OPINION
SINGLETON, Judge.These consolidated cases present a single question: whether the complaining witness to a crime, who prior to trial has been hypnotized to refresh her recollection, is incompetent to testify at trial to a subsequent eyewitness identification of her assailant. We hold that hypnotism prior to trial does not per se render such a witness incompetent. See A.R.E. 601.
I. PROCEEDINGS IN TRIAL COURT
A. CONTRERAS
Joseph Contreras was indicted for five felonies: kidnapping, former AS 11.41.-300(a)(1)(C); assault in the third degree, AS 11.41.220;1 and three counts of sexual assault in the first degree, former AS 11.41.-410(a)(1). The indictment charges Contreras with kidnapping Ms. S.J. and Mr. E.L. with intent to sexually assault them or place them in apprehension that they would be subject to serious physical injury or sexual assault. In addition, Contreras was indicted for assaulting S.J. with a dangerous weapon, and sexually assaulting both S.J. and E.L.
Prior to the arrest of Contreras, S.J. was hypnotized by an Anchorage police officer to help her identify her assailant. She subsequently identified Contreras. Contreras moved for a protective order suppressing evidence of the identification on the ground that the hypnotic session influenced S.J.’s identification of him. After hearing evidence from the parties, including expert testimony regarding the effect of hypnosis on memory, Judge Serdahely granted the requested protective order in part and ruled that S.J. could not testify at trial regarding any matter discussed by her with the hypnotist during the hypnotic session. This ruling effectively suppresses any identification by S.J. of Contreras. The state has petitioned for review, contending that Judge Serdahely erred in suppressing the testimony of S.J. We agree and reverse.
B. GRUMBLES
Ricky Glen Grumbles was charged in an indictment with burglary in the first degree, AS 11.46.300(a)(2)(A); attempted murder in the first degree, AS 11.41.-100(a)(1), former AS 11.31.100; and theft in the second degree, AS 11.46.130(a)(2). The indictment was based upon testimony that Mary Hall discovered Grumbles in the process of burglarizing her residence. Grumbles shot Hall in the thigh and escaped with her property. Hall was hypnotized by a police officer to aid the police in identifying her assailant. She later identified Grumbles. Grumbles moved to suppress her identification on the ground that it was improperly influenced by the hypnotic session. Judge Buekalew, at the request of the parties, reviewed a transcript of the expert testimony introduced in Contreras’ case and after hearing argument denied the protective order. Judge Buekalew specifically ruled that any influence the hypnotic session might have had on Hall’s identification was a matter affecting her credibility to be determined by the jury and not a matter of *795competency to be determined by the court. Grumbles petitions for review of the denial of the protective order. We grant the petition and affirm.
Given the difference of opinion between the trial courts in these two cases, the importance of the issue, and the substantial difference of opinion reflected in decided case law throughout the United States, we have granted the petitions for review to resolve the issue prior to completion of the trials in question. Alaska R.App.P. 402(b)(2).
II. THE RULES OF EVIDENCE
Alaska Rule of Evidence 4022 provides that relevant evidence is admissible unless a specific rule, statute or constitutional provision excludes it. The testimony of an alleged eyewitness is clearly relevant. See A.R.E. 401.3 Consequently, eyewitness testimony is admissible unless expressly excluded. Contreras and Grumbles argue for exclusion based in part on Evidence Rules 4034 and 6015, and in part on the state and federal constitutions.
These arguments may be summarized as follows: Witnesses who have been subjected to hypnosis prior to testifying, although able at the time they testify to understand their duty to tell the truth and express themselves intelligibly, are unusually susceptible to memory distortion through “suggestion”6 and “confabulation.”7 Suggestion may occur intentionally or unintentionally through unconscious cues which the hypnotist conveys verbally or nonverbally to the person hypnotized. Confabulation is a process whereby a person who is under substantial pressure to remember a perception, such as details of the appearance of an assailant, but in fact had no perception to remember, is encouraged to unconsciously manufacture those details from her other experiences or her imagination. “Suggested” details come from outside the subject and may supplant the subject’s actual perceptions, while “confabulated” details generally occur only in the absence of an actual *796perception capable of being remembered. Details that are suggested or confabulated become part of the witness’s memory. Neither the witness nor any expert psychologist or psychiatrist who subsequently interviews her will be able to distinguish those details from her own memory of her actual perceptions.
In addition, it is argued that a hypnotized subject will exhibit a demeanor manifesting a high degree of confidence in her “restructured” memory, including any misinformation she received through suggestion or confabulation. Therefore, the subject achieves substantial insulation from effective cross-examination. She may appear to the jury supremely confident that she accurately remembers, when in fact she merely imagines. In contrast, it is argued that a typical witness who is unsure of an identification will manifest that lack of assurance by appearing hesitant and uncertain. Some au~ thorities have considered this “tampering” with the witness’s demeanor to effectively render the witness permanently unavailable for cross-examination, and conclude that her testimony at trial is the equivalent of hearsay and constitutes a violation of the defendant’s right to confront the witnesses against him.8
Having reviewed the record, and having considered both the arguments of the parties and the authorities cited from legal and psychological literature addressing eyewitness testimony and hypnotism, we decline to adopt a per se rule rendering the complaining witnesses in these cases, S.J. and Hall, incompetent to testify at trial regarding the matters covered in their pretrial hypnotic sessions. See A.R.E. 601.9 We therefore affirm the decision in Grumbles’ case and reverse the decision in Contreras’ case on the issue of competency.10
*797III. STATEMENT OF FACTS
A. CONTRERAS
The following facts were developed before the grand jury and in the evidentiary hearing held on the motion to suppress.
S.J. and E.L. were college students working at a restaurant in Anchorage during their summer vacation in 1980. On the night of August 5,1980, S.J. and E.L. were talking in S.J.’s vehicle outside the restaurant when a man, later identified as Contreras, asked directions to a local cocktail lounge. Neither S.J. nor E.L. were previously acquainted with Contreras. He returned a few minutes later, pulled a knife, and forced his way into the car. As he drove, he held the knife at S.J.’s neck. Several times Contreras threatened to kill them both. During the ride, Contreras demanded that E.L. remove S.J.’s clothing and penetrate her with his finger to get her sexually aroused. E.L. complied with these demands. Contreras eventually turned off the road in the Potter Flats area south of Anchorage. He then demanded that E.L. take off his own clothes so that Contreras could watch E.L. and S.J. have intercourse. He ordered S.J. to perform fellatio upon E.L., and became angry when E.L. did not become sexually aroused. Contreras then required both S.J. and E.L. to perform fellatio upon himself, which they did. He finally ordered them to get out of the car, stating that he was going to kill them. S.J. and E.L. fled naked into the rainy night. Contreras chased them for more than an hour before abandoning his pursuit. Several hours later S.J. and E.L. obtained assist anee.
S.J. and E.L. were interviewed by the police. Each gave a sketchy description of their assailant. At this time, the police had no suspects. Approximately one week later, S.J. underwent hypnosis by Investigator Eugene Parmeter of the Anchorage Police Department for the purpose of obtaining more details regarding the assailant. Investigator Parmeter had deliberately avoided learning any information about the case except the fact of the crime, the fact that S.J. had spent time in the assailant’s company, and the approximate date of the assault. The police artist, Investigator William Casto, attended the session so that he could draw a picture based on S.J.’s descrip*798tion. Investigator Russell, who was primarily responsible for the investigation, was in an adjoining room videotaping the entire session through a two-way mirror.
Under hypnosis S.J. described her conversation with E.L. in her vehicle, her initial contact with the assailant, and the ride through the Anchorage area up to the point where the assailant pulled off the highway. She was not asked to describe the assaults. She was asked to give as much detail as possible about the assailant with her eyes closed while Investigator Casto made a drawing. Then she was asked to open her eyes to comment upon the drawing. She suggested some alterations. After the sketch was completed she was taken out of the hypnotic state. No videotape was made of Investigator Parmeter’s initial contact with S.J. or his parting comments to her after the completion of the hypnotic session.
Within the next several weeks, E.L. and S.J. were shown a number of photographic lineups by the police. During one session, E.L. identified a man other than Contreras, but changed his mind when he saw the man in person. When Contreras was arrested during the latter part of August for a similar rape, his photo was placed in a lineup and shown to S.J. and E.L. S.J. picked Contreras as the one who “looked a lot like” the assailant and made a definite identification after another lineup which contained a more recent photograph of Contreras. E.L. also made a positive identification of Contreras as the assailant.
In the meantime, subsequent to S.J.’s hypnotic session but prior to Contreras’ arrest, Lynette Camfferman phoned S.J. to apologize for the incident. Camfferman told S.J. that she was Contreras’ live-in girlfriend and that on the night in question Contreras had disclosed to her that he had robbed a young couple. He allegedly showed her the property he took from the couple including their driver’s licenses. Camfferman used this identification to locate S.J. to make her apology. During the phone conversation S.J. learned Camffer-man’s address; S.J. turned this information over to the police.
The case was presented to the grand jury. Among the witnesses who testified were the two victims, Comfferman, and Investigator Russell, who testified about the lineups. The grand jury indicted Contreras for the crimes.
B. GRUMBLES
The following facts also developed before the grand jury and in the evidentiary hearing held on Grumbles’ motion for a protective order.
On March 15, 1981, at approximately 12:00 a.m., Mary Hall was at her fiance’s home when she received a telephone call from Alaska General Alarm informing her that the burglar alarm at her residence had been activated. She thought that her dog may have triggered the alarm, so she went home to investigate. When she e.ntered the house she noticed that the kitchen window was broken, and then discovered an intruder whom she later identified as Grumbles. The intruder shot at her several times, hitting her once in the thigh.
Prior to entering her house, Hall had observed a yellow Gremlin with dark pin-strips parked in front of her house. When she looked outside after the assault, it was gone. Several hours later the police located a yellow Gremlin with red stripes on the sides in the parking lot of a cocktail lounge. Hall was brought to the restaurant to identify the vehicle as the one she had observed outside of her residence. Friends of Grumbles testified to the grand jury that Grumbles had been driving the Gremlin on the night of the burglary. After obtaining consent from Grumbles’ live-in girlfriend to search their residence, police discovered a .38 caliber Smith and Wesson revolver taken from Hall’s residence as well as the .22 caliber weapon with which Hall had been shot. When contacted, Grumbles admitted that he knew about the burglary and assault because his friend Scott had told Grumbles that he had committed the offense and had given Grumbles the guns.
On March 17, 1981, Investigator Parme-ter met with Hall for the first time to *799discuss hypnotism as a means of helping her identify her assailant. She expressed concern that the session might make her relive the physical and psychological trauma of the assault. Investigator Parmeter assured her that certain techniques could be used that would insulate her against any fearful situation. This preliminary discussion was not videotaped or recorded.
The next day Investigator Parmeter hypnotized Hall using the same procedures that had been used with S.J. The only difference was that because Hall had exhibited some signs of mental distress, Dr. Clifford Hunt, a licensed Ph.D in psychology, was present during the hypnotic session to care for Hall in the event that emotional distress was triggered. An attempt was made to videotape the session but the camera failed. An audio record of the session was obtained and was reviewed by the trial court.
IY. STANDARD OF REVIEW
Before discussing the expert testimony introduced in Contreras’ case and adopted by reference in Grumbles’ case, it is necessary to determine the standard governing our review of the trial court’s fact findings. We agree with the parties that a trial court’s findings regarding memory and the effects of interrogative techniques on eyewitness testimony and hypnosis in general are questions of legislative rather than adjudicative facts. See State v. Erickson, 574 P.2d 1, 4-6 (Alaska 1978). Consequently, we are free to exercise our independent judgment. In contrast, fact findings regarding the specific cases would be adjudicative facts reviewed under the clearly erroneous standard.11
V. EYEWITNESS TESTIMONY
In order to understand the expert testimony which was given at the Contreras evidentiary hearing and relied upon by Judges Serdahely and Buckalew, it is necessary to briefly summarize recent psychological and legal literature dealing with eyewitness testimony and the use of hypnotism in the investigation of crimes.
Within the last few years, legal and psychological literature has contained a number of books and articles discussing cases in which people have been wrongly convicted of crimes and imprisoned upon the basis of inaccurate eyewitness testimony. In England, a special commission was appointed under Lord Devlin to study the problem of identification evidence in light of two demonstrated miscarriages of justice resulting from faulty eyewitness identification. See Report to the Secretary of State for the Home Department Committee on Evidence of Identification in Criminal Cases, H.C. April 26, 1976. In the United States similar research has been conducted. See E. Loftus, Eyewitness Testimony (1979); P. Wall, Eyewitness Identification in Criminal Cases (1965). See generally Marshall, Marquis & Oskamp, Effects of Kind of Question and Atmosphere of Interrogation on Accuracy and Completeness of Testimony, 84 Harv.L.Rev. 1620 (1971); Levine & Tapp, The Psychology of Criminal Identification: The Gap From Wade to Kirby, 121 U.Pa.L. Rev. 1079 (1973); Note, Did Your Eyes Deceive You? Expert Psychological Testimony on the Unreliability of Eyewitness Identification, 29 Stan.L.Rev. 969 (1977).
