News

In the popular imagination, a criminal trial reaches its apex during an intense and pointed cross-examination. A skilled lawyer catches a witness in a lie or contradiction and voilà: the facts are revealed, and the innocent defendant walks out of the courtroom, liberated by the truth. This scene appears in films such My Cousin Vinny or A Few Good Men, not to mention materializing nightly on courtroom serials such as Law & Order or Suits. 

But to Duke Law Professor Lisa Kern Griffin, these admittedly dramatic encounters obscure cross-examination’s inherent limits in assessing the substance of a witness’s testimony. In a new paper in the Texas Law Review, “False Accuracy in Criminal Trials: The Limits and Costs of Cross-Examination,” she argues that not only does in-person confrontation not consistently guarantee the truth, but the legal system’s overreliance on in-person confrontation can crowd out other important means for evaluating facts and testing the government’s proof. 

The reliance on in-person confrontation dates back to jurists such John Henry Wigmore, who in the early 1900s described cross-examination as “beyond any doubt the greatest legal engine ever invented for the discovery of the truth.” Griffin, the Candace M. Carroll and Leonard B. Simon Distinguished Professor of Law, who herself has performed cross-examination working on criminal trials at the U.S. Department of Justice, believes it can certainly be useful for testing witnesses and pursuing the defense’s theory of the case. 

“I really understand why advocates rely on it, and why it is especially important to criminal defendants who want to challenge every aspect, large and small, of the government's case,” Griffin said. “I just don't think it does all of the work that we rely on it to do.”

This is in large part because of the way in which the Supreme Court has interpreted (and reinterpreted) the right to cross-examine witnesses under the Sixth Amendment’s Confrontation Clause. The Crawford line of cases purported to expand the confrontation right but has effectively limited its scope to “the opportunity for a mechanical assessment of demeanor during some in-court questioning,” Griffin writes, rather than a substantive examination of the validity of the testimony allowed into the court proceedings. Many of those opinions were authored by the late Antonin Scalia, who emphasized the in-person confrontation over the actual information such an encounter elicits. 

Although Crawford demands that declarants be “present at trial to defend or explain,” Griffin argues that a witness doesn’t actually have to defend or explain anything, merely be physically present. For example, someone can appear for cross-examination but say they have lost their memory, as happened in State v. White, a Louisiana case of a fatal shooting whose primary piece of evidence implicating the defendant was a videotaped statement from a witness.

The witness lost his memory in between giving the taped testimony and appearing at trial, and was “effectively a blank slate on the stand,” Griffin writes, only able to answer questions about his age and his address. The videotaped accusation was nevertheless allowed in as evidence, and the defendant, White, was convicted. If such an appearance satisfies the Confrontation Clause, it has effectively emptied the defendant’s right to rebut the government’s case of any meaningful content.

The idea that cross-examining a witness in court will ultimately reveal the truth is founded on the premise that shrewd questioning can elicit flaws in someone’s testimony, or that someone who is lying (even under oath) will exhibit certain telltale physical signs of their dishonesty, such as fidgeting, picking lint off their clothes, or avoiding eye contact.

“A generation of social science and experimental psychology has thoroughly debunked the notion that nonverbal signals are channels that will give us information about whether someone is lying,” Griffin says. Such misplaced trust in our own innate ability to sniff out a liar can “lead to the admission of problematic evidence that has not otherwise been substantively tested,” she writes. 

This is slowly starting to change with respect to some evidence that has long been treated as definitive. Events that were once inconceivable, such as someone falsely confessing to a crime they didn’t commit, are now acknowledged to be real possibilities. But, Griffin said, “It is a glacial progression in the direction of obvious and longstanding findings from social science.” 

One of these obvious findings is that people evolve over time, and memories can change too. A witness at the time of an alleged crime is a different person when they take the stand weeks or months later. “Memory is tricky and complex,” Griffin said. “Trauma changes the way we view events in the past, and the passage of time will color our perceptions that seem like clear memories.” 

The idea of catching someone in a lie through cross-examination is even more problematic when the witness believes they are telling the truth. Eyewitnesses are a classic example of this. 

“They are the quintessential good faith witnesses,” Griffin explained. “They're not looking to frame anyone, they aren't biased or incentivized.” Griffin says that’s not always the case for testimony involving plea deals, noting about half of all wrongful convictions involve compelling trial testimony from witnesses who were offered deals that “may have inspired falsehoods.” However, she says eyewitnesses by contrast “are just very often mistaken.”

This is part of a broader problem in courtroom evidence generated through in-person encounters that Griffin explores in a forthcoming book, Honest Mistakes: Truth, Lies, and the Misleading Law of Questioning. It will be a difficult problem to address, since our societal preference for in-person witnesses runs even deeper than its codification in the law – our capacity to evaluate people in person is essential to navigating the world. 

“I think the durability of the belief in observational lie detection and in telling whether people are trustworthy by looking at them stems from our human condition, which is that we must know who to trust,” Griffin said. Whether in our dealings with friends, family, colleagues, professionals like doctors or teachers, or even witnesses in court, she says, “We want those to be trusting relationships.”

But our adversarial legal system relies too much on labeling witnesses as honest or dishonest, and on using external signals like demeanor cues during cross-examination to do so. What we really need – from cross-examination and from the criminal legal system more generally, Griffin asserts – is more information about the accuracy of investigations and the reliability of evidence.