I've been reading a lot lately about the potential use of fMRI scanning in court trials. For the most part, neuroscientists do not support use of the technology as it currently stands. This hasn't stopped numerous individuals from trying to incorporate fMRI scans into their supporting evidence. Defense attorneys in a 2009 San Diego child abuse case attempted to incorporate fMRI evidence, later denied by the judge. (http://www.wired.com/wiredscience/2009/03/noliemri/) The company that performed the testing, No Lie MRI, claims its data is 90% accurate, much more convincing than a standard polygraph test. No Lie and similar companies base their testing on the idea that the ventrolateral area of the prefrontal cortex is the tell-all for fibbing. Basically, heightened blood flow to this area during questioning indicates that the participant is a dirty, rotten liar.
In 2010, a neuroscience researcher named Kent Kiehl became involved in the trial of a serial murderer named Brian Dugan. (http://www.nature.com/news/2010/100317/full/464340a.html) Dugan had already served twenty years in prison, but was appealing to the court to do away with his death penalty charge. Kiehl used fMRI evidence to show that Dugan was suffering from severe psychopathy; he couldn’t possibly have made clear decisions about murderous acts with a mind so ‘handicapped.’ Though this evidence was allowed in court, the death penalty charge remained unchanged.
How should we feel about the use of this invasive technology? fMRI can visualize the interworking of our brains, completely out of our own control. With enough practice, you can fib on a lie detector test (people do it on Maury all the time). But can you lie to an fMRI scanner? Is this any different from more generalized rights to privacy? fMRI lie detection is a growing industry, with new companies spurting up annually. It doesn’t seem that the current technology can back up promises these companies are making, but what if it could in the future? Should this sort of technique, if based on sound evidence, be considered ethically sound? I think not. Our minds are the only possession that is truly our own. It is our choice who to let in, how to manipulate what emerges. If we take that right away (even from murderous, terrible people), what do we have left? One day, you could show up at a job interview only to be thrown into an fMRI and analyzed by a representative from No Lie MRI. This is a slippery slope, and I foresee some heavy controversy in the near future.
I agree with you, our minds are the only thing we can truly call our own.
ReplyDeleteI suspect that if fMRI or some other type of brain monitoring were to become fairly reliable for truth detection then there'd be a fair bit of research on how to fool it.
Besides, what is truth? In one episode of the TV show Babylon Five, Vir is told a lie by his boss, Lando. Later Vir is asked about what Lando told him. A telepath is present who can read Vir's mind and tell if he's lying or not. Vir says what Lando told him and, since Vir thought it was true, then the telepath reports Vir is telling the truth, even though it actually is not true.
There's the old lawyers' story about the defense attorney who asked one too many questions. He asked the witness: "Did you see my client bite off the victim's ear?" "No I didn't." "Then how do you know the defendant bit the ear off?" "Because I saw him spit it out." If the second question wasn't asked, then the defendant might have walked.
Besides any ethical arguments which could be made about effective "lie detectors," there's practical problems as well.
Upon what does a neurologist base the assumption that heightened blood-flow to the ventrolateral area of the prefrontal cortex indicates lying? Could there exist other reasons for this? Could different people's brains be organized differently? If it's true that it's 90% effective then that means that the blood-flow to this area increases without the subject lying 10% of the time, right? What accounts for this? Could stress or confusion or some other factor cause this blood-flow anomaly? Could an ambiguous or unsettling question also cause this? Could a form of mental illness result in false negatives?
ReplyDeleteIt's an interesting question to consider what might happen if the science someday backs up the claim that fMRI will work as a lie detector. But any skeptic knows that correlation does not prove causation. For that wouldn't you need more evidence independent from the fMRI?
Forcing a (theoretically) devoutly religious Christian to swear on a Bible, if you know they believe it binds them to tell the truth, is the same violation of privacy. The witness only has two practical options: tell the truth or plead the fifth.
ReplyDeleteWere an fMRI lie detector to reach sufficient levels of accuracy, a witness forced to submit to a scan would face the exact same options: tell the truth or plead the fifth.
The fMRI has another extant functional corollary: cross examination, a process designed to detect lies. Does cross examination violate privacy? I know personal questions are often asked if they are relevant, and yet we don't have a problem with that. Now, would it violate their privacy if the cross-examination only took 5 seconds instead of days and was 99% effective? I can't see how it would.
So, I don't see the fMRI scanner as representing an invasion of privacy. Lying under oath is already a crime. All an fMRI scan does is detect when you've broken the law, just like a regular investigation into perjury. It can't force you to speak, or extract information against your will. How is this a bad thing?
There's a huge potential upside: exonerating the innocent. And what if our politicians were forced to pass an fMRI test for corruption on a regular basis? I would gladly relinquish my ability to lie under oath if my elected leaders did too.
This seems like quite a complex issue. Before reading the comments, I was prepared to ask the same questions as Steven: does it produce similar answers for stress (like the polygraph test) or other brain functions? Furthermore, 90% accuracy is technically not that high, as it still means it could be wrong one in ten times.
ReplyDeleteBut Robert also makes good points. In theory, its purpose should be to assist in verifying one's innocence just as much as verifying guilt. While it does seem rather invasive, I would be comfortable with the concept of technology that is capable of providing courts with an accurate recognition of guilt or innocence - but it still doesn't sound as accurate as it should be.
Food for thought, anyway.