/ Research

Temple Grandin and Thinking in Pictures — What Research Supports

2 November 2026 · CognitionType Research Lab

Maybe you read Thinking in Pictures and felt, for the first time, described. The way you replay a scene frame by frame. The way a spoken instruction slides off you until you can see it. The way words feel like a translation of something that happened first in another format.

Or maybe you read it and felt the opposite. You are autistic, or close to someone who is, and your inner world looks nothing like Temple Grandin's film reel. Perhaps it is quiet, or verbal, or barely visual at all. And you wondered whether you were doing autism wrong.

Both reactions are reasonable, and the research explains why. Grandin's account of her own mind has held up remarkably well. The broader story built on top of it — that autistic people, or some large slice of humanity, are simply "picture thinkers" — is where the evidence gets more interesting.

Who is Temple Grandin and what does thinking in pictures mean

Temple Grandin is a professor of animal science at Colorado State University and one of the most widely known autistic people in the world. She was late to speak as a child, and she has spent five decades explaining her inner experience to people who do not share it. Her center-track restraint system is used to handle close to half the cattle in North America, according to her own biography — a design she says she built by running mental simulations of the equipment before a single part was fabricated.

Her 1995 memoir laid out the core claim in plain terms:

"I think in pictures. Words are like a second language to me."

She describes retrieving memories as specific images rather than general concepts. Ask her to think about a church steeple and she does not get an abstract category. She gets particular steeples she has seen, one after another, like flipping through a slideshow.

In later books — The Autistic Brain (2013, with Richard Panek) and Visual Thinking (2022) — Grandin widened the frame. She proposed that minds come in broad types: photo-realistic object visualisers like herself, pattern or spatial visualisers who think in structures and relationships, and word-fact thinkers. She also stresses that most people are a mix.

Did Temple Grandin assume everyone thought in pictures

Yes, and this detail matters more than it seems. In talks over the years, Grandin has described not realising until her late thirties that most people do not think the way she does. She had assumed her film-reel mind was simply how minds work.

That is a common pattern in research on inner experience. Russell Hurlburt at the University of Nevada, Las Vegas, has spent decades showing that people are often wrong about how they think, and that they tend to assume their own format is universal. Grandin's late discovery is not a quirk of autism. It is what happens when you only have access to one mind.

It also explains why her later books moved from "this is how autistic people think" toward "there are several kinds of thinkers." She revised her model as she encountered more data — including people on the spectrum whose minds worked nothing like hers.

Is there scientific evidence that autistic people think in pictures

The most direct test came early. In 1994, Hurlburt teamed up with Francesca Happé and Uta Frith to sample the inner experience of three autistic adults using beepers that sounded at random through the day. All three reported thoughts that took the form of images, primarily or exclusively, with little of the inner speech that is common in non-autistic samples. Three people is a tiny study, but it was the first systematic look, and it matched Grandin's description closely.

Brain imaging added weight. In 2006, Rajesh Kana, Marcel Just and colleagues published an fMRI study in Brain comparing how autistic and non-autistic adults processed sentences. Some sentences were easy to picture; others were abstract. Non-autistic participants recruited the brain's imagery regions mainly for the picturable sentences. Autistic participants recruited parietal and occipital imagery regions for both kinds — as if they were visualising even when the sentence did not call for it.

Then came reasoning. Isabelle Soulières, Laurent Mottron and colleagues gave Raven's Progressive Matrices — a visual pattern-reasoning test — to 15 autistic and 18 non-autistic adults matched on IQ. Accuracy was the same, but the non-autistic group was about 40 percent slower, and the autistic group showed stronger engagement of perceptual brain regions. A 2012 meta-analysis led by Fabienne Samson, pooling 26 imaging studies, found the same general tilt: more activity in visual and posterior regions in autistic participants, less in frontal regions.

What a careful review of the thinking in pictures hypothesis concluded

The most thorough evaluation came from Maithilee Kunda and Ashok Goel at Georgia Tech in 2011. They reviewed evidence across working memory tasks, serial recall, dual-task experiments, Raven's Matrices, visual search, and more. They opened by noting that the hypothesis had received little sustained attention, partly because it "seems ill-defined." They closed with a measured verdict:

"The idea that certain individuals with autism may 'think visually' should be taken seriously as a cognitive model."

The key word is certain. Kunda and Goel did not conclude that autistic people as a group think in pictures. They concluded that a subset may lean on visual representations for tasks other people handle verbally — and that researchers had not yet worked out how to identify that subset.

They also made a quieter point that deserves attention. Two people can produce identical test scores using completely different internal strategies. A verbal working memory task might be solved by one person rehearsing words and by another holding a picture. The behaviour looks the same. The mind underneath does not.

The twist — aphantasia is more common in autistic adults

If thinking in pictures were a core feature of autism, you would expect autistic people to report unusually vivid imagery. Recent studies point, on average, the other way.

In 2021, Alexander Dance, Julia Simner, Adam Zeman and colleagues found that people with aphantasia — no voluntary mental imagery — reported more autistic traits than controls on the Autism Spectrum Quotient. In 2024, Rachel King, Harry Buxton and Ian Tyndall measured imagery vividness in 121 adults. Autistic participants scored lower on average, and about 20 percent of them fell at the aphantasia cut-off, compared with around 6 percent of the non-autistic group.

And yet a 2024 cluster analysis by Clara Bled, Laurent Mottron, Isabelle Soulières and colleagues, testing 43 autistic and 42 non-autistic adults, found that a strongly visual cognitive profile was the most common profile among the autistic participants — while verbal and visuo-spatial profiles were present too.

