Wednesday, April 08, 2009

The lego blocks of language: Hurrah for Talking Brains

It's been a while, I guess, since we had the lowliest of all post types, the 'hey, I found something cool on another blog' post, but somebody called my attention yesterday to this post called "Speech: Not enough distinctions are being made" over at Talking Brains, specifically aimed at treatments of spoken language processing. Those guys are good.

We've got a suddenly thriving group of sound people here on campus, and I have to call them 'sound people' because they include not only phonologists and phoneticians, but engineers and stats people and who knows who all. Some of them have been discussing the 'false parsimony' of linguistic theory lately, a kind of Occam meets slasher flick approach to understanding language. Talking Brains seems (seem? No.) to be on the same wavelength there.

Image from here, and aside from the nice graphic, it's worth visiting just to read 'legolicious' used in context.


Adam Ussishkin said...

As they say on facebook, "I like this."

√v said...

The synchronicity of this post and image is quite amazing. The fashionista was quizzing the boy just yesterday as to why he liked Lego figures so much even though they were much less accurate than regular action figures. The boy's response is reflected in the shadows of the legos in the picture. The legos somehow capture the core aspects of what he is trying to do because they allow creativity that the more accurate action figures do not... a 7-year-old gets the advantages of abstraction in working on stuff, gotta love it.

The guys over at Talking Brains are spot on about their 'not enough distinctions' shtick and they are thinking pretty close to the wetware. One of the most interesting things to observe about the TB guys (IMOHO) is that they are perfectly fine with abstract representations being necessary to account for the operations of the wetware. In other words, they know there is both the 'lego' and the 'shadow' and that we want an explanation that gets both plus the mapping between the two. How this view is lost in the 'middle' speech science/psychology camp and then regained in the abstract theory linguistics camp (sometimes by the good ones) continually baffles me... mabye Mr. Verb needs to send out lego sets to some people...