279 Ga. 435
Ga.
2005

Lead Opinion

BENHAM, Justice.

Appellant J. Brodre Brodes was convicted of two counts of armed robbery in a trial in which the only evidence implicating him was the testimony of the two victims identifying him as the perpetrator. The victims testified they were “absolutely certain” of their identification of him and were repeatedly called upon by the prosecuting attorney to repeat their certainty in the accuracy of their identification. In its instructions to the jury, the trial court gave the pattern jury instruction on eyewitness identification,1 which informs the jurors they may *436consider a witness’s level of certainty in his or her identification in assessing the reliability of the identification. See Suggested Pattern Jury Instructions, Vol. II: Criminal Cases (2nd ed.). We issued a writ of certiorari to the Court of Appeals after it affirmed the judgment of conviction in Brodes v. State, 268 Ga. App. 895 (602 SE2d 895) (2004), expressing as our concern “whether a trial court errs in using the level of certainty’ charge in instructing the jury on assessing the reliability of eyewitness identification.”2

We did not reach the merits of this issue in Jones v. State, 273 Ga. 213 (3) (b) (539 SE2d 143) (2000), but noted the development of a split of opinion among state appellate courts concerning the use of the “level of certainty” portion of the pattern charge. Id., 273 Ga. at 218, n. 17.3 See also Gibbs v. State, 270 Ga. App. 56 (3) (606 SE2d 83) (2004), where the Court of Appeals reversed a conviction due to ineffective assistance of counsel and, citing Jones, noted the trial court could have erred in giving the “level of certainty” charge. The case at bar presents us with an opportunity to address the merits of *437the contention that it is error for a trial court to include an eyewitness’s “level of certainty” in his or her identification of a criminal defendant as a factor for the jury to consider in evaluating the reliability of the eyewitness identification.

In the case at bar, the two victims were softball teammates who worked together. They were talking in a lighted parking area4 near a fast-food restaurant after 11:00 p.m. in October 1996, standing beside their respective vehicles with the driver’s doors open. One victim testified he saw a man approaching them from 40-50 feet away and thought he was a softball teammate. Though he could not make out the approaching man’s features even when he was five-six feet away, the witness realized at that point the man was a stranger. When he got closer, the stranger pulled a gun, cocked it, and demanded money of the victim. The victim noticed the weapon was a revolver and was pointed at his mid-section, and he saw bullets in the gun. After the victim gave his wallet to the robber, the robber turned away from the victim and approached the other victim. The first victim testified 10-20 seconds elapsed from the time the robber walked up to the victims. The second victim testified the robber pointed the gun at him after getting the first victim’s wallet and he gave the robber the cash he had in his pocket. He testified he saw the robber’s face for about seven seconds. He noticed the weapon was a long-barreled revolver in which he could see the bullets.

Both victims described to police the weapon used, the red satin-like Chicago Bulls jacket the perpetrator wore, and the robber’s race. The first victim, who was 6' 1", described the robber as being 5'8"-5'10" and weighing 140-150 pounds. The second victim did not give a description of the robber’s height and weight, and was unable to give any details about the robber’s face to police. The investigating detective testified Brodes was six feet tall and weighed 170 pounds. The first witness was unable to make a positive identification when presented with a photo array at his office two days later, but wanted to see one of the photographed individuals in person. With the assistance of equipment provided by the detectives, the victim then compiled a composite sketch of the robber five-ten minutes after he examined the photo array. When presented with a physical lineup the following day,5 the victim recognized Brodes as the robber. During his testimony, he identified Brodes as the man who pointed a gun at him and robbed him and described his certainty in his identification as *438“positive,” ‘TOO percent certain,” “absolutely confident,” and “I did not pick the wrong man.” The second victim was presented with the photo array right after the first victim and had no hesitation in selecting a photo as being that of the robber. He then viewed the first victim’s composite sketch of the robber and, the next day, he viewed a physical lineup and recognized Brodes as the robber. The investigating detective testified Brodes was the only person who appeared in both the photo array and the physical lineup. The second victim also identified Brodes at trial as the armed robber and described himself as “absolutely certain” and “dead certain” of his identification of Brodes.

