How to Tell If Someone Is Lying: What Truth-Default Theory Actually Says
Why body language fails, why believing other people is usually adaptive, and why evidence beats intuition when the truth matters.
Someone avoids eye contact. Their voice changes. They touch their face. They look nervous. The internet is full of claims that behaviors like these can reveal a liar in real time. The appeal is obvious: if deception had a visible signature, we could learn the signs, watch carefully enough, and protect ourselves.
The research is considerably less reassuring. A major meta-analysis covered 206 documents and 24,483 judges. When people made unaided judgments in real time, they correctly distinguished lies from truths only about 54% of the time. They were also substantially better at recognizing truths than detecting lies: 61% of truthful statements were correctly classified, compared with 47% of lies (Bond & DePaulo, 2006).
A later review of deception research reached a similar conclusion: the nonverbal cues associated with lying are generally faint and unreliable. Nervousness, gaze behavior, or body movements may tell us something about how a person feels. They do not provide a dependable window into whether that person is telling the truth (Vrij et al., 2019).
Timothy R. Levine has spent decades studying this problem and set out his account of it at book length in Duped (2020). His answer is not that humans simply need better lie-detection skills. It is that we may be thinking about deception in the wrong way entirely. In our conversation, he summarized one of his central findings very simply:
“Most lies aren't detected in real time.”
Timothy R. Levine in the interview
Instead, deception is often discovered later, when new information contradicts what we were originally told. That distinction changes the question. Instead of asking How can I read a liar?, it may be more useful to ask: How do I find out what is actually true?
You probably cannot “spot” a liar
Many popular approaches to lie detection assume that deception creates visible leakage. A liar becomes anxious, anxiety produces unusual behavior, and if we become skilled enough at reading that behavior, we can supposedly infer deception.
The problem is that honest people can also be anxious. They may be socially uncomfortable, intimidated, frightened of being disbelieved, struggling to remember something accurately, or simply have an idiosyncratic communication style. Conversely, a deceptive person may be relaxed, articulate, and highly convincing. This is one reason judging truthfulness from demeanor produces so many false positives and false negatives.
Levine describes this in terms of sender demeanor. We do not normally process another person as a checklist of isolated cues. We take in an overall package: confidence, friendliness, tone of voice, fluency, eye contact, gestures, and composure. People who appear confident, friendly, and relaxed tend to be believed more readily. People who appear anxious or socially awkward tend to attract more suspicion. Neither tells us whether the underlying statement is true.
Credibility is not the same as honesty
A person can be credible and honest, credible and deceptive, awkward and honest, or awkward and deceptive. Charm is therefore not evidence of lying, but neither is it evidence of truthfulness. As Levine put it in the interview:
“People tend to believe charming people.”
Timothy R. Levine in the interview
That is a feature of human social perception, not a lie detector.
What is Truth-Default Theory?
Levine's Truth-Default Theory (TDT) begins with a deceptively simple observation: people tend to believe other people. TDT does not treat this tendency as a cognitive defect. It treats it as adaptive.
In his 2014 formulation of the theory, Levine argued that deception detection should be understood primarily through contextualized communication and content, not through supposedly diagnostic nonverbal behaviors. A central premise is that humans normally operate in a truth-default and that this default is functional.
The process is therefore not necessarily:
Someone speaks → I analyze the statement → I decide whether it is true.
It is closer to:
Someone speaks → I generally accept what is being communicated → something may trigger suspicion → only then do I begin evaluating the possibility of deception.
This matters because most of the time, truth and deception are not being consciously evaluated at all. As Levine explained in our conversation, without a trigger, falsity may simply not be “on the radar.”
Why believing other people is usually adaptive
Imagine treating every ordinary communication as a forensic problem. Did your colleague really send the email? Did the waiter actually put in your order? Did your friend really spend the weekend at home? Is the person giving you directions deliberately sending you the wrong way? Did the researcher who published the paper fabricate the data? You could investigate all of these possibilities, and you would accomplish almost nothing else.
Human cooperation depends on our ability to exchange information without independently verifying every statement. Knowledge accumulates because we accept information from teachers, books, scientists, institutions, and other people. Relationships work because every sentence does not need to be proven. Organizations function because most communications can be treated as reliable enough to act upon.
Levine makes exactly this argument in the interview. Effective communication, knowledge transmission, cooperation, and close relationships would become extraordinarily difficult if humans had to continually second-guess everything they were told.
