For professionals
When the feelings question draws a blank
A therapist asks, “How do you feel about that?” The autistic client looks away and says, “I don’t know.” From the outside, this can look like resistance, lack of engagement or emotional distance.
With alexithymia, that answer may be the most precise self-report currently available. The emotion is not missing. The route from body signal to feeling word is delayed, unreliable or inaccessible in that moment. This article is not a general introduction to alexithymia. It examines why a standard therapeutic question can target the wrong stage of processing.
The following answers were generated in German in the app on 29 September 2026, unedited and unabridged. The English text is a faithful translation of those answers.
1. “I don’t know” is an answer
The first answer separates missing verbal access from refusal.
User question
An autistic client can only answer the question ‘How do you feel about that?’ with ‘I don’t know’, even though something is visibly happening in their body. Is that refusal, and what mechanism might lie behind it?
Autistic Mirror (outside view/environment, therapist)
"I don't know" from autistic clients in this context is typically not a refusal. It is an honest answer to a question for which the neural infrastructure currently provides no accessible result.
The primary mechanism is alexithymia. This refers to the difficulty of assigning physically experienced states to an emotional concept. The basis lies in the insula, a brain area that collects body signals and integrates them into an internal state image. In many autistic people, the resolution of this signal is lower, more irregular, or delayed. This does not mean that nothing is happening. Something is happening, and it is also physically visible. But the step from a raw body signal to a categorical term such as "fear", "sadness", or "anger" requires a connection that must be automatic and fast to be retrievable during a conversational moment. This connection is not reliably available in alexithymia. What the client actually experiences in that moment is not an inner emptiness, but a lack of translators between what the body reports and what language provides for it.
The underlying bodily mechanism is interoception. This concerns the perception of the body signal itself, therefore a stage before naming it. If this signal arrives with low strength or only in bursts, alexithymia has little to translate. This explains why something is physically visible at the same time and yet no verbal answer is produced: the signal is externally discernible, but internally not present with sufficient clarity to be deliberately poured into language. Alexithymia should be understood here as a prerequisite, the distance between interoceptive signal and linguistic category increases depending on the quality of interoception in that moment.
Another involved element is the question of phatic communication. The question "How do you feel about that?" although it has a functional intention in therapeutic conversations, operates in the language channel as a relational and opening signal, not just as a factual information request. For a monotropic, information-seeking attentional profile, this is a double processing task: to answer referentially and at the same time to manage the social framework of the question. In conjunction with the reduced interoceptive signal, precisely this overlay can lead to everything else also being blocked.
For therapeutic practice, this does not mean a closed door, but an indication of a different approach. Questions that point more strongly to body signals or to concrete situations are often more accessible: "What is happening in your body right now when you think about it?" or "Where do you feel something?" exclude the alexithymia translation step from the task and address interoception directly. Sometimes, a preliminary slowing down also helps: explicitly giving the client time before a verbal answer is expected, because the signal must first arrive before it can be described.
2. The question assumes a ready-made feeling word
The second answer breaks the route down into body signal and naming.
User question
Why can the standard question asking for a feeling word fail in therapy when alexithymia is present, even though the emotion exists?
Autistic Mirror (outside view/environment, therapist)
When a therapist asks, "What are you feeling right now?", this question presupposes a processing pathway that functions differently or is partially absent in cases of alexithymia.
Alexithymia does not describe the absence of emotions, but rather a limited ability to match what is happening in the body to a feeling word. Two separate steps are involved here, which are often confused: first, the perception of a bodily signal, then the categorisation of this signal as "anxiety", "sadness", or "anger". With alexithymia, it is primarily the second step that is disrupted.
The neurological background lies in interoception, the processing of internal bodily states via the insula. This brain region integrates signals such as heart rate, tension in the chest, stomach constriction, or heaviness in the limbs, providing the raw material from which emotions first arise. In autistic people, this signal often arrives with lower resolution or irregularly. This means: the emotion is physiologically real, it changes pulse, breathing rhythm, muscle tension. But it does not enter consciousness as a clearly contoured feeling, but as diffuse bodily unease or emptiness, without the internal translation "that is sadness" reliably taking place.
