You are mid-conversation when your partner's expression shifts. The warmth you recognised a moment ago is gone, replaced by something flatter, tighter. A colleague who was engaged suddenly becomes quiet, their gaze dropping. Someone smiles through a conversation that is clearly difficult. In each case, something has changed on the face in front of you — and almost immediately, the urge to explain it arrives. They're angry. They're uncomfortable. They're hiding something. The observation may be real. The interpretation is still an interpretation.

I. Why We Read Faces So Quickly

Faces are among the most information-dense surfaces we encounter. A great deal of social life — trust, threat, interest, discomfort — is negotiated through them, and we are strongly disposed to pay attention. When someone's expression changes, we tend to register it rapidly, and we tend to move from something changed to this is what it means with very little pause.

That pause is what this article is interested in. The speed of facial perception is not the problem; the problem is the speed at which we convert perception into certainty. Noticing a change quickly is a genuine social skill. Deciding what it means just as quickly is where most misreading begins.

II. What You See and What You Infer Are Not the Same Thing

Consider how easily three statements that feel identical are actually quite different:

I saw their brows draw together.
They are angry.
They are angry with me.

The first is an observation — a description of something visible. The second is an interpretation, a guess about an internal state. The third is a further interpretation, now layered with a claim about cause and direction. Each step is plausible, but each step is also an inference, and inferences can be wrong at every link in the chain.

Emotional literacy is partly the practice of noticing when you have crossed from one to the next. Most of the time, we do not cross that line deliberately; we simply find ourselves already on the other side, treating they look angry and they are angry as though they were the same kind of statement. They are not. One describes a face; the other describes a mind.

This distinction — between observing and interpreting — is explored in more depth in our Emotional Awareness pathway.

III. What FACS Actually Describes

The Facial Action Coding System, originally developed by psychologists Paul Ekman and Wallace V. Friesen, with a later revision developed with Joseph C. Hager, is a method for describing observable facial movements. It catalogues specific muscle actions — called Action Units — so that a researcher or clinician can record what moved, where, and how strongly.

What FACS does not do is tell you what someone is feeling. It describes facial action, not emotional state. A furrowed brow is a furrowed brow; the leap to anger belongs to the observer, not to the coding system. FACS can support careful description, but it does not prove emotion, intention, honesty, or personality. It cannot, on its own, establish whether someone is being truthful, and it is not a diagnostic instrument.

For a practical reference to common facial-action patterns and their limitations, see our Emotion Atlas.

IV. Why Similar Expressions Are Easy to Confuse

One reason a face cannot be read as a direct emotional readout is that visibly similar movements can accompany quite different states.

Joy-related smile vs social smile

Both involve the corners of the mouth rising. What may differ is that the eye and cheek area may be more involved in smiles associated with enjoyment, while smiles that serve social functions can be more mouth-dominant. But a more mouth-dominant smile is not thereby dishonest. Smiling serves many social functions — politeness, affiliation, easing tension, managing uncertainty — and a smile that serves one of those functions is still a real facial behaviour. The useful question is not real or fake but what is this smile doing in this moment?

Fear vs surprise

Both can involve widened eyes and raised brows. Fear-associated patterns may involve more sustained facial tension, while surprise-associated patterns may involve a more rapid orienting response to something unexpected. These are tendencies, not reliable markers — without knowing what just happened, the two are difficult to distinguish from a single image, and attempting to do so will mislead you often enough to matter.

Anger vs disgust vs contempt

Anger-associated patterns tend to involve brow and eyelid tension, sometimes with tightening in the jaw or mouth. Disgust often involves the nose and upper lip. Contempt is often described in association with Action Unit 14, the Dimpler — a unilateral or asymmetrical movement in which one mouth corner may tighten or pull laterally, while the opposite side remains closer to baseline. As with the other comparisons, this movement does not prove contempt, and reading a single asymmetrical expression as contempt will often be wrong.

V. Baseline and Context Change What a Face Means

A face does not exist in isolation. It belongs to a particular person, with a particular history, in a particular situation — and both the person and the situation change what a given movement means.

People vary considerably in their resting facial presentation. Some carry more habitual tension; others have naturally lower animation. Some have expressive eyebrows that move with nearly every sentence; others have quieter faces that shift little even when emotion is present. A degree of resting asymmetry is normal for many people. If you have never seen someone's face at ease, you do not yet know what their "shifted" looks like — which means you are interpreting without a baseline.

Context reshapes interpretation just as powerfully. What happened in the moments before the change? What is the tone of voice, the posture, the pace of speech? What is the relationship, and what has happened in it before? A furrowed brow across a difficult conversation and a furrowed brow across a difficult spreadsheet may look similar and mean very different things. Cultural, family and professional norms may also influence how openly emotion is expressed, and treating one set of norms as the universal standard will cause you to misread people whose norms differ from yours.

