by Greg Miller
Tuesday night's neuroethics social was billed as a clash between two heavy hitters in cognitive neuroscience over a provocative question posed by session organizer (and referee) Martha Farah of the University of Pennsylvania: Is brain imaging currently capable of something worthy of the term 'mind reading'?
In one corner was John-Dylan Haynes of the Bernstein Center for Computational Neuroscience in Berlin, Germany, weighing in with credentials such as his pioneering work using statistical tools called pattern classifiers to infer mental states from fMRI brain scans. In the other corner was noted neuroskeptic Russell Poldrack of UCLA, who has been an outspoken critic of what he sees as egregious overextrapolation of brain-imaging data in certain academic and commercial circles.
Anyone hoping to see a bloodbath went home disappointed. Haynes argued that current brain-imaging technology is capable of something like "proto-mind reading," noting that his group and others have used pattern classifiers to infer emotions, intentions
, and to identify a photograph
a person has just seen. Poldrack agreed that there have indeed been some very impressive findings, but he noted that they all have come from carefully designed lab experiments with fully cooperative volunteers. And that's a long way from what the average person thinks of as mind reading, he said: "Most people think of mind reading as reading the sentences out of your head as you think them." Poldrack said he doubted neuroimaging would ever accomplish that.
Even so, both contestants agreed that current methods are powerful enough that serious consideration of the ethical implications of their use--and potential misuse--is well warranted.
So who won? Poldrack seemed to get more applause, and Haynes said he was prepared to concede. But Farah called it a draw, citing Poldrack's unfair advantage in going second: "More people had finished their beers and had both hands free to clap."