changing FUNKNET topics

[Martin Haspelmath]

Could we please stop discussing Chomsky all the time on Funknet? It
would be a sad sign for functionalism if we had no more burning issues
to discuss than Chomsky's place in history.

In case you cannot think of any other worthy topic at the moment, here
is a question that I've been asking myself in recent months, without
coming to a conclusive answer:

What is the role of functionality in language change?

Many functionalists have recently stressed that functionally adapted
structures come to be functional through language change. That is, when
we say that some synchronic structure is, say, economically motivated
(e.g. the fact that singulars are mostly zero and plurals are mostly
overt), this doesn't mean that language structure is the way it is
because it's economical. My English plurals are overt not because this
is economical, but because I learned to speak like other English
speakers -- if they had overt singulars and zero plurals, I would have
acquired such a perverse system as well (i.e. markedness universals such
as this one are not due to innate restrictions on acquisition). So
economical structures (and more generally, functional structures) must
arise in language change, but how exactly?

Bill Croft has argued ("Explaining language change", Longman/Pearson
2000) that functional motivation comes in exclusively through the
actuation of language change, i.e. innovations of individual speakers,
which are (or may be) functionally based. The spread of functionally
adapted structures plays no role, acording to Croft, because the
diffusion/propagation of new features is exclusively socially based.
Using biological terminology, we can say that functionality resides in
the mutations, not in the selection. On the other hand, Daniel Nettle
("Linguistic diversity", OUP 1999) and I ("Optimality and diachronic
adaptation", Zs. f. Sprachwissenschaft 1999) have argued that functional
adaptation also comes about through selection, i.e. speakers adopt new
features from other speakers also because they seem structurally useful,
not just because they are socially attractive.

So in brief, the two positions are:

(i) Functional adaptation comes about exclusively through functional
innovation ("mutation") (and social selection is non-functional)
(Croft),

(ii) functional adaptation comes about through functional propagation
("functional selection"), in addition to functional innovation (Nettle,
Haspelmath, and no doubt others).

Although I have argued in print for the second position, I'm not very
sure that Bill Croft is wrong. What do you all think?

Martin

--
Martin Haspelmath (haspelmath at eva.mpg.de)
Max-Planck-Institut fuer evolutionaere Anthropologie, Inselstr. 22
D-04103 Leipzig (Tel. (MPI) +49-341-9952 307, (priv.) +49-341-980 1616)