Body Language & Nonverbal Communication

The conversation under the words

Most of what I know about reading a room I learned by watching the wrong thing. For years I studied the words. Then I started recording the conversations I was in and playing them back on mute, and a second conversation showed up underneath the first, carried by faces, hands, posture, and the distance people kept. This guide is about that second conversation, and about the three researchers who first measured it.

I ran the mute test on a mock interview panel at a university career centre in 2021, recording five candidates and watching the footage without sound. One candidate I had pegged as nervous looked completely composed: steady shoulders, hands resting, a face that answered before the words did. Another I remembered as confident would not stop touching her hair. The panel's written scores tracked the silent footage more closely than my own memory of the room did. We read each other on a channel we are barely aware of.

That channel is what nonverbal communication describes: posture, gesture, facial expression, eye contact, and the space we hold between ourselves and other people. It shapes workplace interactions, presentations, interviews, and the relationships that matter most. And it is worth saying plainly, because the internet keeps getting this wrong: the popular line that "93 percent of communication is nonverbal" is a misreading of a single set of experiments. The real finding is narrower and more useful, and it starts with a psychologist asking what happens when a person's words and face disagree.

Albert Mehrabian published two studies in 1967 that produced the famous numbers. He had people judge how much someone liked them from a single spoken word, delivered with a matching or mismatched tone and face. When the channels conflicted, listeners leaned on the face (55 percent) and the tone (38 percent) and mostly discounted the word itself (7 percent). That is the entire scope of the finding: liking and disliking, through one word, when the signals fight. Mehrabian spent decades telling people to stop applying the equation to a sales pitch or a lecture. If you are explaining a quarterly report, the words carry the report. But the underlying point survives the misquoting. A manager who says "I'm open to feedback" through a clenched jaw has already answered the question, and that gap between the spoken and the shown is the thing you can actually train.

In 2019 I watched that play out in a single moment. A speaker at a leadership session put up posture research, and roughly two hundred executives who had been slumping straightened up, almost in unison, the second posture became the subject. Nobody told them to. The body adjusts the instant it knows it is being watched, including by its owner, and that instinct is the lever this whole guide turns on. See our active listening guide and public speaking tips for the verbal half of the same work.

When the body goes through a screen

Remote work broke a lot of these signals and invented new ones. On a video call you lose the legs, the hands half the time, the lean toward or away from the table. What you gain is a new set of tells that have nothing to do with the body: how fast you reply, whether the camera is on, the difference between an email that ends "Thanks, Sanjesh" and one that just stops. People read those choices the way they once read a crossed arm. A three-hour silence after a direct question lands as a verdict, even when you were only in another meeting.

Tone-detection features now ship inside email and chat apps, flagging a line that might read as cold before you send it. I find them mildly useful and easy to over-trust. They catch the obvious own-goals; they cannot tell you that a colleague in Tokyo and one in Tel Aviv will read the same blunt sentence in opposite ways. For the parts of remote work that still involve a live face, the old fundamentals hold: open posture, a camera you actually look into, and an expression that matches what you are saying. The medium changed. The thing being communicated did not.

A field guide to the common signals

One warning the rest of this guide keeps repeating: a single cue means almost nothing alone. The American Psychological Association and the researchers I cite treat gestures as a cluster, not a code. Crossed arms plus a turned-away torso plus short answers tells you something; crossed arms in a cold room tells you the air conditioning is on. Read the "where people get it wrong" column as carefully as the one beside it.

How I actually changed my own habits

Reading other people is the glamorous half of this subject. Fixing your own signals is the half that pays off, and it is slower than anyone selling a course will admit. Toastmasters International has put more nervous people in front of more rooms than any organisation alive, and its method is repetition and feedback, not tricks. Here is the version I use, in the order it works.

1. Tape yourself, then watch it on mute. Record an ordinary meeting and play it back with no sound. Sound pulls your attention to the words; silence forces you to see the body. The gap between how you felt and how you look is the whole lesson, and it stings a little.
2. Pick the three things that hurt you most. You will spot a dozen habits. Ignore nine. A flat face in every conversation matters more than a tic that only shows up on stage, so rank by how often the habit appears, not by how much it embarrasses you on tape.
3. Change one thing a week. One. If your problem is crossed arms, spend seven days just keeping your hands loose and let everything else stay broken. Fix all three at once and you turn into a stiff, self-watching mess, worse than when you started.
4. Leave yourself reminders where you will trip over them. A sticky note on the monitor that says only "shoulders." A calendar nudge two minutes before a call. The point is to interrupt the autopilot, because the habit you are fighting runs without your permission.
5. Ask one person who sees you a lot. A teammate, a manager, someone in the room every week. The useful questions are narrow: "Did I look like I was listening?" and "Did my face match what I said?" You cannot see your own resting expression. They can.
6. Re-tape after a month and compare. The before-and-after keeps you going, because the change is usually visible enough to feel earned. Then pick three new habits and start over. It never quite ends, which mostly just means you keep getting better.

