this court held that a testifying witness’s out-of-court identification “that cannot be confirmed by an identification [of the defendant] at the trial is insufficient to sustain a conviction in the absence of other evidence tending to connect the defendant with the crime.”
(Id.
at p. 631.) In this case, the Attorney General challenges that holding, contending the requirement is illogical, lacks support in the law, is unsatisfactory as a matter of policy, and is contrary to a statute later enacted by the Legislature (Evid. Code, § 411). For the reasons set forth below, we conclude that we should overrule
Gould’s
holding that an out-of-court identification is in all cases insufficient by itself to sustain a conviction. Instead, the sufficiency of an out-of-court identification to support a conviction should be determined under the substantial evidence test of
People
v.
Johnson
(1980) 26 Cal.3d 557, 578 [162 Cal.Rptr. 431, 606 P.2d 738, 16 A.L.R.4th 1255] that is used to determine the sufficiency of other forms of evidence to support a conviction.
I
Approximately 3 a.m. on July 19, 1992, two gunmen approached an outdoor party of the “Jeffrey Street” gang in an alley in Anaheim, Orange
[258]
County, and shot Vincente Garcia. One bullet hit Garcia in the left arm; another hit his right testicle and lodged in his thigh. Among those present during the shooting were Hugo Guzman, a Jeffrey Street gang member; Oscar Gomez, who considered Jeffrey Street’s members to be his “home boys”; and Carlos Rodriguez, a friend of Jeffrey Street’s members but not a member himself. After Garcia was shot, Rodriguez drove him to the hospital. Doctors removed Garcia’s injured testicle.
Immediately after the shooting, Gomez, Guzman, and others left in Gomez’s car. As they drove off, a passenger in the car was hit by a thrown bottle. Gomez drove his injured passenger to a hospital; there he discovered two fresh bullet holes in his car. Anaheim Police Officer John Orr happened to be at the hospital and separately interviewed Gomez, Guzman, and Gomez’s passengers. Gomez told Officer Orr that two gunmen had walked up to the party and started shooting. Gomez gave a physical description of the two gunmen, one of whom he described as a male Hispanic, age twenty-five, with a heavy build, five feet seven inches tall, wearing a white-and-blue striped shirt, and with a thick mustache and short dark hair.
Gomez went outside to show Officer Orr the bullet holes in his car. Once outside and away from his friends, Gomez whispered to Officer Orr that the shooter was a “West Side Anaheim” gang member known as “Beto,” and that he could identify Beto if he saw Beto again.
When Officer Orr interviewed Guzman at the hospital, Guzman described the gunman as a male Hispanic, age 25, with a heavy build., wearing a light-colored long-sleeved shirt, and having medium-length, slicked-back hair and a mustache. Guzman’s description was consistent with that given by Gomez.
Later, Officer John Kelley interviewed Guzman. At first, Guzman denied being present during the shooting. When the officer reminded Guzman of Guzman’s earlier statements to Officer Orr at the hospital, Guzman replied he could not remember the shooting.
Officer Kelley also interviewed Gomez three days after Garcia was shot. In this interview, Gomez reiterated and expanded his identification of “Beto.” Gomez told Officer Kelley that “Beto” was the shooter. Gomez described “Beto’s” physical appearance as Hispanic, 25 years old, 5 feet 7 inches tall, weighing 175 to 190 pounds, with a bushy mustache and dark medium-length hair that was combed straight back, and wearing a white long-sleeved shirt with blue stripes. Gomez said he became acquainted with “Beto" through a former girlfriend. When Officer Kelley showed Gomez a
[259]
notebook containing 30 to 40 photographs of West Side Anaheim gang members, Gomez selected defendant Robert Munoz Cuevas’s picture as the one depicting “Beto.”
Later, while being interviewed by a defense investigator, Gomez repudiated his earlier identification of defendant as the shooter, claiming it had been motivated by anger at the West Side Anaheim gang for providing evidence against Guzman’s brother, who was charged with murder in an unrelated case.
Defendant was arrested and charged with the attempted murders of Garcia and Gomez and with assault with a firearm on Garcia and Gomez. Defendant was also charged with personally using a firearm in committing these crimes, and with intentionally inflicting great bodily injury on Garcia.
Both Gomez and Guzman testified at defendant’s trial. Gomez recanted his identification of defendant as the gunman. Gomez acknowledged that he knew defendant as “Beto,” that he was previously acquainted with defendant through his former girlfriend, that in his two statements to the police he had identified defendant as the shooter, and that he had selected defendant’s photograph as the gunman he called “Beto." But Gomez claimed he had falsely identified defendant and that he had done so as “pay back” for the West Side Anaheim gang’s role in providing evidence in the (unrelated) murder case against Guzman’s brother. When Guzman testified, he admitted being present during the shooting, but denied having seen any gunman and denied having described the gunman to Officer Orr.
Both Gomez and Guzman testified that they believed it was wrong to “rat off’ a member of a rival gang; that is, to inform the authorities that the gang member has committed a crime. Gary Bushman, a district attorney’s investigator and gang expert, likewise testified that cooperating with police, even if the suspect is in a rival gang, is generally disapproved of in gang culture and that gang members who initially cooperate with police are subject to intimidation to change their testimony at trial.
Officers Kelley and Orr testified that in their separate interviews of Gomez, he had given a physical description of the gunman, whom he identified as “Beto.” In addition, Officer Kelley testified to Gomez’s previous contacts with defendant.
Another witness who had been present at the shooting, Rodriguez, testified that as the gunmen approached immediately before the shooting, Gomez exclaimed: “I know that guy. He’s from West Side Anaheim.” Additional
[260]
evidence was presented regarding defendant’s membership in the West Side Anaheim gang and gang activities in general in Anaheim. Defendant generally met the physical descriptions given by Gomez and Guzman.
Before submission of the case to the jury, defendant moved for a judgment of acquittal (Pen. Code, §1118.1), asserting that because there was no in-court identification of him as the shooter and no other evidence to link him to the crimes, there was insufficient evidence to convict him. The trial court denied the motion, as well as defendant’s subsequent request that the court instruct the jury it could not convict defendant on the basis of an out-of-court identification unless it found the identification was corroborated by other evidence tending to connect defendant to the offense.