These findings are not contradictory. They describe a population that is spread out. Autistic minds appear more likely to sit at the extremes of the imagery spectrum — some extraordinarily visual, some with little or no imagery at all. Grandin is a real data point at one end. She is not the average. If you want to understand the full range from aphantasia to hyperphantasia, autism turns out to be one of the places where that range is widest.

Object visualisers versus spatial visualisers — where the science is strongest

The best-supported part of Grandin's later framework is not actually about autism. It is the distinction between object and spatial visualisers, which she borrows from Maria Kozhevnikov.

In 2005, Kozhevnikov, Stephen Kosslyn and colleagues showed that people who call themselves "visual" split into two groups. Object visualisers generate detailed, colourful, holistic images. Spatial visualisers manipulate structure, location and movement, often in schematic form. Artists tended toward the first; scientists and engineers toward the second.

A 2010 follow-up by Kozhevnikov, Olesya Blazhenkova and colleagues found something stronger: a trade-off. Across five groups of different ages and professions, no group scored above average on both object and spatial visualisation. The authors proposed a developmental bottleneck — getting very good at one kind of visualising seems to come at some cost to the other.

This lines up with Grandin's self-description. She has said she is a vivid object visualiser and that algebra was nearly impossible for her. Her model of herself — vivid pictures, weak abstract symbol manipulation — is consistent with what the lab data predict for someone at the object end of that trade-off.

Where Visual Thinking goes further than the evidence

Grandin's 2022 book makes a broader social argument: that schools and workplaces systematically sideline visual thinkers, and that society loses enormous talent as a result. Parts of that are hard to argue with. Few would dispute that hands-on, mechanical and design skills are undervalued in many education systems.

Other claims rest on thinner evidence. The book cites estimates from psychologist Linda Silverman that roughly a third of students are strongly visual-spatial. Those figures come from questionnaire work rather than objective imagery testing in large, representative samples, and they should be treated as rough. Grandin's enthusiasm for mixing visual and verbal thinkers on teams also deserves nuance: a 2013 study by Ishani Aggarwal and Anita Williams Woolley found that teams mixing object and spatial visualisers were less likely to settle on a shared strategy, and that this was linked to more errors. Diversity of thinking style can be an asset, but it has to be managed.

The biggest risk is a slide into learning styles. Harold Pashler and colleagues' 2008 review found no credible evidence that teaching people in their "preferred" modality improves learning. Grandin does not explicitly endorse that idea, but readers often hear it. The research supports something subtler: people really do differ in the format of their internal representations, but that does not mean visual thinkers can only learn from pictures. We have covered this distinction in depth in why some people think in pictures and others in words.

Thinking in pictures through a cognitive profile lens

Grandin's own account maps onto three cognitive dimensions with unusual clarity.

Visual processing. This is her obvious strength — rich, detailed, high-resolution imagery that lets her test equipment designs in her head. The research on object visualisers suggests this kind of strength is real, measurable, and associated with specific neural resources, not just a turn of phrase.

Memory and sequencing. Grandin describes holding detailed images rather than verbal sequences. In Alan Baddeley's working memory model, that is a mind leaning on the visuospatial sketchpad rather than the phonological loop. Kunda and Goel's review of n-back and serial recall studies suggests some autistic people may route verbal tasks through visual memory — a workaround that can look like normal performance on the outside.

Expression and output. "Words are like a second language" is a statement about translation cost. If your thoughts arrive as images, every conversation, essay and email involves converting them into sequential language. That conversion can be slow, effortful, and draining, even for someone as articulate as Grandin. Many people with strong visual processing and a heavier expression load recognise this experience long before they have words for it.

Seen this way, "thinking in pictures" is not a single trait. It is a particular combination: high visual processing, a memory system that leans visual, and an expression channel that has to work harder to get the picture out.

What you can do if you recognise yourself in Grandin's description

Check your actual format, not your assumed one. Hurlburt's work shows people routinely misjudge their inner experience. Try noticing what is actually in your mind at random moments for a week — images, words, feelings, or something wordless.

Separate vividness from manipulation. You can have vivid pictures and struggle to rotate objects, or the reverse. Knowing which kind of visual thinker you are tells you far more than "I'm visual."

Use visuals as a bridge, not a cage. Sketching ideas before writing, or drawing a diagram before explaining, can reduce the translation cost of expression. But keep building verbal and written skills too — the research does not support avoiding them.

Get a broader picture of your profile. If you want a structured way to see how visual processing, memory and expression interact in your own mind, CognitionType offers a useful tool for understanding your own cognitive profile across seven dimensions. It is not a diagnosis, and it works best alongside, not instead of, formal assessment.

If you suspect autism, seek a proper evaluation. Thinking in pictures is not a diagnostic sign. Plenty of autistic people do not think this way, and plenty of non-autistic people do.

The real lesson of Temple Grandin's mind

Grandin's greatest contribution to this field may not be the claim that she thinks in pictures. It is the insistence that minds differ in kind, not just in degree, and that the differences are worth taking seriously.

The research has largely vindicated her description of herself while complicating the generalisation. Some autistic people think in vivid images; others have almost none. Some visual thinkers are object specialists; others are spatial. The picture that emerges is not two tidy boxes, but a wide spread of cognitive profiles — exactly the kind of variation that labels alone cannot capture.


CognitionType is an informational assessment, not a clinical diagnosis. If you suspect autism, a learning difference, or another specific condition, we encourage you to seek a formal evaluation from a qualified professional.

Discover your own cognitive profile across 7 dimensions.

Take the free assessment