*800The problem appears most forcefully, it is argued, where the only evidence against a defendant is the eyewitness testimony of a stranger. Jurors (and judges) tend to give undue weight to eyewitness testimony even when it is substantially impeached and contradicted.12 However, an eyewitness’s perception of a crime is frequently distorted by the experience. The witness’s memory may be further distorted by subsequent events, particularly the process of interrogation by police officers and lawyers. These risks are aggravated, some psychologists argue, by the general misconception that our eyes function like a video camera and our memory like an attached recorder so that every thing we see, and by extention sense, is somehow recorded and available for recall through a process akin to playback. In fact, modern research indicates that we only selectively store information perceived. More significantly, subsequent experiences may be incorporated into our memory and combined with earlier perceptions making it virtually impossible to sort out the original perceptions from subsequent accretions at a later time.13
*801Information perceived after an event will frequently alter a witness’s memory when it is planted through conscious or unconscious suggestion or when it is merely picked up through reading about the event or hearing others discuss it. In her experiments, Dr. Loftus discovered that misleading information had the greatest effect on memory when it was received substantially after the event perceived:
During the time between an event and a witness’s recollection of that event — a period often called the “retention interval” — the bits and pieces of information that were acquired through perception do not passively reside in memory waiting to be pulled out like fish from water. Rather, they are subject to numerous influences. External information provided from the outside can intrude into the witness’s memory, as can his own thoughts, and both can cause dramatic changes in his recollection.
People’s memories are fragile things. It is important to realize how easily information can be introduced into memory, to understand why this happens, and to avoid it when it is undersirable.
E. Loftus, Eyewitness Testimony at 86-87.
Problems of confabulation and suggestion are magnified by certain investigative techniques. Often the person interviewed will develop a rapport with the interviewer and subconsciously wish to please him or her by giving answers which the suspect is led to believe the examiner desires. This risk is particularly high when the subject has a *802substantial interest in the outcome of the investigation. These answers then become a part of the witness’s memory. The more a witness repeats her story the more fixed it becomes in her mind. In addition, the interviewer may unintentionally supply information that becomes part of the witness’s memory by the manner in which questions are asked.
We have quoted at some length from Dr. Loftus’ books.' Her studies of eyewitness testimony and the effect of various circumstances on an eyewitness’s memory in the absence of hypnotism serve to establish a control so that we can better evaluate the evidence regarding hypnotism and how it affects an eyewitness’s memory. The argument that previously hypnotized witnesses should be disqualified from testifying gains force only if hypnotism creates a risk of distorting memory that is substantially greater than, or qualitatively distinct from, the risk ordinarily posed by interrogating a victim who has rapport with her questioner and who has a vital interest in the identification and conviction of her assailant. In contrast, if improper interrogative techniques and the normal experiences encountered by eyewitnesses account for virtually all instances of memory distortion, then no special rule for hypnotism would appear warranted.
VI. HYPNOTISM
The term hypnotism conjures up a number of images which make a dispassionate consideration of its impact on eyewitness testimony difficult. Who can forget the classic films with John Barrymore as Svengali controlling the helpless Trilby, or Bela Lugosi as Dracula peering into the eyes of his enchanted victims, or Orson Wells as Cagliostro, using mesmerism to steal the diamond necklace which would topple the throne of France. Dwelling on these images will make the difficult task of sorting myth from fact impossible. When we objectively examine the literature, hypnotism becomes far more prosaic and loses its aura of mystery and legerdemain.
There is nevertheless a great deal of controversy as to precisely what hypnotism is. In a much-quoted article in the Encyclopedia Brittannica, Dr. Martin Ome says:
Despite active empirical study of hypnotic phenomena, there is no single generally accepted explanatory theory. Some theorists influenced by Pavlov think of hypnosis as a state of altered consciousness or partial sleep in which a person tends to respond to suggestions automatically and uncritically. While this view does not ignore increased suggestibility, it still holds that a subject not at the moment responding to a suggestion may nevertheless be in a trance state. There may be neurophysiological components of such a state. Such a position would be consistent with the notion, not generally accepts ed, that some animals can be hypnotized, as when a chick, suddenly placed on its back, tends to remain immobile in that position ....
Other theorists stress the social or interpersonal interactions that hypnotic be-haviour involves. They emphasize how an actor willingly and wittingly permits a director to guide him into living a part; how a patient may be relieved of a headache if given a pill that is actually phar-macologically inert (a placebo); how an uncommitted person becomes an enthusiastic supporter of crowd feeling at a demonstration; how a spectator flinches and jabs along with his favourite fighter; and how a student changes his views to those of an admired teacher. Hypnosis is held to consist of nothing more than events like these, there being no need to assume additional special states such as the inhibition of parts of the brain. Hypnotic events are results of interpersonal influences in which various abilities, skills, and response propensities are brought into play.
Evidence can be found to support both major theoretical approaches. Neural changes indeed do occur in hypnosis, but a unique physiological basis for the phenomenon remains to be established. Perhaps this reflects incomplete understanding of physiological alterations that *803may produce psychological change. Social-interaction theorists find it difficult to explain why posthypnotic suggestions are carried out even when the hypnotist neither knows nor apparently cares about the subject’s behaviour. While most current investigators tend to work within one or the other explanatory framework, most feel free to use propositions from both if they seem consistent with experimental or clinical observations.
9 Encyclopedia Britannica, Hypnosis 133, 135 (1974). Illustrative of the first view which assigns special significance to the induction procedure and the trance state is the work of Ernest R. Hilgard. See E. Hilgard, Hypnotic Susceptibility (1965). Illustrative of the second view which assigns special significance to the subject’s interest in the experiment and rapport with the hypnotist is the work of Thomas Barber. See T. Barber, Hypnosis: A Scientific Approach (1969); T. Barber, Suggested (“Hypnotic”) Behavior: The Trance Paradigm Versus an Alternative Paradigm, reprinted in E. Fromm & R.E. Shor, HYPNOSIS: Developments in Research and New Perspectives, 217 (2d ed. 1979); Barber & Calverley, Empirical Evidence for a Theory of Hypnotic Behavior: Effects on Suggestibility of Five Variables Typically Included in Hypnotic Induction Procedures, 29 J. Consulting Psych. 98 (1965).
Those authorities in general agreement with Hilgard use the term “induction” to describe the procedure for placing a person in a “trance.” The hypnotist first establishes rapport with the subject and then creates a passiveness that will make the subject receptive to suggestions, typically by engendering eye fatigue through focusing on a close object such as the hypnotist’s eyes, a swinging time piece suspended on a chain or a jewel rotating in a pendant. Finally, the hypnotist induces a trance through a series of suggestions requiring increased distortion of perception until it is clear that the subject is hypnotized. 9 Encyclopedia Britannica, Hypnosis 133,135 (1974). A person in a trance is believed to be in the “hypnotic state,” which Hilgard and his followers argue is fundamentally different from the waking and sleeping states. A subject is identified as being in the hypnotic state by his susceptibility to suggestions and his ability to perform certain feats which cannot be performed by someone who is awake. The level of hypnosis is partially judged by the subject’s overt and subjective responses to test suggestions, which may include suggestions for limb rigidity, age regression, analgesia, hallucination, amnesia, and post-experimental (“posthypnotic”) behavior. These responses, which for forensic purposes should be expanded to include memory distortion, constitute “hypnotic behavior.” Dr. Diamond, the primary proponent for holding previously hypnotized witnesses incompetent, generally accepts Hilgard’s model.
In contrast, Dr. Barber denies the existence of a “trance state,” and explains hypnotic behavior as a function of the subject’s attitudes, motivations and expectations regarding a given experiment. Essentially, a person exhibiting signs of being under hypnosis is simply one who is highly motivated to cooperate with the hypnotist.14 His ac*805tions “under hypnosis” simply reflect this motivation and his confidence in, and respect for, the examiner. Dr. Barber and his followers argue that they can duplicate any instance of behavior attributed to hypnosis without an induction procedure and without hypothisizing a trance state.15
*806In dividing the world of hypnosis theorists into two camps, we have followed the lead taken by Dr. Orne in his article in the Encyclopedia Britannica. We recognize that this underplays substantial differences between the experts which we have placed within the two camps and that purists might distinguish three or even more substantially distinguishable “camps.” See Orne, The Nature of Hypnosis: Artifact and Essence, 58 J. of Abnormal and Soc. Psych., 277 (1959). In this article Orne identifies three “camps” of hypnosis theorists and indicates that his fourth view is eclectic.16 Nevertheless, for purposes of this case, we believe that our dichotomy is useful.
For our purposes, it is sufficient to recognize that the proponents of the various views concede that their opponent’s views are scientifically tenable. Both Dr. Hil-gard’s and Dr. Barber’s views were mentioned at the Contreras evidentiary hearing. Judge Serdahely concluded that both, among others, were part of the scientific community whose views were relevant to an understanding of the issues presented by Contreras’ motion for a protective order.
VII. THE EFFECT OF HYPNOTISM ON EYEWITNESS TESTIMONY
A number of recent articles discuss the possibility that hypnotizing a witness might lead to distortions in memory. See, e.g., Kroger & Douce, Hypnosis in Criminal Investigation, 27 Int’l J. Clinical & Experimental Hypnosis No. 4, 358 (1979). The writers note that hypnotized subjects are susceptible to the same influences that Dr. Loftus found in studying unhypnotized witnesses. They accept the Hilgard model and use the term hypnotism to include three components: (1) the induction, (2) the resulting trance state, and (3) the subsequent interview or interrogation. Id. at 361-64. Our analysis of the effects of hypnotism on eyewitness testimony requires that we isolate the effects of hypnotism as such, i.e., induction plus trance, from the effects of interrogation techniques which might be as readily used with nonhypnotized subjects. In this context we define “hypnotized subjects” in conformity with the Hilgard model to include those who have been through an induction and believe themselves to be in a trance.
We have only found one article cited in the case law and the literature which purports to support with experimental data the proposition that the memories of hypnotized subjects are more susceptible to distortion through suggestion and confabulation than nonhypnotized subjects. See Putnam, Hypnosis and Distortions in Eyewitness Memory, 27 Int’l J. Clinical & Experimental Hypnosis No. 4, 437 (1979).17 Dr. Putnam chose *807sixteen college students based on presumed equal susceptibility to suggestion. He showed them a videotape of a simulated accident and asked that they pay close attention, but not memorize details. He instructed them to imagine that they had witnessed the “accident” and told them they would be interviewed by the police. He arranged for one-half of the group to be hypnotized, and told them that they would have total recall under hypnosis. He asked the hypnotized students leading questions which included misinformation. He asked the same questions of the control group. The results of the test were as follows:
The Ss [subjects] made more errors when answering leading questions under hypnosis than in a normal waking state. This finding supports the hypothesis that Ss are more suggestible in the hypnotic state and are, therefore, more easily influenced by the leading questions. This result argues strongly that great caution should be exercised when using hypnosis to question eyewitnesses in criminal investigations. The need for caution is further emphasized by the analysis of the confidence ratings that were given by Ss for their responses. Even though the hypnotic Ss made more errors, they were just as confident as the normal Ss who made fewer errors. Moreover, when the hypnotic Ss were asked if they felt that they were more accurate under hypnosis than if they had answered the questions in a waking state, they all felt that they had been more accurate under hypnosis. These Ss also stated that they would prefer to be questioned under hypnosis if they witnessed an actual accident or crime. All of these effects seem to indicate quite clearly that under hypnosis, Ss answer more leading questions incorrectly and they are unaware that their responses are inaccurate.
It is important to note that five of the leading questions consisted of merely changing the article “a” to “the,” but this rather subtle change had substantial effects. For example, asking Ss if they saw “the license plate” when it was not visible at all, not only elicited positive responses from some of the hypnotic Ss, they also offered partial descriptions of the license plate number. One S said it was a California license plate which began with W or V. This constructed information was not obtained under any coercion. There is every reason to believe that the tendency to distort recall would be even greater in a real criminal investigation, where the witness may have a vested interest in recalling the information.
Id at 444.
In evaluating Dr. Putnam’s conclusions, it is wise to bear in mind an observation Drs. Barber and Calverley made in considering the results of an earlier experiment which sought to distinguish between the test results of those who were hypnotized and those who were not:
The above formulation [rejecting the assumption that induction followed by a trance per se accounts for differences between test responses distinguishing hypnotized from nonhypnotized subjects] is based on the following reasoning. When literate subjects in Western culture are told that they have been allocated to a hypnosis group and are to be hypnotized, they are being told implicitly that they are participants in an important and unique experiment in which it will not only be possible but also easy to respond to test suggestions of arm levitation, limb heaviness, amnesia, and so on, and that they are expected to cooperate fully and to try to the best of their ability to experience that which is suggested. Furthermore, since literate individuals in our culture associate suggestions of relaxation, drowsiness, and sleep with hypnosis, such suggestions serve to emphasize to them that the situation is hypnosis in which *808high response to test suggestions is not only possible but is also desired and expected. Conversely, when literate individuals are not given motivational instructions but are simply told that they have been assigned to a “control group” and will not be hypnotized ... they are being told by implication that they are not expected to show a high level of response to test suggestions of the type associated with the word hypnosis, for example, test suggestions of limb and body rigidity, hallucination, and amnesia.
Barber & Calverley, supra at 106 (emphasis in original).