The jury convicted defendant of assault with a firearm on Garcia and found that defendant had personally used a firearm in committing the offense. Defendant appealed. The Court of Appeal concluded that the out-of-court statements of witnesses Gomez and Guzman adequately corroborated each other under
Gould, supra,
54 Cal.2d 621, and that therefore the evidence was sufficient to support the conviction. The court held, however, that the trial court committed prejudicial error in not instructing the jury on the need for corroboration, thus requiring a reversal of the judgment.
The Attorney General petitioned this court for review. The Attorney General contends that
Gould, supra,
54 Cal.2d 621, wrongly held that an out-of-court identification by itself cannot be sufficient evidence to prove guilt beyond a reasonable doubt. Alternatively, the Attorney General argues that even if the
Gould
rule is correct, it is one for an appellate court to apply in reviewing the sufficiency of the evidence supporting a conviction, and not one for a jury to apply. We granted review to consider whether to overrule the
Gould
rule, and, if not, whether the jury should be instructed on its application.
II
It is the prosecution’s burden in a criminal case to prove every element of a crime beyond a reasonable doubt.
(In re Winship
(1970) 397 U.S. 358 [25 L.Ed.2d 368, 90 S.Ct. 1068].) To determine whether the prosecution has introduced sufficient evidence to meet this burden, courts apply the “substantial evidence” test. Under this standard, the court “must review the whole record in the light most favorable to the judgment below to determine whether it discloses
substantial
evidence—that is, evidence which is reasonable, credible, and of solid value—such that a reasonable trier of fact could find the defendant guilty beyond a reasonable doubt.”
(People
v.
[261]
Johnson, supra,
26 Cal.3d at p. 578, italics added; see also
Jackson
v.
Virginia
(1979) 443 U.S. 307, 315-319 [61 L.Ed.2d 560, 570-574, 99 S.Ct. 2781].) The focus of the substantial evidence test is on the
whole
record of evidence presented to the trier of fact, rather than on “ ‘isolated bits of evidence.’ ”
(People
v.
Johnson, supra,
at p. 577.) The substantial evidence test applies both when an appellate court is reviewing on appeal the sufficiency of the evidence to support a conviction and when a trial court is deciding the same issue in the context of a motion for acquittal under Penal Code section 1118.1 at the close of evidence.
There are exceptions, however, to the substantial evidence test. The Legislature has determined that because of the reliability questions posed by certain categories of evidence, evidence in those categories by itself is insufficient as a matter of law to support a conviction. For example, the Legislature has required that the testimony of an accomplice (Pen. Code, §1111), and the testimony of a single witness in a perjury case as to the falsity of the defendant’s perjurous statement (Pen. Code, § 118, subd. (b)), must be corroborated before a conviction can be based on them. This court in
Gould, supra,
54 Cal.2d 621, created a similar corroboration requirement for an out-of-court identification that the witness does not confirm at trial.
In
Gould, supra,
54 Cal.2d 621, a burglary victim testified at trial that she was unable to positively identify the two defendants as the burglars of her home. A police officer then testified that the victim had identified the defendants with certainty in a pretrial photo lineup. This court held that the out-of-court statements of the testifying victim identifying the defendants as the burglars were admissible as “independent evidence of identity.”
(Id.
at p. 626.) Thus, under
Gould,
an out-of-court statement of identification is admissible as substantive evidence when “the witness [who made the identification] is available at the trial for cross-examination.”
held in addition, however, that “[a]n extrajudicial identification that cannot be confirmed by an identification at the trial is insufficient to sustain a conviction in the absence of other evidence tending to connect the defendant with the crime.”
(Gould, supra,
54 Cal.2d at p. 631.) The
Gould
corroboration requirement
1
is a judicially created, per se rule that one type of evidence—an out-of-court identification—is by itself insufficient
as a matter of law
to support a conviction. The continuing vitality of this corroboration requirement is at issue here.
[262]
The Attorney General contends that the
Gould
corroboration requirement was abolished when, after this court’s 1960 decision in
Gould, supra,
54 Cal.2d 621, the Legislature in 1965 enacted Evidence Code section 411, which provides:
“Except where additional evidence is required by statute,
the direct evidence of one witness who is entitled to full credit is sufficient for proof of any fact.” (Italics added.) The Attorney General points out that there is no statute requiring that to prove the perpetrator’s identity, an out-of-court identification of the defendant must be supported by other evidence linking the defendant to the crime.
The Attorney General also points out that in
People
v.
Alcala
(1984) 36 Cal.3d 604 [205 Cal.Rptr. 775, 685 P.2d 1126], a case decided after
Gould, supra,
54 Cal.2d 621, this court relied on Evidence Code section 411 in rejecting a claim that there should be a nonstatutory corroboration requirement, similar to the
Gould
corroboration requirement, for the testimony of jailhouse informants.
relying on Evidence Code section 411 to reject proposed cautionary instruction for immunized witness testimony].)
We note, however, that in a decision predating the cases cited in the preceding paragraph, this court concluded that “the drafters of the Evidence Code did not intend to change the
Gould
rule regarding the sufficiency of evidence to sustain a conviction.”
(In re Johnny G.
(1979) 25 Cal.3d 543, 548 [159 Cal.Rptr. 180, 601 P.2d 196].) The court based this conclusion on the Law Revision Commission comment to Evidence Code section 1238. In that comment, the Law Revision Commission stated that, because Evidence Code section 1235 (which makes prior inconsistent statements admissible) and section 1238, subdivision (c) (which makes prior statements of identification admissible if “the witness testifies that he made the identification and that it was a true reflection of his opinion at that time”) address only the admissibility of out-of-court statements of identification, they “have no effect” on the
Gould
corroboration requirement. (Cal. Law Revision Com. com., 29B pt. 4 West’s Ann. Evid. Code, § 1238 (1995 ed.) p. 250.)
Although neither the Law Revision Commission comment to Evidence Code section 1238 nor this court’s decision in
In re Johnny G., supra,
25 Cal.3d 543, 548, specifically addressed the effect, if any, of Evidence Code section 411 on the
Gould
corroboration requirement, there are additional reasons for doubting the validity of the Attorney General’s assertion that
[263]
Evidence Code section 411 was intended to abolish the
Gould
corroboration requirement. First, it is questionable whether an out-of-court identification that the witness does not confirm in court is “the direct evidence of [a] witness” within the meaning of Evidence Code section 411. Because the out-of-court identification is hearsay and the witness is not testifying to the present truth of the identification, the identification is the evidence of an out-of-court declarant, not a witness. Moreover, in
Gould, supra,
54 Cal.2d 621, 629-630, this court characterized hearsay evidence as indirect, not direct, evidence, although this conclusion was based on the pre-Evidence Code statutory definition of indirect evidence in the Code of Civil Procedure’s former section 1832, which has no current counterpart in the Evidence Code.
