Re: THEORY: Ergativity and polypersonalism

 I've heard people argue that Ergative languages can come about by the
reanalysing of previously more marked (passive usually) structures as
the standard. Perhaps this provides a link? If a language is highly
synthetic, then it's likely that it adjusts valence etc using bound
morphemes rather than with constructions involving separate words... It
seems to me more likely that a language where such endings are bound
could do such a reanalysis than one where separate words are involved.
For instance, pretend English did such a thing:

I went to the cinema and saw a film and then bought popcorn.....
I went to the cinema and A film was seen by me and then pop corn was
bought by me and.....

This obviously adds a hell of a lot of words, so I'm not sure speakers
of English would do it because of the extremely high cost involved. By
contrast, I'd say the typical synthetic language tends to be more
compact than that. For example, in Swahili the passive is -w- I believe,
inserted before the final mood marking vowel. So there's a little less
wastage there.
 So there might be a weak link between highly synthetic languages and
ergative languages maybe? Not because synthetic languages are inherently
more likely to be ergative, but because compact passive (or similar
valence adjusting operation) marking would make such a reanalysis more
likely....

Reply