When the therapist's question for a feeling word now comes, it meets a system that does not have a ready answer. The autistic client may know that something in the body is different, but cannot assign a category to it. This often creates secondary stress: the social demand to produce an answer that is not currently accessible at a neurological level. Many autistic clients then learn to estimate feeling words contextually, i.e., to deduce what one "should" feel in this situation, and to name that. This sounds like openness, but it is compensation, not access.
A more productive approach begins at the physical level: What is happening in the body right now? Is there constriction, warmth, heaviness, or unease anywhere? Where is that located? Such questions bypass the translation step and meet the signal where it is actually present. From this description, patterns can subsequently emerge together, without a feeling word being required in advance that the processing pathway cannot currently deliver.
3. A processing pause and avoidance can look alike
The third answer describes observable differences without turning a single moment into a certain diagnosis.
User question
How can I tell that an autistic client needs more time to assign meaning to body signals rather than emotionally avoiding the subject?
Autistic Mirror (outside view/environment, therapist)
Both phenomena look similar from the outside: a pause, looking away, no verbal feedback. The difference lies in the processing happening behind it.
Interoception describes how body signals such as tension, heart rate, or inner restlessness are processed in the insula and passed on to conscious perception. For many autistic people, this signal arrives with lower resolution, delayed, or in bursts rather than continuously. This means: the information "something is happening inside me right now" is not yet available when you, as the therapist, have already waited and asked. The pause is not a defensive manoeuvre, but a processing step that is still ongoing.
Alexithymia, the difficulty of assigning a feeling-word to an already perceived body signal, often occurs as a co-occurring condition. First, the signal must arrive, then it must be assigned to a category. Both take time, and both are an active cognitive process, not avoidance.
Indicators that processing is taking place and not defence:
The autistic client remains oriented in the room, maintains ground contact, breathes regularly. Their body shows no pattern of contraction or activation towards the exit or a topic change. After the pause, they return to the topic without you having to reintroduce it. What they then say is specifically related to their body ("I feel something in my chest, but I don't know what") instead of abstractly moving on ("that's always been the case"). Even after a longer wait, they do not name anything emotional, but describe body parts, intensities, textures.
Indicators that avoidance is more likely:
The topic jumps, the autistic client introduces new content that has nothing to do with the previous moment. Body language shows activation towards distraction, not a quiet turning inwards. When asked directly, there is no body description, but an explanation or an argument. This pattern consistently appears with certain topics, not generally.
A practical difference in the session: For a processing pause, explicitly stating that there is time, and inviting the autistic client to locate a body part rather than find a feeling-word, helps. With avoidance, the pattern usually does not change through more time alone, but through the context in which the topic appears.
The distinction is never completely certain in one moment. But the pattern over several sessions, especially the question of whether something body-related comes after the pause, is the most reliable indicator.
What is supported in this framing
Alexithymia is an independent construct, not an absence of feelings. Bird and Cook attribute part of the emotional profile once assigned to autism to alexithymia. Kinnaird, Stewart and Tchanturia report a high prevalence of alexithymic traits among autistic adults.
The answers apply those findings to concrete conversations. Statements about the insula, interoceptive resolution and phatic communication are mechanistic models. They are not measurements in an individual person and are not a diagnostic instrument.
The third answer names bodily observations intended to distinguish processing from avoidance. These features offer orientation across several sessions, not a reliable test in a single moment. The final paragraphs of all three answers name concrete question forms and conversational adjustments. They are examples of environmental adaptation, not therapeutic recommendations or substitutes for an individual professional decision. Diagnostic and therapeutic decisions require the person’s full context and the responsible professionals.
Autistic Mirror explains why a feelings question can neurologically draw a blank. It does not diagnose and does not replace therapeutic assessment.
Sources
- Bird & Cook (2013). DOI: 10.1038/tp.2013.61
- Kinnaird, Stewart & Tchanturia (2019). DOI: 10.1016/j.eurpsy.2018.09.004
- Garfinkel, Tiley, O'Keeffe, Harrison, Seth & Critchley (2016). DOI: 10.1016/j.biopsycho.2015.12.003
- Coupland, Coupland & Robinson (1992). DOI: 10.1017/s0047404500015268