Friday, November 20, 2009

Movement and universality

This release was just passed along by a contributor and ultimately comes from an esteemed colleague (so HT to AS). The basic news is about a new MIT Press book, by Shigeru Miyagawa, Why Agree? Why Move? He examines agreement and movement data from Japanese, English and Kinkande.

Some things aren't quite clear from the release, like this:
The existence of similar structures in such otherwise disparate languages, Miyagawa asserts, provides strong evidence that all human languages have a common origin.
Common origin in human cognition, sure, but is the assertion one about monogenesis?

I guess we'll have to read this one ... .


John Cowan said...

Indeed, why bother? The largest language on the planet seems to get along fine without either one.

Coppe said...

He definitely means common origin in cognition. As someone who is familiar with Miyagawa's work, I am pretty sure that comment was just fluff.

The focus of the book is modern generative syntax.

Syntax person said...

Ad John Cowan:

Chinese does show agreement between numerals and nouns and between demonstratives and nouns. That's the famous classifier system.

And there certainly is syntactic movement in Chinese as well. Topicalization, bei-passives, verb-position alternations. Lots.

Mr. Verb said...

Yeah, I know a little about Miyagawa's work and didn't expect anything beyond syntax, but the 'fluff' made me nervous.

John Cowan said...

Disclaimer: I am no Sinologist.

Chinese classifiers strike me as only marginal examples of morphosyntactic agreement: they transparently descend from a purely semantic classifier system such as we find in Burmese. In that language, the use of the 'pair/team' classifier with 'ox' or 'buffalo' is acceptable, but with 'horse' it is not. This is not a mere subcategorizational fact about the word for 'horse', but instead reflects the fact that traditionally the Burmese did not use teams of horses. Likewise 'basket of eggs' and 'basket of chickens' are fine, but 'basket of mosquitoes' is not, again for reasons that are about the real world rather than the language.

In addition, I don't believe that bei-constructions are really passive in the usual sense. Chinese verbs don't really have voice, nor is the language either accusative or ergative, morphologically speaking: context controls all. Thus "George dropped the watermelon and [George] was embarrassed" is not a sign of syntactically accusativity, nor is "George dropped the watermelon and [the watermelon] burst" a sign of syntactically ergativity: these are just the most plausible interpretations, and so they are the ones that Chinese-speakers assume.

Claire said...

Not necessarily. I've been at talks at MIT where there have been comments along the lines of "it's great to see that syntactic research shows that Altaic is actually a family" (because Korean and Japanese share a certain property).

Mr. Verb said...

Wow. 'Family' as a typological rather than genetic entity? Or a failure to distinguish between the two?

Daan said...

Can I just ask the syntacticians here a perhaps silly question?

“If there were no movement in human language, you could not ask questions"

How about the final particle 嗎 ma in Chinese? Would adding that particle to a sentence in order to turn it into a question be considered movement?

Anonymous said...


I don't know what talks at MIT you've attended, but I've been to lots, and never heard anything as silly as you report.