Alexithymia is a relative difficulty identifying and describing your own emotions. The Greek root translates roughly as no words for feelings, which is also how Pierre Sifneos described it in his 1973 paper on psychosomatic patients.1 Sifneos noticed something specific in his clinic: patients who could give precise medical histories went oddly flat when asked what they felt about any of it. The illness had words. The feeling did not.
Half a century later the term has settled into a more careful shape. Alexithymia is a trait dimension, not a disorder, and it is measured on a continuum. Roughly 10% of the general population scores in the alexithymic range on the standard self-report tool. It travels with several other things (autism, eating disorders, certain anxiety presentations) but it is a separate construct from any of them.
Three components, measurable
The most-used measure is the 20-item Toronto Alexithymia Scale, validated by Bagby, Parker and Taylor in 1994.2 The TAS-20 is short, paper-and-pencil, and breaks the construct into three factors. Difficulty identifying feelings: knowing that something is going on inside but not what. Difficulty describing feelings: knowing what it is but not having the words for someone else. Externally-oriented thinking: a habitual focus on what is happening outside rather than what is happening inside.
The cutoff in most studies is a total score of 61 or above. The factor structure has held up across translations and clinical populations, which is one reason the TAS-20 is still the default screening tool thirty years on.
Self-report has the obvious limit. A person who does not notice their feelings well is also the person being asked to rate how well they notice their feelings. Researchers have tried to triangulate with interviews and behavioural tasks, and the construct survives the triangulation, just with appropriate humility about precision.
The autism overlap
The autism connection is the most-discussed piece of the alexithymia literature, and the numbers are worth being specific about. Kinnaird, Stewart and Tchanturia ran a 2019 meta-analysis pooling 14 studies of alexithymia in autistic adults.4 Around half scored in the alexithymic range on the TAS-20, against the ~10% baseline in non-autistic samples. The effect size is large and consistent across studies.
Bird and Cook took the next step in a 2013 Translational Psychiatry paper.5 They argued that several emotional features routinely attributed to autism (difficulty naming one's own feelings, reduced performance on some empathy measures) are better explained by co-occurring alexithymia than by autism itself. When autistic people without elevated alexithymia were compared to non-autistic people without elevated alexithymia, those specific emotional differences shrank or disappeared. The differences load on the alexithymia, not on the autism.
This matters for framing. Autistic adults who struggle to name feelings are not failing at being autistic correctly. They are likely sitting with a separate, measurable trait that happens to be more common in their group. Many autistic people are not alexithymic. Most alexithymic people are not autistic. The labels are useful when they are kept distinct.
The interoception hypothesis
Brewer, Cook and Bird offered a mechanistic account in 2016.3 Their proposal: alexithymia is fundamentally a difficulty reading internal body signals (heart rate, breath, gut, muscle tension, temperature shifts), and emotions are partly composed of those signals plus a label. If the input is faint, the label has less raw material to attach to. Anxious is, mechanically, something like racing heart, tight chest, shallow breath, narrowed attention with a word over the top. Without the body signal, the word floats free.
The interoception story is not the whole picture. Some research finds dissociations between interoceptive accuracy and TAS-20 scores, and the construct has more than one route in. But the hypothesis is useful because it points at where the problem lives: not at vocabulary, not at willingness, but at the input layer.
What this means for everyday tracking
The standard mood-tracking interfaces are quietly hostile to this profile. A 1–10 scale assumes the user can map a faint, confusing internal state onto a number. A vocabulary list of forty emotions assumes the user can produce the right word from scratch. A blank text box assumes the user can put a sentence on something they cannot reliably feel.
The challenge isn't motivation. People with alexithymia are not failing to try. The challenge is input quality, and most apps demand more input precision than the user has available at the moment of logging.
A different design starts lower down. Pick a coarse valence first (a colour, a direction, a rough register) and only narrow if narrowing is possible.
How Colors handles this
Colors is built around a two-tier picker. The first tier is colour: seven valence levels, each with a hue, from a deep red on the worst days to a bright teal on the best. Picking the colour does not require a word. It requires a rough sense of which day is this, which most people can manage even when they cannot say more.
The second tier is the named emotion underneath the colour: anxious, sluggish, lonely, content, focused, restless. If a label fits, it gets attached to the entry. If nothing fits, the colour alone counts as a log. Nothing in the app refuses an entry because the user could not find a word.
The vocabulary is editable. The default emotion list is a starting point, not a constraint. If frustrated feels close but wrong and antsy feels right, antsy gets added and frustrated can be removed. Over time the picker becomes a custom list of the words that actually map onto how this particular user's days go. The granularity ceiling is the user's own vocabulary, not the app's defaults. See emotional granularity for the regulation research that makes precise labelling worth practising.
