A new study in J. Neuroscience failed to replicate a previous finding published in the same journal that linked gesture discrimination deficits to tissue loss in the inferior frontal gyrus, part of the supposed human mirror system. The new study, by Nelissen et al., examined correlations between gesture discrimination (and a range of other language and non-language tasks) and patterns of tissue degeneration in primary progressive aphasia (PPA). They found that gesture discrimination did not correlate with tissue loss in the inferior frontal gyrus, and instead correlated with tissue loss in posterior temporal-parietal regions including portions of the superior temporal gyrus, which is not part of the human "mirror system" (see below which is Fig 2A from Nelissen et al.).
This is from the same group that published the previous stroke-based study that reported an association between tissue damage in the inferior frontal gyrus (part of the "mirror system") and gesture discrimination deficits (see my critique of that study here, or in Hickok, 2009). The same stimuli were used in both studies, so why the difference?
I had pointed out that the stroke study was potentially biased because they used percent correct for their gesture discrimination-brain lesion correlation rather than d' (d-prime), which corrects for response bias. The previous study used a yes/no response paradigm: subjects saw a gesture and had to decide whether it was correctly executed or not. Using percent correct is a problem because some subjects (perhaps as a function of their lesion!) may be biased toward yes or no responses which can skew the results independently of how well they are actually discriminating the gestures. The present study used the same stimuli but with a modified design, 3-alternative forced choice: subjects viewed three gestures and then had to decide which of the three was correctly executed. A still frame of the gesture was left on the screen to minimize working memory effects. This paradigm reduces bias especially when the order and position of the correct item is counterbalanced as was done by Nelissen et al. It appears then, that reducing the bias shifted the brain region that showed a correlation with gesture discrimination performance.
There is much more to this study than what I've highlighted here, including their finding that measures of gesture and language processing are highly correlated. But that's a topic for another blog entry.
Hickok G (2009). Eight problems for the mirror neuron theory of action understanding in monkeys and humans. Journal of cognitive neuroscience, 21 (7), 1229-43 PMID: 19199415
Nelissen, N., Pazzaglia, M., Vandenbulcke, M., Sunaert, S., Fannes, K., Dupont, P., Aglioti, S., & Vandenberghe, R. (2010). Gesture Discrimination in Primary Progressive Aphasia: The Intersection between Gesture and Language Processing Pathways Journal of Neuroscience, 30 (18), 6334-6341 DOI: 10.1523/JNEUROSCI.0321-10.2010