346 Mass. 266

May 6, 1963 - June 24, 1963

Middlesex County


The scientific reliability of polygraph or "lie-detector" tests has not been sufficiently established to justify use of the results of such tests as evidence; and in a criminal case there was no error in excluding evidence offered by the defendant as to such tests administered to him and to witnesses.

INDICTMENTS found and returned September 13, 1961.

Following the decision of this court reported in 345 Mass. 85, there was a second trial of the cases before Tomasello, J., without jury.

F. Lee Bailey for the defendant.

John J. Irwin, Assistant District Attorney (Ruth I. Abrams, Assistant District Attorney, with him), for the Commonwealth.

SPIEGEL, J. These are appeals under G. L. c. 278, Section 33B, from convictions of assault and battery and assault and battery by means of a dangerous weapon. The trial was held before a judge of the Superior Court, the defendant having waived his right to a trial by jury.

There are three assignments of error, [Note 1] each of which relates solely to polygraph tests. We condense the pertinent evidence referred to by the defendant in his brief.

One Joseph F. McAdams was assaulted in his apartment by four men on August 23, 1961, and again on September 4,

Page 267

1961. He later identified the defendant as one of his assailants. On the day of his arrest the defendant requested of the "investigating officers" that he be given a "`lie-detector' test so that he might demonstrate his innocence." No test was given until after the first trial. [Note 2] In January of 1962, while the defendant was in prison, he was given a polygraph test.

Assignments of error numbered 1 and 2 are based on the denial on December 7, 1962, of the defendant's pre-trial motions by a judge of the Superior Court other than the judge who presided at the trial of the cases, which began on January 16, 1963. In view of our subsequent discussion on the issue of the admissibility in evidence of the results of polygraphs tests we see no point in treating with the denial of the defendant's pre-trial motions. It is sufficient to state that there was no error in the denial.

The defendant made an extended offer of proof the substance of which recited the qualifications of several alleged experts, and included the "history, reliability, and use" of the polygraph, a general dissertation on the effectiveness of "lie-detector" tests, and the results of polygraph tests administered to two witnesses and to the defendant.

There is hardly a device which has caused greater controversy among "experts," lawyers, physicians, psychologists, government officials, and the public in general than the polygraph or "lie-detector" as it is colloquially characterized. The question of the admissibility of the results of a "lie-detector" test is one of first impression for this court. Such tests have been described and their judicial history chronicled by the courts of many other States. An excellent opinion for such purposes is State v. Valdez, 91 Ariz. 274. See 23 A. L. R. 2d 1308; A. L. R. 2d, Supp. Serv. (1960), 1998-1999, and subsequent Supplements. It was stated, with accuracy, in State v. Arnwine, 67 N. J. Super. 483, 495, that "there is not a single reported decision where an appellate court has permitted the introduction of the

Page 268

results of a polygraph or lie-detector test as evidence in the absence of a sanctioning agreement or stipulation between the parties." Over the years, appellate courts have repeated the reason given in the 1923 landmark case of Frye v. United States, 293 Fed. 1013 (Ct. App. D. C.). In rejecting a predecessor of the "lie-detector" for evidentiary purposes that court said: "We think the systolic blood pressure deception test has not yet gained such standing and scientific recognition among physiological and psychological authorities as would justify the courts in admitting expert testimony deduced from the discovery, development, and experiments thus far made." Ibid. 1014. The controversy in scientific and legal circles swirling around the polygraph test continues without abatement. In addition to the existing plethora of writings on the subject, well documented articles continue to appear which raise grave doubts as to the scientific reliability of the polygraph tests. See Highleyman, The Deceptive Certainty of the "Lie Detector," 10 Hastings L. J. 47 (1958); Skolnick, Scientific Theory and Scientific Evidence: An Analysis of Lie-Detection, 70 Yale L. J. 694 (1961). A recent and particularly devastating article on the untrustworthiness of the test is titled "Don't Trust the Lie Detector," Harv. Bus. Rev. (1962), 127, by Sternbach, Gustafson and Colier, two of whom are scientists. [Note 3] Numerous other authorities taking an unfavorable view of the "lie-detector" are referred to in the cases cited above. The questions and doubts raised by these sources go, not only to the techniques employed in the administration of the tests, but to the very premises and assumptions upon which they rest. "Emotional unresponsiveness," "[a]bility to `beat' the machine," "[p]hysiological abnormalities," "[m]ental abnormalities," "[n]ervousness or extreme emotional tension," "[t]he misleading nature of available statistics," and the "conflict and disagreement among the

Page 269

examiners and authorities" [Note 4] are some of the variables which vitiate the alleged effectiveness of the tests and thus militate against their admissibility in evidence.