A face should not be interpreted outside the person and the situation it belongs to. When the story you tell yourself hardens into a fact before you have checked it, the face becomes evidence for a conclusion you have already reached.

VI. Why Someone May Smile When They Are Not Happy

The smile is the clearest everyday example of why expression does not equal emotion. People smile when they are happy, yes — but also when they are being polite, when they are nervous, when they are trying to reduce tension, when they want to affiliate, when they feel uncertain, and when they are managing discomfort. A smile can be a greeting, a de-escalation, a hedge, or a habit, sometimes all within the same hour.

These smiles are not necessarily deceptive. A socially regulated smile is a real facial behaviour serving a social function. The mistake is not in noticing the smile; it is in collapsing the smile into a single emotional meaning — they're happy — as though the movement itself had told you everything. It hasn't. It has told you that the corners of the mouth rose. What that rising is for is a separate question, and it usually requires context to answer.

VII. Can Micro-Expressions Reveal Hidden Truth?

Brief facial movements — sometimes called micro-expressions — can occur. A fleeting change may cross someone's face before they settle into a more composed expression. But noticing such a movement does not establish that a concealed emotion is present.

The popular framing — that the face reveals what a person is trying to hide, that micro-expressions expose lies — is not supported in any reliable, standalone way. Micro-expressions are not reliable standalone indicators of deception, and research on human lie detection from nonverbal behaviour has repeatedly found that such cues are faint, inconsistent, and difficult to interpret accurately. A facial cue should not be used as proof of either honesty or dishonesty.

This is where certainty becomes especially hazardous. When people believe they have detected a hidden truth in a fraction of a second, they tend to act on that belief — and the cost of acting on a wrong inference is often higher than the cost of simply not knowing yet. Hypervigilance — scanning constantly for threat signals and reading them with high confidence — can make this pattern worse, not better.

VIII. When Reading a Face Becomes a Relationship Problem

Most of the time, misreading a face is a small, correctable thing. It becomes a relationship problem when it becomes a pattern, and especially when the interpretation is delivered as a statement about the other person's internal state.

"You're angry."

Even if the observation underneath it is accurate — something did change — the sentence leaves no room. It tells the other person what they feel, and it does so in a way that is difficult to disagree with without triggering an argument about whether they are really angry.

"I noticed you became quieter and your expression changed. What happened for you just now?"

This version separates observation from interpretation. It reports what was seen, offers no verdict on what it means, and invites the other person to fill in the gap — which they may do in a way that surprises you. That space for correction can help prevent a single misread expression from becoming a pursuer-withdrawer cycle or an emotional flooding spiral in which both partners end up reacting to something neither of them actually checked.

Repeated certainty about another person's internal state can contribute to defensiveness, reassurance-seeking, escalation or withdrawal. The face becomes a battleground not because it was read, but because the reading was treated as evidence.

IX. The More Useful Skill

The goal of emotional literacy is not to become perfect at reading faces. It is to become better at:

noticing what actually changed
holding several possibilities at once
adding context before concluding
checking assumptions where appropriate

A face can start a question. It should not automatically finish the answer.

If you find that misinterpreting facial cues is a recurring source of tension — in a relationship, at work, or in your own emotional life — the Emotional Awareness pathway and the Emotion Atlas offer a more structured way to practise this. And where repeated misinterpretation is contributing to relationship conflict or emotional difficulty, couples counselling or individual support can help untangle the pattern rather than simply argue about each instance.

Scientific foundations and selected references

Ekman, P., & Friesen, W. V. (1978). Facial Action Coding System: A Technique for the Measurement of Facial Movement. Consulting Psychologists Press.

Ekman, P., Friesen, W. V., & Hager, J. C. (2002). Facial Action Coding System: The Manual.

Barrett, L. F., Mesquita, B., & Gendron, M. (2011). Context in emotion perception. Current Directions in Psychological Science, 20(5), 286–290.

Aviezer, H., Hassin, R. R., Ryan, J., Grady, C., Sutherland, C., Liu, A., et al. (2008). Angry, disgusted, or afraid? Studies on the malleability of emotion perception. Psychological Science, 19(7), 724–732.

Russell, J. A. (1994). Is there universal recognition of emotion from facial expression? A review of the cross-cultural studies. Psychological Bulletin, 115(1), 102–141.

Vrij, A., Hartwig, M., & Granhag, P. A. (2019). Reading lies: Nonverbal communication and deception. Annual Review of Psychology, 70, 295–317.

How Much Can a Face Really Tell You?