On May 14, 2009, Marlyn DeJesus performed her usual duties at the local Overstock Outlet store. She stocked shelves, kept an eye on customers browsing the aisles, and operated the cash register. She was distracted by this latter job when she looked up long enough to see a somewhat familiar customer enter the store, grab some goods and bring them up to the counter for purchase. As Marlyn rang up the items, the vaguely familiar man held a knife to her throat and demanded that she empty the cash register. She complied, the customer fled with the money and she called 911 to report the robbery. She told the police she recognized the perpetrator as one of her "regular customers" who looked "similar to someone she went to high school with." The police placed a suspect fitting the customer's description in a photo lineup and showed it to Marlyn, who quickly identified Stanley Williams.
At trial, Williams's defense attorneys procured the expert testimony of Dr. Steven Penrod to enlighten the jury about the potential fallibility of eyewitness memory (State v. Williams). However, the trial court denied the defense request to use expert testimony, reasoning that in cases involving an identification of a familiar perpetrator, the possibility of misidentification was so remote as to not warrant expert testimony. The Connecticut Supreme Court agreed, stating that since the perpetrator was familiar to the eyewitness, expert testimony "would not be helpful to the jury" when evaluating the reliability of eyewitness evidence.
State v. Williams raises an important question: How should trial courts deal with expert witness testimony when the eyewitness may have seen the perpetrator before the crime?
Courts recognize that jurors may not get adequate education on the fallibility of eyewitness identifications, and they have embraced social science for its ability to show jurors the errors eyewitnesses make (see State v. Guilbert). Yet — as the appeals court in State v. Guilbert noted — the educational value of expert witnesses on familiar eyewitness fallibility takes a back seat because the "identification of a person who is well-known to the eyewitness does not give rise to the same risk of misidentification of a person who is not well-known to the eyewitness." This raises two questions: How reliable are familiar identifications, and — lacking expert witness testimony — do jurors have adequate knowledge about such potential fallibilities on their own?
There is substantial research on the accuracy of familiar eyewitness identifications. Face recognition research finds that participants are accurate in recognizing and identifying previously seen individuals, especially when compared with identifying never-seen-before individuals (Johnston & Edmonds, 2009). However, familiar identifications are far from foolproof, especially if eyewitnesses have only a vague sense of familiarity. For example, Pezdek and Stolzenberg (2013) found that people sometimes falsely believe a target face is familiar even if they have never seen it before. This study and others suggest that the same factors that reduce the reliability of stranger identifications can reduce the reliability of familiar identifications.
Yet it also raises the psychological question, "How familiar do eyewitnesses have to be to trigger the conclusion that they are familiar enough with the defendant to decrease the risk of misidentification?" That question was raised by a dissenting judge in State v. Williams. Countering the majority's reliance on State v. Guilbert (where four eyewitnesses had known the culprit for years, thereby reducing the likelihood of a misidentification), the dissent noted that Marlyn DeJesus — only five months on the job — could not identify how many times she had seen her "familiar customer," raising reliability concerns regarding a bright-line familiarity rule.
The sparse research on the topic suggests that jurors may lack in-depth knowledge of familiar identifications and underappreciate their accuracy. Recent research by Pozzulo and colleagues even suggests that an eyewitness's prior exposure to the perpetrator has little impact on mock jurors' guilt judgments. This suggests that many jurors lack knowledge regarding the general accuracy rates and complex cognitive processes involved in familiar identification cases, particularly in those cases where the eyewitness has little (if any) prior exposure to the perpetrator. Future research should conduct more direct and in-depth examinations of jurors' beliefs about familiar identifications and their effects on relevant legal judgments.
"Judicial Notebook" is a project of APA Div. 9 (Society for the Psychological Study of Social Issues).