The truth-default does create vulnerability to deception. But the alternative would create a different and arguably much larger problem: pervasive uncertainty. The useful question is therefore not How do I stop trusting people? It is: When should I stop relying on the default?
What breaks the truth-default?
Truth-Default Theory proposes that something has to trigger suspicion. A trigger does not prove that somebody lied. It shifts us from simply accepting a communication to recognizing that something may require closer examination.
Several things can create that shift:
You may already know a fact that conflicts with what you are being told.
The account may seem implausible.
You may become aware of a strong motive for deception.
New evidence may contradict an earlier statement.
Something may simply fail to fit with other known information.
Levine also described noticing vague language, omitted information, and conversational shifts. Importantly, even he does not treat these observations as proof of deception.
Suspicion is permission to investigate, not permission to conclude.
There is a large difference between “Something does not add up” and “This person is lying.” The first can be an appropriate response to uncertainty. The second requires evidence.
Most deception is discovered through evidence, not intuition
One of the most useful implications of Levine's work is that lie detection may be the wrong goal. If somebody successfully deceived you yesterday, your task today is not to reconstruct whether their eyebrows moved strangely while they were speaking. Your task is to establish what happened, and that usually requires information from outside the original interaction.
In the interview, Levine described deception as frequently being uncovered after the fact: evidence appears, and suddenly an earlier statement can be reevaluated. This suggests a practical shift.
Stop trying to read the person. Start testing the claim.
Consider what can actually be verified: dates, documents, messages, financial records, actions, and independent accounts. Does the person's current explanation fit what they said previously? Does their observable behavior match the story?
The quality of the conclusion depends, of course, on the quality of the evidence. Levine emphasized precisely this point: evidence can help detect deception, but unreliable evidence simply creates a different route to a false conclusion.
Why a small number of people matter disproportionately
Another finding complicates the reassuring idea that most communication is honest: lying is not evenly distributed. Levine's research on lie prevalence has repeatedly found a strongly skewed pattern. Most people report telling few lies, while a small minority reports telling a great many (Serota, Levine & Boster, 2010).
A later longitudinal study followed 632 participants over 91 days and again found substantial variation in daily lie prevalence, with a relatively small group accounting for a disproportionate share of reported lies (Serota, Levine & Docan-Morgan, 2022). In our conversation, Levine described one study in which almost 80% of participants averaged roughly 1.3 lies per day, while the top one percent averaged around 15.5. Those specific figures are reported here as Levine described them in the interview.
Two things can therefore be true at the same time: most communication is honest, and some individuals lie far more often than the average person. That distinction becomes especially important in close relationships and workplaces. Population averages tell us what generally happens across many people. But nobody dates a population, and nobody reports to the statistical average manager. We interact with specific individuals.
What happens when the liar is your partner or manager?
This was one of the questions I asked Levine directly. Truth-Default Theory is calibrated to a world in which most communication is honest. What happens when someone repeatedly interacts with a person who is unusually deceptive? Does the default eventually adjust?
Levine was careful here: he had not directly studied that question. His expectation was that adjustment may sometimes occur, but that motivated reasoning can get in the way. We may have reasons not to accept that someone is lying. Recognizing the deception could threaten a relationship, a job, our financial security, our view of another person, or our view of ourselves. In those circumstances, contradictory evidence may be rationalized rather than integrated.
This part of the conversation should therefore not be read as a demonstrated finding that victims of prolific liars always remain in the truth-default. Levine explicitly presented it as an expectation, not as a result from studies of romantic or workplace relationships. It still raises an important practical issue: the more consequential the truth is, the harder it may be to evaluate evidence without also confronting what that evidence would mean.
Pathological and prolific lying are not the same as occasional deception
Everyone lies sometimes, and that fact alone tells us very little about personality or pathology. A socially convenient excuse, an attempt to spare someone's feelings, and a repeated pattern of deception for personal gain are not psychologically identical behaviors.
Levine also distinguished unusually prolific lying from pathological lying. He described a small group of people who appear to lie compulsively, including in situations where there is little obvious benefit and substantial potential cost.
Patterns matter. One inconsistent statement does not establish a deceptive personality, and one awkward answer does not establish manipulation. A diagnostic label should not be reverse-engineered from the fact that somebody has lied. The more useful question is often simpler: What is the pattern of behavior, and what evidence supports it?
So how do you actually find out whether someone is telling the truth?
