LINGUIST List 8.826

Tue Jun 3 1997

Disc: Evolution analytic > synthetic

Editor for this issue: Ann Dizdar <annlinguistlist.org>

Directory

1. martin.haspelmathsplit, Disc: Evolution analytic > synthetic

As far as I can tell, Melanie Misanchuk's query

> I once read that the natural evolution of a language is from
> analytic to synthetic. I've been unable to find that assertion
> since, and am wondering if I made it up.

has been answered very inadequately. Since her summary was posted on
LINGUIST, I feel that some corrections are necessary.

Judging from some of the replies Melanie received, the view is still
widespread that linguistic evolution can go in both directions, i.e.
that there is no inherent directionality in language change. This is
wrong. At the same time, we can't say globally that languages tend to
change either from synthetic to analytic (as some 19th century
linguists thought) or vice versa.

What is going on is quite simple: We have to look at CONSTRUCTIONS,
not at LANGUAGES. We then see that there is a universal directionality
of change: Analytic constructions always turn into synthetic
constructions (unless they die out), and synthetic constructions are
always replaced by newly created analytic constructions.) That is,
synthetic constructions never turn into analytic constructions, and
analytic constructions are never replaced by newly created synthetic
constructions. (So both synthetic > analytic and analytic > synthetic
are universal, but in different senses.)

When quite a few synthetic constructions are simultaneously replaced
by analytic constructions, then one can get the impression that the
language as a whole changes from the synthetic type to the analytic
type -- this is what 19th century linguists such as August Schlegel,
Wilhelm von Humboldt, August Schleicher, Max Mueller and others
emphasized. (The cycle mentioned by Rob Pensalfini was first described
by Georg von der Gabelentz in 1891.) At the same time, 19th century
linguists (e.g. Franz Bopp and William D. Whitney) also recognized
that the development from analytic to synthetic (which they called
agglutination, now called grammaticalization) is possible and common,
but they were confuded because they tended to think in terms of
languages rather than constructions.

It is true that 19th century linguists had some preconceptions about
their own languages (actually, Latin and Greek) being the best
languages, but this does not mean that all their insights are totally
discredited.

Martin Haspelmath, Free University of Berlin/University of Bamberg