It is significant that Dr. Putnam’s subjects had all previously been through “hypnosis.” The subjects were students who had no particular stake in the outcome of the simulated investigation except the interest generated by the experiment itself. No effort was made to give the control group an independent incentive to “correctly” recall the videotape by, for example, telling them in advance that “correct” answers would be rewarded in some way. It is not unreasonable to assume that the hypnotized group might be more interested in the “experiment” and therefore pay closer attention to the person supplying information or asking questions than the members of the control group who may have been bored by the whole experiment. Dr. Lof-tus’ conclusions would suggest that if the hypnotized group believed that they had total recall, confabulation (which she calls refabrication) would be encouraged whether or not they were in fact hypnotized, i.e., in an altered state of consciousness. These considerations at least cast doubt on the assumption that hypnosis itself increases the risk of memory distortion where the subject has an independent incentive to provide correct answers, such as when she is the alleged victim of a crime striving to identify and punish her assailant. Conversely, Putnam’s experiments might be persuasive in dealing with eyewitnesses who were otherwise uninvolved bystanders.
Putnam’s experiments might have been more useful for our purposes if his control group had received some incentive to arrive at a correct recollection of the events portrayed on the videotape. In this way the subjects would more closely approximate rape victims who are typically highly motivated to aid in the apprehension of their assailant once they agree to testify. In order to aid in the accurate evaluation of the relative effect of induction and trance on susceptibility to cues in leading questions, the hypnotized group should probably have received similar “motivation.”
In summary, Contreras and Grumbles and the authorities upon which they rely raise three objections to the competency of previously hypnotized witnesses. They argue: (1) that the hypnotist may “suggest” details to the witness which will be incorporated into her memory and become a part of her subsequent testimony; (2) that where the subject has no memory of an event but is pressed to give details and wishes to please the examiner, she will confabulate those details out of unrelated experiences and imagination and incorporate the confabulated details into her subsequent testimony; and (3) that the process of hypnosis will give the subject such confidence in her “restructured memory” that she will become immune to cross-examination because her demeanor will exude confidence and mislead the jury.
Dr. Martin Orne suggests certain standards as safeguards. Orne, The Use and Misuse of Hypnosis in Court, 27 Int’l J. Clinical & Experimental Hypnosis No. 4, 311 (1979). These standards address the risk of suggestion and confabulation by requiring that the examiner have no knowledge of the event to communicate to the subject and requiring that the procedures used avoid pressuring the subject to provide details. In addition, Ome’s standards recognize that a hypnotic session may result in mental or emotional injury to the person hypnotized and seek to limit that risk.18
*810VIII. THE EXPERT TESTIMONY AT THE CONTRERAS EVIDENTIARY HEARING
Three expert witnesses testified at the Contreras evidentiary hearing regarding use of hypnosis as an investigative tool. The state called Dr. Martin Reiser.19 He indicated that he had reviewed the videotape and was satisfied that no impermissible suggestion had taken place. He noted that several times S.J. responded, “I don’t know” to questions, indicating resistance to confabulation. Finally, Dr. Reiser criticized the proposed standards for conducting an investigative hypnosis session developed by Dr. Orne.
Dr. Bernard L. Diamond testified on Contreras’ behalf.20 Dr. Diamond testified that a prospective witness who is hypnotized tends to produce a memory that is a mixture of real experience and confabulation or fantasies. While he conceded that the normal individual has a tendency at times to embellish his memories by filling in missing details and organizing them into a coherent whole, this process is normally subject to self-criticism. An honest individual, Dr. Diamond said, will readily acknowledge and communicate doubts. Hypnosis by its very nature interferes with this process of self-criticism, not necessarily by eliminating it, but by distorting it so that the individual loses the capacity to make distinctions between what she really remembers and what she has drawn from other thoughts or experiences. In effect, Dr. Diamond cautioned, a hypnotized person develops “source amnesia.”21 Dr. Diamond stressed that an individual, particularly an eyewitness submitting to hypnosis, is under substantial psychological pressure to please the hypnotist and will try to produce what is expected by the interviewer even if it means confabulating all of the details.
Dr. Diamond was particularly critical of the procedure used by Investigator Parme-ter and recommended by Dr. Reiser, i.e., telling the witness that her mind is like a camera with sophisticated lenses that permit her to “freeze motion” or go back in time to view past experiences in a detached manner. Dr. Diamond explained that since the average individual’s memory is faulty, and since a witness probably does not perceive all the relevant details of a description, a belief in total recall and a desire to supply details to satisfy the examiner will necessarily result in confabulation. He concluded that once this process is complete, most hypnotized subjects have the persist *811ent belief that they knew the confabulated material all along. Dr. Diamond testified: “[I]t is my conviction that once an individual has been hypnotized, their memory has been irrevocably contaminated in such a way that it cannot anymore be decontaminated .... ”
According to Dr. Diamond it is impossible, either through additional hypnosis, analysis by an expert, or any other technique available in a legal setting, to distinguish between the facts that would have been obtained if the individual had not been hypnotized and those that are attributable to hypnosis. Dr. Diamond described hypnosis as “a form of systematic tampering with the memory.” In his opinion, the most serious contaminating effect of hypnosis is its role in altering the demeanor of a witness by creating subjective confidence in false memory.
On cross-examination, Dr. Diamond conceded that psychological problems exist with all eyewitness testimony. He could not deny that some of the risks associated with hypnotically induced evidence may also apply to the testimony of any witnesses who have been interrogated but not hypnotized. He insisted, however, that there are quantitative and qualitative differences between the testimony of previously hypnotized witnesses and the testimony of witnesses who have merely been coached or subjected to suggestions or leading questions.22 Dr. Diamond admitted that his *812view, which would render any witness hypnotized during a criminal investigation incompetent to testify, is perhaps the most extreme view held by experts in the field.
The heart of the problem presented by Investigator Parmeter’s hypnosis of S.J., as Dr. Diamond perceived it, was not that facts may have been implanted in her mind through suggestions, but rather that there was a risk of confabulation. S.J. was put in a position where (1) she was made to feel that she should supply details about the crime in response to persistent questioning, and (2) she was made to believe that by virtue of hypnosis she was capable of giving those details. She was prompted to provide the details demanded even though she may not have had an accurate memory of those details. Thus, Dr. Diamond hypothesized, S.J. supplied details from her general experience. Although some of them may have reflected the incident, some may not have; there was no way to determine which details were accurate.
On cross-examination Dr. Diamond was asked, “Is it true that confabulation, source amnesia and fantasy which [the witness now believes] to be true, occur in persons other than those who have been hypnotized?” He answered affirmatively. Dr. Diamond had read the literature concerning eyewitness testimony in general. Although he had not reached any general conclusions, he did say, “I think in fairness one has to say that every contaminant that can occur in hypnotized witnesses can under some cir-eumstances occur in nonhypnotized witnesses, but what I don’t know is how frequent ly, how often, under what circumstances, when, how you detect it, and what you do about it.” On further cross-examination, Dr. Diamond conceded that he could not be certain that S.J.’s testimony would be unreliable or false. He emphasized, however, that it was “not possible to accurately determine how much or how little contamination occurred as a result of the hypnotic experience.” When asked why witnesses are allowed to testify at all when it is also impossible to determine whether a nonhyp-notized witness is telling the truth, Dr. Diamond replied: “Well, you’re getting into the whole problem of eyewitness testimony, it has been proposed that no one should be able to testify. As I say, this is a serious problem. I’m not yet prepared to give an opinion about this.”
The final expert witness to testify was Dr. Donald Rossi, who was called by the state in rebuttal.23 Dr. Rossi discussed Dr. Orne’s guidelines and concluded that they had been met in Investigator Parmeter’s hypnosis of S.J. Dr. Rossi was convinced after viewing the videotape that the risk of suggestion was minimal. He mentioned that the risk that Parmeter would “cue” S.J. was minimized when S.J. was requested to keep her eyes closed during hypnosis.
Dr. Rossi stressed that confabulation occurs as a byproduct of all interviews and interrogations when a witness is pushed for *813details whether or not hypnosis is used. In fact, he stated that in an interview conducted without a “trance,” there may be more done to produce confabulation than in a properly conducted hypnotic session where measures are taken to avoid it. Dr. Rossi disputed the suggestion that a confabulation automatically becomes a part of a subject’s memory. He asserted that a hypnotized subject may remember a confabulation later and correct it. Dr. Rossi pointed out that there is no generally accepted theory about how the' memory works. He stressed that, in conducting an interview with a person in a fully awake state, confabulated details can only be recognized by comparing the information the witness furnishes with corroborating circumstances. Dr. Rossi conceded that hypnosis will not guarantee truth. He also conceded that a trance state introduces variables which create a greater likelihood of affecting credibility if the hypnotist is not careful. Finally, Dr. Rossi stated that the effect of hypnosis on a person’s subsequent demeanor depends on the nature of any suggestions implanted. After reviewing the videotape of Investigator Parmeter’s hypnosis of S.J., Dr. Rossi concluded that he did not observe anything that would have caused the contamination of SJ.’s memory.
Dr. Rossi stated that Dr. Diamond was in a small minority of those familiar with hypnosis who would argue that a person whose memory was refreshed by hypnosis should be totally incompetent to testify. He objected to the theory that previously hypnotized witnesses should be incompetent to testify because of their susceptibility to suggestion and confabulation because it does not acknowledge that the same risks apply to unhypnotized witnesses who are nevertheless presumed competent to testify. Dr. Rossi also stated that if all those knowledgeable about hypnosis and its forensic use were placed on a continuum, Dr. Reiser would be at one end concluding that no safeguards were needed, and Dr. Diamond at the other concluding that no safeguards were satisfactory. The majority would be in the middle, concluding that the same safeguards utilized to prevent distortion of memory in witnesses who were not hypnotized should be satisfactory to prevent contamination when used with hypnotized witnesses.
IX. THE CASE LAW CONSIDERING THE COMPETENCY OF PREVIOUSLY HYPNOTIZED WITNESSES
The substantial diversity of opinion about forensic use of hypnosis demonstrated by the expert testimony in the Contreras case is also reflected in the reported cases. Essentially, three positions have developed. The first group of cases views hypnosis as affecting a witness’s credibility but not her competency or the admissibility of her testimony. See Clark v. State, 379 So.2d 372, 375 (Fla.App.1979); People v. Smrekar, 68 Ill.App.3d 379, 24 Ill.Dec. 707, 712-14, 385 N.E.2d 848, 853-55 (1979); State v. Wren, 425 So.2d 756, 759 (La.1983); State v. Greer, 609 S.W.2d 423, 436 (Mo.App.1980), vacated on other grounds, 450 U.S. 1027, 101 S.Ct. 1735, 68 L.Ed.2d 222 (1981); State v. McQueen, 295 N.C. 96, 244 S.E.2d 414, 427-29 (1978); Chapman v. State, 638 P.2d 1280, 1284 (Wyo.1982). See also Harding v. State, 5 Md.App. 230, 246 A.2d 302, 306 (1968), cert. denied, 395 U.S. 949, 89 S.Ct. 2030, 23 L.Ed.2d 468 (1969), overruled, Collins v. State, 52 Md.App. 186, 447 A.2d 1272, 1283 (1982) (adopting approach articulated in State v. Hurd, 86 N.J. 525, 432 A.2d 86 (1981)). Cf. State v. Jorgensen, 8 Or.App. 1, 492 P.2d 312, 315 (1971).
The second line of cases accepts the possibility that hypnosis could affect competency as well as credibility but permits a previously hypnotized witness to testify if the state establishes a foundation to show that the hypnosis occurred under stringent safeguards. See United States v. Adams, 581 F.2d 193, 198-99 (9th Cir.1978), cert. denied, 439 U.S. 1006, 99 S.Ct. 621, 58 L.Ed.2d 683 (1978); State v. Hurd, 86 N.J. 525, 432 A.2d 86, 89-92 (1981) (adopting Dr. Orne’s guidelines, see supra n. 18); State v. Beachum, 97 N.M. 682, 643 P.2d 246, 253-54 (N.M.App.1981); State v. Long, 32 Wash.App. 732, 649 P.2d 845, 846-47 (1982); State v. Armstrong, 110 Wis.2d 555, 329 N.W.2d 386, 394 *814(1988), cert. denied, - U.S. -, 103 S.Ct. 2125, 77 L.Ed.2d 1304 (1983). See also Commonwealth v. Juvenile, 381 Mass. 727, 412 N.E.2d 339, 344 (1980).
Finally, there are a number of recent cases holding that participation as a subject in a hypnotic session for investigative purposes so distorts a witness’s memory that she is rendered incompetent and cannot testify regarding any fact elicited for the first time during the hypnotic session. See State v. Stolp, 133 Ariz. 213, 650 P.2d 1195, 1196 (1982); State ex rel. Collins v. Superior Court, 132 Airz. 180, 644 P.2d 1266, 1293-95 (1982) (hypnotized witness may only testify to facts demonstrably recalled prior to hypnosis); People v. Shirley, 31 Cal.3d 18, 181 Cal.Rptr. 243, 250, 641 P.2d 775, 804 (1982), cert. denied, - U.S.-, 103 S.Ct. 13, 73 L.Ed.2d 1400 (1982) (witness who has been hypnotized incompetent to testify on any matter related to subject discussed under hypnosis); People v. Gonzales, 108 Mich. App. 145, 310 N.W.2d 306, 314 (1981); State v. Mack, 292 N.W.2d 764, 772 (Minn.1980); State v. Patterson, 213 Neb. 686, 331 N.W.2d 500, 504 (1983); State v. Palmer, 210 Neb. 206, 313 N.W.2d 648, 654 (1981); People v. Hughes, 59 N.Y.2d 523, 466 N.Y.S.2d 255, 266, 453 N.E.2d 484, 495 (N.Y.1983); Commonwealth v. Nazarovitch, 496 Pa. 97, 436 A.2d 170, 176-78 (1981).
