What Are Microexpressions and Can You Really Learn to Read Them?

The first time I saw a microexpression in a live context, I almost missed it. I was watching a practitioner conduct an interview, and the subject's face did something around the eyes and the corners of the mouth for what I would estimate at roughly a fifth of a second before the composed expression came back like nothing had happened. If you had glanced down at a notepad, it was gone. But if you caught it, what you saw was contempt, sharp and unmistakable, sitting right underneath a polite answer about a topic the person claimed not to care about. That gap, between what the face offered for an instant and what the words said a moment later, is what microexpression reading is actually about.

01 · The definitionWhat is a microexpression?

A microexpression is a brief, involuntary facial expression that flashes across the face when someone experiences a strong emotion they are trying to conceal or suppress. The duration is what separates it from ordinary expression. Where a deliberate social smile or a composed look of concern might hold for several seconds, a microexpression typically lasts somewhere between one twenty-fifth and one fifth of a second. At that speed it is nearly invisible without training.

The involuntary part matters. Microexpressions are produced by the same subcortical neural pathways that generate all genuine emotional expression, circuits that run largely outside conscious control. When a strong enough emotion fires, it leaks through before the deliberate mind can fully compose the face it intended to show. You can decide what expression to wear, but you cannot always decide fast enough to stop the one that escaped first.

That leak is the signal. It is not the whole story, and I will come back to what it does and does not mean, but it is a real signal from a part of the brain that is not in the business of managing impressions.

02 · The scienceWhere does this actually come from?

Paul Ekman's work is the foundation of everything in this field, and the actual history is worth knowing rather than the compressed pop-science version. Ekman and his collaborator Wallace Friesen spent years in the 1960s and 1970s examining whether certain core emotional expressions were truly universal, recognizable across cultures that had not been exposed to the same media or social conventions. Working with populations in Papua New Guinea with minimal prior contact with Western culture, they found that the core emotional displays were consistent. People who had never seen a photograph from another continent could identify the same basic emotions.

That research produced what the field calls the seven universal emotions, each with a specific facial muscle signature:

• Anger. Brows drawn down and together, the upper eyelids tighten, lips press or thin. The look of someone whose limits have been crossed.
• Contempt. The single most distinctive expression in the set, because it is the only asymmetrical one. One corner of the lip pulls up and back. It signals a sense of moral superiority or dismissal and it is one of the most potent signals you can catch in a conversation.
• Disgust. The nose wrinkles, the upper lip rises. Originally a response to something literally foul, but it shows up whenever something violates a person's values or sense of what is acceptable.
• Fear. Brows rise and draw together, the upper eyelids lift, the lower eyelids tighten. The freezing moment before the threat response fires.
• Happiness. The only expression in the set that is reliably different when it is genuine. A real smile, called a Duchenne smile, involves the orbicularis oculi, the muscle around the eye, contracting. Most people cannot voluntarily contract that muscle, which is why fake smiles look like fake smiles.
• Sadness. The inner corners of the brows draw up, the upper lip may quiver. Often the hardest to suppress, which is why it leaks in microform when people are trying to hold it together.
• Surprise. Brows raise and curve, the jaw drops slightly, eyes widen. It is the fastest expression in the set and almost always involuntary. In isolation it tells you almost nothing beyond that the person did not expect what just happened.

Ekman and Friesen went on to develop the Facial Action Coding System, or FACS, which maps every movement the face can make to specific muscles and action units. FACS is still used in animation studios, clinical psychology, and security research as a systematic description of what a face is doing. Microexpressions, in Ekman's framework, are the involuntary version of these universal signals appearing briefly on the surface when emotional control slips for an instant.

03 · The difficultyWhy are microexpressions so hard to catch?

Speed is only part of the problem. The deeper issue is that most people are not trained to observe faces at the right level of detail. In normal conversation we spend the majority of our visual attention on the eyes and the mouth, and we are looking mainly for content and social cues: whether the person is being friendly, whether they agree, whether they are done talking. We are not tracking the specific micro-movements of the muscles around the orbital area or monitoring the nasal wings for the flare that signals contempt.

The face is capable of producing dozens of distinguishable movements, and in everyday interaction we register a fraction of them consciously. The rest goes past us or gets absorbed into a vague gut feeling we cannot explain. Microexpressions are exactly the kind of signal that falls into that unprocessed channel unless you have been trained to look for them at the right layer.

