literal thinking

Literal thinking is a precision-oriented processing style common in autistic people, where words, questions, and instructions are interpreted according to their exact meaning rather than their implied or intended meaning. It is one of the most commonly misunderstood autistic traits — both by neurotypical people who assume it means autistic people cannot grasp metaphors or jokes, and by autistic people themselves who dismiss it because they understand figurative language perfectly well.

Many autistic adults comprehend metaphors, sarcasm, and idioms with ease, but still respond very precisely to the literal content of questions, miss the unstated social layer attached to a comment, or get stuck on vague terms like “often” that don’t contain enough information for an accurate answer. Literal thinking shows up most clearly when communication leaves gaps that the listener is expected to fill in — and it becomes far less of a factor when the information provided is clear, specific, and explicit.

What is literal thinking?

Autistic literal thinking is better understood as a processing sequence.

Research shows that many autistic adults comprehend metaphors, idioms, and figurative expressions at similar accuracy levels to neurotypical adults, though often with a measurable processing delay — the literal meaning fires first, then the figurative meaning catches up¹². The brain reaches for the precise, concrete interpretation first because that is the safest, most reliable reading of any input. Contextual, implied, and socially inferred meanings require an additional step that neurotypical processing often skips because it happens automatically.

Taking literal thinking literally

Many autistic adults don’t recognise literal thinking in themselves until specific examples make the pattern visible. This is, somewhat ironically, itself an example of the pattern: “literal thinking” is taken literally to mean “interprets everything literally, all the time,” and since that clearly doesn’t apply, the concept is dismissed. But literal thinking in autism is more nuanced than that, and it shows up in forms that are easy to miss — especially when you’ve been compensating for them your whole life.

Examples of literal thinking in everyday life

Literal thinking in autism takes several distinct forms, each with its own mechanism and its own practical consequences. Let’s see a handful of examples for each.

Precision of interpretation

Many autistic people respond to the exact question that was asked, rather than the question that was intended.

When asked “Do you have a sibling?”, the answer is “yes” — a complete and accurate response to the question as stated. The neurotypical expectation that you’d volunteer additional information (“Yes, I have a brother, he’s three years older, he lives in Manchester”) is an unspoken social convention layered on top of the literal question. The autistic response is precise — it answers exactly what was asked.

This precision shows up constantly in medical settings.

• A physiotherapist asks a patient to raise their arm as high as they can. The patient pushes through the pain and lifts it all the way, because they can raise it that high. The physio concludes a full range of motion. The patient goes home without a referral — not because they hid the pain, but because “as high as you can” was answered precisely. They could. It just cost them.
• A GP asks whether a patient experiences dizziness. The patient says no, because what they experience is the room tilting slightly when they stand up too fast. That doesn’t feel like dizziness — dizziness is surely something more dramatic, more spinning. The symptom goes unreported for years because the clinical word and the lived experience didn’t overlap enough for the patient to recognise them as the same thing.
• A neurologist asks a migraine patient about “spots” or “auras.” The patient says no, because what they experience is their screen getting oddly shiny in one area, or a subtle visual distortion that doesn’t match the textbook description of spots or auras. The symptom goes unreported, not because the patient is withholding information, but because the label didn’t fit the experience precisely enough to trigger recognition.

It shows up just as often in everyday life at home.

• Your roommate asks you to clear the table after dinner. You clear the table. The dishes are now on the kitchen counter, the table is clear, and your roommate is staring at you — because “clear the table” apparently meant “clear the table, rinse the plates, load the dishwasher, and wipe the surfaces.” You did exactly what was asked. The remaining steps were implied by a context that your brain processed as a specific instruction rather than a category of activity.
• Your partner asks you to bring down the trousers from upstairs so they can put a wash on. You bring the trousers. Just the trousers. It turns out “I’m putting a wash on” was meant to signal “bring down everything that needs washing,” but the request named trousers, so trousers are what arrived.
• A new acquaintance “Do you have any hobbies?” and the answer is “Yes.” Again, a full and accurate answer to the question, as it was literally asked. The expected elaboration — naming the hobbies, describing how often you do them, showing enthusiasm — is assumed by the question but not contained in it.

Every one of these is an extremely precise response to the exact information provided. Nothing more, nothing less.

Learned meaning versus contextual meaning

Some forms of literal thinking involve processing a phrase’s component words rather than its conventional meaning.

These interpretations make perfect structural sense — they are logical readings of the words as presented. The conventional meaning is a layer of cultural knowledge that has to be learned separately, and until that learning happens, the literal reading is the only one available.

Once the conventional meaning is learned, most autistic people have no difficulty using it. Understanding comes easily once the conventional meaning has been encountered. The initial acquisition is where the difference lies. This is the same explicit-learning pathway described in our autism entry’s social communication section: neurotypical people absorb conventional meanings through exposure without conscious effort, while autistic people often need to encounter, study, and deliberately learn them.

This also explains why many autistic people are perfectly comfortable with familiar figurative language but can be thrown by novel or ambiguous expressions. If you’ve learned that “raining cats and dogs” means heavy rain, you’ll never think animals are falling from the sky. But encountering a new idiom for the first time can produce a brief moment where the literal image arrives before the intended meaning does.