X. THE FRYE TEST
The disagreement in the cases is in part based on their treatment of the so called Frye test. See Frye v. United States, 293 F. 1013 (D.C.Cir.1923). The cases which exclude testimony of eyewitnesses who have been hypnotized do so based on their application of the Frye test. Those cases that find the Frye test inapplicable tend to admit the testimony of hypnotized witnesses.
Frye was charged with second-degree murder. He sought to introduce expert testimony that he had been given a systolic blood pressure deception test and passed it. He reasoned that this evidence would serve to strengthen his credibility with the jury. The testimony was rejected and Frye was convicted. On appeal, the Court of Appeals affirmed. Id. The court cited no authority and made no attempt to fit the rule it announced into the rules of evidence. The court merely said:
Somewhere in this twilight zone [between the experimental and demonstrable stages] the evidential force of the [scientific] principle must be recognized, and while courts will go a long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs.
Id. at 1014.
The Frye test has been frequently criticized. See, e.g., E. Cleary, McCormick on Evidence, § 203, at 489-91 (2d ed. 1972); Strong, Questions Affecting the Admissibility of Scientific Evidence, 1970 U.Ill.L.F. 1, 10-15 (1970) (the considerations which have moved courts to apply Frye to various scientific principles may be adequately accommodated within the usual rules of evidence, and within the adoption of special rules of limited ambit). Professor Wright has argued that the Frye test cannot survive the enactment of the Federal Rules of Evidence. See C. Wright & K. Graham, Federal Practice and Procedure: Evidence § 5168, at 91 (1978). He reasons that the Federal Rules of Evidence control the issue of admissibility of expert testimony and leave no place- for Frye to operate. Accord 3 J. Weinstein Evidence ¶ 702[03], at 702-16-18 (1982). But see Gianelli, The Admissibility of Novel Scientific Evidence: Frye v. United States, a Half-Century Later, 80 Colum.L.Rev. 1197, 1229 (1980). One federal court seems to agree with Wright and Weinstein. See United States v. Williams, 583 F.2d 1194, 1197-98 (2nd Cir.1978), cert. denied, 439 U.S. 1117, 99 S.Ct. 1025, 59 L.Ed.2d 77 (1979). Accord State v. Hall, 297 N.W.2d 80, 84-85 (Iowa 1980), cert. denied, 450 U.S. 927, 101 S.Ct. 1384, 67 L.Ed.2d 359 (1981) (limiting Frye under state rules patterned on federal rules).
*815We are concerned that those cases using the Frye analysis to justify a per se rule rendering incompetent any eyewitness whose memory has been “refreshed” by pretrial hypnosis suffer from at least four defects in their treatment of Frye.
First, they assume the Hilgard model for hypnosis without subjecting this assumption to the Frye standard which requires general acceptance within the scientific community. See, e.g., State v. Mena, 128 Ariz. 226, 624 P.2d 1274, 1276-77 (1981), modified, State ex rel Collins v. Superior Court, 132 Ariz. 180, 644 P.2d 1266, 1269, 1286 (1982); People v. Gonzales, 310 N.W.2d at 309; State v. Palmer, 313 N.W.2d at 654; Commonwealth v. Nazarovitch, 436 A.2d at 173. This assumption overlooks the work of Barber and his followers and thus neglects to address a highly significant inquiry central to the validity of their use of the Frye test to support a per se rule rendering incompetent eyewitnesses who have undergone pre-trial hypnosis.
Second, in considering memory distortion, they fail to distinguish the effects of the interrogation procedure alone from the effects of induction, trance and interrogation together. Putnam is the only authority for the proposition that hypnotized witnesses are more susceptible to memory distortion than nonhypnotized witnesses and his single experiment hardly qualifies as general community acceptance.
Third, they erroneously assume that a per se rule, in preference to a case-by-case analysis under local equivalents to A.R.E. 403, will avoid both a time consuming and expensive battle of experts and the confusion engendered by diverting the jury’s attention from the defendant’s guilt to possible hypnotic distortions of a witness’s memory. See State ex rel. Collins v. Superior Court, 644 P.2d at 1283; People v. Shirley, 641 P.2d at 787-88; State v. Mack, 292 N.W.2d at 766; State v. Palmer, 313 N.W.2d at 654-55. It appears that these eases have only considered the tip of the iceberg. Under a per se rule any witness who undergoes pretrial hypnosis to refresh her recollection is incompetent to testify (in Arizona regarding facts not demonstrably recalled prior to hypnosis and in California regarding all matters relating to events in issue). Since defense counsel will always be motivated to prevent a crime victim from testifying, they will naturally try to show that the victim has been interrogated under hypnosis, and the trial court must allow some opportunity to make this showing. As we have seen, one tenable paradigm of hypnosis, ascribed to by Barber and his allies, includes within it any subject, whether undergoing an induction and trance or not, who enters into a special relationship of trust with the interviewer and has a substantial incentive to see the investigation succeed. Does the per se rule disqualify witnesses exposed to pretrial hypnosis as defined by Barber? The interaction between the crime victim and the investigating officer or prosecutor would frequently appear to satisfy this test. In any event, one can easily foresee that consideration of the question would lead to a battle of experts.
It is instructive that proponents of the Hilgard view criticize the results of Barber’s experiments by pointing out that the task motivational language, which he uses to build rapport and capture the subject’s interest in the experiment, could itself constitute an induction and induce a “trance.” P. Sheehan and C. Perry, supra n. 16 at 106. Experts in hypnosis from the Hilgard camp might well make the same contention wherever it is shown that an eyewitness with a substantial interest in the outcome of a case is subjected to pressure to identify her assailant.
Even if we assume that the courts will pragmatically, if perhaps illogically, limit the per se rule to those who have gone through an induction arguably resulting in a trance, counsel may still ask to be permitted to show that someone who was not in a trance was nevertheless subjected to the same kind of memory distorting experiences and therefore subject to the per se rule. It appears that Dr. Loftus and her colleagues are prepared to give testimony about factors leading to distortion in eyewitness tes*816timony in general which is virtally indistinguishable from Dr. Diamond’s testimony regarding memory distortion in hypnotized witnesses. See State v. Chapple, 135 Ariz. 281, 660 P.2d 1208, 1217-19 (1983) (holding exclusion of such expert testimony an abuse of discretion under the Arizona counterpart to A.R.E. 403, where an identification by an eyewitness who was not hypnotized was not corroborated). It would appear, therefore, that the per se rule will increase rather than minimize the battle of the experts.
Fourth, proponents of the per se rule fail to consider the possibility of corroboration as a validator of an eyewitness identification by a previously hypnotized witness. See People v. Gonzales, 310 N.W.2d at 310 n. 4 (quoting affidavit of Dr. Martin Orne which suggests external corroboration can be used to distinguish between confabulation and accurate recall). Since the cases adopting per se rules are frequently cases in which the eyewitness testimony was not corroborated, it is strange that the courts do not even consider limiting the per se rule to cases in which no corroboration exists. See State v. Mack, 292 N.W.2d at 772; People v. Shirley, 641 P.2d at 776-77. It is possible that the Shirley court rejected this alternative because it thought that hypnotism tended to clothe the witness’s entire testimony in an artificial but impenetrable aura of certainty, see 641 P.2d at 806, b,ut this seems to be a misunderstanding of the effect of distortion on memory and confidence levels. As we have seen, whether hypnotized or not, most witnesses whose memories have been distorted are confident about both false information and actual perception and are unable to tell the difference between them. It is significant that Putnam, the only expert who has attempted to test this assumption, found that the hypnotized subjects, while more apt to accept false information than the control group, were equally as confident, but no more confident, than the unhypnotized control group about the false information accepted. Putnam, Hypnosis and Distortion in Eyewitness Memory, 27 Int’l J. Clinical and Experimental Hypnosis No. 4, 437, 444 (1979). Thus both the California Supreme Court and the New York Court of Appeals appear to have misread Putnam when they cite him as proving that hypnotized witnesses are more confident about those parts of their memories proved to be distorted than the unhypnotized control group is about equally false memories they hold. See People v. Shirley, 641 P.2d at 803-04; People v. Hughes, 466 N.Y.S.2d at 260, 453 N.E.2d at 489.
Further, to apply the Frye test here would require more than a finding that it survives enactment of the Alaska Rules of Evidence. Strictly speaking, no expert is involved; S.J. and Hall would simply testify like any other witnesses. There would be no need to mention that they had been hypnotized. Courts relying on the Frye test to exclude such testimony attempt to avoid this problem by finding that the induction process and the “trance” which presumably follows it somehow produces the hypnotized witness’s subsequent testimony. See People v. Shirley, 641 P.2d at 796. But this seems plainly wrong. All the authorities on hypnotism reject the contention that the subject is simply a ventriloquist’s dummy parroting the hypnotist’s testimony. See 9 Encyclopedia Britannica, Hypnosis 133, 139 (1974).
Justice Mosk seeks to justify use of the Frye test by arguing:
it appears that hypnotizing a witness to improve his memory is not in fact like “any other method” of refreshing a witness’ recollection. These sources reveal that the hypnotic process does more than permit the witness to retrieve real but repressed memories; it actively contributes to the formation of pseudomemories, to the witness’ abiding belief in their veracity, and to the inability of the witness (or anyone else) to distinguish between the two. In these circumstances, as noted above, the resulting recall of the witness “is dependent upon, and cannot be disassociated from” the underlying hypnosis. And if the testimony is thus only as reliable as the hypnotic process itself, it must be judged by the same standards of admissibility.
*817641 P.2d at 795-96 (citation omitted). Justice Mosk fails to distinguish between hypnotized and nonhypnotized witnesses, ignoring the distinction between the effect of interrogation techniques and the effect of the induction and the trance state on the subject’s memory. As we have seen, the danger of confabulation, the danger that information supplied through suggestion will become a part of the witness’s memory, and the danger that the witness will be as confident about the inaccurate recollections as the accurate recollections, are possible side effects of the interrogation process to which any witness is vulnerable regardless of induction and trance.
Further, if Dr. Barber is correct, any hypnotic effect will be duplicated with a subject who has not been hypnotized, so long as the subject has sufficient interest in the success of the experiment and rapport with the experimentor. Induction and trance under Barber’s view are thus irrelevant even with those who are hypnotized, except to the extent that they serve to enhance the subject’s interest in the outcome of the investigation and create an attitude of cooperation. It would appear that a rape victim who had overcome the embarrassment and pain caused by the incident so that she was willing to testify, even without hypnotism, would be highly motivated to cooperate, would have a substantial interest in the outcome of the investigation, and would probably identify with the police and prosecutors seeking to convict her assailant. In summary, Dr. Barber’s views are part of the relevant scientific community as Judge Serdahely found. Moreover, unless Dr. Barber’s views are rejected, it would be inappropriate to say that a previously hypnotized witness’s trial testimony was any more the product of the hypnotic session that a nonhypnotized witness’s trial testimony is the product of the interviews she previously had with police and prosecutors.
More significantly, in Shirley, Justice Mosk seems to turn the Frye test on its head. As originally applied, the Frye test prevented an expert from vouching for the credibility of a witness and then supporting his (the expert’s) testimony by reference to some scientific principle that purportedly proved that the initial witness was telling the truth, i.e., truth serum. Later, the Frye test was extended to prevent an expert from attacking the credibility of a witness and then justifying the attack by reference to a scientific principle. See, e.g., United States v. Fosher, 590 F.2d 381, 383 (1st Cir.1979) (trial court did not err in refusing to admit expert testimony on unreliability of eyewitness identification because such testimony did not clearly meet the Frye standard of reliability); United States v. Amaral, 488 F.2d 1148, 1152-53 (9th Cir.1973).
In Shirley the California Supreme Court permitted an expert, Dr. Diamond, to attack the credibility of a class of witnesses, i.e., those who had been hypnotized prior to trial, based on scientific principles derived from the results of studies of the impact of improper interrogation techniques on the memory of hypnotized subjects. Based on this testimony, the California Supreme Court disqualified previously hypnotized eyewitnesses as a class.24 Curiously, the court did not examine whether Dr. Diamond’s views had received general scientific acceptance. Dr. Diamond readily concedes that his view regarding the effect of hypnotism on memory distortion in eyewitnesses is a minority view among experts familiar with the subject.
XI. THE COMPETENCY OF S.J. AND HALL
The parties agree that the cases of Contreras and Grumbles, when considered together, provide a necessary and convenient vehicle for choosing among the alternatives discussed in the cases cited above. The parties ask us to establish general rules *818to govern the admissibility in Alaska of eyewitness testimony when the witness’s memory has been “refreshed” by hypnosis prior to trial. We decline to do so. It is not necessary to decide the extent to which Frye survived the enactment of the Alaska Rules of Evidence or, if viable, Frye ⅛ relevance in a case of pre-trial hypnosis. We believe that the specific facts of these cases require a conclusion that Hall and S.J. should be permitted to testify subject to cross-examination. This holding has limited application beyond the facts of these cases, however, because of the substantial corroboration serving to establish the accuracy of their identifications. Such corroboration may not be present in other cases.
