Who Are the Brain Police?

Polygraphs, contrary to TV cop shows (or Dr. Phil, even), are notoriously unreliable and are inadmissible in court. But what if there was a better lie trap out there? Perhaps something that literally reads your brain waves? It's on the way:

The technique he calls 'brain fingerprinting' is an electronic test of a specific kind of brain wave that he says can identify incriminating information despite an individual's attempt to conceal the knowledge.

* * *

As part of this research, Farwell ran across what would become the scientific basis of brain fingerprinting. It is a type of signal in the brain known as a P300 wave, so-called because it is an involuntary response to a recognized object or piece of information that happens within 300 milliseconds.

It's been a well-known and widely accepted phenomenon within neuroscience. What Farwell did is connect it with another related electrical brain response that he dubbed a MerMer (for 'memory and encoding related multifaceted electroencephalographic response') that he contends provides a foolproof method for testing an individual's knowledge -- or lack of knowledge -- of a criminal act.

'It's 100 percent reliable and has been ruled admissible in court,' Farwell said.

I doubt Farwell's confidence (is anything 100 percent reliable?), but the technology has already made a couple of court appearances (one leading to an acquittal in a murder case).

If it really does work that well, what could it mean for our traditional adversarial system of justice? The Fifth Amendment, of course, protects your right to remain silent and to not be compelled to testify against yourself. That's fairly rare in the realm of global justice. There are at least two justifications for it. First, it simply makes the state's job of putting you in prison more difficult, which is a good thing in and off itself. Second, people compelled to testify against themselves can't necessarily be trusted to be accurate and may falsely incriminate themselves.

Assuming most folks think the rule rests on the second justification (as most do, I think), what support would the rule have if there was a full proof means of determining if the defendant was telling the truth? Sure, you'd have some folks who really believed they were telling the truth, but weren't, but would that be an acceptable margin of error? Something to ponder.

UPDTATE: Via Volokh, apparently brain scanning is already at work in the courts in India.