Second, Evidence Code section 411 was intended to “restate[] the substance of and supersede[] [former] Section 1844 of the Code of Civil Procedure.” (Cal. Law Revision Com. com., 29B pt. 1 West’s Ann. Evid. Code, §411 (1995 ed.) p. 388.) The Code of Civil Procedure’s former section 1844 was the law at the time
Gould
was decided; similar to Evidence Code section 411, it provided that “The direct evidence of one witness who is entitled to full credit is sufficient for proof of any fact except perjury and treason.” Because the
Gould
corroboration requirement coexisted with the Code of Civil Procedure’s former section 1844, the mere restatement of former section 1844 in Evidence Code section 411 would not necessarily abolish the
Gould
corroboration requirement.
Even if, however, consistent with this court’s holding in
In re Johnny G., supra,
25 Cal.3d 543, we were to conclude that Evidence Code section 411 did not abrogate the
Gould
corroboration requirement, we would still face the question of whether to overrule that judicially created rule determining the sufficiency of an out-of-court identification to support a conviction. Because, for the reasons stated below, we conclude that there are compelling reasons for overruling that portion of
Gould
which establishes the corroboration requirement, we need not ultimately resolve what effect Evidence Code section 411 had on that requirement.
III
As discussed above, the holding in
Gould, supra,
54 Cal.2d 621, 631, that an out-of-court identification is insufficient to support a conviction in the absence of other corroborating evidence linking the defendant to the crime is simply a rule for determining the sufficiency of one type of evidence—an out-of-court identification—to support a criminal conviction. The
Gould
corroboration requirement differs, however, from the “substantial
[264]
evidence” test ordinarily applied to determine the sufficiency of evidence to support a criminal conviction. The
Gould
corroboration requirement deems an uncorroborated out-of-court identification insufficient evidence
as a matter of law
to support a conviction regardless of the identification’s probative value or the existence of other evidence in the record. Under the
Gould
corroboration requirement, therefore, a conviction must be reversed on appeal even if “the whole record in the light most favorable to the judgment . . . discloses substantial evidence—that is, evidence which is reasonable, credible, and of solid value—such that a reasonable trier of fact could find the defendant guilty beyond a reasonable doubt.”
(People
v.
Johnson, supra, 26
Cal.3d at p. 578.) Moreover, contrary to the substantial evidence test, the
Gould
corroboration requirement focuses on an “ ‘isolated bit[] of evidence’ ”
(id.
at p. 577)—the out-of-court identification—to the exclusion of other evidence in the record.
Because the
Gould
corroboration requirement conflicts with the substantial evidence test set forth in
People
v.
Johnson, supra, 26
Cal.3d 557, 576-578, it would bear a heavy burden of justification were we deciding for the first time whether to adopt it. (See generally,
People
v.
Towler
(1982) 31 Cal.3d 105, 118 [181 Cal.Rptr. 391, 641 P.2d 1253] [rejecting claim that sufficiency of circumstantial evidence should be subject to a higher standard than the substantial evidence test of
People
v.
Johnson].)
Corroboration requirements are the exception, not the rule. Professor Wigmore has observed “that our whole presumption should be against any specific rule requiring a number of witnesses, or corroboration of a single witness; that such arbitrary measures are likely to be of little real efficacy and to introduce disadvantages greater than those which they purport to avoid; and that therefore any such rule, . . .
must justify itself by experience as overwhelmingly useful and efficacious.”
(7 Wigmore, Evidence (Chadbourn ed. 1978) § 2034, p. 342, italics added.) Such justification is lacking for the
Gould
rule requiring corroboration of an out-of-court identification that is not confirmed at trial.
Gould
did not explicate the reasons for its rule.
(Gould, supra,
54 Cal.2d at p. 631.) Instead,
Gould
cited a federal case,
Reamer
v.
United States
(6th Cir. 1956) 229 F.2d 884.
Reamer,
however, did not hold that an out-of-court identification is always insufficient as a matter of law to support a conviction. The defendant in
Reamer
was charged with bank robbery. The robber obscured his face during the robbery with a hat pulled down low and a handkerchief.
(Id.
at p. 885.) Two bank tellers independently identified the defendant as the robber during a lineup in which the participants obscured their faces with hats and handkerchiefs while they gestured as the robber had gestured and spoke the words that the robber had spoken during the holdup.
[265]
(Ibid.)
The tellers based their identifications on the voice and the “ ‘whole appearance’ ” of the defendant in the lineup.
(Ibid.)
At trial, the tellers testified to and reaffirmed their belief in the accuracy of their out-of-court identifications.
(Ibid.)
Unlike
Gould,
nothing in the federal court’s opinion in
Reamer
indicates that the identifying witnesses there attempted but failed to make an in-court identification or questioned their prior out-of-court identifications.
(Reamer
v.
United States, supra,
229 F.2d at pp. 885-886.)
Reamer
held that the tellers’ out-of-court identifications
in that case
were not “dependable substantial proof’ to support the defendant’s conviction for bank robbery.
(Reamer
v.
United States, supra,
229 F.2d at p. 886.)
Reamer
did not purport to establish a rule that any out-of-court identification is by itself insufficient
as a matter of law
to support a conviction, much less did it offer any reasons in support of such a rule.
Thus, neither this court’s decision in
Gould, supra,
54 Cal.2d 621, nor the federal case it relied upon,
Reamer
v.
United States, supra,
229 F.2d at page 886, offers convincing justification for the rule that an out-of-court identification is always insufficient as a matter of law to support a conviction. Nor can the
Gould
corroboration requirement be justified on the ground that out-of-court identifications are inherently unreliable.
Gould
itself recognized that an out-of-court identification generally has
greater
probative value than an in-court identification, even when the identifying witness does not confirm the out-of-court identification: “[T]he [out-of-court] identification has greater probative value than an identification made in the courtroom after the suggestions of others and the circumstances of the trial may have intervened to create a fancied recognition in the witness’ mind. [Citations.] The failure of the witness to repeat the [out-of-court] identification in court does not destroy its probative value . . . .”
(Gould, supra,
54 Cal.2d at p. 626; accord,
Gilbert
v.