The Friends feature handles the social side without requiring an explanation. Sharing the colour shows the people who matter where the day is sitting, without asking the user to put the feeling into a sentence. I am a grey today is sometimes all that is available, and that is a usable signal.
A reasonable framing
Alexithymia does not get fixed by an app. The trait is stable, the body-signal layer is what it is, and the words come or do not come on their own schedule. What does change is the daily cost of trying to track anything at all. A colour without a word is a real entry. A custom vocabulary is more honest than a borrowed one. A friend who can read a colour does not need an explanation.
That is what the input-side accommodations are for. Lower the friction at the moment of logging, keep the word optional, and let the rest of the system run on whatever signal is available. For people sitting in the alexithymic range, that is the difference between a tool they keep using and one they quietly stop opening.
Frequently asked questions
What is alexithymia?
Alexithymia, from the Greek for 'no words for feelings', is a trait dimension describing relative difficulty identifying and describing your own emotions, along with a tendency to focus on external events rather than inner states. Pierre Sifneos coined the term in 1973 after working with psychosomatic patients who could describe their illness in detail but struggled to put their feelings into words. It is not a diagnosis. It is measured on a continuum, and roughly 10% of the general population scores in the alexithymic range.
How is alexithymia measured?
The standard tool is the 20-item Toronto Alexithymia Scale (TAS-20), validated by Bagby, Parker and Taylor in 1994. It scores three factors: difficulty identifying feelings, difficulty describing feelings, and externally-oriented thinking. A score of 61 or above is the conventional cutoff for the alexithymic range. Self-report tools have well-known limits — people who do not notice their feelings may not notice they do not notice — so clinical assessment usually pairs the TAS-20 with an interview.
Is alexithymia the same as autism?
No. They are separate constructs that overlap. A 2019 meta-analysis by Kinnaird, Stewart and Tchanturia pooled 14 studies and found that around half of autistic adults score in the alexithymic range, against the ~10% baseline. Bird and Cook argued in 2013 that several emotional difficulties commonly attributed to autism — trouble naming one's own feelings, reduced empathy on certain measures — are better explained by co-occurring alexithymia than by autism itself. Many autistic people are not alexithymic, and most alexithymic people are not autistic.
Can alexithymia be treated?
There is no quick fix, and framing it as something to cure misses the point — it is a trait, not a disease. What can change is the friction around it. Building a personal emotion vocabulary, practising interoceptive awareness (noticing body signals like heart rate, breath, and muscle tension), and using tools that do not require the right word to log a feeling all reduce the daily cost. Therapy that targets emotion identification specifically, such as some adaptations of CBT and DBT, has reasonable evidence; results are gradual.
Why does a colour-first picker help with alexithymia?
Producing the correct emotion word from scratch is the hardest step. Picking a coarse colour for how a moment felt — a red, a grey, a soft blue — does not require the word, only a rough sense of valence. Once the colour is logged the entry already counts. If a more specific label comes, it gets attached; if not, the colour stands alone. This matches the input profile of someone with alexithymia better than a 1–10 scale or a blank text box.
Not medical advice
This article is for informational and educational purposes only. It does not constitute medical advice and should not replace consultation with a licensed mental health professional. If you are in crisis, please contact emergency services in your country immediately.
Crisis lines: US — 988 Suicide & Crisis Lifeline · UK / Ireland — Samaritans 116 123 · EU — Befrienders Worldwide
Last reviewed: May 2026.
References
- Sifneos, P. E. (1973). The prevalence of 'alexithymic' characteristics in psychosomatic patients. Psychotherapy and Psychosomatics, 22(2–6), 255–262. doi:10.1159/000286529
- Bagby, R. M., Parker, J. D. A., & Taylor, G. J. (1994). The twenty-item Toronto Alexithymia Scale: I. Item selection and cross-validation of the factor structure. Journal of Psychosomatic Research, 38(1), 23–32. doi:10.1016/0022-3999(94)90005-1
- Brewer, R., Cook, R., & Bird, G. (2016). Alexithymia: a general deficit of interoception. Royal Society Open Science, 3(10), 150664. doi:10.1098/rsos.150664
- Kinnaird, E., Stewart, C., & Tchanturia, K. (2019). Investigating alexithymia in autism: A systematic review and meta-analysis. European Psychiatry, 55, 80–89. doi:10.1016/j.eurpsy.2018.09.004
- Bird, G., & Cook, R. (2013). Mixed emotions: the contribution of alexithymia to the emotional symptoms of autism. Translational Psychiatry, 3(7), e285. doi:10.1038/tp.2013.61