Defendant was convicted of a burglary and theft at the home of Mrs. Barth. At trial, Mrs. Barth was unable to identify defendant, but a police officer testified that she had selected defendant’s picture from a set of seven photographs shown to her seven months after the burglary. On appeal, defendant contended that the trial court should have suppressed this evidence because the circumstances of the pretrial photographic identification were so suggestive as to destroy its reliability. The Court of Appeals rejected defendant’s argument for suppression under Manson v. Brathwaite, 432 US 98 (1977), and affirmed the conviction, Schwab, C.J., dissenting. 31 Or App 683, 571 P2d 527 (1977). We allowed review.
I.
Before turning to the evidence, we review the applicable law. As far as the constraints of "due process” imposed by the federal 14th amendment are concerned, the state of the law may be summarized as follows.
In a trilogy of cases decided in 1967, the United States Supreme Court recognized that the process by which a witness is given an opportunity to identify a suspect represents a "critical stage” in a criminal prosecution. United States v. Wade, 388 US 218 (1967); Gilbert v. California, 388 US 263 (1967); Stovall v. Denno, 388 US 293 (1967).1 While Wade and Gilbert involved the protection of the presence of defense counsel at this "critical stage,” Stovall stated that evidence of a pretrial identification must be excluded if, under "the totality of the circumstances surrounding it,” the identification procedure was "so unnecessarily suggestive and conducive to irreparable mistaken identification” that it fell short of due process of law. 388 US at 302. Consistent with the premise that *224the "process” which must be "due” includes the "critical” pretrial procedures of investigation and prosecution, this formulation focused on the question whether the circumstances of the identification needlessly created a risk of misidentification, not whether a misidentification had occurred.
The following year, the Supreme Court applied the Stovall formulation to a case of pretrial identification by photograph. Simmons v. United States, 390 US 377 (1968). However, after having found the particular procedure justified under the circumstances of the case, the Court added that there was "little chance that the procedure utilized led to misidentification of Simmons,” and that subsequent testimony left "little room for doubt that the identification of Simmons was correct.” 390 US at 385-386. These phrases in Simmons could be read to add a "harmless error” qualification to the assessment of the original identification procedure; even though that procedure fell short of due process, it would not require reversal if the ability of the witnesses to identify defendant on other occasions was plainly independent of the suggestiveness of the challenged procedure. But apart from such independent confirmation, the legitimacy of the initial identification process by Stovall standards would still be judged as of the time when this "critical stage” in the prosecution was conducted.
Neil v. Biggers, 409 US 188 (1972), further shifted the focus from the propriety of the initial procedure to the reliability of the identification. The part of the Stovall formulation that called for considering the "totality of the circumstances” in testing the identification procedure was transferred to testing whether the identification was reliable despite the needless use of a suggestive procedure. 409 US at 199. BiggershaA arisen before the decision in Stovall, which the Supreme Court had declined to apply retroactively. But in Manson v. Brathwaite, 432 US 98 (1977), the Court applied also to post-Stovall prosecutions the *225shift from the premise that the "critical stage” of investigatory identification must comply with due process to the position that due process is satisfied if a court can find that the proffered identification "possesses certain features of reliability.” The stated reason, adopting an explanation offered by Judge Leventhal of the U.S. Court of Appeals, was that the exclusion of an improperly obtained identification is required only to protect the defendant’s "evidentiary interest,” that is to say, his right not to be convicted on unreliable evidence partly created by the prosecution itself, rather than to enforce observance of a legal right standing on its own feet, such as the right against an illegal arrest, search, or seizure.2 The "features of reliability” for federal purposes, in turn, were to be sought in scrutinizing each case "on its own facts,” Simmons v. United States, 390 US at 384, for the relative importance of five "factors” bearing on the probable reliability of the identification. Neil v. Biggers, 409 US at 199; Manson v. Brathwaite;432 US at 114. We consider those factors below.
