Corpora: Chomsky/Harris - one more fun question.

[Mike Maxwell]

Carl Mills wrote:

>But the virtual "data" that generative linguists use to
>"test" their theories and resolve "empirical"
>questions turn out to be the linguist's own intuitions
>about the grammaticality of their made-up sentences.
>I have spent 25 years marvelling at generative linguists'
>grammaticality judgments and at their refusal to change
>them in the face of native-speaker denials.  If
>probabilistic linguistics and MIT linguists need
>to come together it is in the area of grammaticality/
>acceptability judgments.

I have no argument with that.  It certainly is a weak point of some
generative linguists' papers, and has been commented on (especially by
generativists whose theories conflict with the cited judgments :-)).  I
don't think it's a real common problem, but it has probably become more
common in recent years as theories have become more observationally
adequate, and therefore linguists have to look harder for evidence to decide
between theories.

For the same reason (it's getting harder to distinguish theories
empirically), the use of syntactic evidence from languages other than
English appears to me to be more common. (Which is not to say that it wasn't
done before--see Ross's dissertation.)  And of course if theories are indeed
more observationally adequate, it may also be harder to find reliable corpus
evidence: rare data is potentially unreliable data, since it could simply be
a missteak on the part of the speaker/ writer (or intentional, like a pun).

Occasionally linguists elicit opinions from other linguists.  Trying to do
the same with your mother-in-law, assuming she is not a linguist, will
probably get you an opinion, but it may not be the one you're looking for!
All seriousness aside--asking for grammaticality (I should say
"acceptability") judgements from the Man On the Street can be difficult,
since people often focus on other things than grammaticality.

I could go on, but I think there's actually a literature on this issue, so
I'll defer to someone who is familiar with it.

                                         Mike Maxwell
                                         Summer Institute of Linguistics
                                         Mike_Maxwell at sil.org

P.S. Thank you, Carl and Jim, for the nice words!