LINGUIST List 4.629

Tue 24 Aug 1993

Disc: The Theoretical Status of Marginal Utterances

Editor for this issue: <>

Directory

1. Mark Seidenberg, theoretical status of marginal utterances

As someone who doesn't actually do linguistic theory but is an
interested bystander, I have a question about the theoretical role of
utterances (typically complex ones) whose grammaticality is uncertain.

Within the Chomskyan tradition, there seem to be two common attitudes
toward such sentences.

One is that rather than being a problem for grammatical theory, such
utterances play an essential role in theory development Any theory can
handle the simple cases, the argument goes, and it's only in regard to
their ability to handle subtle aspects of language structure that
grammars are interesting, revealing, etc..

The other attitude is that grammar is a theory of the well-formedness
of utterances and therefore we should use the theory as a basis for
disambiguating marginal cases. Thus, an adequate theory should provide the
basis for deciding if the suspect utterance is grammatical or not.

I do not understand how these assumptions fit together. If
the theory is supposed to decide the unclear cases, how can the
unclear cases provide a basis for developing the theory?

I realize that there is another school of thought that holds
that these issues only arise if your approach to linguistic
theory assumes a strong competence-performance distinction and a
binary notion of grammaticality. So, if you don't hold these
assumptions, there's no problem. But what if you do?

Mark Seidenberg