Evolution, and 'functional' + 'social'

[Dan Everett]

If functional linguistics is right to believe that (most of) syntax is
motivated by communicative needs, then Ellen's quote below should be
underscored cannot be ignored by people seriously concerned with the
relationship between syntax and communicative function (I interpret
Ellen's quote as ironic), e.g. the readership of this list.

That is, saying that Creole languages are equally communicative as
nonCreoles, while simultaneously being syntactically simpler, is likely
oxymoronic. To say it is both is trying to have your cake and eat it too,
at least if you believe that communicative function accounts for much of
syntax.

Of course, to compare languages' relative syntactic complexity is not
merely a matter of adding up morphemes. Such a comparison is very hard.
Though I haven't yet read DeGraff's contribution, it sounds more plausible
on the surface of things than Parkvall's remarks, though, because it
seems to take syntax more seriously than what I could gather from
Parkvall's quote.

I suppose that, if one could show that there is a lot of syntax which is
nothing more than diachronic detritus, with no contribution to
communication, then one might maintain that Creoles, being newer, are
'simpler' in having less 'trash'. But I suspect that no one is going to
find this a very useful way of understanding the residue of historical
change, i.e. that any significant portion of it has no communicative
value. (Moreover, Bill Croft's theory of historical change, via the
evolutionary model it employs, is in a particularly good position to say
why this is.)

So is Creole syntax (not the number of inflectional or derivational
morphemes it has, which is a non-useful metric) as complex as non-Creole
syntax? That seems to be part of DeGraff's research results and programme
and a worthwhile, complex endeavor.

-- Dan Everett

> Ellen Prince said:
********
> Hint: It's the syntax -- i.e. those pesky things that 'are not
> motivated by communicative needs' -- that are at issue here.
>