At trial, Brodes presented the expert testimony of Dr. Steven Cole, a research psychologist who has been studying, teaching, and publishing papers on eyewitness identification since 1985. Dr. Cole testified that eyewitness identification was a “subarea” of human memory and perception that involved three phases of memory: encoding (where information is taken in by human senses), storage (where perceptions are integrated and stored in the brain), and retrieval (where information is moved out of storage and into consciousness). The encoding phase can be negatively impacted by, among other things, stress or violence, the presence of a weapon, and the presence of a cross-racial identification situation. “Storage” is affected by the passage of time and the “assimilation factor,” whereby a witness subconsciously incorporates into his identification information gathered after the event and confused with the event, including photo arrays and live lineup procedures. Dr. Cole informed the jury of two counter-intuitive findings repeatedly established by researchers: (1) the ability to perceive and remember decreases in a “high stress” situation and the presence of a gun during the commission of a robbery makes the situation “high stress”; and (2) there is not a good relationship between a witness’s level of confidence in his identification and the accuracy.

Jury instructions are the lamp to guide the jury’s feet in journeying through the testimony in search of a legal verdict. The office of a charge by the court is to give to the jury such instruction touching the rules of law pertinent to the issues involved in a pending trial as will enable them intelligently to apply thereto the evidence submitted, and from the two constituents law and fact make a verdict. [Cit.]

(Citations and punctuation omitted.) Chase v. State, 277 Ga. 636, 639 (2) (592 SE2d 656) (2004). The creation ofthe pattern jury instruction regarding the assessment of reliability of eyewitness identification testimony reflects the studied conclusion that judicial guidance to the *439jury on the topic of eyewitness identification is warranted6 • — • our task today is to determine whether it is appropriate to continue to highlight to the jury a witness’s certainty in that identification as a factor that bears on the reliability of the witness’s identification.

The vagaries of eyewitness identification are well-known; the annals of criminal law are rife with instances of mistaken identification. [Cit.] Mr. Justice Frankfurter once said: “What is the worth of identification testimony even when uncontradicted? The identification of strangers is proverbially untrustworthy. The hazards of such testimony are established by a formidable number of instances in the records of English and American trials. These instances are recent — not due to the brutalities of ancient criminal procedure.” [Cit.]

United States v. Wade, 388 U. S. 218, 228 (87 SC 1926, 18 LE2d 1149) (1967).

*440Several of the criteria for assessing reliability of an eyewitness identification set forth in the pattern instruction appear to have their roots in Neil v. Biggers, 409 U. S. 188 (93 SC 375, 34 LE2d 401) (1972). There, the U. S. Supreme Court addressed a habeas petitioner’s contention that he was denied due process of law when the trial court admitted into evidence eyewitness identification testimony derived from a showup, a suggestive identification procedure. The Court determined that pre-trial identification testimony could be admitted, despite a suggestive procedure, if the trial judge were able to conclude that the identification was reliable. To aid the trial judge in making that assessment, the Supreme Court listed

factors to be considered in evaluating the likelihood of misidentification [which] include [d] the opportunity of the witness to view the criminal at the time of the crime, the witness’ degree of attention, the accuracy of the witness’ prior description of the criminal, the level of certainty demonstrated by the witness at the confrontation, and the length of time between the crime and the confrontation.

409 U. S. at 199. Our pattern charge takes three of the criteria used by a trial judge to determine the admissibility of evidence, including the “level of certainty’ factor, and passes them on to the jury to be used to assess the reliability of an eyewitness identification.

In the 32 years since the decision in Neil v. Biggers, the idea that a witness’s certainty in his or her identification of a person as a perpetrator reflected the witness’s accuracy has been “flatly contradicted by well-respected and essentially unchallenged empirical studies.” State v. Long, 721 P2d 483, 491 (Utah 1986).7 *441State v. Dyle, 899 SW2d 607, 609 (Tenn. 1995). A Florida court of appeals recently noted:

*440“On the one hand, eyewitness testimony is very believable and can wield considerable influence over the decisions reached by a jury; on the other hand, eyewitness testimony is not always reliable. It can be flawed simply because of the normal and natural memory processes that occur whenever human beings acquire, retain, and attempt to retrieve information.” Elizabeth Loftus, Eyewitness Testimony 7 (1979).
*441While a great deal of credibility is given to eyewitness identification, empirical studies have shown that eyewitness identification can actually be extremely unreliable.... What makes eyewitness identification unreliable? When crime victims attempt to recall faces of strangers they have seen for only a brief period of time, many factors affect their ability to accurately remember what they have seen. Factors that may affect reliability of the identification include: lighting conditions; the duration of the event; violence; the age, sex and race of the perpetrator; the length of time between the event and the identification of the perpetrator, and the acquisition of post-event information that may distort the memory.

