A Truth Machine

Can brain-scanning technologies stop terrorists -- or just threaten privacy?

"It's happening much faster than I thought it would," says James Halperin, author of the 1996 science fiction novel The Truth Machine. The novel describes how humanity would react to the invention of an infallible lie detector in the year 2024. "When I was talking about the concept of a truth machine back in the 1990s, a neuroscientist friend told me that his best guess was that it would be 50 years, if ever, before such a thing could be created," says Halperin. "I picked 2024 as the date so that the idea wouldn't seem too ridiculous."

However, recent advances in brain-scanning techniques may bring about the development of a kind of truth machine sooner rather than later. These techniques, if validated through more research, could replace fallible polygraph tests, which some argue are no more accurate than chance in determining guilt or innocence. Self-described "polygraph victims" George Maschke and Gino Scalabrini explain how federal agencies ruin lives by relying on polygraphy to screen applicants for sensitive jobs despite the fact that polygraph tests are so bad that U.S. courts refuse to allow their results to be admitted as evidence in trials.

But the science of detecting deception might be coming of age. Dr. Lawrence Farwell, a psychiatrist who heads the Human Brain Research Laboratory in Fairfield, Iowa, has developed a technique he calls "brain fingerprinting."

"It's not actually a lie detector," explains Farwell. "Instead it detects whether or not certain information is stored in a person's brain." He likens brain fingerprinting to finding fingerprints or DNA traces at a crime scene. The presence of a person's fingerprints or DNA at a crime scene does not tell investigators whether the person is guilty or not. After all, there may well be an innocent explanation for how they got there. Similarly, brain fingerprinting does not tell investigators whether a suspect is guilty or not, just that specific information is or is not present in his or her brain.

But brain fingerprinting has one big advantage over DNA and fingerprinting, notes Farwell. "Criminals leave fingerprints or DNA at crime scenes only about one percent of the time," Farwell says. "But their brains are always there, planning, executing, and recording the crime."

Brain fingerprinting works by flashing words or pictures relevant to a crime on a computer screen along with other, irrelevant, words or pictures. When a person recognizes information, specific electrical brain impulses that Farwell calls memory and encoding related multifaceted electroencephalographic responses (MERMERS) are involuntarily elicited. MERMERS are measured using a headband equipped with sensors. When details of a crime that only the perpetrator and investigators would know are presented, the perpetrator's brain emits a MERMER, but the brain of an innocent person does not.

Farwell claims that his MERMER device has been 100 percent accurate in tests. For example, in one test he was able to correctly identify 17 FBI agents out of 21 people tested.

Brain fingerprinting could also be used not only to detect the absence or presence of criminal knowledge, according to Farwell, but also to detect possible terrorists. For example, it could be used to screen travelers from Afghanistan to see if they have or do not have specific information about how Osama bin Laden trains recruits at his terrorist camps. Asked if he has been talking with U.S. law enforcement agencies since the World Trade Center atrocities, Farwell says only, "There have been some discussions. I'll leave it at that."

However, Farwell's work is by no means universally accepted by academic researchers. J. Peter Rosenfeld, a professor of psychology at Northwestern University, believes that Farwell claims far too much for his MERMER technique.

"I simply don't believe that he is getting 100 percent correct results on his tests," Rosenfeld says. Rosenfeld himself gets about an 80 percent correct identification rate from similar brainwave studies. Rosenfeld is also pursuing another line of research aimed at identifying the physical correlates in the brain that correspond to a person's awareness that he is lying. This research goes beyond identifying whether or not a person has specific knowledge. Asked if he thought that current brain research on detecting deception would lead to ways to determine accurately whether a person is lying or not, Rosenfeld replied: "Absolutely. Within 10 years, maybe even five."

Besides brainwave research, Dr. Daniel Langleben and Dr. Ruben Gur from the University of Pennsylvania reported this week the results of their study, "Functional Magnetic Resonance Imaging (fMRI) of the Brain During Deception," at the annual conference of the Society for Neuroscience. The study used fMRI as a kind of lie detector. Eighteen volunteers were asked to lie about which playing card they held while inside an MRI machine that measured blood flow to various regions of their brains. Using fMRI, the researchers identified the anterior cingulated cortex (located three or so inches behind the forehead) and the left premotor cortex (located near the ear) as regions activated when the volunteers were lying.

Langleben and Gur suggest that these brain regions, which are involved in attention and judgment, are activated to suppress or inhibit the "truth." Gur thinks that the fMRI data "are detecting the act of suppression, the act of lying," not just the presence or absence of particular knowledge. It is important to keep in mind that the fMRI results are preliminary group data that need to be replicated and cannot be applied to determining whether or not any individual is being deceptive at this time.

However, Gur sees the possibility of using fMRI studies to validate and improve brainwave detection techniques by identifying the physical locations in the brain that generate the brainwaves being studied by researchers like Farwell and Rosenfeld.

Unlike trying to determine the presence or absence of specific information in the brain, trying to detect the act of lying brings with it the complication that the person must be aware of the fact that they are lying. Take a problematic example--some people might delude themselves into thinking that they are telling the truth, e.g., eyewitnesses to a crime who are mistaken.

The Fifth Amendment guarantees that no person "shall be compelled in any criminal case to be a witness against himself." Does brain fingerprinting raise Fifth Amendment concerns about self-incrimination? It seems wrong to treat finding guilty knowledge in one's brain the same as finding fingerprints or DNA traces at crime scenes, as Farwell urges us to do.

For his part, science fiction writer Halperin notes an interesting convergence in current fMRI and brainwave research since his fictional "Cerebral Image Processor" measured a combination of electrical activity and blood flow. In The Truth Machine, Halperin illustrates the benefits and problems that the pervasive availability of an infallible lie detector would cause society. It is easy to see some of the benefits -- detecting would-be terrorists, finding politicians who tell the truth during campaigns, detecting honesty in meeting contractual obligations. But what about those areas of life we would like to keep private, say, one's sexual orientation, or unusual religious beliefs, or drug habits, or taste in pornography? Halperin suggests that right now, many of us tolerate laws and regulations on many of these private activities because we know that we are not likely to be caught when we violate them. In a world where the truth can be known absolutely, Halperin thinks laws regulating many private activities would be repealed and there would be areas of life in which the use of a truth machine itself would be banned.

Let's hope he's right.