There is no universal behavioral checklist. There are, however, ways to improve the quality of the information you are working with.
1. Ask questions whose answers you already know
One practical strategy Levine described in the interview is deceptively simple: ask a person something to which you already know the answer. He referred to this as a secret test. You are not measuring whether the person scratches their nose. You are comparing what they tell you with independently established information.
Suppose someone tells you they were not present at a meeting, and you have reliable evidence showing they were. Asking where they were does not magically reveal their internal state. But their answer can be tested against known facts, which is a much stronger basis for judgment than whether they appeared nervous.
This interview advice is related to, but distinct from, another strand of Levine's research. In a series of experiments on diagnostic questioning, Levine and colleagues found that relatively subtle differences in the questions asked could produce large changes in deception-detection accuracy. In the first three studies, diagnostically useful questioning produced accuracy above 70%; in another experiment, accuracy ranged from 32% to 73% depending on the questions (Levine et al., 2014). The broader lesson is not that there is one magic question. It is that better questions can be more useful than better observation.
2. Look for independent evidence
If the issue matters, move beyond the person's account. Can part of the story be checked independently? Are there documents, messages, or records? Are there other witnesses who are not simply repeating the same original source? Does subsequent behavior support the explanation?
The aim is not to assemble a prosecution file every time someone cancels dinner. It is to recognize that when accuracy matters, verification is stronger than impression.
3. Do not disclose everything you know immediately
Research on investigative interviewing offers another useful principle. The Strategic Use of Evidence (SUE) approach examines what happens when interviewers hold evidence but do not immediately disclose all of it. This can make inconsistencies between a deceptive account and the existing evidence easier to see. In one study, police trainees who had learned to use evidence strategically achieved substantially higher deception-detection accuracy than untrained interviewers, partly because deceptive accounts became more inconsistent with the evidence (Hartwig et al., 2006).
This does not mean turning your marriage, friendship, or office into a police interrogation. The research comes from investigative settings with very different goals, procedures, and safeguards. But the underlying logic is useful: if you immediately tell somebody everything you know, you also give them what they need to construct an explanation around it. Sometimes asking an open question first gives you more information.
4. Withhold judgment when the evidence is incomplete
Humans seem strangely uncomfortable with one of the most useful possible answers: I don't know yet. There are at least three reasonable positions: I think this is true; I think this is false; I do not currently have enough information to decide. Suspicion does not require an immediate verdict.
Levine's description of leaving the truth-default explicitly allows for this intermediate state, in which someone becomes suspicious, looks for further evidence, and withholds judgment. That protects against two opposite errors: blindly accepting something because you want it to be true, and deciding someone is deceptive because something merely feels wrong.
5. Look for patterns, not isolated awkwardness
One contradiction deserves attention. A recurring pattern deserves more. Repeated statements that conflict with established facts, shifting explanations, selective omissions, and deception that appears whenever the truth obstructs an important goal all provide a different kind of information than a single nervous conversation.
Even then, describe what you actually know. “This explanation contradicted the evidence” is stronger than “I can tell this person is a pathological liar.” The first is an observation that can be supported. The second may be a conclusion you do not have sufficient evidence for.
6. Separate confidence from accuracy
Confident people are persuasive, but that is not the same as being correct. Anxious people can be truthful, and that is not the same as being deceptive. Whenever your judgment rests heavily on impressions like “He seemed completely sincere” or “Something about her looked guilty,” it is worth asking what evidence exists independently of demeanor. This matters especially when the communicator is highly charismatic, socially skilled, or authoritative.
7. Decide whether the truth matters enough to investigate
Not every potential deception deserves your attention. People tell small lies. They simplify, protect their privacy, and avoid uncomfortable conversations. They say “I have plans” because “I would rather spend tonight alone” feels unnecessarily harsh.
Levine describes calibrating his attention according to the stakes. When something is important, he pays closer attention; when little depends on the answer, he is often willing to let it go. If something carries meaningful personal, relational, professional, financial, or safety consequences, finding out what is true may matter enormously. If nothing meaningful depends on the answer, exhaustive investigation can itself become destructive. The goal is not permanent vigilance. It is appropriate verification.
Speaking up is part of the social defense against deception
Humans do not defend themselves against deception only through individual detection. We also exchange information, and once a deceptive pattern becomes visible, other people can be warned. Levine described this as one reason prolific deception carries social costs: the truth may emerge over time, and other members of a community can share what they have learned.
