Eyewitness Memory and the Reliability of Testimony

The constructive nature of human memory

The dominant theoretical framework in contemporary memory research is the constructivist account, which holds that human memory is not a passive recording device that faithfully preserves experienced events for subsequent retrieval but an active and reconstructive process in which past events are partially encoded and reconstructed at retrieval in ways that are influenced by post-event information, general knowledge, and the specific context of recall. Bartlett (1932) established the foundational insight of constructivist memory theory through his cross-cultural studies of narrative recall, demonstrating that participants' reproductions of an unfamiliar folk story changed systematically over time in the direction of greater conformity with culturally familiar narrative schemas.

Eyewitness testimony and the misinformation effect

Loftus and Palmer (1974) provided one of the most influential demonstrations of memory's vulnerability to post-event suggestion: participants who watched film footage of a car accident and were asked how fast the cars were going when they "smashed" into each other gave significantly higher speed estimates than those asked how fast the cars were going when they "contacted" or "hit" each other, and were significantly more likely to report broken glass in a subsequent memory test, even though no broken glass was visible in the footage. Loftus et al. (1978) demonstrated that exposure to misleading post-event information about a peripheral detail of a witnessed event could replace the original memory of that detail in subsequent recall, a finding with obvious and troubling implications for police interviewing practices that use leading questions. The misinformation effect has been replicated extensively across a wide range of experimental paradigms and has been demonstrated to affect the full range of eyewitness memory content, including descriptions of physical appearance, the presence or absence of objects, and the identity of perpetrators.

Evidence-based investigative procedures

The cognitive interview, first described by Geiselman et al. (1985) and subsequently refined by Fisher and Geiselman (1992), is grounded in the encoding specificity principle: retrieval is facilitated when the conditions at recall approximate the conditions at encoding. The cognitive interview instructs interviewers to reinstate the environmental and psychological context of the event for the witness through mental imagery, to allow witnesses to report all remembered information without imposing chronological or thematic structure, and to attempt recall from multiple temporal perspectives. Meta-analyses of the cognitive interview research demonstrate significant improvements in the quantity of correctly recalled information relative to standard police interviewing techniques, without corresponding increases in the reporting of incorrect information (Kohnken et al., 1999). Wells and Olson (2003) identify a range of system variables, procedural factors under the control of the justice system, that affect the accuracy of lineup identifications. Sequential lineup presentation, in which lineup members are presented one at a time with a forced yes-or-no decision before the next member is presented, encourages absolute judgments and has been shown to reduce false identifications (Steblay et al., 2011).

Conclusion

The research literature on memory and eyewitness testimony demonstrates convincingly that human memory is fallible, malleable, and vulnerable to distortion in ways that have direct implications for the reliability of eyewitness evidence in criminal proceedings. The constructivist account of memory, supported by experimental evidence from Bartlett through Loftus and beyond, provides a theoretical framework that explains the mechanisms of memory distortion while resisting the conclusion that eyewitness testimony is uniformly unreliable. Evidence-based investigative practice, incorporating the cognitive interview, sequential lineup procedures, and double-blind administration, can substantially improve the accuracy of eyewitness evidence while reducing the risk of false convictions based on mistaken identification.

Children as eyewitnesses: special considerations

The reliability of children as eyewitnesses raises distinct considerations that go beyond the general vulnerabilities of adult memory identified above. Children are significantly more susceptible to suggestion than adults, a vulnerability that increases with younger age and that is particularly pronounced when the suggestion comes from an adult authority figure. Ceci and Bruck (1993) conducted a series of influential experimental studies demonstrating that repeated, suggestive questioning by adult interviewers could induce false memories in children for events that had never occurred, including physically implausible events involving bodily touch. The implications for child sexual abuse investigations, which rely heavily on child testimony obtained through interview processes that may involve repeated questioning by multiple professionals, are profound and have driven the development of specialised interview protocols including the NICHD Protocol, which structures investigative interviews with children to minimise suggestive questioning and maximise the amount of information elicited through open-ended prompts.