We granted certiorari in this statutory rape case to consider whether the results of lie detector tests of the defendant and the prosecutrix were properly admitted at trial when it was stipulated between the state and the defendant before the tests that the results would be admissible.
Chambers was convicted of the statutory rape of a young girl and sentenced to 20 years. Polygraph tests were administered to him and to the prosecutrix pursuant to an agreement between the state and the defendant that the results would be admissible whatever they showed. The results were admitted, and the examiner testified that he interpreted the results to mean that Chambers was lying in denying the crime, and the victim was truthful in accusing him. The Court of Appeals reversed the conviction on the ground that under Famber v. State, 134 Ga. App. 112 (213 SE2d 525) (1975)
, polygraph results are inadmissible and without probative value, and thus the victim's charge failed for lack of corroboration. The Court of Appeals correctly noted that doubt has been cast on the continued viability of Famber, see Scott v. State, 238 Ga. 30 (230 SE2d 857) (1976)
(and dissenting opinion), but it has not heretofore been overruled.
We rule today that upon an express stipulation of the parties that they shall be admissible, the results of a lie detector test shall be admissible as evidence for the jury to attach to them whatever probative value they may find them to have. Famber v. State, supra, is hereby overruled.
The Court of Appeals a few years ago stated the law correctly in Cagle v. State, 132 Ga. App. 227
, 229 (207 SE2d 703
) (1974): "Neither an agreement to take a polygraph, nor the taking of one, constitutes a waiver of a right to object to the admission of its results into evidence, absent an express stipulation of the parties as to its admissibility." (Emphasis supplied.) See also Scott v. State, supra (dissenting opinion).
Numerous other jurisdictions have now arrived at the conclusion that these results may be admitted by agreement. E.g., Herman v. Eagle Star Ins. Co., 396 F2d 427 (9th Cir. 1968), affirming 283 FSupp. 33 (CD Cal. 1966); Commonwealth v. A Juvenile, 365 Mass. 421 (313 NE2d 120) (1974); State v. Woo, 84 Wash. 2d 472 (527 P2d 271) (1974) (rule recognized); State v. Alderete, 86 N.M. 176 (521 P2d 138) (1974); State v. McDavitt, 62 N. J. 36 (297 A2d 849) (1972); State v. Williams, 108 Ariz. 382 (499 P2d 97) (1972); State v. Rowley, 15 Utah2d 4 (386 P2d 126); People v. Houser, 85 Cal.App.2d 686 (193 P2d 937) (1948); State v. Lowry, 163 Kan. 622 (185 P2d 147) (1947) (rule recognized). See Annot., Admissibility of Lie Detector Test Taken upon Stipulation that the Result will be Admissible in Evidence, 53 ALR3d 1005 (1973).
We acknowledge that doubt exists as to the complete reliability of lie detector tests, and we share at least a modicum of that doubt. Operators may be unskilled, and results may be ambiguous and subject to arbitrary characterization. Of course, by cross examination counsel may show any vagueness of the electronic indications or any subjectiveness of the examiner's interpretations, as well as exploring conditions other than the subject's untruthfulness which could have produced such responses.
However, despite some problems posed by polygraphs, McCormick warned more than 20 years ago that "We cannot in our hearts be so confident of the reliability of the present system of resolving conflicts in testimony by impeachment, cross examination and inferences from demeanor, that we can afford to reject scientific aid in the task." McCormick, Evidence 369-370, 174 (1954 Ed.). A more recent writer has phrased this thought in more urgent language: "If the judicial system is to fulfill its duty of searching for truth and maintaining integrity, it must commence a war against perjury. The war cannot be won with weapons restricted to cross examination, inferences from demeanor, and other relics from the crossbow era of Henry II " '. . . [T]here is no tenable reason why qualified polygraphers should not be welcomed by courts confronting credibility questions; clearly, 'polygraphy appears to have something valuable to add to the administration of justice.' " Tarlow, Admissibility of Polygraph Evidence in 1975, 26 Hastings L. J. 917, 920-921 (1975). (Footnotes omitted.)
