pseudo-generalizations

[Eitan Grossman]

Regarding the matter of 'explanations' in linguistics, especially of the
cognitive or functional sort, I think that they are not all that different
from generativist 'explanations.' Whether one looks for motivation or
explanation in some imaginary mental 'organ' or in some version of the
economy principle, one is making one's job much easier, by refusing to treat
one of the more intriguing and demanding aspects of language: it
systematicity. After all, after one has decided that a given language has a
certain pattern or feature, it can be compared superficially with other
languages that have the same pattern or feature. However, the problem of
structural value must also be addressed. It is not enough to note that two
given languages have, for example, a nominal-rheme sentence pattern; one
should also determine the value of the nominal sentence, its oppositions and
neutralizations vis-a-vis other nexus patterns, as well as its internal
structure. It is only at this point that valid and important
cross-linguistic comparisons can usefully be made. I realize that this would
slow down the accumulation of cross-linguistic data and complicate the
identification of valid cross-linguistic generalizations -- but one would
gain a much sharper resolution as a result. Additionally, I would venture to
guess that many 'pseudo'-generalizations would be winnowed out.
I realize that such blatant structuralism is probably seen as somewhat
outdated, but judging from what I have read in contemporary typological
research, it would probably do no harm.

Eitan Grossman
Jerusalem

_________________________________________________________________
Add photos to your e-mail with MSN 8. Get 2 months FREE*.
http://join.msn.com/?page=features/featuredemail