Lee v. State, 873 S2d 582, 584 (Fla. App. 2004) (quoting Mayer, Due Process Challenges to Eyewitness Identification Based on Pretrial Photographic Arrays, 13 Pace L. Rev. 815, 819 (1994)). “An important body of psychological research undermines the lay intuition that confident memories of salient experiences ... are accurate.... [T]he mere fact that we remember something with great confidence is not a powerful warrant for thinking it true.... [Accuracy of recollection is not highly correlated with the recollector’s confidence....” Krist v. Eli Lilly & Co., 897 F2d 293, 296 (7th Cir. 1990) (addressing the reliability of memory). “ ‘The scientific validity of the studies confirming the many weaknesses of eyewitness identification cannot be seriously questioned at this point.’ Abney, Expert Testimony and Eyewitness Identification, 91 Case & Comment, 26, 29 (March/Apr. 1986).” United States v. Moore, 786 F2d 1308, 1312 (5th Cir. 1986). “[T]here appears to be widespread agreement among the experts that the accuracy of eyewitness identification may be affected by several factors.” State v. Dyle, supra, 899 SW2d at 610. Since Neil, “additional social science research, as well as actual cases, have taught us much about the fallibility of eyewitness identification and the danger of relying too heavily on eyewitness identification as absolute proof of a defendant’s guilt.” Rimmer v. State, 825 S2d 304, 337 (Fla. 2002) (Pariente, J., dissenting). The Indiana Court of Appeals recognized the average juror is not aware of the likelihood or frequency with which misidentifications are made by seemingly unequivocal witnesses (Reed v. State, 687 NE2d 209, 213 (Ind. App. 1997)), and more recently noted “the growing body of scientific work which has challenged many commonly held beliefs about eyewitness credibility.” *442Jones v. State, 749 NE2d 575, 586 (Ind. App. 2001). See Commonwealth v. Santoli, 424 Mass. 837 (680 NE2d 1116) (1997) (acknowledging the body of scholarly doubt as to the soundness of the assumption that the confidence with which one makes an identification directly correlates with its accuracy).

“[T]he law will always lag behind the sciences to some degree because of the need for solid scientific consensus before the law incorporates its teachings....” State v. Long, supra, 721 P2d at 491. Appellate courts have a responsibility to look forward, and a legal concept’s longevity should not be extended when it is established that it is no longer appropriate. Partridge v. State, 218 Ga. App. 580, 583 (462 SE2d 415) (1995) (Ruffin, J., concurring specially). When identification is an essential issue at trial, appropriate guidelines focusing the jury’s attention on how to analyze and consider the factual issues with regard to the reliability of a witness’s identification of a defendant as the perpetrator are critical. See State v. Robinson, 165 N.J. 32, 41 (754 A2d 1153) (2000). In light of the scientifically-documented lack of correlation between a witness’s certainty in his or her identification of someone as the perpetrator of a crime and the accuracy of that identification, and the critical importance of accurate jury instructions as “the lamp to guide the jury’s feet in journeying through the testimony in search of a legal verdict,” we can no longer endorse an instruction authorizing jurors to consider the witness’s certainty in his/her identification as a factor to be used in deciding the reliability of that identification.8 Accordingly, we advise trial courts to refrain from informing jurors they may consider a witness’s level of certainty when instructing them on the factors that may be considered in deciding the reliability of that identification.

Including the “level of certainty” portion of the charge on reliability of eyewitness identification in the case at bar was harmful error. As stated earlier, the only evidence connecting Brodes to the crimes was the eyewitness identification of him by the two victims, one of whom was unable to pick Brodes’ photo in a photo array, and the other of whom was able to describe to police the weapon used in the crimes but was unable to give any physical characteristics of the perpetrator. We cannot say that the trial court’s instruction to the jury that it might consider the witnesses’ assurances of their certainty that Brodes was the perpetrator did not affect the outcome of the trial. Accordingly, we reverse the Court of Appeals’ affirmance of the judgment of conviction and remand the case with direction that *443the judgment of conviction be reversed and the case remanded to the trial court for a new trial.