Once it has been decided to admit polygraph evidence, it becomes necessary to decide what kind of evidence it is. As was true at Chambers' trial, such evidence usually has two parts. One part is the actual graphs made by the machine showing the subject's physical responses to questions. The other part is the opinion of the examiner as to what those responses indicate.
Authorities are in some disagreement in characterizing this evidence. In Schmerber v. California, 384 U. S. 757, 764 (86 SC 1826, 16 LE2d 908) (1966), the United States Supreme Court found it difficult to say whether it was more like testimonial evidence (that is, "testimony" by the one subjected to the test), or physical evidence. A Massachusetts court in Commonwealth v. A Juvenile, 365 Mass. 421, supra, said it was testimonial evidence and that "the trial judge, . . . may . . . admit the results, not as binding or conclusive evidence, but to be considered with all other evidence as to innocence or guilt." P. 426. It is sometimes said, for example, that "Polygraph evidence is introduced as expert opinion testimony." Note, Pinocchio's New Nose, 48 N. Y. U. L. Rev. 339, 361 (1973). McCormick treats it as scientific evidence. McCormick, Evidence 421, 207 (1972 Ed.). Accord, Tarlow, Polygraph, 26 Hastings L.J. 917 (1975). One federal court concluded that it was generally opinion evidence, but could sometimes be direct evidence on the point of the examinee's belief in his answer: "Since this is a perjury case, the issue is -- was the defendant lying? The opinion of the polygraph examiner based on a properly conducted examination is more than character evidence, it is direct evidence on this point and may be offered by either side regardless of whether the accused takes the stand or puts his character in issue." United States v. Ridling, 350 FSupp. 90, 98 (E.D. Mich. 1972). (Emphasis supplied.) "Whether the polygraph test results are admitted in a criminal or civil case, the judge should instruct the jury that they should consider the test results along with the other evidence, but that the opinion of the examiner as to whether the subject gave truthful or deceptive responses is not conclusive and should be only given whatever weight they think it deserves." Note, The Role of the Polygraph in Our Judicial System, 20 S. C. L. Rev. 804, 824 (1968).
In a context different from the instant case, a Washington court has admitted polygraph evidence for corroboration: "While all courts are not in complete agreement, we are persuaded that the better rule is that the results of a polygraph test are admissible for the purpose of corroboration under [certain circumstances] . . ." State v. Ross, 7 Wash. App. 62 (497 P2d 1343, 1347) (1972).
It is a question of corroboration which is presently before us, and it has long been the law in Georgia that only slight circumstances can suffice to corroborate the accusation of a rape victim.
"In a prosecution for rape, corroborating evidence need not be of itself sufficient to convict the accused. Slight circumstances may be sufficient corroboration. [Cits] . . . 'Therefore, if there be any corroborative evidence at all, this court will not pass upon its probative value, since the verdict of the jury under proper instructions has resolved that issue.' [Cits.]." Johnson v. State, 239 Ga. 116
, 117 (236 SE2d 65
It is not necessary in this opinion to answer all questions which future cases may present. It suffices to decide this case that the polygraph results were at least some evidence, though not direct evidence, of Chambers' guilt. The Court of Appeals erred in ruling that polygraph results are always inadmissible and without probative value. Accord, Cullin v. State (Wyo.), 565 P2d 445, 449 (1977); State v. Sims, 21 Crim. L. Rptr. 2190 (April 25, 1977) (Cuyahoga, Ohio, Common Pleas Court). The polygraph results were adequate to provide the corroboration of the victim's testimony required by Code Ann. 26-2018.
