http://www.ccnl.emory.edu/greg/Berns%20Conformity%20final%20printed.pdf
Asch’s paradigm is based on social conformity: when asked to compare line lengths, a surprisingly high percentage of subjects agree with the blatantly erroneous consensus of their peers. Berns et. al. used fMRI to analyze a related decision making process. The task involved looking at a pair of 3D shapes, and judging whether they are rotated but identical - “Same“, or rotated and mirror imaged (not super imposable) - “Different“. Each test subject had four “peers”, who were conspiring with the scientists. The subjects had a 14% average baseline error rate, in the absence of external information from their peers. The error rate increased to 41% when the four peers provided incorrect judgments. In other trials the subject thought the four other answers were generated by a computer program. Subjects agreed with a computer generated, incorrect consensus 32% of the time.
Berns et. al. then compared the fMRI scans between subjects while they were conforming to incorrect peer judgments, versus when the subjects conformed to incorrect computer generated results. The goal was to notice effects of social conformation, isolated from the effects of confusion due to erroneous external data. Interestingly, social conformation correlated with greater activity in parts of the visual cortex, which was not present when subjects were conforming to computers.
Asch’s initial experiment was simple, and there could be little doubt that the subjects were aware of the correct answer, but deliberately replied otherwise. In this study however, the task is more complex, and the different activation of visual cortex neurons might indicate altered perception. Ultimately, maybe having other people tell you what to see might just make you see it. If information cascades can alter not just social behaviour, but also perceptions at a biological level, the implications would be phenomenal. Elections, religions, and even Hush Puppies might have to be seen in a whole new light. (It’s two AM… please excuse the puns.)
Of course, there are alternative explanations. Perhaps the altered neural activity is just from giving the image more consideration, maybe from paying more attention to it or trying to see it from another’s perspective. There is always the chance that a subject will, intentionally or not, lie when questioned about their decision making process. Luckily fMRI can come to the rescue, yet again.
http://technology.newscientist.com/article/dn13415 Researchers at Berkley have developed an algorithm that can decode what image a person is looking at. It is still relatively rudimentary, and functions well only on given sets of a hundred or so pictures. Coincidentally, this would be more than sufficient for an Asch-type experiment. Berns study involved spatial rotations, to create the possibility of credible errors in perception. A future study could limit itself to using only 2D images, whose details the visual decoding algorithm can differentiate between. If it turns out that we really, physically see what we’re told is there… Well, your imagination is as good as mine.