It wasn't supposed to happen—not yet at least—but it did: This past June, a judge in the Indian state of Maharashtra convicted a woman of killing her ex-fiance, citing as proof an EEG scan showing "experiential knowledge" of the crime. Many people do think there's something to this, that an EEG or MRI scan of the noggin can depict lies and truth if read correctly, but in the US it's agreed that this is experimental science at best, and snake-oil sales at worst. The story tells of a woman who lived in the town of Pune, engaged to Man A. One day, she up and runs off to Delhi with Man B. She returns to Pune, meets Man A at a McDonald's, and later on, he dies. Of arsenic poisoning. When the woman was brought in accused of murdering Man A, she denied the allegation. When investigators hooked her up to an EEG and read aloud facts of the crime, however, software interpreting the electrical impulses in her brain told a different story. Says the NYT: "The relevant nooks of her brain where memories are thought to be stored buzzed when the crime was recounted." Unlike in previous cases, there was little or no corroborative evidence here, but the judge sentenced the woman to life in prison anyway, and went on to write a 9-page lovesong to this particular Brain Electrical Oscillations Signature test, even though it has yet to be "validated by any independent study and reported in a respected scientific journal." (Peer review, who needs it?) The US is leading this burgeoning field of study, but the only time it's used in court is when the accused pays to have a study performed as evidence of innocence. The New Yorker ran an amazing expose on this shady business a year ago, and it's still well worth the read. What happens in an Indian courtroom doesn't set precedent in the US, but this technology certainly isn't going to go away, so it's important either to rule it out as faux science, or tighten up the applied methodology quickly, so that we can all get on to the business of reading each others' minds in court. Course then we'd really start killing each other. [New York Times]
Well, the Guilty Knowledge Test (the one performed here) is generally accepted to be by far the most dependable lie-detector test. So, I'd say let the woman fry. Or face life in prison, or get her ten lashings, or have to apologize profusely to his (other?) loved ones. Or whatever the punishment is in India. I don't really care.