False Memories: Impact on Justice Systems


Since the first DNA exoneration in 1989, 312 convictions have been overturned [1]. Nearly 75 percent of the cases were determined based on faulty eyewitness testimonies, and one-third involved two or more incorrect recollections [2]. Each exoneree, an average of 27 years old upon conviction, served around 13.5 years in prison [1]. Furthermore, 18 exonerees served time on death row, and 29 innocent inmates were driven to plead guilty out of desperation [1]. Aside from post-conviction acquittals, tens of thousands of suspects have been pursued because of incorrect witness identification, only to be cleared of charges by virtue of DNA testing [1]. In an effort to alleviate the incarcerated victims’ suffering and subdue their resentment towards the legal system, 65 percent of the exonerees were financially compensated for their prison stints [1]. However, monetary awards do not serve justice, and the true problem, the false memory phenomenon, remains unsolved.

False memories have been studied extensively since DNA testing first uncovered the inaccuracy of witness reports, but scientists have only recently confirmed major hypotheses in this complex area of cognitive neuroscience. A study conducted by researchers at Northwestern University revealed that old memories are constantly updated with sensory information from the current environment [3]. In some cases, the new information corroborates the former, but most times memory is altered, blending together old and new knowledge.

The three main errors in this process are confabulation, misinformation effect, and source-monitoring error. Confabulation is the fabrication of imaginary experiences to fill in blanks in memory; the misinformation effect is the filtration and incorporation of post-event information; and source-monitoring error is the muddling of a memory and its source (e.g. where/when an event occurred).

Illusory recall is easily developed through confabulation during a criminal event. Aside from complications imposed by estimator variables—simple factors like low lighting or distance—a bystander normally does not realize that a crime is occurring until it is well underway. Thus, confabulation is invoked to combine prior observations and current hypotheses to develop an idea of the pre-event proceedings. Confabulation is also colored by personal beliefs and biases, creating an overall picture that is far distanced from observed reality. Meanwhile, the witness begins storing sensory information about the present, lending himself to additional confabulation and reverse suggestibility, whereby pre-event and in medias res observations are merged as a result of simultaneous processing [4].

Assuming the witness’ observations are reasonably reliable, media reports following a crime directly contribute to the misinformation effect and source-monitoring error. The media presents an agglomeration of information taken from various bystanders, crime scene investigators, and surveillance videos. Often times these viewpoints differ from the witness’ recollection, but the witness, still attempting to digest what he has recently seen, accepts the other perspectives and incorporates them into his own memory. According to Dr. Stephen Lindsay, “source confusions are more common when two or more potential sources give rise to highly similar memory records,” so if media reports nearly corroborate witness memories, the specific details are more likely to remain ambiguous [4]. Furthermore, if the crime was fleeting or the witness had difficulty discerning the happenings, the media recounting is likely to resonate more strongly with him because he would have had more time to absorb the information. Although implanted memories coexist with—not replace—the true memories, source-monitoring error can hinder the witness from differentiating between the two.

The repetitive questioning eyewitnesses undergo before a trial leaves them susceptible to retrieval-enhanced suggestibility (RES). Although questioning can help solidify a witness’ memory of a criminal occurrence, any doubts that the witness has are preyed upon. A witness’ susceptibility to misleading information increases with the number of questioning sessions, such as police and media testimonies [5]. However, RES can be hampered by allowing for a larger time gap, ideally a few days, between the crime and the questioning [5]. This benefit cannot commonly be afforded to eyewitnesses however, as police and media need information quickly to formulate reports and launch investigations, but questioning should be minimized.

The final factor that can compromise a witness’ recollection is shock, a common psychological response to an unusual or terrifying event. Shock is an acute stress reaction, so it shares many of the same aftereffects as stress. According to a recent study (Karel), “stress itself reactivates memories even if the memory is unrelated to the stressful experience” [6]. These memories are then rendered moldable and can be enhanced by current experiences and vice versa, abetting false memory development [6].