Delayed social inference

Literal thinking often has an element that you experience long after the interaction is over – the realisation of “oh, this is what they meant!” or “dang it, I sounded uninterested.“

This form of literal thinking is specifically about the layer of social expectation that sits on top of language. Rhetorical questions are answered because they are structured as questions. Teasing is taken at face value because the words say one thing, and the tone (which requires separate, conscious processing) says another. Someone being unkind is not recognised in the moment because the literal content of what they said was neutral — the cruelty was in the subtext, which takes longer to decode.

Social conventions that rely on saying one thing and meaning another are particularly opaque.

For example, a friend hosting a dinner says “Don’t worry about bringing anything, just bring yourself.” You then, as asked, arrive with nothing. You were, after all, told not to bring anything. It turns out “just bring yourself” is a politeness convention that most people interpret as “you don’t have to, but a bottle of wine would be nice.” The literal instruction and the social convention were two different messages, and you received the literal one.

Many autistic adults describe the experience of replaying conversations hours or days later and suddenly understanding what was actually happening. The processing took longer because it had to be done consciously, through reflection, rather than automatically.

Information clarity and benchmarks

There are situations when the words are vague, and without clarity, they are hard to answer factually.

A doctor asks: “Do you experience this symptom often?” This question assumes a shared, intuitive understanding of what “often” means. For many autistic people, it doesn’t land that way. Is daily often? Weekly? For some symptoms, even once a month could be too often. Without a benchmark — a number, a frequency, a comparison point — the question is genuinely unanswerable rather than vaguely answerable. Also, “often” compared to whom? More often than usual compared to my own benchmark, or more often in general, compared to some outer metric?

This, of course, also extends well beyond medical settings. Workplace instructions that rely on implied scales (“try to get this done fairly soon”), emotional check-ins that use undefined terms (“are you feeling okay?”), and questionnaires that offer vague frequency options (“rarely / sometimes / often”) can all produce inaccurate responses — not because the autistic person doesn’t know their own experience, but because the measurement tool is too ambiguous for precise processing to engage with.

The autistic brain wants to give an accurate answer and recognises that the question doesn’t contain enough information to produce one. Neurotypical respondents tend to approximate, going with a gut feeling about what “often” probably means. Autistic respondents are more likely to get stuck on the ambiguity, give an overly cautious answer, or interpret the question more narrowly than intended.

Metaphors, sarcasm, and jokes

One of the reasons many autistic people don’t recognise literal thinking in themselves is that the popular description focuses heavily on figurative language — metaphors, sarcasm, idioms — and many autistic adults handle these competently. Research confirms this: autistic adults often achieve comparable accuracy to neurotypical adults on metaphor comprehension tasks¹. The difference tends to show up not in whether the figurative meaning is understood but in how it’s reached. Eye-tracking studies show a measurable initial bias toward the literal interpretation, followed by successful arrival at the figurative meaning². The metaphor is understood just as well, it just takes a slightly different route to get there.

For many autistic people, figurative language is not only understood but actively valued. Metaphors, analogies, and comparisons can be powerful tools for understanding precisely because they connect an unfamiliar concept to a framework that already makes sense. Having something likened to a known structure provides the kind of concrete anchor that autistic processing works well with. The appeal of a good metaphor is that it is precise — it maps one structure onto another in a way that illuminates both.

Sarcasm is often cited as a particular difficulty, and the research does find more consistent challenges here than with metaphor³. But many autistic adults have learned sarcasm through deliberate exposure — growing up with sarcastic family members, watching comedy, studying the tonal and contextual patterns that signal “the opposite of what I just said is what I mean.” Some become fluent in it. Others find it consistently unreliable, particularly from people they don’t know well, because sarcasm depends heavily on reading tonal cues in real time from a specific person in a specific context. The pattern is harder to learn when the variables keep changing.

The recognition that matters most for late-identified autistic adults is that literal thinking and figurative language comprehension can coexist in the same person without contradiction. You can be a poet who excels at metaphor and still answer assessment questions too precisely. You can be fluent in sarcasm and still miss that a colleague was teasing you. You can understand every idiom in the English language and still take “do you have any questions?” as a genuine enquiry when it was actually a signal that the meeting is over. These are different cognitive operations, and being skilled at one tells you nothing about the others.

What reduces literal thinking?

Clearer communication.

A consistent thread runs through every form of literal thinking described in the examples: the gap between what was said and what was meant.

• When the physiotherapist says “Raise your arm as high as you can,” there is a gap between the instruction and the intent.
• When a friend says “Just bring yourself,” there is a gap between the words and the social convention.
• When a doctor asks about symptoms “often,” there is a gap between the question and any answerable meaning.

Literal thinking steps into these gaps. It bridges missing context by defaulting to the safest, most concrete interpretation available — the actual words. In many ways, this is a rational response to ambiguity. When you can’t be sure what someone means beyond what they said, taking them at their word is the most reliable option.

This means that literal thinking becomes less of a factor when communication is clear, direct, and explicit.

• When the physiotherapist says, “Raise your arm until it starts to hurt, then stop,” the gap disappears.
• Asking “How many days per week do you experience this?” rather than “Do you experience this often?” means there is enough information for precise processing to work with.
• When you want to know about someone else’s hobbies, ask “What is something exciting you recently learned about one of your hobbies?” instead of do they have one.

When someone says what they actually mean, the need to infer what they might have meant goes away.

For autistic adults, recognising this can be a relief. The difficulty was always the constant demand to bridge gaps that other people’s communication left open, rather than any lack of comprehension. Clearer information means less ambiguity, less intolerance of uncertainty, and less cognitive load spent on guessing what was really meant.