The Court of Appeals noted that as a result of this ad hoc approach to the admissibility of identification evidence "every case is now of constitutional significance” and expressed some doubt whether an instruction to consider each case "on its own facts” states much of a constitutional standard. Real as these concerns are, they put the cart before the horse. The Supreme Court of the United States pronounces rules governing the admissibility of evidence in the federal courts as a matter of federal decisional law in the absence of statute, or in applying a statute or rule. See, e.g., Weeks v. United States, 232 US 383 (1913) (excluding evidence seized in violation of fourth amendment); Nardone v. United States, 302 US 379 (1937), 308 US 338 (1939) (excluding the fruits of wiretaps under the Federal Communications Act); Mallory v. United States, 354 US 449 (1957) (excluding *226admissions made while in custody in violation of Rule 5 of FRCP). Thus in Simmons itself, the Court necessarily considered the identification procedure under its "supervisory authority” before reaching a constitutional issue. 390 US at 386. In this respect, it occupies the same position as this court. See, e.g., State v. Fairley, 282 Or 689, 580 P2d 179 (1978); State v. Flores, 280 Or 273,570 P2d 965 (1977); State v. Valdez, 277 Or 621, 561 P2d 1006 (1977); State v. Laundy, 103 Or 443, 204 P 958, 206 P 290 (1922). But the Supreme Court does not purport to make the law of evidence for the states. The Court’s decisions under the 14th amendment only pronounce constitutional tests which a state’s rules of evidence, and their application in a particular case, may not fail; but these decisions assume that there is an applicable state rule in advance of the issue of its constitutionality. The rules governing the admissibility of evidence in state courts are the responsibility of the states before a Supreme Court decision and remain so afterwards, within the constitutional limits laid down in the decision. See Cooper v. California, 386 US 58, 62 (1967); Johnson v. New Jersey, 384 US 719, 733 (1966).
Evidence law has long provided for excluding certain evidence as a class when its questionable reliability vitiates the value of its possible truthfulness in the particular case, apart from any question of constitutional law. One familiar example is the exclusion of coerced confessions; others are the exclusion of evidence obtained by polygraph tests, hypnosis, or "truth drugs.” See State v. Wintzingerode, 9 Or 153 (1881) (confessions obtained by promises or by threats); State v. Green, 271 Or 153, 164-171, 531 P2d 245 (1975) (polygraph evidence); People v. Harper, 111 111 App 2d 204, 250 NE2d 5 (1969) (sodium amobarbital); Jones v. State, 542 P2d 1316 (Okla Crim App 1975) (hypnosis). These rules of evidence are not dependent on federal law. It is necessary, therefore, to consider application of their principle to the use of the challenged identification evidence.
*227The rule must take into account, on the one hand, the importance of eyewitness testimony in identifying (or exculpating) a suspected offender and, on the other hand, the widely recognized risk that such identification may often be unreliable at best and at worst may be the psychological product of the identification procedure itself. The extensive research and commentary by psychologists and jurists on the dangers of misidentification and ways to minimize them stretches back at least half a century. Courts as well as commentators have been concerned that the risk of mistake is greater when a witness identifies a suspect from a photograph than in person. Some years before the federal cases, Justice Traynor of the California Supreme Court noted that "identification from a still photograph is substantially less reliable than identification of an individual seen in person.” People v. Gould, 54 Cal2d 621, 354 P2d 865 (1960), citing 3 Wigmore, Evidence §786a (3d ed 1940) 164-166. In People v. Anderson, 389 Mich 155,186, 205 NW2d 461 (1973), the Supreme Court of Michigan, upon a review of the experience of Michigan courts as well as the research literature, concluded that "eyewitness identification through photographs is at least as hazardous as corporeal identification and probably is more hazardous to the securing of correct identifications,” and that identification by photograph should normally not be used when the suspect is in custody and available for personal identification procedures. The opinion appends extensive quotations from legal and psychological sources that we shall not reproduce here. See 205 NW2d at 479-494. The Supreme Court of Louisiana similarly held that in view of the greater reliability of a legally conducted lineup, "after a suspect is in police custody, there is seldom any justification for employing photo-identification.” State v. Wallace, 285 S2d 796, 801 (La 1973).