Judgment reversed.

All the Justices concur, except Corley, Thompson and Hines, JJ., who dissent.

Notes

The trial court instructed the jury:

Identity is a question of fact for determination by you, the jury. It is dependent upon the credibility of the witness or witnesses offered for this purpose, and you have the right to consider all the facts previously charged regarding credibility of witnesses [“all the facts and circumstances of the case, the witness’ manner of testifying, their intelligence, their interest or lack of interest, their means and opportunity for knowing the facts which they testify about, the probability or improbability of their testimony and of the occurrences which they testify about”]. Some, but not all of the *436factors you may consider in assessing reliability of identification are the opportunity of the witness to view the alleged perpetrator at the time of the alleged incident, the witness’ degree of attention toward the alleged perpetrator at the time of the alleged incident, the level of certainty shown by the witness about his identification, the possibility of mistaken identity, whether the witness’ identification may have been influenced by factors other than the view that the witness claimed to have had, and whether the witness on any prior occasion did not identify the defendant in this case as the alleged perpetrator.

The propriety of the ‘level of certainty’ language in the pattern charge was presented in Brown v. State, 278 Ga. 544 (11) (604 SE2d 503) (2004); Zellars v. State, 278 Ga. 481 (7) (604 SE2d 147) (2004); DeLoach v. State, 272 Ga. 890 (2) (536 SE2d 153) (2000); and Jones v. State, 273 Ga. 213 (3) (b) (539 SE2d 143) (2000), but the merits were not reached because the defendant’s request that the charge be given waived his right to complain on appeal in Brown, Zellars, and DeLoach, and any error in giving the charge in Jones was deemed harmless due to the “significant evidence” of Jones’s guilt other than the eyewitness identification testimony. In Johnson v. State, 272 Ga. 254, 260, n. 6 (526 SE2d 549) (2000), we were not called upon to assess the propriety of including in the pattern charge on eyewitness identification the directive that the jury use an eyewitness’s “level of certainty’ in his or her identification to assess the reliability of the eyewitness identification because the appellant had not asserted error in the giving of the pattern charge or the inclusion of the “level of certainty’ language in the charge. Because we did not address the merits and therefore did not disapprove the charge in DeLoach and Jones, the Court of Appeals has rejected the argument that giving the charge is error. See Darnell v. State, 257 Ga. App. 555 (5) (571 SE2d 547) (2002); Armstead v. State, 255 Ga. App. 385 (2) (565 SE2d 579) (2002). See also Smith v. State, 273 Ga. App. 107 (614 SE2d 219) (2005); Wright v. State, 265 Ga. App. 855 (7) (c) (595 SE2d 664) (2004), and Cummings v. State, 261 Ga. App. 281 (6) (c) (582 SE2d 231) (2003) (failure of trial counsel to seek exclusion of “level of certainty’ charge is not a basis for a finding of ineffective assistance of counsel); Tenorio v. State, 261 Ga. App. 609 (2) (583 SE2d 269) (2003) (giving pattern charge was not error when defendant did not request modifications thereto).

Our observation caused the editors of the third edition of the criminal volume of the Suggested Pattern Jury Instructions to insert a cautionary note concerning the use of the ‘level of certainty’ portion of the charge. Suggested Pattern Jury Instructions, Vol. II: Criminal Cases, Par. 1.35.10 (c), pp. 27-28 (3rd ed. 2003).

The general manager of the restaurant, both at the time of the armed robbery and at trial, testified new lighting was installed in the parking lot in 2002, replacing two street lights that formerly covered the parking lot.

The investigating detective testified the physical lineup was the first one ever conducted by the City of Hapeville.

