post-Chomskyan era.

Dan

I hasten to add to my last posting that much of what I say on nonobjects of
syntactic study will be old information to linguists like George Lakoff,
Paul Hopper, Tom Givon, Ron Langacker, Syd Lamb, Joan Bybee, Haj Ross, etc.
But to the degree that their own research falls within the 'scientific
realism' tradition, then what bothers me in Chomskyan structuralism will
continue to bother me in Cognitive Linguistics, etc. The bother is of a
lower order of magnitude, however.

Dan Everett