At the same time, as noted by the Council of Judges of the National Council on Crime and Delinquency, when standards for pre-arraignment practices in the *228investigation of crime and the identification of suspects must be sought in the "piecemeal decisions” of appellate courts in an intermittent series of unique factual disputes, they provide neither clear nor effective guidance to the law enforcement officials responsible for the investigatory processes.3 In its decisions on lineup procedure, the Supreme Court stressed the *229advantages of "[^legislative or other regulations, such as those of local police departments, which eliminate the risks of abuse and unintentional suggestions” and disclaimed any intention to place the details of identification procedure in a "constitutional strait jacket.” United States v. Wade, 388 US at 239. 4 As Justice Stevens wrote in Manson,
the arguments in favor of fashioning new rules to minimize the danger of convicting the innocent on the basis of unreliable eyewitness testimony carry substantial force. Nevertheless, . . . this rulemaking function can be performed "more effectively by the legislative process than by a somewhat clumsy judicial fiat,” . . . and . . . the Federal Constitution does not foreclose experimentation by the States in the development of such rules.
432 US at 117-118 (concurring opinion). Lineup procedure is in fact often prescribed in police regulations, at least in large cities. See Read, Lawyers at Lineups: Constitutional Necessity or Avoidable Extravagance?, 17 UCLA L Rev 339 (1969); Note, Protection of the Accused at Police Lineups, 6 Colum J L Soc Prob 345, 360-373 (1970) (citing many sources). A recent set of model rules published by the Police Foundation covers *230both confrontation and photographic procedures. The model rules provide that photographic identification "is permissible only when a live identification procedure is impractical,” rarely after the suspect’s arrest, and describes how it should be conducted.5 Another systematic effort to state procedures designed to facilitate the important role of eye-witness identification while minimizing the risks of suggestiveness and mistake is the American Law Institute’s Code of Pre-Arraignment Procedure.6 The commentary to the ALI code states:
*231The Code does not include provisions dealing with pre-trial identification simply as a response to the Supreme Court’s past and present decisions in this area. Whatever the present state of the law in the Court, the fact remains that a wide variety of experienced persons consider and have considered the pre-trial identification as a crucial factor in the fair and accurate determination of guilt or innocence, and a factor as to which certain kinds of error, once committed, are particularly hard to remedy and peculiarly likely to lead to unjust results.
The premise of the code is that
where policies exist in respect to the conduct of law enforcement agencies at the pre-arraignment stage, and guidance — in the form of statutes and regulations — can be given to law enforcement, then such policies should be formulated and guidance given in an explicit form. Plainly many of the due process questions relating to pre-trial identifications that might arise in a criminal prosecution could be avoided or at least mitigated if the police in the conduct of their investigations had a clear mandate regarding *232how they are required to proceed, preferably a mandate which they had some part in formulating. . . .
Id. at 422, 423.
Although we agree with the United States Supreme Court’s statement in Wade, supra, about the value of clear procedural rules, of course we do not hold that any version of the foregoing specific procedures is mandated in the absence of statute. We cite these sources to show that the unreliability of eyewitness identification under suggestive circumstances is widely recognized, and that the procedures used to minimize this unreliability bear on the admissibility of evidence of such identification. As a practical matter, in the context of a motion by a defendant to suppress identification evidence on the ground that it is the product of a suggestive procedure, the decision on its admissibility involves two steps. First, the court must determine whether the process leading to the offered identification was suggestive or needlessly7 departed from procedures prescribed to avoid such suggestiveness. If so, then the prosecution must satisfy the court that "the proffered identification has a source independent of the suggestive confrontation” or photographic display, see Commonwealth v. Botelho, 343 NE2d 876, 881 (Mass 1976) (citing cases), or that other aspects of the identification at the time it was made substantially exclude the risk that it resulted from the suggestive procedure.