While Georgia has decided that a jury instruction on eyewitness identification should be given when testimony warrants, other states see such a charge as superfluous when general instructions on witness credibility and burden of proof are given, or reject such an instruction as an impermissible judicial comment on the evidence. See, e.g., Buchanan v. State, 561 P2d 1197, 1207 (Alaska 1977) (not necessary where jury adequately instructed on burden of proof); State v. Valencia, 118 Ariz. 136, 137-138 (575 P2d 335) (1977) (a constitutionally-prohibited comment on the evidence; general instructions sufficient); Conley v. State, 270 Ark. 886 (607 SW2d 328) (1980) (impermissible comment on the evidence); People v. Palumbo, 192 Colo. 7, 11 (555 P2d 521) (1976) (not necessary where jury adequately instructed on witness credibility); State v. Freeman, 380 S2d 1288 (Fla. 1980) (not necessary where jury adequately charged on the State’s burden of proof); State v. Vinge, 81 Haw. 309 (916 P2d 1210, 1217) (1996) (instruction superfluous in light of adequate jury instructions); Brown v. State, 468 NE2d 841, 843 (Ind. 1984) (Indiana law “distinctly biased against jury instructions which single out eyewitness testimony’ for emphasis; general instructions on weighing testimony sufficient); Jones v. Commonwealth, 556 SW2d 918, 921 (Ky. App. 1977) (jury instruction on reasonable doubt covered the topic); State v. Artis, 391 S2d 847, 848 (La. 1980) (instruction unnecessary in light of adequate instructions on general credibility of witnesses); Nevius v. State, 101 Nev. 238, 248-249 (699 P2d 1053, 1060) (1985) (eyewitness instructions are duplicitous of general instructions on witness credibility and proof beyond a reasonable doubt); State v. Paige, 316 N.C. 630, 663 (343 SE2d 848) (1986) (general charge of burden of proof sufficient); State v. Ferguson, 391 NW2d 172 (N.D. 1986) (instructions on burden of proof and witness credibility were sufficient); State v. Classen, 31 Or. App. 683, 693 (571 P2d 527) (1977), rev’d on other grounds, 295 Or. 221 (590 P2d 1198) (1979) (suggested instruction overemphasized the issue and was a comment on the evidence); State v. Payette, 557 A2d 72 (R.I. 1989) (charge on burden of proof of identification sufficient); State v. Robinson, 274 S.C. 198 (262 SE2d 729) (1980) (state constitution forbids a jury charge on facts; charge on burden of proof re identification sufficient); Roy v. State, 627 SW2d 488, 490 (Tex. App. 1981) (charge unnecessary where adequate instructions on burden of proof given); State v. Edwards, 23 Wash. App. 893, 896-897 (600 P2d 566) (1979) (requested charge calls into question witness credibility; charge on burden of proof sufficient); Pearson v. State, 811 P2d 704, 710 (Wyo. 1991) (general instructions on reasonable doubt and witness credibility sufficient). But see State v. Tatum, 219 Conn. 721 (595 A2d 322) (1991) (Connecticut permits a charge on eyewitness identification, noting that trial courts may comment on evidence in Connecticut).

In State v. Long, 721 P2d at 490-491, the Supreme Court of Utah listed a significant number of scholarly articles on the subject. See also Connie Mayer, Due Process Challenges to Eyewitness Identification Based on Pretrial Photographic Arrays, 13 Pace L. Rev. 815 (1994), in which Mayer, a law professor, reviewed 12 articles or studies on the reliability of eyewitness identification.

The suggested instruction attached to United States v. Burrous, 934 FSupp. 525 (E.D.N.Y. 1996) contains several instructive principles on this subject.



Dissenting Opinion

CARLEY, Justice,

dissenting.

In this case, the trial court gave an extensive charge on identity and the assessment of the reliability of eyewitness identification evidence. As part of that comprehensive instruction, the trial court enumerated the following as “[s]ome, but not all of the factors” the jury would be authorized to consider in making its assessment:

the opportunity of the witness to view the alleged perpetrator at the time of the alleged incident, the witness’ degree of attention towards the alleged perpetrator at the time of the alleged incident, the level of certainty shown by the witness about his identification, the possibility of mistaken identity, whether the witness’ identification may have been influenced by factors other than the view that the witness claimed to have had, and whether the witness on any prior occasion did not identify the defendant in this case as the alleged perpetrator.

This list of relevant factors is consistent with that previously recognized by the Supreme Court of the United States, as well as by this Court. Neil v. Biggers, 409 U. S. 188, 199 (III) (93 SC 375, 34 LE2d 401) (1972); Thomason v. State, 268 Ga. 298, 303-304 (3) (486 SE2d 861) (1997). Because these factors have been approved by controlling federal and state authority as relevant to the determination of eyewitness reliability, they have been incorporated into the pattern jury charge on that issue. Suggested Pattern Jury Instructions, Vol. II: Criminal Cases, Par. 1.35.10 (c), pp. 27-28 (3d ed. 2003). Today’s opinion does not hold that it was error to give this pattern charge in its entirety. Compare Renner v. State, 260 Ga. 515, 517 (3) (397 SE2d 683) (1990) (disapproving a charge on flight altogether). Instead, the majority singles out the eyewitness’ “level of certainty” as the only factor which the trial court should have omitted from the instruction. I submit that this holding is not only insupportable under controlling Georgia authority, it also is contrary to the established law in most other jurisdictions. Therefore, I respectfully dissent.

