Why Everything You Think You Know About Body Language Is Wrong

There is an industry worth hundreds of millions of dollars built on a single, seductive idea: that the human body is a reliable decoder ring. That crossed arms mean defensiveness. That looking up and to the right means lying. That a firm handshake predicts success.

It is almost entirely fiction.

I spent thirteen years as a criminal intelligence analyst with the Australian Federal Police. I have been trained in behavioural analysis. I have applied it in interview rooms, court proceedings, and intelligence assessments where getting it wrong had serious consequences.

And I can tell you with absolute confidence: the body language you have been taught to read is not body language. It is folk mythology dressed in scientific-sounding language and sold to you by people who have never been in a room where accuracy actually mattered.

Here is what the science actually says — and what it means for how you lead.

The myths, and why they persist

Myth one: Crossed arms mean you are closed off or defensive.

This is perhaps the most widely repeated piece of body language advice in the corporate world, repeated in management training programmes, negotiation workshops, and countless books that should know better.

The research does not support it.

Crossed arms are one of the most context-dependent gestures in the human repertoire. They can indicate defensiveness. They can also indicate that a person is cold. That they are concentrating. That they find the posture comfortable. That they are managing mild anxiety about an unrelated topic. That it is simply a habitual resting position they have adopted since adolescence. OR THEY ARE COLD?!

The gesture alone tells you almost nothing. Without context — baseline behaviour, verbal content, situational factors, and supporting signals — interpreting crossed arms as defensiveness is not reading body language. It is guessing, with the confidence of someone who read a book about it.

Myth two: Looking up and to the right means someone is lying.

This one has been thoroughly and repeatedly dismantled by academic research — most definitively by a 2012 study published in PLOS ONE by Wiseman, Watt, ten Brinke, Porter, Couper, and Rankin, which found no relationship whatsoever between eye direction and lying.

It also contradicts basic neuroscience. The notion that eye movement reliably indicates which cognitive process a person is engaged in — a claim originating in certain schools of Neuro-Linguistic Programming — has not survived scientific scrutiny.

And yet it persists, because it is simple, memorable, and satisfying. It turns a complex human phenomenon into a parlour trick. People want parlour tricks. They are much easier to sell than the truth.

Side note: they are probably recalling information. If you want an actual parlour trick, I recommend asking someone mid conversation their front door colour. They will often look away from you and up - recalling information.

Myth three: You can detect lying from body language.

This is the most damaging myth of all, because it is the one with the highest stakes.

The research on deception detection is both extensive and humbling. A landmark meta-analysis by Bond and DePaulo, examining over two hundred studies and more than twenty thousand participants, found that human beings detect lies at a rate only marginally better than chance — approximately 54 percent, when 50 percent would be random.

Law enforcement professionals, trained specifically in deception detection, perform only fractionally better than untrained civilians. Judges, customs officers, polygraph operators — none of the groups you might expect to be skilled at catching lies show meaningful superiority in controlled studies.

The reason is important: there is no reliable, universal behavioural indicator of deception. There is no single gesture, expression, or vocal pattern that consistently distinguishes a person who is lying from a person who is telling the truth.

What there is — and this is what intelligence training actually develops — is the ability to detect departures from an individual's established baseline, combined with the ability to generate hypotheses about what might explain those departures, and then test those hypotheses rigorously before reaching any conclusion.

That is a far slower, more effortful, and more uncertain process than "they looked to the right so they're lying." Which is precisely why it is not the version that gets sold at conferences.

What television gets catastrophically wrong

The genre of crime and investigation television has done extraordinary damage to public understanding of behavioural analysis.

Shows built around the premise of a protagonist who can detect deception through observation have conditioned millions of people to believe that expert body language reading is fast, reliable, and produces certain conclusions. The protagonist watches someone for a few seconds, catalogues three or four signals, and announces with confidence: lying.

This is not how it works. It has never been how it works.

In real intelligence and investigative settings, behavioural observation is one input among many. It informs hypotheses. It generates questions. It provides context for evaluating other evidence. It never, on its own, produces certainty.

The investigators and analysts I worked alongside who were most skilled at behavioural assessment were also the most cautious about their conclusions. The correlation is not accidental. Understanding the genuine complexity of human behaviour — how many variables are operating simultaneously, how context shapes expression, how individual differences confound generalisation — produces humility, not confidence.

The most dangerous person in any intelligence assessment is the one who is certain they can read people. Certainty of that kind is not skill. It is a failure of understanding dressed as expertise.

What actually works

If the folk mythology does not hold up, what does? What can you actually learn from observing the people around you?

Context is everything. The same gesture means different things in different situations, from different people, at different moments in an interaction. Body language is not a dictionary. It is a language — and like any language, words change meaning depending on the sentence, the speaker, and the conversation they are embedded in.

Baseline is the starting point. The only behaviour that is meaningful to observe is behaviour that deviates from a person's established norm. Before you can identify deviation, you must understand what normal looks like for that specific individual. This takes time, attention, and deliberate observation — none of which is compatible with the quick-read fantasy that television and pop psychology sell.

(Ever been through Customs at the airport and gotten a really random question? Something like what did you eat on the flight or what is the hotel you are staying at? They are looking for a baseline.)

Clusters, not signals. A single signal tells you almost nothing. A cluster of signals — multiple channels pointing in the same direction, sustained across time, occurring in a context that makes them plausible — begins to tell a story. Even then, that story is a hypothesis, not a conclusion.

Verbal and nonverbal together. The most useful information does not come from nonverbal behaviour in isolation. It comes from the relationship between what someone says and what their body does when they say it. Inconsistency between verbal and nonverbal content is more informative than any individual gesture.

Leakage is more reliable than deliberate expression. The most honest information comes not from intentional communication but from the behaviour people do not realise they are producing: the micro-expression before the composed face reassembles itself, the brief postural shift in the moment a difficult question lands, the vocal quality change that occurs independently of the words being spoken.

What this means for how you lead

The implication of all this for leadership is not that behavioural observation is useless. It is that it requires genuine skill — and that genuine skill looks nothing like the confident parlour trick the self-help industry has sold you.

Stop looking for shortcuts. The leader who sits across from someone and thinks they can determine their honesty from a three-second scan has not developed a skill. They have developed a bias — one that will cause them to see what they expect to see rather than what is actually there.

Invest in genuine observation. Slow down. Pay attention to baseline behaviour. Notice the relationship between what people say and how they say it. Treat what you observe as information that requires interpretation, not as conclusions that have already been reached.

Be sceptical of certainty — including your own.

The intelligence analyst who told you they could read a person in thirty seconds was either lying or had not been doing it long enough to understand how wrong they could be.

The best behavioural readers I have ever encountered were the ones who were most aware of the limits of what they could know.

That is not a limitation. That is the skill.