California
(1967) 388 U.S. 263, 272-273, fn. 3 [18 L.Ed.2d 1178, 1186, 87 S.Ct. 1951].) It is paradoxical, as other courts have recognized, to acknowledge that an out-of-court identification has greater probative value than an in-court identification, and yet hold that an in-court identification is sufficient evidence on which to base a conviction but an out-of-court identification is not.
Also, there is no logic to requiring corroboration of out-of-court identifications, but not of other types of hearsay that might be offered as evidence of guilt. Like the United States Supreme Court, we reject the proposition that “out-of-court statements of identification are inherently less reliable than other out-of-court statements.”
(United States
v.
Owens
(1988)
[266]
484 U.S. 554, 561 [98 L.Ed.2d 951, 959, 108 S.Ct. 838].) Logically, therefore, out-of-court identifications and other out-of-court statements should be measured by the same standard in judging their sufficiency to support a conviction. The
Gould
corroboration requirement goes against this logic, for while it prohibits in all cases the use of an out-of-court identification as the sole evidence of guilt, other types of out-of-court statements may serve as the sole evidence of guilt if they satisfy the substantial evidence test.
The
Gould
corroboration requirement also finds no support in the law of other jurisdictions, federal or state, that admit out-of-court statements of identification. Like
Gould,
the federal courts and the great majority of state courts admit out-of-court identifications when the identifying witness testifies and is subject to cross-examination.
2
But no jurisdiction that admits out-of-court identifications has adopted the broad holding in
Gould, supra,
54 Cal.2d 621, 631, that an out-of-court identification can
never
by itself be sufficient to support a conviction.
3
To the contrary, numerous courts have held that an out-of-court identification can be sufficient by itself to support
[267]
a conviction even in the absence of other evidence connecting the defendant to the crime.
(Acosta
v.
State
(Del. 1980) 417 A.2d 373;
Watkins
v.
State
(Ind. 1983) 446 N.E.2d 949, 957;
Bedford
v.
State, supra,
443 A.2d at p. 84;
Nance
v.
State
(1993) 331 Md. 549 [629 A.2d 633, 639];
State
v.
Mancine
(1991) 124 N.J. 232, 256 [590 A.2d 1107, 1119];
State
v.
Robar
(1991) 157 Vt. 387, 395 [601 A.2d 1376, 1380];
State
v.
Hendrix
(1988) 50 Wn.App. 510, 514-516 [749 P.2d 210, 212-213].) In
Bedford
v.
State, supra,
443 A.2d 78, 84, Maryland’s highest court, noting the inconsistency between
Gould's
recognition that an out-of-court identification may possess great probative value and its holding that an out-of-court identification can never be sufficient evidence to support a conviction, concluded that “we regard the opinion in
Gould
as not soundly premised.”
The
Gould
corroboration requirement, by inflexibly requiring corroboration of all unconfirmed out-of-court identifications regardless of their probative value, does not take into account the many varied circumstances that may attend an out-of-court identification and affect its probative value. These circumstances include, for example: (1) the identifying witness’s prior familiarity with the defendant; (2) the witness’s opportunity to observe the perpetrator during the commission of the crime; (3) whether the witness has a motive to falsely implicate the defendant; and (4) the level of detail given by the witness in the out-of-court identification and any accompanying description of the crime. (See also CALJIC No. 2.92 (5th ed. 1988) [listing factors relevant to reliability of eyewitness identification].) Evidence of these circumstances can bolster the probative value of the out-of-court identification by corroborating both that the witness actually made the out-of-court identification (e.g., testimony by the police officer or other person to whom the statement was made) and that the identification was reliable (e.g., evidence that the witness was present at the scene of the crime and in a position to observe the perpetrator, evidence that the witness had a prior familiarity with the defendant, or evidence that the witness had no self-serving motive to implicate the defendant). Such evidence can show the accuracy and reliability of the out-of-court identification, even though none of it is independent evidence connecting the defendant to the crime, the corroboration required by
Gould.
Nor does the
Gould
corroboration requirement take account of the many varied circumstances relating to the witness’s failure to identify the defendant at trial. Among these circumstances áre: (1) whether the identifying
[268]
witness admits, denies, or fails to remember making the out-of-court identification; (2) whether the witness remembers the underlying events of the crime but no longer believes in the accuracy of the out-of-court identification; (3) whether, if the witness claims the identification was false or erroneous, the witness offers an explanation for making a false or erroneous identification; (4) whether, if the witness claims a failure of recollection, there are reasons supporting the loss of memory; (5) whether there is evidence that the witness’s failure to confirm the identification in court resulted from the witness’s appreciation that doing so would result in the defendant’s conviction; or (6) whether there is evidence that, as the Attorney General suggests occurred here, the witness’s failure to confirm the identification arises from fear or intimidation.
As
Gould
itself recognized, the probative value of a witness’s out-of-court identification can vary tremendously depending on these surrounding circumstances.
(Gould, supra,
52 Cal.2d at p. 631 [“[T]he probative value of an identification depends on the circumstances under which it was made.”].) Two examples will illustrate the degree to which the probative value of an out-of-court identification can vary. In the first example, an identifying witness who is incarcerated allegedly tells his cellmate that he saw the defendant commit the crime. The alleged statement of identification to the cellmate contains no assertion of a prior acquaintance between the witness and the defendant, and does not describe the circumstances surrounding the crime or any other details subject to corroboration. When the witness is called to testify at trial, the witness denies making the identification to his cellmate, denies the truth of the identification, denies knowing anything about the defendant’s crime, and denies ever having known or seen the defendant. The cellmate testifies to the witness’s out-of-court identification as a prior inconsistent statement. (Evid. Code, § 1235.) The defense then presents unrebutted evidence showing that the witness was in another state at the time of the crime and the cellmate has been promised a reduced sentence in exchange for his testimony. Under these circumstances, the probative value of the out-of-court identification would be very low.
By contrast, consider a witness who gives a videotaped statement to the police immediately after observing the commission of the crime. The witness has known the defendant for many years, identifies the defendant as the perpetrator, and gives many details of the crime that only someone who observed the crime would likely know. Before trial, the witness is in an automobile accident and loses all memory of the crime. The prosecution introduces the videotaped statement as the witness’s past recollection recorded (Evid. Code, § 1237), and offers other evidence showing that the witness was present at the scene of the crime. The defense presents no
[269]
evidence to impeach the witness’s identification. Under these circumstances, the probative value of this out-of-court identification would be high.