It is in making this second determination that facts of the kind mentioned by the Supreme Court in Manson v. Brathwaite are relevant. These include the opportunity that the witness had at the time to get a clear view of the persons involved in the crime and the attention he or she gave to their identifying features, the timing and completeness of the description given *233by the witness after the event, the certainty expressed by the witness in that description and in making the subsequent identification, and, of course, the lapse of time between the original observation and the subsequent identification. These are not to be taken as a mechanical checklist of "constitutional” facts. Obviously other facts may also be important, such as the age and sensory acuity of the witness, see State v. Bush, 29 Or App 315, 563 P2d 747 (1977), or a special occupational concern with people’s appearance or physical features, or the frequency of his or her contacts with individuals sharing the general characteristics of the person identified, see Manson v. Brathwaite, 432 US at 115. Listing these and other relevant inquiries must not distract attention from the ultimate issue whether an identification made in a suggestive procedure has nevertheless been demonstrated to be reliable despite that suggestiveness.8 Of course, when the evidence survives this judicial scrutiny, its reliability and probative force under the circumstances of the identification remain questions for the jury-
II
The Court of Appeals summarized the facts of the crime as follows:
On May 15, 1975, the complaining witness hired two men, one black and one white, who had come to her home soliciting yard work. After the two men had worked for a few minutes while their employer remained with them working and answering questions, the black man asked to use the toilet. He went into the house, and the complainant remained outside with the other man. After several minutes she became suspicious and went into the house to find out why the black man had not returned. He was gone. She quickly discovered that several valuable items were missing. She then rushed back outside and found that the white man had also disappeared. *234Shortly after the incident, complainant gave descriptions of the two men to a police officer. . . .
In his written report of the interview on May 15, the officer, Deputy Sheriff Detloff of Clackamas County, wrote that Mrs. Barth described the white man as being five feet four inches tall, slender, weighing 150 pounds, with short light brown hair and a Vandyke beard. The photographic identification at issue here took place on a visit to Mrs. Barth’s home by another officer, Detective Christensen of the Multnomah County Sheriff’s office, some seven months later, on or about December 18, 1975. A hearing on defendant’s motion to suppress the evidence of this photographic identification was held on June 21, 1976, six months after the challenged identification and 13 months after the crime and Officer Detloffs initial interview with Mrs. Barth. At this hearing the following evidence was taken relevant to the asserted suggestiveness of the identification procedure and to the reliability of the identification despite any such suggestiveness.
Detective Christensen testified that he first heard of the May 15 episode from Mrs. Barth in August, 1975, when he visited her in connection with a different burglary. Mrs. Barth’s description led him to suspect that the black man involved might be one Richard Fuller. Either on that occasion or in December Christensen showed Mrs. Barth one or two pictures of Fuller. She positively identified him as the black participant in the May 15 crime. In December, Christensen showed Mrs. Barth seven "mug shots” of white men. All had mustaches. One had additional facial hair along the jaw from temples to chin which in Christensen’s view was something less than a "beard.” Christensen had not read the earlier report of the Clackamas County deputy concerning Mrs. Barth’s description of the two men involved. He did not himself ask Mrs. Barth to describe them nor describe to her the height or weight of the men whose faces appeared in the pictures.
*235The officer testified that he told Mrs. Barth that "I suspected that the white subject with the Negro subject from May 15 was in this group.” Mrs. Barth first segregated two of the seven pictures and, after further study, selected the picture with the additional facial hair.9
Mrs. Barth’s recollections at the suppression hearing diverged somewhat from Christensen’s. She testified that she was shown more than seven pictures, including several of Negroes, and that the photographs presented at the hearing were not those shown her at her home six months earlier. She recalled narrowing the choice to two of the white men before selecting one. Mrs. Barth also recalled having been told that the pictures included the men who had come to her home the previous May.