S.J.’s identification of Contreras is corroborated by the identification by E.L. who was not hypnotized. More significantly, Camfferman, Contreras’ girlfriend, testified before the grand jury that the night of the incident Contreras came home, and told her that he had just robbed a couple. Camffer-man, after identifying the victims from their stolen driver’s licenses, called S.J. to apologize for Contreras’ actions.
Hall’s identification of Grumbles was also substantially corroborated. Prior to any hypnosis, Hall told the police that an orange or yellow Gremlin with dark pinstripes had been parked near her residence when she arrived, but that it was gone after the incident. Subsequent to Hall’s hypnotism Grumbles was arrested on unrelated grounds as a fugitive from justice. Because Grumbles matched the description provided by Hall, police officers visited his residence and obtained consent to search from his girlfriend. Two firearms were seized and the serial number on the .38 caliber revolver matched that of the gun stolen during the burglary of Hall’s residence. The other gun was later identified by Hall as similar to the one used by her assailant. Through weapons identification technology, comparing this weapon and the shell casings found at the scene of the burglary, the F.B.I. determined that the gun seized from Grumbles’ apartment was fired in the Hall residence. Further, a friend of Grumbles testified that she had loaned her orange Gremlin to Grumbles the night of the burglary. She also testified that Grumbles had been wearing a brown leather jacket. A brown leather jacket seized at Grumbles’ apartment was identified by Hall as being worn by her assailant. A glove found at Hall’s residence was identified as similar to gloves known to belong to Grumbles. When interviewed, Grumbles denied being the assailant but claimed to be aware of the burglary and assault because it had been committed by his friend, Scott, who had given the guns to him. Police officers testified that they interviewed Scott and that he differed substantially in appearance from Grumbles and from the description given by Hall.
There is nothing in either record to suggest that any of the corroborating evidence which we have mentioned was in any way affected by the fact of S.J.’s and Hall’s hypnosis. > More significantly, nothing indicates that Officer Parmeter was aware of the corroborating evidence prior to hypnotizing either S.J. or Hall.
While cases can be hypothesized in which the only evidence against the defendant is his alleged victim’s eyewitness identification and in which the arguments against permitting a hypnotized witness to testify might be extremely forceful, the facts of these cases deprive these arguments of any force they might otherwise have had. Neither Contreras nor Grumbles were suspects at the time S.J. and Hall were hypnotized. Thus, it is highly unlikely that Investigator Parmeter intentionally fed either S.J. or Hall details regarding the appearance of Contreras or Grumbles in an effort to structure an identification. There is no suggestion that S.J. knew Contreras or that Hall knew Grumbles before the incidents. Consequently, there is no basis for an inference that either would have had a conscious or unconscious bias against either assailant; nor is there a basis to find motivation for a conscious or unconscious misidentification.
The records do not show that either victim had previously seen her alleged assailant near the time of the incident under *819circumstances which might make her sufficiently familiar with his appearance that she would mistakenly associate that recollected appearance with the appearance of her assailant. Even if we partially discount S.J.’s and Hall’s identifications because of the obviously stressful situations in which their observations took place, and the risk that their attention might have been focused on their assailants’ weapons rather than on the features of the assailants’ countenances, the records establish that both Hall and S.J. spent enough time in the presence of their assailants to observe and form distinct impressions of their features.
In summary, the case for a misidentification comes down to a risk that Investigator Parmeter would unintentionally communicate cues to S.J. and Hall which, in combination with their own confabulation, would lead them to falsely identify two men who coincidentally had possession of property stolen from them. Camfferman testified that Contreras confessed the crime to her the night it allegedly occurred and corroborated his confession by showing her S.J.’s identification. Grumbles admitted knowledge of the robbery and sought to blame his friend, but the evidence suggests no similarity in appearance between Scott and Grumbles. It is unlikely that Hall coincidentally confabulated the details of Grumbles’ appearance rather than Scott’s. By the same token, it is improbable that S.J. confabulated details regarding the appearance of a man Camfferman was coincidentally preparing to falsely accuse of assaulting S.J. Under the peculiar facts of these cases, we are satisfied that the risk of misidentification is virtually nonexistent. Consequently, we conclude that the probative value of these identifications is very high.
In addition, we are satisfied that the prejudice to the state if the identifications are excluded would be great. Both victims had substantial exposure to their assailants. A jury would naturally expect an explanation for any inability they might have to identify those assailants. The jury would naturally wonder why E.L. could identify Contreras and S.J. could not. Foreseeable explanations would either leave the jury with a mistaken belief that the victims were unable to identify their assailants, which would be unfair to the state or, alternatively, that the court was keeping a proper identification out of evidence on a technicality, which would be unfair to Contreras. Cf. People v. Hughes, 466 N.Y.S.2d at 266, 453 N.E.2d at 495. Weighing the probative value of the evidence against any possible prejudice through misidentification, we hold that the identification in both cases is admissible evidence. A.R.E. 403.
XII. THE CONSTITUTIONAL RIGHT TO CONFRONTATION
We are equally satisfied that the substantial corroboration supporting S.J.’s identification of Contreras and Hall’s identification of Grumbles serves to eliminate any confrontation problem. The United States Supreme Court has held that hearsay may be admitted against a defendant without violating the confrontation clause when a two-tier test is met. U.S. Const, amend. VI. Ohio v. Roberts, 448 U.S. 56, 65-66, 100 S.Ct. 2531, 2538-2539, 65 L.Ed.2d 597, 607-08 (1980). First, the declarant must be unavailable; second, the hearsay statements must bear adequate “indicia of reliability.” Here, S.J. and Hall are indisputably available for purposes of cross-examination on all issues except identification. Construing the record most favorably to Contreras and Grumbles, S.J. and Hall are unavailable on the issue of identification only to the extent that their demeanor might have been affected by the hypnotic sessions. Cross-examination must be fully allowed to inquire into any conflict between their testimony and other evidence, including any prior statements they may have made. Finally, to the extent that they are unavailable on this limited issue, and their in-court identification testimony is the conceptual equivalent of hearsay, it bears adequate indicia of reliability given Contreras’ admissions to his girlfriend, Grumbles’ in-culpatory statements, and the fact that both Contreras and Grumbles were found in possession of their victims’ property. We therefore find no violation of the confronta*820tion clause. See Van Hatten v. State, 666 P.2d 1047, 1051-54 (Alaska App.1983).25
XIII. SUGGESTIVE IDENTIFICATION
Contreras and Grumbles make two additional closely related arguments. First, they contend that the hypnotic sessions involving S.J. and Hall were so unnecessarily suggestive and so conducive to irreparably mistaken identifications that they were denied due process of law. See Stovall v. Denno, 388 U.S. 293, 301-02, 87 S.Ct. 1967, 1972, 18 L.Ed.2d 1199, 1206 (1967). In Howe v. State, 611 P.2d 16, 17-19 (Alaska 1980), the Alaska Supreme Court discussed an analogous situation:
[E]vidence of suggestive pre-trial identification procedures has not been subject to strict exclusionary rules either in the United States Supreme Court or in this court. Suggestiveness alone does not require exclusion. The test is whether, under the totality of the circumstances, the identification is reliable. In Manson v. Brathwaite, 432 U.S. 98, 114, 97 S.Ct. 2243, 2252, 53 L.Ed.2d 140, 154 (1977), the Supreme Court held that “reliability is the linchpin in determining the admissibility of identification testimony ... . ” Recently, in Holden v. State, 602 P.2d 452 (Alaska 1979), we followed Manson and its suggested criteria for determining reliability.
Id. at 18 (footnote omitted). The criteria alluded to by the Howe court include:
the opportunity of the witness to view the criminal at the time of the crime, the witness’ degree of attention, the accuracy of his prior description of the criminal, the level of certainty demonstrated at the confrontation, and the time between the crime and the confrontation.
Id. at 18 n. 2 (quoting Holden v. State, 602 P.2d 452, 456 (Alaska 1979), and Manson v. Brathwaite, 432 U.S. 98, 114, 97 S.Ct. 2243, 2252, 53 L.Ed.2d 140, 154 (1977)). See also Neil v. Biggers, 409 U.S. 188, 200, 93 S.Ct. 375, 382, 34 L.Ed.2d 401, 411 (1972).
Judge Buckalew applied this test in Grumbles’ case and concluded, first, that the hypnotic session was not “unduly suggestive,” and second, that Hall’s identification of Grumbles was reliable. The record supports this conclusion. In Contreras’ case Judge Serdahely applied a per se rule of exclusion and, consequently, did not consider the suggestiveness of the hypnotic interview. We have examined the record and conclude as a matter of law that S.J.’s identification was sufficiently reliable to warrant its introduction. Her credibility and the weight her identification testimony should receive are jury questions. The record reflects that S.J. spent a substantial period of time with her assailant, that she observed him at close range, and that she was confident about her identification. We recognize that the hypnosis itself might have influenced her confidence. Nevertheless, viewing the totality of the circumstances, including E.L.’s independent identification of Contreras and the substantial evidence corroborating the identification, we find that S.J.’s identification was sufficiently reliable to warrant its admissibility.
XIV. FAILURE TO PRESERVE EVIDENCE
Contreras and Grumbles also argue that the state destroyed material evidence favorable to the defense and therefore denied them due process of law. They argue that, because hypnosis permanently alters a witness’s demeanor, meaningful cross-examination was effectively precluded. In addition the appellants contend that the failure of the police to videotape the hypnotic interviews in their entirety deprived them of evidence necessary to challenge the respective credibility of S.J. and Hall at trial by showing police suggestions and inducements to confabulation. Cf. Catlett v. State, 585 P.2d 553, 556-58 (Alaska 1978); Torres v. *821State, 519 P.2d 788, 793-97 (Alaska 1974). Investigator Parmeter’s initial discussions with S.J. and Hall and any dialogue occurring after S.J. and Hall were brought out of their “trances” were not recorded. In addition, the videotape machine failed during the entire hypnotic interview Parmeter conducted with Hall.
The defendants’ argument has two facets. The first involves a defendant’s right to confront opposing witnesses. U.S. Const, amend. VI. We have previously discussed that clause and rejected the argument that it applies to the facts of the cases before us. Our discussion is equally applicable here. The second facet of the argument rests on due process protections. See U.S. Const. amend. XIV, § 1; Alaska Const. art. 1, § 7. In Torres v. State, the Alaska Supreme Court adopted the following test from a federal decision:
[I]t is the law in this circuit that the due process requirement applies to all evidence which ‘might have led the jury to entertain a reasonable doubt about [defendant’s] guilt,’ and that this test is to be applied generously to the accused when there is ‘substantial room for doubt’ as to what effect disclosure might have had.
519 P.2d at 795 (footnote omitted) (quoting United States v. Bryant, 439 F.2d 642, 648 (D.C.Cir.1971).
Alaska cases discussing the prosecution’s failure to preserve evidence require a finding that the evidence would have affected the outcome before a sanction, including suppression of evidence, is warranted. Where a specific rule or court decision requires preservation of certain evidence, and the defendant makes a timely request for that evidence but it is missing, prejudice to the defendant will be presumed unless the state sustains the burden of proving that it acted in good faith and that the evidence, if preserved, would not have helped the defendant. If, after considering the whole record, the court is in doubt regarding prejudice to the defendant or the prosecution’s good faith, then a sanction is warranted. See Putnam v. State, 629 P.2d 35, 44 n. 18 (Alaska 1980). Where no specific rule or court decision requires the prosecution to preserve the evidence in question, however, the defendant bears the burden of proving that, if preserved, the evidence would have been exculpatory, see Alaska R.Crim.P. 16(b)(3), and that the prosecution should have recognized that it might be exculpatory and preserved it. See Nicholson v. State, 570 P.2d 1058, 1064 (Alaska 1977). Similarly, where there is no timely request for the evidence, the defendant must establish the materiality of the missing evidence. See Maloney v. State, 667 P.2d 1258, 1263-67 (Alaska App.1983); Carman v. State, 658 P.2d 131, 139-40 (Alaska App.1983).
Contreras and Grumbles made timely requests for videotapes of the hypnotic interviews with S.J. and Hall. No rule or court decision specifically requires the police to tape interviews with eyewitnesses. Any tapes made, however, are witness statements which must be made available to the defendant. Alaska R.Crim.P. 16(b)(l)(i). Contreras and Grumbles therefore bore the burden of proving that more complete recordings would have been exculpatory, i.e., that they could lead a jury to entertain a reasonable doubt about their guilt. See Torres v. State, 519 P.2d 788, 795 (Alaska 1974). Where the missing evidence is the only evidence bearing upon the issue in question, the defendant sustains his burden if he shows that the subject matter of the evidence addresses the issues in the case. Lauderdale v. State, 548 P.2d 376, 381 (Alaska 1976). Where the record contains other evidence addressing the same subject, however, the likelihood that the missing evidence would have been favorable to the defendant must be determined from a review of the total record. See Putnam v. State, 629 P.2d 35, 43 (Alaska 1980). Thus, the question is: would better records of the hypnotic sessions create a greater possibility that a jury would entertain a reasonable doubt regarding the identifications? See United States v. Agurs, 427 U.S. 97, 107-110, 96 S.Ct. 2392, 2399-2400, 49 L.Ed.2d 342, 352-53 (1976). Given the substantial *822evidence corroborating the identifications in question, which will be presented to the jury, we are satisfied that no due process violation has occurred. See Catlett v. State, 585 P.2d 553, 557-58 (Alaska 1978).26
In summary, we are satisfied that the various contentions of Contreras and Grumbles all present essentially the same question: do the probative qualities of the identifications in these cases sufficiently outweigh the risks of inaccuracy through hypnotic distortion so that their value is greater than the possibility of misidentification? A.R.E. 403. Having thoroughly reviewed the records in both cases, we are satisfied, as a matter of law, that the probative value of the eyewitness identifications outweighs any possible prejudice from misidentification and that the jury should be permitted to hear that evidence.