No one disputes that “[t]he vagaries of eyewitness identification are well-known; the annals of criminal law are rife with instances of mistaken identification.” United States v. Wade, 388 U. S. 218, 228 (87 SC 1926, 18 LE2d 1149) (1967). For precisely this reason, Georgia *444permits expert testimony to “assist the jury in assessing the reliability of eyewitness testimony . . ..” Johnson v. State, 272 Ga. 254, 257 (1) (526 SE2d 549) (2000). In fact, the trial court’s erroneous refusal to admit such expert testimony resulted in the reversal of Brodes’ original convictions, and necessitated a new trial on the armed robbery charges. Brodes v. State, 250 Ga. App. 323, 324 (1) (551 SE2d 757) (2001). At the retrial, Brodes presented the testimony of an

expert [who] testified that, among other things, his research shows that there is no correlation between the degree of an eyewitness’s certainty about his identification of a subject and the accuracy of the degree of the eyewitness’s identification. In other words, someone who is highly confident that he has identified a suspect is not more likely to identify the right person than someone who is not as confident.

Brodes v. State, 268 Ga. App. 895, 896 (602 SE2d 895) (2004).

Notwithstanding the acknowledged “vagaries of eyewitness identification” as exemplified by the testimony of Brodes’ expert, an eyewitness is not invariably mistaken. Some eyewitnesses may be wrong when they identify the accused, but not all defendants are misidentified by their victims or by others who had the opportunity to see the crime as it occurred. “ ‘Where [a] witness is unable to positively identify the defendant, his opinion and best judgment were relevant.’ [Cit.]” Randall v. State, 73 Ga. App. 354, 367 (2) (36 SE2d 450) (1945). See also Sanford v. State, 225 Ga. App. 898, 899 (2) (485 SE2d 233) (1997). Therefore, the victims in this case were properly allowed to testify as to the circumstances surrounding the crimes and the certainty of their identification of Brodes as the armed robber. Thus, there was a clear conflict in the evidence as to the reliability of the victims’ testimony, and the jury was required to determine whether to believe them or the defense expert. To assist the jurors in making that determination, the trial court gave the pattern instruction on the factors that they were authorized to consider, including the “level of certainty” shown by the victims in their identification testimony. This instruction was completely neutral, permitting the jury to credit Brodes’ expert as to the “level of certainty’ and thereby discount the victims’ testimony or, on the other hand, to believe the victims’ testimony as to their “level of certainty” and reject the expert’s opinion. Having heard the victims and the defense expert,

“[t]he jury, as in cases of opinion evidence, could have said either that we, upon consideration of the facts detailed by you upon which you base your opinion, do not accept that opinion and will, therefore, disregard it, or that we will give *445it credence. (Cit.)” [Cit.]

Sanford v. State, supra. The expert expressed an opinion as to the unreliability of eyewitness identification generally, and the jury was required to determine whether the specific identification testimony of victims in this case came within or without that general rule.

The majority does not hold that evidence as to a witness’ level of certainty is inadmissible for either the defense or the prosecution. To do so would require overruling Johnson v. State, supra, Brodes v. State, 250 Ga. App., supra at 324 (1), and other cases. What today’s opinion does hold is that a trial court errs when it charges jurors that they may consider that evidence for its intended purpose. The longstanding rule in this state is that “ ‘[w]here there is any evidence, however slight, upon a particular issue, it is not error for the court to charge the law in relation to that issue. (Cits.)’ [Cit.]” Rhodes v. State, 271 Ga. 481, 483 (3) (521 SE2d 579) (1999). The majority completely ignores that principle. Now a trial court errs if it charges the jury that it is authorized to consider the witness’ “level of certainty” as a factor in assessing reliability, notwithstanding the admission of evidence relevant to that very matter. The introduction of evidence, coupled with the prohibition against the trial court’s giving of any instruction as to its relevancy, will undoubtedly create confusion for the jury and leave a gap which the attorneys for the accused and for the State will be tempted to fill during closing arguments. For example, defense counsel could truthfully point out that, although an eyewitness to the crime gave unshakable testimony that he was positive in his identification, the trial court’s instruction will not include that as a possible factor for the jury’s consideration. On the other hand, the prosecutor could argue in all honesty that, even though the defense produced an expert who questioned the general reliability of eyewitness identification, the trial court’s charge will be silent as to the relevancy of that expert testimony. The jury should receive a full and comprehensive instruction on the law from the trial court in the form of an objective charge, not from advocates who are seeking to sway the jury to reach a particular outcome. The majority acknowledges that