We conclude that two elements of the process that led to her selection of defendant’s photograph must be characterized as suggestive. One is the statement to the witness that the suspect is one of the persons among whose pictures the witness is to choose. As the Supreme Court of Indiana has stated:
A witness may thus be lead to feel that he has an obligation to choose one of the participants in the display since the police evidently are satisfied that they have apprehended the criminal. The result may be that the witness strains to pick someone with familiar characteristics or someone who most resembles the actual criminal or the result may be that the witness will choose the one least dissimilar by the process of elimination. . . .
*236Sawyer v. State, 260 Ind 597, 602, 298 NE2d 440, 443 (1973). This has been held to make admission of the identification evidence reversible error in some cases. See State v. Wallace, supra. The second suggestive element is the fact that only one of the pictures showed a man with facial hair beyond a mustache. In her report of the incident, Mrs. Barth had made a particular point of the white man’s "Vandyke” beard, and without getting into speculation whether the picture Mrs. Barth chose showed a "beard” or "four or five days’ growth,” as Mr. Christensen thought, an examination of the photographs shows that it alone displayed this feature that was important to her recollection of a man seen seven months earlier.
As stated above, these suggestive aspects would not require the court to suppress the evidence that Mrs. Barth had identified defendant’s photograph if the prosecution was able to negate any substantial risk that this identification was stimulated by the suggestive procedure. The prosecution did not do so at the suppression hearing. Mrs. Barth’s testimony at the hearing was brief and was not directed to the kind of showing of reliability in the face of a suggestive procedure discussed above. On the record made at the suppression hearing, it was error not to suppress the evidence of the photographic identification.
However, although the evidence should have been suppressed, the question remains whether its use at trial was reversible error. In concluding that it was not, the Court of Appeals took into account testimony at the trial itself as bearing on the question whether Mrs. Barth had reliably identified defendant’s photograph independently of the suggestive manner in which it was shown to her.10 The court stressed Mrs. *237Barth’s extended opportunity to observe the two men when they came to work for her, in daylight, under conditions when she would give her attention to them, and without distracting elements of fear or stress.11 We agree that her opportunity to form an impression of the men was better than in most reported cases in which a subsequent identification is challenged, and it would lend reliability to any otherwise positive identification she might have made shortly after the events, either by confrontation or, if the suspect was not in custody, by recognizing a photograph. However, the difficulty arises because Mrs. Barth did not see the photographs until seven months after the only time she had seen the suspects. On her own testimony, her identification of defendant’s photograph at that time cannot be described as very positive. She had already been told that the officer believed the white suspect to be one of the group of pictures he showed her. She expressed doubts because she remembered the "Vandyke beard,” and only one of the pictures had facial hair beyond a mustache. As Chief Judge Schwab wrote in dissent below, under these circumstances it is not surprising that Mrs. Barth selected defendant’s picture.
*238Actually, Mrs. Barth testified that she selected two pictures and said "it’s got to be one of these two.” When asked whether her memory was better at the time she saw the photographs than it was six months later at the hearing, Mrs. Barth answered: "Not about most things, but at the time I just didn’t pay that much attention. I didn’t even think of it coming to trial. I didn’t pay that much attention.”
On this record, we cannot conclude that Mrs. Barth would have identified one of the photographs out of seven if several of them had shown men with similar beards and if she had been told that the police did not know whether the man she had seen was among those in the set of photographs, that is to say, in the absence of the suggestive procedure. The state has not carried that burden. Accordingly, the defendant is entitled to a new trial.
Reversed and remanded.
The quoted phrase is from Wade, 388 US at 224, referring to the need for the assistance of counsel.
432 US 113, citing Clemons v. United States, 408 F2d 1230, 1251 (1968)(concurring opinion).
Council of Judges of the National Council on Crime and Delinquency, Model Rules of Court on Police Action from Arrest to Arraignment 1 (1969). The Council proposed a rule that would allow photographic identification procedures only when a line-up, as specified in the rule, cannot be followed. See proposed Rule 10, id at 17.