There is also the matter of baseline. Before you can detect an unusual expression, you need to know what a person's face looks like in a neutral, low-stakes moment. Without that reference point, a fleeting expression is noise. Establishing a baseline before anything important happens is one of the first things formal training teaches, and it is one of the things that separates a trained reader from someone who just watched a few YouTube videos about it.

You can decide what expression to wear. You cannot always decide fast enough to stop the one that escaped first.

04 · The training questionCan you actually learn to read them?

Yes. This is probably the most contested and most misrepresented question in the field, so I want to be careful with the honest answer rather than the satisfying one.

Training works, but not in the way most people imagine. The approach that actually builds the skill is repeated, deliberate exposure to a wide variety of faces at slowed-down speeds, combined with gradual reduction of the delay until you can catch the signal at close to real-time. Ekman's training tools were built on exactly this logic: you see an expression slowed down until you can name it reliably, and then the speed comes up incrementally. The goal is not to develop a superhuman ability to freeze time. The goal is to calibrate your attention so that you catch more of the signals that are already visible to a trained eye.

Even with solid training, catching every microexpression in a live conversation is not realistic. The conditions are different depending on the person, the environment, and what you are doing at the moment they flash. What training gives you is a significant improvement in hit rate and, more importantly, a language. Before I had formal training, I was sometimes aware that a face had done something, but I could not tell you what. I just had a vague uncomfortable feeling in response to it. Training gave me a vocabulary: this was the brow compression and the pressed lips of suppressed anger; that was the asymmetric lip pull of contempt. Once you have the vocabulary, even a partial catch becomes interpretable rather than just a feeling.

05 · The limitsWhat can a microexpression actually tell you?

Here is where a lot of people, and unfortunately some of the popular media around this topic, go wrong. I would rather give you the accurate answer than the more exciting one.

A microexpression tells you that someone experienced a strong emotion in that moment, specifically one they were not choosing to show. It does not tell you why they felt it. It does not tell you whether they are lying. It does not tell you what they are about to do. Those are inferences, and inferences have to be grounded in everything else you are observing in context.

Consider fear. A candidate interviewing for a senior role might show a flicker of fear when a compensation question comes up. That could mean they are misrepresenting their salary expectations. It could also mean they are genuinely anxious about asking for too much and losing the opportunity. It could mean that saying the number out loud was the first moment the reality of what they were asking for hit them. Without additional signals and context, fear just tells you that fear was there. The emotion is a flag, not a verdict.

This is why trained behavioral analysts work with clusters rather than single signals. One microexpression is meaningful. A cluster is a conclusion: the microexpression, plus a posture shift, plus a change in speech rhythm, plus a self-soothing gesture a beat later, all pointing in the same direction. No responsible practitioner in this field makes a judgment call on one read. The signals stack, and the stack is what tells you something.

What microexpressions do tell you, with real confidence, is where the emotion lives. When you catch one during a conversation, you know you just touched a nerve, something this person was not planning to let surface. That is valuable, not as a verdict against the person, but as a direction. You know where to look next. You know which topic carries weight for them. You know what is worth coming back to with a softer approach.

06 · The applicationHow I use this in practice

My background in body language and behavioral analysis was built by people from the intelligence and law enforcement world, and I apply it training executives, sales teams, and leaders to read people and influence outcomes more effectively. Microexpressions are one channel in a broader system, never a standalone oracle.

Where I find them most useful is in calibrating the emotional temperature of a conversation in real time. If I catch a flash of contempt when someone verbally agrees with a proposal, I take that agreement as conditional at best. If I see suppressed fear underneath a statement of confidence, I know there is a gap between the stated position and the actual feeling, and my job becomes creating enough psychological safety for the real thing to surface rather than pushing on what I was just told.

Joe Navarro, whose work on the limbic basis of nonverbal communication has influenced how I think about all of this, consistently makes the same point: what you see is in service of the conversation, not a verdict against the person you are reading. The purpose of catching an emotional signal is not to catch someone out. It is to understand them more accurately than their words alone allow, and then to respond to the real situation rather than the managed version of it. That framing is what keeps this skill useful and ethical rather than a tool for manipulation.

Contempt is the signal I see mishandled most often in the corporate world. Leaders and sales professionals catch a flash of it in a meeting and feel the need to address it directly or become defensive. The better move is almost always to note it quietly, soften your own tone, and find a way to give the other person's real concern room to be named. The face told you where the nerve is buried. Your job is to make it safe to speak out loud.