Some people know that something is going on inside them but cannot say what. The heart beats faster, the neck is tight, the day weighs, and the question "what are you feeling?" gets no answer. Alexithymia is the word for that difficulty in recognising one's own emotions and putting them into words. It is not a diagnosis, not coldness of character, and above all not an absence of feeling.
What the word describes
The term was coined in the early 1970s by the psychiatrist Peter Sifneos, from the Greek: literally, no words for emotions. He was describing patients who could recount their day in exhaustive factual detail but stalled the moment they had to say what that day had done to them.
Alexithymia appears neither in the DSM-5 nor in the ICD-11 as a standalone disorder. It is a transdiagnostic trait: a characteristic spread by degrees across the population, found across very different diagnoses without belonging to any of them. It is a matter of more or less, and the cutoff separating "high" from the rest is a statistical convention. General-population estimates often land around one person in ten, with wide variation by instrument and country.
Three difficulties, one questionnaire
The most widely used instrument is the TAS-20, a twenty-item self-report scale that breaks the trait into three factors.
- Identifying feelings: telling anger from fear, sadness from tiredness, and separating the emotion from the bodily sensations around it.
- Describing feelings to others: finding the words, at the right moment, for another person. Some people do reasonably well internally and get stuck only here.
- Externally oriented thinking: a marked preference for the concrete, for facts and mechanisms, over an inner world.
The three strands do not travel together: someone can run high on the first two and low on the third, which is also the factor whose psychometric properties are most disputed.
One objection comes up regularly, and it is a strong one: asking someone by questionnaire to rate how well they recognise their emotions means asking them to use the very tool they lack. Hence the observer-rated and interview-based versions. A score only means something when it is discussed with a professional.
It is not an absence of emotion
This is the central confusion, and it does real damage. The emotion is there: the body reacts, the mood shifts. What stalls is the recognising and naming step, not the production of feeling. Someone high in alexithymia may cry at a funeral, form deep attachments, and still be unable to explain what just happened inside them.
Stubborn misreadings follow. Silence gets read as disinterest, a flat face as contempt, and "I don't know" as a refusal to cooperate, when it is often the most honest answer available.
The most stigmatising myth turns alexithymia into a synonym for lack of empathy, or a marker of psychopathy. That is wrong and deserves saying plainly. The difficulty concerns one's own emotions first; the slowness in reading other people's, frequent too, looks like delayed decoding rather than indifference to what happens to them. None of it excuses anyone from responsibility: being unable to name what you feel does not license hurting someone.
The body before the words
Many people describe the same route: a physical signal first, the emotion worked out afterwards, sometimes hours later. A clenched jaw turns out to have been anger once the meeting is over. Unexplained tiredness turns out to have been sadness the next day.
That bodily route explains why alexithymia has long been studied on the psychosomatic side of medicine. It also explains a practical risk: without an early emotional signal, you keep going for as long as the body holds, and you discover exhaustion once it has set in.
The idea of a link with interoception, the perception of internal bodily signals, is heavily researched, with findings that are real but heterogeneous: a serious lead, not a settled explanation.
What it overlaps with
Alexithymia is markedly more common among autistic people than in the general population, in a proportion several reviews put at around half. It is not, for all that, an autistic trait: most people high in alexithymia are not autistic, and many autistic people are not alexithymic at all. One influential and still-debated hypothesis goes further, proposing that part of the emotional difficulty long attributed to autism actually belongs to the alexithymia that often accompanies it.
It also turns up at elevated rates in ADHD, depression, anxiety disorders and post-traumatic stress. Hence a useful distinction: so-called secondary alexithymia appears or worsens in the wake of depression, trauma or illness and can recede afterwards, while the primary form looks more like a baseline way of functioning. Only time tells them apart.
Two neighbours worth not confusing it with: anhedonia, the loss of pleasure rather than the loss of words, and the emotional blunting some medications can produce, which is reversible and belongs in a conversation with the prescribing doctor.
... and romantic life
This is where the trait shows up most, because life as a couple leans on the one question alexithymia answers badly: "what are you feeling?". Asked openly, it often produces a blank, then irritation on both sides.
What genuinely helps is rarely dramatic. Closed questions rather than open ones: a choice between two words, or asking whether it is more like tiredness or more like annoyance. Time: accepting that an answer may arrive the next day, without treating the delay as evasion. Writing, which allows the search for words without the weight of being watched. And going through the concrete: what happened, what was hard, what is missing, rather than the name of the emotion. Many couples end up building a vocabulary of their own, sometimes a plain number scale for intensity when the word will not come: not a crutch, a shared language.
The label itself predicts almost none of what matters: not attachment, not faithfulness, not the capacity to take care of someone. Nor does it say who should do the work. Two symmetrical traps wait here: the partner who becomes a permanent translator and burns out, and the person concerned who hides behind the word and stops trying. The trait is dimensional, so it can be worked on: emotional vocabulary does grow with practice or with support, slowly, with no guaranteed result but rarely with none at all.
As for raising it, there is no right moment decided in advance. Naming a concrete need is usually enough: saying you need a delay before answering hard questions tells someone more, and more usefully, than a technical term dropped on the table at a first meeting.