The importance of spelling out rules for law enforcement practices below the level of constitutional litigation is also stressed by the American Bar Association’s Project on Standards for Criminal Justice. The ABA Standards Relating to the Urban Police Function (Supp. 1973) include the following:
4.3 Administrative rule-making.
Police discretion can best be structured and controlled through the process of administrative rule-making by police agencies. Police administrators should, therefore, give the highest priority to the formulation of administrative rules governing the exercise of discretion, particularly in the areas of selective enforcement, investigative techniques, and enforcement methods.
4.4 Contribution by legislatures and courts.
To stimulate the development of appropriate administrative guidance and control over police discretion, legislatures and courts should actively encourage police administrative rule-making.
(a) Legislatures can meet this need by delegating administrative rule-making responsibility to the police by statute.
(b) Courts can stimulate administrative development in several ways including the following:
(i) Properly-developed and published police administrative policies should be sustained unless demonstrated to be unconstitutional, arbitrary, or otherwise outside the authority of the police;
(ii) To stimulate timely and adequate administrative policy-making, a determination by a court of a violation of an administrative policy should not be a basis for excluding evidence in a criminal case unless the violation of administrative policy is of constitutional dimensions or is otherwise so serious as to call for the exercise of the superintending authority of the court. A violation per se should not result in civil liability; and
(iii) Where it appears to the court that an individual officer has acted in violation of administrative policy or that an administrative policy is unconstitutional, arbitrary, or otherwise outside the authority of the police, the court should arrange for the police administrator to be informed of this fact, in order to *229facilitate fulfillment by the police administrator of his responsibility in such circumstances to reexamine the relevant policy or policies and to review methods of training, communication of policy, and supervision and control.
5.4 Need for administrative sanctions and procedures.
In order to strengthen administrative review and control, responsibility should formally be delegated to the police for developing comprehensive administrative policies and rules governing the duties and responsibilities of police officers together with procedures and sanctions for ensuring that these duties and responsibilities are met ....
Id. at 8-11.
The Court quoted proposals for such safeguards from Wigmore, The Science of Judicial Proof 541 (3d ed 1937) and Murray, The Criminal Lineup at Home and Abroad, 1966 Utah L Rev 610.388 US at 236, n. 26, id at 239, n. 30. For an example of a statutory rule in Oregon, see ORS 136.675 —136.695, requiring recording and other safeguards in obtaining testimony under hypnosis.
See Project on Law Enforcement Policy and Rulemaking, College of Law, Arizona State University and The Police Foundation (1974), Model Rules for Law Enforcement: Eyewitness Identification, reprinted in American Law Institute, A Model Code of Pre-Arraignment Procedure at 657 (1975). Rules 301-302 provide:
Rule 301. When Photographs May Be Used. The use of photographs to identify criminal suspects is permissible only when a live identification procedure is impractical. Example: A live identification procedure would be impractical where there is no suspect; there is a suspect but probable cause to arrest him is absent; the suspect cannot be found; a suspect or witness refuses to cooperate; or the suspect is in custody some distance from the prospective viewer.
Rule 302. Saving Witnesses To View Lineup. Whenever a witness makes a positive identification from a photograph, and probable cause to arrest the suspect is thereby established, photographs should not be displayed to other witnesses if they will later view the suspect at a lineup.
Subsequent rules provide that photographic displays may be used after the arrest of the suspect only when there are specified reasons why a lineup is impractical, but an adequate record must be made of the photographic display, and that the presence of defense counsel is not a prerequisite for an admissible pictorial identification. Rules 305-401. ALI Code of PreArraignment Procedure at 661-663.
