Brain fingerprinting: Encyclopedia II - Brain fingerprinting - Criticisms

Brain fingerprinting - Criticisms

Some believe brain fingerprinting needs far more refinement before its use becomes widespread and court cases are decided on its evidence. A large proportion of the scientific community believes that this new test will have the same limitations as the older polygraph test. Several studies have shown it is possible to control electroencephalographic response in much the same way as it is possible to fool a polygraph test; however, a higher level of training and discipline is required.

Neither American or European courts admit "brain fingerprints" as evidence.

MERMER is inappropriate for use in crimes that do not rely upon information being present while the crime was committed, since both people have equal knowledge of the crime. Farwell also acknowledges that where there is no information that law enforcement officials possess that the criminal does not also possess, the test is not appropriate.

The MERMER system exploits the known Oddball-evoked P300, which is similar in function to the N400 and P600 that have been documented by Kutas and Hillyard of Sandiego, and Osterhout and Holcomb. These two ERPs are used to measure brain wave response to unexpected or “oddball” anomalous words or grammar formations, but the N400 and P600 waves are measured in the same locations and using very similar methodology to the MERMER system. In the comparative analysis, there is evidence that this controlled system of 99.9% accuracy in detecting “information present” have been refuted, rendering the MERMER system to be irrelevant. The P300 has been shown to be related to “information present” analysis, in addition to a multitude of unknown factors. Farwell’s documented analysis of 99.9% accuracy with no false positives is entirely based on a well-chosen set of participants and artfully designed questioning, not on the accuracy of the system in real-world scenario with unknown suspects.

Brain fingerprinting - Ethical considerations

Farwell ignores the ethical considerations of the ability to interpret someone’s mind to test if they possess specific knowledge. The designers and supporters of brain fingerprinting do not seem concerned with the unethical uses of this technology, and have made it clear that this test is designed for, and created with the aid of, government agencies. Some feel that a discussion of the humanitarian use in reducing stress for innocent people that are accused is unethical when the ethics of the technology itself is not discussed.

Brain fingerprinting - Conflicting evidence

Johnson and Rosenfeld of Northwestern University have published two papers that refute the MERMER system’s accuracy. The first test was a recreation of P300 guilty-knowledge testing similar to a simplified MERMER, however, it did not require a rehearsal of the Target and Probe stimulus. Subjects were required to complete a checklist in advance of their crimes in terms of “innocent/not innocent,” which implies that this test has not yet been applied by a tester that does not possess foreknowledge of the subject’s understanding or perception of his own crime. They were able to correctly classify 89% of the subjects.

In another paper by Johnson and Rosenfeld, they developed specific countermeasures to generate a false positive to the p300 test as proposed by Donchin and Farwell in their collaborative paper that initially proposed the connection between the P300 and lie detection. Johnson and Rosenfeld beat Farwell and Donchin’s proposed lie detector test by imagining an emotional event at the same time as the target and probe stimulus was presented. The subjects trained in countermeasures were able to defeat the P300 lie detector test’s attempts to classify them by 90% after 5 practice trials. They were completely unable to produce false positives.

This would imply that the MESMER test, which is an expansion of the P300 Oddball-Evoked Guilty Knowledge Test, is also under attack, as there is no chance of a 100% correct guilty or “information present” classification. These two papers imply that a MESMER test cannot correctly identify a subject with knowledge of a crime, and definitely not within 99.9%. The Farwell MERMER system is in serious jeopardy in light of the recent publications by Johnson and Rosenfeld, and therefore, cannot be given any credence until further tests in favour of the MERMER system are administered using the techniques carried out by Rosenfeld and Johnson.

Currently, there has been only one field study published that has used a P300 Guilty Knowledge Test to verify its accuracy, but this group reported chance accuracy, definitely inconsistent with Farwell’s reported 100% accuracy. It has also been shown in Rosenfeld and Johnson’s countermeasures paper that while defeating the MERMER system is possible, it will necessarily require practice and an educated guess about the nature of the probes that will be presented. It is noted that common affective murderers and similar criminals may not have access to psychophysiology experts, it is conceivable that foreign terrorists will have such access and ability, rendering the MERMER system useful only against the uninitiated. After receiving countermeasure training, Johnson and Rosenfeld reported a change from 82% correctly classified “guilty” subjects to 18% after countermeasure training.

Adapted from the Wikipedia article "Criticisms", under the G.N U Free Docmentation License. Please also see http://en.wikipedia.org/wiki