Judge Buckalew provisionally held that Grumbles could not introduce expert testimony purporting to show that hypnotized witnesses are more prone to memory distortion than nonhypnotized witnesses subjected to the same interrogation techniques. See A.R.E. 403; A.R.E. 702. Judge Buckalew reasoned that the evidence would be more prejudicial than probative since Hall’s identification was substantially corroborated and the testimony would unduly prolong the trial and confuse the issues. Judge Buckalew also relied on a number of cases rejecting expert psychological testimony pointing out problems with the distortion of the memories of eyewitnesses in general. See United States v. Amaral, 488 F.2d 1148, 1152-53 (9th Cir.1973). But see State v. Chapple, 135 Ariz. 281, 660 P.2d 1208, 1218-24 (1983).
A trial judge clearly has the authority to permit expert testimony from a knowledgeable psychologist or other qualified expert on eyewitness susceptibility to memory distortion, whether or not hypnosis has occurred, if the testimony is based on information not generally understood by lay people sitting on juries. See A.R.E. 401; A.R.E. 702. See also State v. Chapple, 660 P.2d at 1220-21. Where the evidence regarding an identification is weak and uncorroborated or sharply disputed, it may be an abuse of discretion to exclude such testimony. The same rules should apply where pre-trial hypnosis is used to refresh or enhance a witness’s recollection and the defendant presents a qualified expert prepared to testify, subject to cross-examination, that hypnosis increases the risk of a mis-identification.
Grumbles did not question Judge Bucka-lew’s ruling in his petition for review, though he did mention it critically in his subsequent briefs. We assume that Judge Buckalew would be willing to reconsider his ruling, if after hearing the state’s case-in-chief, it appears that such testimony might be helpful to the jury in evaluating Hall’s identification. Such testimony might take on added relevance if other evidence introduced by Grumbles in support of his theory of the case independently casts doubt on the accuracy of Hall’s identification.27
*823The judgment of the superior court in the Contreras case is REVERSED and the case is REMANDED for further proceedings consistent with this decision.
The judgment of the superior court in the Grumbles case is AFFIRMED.28
. We note that AS 11.41.220 was not in effect at the time of the alleged offense. The indictment charged that Contreras “did unlawfully and recklessly place S.A.J. in fear of imminent serious physical injury by means of a dangerous instrument....” Under the statutes existing at the time of the offense, this would have described reckless endangerment, AS 11.41.-250.
. Alaska Rule of Evidence 402 provides:
Relevant Evidence Admissible — Exceptions — Irrelevant Evidence Inadmissible.
All relevant evidence is admissible, except as otherwise provided by the Constitution of the United States or of this state, by enactments of the Alaska Legislature, by these rules, or by other rules adopted by the Alaska Supreme Court. Evidence which is not relevant is not admissible.
. Alaska Rule of Evidence 401 provides:
Definition of Relevant Evidence.
Relevant evidence means evidence having any tendency to make the existence of any fact that is of consequence to the determination of the action more probable or less probable than it would be without the evidence.
. Alaska Rule of Evidence 403 provides:
Exclusion of Relevant Evidence on Grounds of Prejudice, Confusion, or Waste of Time.
Although relevant, evidence may be excluded if its probative value is outweighed by the danger of unfair prejudice, confusion of the issues, or misleading the jury, or by considerations of undue delay, waste of time, or needless presentation of cumulative evidence.
. Alaska Rule of Evidence 601 provides:
Competency of Witnesses.
A person is competent to be a witness unless the court finds that (1) the proposed witness is incapable of expressing himself concerning the matter so as to be understood by the court and jury either directly or through interpretation by one who can understand him, or (2) the proposed witness is incapable of understanding the duty of a witness to tell the truth.
In the official commentary to this rule, Professor Saltzburg suggests that:
The policy of [A.R.E. 601] “is that matters of the witness’s opportunity for perception, knowledge, memory, experience and the like go to the weight to be given to his testimony rather than to his right to testify.” Commissioner’s Note to Uniform Rule 17 (1953). But the rule recognizes that some witnesses should not be permitted to appear before the trier of fact because their testimony is entitled to no consideration.
E.R.C. 151.
. Suggestion is defined as “the act or process of impressing something ... upon the mind of another” and “a means or process of influencing attitudes and behavior hypnotically.” Webster’s New International Dictionary, 2286 (3d ed. 1966).
. Confabulation is defined as, “a filling in of gaps in memory by free fabrication.” Id. at 475. This phenomenon has also been called refabrication. See infra note 13.
. The Sixth Amendment to the United States Constitution provides in part: “In all criminal prosecutions, the accused shall enjoy the right ... to be confronted with the witnesses against him .... ” See also Alaska Const, art. I, § 11.
. In so holding, we recognize that a particular hypnotic session may be so suggestive that a subsequent identification may be inadmissible because its probative value is outweighed by the danger of unfair prejudice. A.R.E. 403.
Where a defendant learns that a witness has been subjected to pretrial hypnosis or otherwise subjected to suggestive procedures raising a substantial risk of memory distortion, he should file a pre-trial motion to suppress the witness’s testimony supported by affidavits setting out the facts upon which he relies. The state may file counter-affidavits, and if material facts are in dispute the court may hold an evidentiary hearing. In deciding the matter the trial court should consider the extent to which the danger of suggestion or confabulation exists and issue appropriate orders. In reaching its conclusions the trial court should consider the Ome standards, see infra note 18, the Neil v. Biggers test, see infra Part XIII of this decision, and the extent to which the witness’s testimony is corroborated by other evidence. Ultimately, the question will be: Is the risk that the testimony reflects memory distortion sufficiently great that the probative value of the testimony is destroyed?
. Because of the diversity of issues addressed in the reported cases it is important to stress the limited nature of our holding. We are not presented with the issue of the desirability of hypnotizing witnesses to refresh their recollection. We thus express no opinion on the efficacy of hypnotism for this purpose. S.J. and Hall were hypnotized before their cases came to trial. The issue is whether that hypnotism renders them incompetent to testify.
Second, our attention is focused on the effect of hypnotism, not on the effect of interrogation techniques as such. “Hypnotism,” as that term was defined in the evidentiary hearing in Contreras is: a process of induction followed by the subject’s belief that she is in a trance. We are thus not called upon to consider any particular investigative techniques or interview procedures which might serve to distort the memory of someone whether or not she is “hypnotized” as the trial courts defined hypnotism.
We recognize that a number of the cases addressing these issues approach them from a slightly different perspective. They view the case sub judice purely as a vehicle for announcing a prophylactic rule of general application. To reach a conclusion these cases ask whether pretrial hypnotism of prospective eyewitnesses to refresh recollection is a good or bad practice as a general rule. They conclude that hypnotism has not been shown to invariably produce accurate information that could not be obtained through normal investigation and interrogation, that any otherwise unattainable information produced is therefore probably false, that hypnotism does not guarantee the truthfulness of a witness’s testimony, and that there is at least a risk that hypnotism might distort a witness’s memory to a greater extent than other means of interrogation and investigation.
*797These conclusions motivate these courts to establish a “per se” rule barring testimony by witnesses hypnotized prior to trial. To the limited extent that the concurring opinion adopts a per se rule, it reflects that view. In the abstract, we express no opinion about such a rule. Unfortunately, the cases before us are not solely vehicles for the announcement of general rules of evidence. They are equally concerned with whether Contreras sexually assaulted S.J. and E.L. and whether Grumbles burglarized Hall’s residence and assaulted her.
The Alaska Supreme Court has adopted evidence rules of general application. In Alaska Rule of Evidence 402, the court expressed a preference that changes in the rules of evidence, unless they were mandated by a statute or constitutional provision, come through rule revision rather than common law court decision. A reporter has been hired and committees formed to periodically upgrade those rules in response to developments unforeseen at the time the rules were originally drafted. The evidence rules committee could take testimony at leisure and perhaps resolve some of the factual questions raised in this opinion regarding the assumptions upon which those courts adopting a per se rule rely. Perhaps those courts are correct in their assumptions. If they are, an evidence rule which has only prospective effect would appear to be the proper vehicle for generating a rule of general application. Such a rule might also address the questions of eyewitness testimony in general: the relative desirability of giving cautionary instructions and the possibilities of permitting expert testimony or excluding evidence where problems arise. See United States v. Telfaire, 469 F.2d 552 (D.C.Cir.1972). See also E. Loftus, Eyewitness Testimony, 187-203 (1979). Loftus discusses the relative value of four solutions to the problem of unreliable eyewitness testimony: (1) exclusion, (2) corroboration, (3) cautionary instructions, and (4) expert testimony. Loftus favors the last alternative which would permit experts to testify regarding factors which may make the testimony unreliable. The evidence rules committee might productively consider these various alternatives in adopting rules governing eyewitness testimony of those subjected to interrogation and investigatory procedures likely to distort their memories whether or not the eyewitnesses were previously hypnotized. The supreme court could amend the rules on the basis of such a report. The committee could also suggest modifications in the recently promulgated from jury instructions for criminal cases.
. See Davis, An Approach to Problems of Evidence in the Administrative Process, 55 Harv.L. Rev. 364, 402-10 (1942). See also, Davis, Judicial Notice, 55 Colum.L.Rev. 945, 952 (1955):
Stated in other terms, the adjudicative facts are those to which the law is applied in the process of adjudication. They are the facts that normally go to the jury in a jury case. They relate to the parties, their activities, their properties, their businesses. Legislative facts are those which help the tribunal to determine the content of law and policy and to exercise its judgment or discretion in determining what course of action to take. Legislative facts are ordinarily general and do not concern the immediate parties. In the great mass of cases decided by courts and by agencies, the legislative element is either absent, unimportant, or interstitial, because in most cases the applicable law and policy have been previously established. But whenever a tribunal is engaged in the creation of law or of policy, it may need to resort to legislative facts, whether or not those facts have been developed on the record.
(quoted in State v. Erickson, 574 P.2d 1, 4 n. 14 (Alaska 1978)).
. Glanville Williams describes the problem as follows:
The general characteristic of these episodes [of mistaken identification leading to conviction of the innocent] was a naive belief by the police, judges and juries in the accuracy of visual identification, even by a single witness. It was in vain that the defendant gave evidence of an alibi: such a defence, even if supported by witnesses, generally had no power to countervail a witness’s identification. The court took no account of discrepancies between the descriptions first given by witnesses and the person ultimately charged, or of the fact that forensic science evidence was negative (when if the defendant had been the offender some scientific confirmation might have been expected). When the police had evidence of identification they sometimes did not bother to check alibi witnesses named by the suspect — even when the offence charged was an ordinary driving of-fence and the suspect furnished the details of the alibi immediately upon being interviewed by the police.
Williams, Evidence of Identification: The Dev-lin Report, Crim.L.Rev. 407 (1976) (footnote omitted). Williams continues his discussion:
The defects in the identification evidence in these cases [wrongful convictions based on faulty eyewitness testimony] may be briefly catalogued. Amongst the identifications accepted as reliable were (1) a hesitating identification by one witness after the defendant had been on parade before 15 witnesses without being picked out by one of them; (2) identification by a witness who picked out the wrong man on one parade and walked past the defendant twice on the second parade before picking him out; (3) identification by one man three months after the incident; (4) identification by a girl who had seen the offender only by the light of the moon; (5) in a case where the offender had grey hair, identification of the defendant, a man with nearly white hair, in a parade where all the other 11 men had dark hair. In one case, where a member of the Metropolitan police attempted to intercede for a neighbour of unblemished character who was charged by the City of London police on implausible identification evidence, the result was that he himself was put on parade, “identified,” and immediately suspended from duty as being under suspicion of being concerned in the same offence! Fortunately the whole prosecution was afterwards withdrawn, but the Commissioner of the City of London police said that he had no reason to think that there was any justifiable criticism of any of his officers concerned in the case. Subsequently, a disciplinary inquiry was made into the affair but the report was not published. The mulishness of the police in this case is paralleled by the conduct of prosecuting counsel in another case who tried to uphold on appeal a conviction based on identification evidence although another man had subsequently pleaded guilty to the same crime and had made a full statement that completely exonerated the defendant. One is somewhat relieved to note that the Court of Appeal described the conduct of counsel as “utterly wrong.”
Id. at 408 (footnotes omitted).
. See E. Loftus, Memory at 37 (1980):
Memory is imperfect. This is because we often do not see things accurately in the first place. But even if we take in a reasonably accurate picture of some experience, it does not necessarily stay perfectly intact in memory. Another force is at work. The memory traces can actually undergo distortion. With the passage of time, with proper motivation, with the introduction of special kinds of interfering facts, the memory traces seem sometimes to change or become transformed. These distortions can be quite frightening, for they can cause us to have memories of things that never happened. Even in the most intelligent among us is memory thus malleable.
