form, function, data, description

[Matthew S Dryer]

In response to Alan Dench, what he says is largely consistent with what I
said, so let me repeat it and add some clarifying remarks: I said that I
was not aware of "any instance in which the descriptions provided by
formal linguists have been of value" to functionalists.  I did not say,
nor would I say, that there are no instances in which descriptive work has
been informed by work by formal linguists.  In particular, I think that
what Bob Dixon calls basic linguistic theory, the theoretical framework in
which most descriptive work has been done, has clearly been influenced by
various forms of generative theory, although I think the primary influence
in syntax has come from earlier generative theory (pre-1970).  I should
also qualify my claim so as not to exclude descriptive work by formal
linguists who use basic linguistic theory rather than current (or
then-current) formal theoretical frameworks.  For example, the descriptive
work by Dan Everett or Keren Rice's grammar of Slave are little different
in orientation from descriptive work by functional linguists, such as Li
and Thompson's grammar of Mandarin Chinese or John Haiman's grammar of
Hua, and are of equal value.  My claim is that there is very little if any
description using current or recent formal theoretical frameworks that is
of value to functional linguists.  I concede that if we go far enough back
in time, we do find partial exceptions to this, like Bob Dixon's grammar
of Dyirbal.

Matthew Dryer