Not Every Thought Uses Words

What inner speech research reveals about visual thinking, quiet minds, and why a lack of constant narration does not mean a lack of thought

Thank you for reading, sharing, commenting on, and ***supporting my work. If you enjoy evidence-based explorations of autism, ADHD, AuDHD, inner speech, visual thinking, communication, and the many forms human cognition can take, consider following my work. If this article helped you understand why thought does not always arrive as a sentence, you may also enjoy my book [Evolution in Progress: Rethinking Neurodivergence](https://www.amazon.com/dp/B0F3Y6CCKC)***, where I explore how differences in perception, language, attention, and nervous system processing shape the ways we think, understand ourselves, and navigate the world.

Cover image concept and art direction by J.K. Hamilton. Created using OpenAI’s image-generation tool.

1. Silence Is Not Empty

I used to assume that thinking meant hearing language somewhere inside the mind. A problem appeared, a sentence described it, and another sentence argued with the first. That experience can feel so ordinary that it becomes a theory about everyone. If my thoughts sound verbal to me, I may imagine that another person’s quiet internal world contains fewer thoughts, less reflection, or less self-awareness. The assumption is understandable, but it confuses one form of access with the existence of cognition itself.

Research on everyday inner experience has repeatedly found substantial variation. Descriptive Experience Sampling asks people to record what was present in randomly selected moments and then explores those reports in detailed interviews. In one account of this work, inner speaking ranged from nearly absent to present in most sampled moments, with important differences in whether it felt expanded, condensed, dialogic, or voice-like (Hurlburt et al., 2013). These findings do not give researchers direct access to private thought, and introspective methods have real limitations. They do, however, undermine the claim that a continuous verbal narrator is universal.

My own experience can shift within a single hour. I may think in complete sentences while drafting, in fragments while moving through a familiar routine, and in spatial relationships when trying to understand how several ideas connect. Sometimes I know that something is wrong before I can name what the problem is. That sequence is an anecdote, not proof, but it illustrates the question the science raises. If language is only one possible form thought can take, then silence inside the mind cannot be treated as evidence that nothing is happening.

What else can a thought be before words arrive?

2. Inner Speech Varies

Inner speech is often described as the subjective experience of language without audible articulation. Even that definition contains several possibilities. A thought may sound like a complete statement, appear as a compressed phrase, unfold as a conversation, borrow the remembered tone of another person, or function more like a wordless intention with a verbal edge. The revised Varieties of Inner Speech Questionnaire separates dialogic, condensed, evaluative or critical, positive or regulatory, and other-people-in-inner-speech features rather than treating the inner voice as one uniform phenomenon (Alderson-Day et al., 2018).

A 2024 study proposed the term anendophasia for very low self-reported inner speech. Across four studies, adults reporting low inner speech performed less well on verbal working-memory and rhyme-judgment tasks than adults reporting high inner speech, while task switching and categorical perception did not differ in the predicted ways (Nedergaard & Lupyan, 2024). The sample was modest, the groups were selected from the ends of a self-report distribution, and the study did not establish that anyone literally never thinks in language. A subsequent scholarly commentary challenged whether the evidence justified the stronger claim that some people have no inner voice at all (Lind, 2025). The careful conclusion is that reported inner speech varies widely and may relate to some verbal tasks, not that researchers have discovered two fixed kinds of mind.

This caution matters because labels can turn variation into a new hierarchy. Constant narration is not automatically deeper thought, and quiet is not automatically freedom from worry. A person with frequent inner speech may still struggle to organize it. A person with little verbal narration may reason through images, relations, emotions, movement, or abstract structure. Frequency tells us something about experience, but it does not rank the mind experiencing it.

If inner speech is not one thing, what other forms share the work of thinking?

3. Thought Has Other Forms

Thought can include visual imagery, spatial relationships, bodily sensations, emotional tendencies, remembered sounds, motor plans, and abstract representations that do not present themselves as sentences. Researchers cannot simply place these experiences in separate mental boxes because they interact. In behavioral and brain-imaging experiments, deliberate verbal thinking produced stronger verbal representations than deliberate visual thinking, while visual imagery appeared during both kinds of tasks. The authors described an asymmetry rather than two isolated systems (Amit et al., 2017).

Evidence from neuropsychology also shows that substantial cognition can remain when language is severely impaired. People with profound aphasia have demonstrated abilities in arithmetic, logic, navigation, music, and reasoning about other minds despite major difficulty producing or understanding language. Neuroimaging likewise finds that many nonlinguistic cognitive tasks rely on networks that are separable from the brain’s language system (Fedorenko & Varley, 2016). This does not make language unimportant. Language can organize, transmit, and reshape thought. It means that language and thought cannot be treated as interchangeable.

