“You have the right to remain silent.”
It’s one of the most familiar lines in American law, spoken daily in police stations across the country.
There are more than 900 versions of the Miranda warning in use in the United States. All of them are meant to do the same thing: inform individuals in police custody of their constitutional rights—including the right to remain silent and the right to an attorney—before questioning begins. If those rights aren’t clearly conveyed, or clearly understood, anything a person says cannot be used in court.
But comprehension can’t be assumed, especially for people who speak English as a second language.
Scott Jarvis, professor of applied linguistics at Northern Arizona University and holder of the Stoller-Grabe Endowed Distinguished Chair, recalls one interrogation video of a Pakistani man being read his rights: “As soon as [the officer] said, ‘You have the right,’ this man started looking on the table to try to find a pen,” Jarvis said. Like many non-native speakers, the man had heard “right” as “write.”
These misunderstandings aren’t rare, and they have real consequences.
In 2005, Jarvis’s longtime collaborator, University of York linguist Aneta Pavlenko, was asked by a defense attorney to review the interrogation video of a Russian-speaking woman charged with murder. Pavlenko prepared a detailed expert report showing the woman likely hadn’t understood the Miranda warning.
But the judge refused to admit it. “I can tell myself whether the person understood,” he said.
Pavlenko was stunned. She suspected the real issue wasn’t the quality of her analysis, but rather the lack of published research to support it. She reached out to Jarvis and asked if he would help build that foundation.
Together, they designed a study to investigate how well people actually comprehend Miranda warnings. Their findings, published in 2019, revealed that both native and non-native English speakers often misunderstand the Miranda warning, mistakenly believing they comprehend their rights when they do not. Their work earned the Article of the Year Award from the American Association for Applied Linguistics.
In addition to non-native speakers, juveniles and individuals with cognitive impairments, including intellectual disabilities and mental illness, often struggle to grasp the rights being read to them and make consequential legal decisions based on misunderstandings. Children, for instance, may technically understand that they have the right to remain silent but incorrectly assume that if they don’t speak up, they’ll look guilty.
In 2023, Jarvis partnered with fellow NAU linguist Jesse Egbert to expand the scope of their research. This time, they turned to jury instructions.
In most courtrooms, judges read pre-written scripts to jurors explaining the legal standards they must apply in deciding a case. But those scripts are often written by judges and lawyers using complex legalese not intended for the masses.
To address this, Utah’s Supreme Court has established several standing committees to revise and simplify legal language. One of those groups, the Model Utah Jury Instructions Committee, or MUJI, agreed to work with Jarvis and Egbert to test whether their jury instructions were actually comprehensible to the public.
The researchers focused on 11 jury instructions commonly used in civil cases in Utah.
Jarvis and Egbert’s process began with a technique called corpus analysis, which uses massive databases of real-world language—books, newspapers, transcripts, online forums—to measure how common or uncommon particular words and phrases are. If a term appears only once every million words, for example, it’s flagged as potentially difficult for the average person to understand.
Next, they incorporated findings from studies that measure how quickly someone can recognize a word as real. “Turns out, some words take longer to identify than others,” said Jarvis. “And if a judge uses a word that makes you pause for even a second, you’ve already missed the next few words.”
Then came user testing. In surveys, participants were asked to identify terms or phrases they found confusing. Finally, participants would read a simplified case summary, review a sample jury instruction, and decide what verdict they would reach based on the language provided.
The results were troubling, if eye-opening.
Key legal concepts like “preponderance of the evidence,” which simply means a claim is more likely true than not, were widely misunderstood. Participants guessed that it required anywhere from 30 to 95 percent certainty (The correct answer is anything above 50). The phrase “clear and convincing evidence” fared no better. Across the board, people struggled to interpret the thresholds they were being asked to apply.
In the courtroom, misunderstandings like these compromise justice.
Jarvis has served as an expert witness in multiple cases, and, despite the rigor of his research, his testimony is sometimes dismissed.
“In 2020, I did my first case as an expert witness,” he said. “I wrote this very thorough analysis, and the judge said, ‘I don’t really care what the linguist said. I can tell myself.’”
That belief—that comprehension is something a judge can intuitively assess—is precisely what Jarvis is trying to challenge.
Ultimately, Jarvis wants courts to treat linguistic analysis as credible, admissible evidence. He’s building tools and tests that other forensic linguists can use, and he’s training the next generation to do the work, too.
NAU graduate students have played a central role in every stage of his research—from designing studies and recruiting participants to co-authoring publications and presenting findings at national conferences. Students who participate in research are more likely to graduate, have stronger employment outcomes, and have higher entry-level salaries.
That level of support is possible in part because of the Stoller-Grabe Endowed Distinguished Chair, which provides Jarvis with a dedicated research budget. He’s used it to pay student researchers, fund participant stipends, bring guest speakers to campus, and cover tuition for graduate students. He also plans to hire a postdoctoral researcher.
At NAU, Jarvis is part of a department that punches far above its weight.
“The applied linguists in this department are some of the top experts in the world,” he said. “There’s a strong sense of community here, and the faculty are really here for the students and here to prepare them for their future careers.”
He described the atmosphere with a Finnish term: työnilo, meaning “the joy of work.” And when your work helps uphold justice for everyone, that joy is contagious.