“[t]he office of a charge by the court is to give to the jury such instruction touching the rules of law pertinent to the issues involved in a pending trial as will enable them intelligently to apply thereto the evidence submitted, and from the two constituents law and fact make a verdict. (Cit.)” [Cit.]

Majority opinion, p. 438. In accordance with that principle, so long as evidence regarding an eyewitness’ “level of certainty” remains admissible as a factor in determining the reliability of identification *446testimony, the trial court should be authorized to give an instruction that the jury may consider it along with the other factors which are set forth in the pattern charge.

The majority correctly recognizes the existence of a split of authority among the jurisdictions as to whether any instruction on eyewitness identification should ever be given. Majority opinion, p. 439, fn. 6. Some states hold that a charge should be given, whereas others consider such an instruction superfluous or an impermissible judicial comment on the evidence. However, this divergence of opinion does not affect our resolution of this case, since Georgia has long recognized that the jury should be instructed on that issue. Compare Renner v. State, supra (no charge should be given on flight). The question presented for resolution here is how that instruction should be phrased. In holding that the trial court erred, the majority relies upon certain foreign cases which, I submit, actually support the instruction given in this case. See State v. Dyle, 899 SW2d 607,612 (V) (Tenn. 1995) (approving a charge authorizing the jury to consider “[t]he degree of certainty expressed by the witness regarding the identification”); State v. Long, 721 P2d 483, 493, fn. 7 (Utah 1986) (approving a charge that permits the jury to “take into account both the strength of the identification, and the circumstances under which [it] was made”). Those cases do contain language citing scientific opinions expressing doubt as to the reliability of eyewitness identification. However, that language was actually cited in those cases as authority for joining those states, like Georgia, which have authorized trial courts to give jury instructions, including the “degree of certainty,” on the factors which a jury may consider in making its determination. See State v. Dyle, supra at 612 (V) (approving the charge on “degree of certainty expressed by the witness,” after finding that “the pattern identity instruction traditionally given in Tennessee is not adequate in cases where identity is a material issue”); State v. Long, supra at 494 (holding that, even though an instruction on “the strength of the identification” “incorporates some of the fallacious assumptions expressed by the Supreme Court in Neil v. Bigger s[, supra, it] would seem to suffice as an adequate cautionary instruction under most circumstances”). See also State v. Robinson, 754 A2d 1153 (N.J. 2000), which the majority cites on p. 442 as recognizing the necessity for “appropriate guidelines focusing the jury’s attention on how to analyze and consider the factual issues with regard to the reliability of a witness’s identification.” Robinson, supra at 1159 (III) actually approves a “model charge” that lists as a factor “the degree of certainty expressed by the witness in making the identification.”

Of the remaining foreign cases cited by the majority, most have no material bearing on the issue presented for review here, since they *447do not address the giving of a jury charge and merely contain language or citations to treatises which argue the unreliability of eyewitness identification in other contexts. See Krist v. Eli Lilly & Co., 897 F2d 293 (7th Cir. 1990) (no charge issue presented); United States v. Moore, 786 F2d 1308 (5th Cir. 1986) (no charge issue presented); United States v. Burrous, 934 FSupp. 525, 531, 533 (Appendix) (E.D. N.Y. 1996) (which does not expressly disapprove an instruction on “level of certainty,” and which does endorse charge that “it is not essential that the witness him or herself be free from doubt as to the correctness of his or her identification of the defendant”); Lee v. State, 873 S2d 582 (Fla. App. 2004) (no charge issue presented); Rimmer v. State, 825 S2d 304, 316 (Fla. 2002) (no charge issue presented, but also recognizing “ ‘the level of certainty demonstrated by the witness at the confrontation’ ” as a viable factor); Reed v. State, 687 NE2d 209 (Ind. App. 1997) (no charge issue presented).

