Persuasion in the Courtroom: Stories, Memory, and the Limits of Lie Detection

This article is educational and is not legal advice. Procedure and evidence rules differ between jurisdictions. If you are involved in a legal matter, speak to a qualified lawyer.

A trial is one of the few settings in which persuasion is conducted under written rules, in front of a referee, with both sides given equal time. Decades of research in psychology and law show that the courtroom does not suspend ordinary human cognition. It channels it. Understanding how is useful if you are ever called as a juror or a witness, and it makes you a far better reader of trial coverage.

The formal model of a trial is additive: each item of evidence nudges the probability of guilt up or down, and the juror compares the final figure with the standard of proof. Nancy Pennington and Reid Hastie tested that picture in the 1980s by asking mock jurors to think aloud. What they found became known as the Story Model. Jurors take evidence that arrives in a scrambled order, from different witnesses and over several days, and assemble it into a narrative of what happened, with characters, motives, and a causal chain. They fill gaps with their own expectations about how people behave. They then match the finished story to the verdict options the judge describes. Jurors who reached different verdicts had usually not weighed the same story differently. They had built different stories.

In a follow-up experiment Pennington and Hastie held the evidence constant and varied only the order of presentation. When one side's evidence was arranged in story order and the other side's was arranged witness by witness, verdicts shifted sharply toward the side whose case was easier to assemble into a narrative. Nothing about the facts had changed.

This is why trial advocacy manuals tell lawyers on every side, prosecution and defence, plaintiff and defendant, to give the jury a theme and a story in the opening statement, and why the first and last things jurors hear are planned so carefully. It is legitimate advocacy. It also means that a coherent story can feel more proven than it is. The juror's discipline is to ask which links in the narrative were actually supported by evidence and which were supplied by the juror's own sense of how such stories usually go.

Irving Younger's "Ten Commandments of Cross-Examination," delivered in lectures to American lawyers in the 1970s and still taught, reads like a manual for keeping the story in the lawyer's hands. Be brief. Use short questions and plain words. Ask only leading questions. Never ask a question to which you do not already know the answer. Listen to the answer. Do not quarrel with the witness. Do not let the witness repeat their direct testimony. Never allow the witness to explain. Avoid the one question too many. Save the final point for closing argument.

Read from the witness box, the list is a description of what is about to happen to you. The cross-examiner is, in effect, testifying through a series of statements to which you are invited to say "yes." If you are a witness, the practical lessons are modest. Listen to the whole question. Answer the question that was asked, truthfully. If a yes or no would be misleading, say so, and trust that the lawyer who called you will have a chance to ask you to explain on re-examination.

Eyewitness testimony is persuasive because it is delivered by a sincere person who was there. Elizabeth Loftus's research programme, summarized in her 2005 review of thirty years of work on the misinformation effect, shows why sincerity is not enough. Memory is not a recording. It is rebuilt at each retrieval, and information encountered after the event, including the wording of a question, can be folded into it. In the classic study by Loftus and Palmer, people who watched a film of a car accident gave higher speed estimates when asked how fast the cars were going when they "smashed into" each other than when the verb was "hit," and a week later were more likely to report having seen broken glass that was not in the film.

Later studies showed that entire events that never happened can be planted in a minority of participants through suggestion and repetition. The implication for the courtroom is uncomfortable: a witness can be honest, confident, detailed, and wrong, and their confidence on the stand may have grown with each retelling and each piece of confirming feedback. See the entry on eyewitness suggestibility. Confidence recorded at the time of a first, properly conducted identification is a more useful signal than confidence displayed months later at trial.

Few things are as compelling to a jury as a confession. It seems obvious that no innocent person would confess to a serious crime. The record of DNA exonerations says otherwise. In a 2010 review written as a scientific white paper for the American Psychology-Law Society, Saul Kassin and colleagues reported that false confessions had contributed to a substantial minority of the wrongful convictions later overturned by DNA evidence.

The review identifies the risk factors. Some belong to the suspect: youth, intellectual disability, and certain mental health conditions. Others belong to the situation: very long interrogations, sleep deprivation, the presentation of false evidence ("your fingerprints were on the weapon"), and minimization, in which the interrogator offers a morally face-saving version of the crime that implies leniency without ever promising it. These tactics are associated with accusatorial approaches such as the Reid technique. Innocent suspects are, paradoxically, at particular risk, because they tend to waive their rights and keep talking in the belief that the truth will clear them.

The review's central recommendation is that custodial interrogations be electronically recorded from start to finish, so that judges and jurors can see how a statement was obtained rather than only its final form. It also notes experimental evidence that mock jurors are swayed by confessions even when they recognize them as coerced and are instructed to disregard them.

Jurors are told that judging credibility is their job, and many people believe they can tell when someone is lying by watching them. Charles Bond and Bella DePaulo pooled the results of more than two hundred studies and found that people distinguish lies from truths correctly about 54 percent of the time, where chance is 50. The finding held broadly across groups, and professionals such as police officers were not reliably better than students, though they were often more confident.

The cues people rely on, such as averted gaze, fidgeting, and hesitation, are at best weakly related to deception, and are strongly related to nervousness, which is the ordinary condition of a person in a witness box. A juror who discounts a witness for "looking shifty" is measuring anxiety. Better questions are about content: Is the account internally consistent? Does it fit the physical and documentary evidence? Did it change over time, and if so, when and why?

Two further effects are worth knowing. In civil trials, the amount a lawyer asks for acts as an anchor: experimental work on damages has repeatedly found that higher requests produce higher awards, including when the request is broken into a small daily figure (see per diem anchoring). And expert witnesses carry the weight of authority, so that credentials and fluency can stand in for the quality of the underlying method. The juror's question is not "who sounded more impressive?" but "what was actually measured, how, and with what known error rate?"

• Notice the story you have formed, then list which parts were proven and which you filled in.
• Try building the other side's story from the same evidence. If it is also coherent, ask what evidence separates the two, and remember who bears the burden of proof.
• Treat eyewitness confidence at trial with caution; ask how and when the first identification was made.
• For a confession, ask how long the interrogation lasted, whether it was recorded, and whether it contains details that only the perpetrator could have known and that the police did not already have.
• Judge testimony on consistency and corroboration, not demeanour.
• When a number is suggested to you, ask where it came from before you adjust from it.

Courts know about most of these problems. Rules of evidence, jury instructions, and the adversarial structure exist partly to counter them, and reforms such as recorded interrogations and double-blind lineups have spread because of this research. The remaining safeguard is a juror who understands how persuasion works on them. For related tactics, see the Defenses section and the companion post on arguing honestly.