We note that in a different case in which the only evidence of guilt is defendant's performance on a lie detector test, the state's case would not meet the stringent test which is applied by Code Ann. 38-109 to circumstantial evidence: "To warrant a conviction on circumstantial evidence, the proved facts shall not only be consistent with the hypothesis of guilt, but shall exclude every other reasonable hypothesis save that of the guilt of the accused." Today's decision to allow these results into evidence is limited by our concurrent conclusion that the results of the lie detector tests are insufficiently reliable to exclude every other reasonable hypothesis in conflict with their conclusion. This means that when the only evidence of guilt is defendant's performance on a lie detector test, a conviction may not stand.
When polygraph results are admitted at trial, either party is entitled, upon request, to have the jury charged concerning the meaning of this evidence. In giving the charge the judge should state that the examiner's opinions may only be used to indicate whether at the time of the polygraph examination the person examined believed that he was telling the whole truth; that the jury are not bound by the polygraph examiner's conclusions and his (or her) testimony is not controlling on the issues and may even be entirely disregarded; and that it is for the jury to decide what weight should be given this evidence.
In reversing on the ground that, absent corroboration, the evidence was inadequate to convict, the Court of Appeals did not consider the other enumerations of error. Accordingly, we must remand the case to the Court of Appeals for their consideration of the remaining issues.
JORDAN, Justice, dissenting.
The law has been settled for years in Georgia that results of polygraph tests are inadmissible and without probative value. This is based on the lack of scientific proof of the accuracy and reliability of such tests. Now, without any scientific authority or other rational basis, this court would open the door for the admissibility of such tests. While some jurisdictions have arrived at the same result, the rule in most jurisdictions comports with the Georgia view. See Stack v. State, 234 Ga. 19
,21 (214 SB2d 514).
There is simply no "lie detector," machine or human. The first recorded lie detector test was in ancient India where a suspect was required to enter a darkened room and touch the tail of a donkey. If the donkey brayed when his tail was touched the suspect was declared guilty, otherwise he was released. Modern science has substituted a metal electronic box for the donkey but the results remain just as haphazard and inconclusive.
The state should not be able to convict nor the defendant to gain acquittal on the basis of such tests, even though both are willing to stipulate its admissibility.
I respectfully dissent.
BOWLES, Justice, dissenting.
I cannot agree with the majority opinion in this case. An express stipulation of the parties that the results of a polygraph test shall be admissible in evidence does not m my opinion make the results competent evidence. The majority opinion suggests that these scientific devices may have improved in recent years sufficiently to now aid in the search for truth. Certainly there may have been improvements both in the machines themselves and in the abilities of various examiners. I note, however, that the decisions cited in the majority opinion deal with the result being admitted by agreement only and there is practically no authority, statutory or otherwise permitting such evidence to be introduced by a party as a matter of right. My personal experience from many long years at the bar leads me to appreciate many, many faults with polygraph results and procedures, and little or no tangible benefits. I cannot agree that evidence of any kind has probative value merely because counsel for both parties stipulate as to its admissibility. The courts of many states agree. State v. Corbin, 285 S2d 234 (La. 1973); Lewis v. State, 500 SW2d 167 (Tex. 1973); Romero v. State, 493 SW2d 206 (Tex. 1973); Pulakis v. State, 476 P2d 474 (Alaska 1970); Conley v. Commonwealth, 382 SW2d 865 (Ky. 1964); People v. Zazzetta, 27 Ill. 2d 302 (189 NE2d 260) (1963); Stone v. Earp, 331 Mich. 606 (50 NW2d 172) (1951); Le Fevre v. State, 242 Wis. 416 (8 NW2d 288) (1943). Courts are usually reluctant to exclude any matter consented to by both parties. This is based on the basic principles of fair play. However, polygraph results have not been demonstrated as reliable. I cannot feel comfortable leaving the guilt or innocence of an individual based on "evidence" which almost every court in America has concluded is lacking as competent evidence because of insufficient scientific reliability. Defendants do not know or understand the trustworthiness. Jurors are not experts and usually have had little or no experience along these lines. On the other hand, juries could be greatly persuaded in one direction or the other by such unreliable scientific evidence. To approve its use, in my opinion, is a step backward in the judicial process.
Therefore, I respectfully dissent.