Indeed, of all the foreign cases cited by the majority, only Commonwealth v. Santoli, 680 NE2d 1116, 1121 (Mass. 1997) actually disapproves an instruction on the “strength of the identification.” However, that case is distinguishable. In Santoli, unlike in this case, the defense was not allowed to present expert testimony which questioned the reliability of the eyewitnesses’ identification testimony. Thus, there, unlike here, the eyewitness testimony was uncontested. The court in Santoli, supra at 845, noted that “[i]t is probably true that the challenged instruction has merit in so far as it deals with the testimony of a witness who expressed doubt about the accuracy of her identification. . . .” Likewise, an instruction on the “level of certainty” would be beneficial to the jury where, as here, there is expert testimony expressing doubt about the reliability of the eyewitnesses’ identification testimony. Such an instruction permits the jury to consider all of the relevant testimony and determine for itself whether the accused has been misidentified.

The majority appears to be motivated by a concern that Georgia is somehow out of step with the current scientific thinking, noting on p. 442 that “ ‘[t]he law will always lag behind the sciences to some degree because of the need for solid scientific consensus before the law incorporates its teachings. . . .’ [Cit.]” The error in relying upon this as a rationale for today’s decision is that no scientific proof currently exists to show that eyewitness testimony is always incorrect. “Scientific studies have amply demonstrated the dangers of mistake in human perception and identification. Of course, this does not mean that the identification in this case is incorrect.” United States v. Burrous, supra at 532 (Appendix) (which was cited with approval by the majority). An expert who lacks personal knowledge can never testify with absolute certainty that any eyewitness is mistaken in his identification of the accused. All that the expert can *448express is an opinion about the reliability of eyewitness identification generally. That opinion does not establish the unreliability of the eyewitness’ identification, but merely provides a frame of reference to apply in considering whether it is reliable. Therefore, the credibility of eyewitness identification testimony should remain, as it always has, a matter for the jury to decide in accordance with legal instruction from the trial court. The law of this state already fully accommodates all scientific views in that regard by permitting the introduction of expert testimony to assist the jurors in assessing for themselves the reliability of the eyewitness. See Johnson v. State, supra. The pattern charge reflects that scientific accommodation.

It is within the common knowledge of a juror that the certainty of a witness’s identification does not necessarily reflect its accuracy, and the language of the instruction makes this clear. The current pattern jury instruction provides that the juror “may consider,” not “must consider,” the certainty of identification. Certainty is only one of the six factors in the pattern jury instruction. Another factor the jury may consider is the possibility of mistaken identity. The juror determines how much weight to give each of the factors. The charge also provides that identification is dependent on the credibility of the witness and that the testimony, facts, and circumstances of the case must identify the defendant beyond a reasonable doubt as the perpetrator.

Armstead v. State, 255 Ga. App. 385, 388 (2) (565 SE2d 579) (2002).

Although today’s decision deals only with the “level of certainty” as a factor in eyewitness identification, it sets a very dangerous precedent. It allows expert opinion testimony to determine the permissible scope of the trial court’s authority to instruct the jury on the legal principles applicable in a case. In effect, the expert now becomes the judge whenever he expresses a general opinion based upon a “ ‘solid scientific consensus____’ [Cit.]” Majority opinion, p. 442. When such an expert opinion is given, the existence of contrary lay testimony based upon personal knowledge becomes immaterial and the trial court is compelled to refrain from instructing the jury on its responsibility for resolving that conflict. I do not believe that one isolated foreign case, which is itself distinguishable, justifies taking such a radical step. The majority may act in the belief that it is placing Georgia in the mainstream of current scientific thought, but it has, in fact, steered us far outside the mainstream of current legal thinking, as exemplified by the very cases cited as authority for its ill-advised *449change of course. See State v. Robinson, supra; State v. Dyle, supra; State v. Long, supra. The judgment of the Court of Appeals should be affirmed.

Decided June 16, 2005. Cromwell & Hibbert, Henry A. Hibbert, for appellant. PaulL. Howard, Jr., District Attorney, Alvera A. Wheeler, Anne E. Green, Assistant District Attorneys, for appellee.

I am authorized to state that Justice Thompson and Justice Hines join in this dissent.