Monday, November 17, 2008
Jeff Binder visits Talking Brains West
I recently had the pleasure of hosting Jeff Binder during his visit to our Center for Cognitive Neuroscience here at UC Irvine. Jeff, of course, was among the first to use fMRI to study the auditory system and has published several important papers in the field. I had met Jeff before at conferences, and had many previous email correspondences, but never had the chance previously to hang out and chat, so it was a fun visit.
His talk was on the neural basis of semantics -- not action semantics, or the semantics of fruits and vegetables, just semantics, broadly construed. Not to steal his thunder when the work eventually gets published but... He presented a meta-analysis of a boatload of imaging studies on "semantic processing." Lots of different kinds of stimuli and tasks were included in the meta-analysis, so the findings are necessarily going to be relevant to a very broad definition of semantics. I can imagine someone critiquing the approach based on this loose definition of semantics, and in one sense I wouldn't disagree: the analysis isn't going to tell you anything about the details of semantic representation or processing. On the other hand, I personally found it very useful as a guide to distribution of brain regions involved in semantic processing. And guess what? It wasn't just motor cortex, or the anterior temporal lobe, or the posterior temporal lobe, or the angular gyrus that was involved. It was a fairly extensive distribution of regions that included many of these "semantic" areas and more. (We'll have to wait for the paper to get the details -- my memory isn't that good.) One thing I found particularly interesting was that the distribution of these semantic areas was virtually identical to the distribution of the "default network" -- the set of brain areas that seem to show increased activity during "rest" periods in functional imaging studies (i.e., when subjects are thinking about any number of things they're not supposed to be thinking about). This, of course, has important implications for how we design our experiments because the resting baseline is really more like a contrasting our task of interest with a semantic task.
Thanks to Jeff for a fun and informative visit!