The Code, drafted in the form of a model statute, includes the following section relevant to this case:
Section 160.2. Conduct of Identification Procedures
(1) Personal Identifications: Line-ups, Confrontations, Unknown Suspects. The identification of a suspect by having a witness view the suspect shall be from an array of several persons, a reasonable number of whom are similar in dress and appearance to the person to be identified, except such line-up or other array is not required in any of the following circumstances:
(a) Confrontation promptly after commission of a crime. A law enforcement officer arranges a confrontation by a witness to a *231crime promptly following the commission of that crime, unless there is reason to believe that a line-up or other procedure is necessary to make the identification reliable.
(b) Suspect unknown or at large. The identity of the suspect is unknown, or the suspect is at large and his presence cannot practicably be secured pursuant to Article 170 or otherwise.
(c) Consent of counsel. Counsel for the suspect consents to such other identification procedure as is carried out.
(2) Photographic Identifications. If the identification is to be made from photographs, drawings or other representations, the witness shall be presented with an array of representations of several persons. Such array shall so far as practicable include a reasonable number of persons similar to any person then suspected whose likeness is included in the array.
(4) Multiple Identifcation Procedures . . .
(5) Avoidance of Suggestion. Any identification procedure pursuant to subsections (1) through (4) including any multiple identification procedure, shall be conducted to avoid so far as practicable any suggestion to a witness at or prior to the procedure that a particular person is the person suspected of involvement in the crime under investigation.
Id. at 76-78.
Whether a questioned identification procedure was necessitated by the circumstances or needless does not affect its suggestiveness, as far as evidence law is concerned, but is relevant if a due process issue is reached under Simmons, Biggers, and Manson, supra.
Compare ALI Code of Pre-Arraignment Procedure, §160.7, cited in Commonwealth v. Botelho, 343 NE2d at 885. See also United States v. Sanders, 479 F2d 1193 (DC Cir 1973).
Christensen also testified that, although defendant was in custody in Portland at the time and represented by counsel, he did not consider asking Mrs. Barth to identify him in a lineup procedure. He explained that he never used used lineup identification, that the facility at Rocky Butte jail was antiquated and rarely used, and that the "biggest hassle and one I try to avoid is getting all attorneys there present at the same time.” That is, of course, not a proper reason for circumventing the protection of the presence of counsel mandated by Wade and Gilbert, supra. See, e.g., People v. Anderson, supra, n. 3. If the photographic identifcation was suggestive, its substitution for a lineup bears on whether it was needlessly so for constitutional purposes under Simmons, Biggers, Manson, supra. However, this substitution does not bear on the issue whether the procedure was suggestive.
In this case, there was the additional complication that there were two trials, the first having ended in a mistrial. A second motion to suppress was denied on the basis of the first suppression hearing and the first trial without taking new evidence. At the second trial, defendant entered no objection to the introduction of evidence for the purpose of rehabilitating the identification, so we need not consider whether evidence not brought out at the suppression hearing may be introduced at trial for that purpose.
The last mentioned element illustrates the need for caution in judicial ventures into psychology. As Chief Judge Schwab, dissenting below, pointed out, sometimes the stress of confrontation with a criminal has been cited as malting for a high degree of attention and impressing the defendant’s picture on the victim’s memory. 31 Or App at 699. Cf. State v. Strickland, 36 Or App 119,123,584 P2d 310 (1978); Neil v. Biggers, 409 US at 200-201. By contrast, Mrs. Barth did not know a crime was in progress (though she testified that she did grow suspicious) and had no special reason to give her attention to the men’s appearance.
It is important not to turn the facts of specific reported cases into "precedents” to be cited for generalizations of the judicially expected behavior of witnesses. Probably the physical conditions of the observation —light or darkness, the presence or absence of a disguise or of unusual identifying characteristics, speech or silence, distance and elapsed time of observation — and the explanation given by the witness for his or her own certainty or tentativeness about the identification in the specific instance will matter more than the assumptions of appellate courts about the effect of calm or crisis on witnesses generally. However, this also emphasizes the value of express trial court findings of the relevant historical facts, insofar as they are not undisputed. See State v. Warner, 284 Or 147,158-160, 585 P2d 681 (1978).