The face that gives you away

The most surprising thing I have learned in this field is that a handful of expressions mean the same thing almost everywhere, and Paul Ekman is the reason we know it. In the late 1960s he and Wallace Friesen travelled to the Southeast Highlands of New Guinea to work with the Fore, a people who had had almost no contact with Western media. If facial expressions were learned from films and magazines, the Fore should not recognise an American's surprise or disgust. They did. Ekman and Friesen published the result in 1971 as "Constants across Cultures in the Face and Emotion," and it held up: a short list of emotions, including happiness, sadness, anger, fear, surprise, and disgust, shows up on the face in ways people recognise across wildly different cultures.

Ekman then built a system precise enough to argue about. With Friesen he released the Facial Action Coding System in 1978, which breaks every visible facial movement into numbered "action units": the inner brow raise, the lip corner pull, the cheek raiser that separates a real smile from a polite one. FACS lets a researcher say a specific muscle fired for a specific fraction of a second, and it underpins most credible work on expression today, including the software that now claims to read faces automatically.

Out of that work came the idea most people have heard of: the micro-expression. When someone suppresses a feeling, the true emotion can leak out for as little as 1/25th of a second before the social mask snaps back. In a tense conversation or an interview, that flash of doubt is real information. But this is where the subject gets oversold. Spotting a micro-expression tells you a feeling occurred; it does not tell you why, and it certainly does not let you detect lies, whatever a television drama implied. Ekman himself has been blunt about that limit. Treat a fleeting expression as a question to ask, never a verdict to deliver, and you will use it the way the research supports while you calibrate your own approach in real time.

The space between people

In 1966 the anthropologist Edward T. Hall published "The Hidden Dimension" and gave the field a word it had been missing: proxemics, the study of how we use the space around our bodies to communicate. Hall noticed that Americans carry four invisible bubbles. The intimate zone runs from roughly six to eighteen inches and belongs to lovers, family, and the crowded elevator we all silently endure. The personal zone, about a foot and a half to two and a half feet, is the bubble of friends and easy conversation. Beyond it sit the social and public zones, for colleagues and strangers and anyone addressing a room. What fascinated Hall was that these distances are not universal. They are cultural, learned so early we mistake them for instinct, and they collide constantly when cultures meet.

That collision is easy to watch, and slightly comic, once you know what you are seeing. A colleague from a closer-talking culture steps in; a colleague from a more distant one steps back; the first reads the retreat as coldness and closes the gap again, and the two of them drift across the room in a slow waltz neither one chose. Gestures betray us the same way; the thumbs-up that means "well done" across much of the West is an insult in parts of the Middle East. Eye contact runs on the dial Argyle and Dean measured in 1965, when they found people held more eye contact the farther apart they sat, trading distance for gaze to keep intimacy comfortable. In one culture a dropped gaze reads as respect for a senior; in another it reads as evasion.

I gave up memorising the rules years ago. There are too many, and the list is a trap that makes you stiff. What works is simpler: watch the person in front of you and follow their lead. If they stand close, let them; if they hold your gaze less than you expect, do not chase it. That one habit, mirror rather than correct, covers more ground than any chart of national norms and pairs naturally with strong listening skills for anyone leading a team across cultures.

Reading the room when the room is a webcam

Everything above was worked out for people sharing physical air. A webcam keeps the science and breaks the staging. You are a head and shoulders in a box, your hands usually amputated below the desk, and a fraction of a second of lag quietly wrecks the timing conversation runs on. Hall's distance zones stop applying when the other person is two feet away and a thousand miles away at once. Three small adjustments recover most of what you lose. Get the camera to eye level so you are not filming up your own chin. Sit back far enough that your hands return to the frame when you talk, because a gesture nobody can see may as well not happen. And look into the lens, not the face on your screen, when it is your turn to speak; it feels wrong and reads to the other person as eye contact, which is the trade you want.

The face does more work on camera than in a room, simply because it fills more of the picture. A neutral resting expression that vanishes around a conference table can read as boredom when it is the only thing on screen. I dial my reactions up a notch on calls, a little more nodding and a clearer "yes, I'm with you" on my face, not to perform but to push back against the flattening the medium does to everyone. Our remote communication guide goes deeper on the virtual toolkit. The screen does not excuse you from the body; it just demands a more deliberate version of it.

Questions people actually ask me

Nonverbal cue interpretations vary across cultures and individuals. Use these frameworks as starting points, not absolutes. Read our terms.

Last reviewed: May 24, 2026