That creates an uncomfortable tension. Speaking about deception may be costly for the person who was deceived. There may be shame, fear of not being believed, fear of retaliation, or simply a wish not to get involved. But silence also has social consequences. As Levine put it:
“Being [an] honest witness is very good for the larger society.”
Timothy R. Levine in the interview
The point is not that every interpersonal conflict should become a public accusation. It is that reliable information has social value. If people systematically conceal what they know about harmful deception, everyone else loses information that might help them make better decisions.
The bigger problem: when falsehood becomes the consensus
Truth-Default Theory becomes particularly interesting when we move beyond a single conversation. What happens when false information circulates through a group, organization or institution?
In our interview, Levine described two important trigger moderators. The first is what I already believe. The second is what my social group believes. When new information agrees with both, it may generate very little suspicion. In Levine's words:
“When information aligns with what I believe and what everybody in my social circle believes, I am almost certain to believe it, regardless of whether it's true or false.”
Timothy R. Levine in the interview
This creates a more difficult form of vulnerability. When one person tells us something implausible, the contradiction can trigger scrutiny. When the same claim is repeated by people we trust, fits what we already believe, and appears to represent a social consensus, the mechanisms that would normally trigger doubt can weaken.
The problem is then no longer simply whether a particular individual is lying. It becomes: What happens when an entire information environment makes a false claim feel normal? That question cannot be solved by reading facial expressions. It requires independent evidence, competing sources, epistemic humility, and a willingness to revise beliefs that are socially reinforced.
The goal is not distrust. It is better reality testing.
Truth-Default Theory does not tell us that trusting people is foolish. Almost the opposite: trust is what makes ordinary human communication possible, and most of the time, the system works. The danger appears when we confuse that useful default with an obligation to keep believing regardless of contradictory evidence.
You do not need to become hypervigilant. You do not need to learn mythical body-language tells, and you do not need to treat every inconsistency as proof of manipulation. But when something important no longer adds up, you can change strategies:
Ask better questions and check claims against facts.
Look for independent evidence.
Separate confidence from credibility.
Watch patterns rather than isolated moments.
Allow yourself to say I don’t know yet.
And when the evidence becomes strong enough, be willing to update what you believe. The best defense against deception may therefore be less dramatic than learning how to “spot a liar.” It is learning how to find out what is true.
Sources and further reading
Bond, C. F., Jr., & DePaulo, B. M. (2006). Accuracy of deception judgments. Personality and Social Psychology Review, 10(3), 214–234. https://doi.org/10.1207/S15327957PSPR1003_2
Hartwig, M., Granhag, P. A., Strömwall, L. A., & Kronkvist, O. (2006). Strategic use of evidence during police interviews: When training to detect deception works. Law and Human Behavior, 30(5), 603–619. https://doi.org/10.1007/s10979-006-9053-9
Levine, T. R. (2014). Truth-Default Theory (TDT): A theory of human deception and deception detection. Journal of Language and Social Psychology, 33(4), 378–392. https://doi.org/10.1177/0261927X14535916
Levine, T. R. (2020). Duped: Truth-Default Theory and the social science of lying and deception. University of Alabama Press.
Levine, T. R., Clare, D. D., Green, T., Serota, K. B., & Park, H. S. (2014). Diagnostic utility: Experimental demonstrations and replications of powerful question effects in high-stakes deception detection. Human Communication Research, 40(2), 262–289. https://doi.org/10.1111/hcre.12021
Serota, K. B., Levine, T. R., & Boster, F. J. (2010). The prevalence of lying in America: Three studies of self-reported lies. Human Communication Research, 36(1), 2–25. https://doi.org/10.1111/j.1468-2958.2009.01366.x
Serota, K. B., Levine, T. R., & Docan-Morgan, T. (2022). Unpacking variation in lie prevalence: Prolific liars, bad lie days, or both? Communication Monographs, 89(3), 307–331. https://doi.org/10.1080/03637751.2021.1985153
Vrij, A., Hartwig, M., & Granhag, P. A. (2019). Reading lies: Nonverbal communication and deception. Annual Review of Psychology, 70, 295–317. https://doi.org/10.1146/annurev-psych-010418-103135
About the interview
In this episode of Reclaim Your Reality, I speak with Timothy R. Levine about deception, Truth-Default Theory, why most lies go undetected in real time, what triggers suspicion, prolific and pathological lying, trust, relationships, and how falsehoods spread through groups and systems.
Comments