See also E. Loftus, Eyewitness Testimony, supra, at 111.
Dr. Loftus describes the process of distortion as follows:
*801The same sorts of distortions occur when people’s memories are tested in a controlled laboratory situation. A classic example is the case where subjects were shown an illustration of several people on a subway car, including a black man with a hat and a white man with a razor in his hand. Using the method of serial reproduction where one person describes the picture to another, who describes it to another — much like the childhood game of “telephone” — two investigators found that the razor tended to migrate in memory from the white man to the black man. One person reported, “This is a subway train in New York headed for Portland Street. There is a Jewish woman and a Negro who has a razor in his hand. The woman has a baby or a dog. The train is going to Dyer Street and nothing much happened.” How does someone create so detailed an image? And why do people remember the razor in the hands of the black man? In this case, sterotypes are affecting what they see and remember.
When we try to remember something that happened to us, these sorts of “constructive” errors are common. We can usually recall a few facts, and using these facts we construct other facts that probably happened. We make inferences. From these probable inferences, we are led to other “false facts” that might — or might not — have been true. To paraphrase C.S. Morgan, we fill up the lowlands of our memories from the highlands of our imaginations. This process of using inferences and probable facts to fiil in the gaps of our memories has been called “refabrication,” and it probably occurs in nearly all of our everyday preceptions. We supply these bits and pieces, largely unconsciously, to round out fairly incomplete knowledge.
We fill in gaps in our memory using chains of events that are logically acceptable. There are so many real-life illustrations of this to choose from. For example, a taxi traveling on a busy street makes an emergency stop. A passenger in the taxi sees that a red Buick in front has stopped abruptly, its right-hand door is open, and an elderly man is lying unconscious on the street. The passenger assumes that the man either fell or was thrown out the door of the Buick and remembers seeing this happen. In fact, the driver of the Buick had stepped on his brakes suddenly so as to avoid hitting the older man who had wandered into the intersection. The collision was unavoidable and the man was knocked to the ground. The only thing the passenger actually saw was the unconscious man lying on the ground and the open door of the Buick. These fragments were then integrated into a logical sequence and a new “memory” was created. Our biases, expectations, and past knowledge are all used in the filling-in process, leading to distortions in what we remember.
E. Loftus, Memory, supra, at 39-40 (footnote omitted; emphasis added). Finally, she concludes:
The malleability of human memory represents a phenomenon that is at once perplexing and vexing. It means that our past might not be exactly as we remember it. The very nature of truth and of certainty is shaken. It is more comfortable for us to believe that somewhere within our brain, however well hidden, rests a bedrock of memory that absolutely corresponds with events that have passed. Unfortunately, we are simply not designed that way.
Id. at 190.
. Dr. Barber uses the following example to illustrate his point:
There is another way of viewing responsiveness to test suggestions that does not involve special state constructs such as “hypnosis,” “hypnotized,” “hypnotic state,” or “trance.” This alternative paradigm does not see a qualitative difference in the “state” of the person who is and the one who is not responsive to test suggestions. Although the alternative paradigm has many historical roots (discussed by Sarbin, 1962), it derives primarily from my more recent theoretical endeavors and those of Sarbin .... An analogy to members of an audience watching a motion picture or a stage play may clarify the paradigm.
One member of an audience may be attending a performance with the purpose of having new experiences. His attitude is that it is interesting and worthwhile to feel sad, to feel happy, to empathize, and to have the other thoughts, feelings, and emotions the actors are attempting to communicate. He both desires and expects the actors to arouse in him new or interesting thoughts and emotions. Although he is aware that he is watching a contrived performance and that he is in an audience, he does not actively think about these matters. Since this mem*804ber of the audience has “positive” attitudes, motivations, and expectancies toward the communications emanating from the stage, he lets himself imagine and think with the statements and actions of the actors; he laughs, weeps, empathizes and, more generally, thinks, feels, emotes, and experiences in line with the intentions of the actors.
Another member of the audience had an anxious and tiring day at the office, wanted to go to bed early in the evening, and unwillingly came to the performance in order to avoid an argument with his wife. He is not interested in having the emotions and experiences the actors are attempting to communicate. He does not especially desire and does not expect to feel empathic, happy, sad, excited, or shocked. He is continually aware that he is in an audience and that he is observing a deliberately contrived performance. Given this set of attitudes, motivations, and expectancies, this member of the audience does not let himself imagine and think with the statements and actions of the actors; he does not laugh, weep, empathize, or, more generally, think, feel, emote, and experience in line with the communications from the actors.
The implications of this analogy are:
1. The experimental subject who is highly responsive to test suggestions resembles the member of the audience who experiences the thoughts, feelings, and emotions that the actors are attempting to arouse. The very suggestible subject views his responding to test suggestions as interesting and worthwhile; he desires and expects to experience those things that are suggested. Given these underlying “positive” attitudes, motivations, and expectancies, he lets himself imagine and think with the things suggested and he experiences the suggested effects.
2. The experimental subject who is very unresponsive to test suggestions resembles the member of the audience who does not experience the thoughts, feelings, and emotions that the actors are attempting to arouse. The very nonsuggestible subject views his responding to test suggestions as not desirable; he neither wants nor expects to experience those things that are suggested. Given these underlying “negative” attitudes, motivations, and expectancies, he does not let himself imagine and think with the things suggested and he does not experience the suggested effects.
3. It is misleading and unparsimonious, to label the member of an audience who is thinking, feeling, and emoting in line with the communications of the actors, as being in a special state (hypnotic trance) that is fundamentally different from the waking state. In other words, it is misleading, and unparsimo-nious, to restrict our conceptions of normal conditions or waking conditions to such an extent that they exclude the member of the audience who is having various experiences as he listens to the communications from the stage. Furthermore, since the member of the audience, who is responding to the words of the actors, and the experimental subject, who is responding to the words (test suggestions) of the experimenter, do not differ in any important way in their attitudes, motivations, and expectations toward the communications or in the way they think along with the communications, it is also misleading and unpar-simonious to label the subject who is responding to test suggestions as being in a special state (hypnotic trance).
4.Although the member of the audience who is responding to the words of the actors and the experimental subject who is responding to the test suggestions of the experimenter have similar attitudes, motivations, and expectations toward the communications and are similarly “thinking with” the communications, they are being exposed to different types of communications. The messages or communications from the actors are intended to elicit certain types of thoughts, feelings, and emotions — to empathize, to feel happy or sad, to laugh or to weep, to feel excited or shocked — whereas the messages or communications (test suggestions) from the experimenter are intended to elicit somewhat different types of thoughts, feelings, or emotions — to experience an arm as heavy, to experience oneself as a child, to forget preceding events, and so forth. From this viewpoint, the member of the audience and the subject who is responding to test suggestions are having different experiences, not because they are in different “states” but because they are receiving different communications.
The above analogy exemplifies some of the basic assumptions underlying the alternative paradigm. These assumptions include the following:
1. It is unnecessary to postulate a fundamental difference in the “state” of the person who is and the one who is not responsive to test suggestions.
2. Both the person who is and the one who is not responsive to test suggestions have attitudes, motivations, and expectancies toward the communications they are receiving.
3. The person who is very responsive to test suggestions has “positive” attitudes, motivations, and expectancies toward the communications he is receiving. That is, he views his responding to test suggestions as interesting or worthwhile and he wants to, tries to, and expects to experience the sug*805gested effects. Given these “positive” attitudes, motivations, and expectations, he lets himself think with and imagine those things that are suggested.
4. The person who is very unresponsive to test suggestions has “negative” attitudes, motivations, and expectations toward the communications he is receiving. That is, he views his responding to test suggestions as not interesting or worthwhile and he neither tries to nor expects to experience the suggested effects. Given these “negative” attitudes, motivations, and expectations, he does not let himself imagine or think with the suggestions; instead, he verbalizes to himself such statements as, “This is silly,” or, “The suggestion won’t work.”
5. The three factors — attitudes, motivations, and expectations — vary on a continuum (from negative, to neutral, to positive) and they converge and interact in complex ways to determine to what extent a subject will let himself think with and imagine those things that are suggested. The extent to which the subject thinks with and vividly imagines the suggested effects, in turn determines his overt and subjective responses to test suggestions.
6. Concepts derived from abnormal psychology — such as “trance,” “somnambulism,” and “dissociation” — are misleading and do not explain the overt and subjective responses. Responsiveness to test suggestions is a normal psychological phenomenon that can be conceptualized in terms of constructs that are an integral part of normal psychology, especially of social psychology. Social psychology conceptualizes other social influence processes, such as persuasion and conformity, in terms of such mediating variables as attitudes, motivations, expectancies, and cognitive processes. In the same way, the mediating variables that are relevant to explaining responsiveness to test suggestions include attitudes, motivations, expectations, and cognitive-imaginative processes.
7. The phenomena associated with test suggestions are considered to be within the range of normal human capabilities. However, whether or not the suggested phenomena are similar to or different from phenomena occurring in real life situations that bear the same name, is viewed as an open question that needs to be answered empirically. For example, such questions as the following are open to empirical investigation: To what extent is suggested analgesia similar to the analgesia produced by nerve section or by anesthetic drugs? What are the similarities and differences between suggested and naturally occurring (nonsuggested) blindness, col- or blindness, deafness, hallucinations, dreaming, and amnesia? The empirical evidence at present indicates that, although there are some similarities between the suggested and nonsuggested phenomena, they also differ in very important respects. For example, suggested color blindness has only superficial resemblances to actual color blindness, and suggested amnesia is much more labile or transient than actual amnesia....
Which paradigm is more successful in explaining responsiveness to test suggestions for limb rigidity, age-regression, analgesia, hallucination, amnesia, etc. — the traditional one that postulates a special state that is fundamentally different from the waking state, or the alternative one that focuses on attitudes, motivations, expectancies, and thinking with and imagining those things that are suggested? I will next summarize experimental data, pertaining to responsiveness to test suggestions under control (“waking”) conditions, which indicate that the alternative paradigm provides a more successful and more parsimonious explanation.
T. Barber, Suggested (“Hypnotic”) Behavior: The Trance Paradigm Versus an Alternative Paradigm, reprinted in E. Fromm & R.E. Shor, HYPNOSIS: Developments in Research and New Perspectives, 217, 221-24 (2d ed. 1979) (emphasis in original; footnote and citations omitted).
. The New York Court of Appeals was impressed by the experiments where subjects, under hypnosis as Hilgard and his disciples define it, experienced hypnotic age regression:
These effects have been demonstrated under experimental conditions. In one series of experiments, involving age regression, hypnotic subjects were able to recall, with great detail and convincing manner, events or incidents from their childhood which upon investigation often proved to be totally unfounded. A more dramatic illustration is provided by experiments with age progression in which the hypnotized persons recalled, with equal specificity and conviction, events which “happened” in the future. In these instances the defects in hypnotically induced recollection were easily detected by reference to known facts. When such verification is not available, there is no currently accepted method for scientifically determining the reliability of hypnotically induced recollection.
People v. Hughes, 59 N.Y.2d 523, 466 N.Y.S.2d 255, 260-61, 453 N.E.2d 484, 489-90 (1983) (footnotes omitted). What the court overlooks is the fact that Barber has achieved the same results, i.e., subjects recalling events from their childhood in great detail and in a convincing *806manner and professing absolute confidence in the details recalled, without utilizing an induction procedure or postulating an altered state of consciousness. Of course he had to persuade the subject that age regression was possible. We are satisfied that the average investigator will have little difficulty in persuading the average rape victim that recall of her assailant’s appearance is at least possible. Barber concludes his discussion of age regression as follows:
Although various theoretical formulations might possibly account for the foregoing data, one formulation that can parsimoniously explain the results is as follows: When it is suggested to “hypnotic trance” subjects or to “awake” subjects that they are in the past (or in the future), (a) some “hypnotic trance” subjects and also some “awake” subjects try to the best of their ability to think about continuously and to imagine vividly that they are in the past (or in the future), (b) some of the “hypnotic trance” subjects and also some of the “awake” subjects succeed in focusing imaginatively on the past (or future), and (c) when thinking about and vividly imagining themselves in an earlier time (or in a future time), some subjects in both groups feel as if they are in the past (or in the future) and tend to behave to a certain limited degree as if they are in the past (or future).
T. Barber, Suggested (“Hypnotic”) Behavior: The Trance Paradigm Versus an Alternative Paradigm, reprinted in E. Fromm and R.E. Shor, HYPNOSIS: Developments in Research and New Perspectives, 217, 242.
. The theories of Drs. Orne, Hilgard, Barber and Sarbin (who is mentioned in our quotations from Barber’s work) and a number of other theorists are discussed and critiqued in P. Shee-han & C. Perry, Methodologies of Hypnosis: A Critical Appraisal of Contemporary Paradigms of Hypnosis (1976).
. Most of the cases adopting a per se rule of exclusion of eyewitness testimony allegedly refreshed by pre-trial hypnosis rely on Dr. Bernard Diamond for the proposition that hypno*807tism significantly increases the risk that a witness’s memory will be distorted. Dr. Diamond does not support his views on this point with any scientific evidence. See infra note 22. The per se cases also implicitly assume the truth of the Hilgard model without discussing the Barber model or considering its implications if a per se rule is adopted for criminal litigation.