Notwithstanding the disparity in probative value between the out-of-court identifications in these two examples, under the
Gould
corroboration requirement both are equally insufficient to support a conviction. If we were deciding the matter as a question of first impression, we would conclude that the
Gould
corroboration requirement is unacceptably overbroad. Given the wide range of relevant circumstances affecting the probative value of an out-of-court identification, it is unwise to attempt to subsume the whole spectrum of out-of-court identifications under a rule that deems insufficient
as a matter of law
any out-of-court identification uncorroborated by other evidence connecting the defendant to the crime. Instead, the case-by-case analysis that we apply in reviewing the sufficiency of other types of evidence under the substantial evidence test is preferable because it permits an individualized assessment of the probative value of the particular out-of-court identification at issue. (See
State
v.
Mancine, supra,
124 N.J. at p. 256 [590 A.2d at p. 1119] [“A court should weigh substantive evidence found in a prior inconsistent statement [of identification] on the same scale as any other evidence when determining whether sufficient evidence to support a guilty verdict exists.”].)
Nevertheless, the
Gould
corroboration requirement is not a question of first impression. Accordingly, it is not enough that a different rule might seem preferable to us now; before deciding to overrule it, we must take into account the doctrine of stare decisis. The
Gould
decision was written by Justice Traynor for a unanimous court, and has been the law for 35 years. We do not lightly overrule such an established precedent. As we have observed in the past: “ ‘It is, of course, a fundamental jurisprudential policy that prior applicable precedent usually must be followed even though the case, if considered anew, might be decided differently by the current justices. This policy, known as the doctrine of stare decisis, “is based on the assumption that certainty, predictability and stability in the law are the major objectives of the legal system; i.e., that parties should be able to regulate their conduct and enter into relationships with reasonable assurance of the governing rules of law." [Citation.] [¶][] It is likewise well established, however, that the foregoing policy is a flexible one which permits this court to reconsider, and ultimately to depart from, our own prior precedent in an appropriate case. [Citation.] As we stated in
Cianci
v.
Superior Court
(1985) 40 Cal.3d 903, 924 [221 Cal.Rptr. 575, 710 P.2d 375], “[although the doctrine [of stare decisis] does indeed serve important values, it nevertheless should not shield court-created error from correction.” ’ ”
A central factor in the stare decisis analysis is whether there are private or legislative reliance interests that have sprung up in dependence on the existing rule and, if so, the costs that would result to those interests if the rule were changed.
(People
v.
Latimer, supra,
5 Cal.4th at pp. 1212-1216;
Moragne
v.
States Marine Lines
(1970) 398 U.S. 375, 403 [26 L.Ed.2d 339, 358, 90 S.Ct. 1772] [reliance interests are “often considered the mainstay of
stare decisis”].)
Here, we conclude that no reliance interests exist that would preclude us from overruling the
Gould
corroboration requirement. Although a defendant has a legitimate interest in not being convicted on evidence that is insufficient to prove the elements of the charged offense beyond a reasonable doubt to a rational jury (the substantial evidence standard of
People
v.
Johnson, supra,
26 Cal.3d 557, 578), no defendant has a legitimate interest in holding the prosecution to a greater burden of proof. A defendant cannot cognizably claim to have relied on the expectancy that he would not be convicted because the only evidence against him, although sufficient under
People
v.
Johnson,
was an out-of-court identification which he anticipated would not be confirmed at trial. Nor are there any property interests that would be affected by overruling
Gould.
In addition, the
Gould
corroboration requirement is a common law rule created by this court, not a question of statutory interpretation which would give us greater pause before overruling. (See
People
v.
Latimer, supra, 5
Cal.4th at p. 1213.) Indeed, as we noted above, the Legislature has, with few exceptions (see Pen. Code, §§ 118, subd. (b), 1111) left to the courts the task of defining standards for the sufficiency of evidence. Nor, unlike
People
v.
Latimer, supra, 5
Cal.4th at pages 1212-1216, has the Legislature enacted statutes in reliance on the
Gould
corroboration requirement.
Moreover, this court has previously qualified the
Gould
corroboration requirement in two respects. In
People
v.
Chavez, supra,
26 Cal.3d 334, 362-364, and
People
v.
Ford
(1981) 30 Cal.3d 209, 215 [178 Cal.Rptr. 196, 635 P.2d 1176], the court held that a pretrial identification made under oath at a preliminary hearing is sufficient evidence to support a conviction in the absence of other evidence linking the defendant to the crime, even if at trial the witness repudiates the prior identification. In
People
v.
Lucky
(1988) 45 Cal.3d 259, 289 [247 Cal.Rptr. 1, 753 P.2d 1052], the court held that an out-of-court identification repudiated at trial can be corroborated by the
[271]
repudiated out-of-court statement of another witness linking the defendant to the crime.
4
In
People
v.
Chavez, supra,
26 Cal.3d 334,
People
v.
Ford, supra,
30 Cal.3d 209, and
People
v.
Lucky, supra,
45 Cal.3d 259, this court focused on the specific factors affecting the probative value of the particular out-of-court identification at issue and noted that under the circumstances of the case a reasonable trier of fact could have credited the out-of-court identification.
(People
v.
Lucky, supra,
at p. 289 [“there were ample concrete bases on which the jury could evaluate the relative credibility of the prior [out-of-court] and current [in-court] statements”];
People
v.
Ford, supra,
at p. 214 [“in the circumstances of this case the repudiation [of the out-of-court identification] itself provided the jury with a basis on which it could have discredited [the identifying witness’s] trial testimony in favor of his [out-of-court identification]”];
People
v.
Chavez, supra,
at p. 364 [“unlike
[Gould, supra,
54 Cal.2d 621, and other] cases in which the record contained no concrete basis to credit the witness’ prior identification over his trial testimony, the prosecution in the instant case presented evidence which both explained and discredited the witness’ inconsistent testimony at the time of trial”]; see also
People
v.
Armijo, supra,
221 Cal.App.3d 271, 278-279 [reviewing the indicia of reliability present in the circumstances surrounding three repudiated out-of-court identifications of the defendant];
In re Richard W.
(1982) 136 Cal.App.3d 733, 736 [186 Cal.Rptr. 409] [witness’s uncorroborated statement of identification was sufficient to support finding that burglary occurred, where at time of trial witness no longer remembered what burglars looked like but remembered prior identification was accurate and reaffirmed its accuracy].)
