Monday, March 16, 2009

Does speech perception rely on motor cortex? Of course not!

D’Ausilio and colleagues report a very nice new study showing that stimulation of human motor cortex (via TMS) directly affects the perception of speech sounds. TMS was applied to the lip or tongue areas of M1 while participants were asked to identify speech sounds that either involved prominent lip articulation, [b] and [p], or prominent tongue articulation, [d] or [t]. They found a double-dissociation: relative to a non-stimulation baseline, participants were faster to indicate that they heard a lip-related sound when TMS was applied to motor lip areas, and faster to indicate that they heard a tongue-related sound when TMS was applied to motor tongue areas. The authors conclude that “motor structures provide a specific functional contribution to the perception of speech sounds” and go on to propose “a modified ‘motor theory of speech perception’ according to which speech comprehension is grounded in motor circuits…”.

I strongly disagree with this conclusion and I have convinced the editors of Current Biology to let me tell you why. Since I don't have the cool flashy experiment to report, I only get my piece on their website, not in print, but I'll take it nonetheless. Fadiga and company are writing a response to my arguments. Have a look when it comes out and let me know what you think.

D'Ausilio, A., PulvermĂ¼ller, F., Salmas, P., Bufalari, I., Begliomini, C., & Fadiga, L. (2009). The Motor Somatotopy of Speech Perception Current Biology, 19 (5), 381-385 DOI: 10.1016/j.cub.2009.01.017

6 comments:

James Dean said...

Would you mind posting a quick summary of why you don't like their interpretation?

Greg Hickok said...

Sure. Because damage to the motor system does not cause speech perception deficits. See this post.

Brad Buchsbaum said...

to play devil's advocate ...


it could be that lesions that cause Broca's aphasia tend to be more inferior than the more dorsal premotor/motor regions stimulated in the TMS study.

In other words, it's possible(?) that this is a localization issue in that dorsal premotor lesions may not cause Broca's aphasia but may nevertheless cause impairments in speech perception (just as posterior temporal lesions can leave perception intact but nevertheless affect speech production).

Greg Hickok said...

Right, that is possible, which is why I tried to collect evidence from disruptions to a range of levels within the motor system including severe Broca's aphasia (large fronto-parietal lesions), Wada procedures (entire left hemisphere), motor cortex lesions affecting face/lip areas, bilateral lower motor and premotor cortex lesions, and bilateral Broca's area lesions. In addition, I pointed out that speech perception/comprehension is possible even before motor speech abilities develop (1-month-old infants), or where motor speech fails to develop (congenital or early-onset acquire anarthria) or cannot develop (chinchillas).

These sorts of arguments are exactly those that lead to the rejection of the motor theory of speech perception years ago. People seem to have forgotten all this work though.

tom said...

The effects of TMS on RTs look quite convincing (Fig 2), but the TMS effects on accuracy look pretty mild and inconsistent - TMS to Tongue M1 introduces an error bias of around ~ 5% in favour of tongue-phoneme detection, but there's no effect on accuracy when Lip M1 is stimulated (Fig 3).

Overall non-TMS accuracy rates in the pilot were about 75%, so this was quite a hard task (because the speech sounds were embedded in noise). Motor structures are not necessary for speech perception under normal conditions (Greg has argued this convincingly) but the RT data from this paper suggests to me that they maybe of some use sometimes in speech perception, at least under conditions of uncertainty - e.g. in noise (as in this paper) or perhaps during recovery of language function after stroke.

The Sheriff said...

Is there a url for Current Biology where your comment can be found?