By the time witnesses are called to testify, their memories of the crime are tainted by personal beliefs, secondhand media reports, and prior questioning. Their recollections are also distorted by each retelling, as they cement the false stories in their memories by memorizing and regurgitating the same story each time. Consequently, confidence in responses, which is used as an indicator of accuracy, is misleading [2,7].

A 2005 study (Tetterton) reveals that judges are biased towards confident witnesses [8]. Participants in the study were divided into two groups: one was given no instructions, and the other was told not to judge based on confidence levels [8]. Regardless of instructions, both groups rated the most confident witnesses as the most credible [8].

Though confidence was an official criterion for witness accuracy in the Supreme Court 30 years ago, its weight has gradually decreased with the emergence of new research. However, a majority of the lower and mid-level courts still count confidence as an almost infallible indicator of a reliable testimony, leaving the door open for false memories to take the floor [8].

Eyewitness accounts remain one of the most essential components of modern criminal justice cases. However, they are susceptible to false memories implanted by imagination, misleading information, repeated questioning, and shock. Confidence in responses can also be merely a reflection of a witness’ conviction in illusory recollections, and should therefore not be given excess importance in the final ruling. We cannot risk wrongly incarcerating suspects, yet neither can we criticize the witnesses. The involuntary false memory phenomenon is the Achilles’ heel of our justice systems.



[1] Innocence Project. “Fact Sheets.” The Innocence Project. Accessed February 14, 2014.

[2] Arkowitz, Hal, and Scott O. Lilienfeld. “Why Science Tells Us Not to Rely on Eyewitness Accounts.” Scientific American. Last modified January 8, 2009. Accessed February 14, 2014.

[3] Bridge, Donna J., and Joel L. Voss. “Hippocampal Binding of Novel Information with Dominant Memory Traces Can Support Both Memory Stability and Change.” Abstract. The Journal of Neuroscience 34, no. 6 (February 5, 2014). Accessed February 14, 2014. doi:10.1523/JNEUROSCI.3819-13.2014.

[4] Lindsay, Stephen, Dr. “Memory Source Monitoring and Eyewitness Testimony.” In Adult Eyewitness Testimony: Current Trends and Developments, edited by David Frank Ross, J. Don Read, and Michael P. Toglia, 27-55. Cambridge: Cambridge University Press, 1994.

[5] Chan, J. C., and J. A. Lapaglia. “The dark side of testing memory: repeated retrieval can enhance eyewitness suggestibility.” J Exp Psychol Appl.
Abstract in J Exp Psychol Appl 17, no. 4 (December 2011). Accessed March 10, 2014. doi:10.1037/a0025147.

[6] Jezek, Karel, Benjamin B. Lee, Eduard Kelemen, Katharine M. McCarthy, Bruce S. McEwen, and André A. Fenton. “Stress-Induced Out-of-Context Activation of Memory.” PLoS Biol 8, no. 12 (December 21, 2010). Accessed March 10, 2014. doi:10.1371/journal.pbio.1000570.

[7] Penrod, Steven, and Brian Cutler. “Witness confidence and witness accuracy: Assessing their forensic relation.” Psychology, Public Policy, and Law 1, no. 4 (December 1995): 817-45. Accessed February 14, 2014. doi:10.1037/1076-8971.1.4.817.

[8] Tetterton, Veronica S., and Amye R. Warren. “Using Witness Confidence can Impair the Ability to Detect Deception.” Criminal Justice and Behavior 32, no. 4 (August 2005): 433-51. Accessed March 10, 2014. doi:10.1177/0093854805276406.

Image Credit:

[1] Creative Commons Attribution-Noncommercial 2.0: Gavel. Photograph. Flickr. January 2, 2006. Accessed May 25, 2014.

[2] Wikimedia Commons and property of US Federal Government: Graner, John. FMRI Scan During Working Memory Tasks. Photograph. Wikimedia
Commons. July 23, 2010. Accessed May 25, 2014.


Anika Mohindra is a sophomore at The Harker School. Follow The Triple Helix Online on Twitter and join us on Facebook.