Kanji processing in the brain

These days fMRI is used to figure out what parts of your brain light up when doing everything from meditating to taking a crap, so why not take a peek at the neurophysiology of reading Chinese characters? That’s what a group of Taiwanese researchers did in this study (warning, long, boring PDF). Strapped into the big fMRI machine, the hapless subjects peered through a mirror at pairs of huge Chinese characters projected at their feet, attempting to determine if they were homophones, a task requiring the so-called orthography-to-phonology mapping, or identical shapes, a purely geometric task. This paper comes complete with those de rigeur pictures of brains with their activated red and orange areas (the one here showing people at work on the homophone task).

The only problem is, I can’t figure out what their conclusions were:

While the left occipitotemporal region, left dorsal processing stream, and right middle frontal gyrus constitute a network for orthogrpahic processing, the regions of the left premotor gyrus, left middle/inferior frontal gyrus, medial frontal cortex, and the left temporopariental region work in concert for phonological processing of Chinese…The engagement of sets of regions for different levels of Chinese orthographic and phonological processing is consistent with the notion of distributed parallel processing. Our knowledge of characters arises from concurrent interaction between orthographic, phonological, and semantic processing.

Well, OK.

As an aside, I’m fascinated by this excerpt from the paper:

Engagement of the left post-central gyrus, medial superior frontal gyrus (SMA, spatially extended to cingulate cortex), thalamus, and cerebellum was mostly due to subjects’ voluntary movement of right index and middle fingers in response to the tasks.

Now tell me, what were those folks doing wiggling their fingers in that big old fMRI machine?

Overall, this is an intruiging topic, but I am dismayed by the scientific level of this paper. Whatever happened to the good old scientific method stuff—having a hypothesis, making predictions, and designing experiments to validate them?

Leave a Reply