. Dr. Orne describes and justifies the standards as follows:
The use of hypnosis and related techniques to facilitate memory raises profound, com*809plex questions, and it is likely that the individual will be protected only if these issues are dealt with at the highest level of our court system. There are instances when hypnosis can be used appropriately provided that the nature of the phenomenon is understood by all parties concerned. It must be recognized, however, that the use of hypnosis by either the prosecution or the defense can profoundly affect the individual’s subsequent testimony. Since these changes are not reversible, if individuals are to be allowed to testify after having undergone hypnosis to aid their memory, a minimum number of safeguards are absolutely essential. Based upon extensive review of the field and my own experiences in a considerable number of circumstances, I have proposed the following minimal safeguards in an affidavit [filed in a case pending before] the Supreme Court of the United States.
1. Hypnosis should be carried out by a psychiatrist or psychologist with special training in its use. He should not be informed about the facts of the case verbally; rather, he should receive a written memorandum outlining whatever facts he is to know, carefully avoiding any other communication which might affect his opinion. Thus, his beliefs and possible bias can be evaluated. It is extremely undersirable to have the individual conducting the hypnotic sessions to have any involvement in the investigation of the case. Further, he should be an independent professional not responsible to the prosecution or the investigators.
2. All contact of the psychiatrist or psychologist with the individual to be hypnotized should be videotaped from the moment they meet until the entire interaction is completed. The casual comments which are passed before or after hypnosis are every bit as important to get on tape as the hypnotic session itself. (It is possible to give suggestions prior to the induction of hypnosis which will act as post-hypnotic suggestions.)
Prior to the induction of hypnosis, a brief evaluation of the patient should be carried out and the psychiatrist or psychologist should then elicit a detailed description of the facts as the witness or victim remembers them. This is important because individuals often are able to recall a good deal more while talking to a psychiatrist or psychologist than when they are with an investigator, and it is important to have a record of what the witness’s beliefs are before hypnosis. Only after this has been completed should the hypnotic session be initiated. The psychiatrist or psychologist should strive to avoid adding any new elements to the witness’s description of his experience, including those which he had discussed in his wake state, lest he inadvertently alter the nature of the witness’s memories — or constrain them by reminding him of his waking memories.
3. No one other than the psychiatrist or psychologist and the individual to be hypnotized should be present in the room before and during the hypnotic session. This is important because it is all too easy for observers to inadvertently communicate to the subject what they expect, what they are startled by, or what they are disappointed by. If either the prosecution or the defense wish to observe the hypnotic session, they may do so without jeopardizing the integrity of the session through a one-way screen or on a television monitor.
4. Because the interactions which have preceded the hypnotic session may well have a profound effect on the sessions themselves, tape recordings of prior interrogations are important to document that a witness had not been implicitly or explicitly cued pertaining to certain information which might then be reported for apparently the first time by the witness during hypnosis.
Ome, The Use and Misuse of Hypnosis in Court, 27 Int’l J. Clinical & Experimental Hypnosis No. 4, at 335-36 (1979) (footnotes omitted; citations omitted). The New York Court of Appeals and the Supreme Court of California criticize the Orne standards, alleging that the standards only address the risk of suggestion and do not deal with the risk of confabulation, i.e., the risk that the witness will incorporate impressions from her other experiences and imagination into her memory as a result of hypnotism. See People v. Shirley, 31 Cal.3d 18, 181 Cal.Rptr. 243, 641 P.2d 775, 786-87 (1982) cert denied - U.S. -, 103 S.Ct. 13, 73 L.Ed.2d 1400 (1982); People v. Hughes, 59 N.Y.2d 523, 466 N.Y.S.2d 255, 265-66, 453 N.E.2d 484, 494 — 95 (1983). This criticism is unfounded. As we have seen, confabulation occurs when (1) someone without any perception of an event and therefore no memory to recall is (2) placed under pressure to recall a memory by someone she wishes to please. The Orne standards address the second requisite for confabulation by attempting to minimize the pressure on the witness to have memories. It should be noted that the test developed by the United States Supreme Court in Neil v. Biggers, 409 U.S. 188, 199, 93 S.Ct. 375, 382, 34 L.Ed.2d 401, 411 (1972), addresses the first requisite, i.e., the absence of a memory. The test requires verification of an identification subject to a suggestive lineup or showup by considering;
the opportunity of the witness to view the criminal at the time of the crime, the witness’ degree of attention, the accuracy of the witness’ prior description of the criminal, the level of certainty demonstrated by the witness at the confrontation, and the length of time between the crime and the confrontation.
This test, if satisfied, will help assure that the witness had a memory to recall regarding her identification of her alleged attacker. Thus the *810Ome standards considered together with the Neil test would provide protection against both suggestion and confabulation.
Finally, corroboration of an identification through discovery of incriminating evidence in the possession of the person identified would not rule out “suggestion” because the police may have known of such evidence before hypnotizing the witness. It would rule out confabulation, however, since it is highly unlikely that the victim would “imagine” the true assailant in the absence of some external reason for suspecting him.
.Dr. Martin Reiser has a doctorate in educational psychology. He is the director of the Law Enforcement Hypnosis Institute in Los Angeles, California, and he is a consultant to the Los Angeles Police Department. Dr. Reiser is a leading proponent of hypnosis in criminal investigation. He has written a training manual to teach police to use hypnosis as a tool. See M. Reiser, Handbook of Investigative Hypnosis (1980). Dr. Reiser validated the procedures used by Investigator Parmeter in hypnotizing S.J. Dr. Reiser’s views have been sharply criticized. See Book Review, Int’l J. Clinical & Experimental Hypnosis No. 4, 443 (1982) (C. Perry and J. Laurence criticizing Reiser’s Handbook).
. Dr. Diamond is a professor of law at the University of California, Berkeley. He is also a clinical professor of psychiatry at the University of California Medical Center in San Francisco. He has authored a number of articles relating psychiatry to the law. See, e.g., Diamond, Inherent Problems in the Use of Pretrial Hypnosis on a Prospective Witness, 68 Calif.L.Rev. 313 (1980) (frequently quoted article criticising hypnosis of witnesses as an aid to investigation).
. Dr. Diamond uses the term “source amnesia” as follows: After a witness to a crime or accident has been hypnotized to refresh her recollection, she may remember what was discussed during hypnosis. However, the witness may be unable to differentiate between what she actually remembers of the incident and what she was told during hypnosis, what she read in a newspaper, or what she simply imagined.
. Most of the authorities that adopt a per se rule rendering incompetent any witness who has been hypnotized rely on Dr. Diamond’s article, Inherent Problems in the Use of Pretrial Hypnosis on a Prospective Witness, 63 Calif.L. Rev. 313 (1980). In that article, Dr. Diamond contends that hypnotized witnesses are more prone to distortion than nonhypnotized witnesses. His entire treatment of the issue is as follows:
Are the uncertainties of ordinary eyewitness testimony and those of hypnotically enhanced recall sufficiently similar that the legal rules of procedures designed for coping with the one are sufficient for the other?
Recent research has conclusively demonstrated that conventional eyewitness testimony is fraught with grave problems of inaccuracy and distortion. 121 People quite commonly fail to have seen things that happened and claim to have seen things that did not happen. Particularly, the key words used by the interrogator exert a powerful influence on the memory of the subject, and the distortions introduced by such suggestion may last indefinitely, with the subject unaware that his memory has been so altered.122 Even Bernheim, nearly 100 years ago, knew that the distortions of suggestion were not limited to subjects in a state of hypnosis, and he described cases of unhypnotized witnesses succumbing to suggestion and producing imaginary evidence.123
Nevertheless, the hypnotic subject is a great deal more vulnerable to suggestion than is the normal person, and the hypnotic distortions persist into the posthypnotic period with much greater force. The accepted view, that the law cannot exclude the usual eyewitness testimony where there may have been some unreliable distortions, should not justify the admission of testimony that is known to have been subject to the inevitable distortions of the hypnotic process. The present legal attitude expressed in most case law on hypnosis is analogous to a claim that we should not even attempt to eliminate the disease of smallpox because chicken pox is so common and cannot be controlled.
I am in complete agreement with the following statement made by a public defender in the brief for a recent unpublished case:
There is no practice in which false testimony is more “apt to harden” than hypnotic “memory enhancement”.... [T]he state of hypnosis is one of high suggestibility for the willing subject. The tendency for such a subject to confabulate, or manufacture a memory, even with very strong subjective conviction, is inherent in the hypnotic technique .... The real danger is that if the subject confabulates, and the recalled memory is “eminently plausible,” then it is virtually impossible for even a skilled therapist to detect such deception; and the process is irreversible.124
Pretrial hypnosis of a witness cannot be considered a harmless form of “coaching” or legitimate preparation of the witness for the courtroom experience. In some respects it is worse than ordinary subornation of a witness for it accomplishes the same effect, yet allows the perpetrator, the witness, and the trier of fact to remain unaware that perversion of the evidence has occurred.
. See Woocher, Did Your Eyes Deceive You? Expert Psychological Testimony on the UnreUability of Eyewitness Identiñcation, 29 Stan.L.Rev. 969 (1977).
. Fishman & Loftus, [Fishman & Loftus, Expert Psychological Testimony on Eyewitness Identification, 4 L. & Psych.Rev. 87 (1973).]
. H. Bernheim, [H. Bernheim, Suggestive Therapeutics: A Treatise on the Nature and *812Uses of Hypnotism 111 (trans. from 2d rev. French ed. 1886).]
. Brief for Appellant at 29, People v. Jones, No. 67183 (Super.Ct. Alameda County, California), quoted in T. Worthington, [T. Worthington, Hypnotically Induced or Enhanced Testimony 18, 65 (unpublished manuscript on file with the California Law Review).]
Id. at 341-42.
It is significant that in a law review article replete with citations to the psychiatric and psychological literature, Dr. Diamond’s only authority for this highly important point was a quotation from a brief filed by a public defender seeking to have a witness’s testimony ruled incompetent.
Dr. Diamond testified in the Contreras hearing substantially after publishing the article in question. At that time he conceded that he had not done a comparative analysis of hypnotized and unhypnotized witnesses to verify the proposition that the former were more subject to memory distortion through suggestion and confabulation than the latter.
. Dr. Rossi is the director of the behavioral science section of the Michigan Department of State Police. Dr. Rossi has a Ph.D. in clinical psychology. He maintains a private practice in clinical psychology and serves on the faculty of the American Society of Clinical Hypnosis. In People v. Gonzales, 108 Mich.App. 145, 310 N.W.2d 306, 309 n. 3 (1982), the Michigan Court of Appeals questioned Rossi’s independence because of his association with the state police. We have considered this factor.
. Lest we be misunderstood, we wish to make it clear that we are referring to Dr. Diamond’s testimony regarding the effect of hypnotism on memory distortion in eyewitnesses, not his view about appropriate rules of law. Both are minority views. The Frye test, requiring general acceptance, applies to scientific principles not legal principles.
. In reaching this conclusion, we note the close similarity between “indicia of reliability” for purposes of confrontation and the reliability of an identification under Neil v. Biggers, 409 U.S. 188, 196-200, 93 S.Ct. 375, 380-382, 34 L.Ed.2d 401, 409-11 (1972) (discussed infra section XIII).
. While we decline to adopt a rule regulating preservation of evidence of hypnotic interviews at this time, in subsequent cases in which there is expert evidence directly addressing the concerns we voice in this opinion we might be led to consider adopting standards similar to those of Dr. Orne. The state, to the extent it intends to rely in the future on the testimony of witnesses hypnotized before trial, should carefully consider the risk that stringent rules may well be adopted and take care to insure that evidence of the hypnotic interview is available if needed.
We are not prepared at this time to decide exactly what evidence the state should preserve in the event that the state elects to have a witness hypnotized. However, we believe that it makes sense to have a full statement from a witness before the witness undergoes hypnosis. It also appears that the entire hypnotic session should be recorded, on videotape if possible. This procedure would allow any expert witnesses, counsel, the court and jurors to know what the witness’s statement was before hypnosis, and to compare that statement to the witness’s statements after hypnosis. The hypnotic session itself could also be examined to gauge what effect the hypnosis might have had on the witness’s testimony.
. The New York Court of Appeals suggests that a jury instruction might minimize the risk of prejudice. People v. Hughes, 59 N.Y.2d 523, 466 N.Y.S.2d 255, 268, 453 N.E.2d 484, 497 (1983).
. The per se rule advocated by the concurrence avoids some of the problems attending the per se rules criticized in this opinion. It nevertheless is subject to the criticism that it disqualifies a class of people from testifying before it is known who occupies the class. It is unclear whether Chief Judge Bryner’s proposal applies only to those who have been “hypnotized” as Hilgard defines the term or whether it extends as well to those who meet Barber’s definition. We believe that A.R.E. 402 suggests that any per se rule governing the admission of evidence must be mandated either by statute, constitution, or rule of evidence. The rule refsion procedure provides an appropriate remedy if a special rule governing witnesses subjected to pretrial hypnosis is warranted. We therefore decline to adopt any per se rule however qualified. We concede an inability to distinguish between our interpretation of A.R.E. 403 as it would apply to these cases and the application of the Chief Judge’s rule because: (1) both require case by case application, (2) both require that the trial court consider the same criteria, and (3) it is unlikely that any testimony that would pass muster under A.R.E. 403 would be excluded under Chief Judge Bryner’s per se rule.