We conclude that, considerations of stare decisis notwithstanding, this approach—individually assessing the circumstances of the out-of-court identification to determine whether it is sufficient to support a criminal conviction—is a sound one and should be applied to all out-of-court identifications. Accordingly, we hereby overrule that portion of
Gould, supra,
54 Cal.2d
[272]
621, 631, holding that an out-of-court identification is in all cases insufficient by itself to sustain a conviction and must be corroborated by other evidence linking the defendant to the crime. Instead, the substantial evidence test set forth in
People
v.
Johnson, supra,
26 Cal.3d 557, 578, should be used to determine whether an out-of-court identification is sufficient to support a criminal conviction.
In overruling the corroboration requirement of
Gould, supra,
54 Cal.2d 621, 631, we intend no disrespect to the court that created it. Although
Gould
does not identify the reason for its corroboration requirement, it is plausible that the requirement arose out of a well-placed sense of caution and prudence on the part of the court. In
Gould,
the court for the first time authorized the admission of out-of-court statements of identification as substantive evidence of identity, and furthermore authorized doing so even when the identification is not confirmed at trial. The
Gould
court may reasonably have been concerned that this greatly expanded admissibility carried with it the potential for the presentation of false or unreliable out-of-court identifications that might nonetheless carry excessive evidentiary weight in a jury’s eyes. The
Gould
corroboration requirement, by precluding reliance on an unconfirmed out-of-court identification as the sole evidence of guilt, circumscribed the impact of the new rule of admissibility that the court was simultaneously adopting and served as a protection against any abuses of the new rule that might occur.
We now have had the benefit, as the
Gould
court did not, of 35 years of experience with the admission of out-of-court identifications as substantive evidence. In our view, the rule admitting out-of-court identifications has not been subject to widespread abuse. In addition, as we discuss more fully in the next part, even in the absence of the
Gould
corroboration requirement, the other protections provided by the trial and appellate process are adequate to guard against any such abuses.
IV
In urging us to retain the
Gould
corroboration requirement, defendant argues that the rule is necessary to protect defendants from unreliable out-of-court identifications not made under penalty of perjury. There are, however, already a number of existing safeguards at both the trial and appellate levels that adequately ensure the reliability of such identifications. One safeguard is the availability of the identifying witness for cross-examination, which is a condition of admitting the out-of-court identification.
(Gould, supra,
54 Cal.2d at p. 626.)
Defendant contends that cross-examination of a witness concerning an out-of-court identification that the witness does not confirm at trial is
[273]
necessarily ineffective. We disagree. Unconfirmed out-of-court identifications generally fall into one of two categories: Either the witness repudiates the out-of-court identification or the witness testifies that he or she lacks the recollection to either confirm or deny the out-of-court identification. In either situation, cross-examination can be effective in shedding light on the reliability and veracity of the out-of-court identification.
This court has previously noted with respect to the first category—a witness who disowns the out-of-court identification—that the “[defendant retains the opportunity to question the declarant as to the circumstances surrounding the prior statement^ and to elicit from the declarant an explanation for the inconsistencies in his prior statement and his on-the-stand testimony. Through such questioning, the defendant can test the credibility of the witness’ statements on the witness stand before the trier of fact.”
(People
v.
Chavez, supra,
26 Cal.3d at pp. 360-361.)
This view finds substantial support. Many years ago, Judge Learned Hand observed that juries are capable of determining the credibility of out-of-court statements that are inconsistent with a witness’s trial testimony by observing the witness’s in-court demeanor: “If, from all that the jury see of the witness, they conclude that what he says now is not the truth, but what he said before, they are none the less deciding from what they see and hear of that person and in court. There is no mythical necessity that the case must be decided only in accordance with the truth of the words uttered under oath in court.”
(Di Carlo
v.
United States
(2d Cir. 1925) 6 F.2d 364, 368.)
A prominent treatise on the law of evidence makes a similar point: “The witness who has told one story aforetime and another today has opened the gates to all the vistas of truth which the common law practice of cross-examination and re-examination was invented to explore. The reasons for the change of face, whether forgetfulness, carelessness, pity, terror, or greed, may be explored by the two questioners in the presence of the trier of fact, under oath, casting light on which is the true story and which the false. It is hard to escape the view that evidence of a prior inconsistent statement, when declarant is on the stand to explain it if he can, has in high degree the safeguards of examined testimony.” (2 McCormick, Evidence (4th ed. 1992) § 251, p. 120.) Moreover, the witness’s repudiation of the out-of-court identification is by its very nature testimony favorable to the defendant. The prosecution must then demonstrate that the out-of-court identification, and not the in-court repudiation, was correct.
Cross-examination can also be effective with regard to cases falling into the second category—that is, cases in which the witness’s reason for not
[274]
confirming the out-of-court identification is failure of recollection. As the United States Supreme Court has observed: “[T]he opportunity to bring out such matters as the witness’ bias, his lack of care and attentiveness, his poor eyesight, and even (what is often a prime objective of cross-examination [citation]) the very fact that he has a bad memory. . . . suffices [to satisfy the confrontation clause’s requirement of cross-examination] when the witness’ past belief is introduced and he is unable to recollect the reason for that past belief. . . . [T]he foundation for the belief . . . cannot effectively be elicited, but other means of impugning the belief are available. . . . The weapons available to impugn the witness’ statement when memory loss is asserted will of course not always achieve success .... They are, however, realistic weapons . . . .”
(United States
v.
Owens, supra,
484 U.S. at pp. 559-560 [98 L.Ed.2d at p. 958].)
In addition to cross-examination, a further safeguard against convictions based on unreliable out-of-court identifications lies in the ability of the defendant to offer other evidence casting doubt on the identification, such as evidence that the identifying witness was not present at the scene of the crime, was not previously familiar with the defendant, or had a motive to implicate the defendant. Moreover, the defendant is free to seek appropriate instruction directing the jury’s attention to the issue of identification and to the factors affecting the probative value of the out-of-court identification. (See CALJIC Nos. 2.91, 2.92;
Finally, the substantial evidence test used to determine the sufficiency of the evidence supporting a conviction provides additional protection against a dubious out-of-court identification. Under this standard, the probative value of the identification and whatever other evidence there is in the record are considered together to determine whether a reasonable trier of fact could find the elements of the crime proven beyond a reasonable doubt. (See
People
v.
Johnson, supra,
26 Cal.3d at pp. 576-578;
Jackson
v.