The defendant, in his offer of proof, would have a polygraph examiner testify that "he is now able to reach a definite conclusion in better than 95% of cases tested." The accuracy of this statistic is very much in dispute. [Note 5] It is precisely this sort of controversy which creates the danger that on the introduction of such evidence a trial could descend into a battle of experts on the probative value of the polygraph test rather than a determination of the guilt or innocence of a defendant. The end result, in all likelihood, would be confusion instead of enlightenment.

Judicial acceptance of a scientific theory or instrument can occur only when it follows a general acceptance by the community of scientists involved. When supported by substantial authority establishing scientific reliability, this court has not hesitated to accept the benefits of science. Blood tests excluding paternity were held admissible in Commonwealth v. Stappen, 336 Mass. 174, and held conclusive as to nonpaternity in Commonwealth v. D'Avella, 339 Mass. 642. Both opinions contain substantial documentation attesting to the acceptance of blood tests as a scientific

Page 270

fact. Scientific evidence as to the amount of alcohol in an individual's blood is a further example. Commonwealth v. Capalbo, 308 Mass. 376, 380. Evidence of fingerprints and evidence of ballistics are additional illustrations. In each instance judicial acceptance followed general scientific recognition, a recognition which, patently, has not yet been accorded the polygraph test. We do not hold that such recognition must be universal or that the test must be proven infallible, but rather that the substantial doubts which presently revolve about the polygraph test must be removed. The trial judge properly excluded all evidence concerning the "lie-detector."

Judgments affirmed.


[Note 1] The defendant's assignments of error are as follows: "1. The Court erred in denying defendant's motion that complainant be subjected to a polygraph test. 2. The Court erred in denying defendant's motion for impartial polygraph test. 3. The Court erred in excluding the results of polygraph tests upon the defendant, Frank Esposito, and Nathan Weiner as set forth in defendant's written offer of proof, B for identification."

[Note 2] The defendant was originally tried before a jury, under the same indictments, in October, 1961. See Commonwealth v. Fatalo, 345 Mass. 85.

[Note 3] Richard A. Sternbach, Assistant Psychologist at Massachusetts General Hospital, and Lawrence A. Gustafson, Research Associate in Psychology at Massachusetts Mental Health Center, are both in the Department of Psychiatry, Harvard Medical School. Ronald L. Colier is a graduate student of industrial management at Massachusetts Institute of Technology.

[Note 4] See Highleyman, The Deceptive Certainty of the "Lie Detector," 10 Hastings L. J. 47-64 (1958), and Skolnick, Scientific Theory and Scientific Evidence: An Analysis of Lie-Detection, 70 Yale L. J. 694-728 (1961).

[Note 5] Regarding the validity of such statistics we quote in part from Harv. Bus. Rev., November-December, 1962. "Some commercial operators, in approaching prospective buyers of their services, claim up to 95% accuracy and less than 1% error, with remaining cases being called `inconclusive.' In their book, Lie Detection and Criminal Interrogation, Fred E. Inbau and John E. Reid offer some figures purporting to justify such claims, but the figures are extremely misleading. For example, they state that out of 4,280 cases run over a five-year period, 64.5% were reported innocent of the crime, 31.1% were reported guilty, and 4.4% were indefinite (page 129). The reports were based solely on the polygraph records. Of the group of persons reported guilty, only 791 (59.3%) were `interrogated with the aim of obtaining a confession,' and of these only 61.4% did confess. In other words, only 36% of the individuals who were reported as guilty on the basis of their polygraph records were actually verified as such. And only 11.7% of the cases reported as innocent were later verified as such. Together there were 18.9% of total cases verified as correct. The percentage of verified errors was 0.07% (the figure of 0.0007% cited in the book is incorrect). These figures leave us with no information at all as to the correctness of `reported guilty' in about 40% of the cases, and with no such information at all in almost 90% of the `reported innocent' cases. This is a far cry from the claim of `95% accuracy'" (pp. 129-130).