Virginia, supra,
443 U.S. at pp. 318-319 [61 L.Ed.2d at pp. 573-574].) The defendant can raise the issue of whether substantial evidence exists to support conviction both in the trial court by means of a motion at the close of the evidence under Penal Code section 1118.1 (as defendant did here) and on appeal.
We conclude therefore that the availability of the identifying witness for cross-examination, the opportunity of the defense to present other evidence questioning the reliability of the out-of-court identification and to request appropriate jury instructions, and the requirement that substantial evidence support the conviction are adequate safeguards against the unjust conviction
[275]
of a defendant solely on the basis of an unreliable out-of-court identification.
5
V
Having concluded that the sufficiency of an out-of-court identification to support a conviction should be judged by the substantial evidence test and that the
Gould
corroboration requirement should be rejected, we must now determine whether substantial evidence supports defendant’s conviction. Before doing so, however, we address defendant’s contention that due process precludes us from applying our holding retroactively to his case. He relies on the principle that the due process guaranty prohibits courts from retroactively applying judicial decisions that expand criminal liability. (See
People
v.
Escobar
(1992) 3 Cal.4th 740, 752 [12 Cal.Rptr.2d 586, 837 P.2d 1100] [“a state Supreme Court, no less than a state Legislature, is barred from making conduct criminal which was innocent when it occurred, through the process of judicial interpretation”];
Bouie
v.
City of Columbia
(1964) 378 U.S. 347, 352-354 [12 L.Ed.2d 894, 899-900, 84 S.Ct. 1697].) Our decision in this case to adopt the substantial evidence test as the standard for judging the sufficiency of an out-of-court identification to support a conviction, however, does not expand criminal liability; the conduct for which defendant was convicted was criminal long before this case arose. Thus, there is no due process obstacle to retroactive application of our decision. (See
ibid.)
Defendant’s claim that it would violate equal protection not to apply the
Gould
corroboration requirement to his case because previously it has been applied to other defendants also lacks merit, for the fact that retroactivity results in similarly situated defendants being treated differently “is an unavoidable consequence of dispensing justice on a societal basis.”
(People
v.
Carrera
(1989) 49 Cal.3d 291, 328 [261 Cal.Rptr. 348,
111
P.2d 121].)
Nor is it inequitable to apply our decision to the facts of this case. As we observed earlier, so long as there is substantial evidence supporting his
[276]
conviction, defendant has no cognizable interest in escaping conviction through the operation of a rule that would have required additional evidence. Moreover, defendant has not asserted that he would have pursued a different trial strategy or offered different evidence had the
Gould
corroboration requirement been overruled before he was tried. (See
People
v.
Welch
(1993) 5 Cal.4th 228, 237-238 [19 Cal.Rptr.2d 520, 851 P.2d 802] [new rule requiring trial objection to probation condition not applied retroactively because at time of trial defendant had no notice that he was required to object].)
For the reasons set forth above, we apply to defendant’s case our decision that the substantial evidence test should be used to determine the sufficiency of an out-of-court identification to support a criminal conviction. The substantial evidence test is satisfied here because a reasonable jury could find from the evidence presented that the prosecution had shown beyond a reasonable doubt that defendant had assaulted the victim, Garcia, with a firearm. As set forth in greater detail in part I above, the shooting occurred during an outdoor party in an alley. It is undisputed that witnesses Gomez and Guzman were present at the party when Garcia was shot. Gomez was first interviewed by Officer On at the hospital immediately after the shooting and was interviewed again by Officer Kelley three days later; on both occasions he unequivocally identified defendant (under the name “Beto”) as the shooter and gave a physical description of him. Although at trial Gomez recanted his identification of defendant, he admitted that he had made these two prior identifications of defendant to the police. Gomez also admitted at trial that he was acquainted with defendant before the shooting. And there was testimony by eyewitness Rodriguez that as the gunman approached, Gomez exclaimed: “I know that guy. He’s from West Side Anaheim.” In addition, when Officer Orr interviewed witness Guzman at the hospital immediately after the shooting, Guzman gave a physical description of the gunman that was both similar to Gomez’s independent description of the gunman and generally consistent with defendant’s appearance.
Although witnesses Gomez and Guzman both disowned their out-of-court statements when testifying at trial, the prosecution offered evidence that they had a motive to falsely recant their statements: Gomez and Guzman both testified that they believed it was wrong to accuse a member of a rival gang of committing a crime, and a gang expert testified that gang members disapprove of testifying against rival gang members and intimidate witnesses who initially cooperate with the police into changing their testimony.
From this evidence, a reasonable jury could have concluded that Gomez and Guzman were telling the truth when they made their out-of-court
[277]
statements to police officers and that they recanted those statements in court for gang-related reasons. Those out-of-court statements are “substantial evidence—that is, evidence which is reasonable, credible, and of solid value—”
(People
v.
Johnson, supra,
26 Cal.3d at p. 578) from which a reasonable jury could conclude beyond a reasonable doubt that defendant assaulted Garcia with a firearm.
Conclusion
For the reasons stated above, we conclude that we should overrule
Gould, supra,
54 Cal.2d 621, and reject its rule that an out-of-court identification is by itself always insufficient evidence to support a conviction. Instead, the sufficiency of an out-of-court identification to support a conviction should be judged by the substantial evidence standard of
People
v.
Johnson, supra,
26 Cal.3d at page 578. Because in this case a reasonable juror could have concluded beyond a reasonable doubt from the evidence presented that defendant had assaulted the victim Garcia by personally using a firearm, there was sufficient evidence to support his conviction. Accordingly, we reverse the judgment of the Court of Appeal, which held that defendant’s conviction should be reversed because the jury was not instructed in accordance with
Gould
that it could not rely on an uncorroborated out-of-court identification as the sole evidence on which to convict defendant.
Lucas, C. J., Mosk, J., Arabian, J., Baxter, J., George, J., and Werdegar, J., concurred.
1
We use the term
“Gould
corroboration requirement” to refer to
Gould’s
requirement that an unconfirmed out-of-court identification be corroborated by “other evidence tending to connect the defendant with the crime.”
(Gould, supra,
54 Cal.2d at p. 631.)
2
Federal Rules of Evidence, rule 801(d)(1)(C) (28 U.S.C.) provides: “A statement is not hearsay if . . . [*]□ [t]he declarant testifies at the trial or hearing and is subject to cross-examination concerning the statement, and the statement is . . . one of identification of a person made after perceiving the person.” Twenty-seven states have adopted evidence statutes or rules essentially identical to Federal Rules of Evidence, rule 801(d)(1)(C). (Alaska Rules Evid., rule 801(d)(1)(C); Ariz. Rev. Stat. Ann., Rules Evid., rule 801(d)(1)(C); Ark. Rules Evid., rule 801(d)(l)(iii); Colo. Rules Evid., rule 801(d)(1)(C); Del. Unif. Rules Evid., rule 801(d)(1)(C); Fla. Ann. Stat., Evid. Code §90.801, subd. (2)(c); Hawaii Rules Evid., rule 802.1(3); Idaho Rules Evid., rule 801(d)(1)(C); Iowa Rules Evid., rule 801(d)(1)(C); 111. Ann. Stat. ch. 725, §5/115-12 (Smith-Hurd 1995); Ky. Rules Evid., rule 801A(a)(3); Me. Rules Evid., rule 801(d)(1)(B); Mich. Rules Evid., rule 801(d)(1)(C); Miss. Rules Evid., rule 801(d)(1)(C); Mont. Rules Evid., rule 801(d)(1)(C); N.H. Rules Evid., rule 801(d)(1)(C); N.M. Rules Evid., rule ll-801(D)(l)(c); N.D. Rules Evid., rule 801(d)(l)(iii); Or. Rules Evid., rule 801(4)(a)(C); R.I. Rules Evid., rule 801(d)(1)(C); S.D. Codified Laws Ann. § 19-16-2(3); Tex. Rules Grim. Evid., rule 801(e)(1)(C); Utah Rules Evid., rule 801(d)(1)(C); Vt. Rules Evid., rule 801(d)(1)(C); Wash. Rules Evid., rule 801(d)(l)(iii); W. Va. Rules Evid., rule 801(d)(1)(C); Wyo. Rules Evid., rule 801(d)(1)(C).)
Minnesota has adopted the federal rule with the additional qualification that the circumstances demonstrate the out-of-court identification to be reliable; two other states—Wisconsin and Nevada—have adopted it with the qualification that the out-of-court identification be made soon after the crime; and Ohio has adopted the federal rule with both these qualifications. (Minn. Rules Evid., rule 801(d)(1)(C); Nev. Rev. Stat. tit. 4, §51.035, subd. (2)(c) (1995); Ohio Rules Evid., rule 801(D)(1)(c); Wis. Stat. Ann. § 908.01, subd. (4)(a)(3) (West 1993).) Other states have made out-of-court statements of identification admissible by judicial decision.
(Modesitt
v.
State
(Ind. 1991) 578 N.E.2d 649, 654;
Bedford
v.
State, supra,
443 A.2d at pp. 80-82;
Commonwealth
v.
Fitzgerald
(1978) 376 Mass. 402 [381 N.E.2d 123, 130];
Commonwealth
v.
Doa
(1989) 381 Pa.Super. 181 [553 A.2d 416].)
3
Massachusetts permits an out-of-court statement of identification to serve as the sole evidence to support a conviction unless the identification is inconsistent with the identifying witness’s trial testimony.
(Commonwealth
v.
Vitello, supra,
376 Mass. 426 [381 N.E.2d 582, 601] [“an extrajudicial identification may in an appropriate case provide a jury with a sufficient basis upon which to find a defendant guilty beyond a reasonable doubt”].) If the
[267]
out-of-court identificatibn is inconsistent with the identifying witness’s trial testimony, then Massachusetts requires additional evidence to support the conviction, as does Florida.
(Commonwealth
v.
Daye
(1984) 393 Mass. 55 [469 N.E.2d 483, 495] [prosecution “must produce identification evidence in addition to a prior inconsistent statement in order to meet its burden of proof’];
State
v.
Moore
(Fla. 1986) 485 So.2d 1279, 1285 [prior inconsistent statement insufficient to support a conviction]; see also
The Court of Appeal relied on this rationale to hold that the
Gould
corroboration requirement was met in this case. It reasoned that eyewitness Guzman’s out-of-court description of the gunman’s appearance, which was generally consistent with defendant’s physical appearance, was “other evidence tending to connect the defendant with the crime”
(Gould, supra,
54 Cal.2d at p. 631) that corroborated eyewitness Gomez’s out-of-court identification. (See
People
v.
Lucky, supra,
45 Cal.3d at p. 289 [out-of-court identification repudiated at trial was corroborated by second witness’s out-of-court statement, also repudiated at trial, “that she saw defendant leave with a gun right after the shots were fired”];
People
v.
Armijo
(1990) 221 Cal.App.3d 271, 278 [270 Cal.Rptr. 496].) Because of our conclusion that the better course is to dispense with the
Gould
corroboration requirement altogether, we need not decide whether, under
Gould
and
People
v.
Lucky,
Guzman’s out-of-court description would be adequate corroboration of Gomez’s identification.
5
Because we overrule the
Gould
corroboration requirement and instead hold that the sufficiency of an out-of-court identification to support a conviction should be determined by applying the substantial evidence test, the question whether, as the Court of Appeal held in this case, the jury should have been instructed on the
Gould
corroboration requirement is moot. Furthermore, we disapprove the holding of
People
v.
Marquez
(1993) 16 Cal.App.4th 115, 121 [20 Cal.Rptr.2d 365] that the trial court must instruct the jury sua sponte on the
Gould
corroboration requirement. We also disapprove CALJIC No. 3.11.1 (5th ed. 1995 supp.), which embodies the
Marquez
holding.
In addition, over the years a number of cases have applied the
(1981) 116 Cal.App.3d 341, 345 [172 Cal.Rptr. 44].) We disapprove those decisions to the extent they are inconsistent with our holding in this case.
AI Brief
AI-generated · verify before citing
Holding. The court overruled the rule from People v. Gould (1960) 54 Cal.2d 621, which held that an out-of-court identification is insufficient as a matter of law to sustain a conviction without corroborating evidence. Instead, the sufficiency of such evidence must be determined by the substantial evidence test.
Issues
Whether an out-of-court identification that is not confirmed at trial is insufficient as a matter of law to sustain a conviction without corroborating evidence.
Whether the substantial evidence test should apply to out-of-court identifications.
Disposition. remanded
Quotations verified verbatim against the opinion
“we conclude that we should overrule Gould’s holding that an out-of-court identification is in all cases insufficient by itself to sustain a conviction.”
“the sufficiency of an out-of-court identification to support a conviction should be determined under the substantial evidence test”