That difference helps explain an experience many people recognize. I can understand the arrangement of a room before describing it, feel the structure of an argument before finding its vocabulary, or recognize a pattern before I know which details produced the recognition. The knowing may be incomplete or mistaken, but it is not necessarily empty. Words often make the thought easier to inspect and share. They may also force a simultaneous, multidimensional representation into a sequence. What feels immediate inside the mind must become one phrase after another outside it.

What changes when language begins organizing a thought that already exists?

4. Words Reshape Thought

Words do more than report mental content. They can cue categories, highlight relevant features, hold instructions available, and make a vague experience easier to revisit. A major review of inner speech connects verbal thinking with working memory, executive control, metacognition, and self-regulation while emphasizing that the contribution changes across people and tasks (Alderson-Day & Fernyhough, 2015). Language can therefore function as a tool the mind uses on itself.

The effect is visible in ordinary problem solving. Saying “the appointment is Thursday at two” packages several details into a repeatable unit. Naming an emotion can separate it from the entire self. Describing the next step can reduce a cloud of possibilities to one usable action. Theoretical work argues that words act as powerful cues that activate and organize mental representations, allowing people to guide learning and behavior with unusual flexibility (Lupyan & Bergen, 2016). That argument does not mean language creates every concept or that thinking without words is primitive. It means learned language can change what becomes easy to notice, maintain, and coordinate.

I can experience this transformation while writing. An idea may begin as a felt relationship among several observations. Once I name the relationship, I can test it, revise it, and discover where it fails. The sentence does not merely reveal a finished thought. It participates in making the thought more explicit. That is why two claims can both be true: meaningful cognition can occur without language, and language can still expand what cognition is able to do. The tool matters without becoming the whole workshop.

If words can sharpen thought, can they also distort it?

5. Language Can Interfere

Putting an experience into words can improve memory, understanding, and control, but verbalization is not automatically helpful. Some information is difficult to translate without losing the features that made it recognizable. In classic verbal-overshadowing experiments, participants who described a previously viewed face were later less accurate at recognizing it than participants who had not described it. The proposed explanation was not that the visual memory vanished, but that the verbal description biased how people searched the original representation (Schooler & Engstler-Schooler, 1990).

The finding has limits. Verbal overshadowing varies with task design, and replication work has produced more qualified estimates than the original experiments suggested. It should not be turned into a rule that talking ruins memory. The broader lesson is narrower and more useful: a verbal representation may emphasize features that are easy to name while neglecting information that was carried in another format. The description can become easier to retrieve than the experience it was meant to describe.

This may be why explaining a complex intuition sometimes makes me feel less certain. The original understanding was relational and simultaneous. The explanation must select a starting point, decide what to omit, and arrange the rest in sequence. If the sentence feels inadequate, that does not prove the thought was profound or correct. It may reveal that translation is doing real cognitive work. Language can clarify confusion, but it can also create an illusion that whatever cannot be stated quickly was never understood. Respecting nonverbal thought means allowing words to be a model of experience rather than demanding that they reproduce every dimension of it.

What happens when different minds rely on these forms in different proportions?

6. Neurodivergence Complicates Assumptions

Autism, ADHD, and AuDHD do not produce one characteristic inner world. Autistic people may report vivid verbal narration, primarily visual or sensory thought, highly compressed language, difficulty identifying inner speech, or different forms at different times. A theoretical review of neurodiversity and inner speech argues that research should expect variability rather than use one presumed neurotypical pattern as the standard against which every other mind is measured (Alderson-Day & Pearson, 2023).

The same caution applies to visual-thinking stereotypes. Some autistic people describe powerful imagery or spatial reasoning, but autism does not guarantee a picture-based mind. Some ADHD experiences may make external language useful for maintaining a goal, but ADHD does not mean the mind lacks nonverbal thought. Diagnostic labels describe patterns at a population level. They cannot tell us whether a particular person experiences a narrated sentence, a felt direction, a moving image, or several layers at once.

My lived experience can contribute a question without becoming a conclusion. I may speak an instruction because language stabilizes it, then solve the underlying problem through pattern recognition that never becomes fully verbal. Another AuDHD person may experience the sequence differently. Treating either account as the hidden reality of AuDHD would replace one stereotype with another. The more defensible approach asks which forms of representation are available, when they become useful, and what environmental demands make one form easier to access than another. Neurodivergence does not reveal a single alternative to verbal thought. It reveals why a single model of normal thinking was inadequate in the first place.

How much harm begins when visible fluency is mistaken for invisible capacity?

7. Quiet Is Not Deficit

People often judge cognition through its most public output. A quick explanation appears intelligent. A polished verbal answer appears confident. Silence, hesitation, or an unconventional description may be read as weak understanding. Yet evidence separating language from other forms of cognition shows why this inference is unsafe. Many complex abilities can remain available even when language production or comprehension is severely disrupted (Fedorenko & Varley, 2016).

The same principle applies less dramatically in everyday life. A person may need time to translate an image into language, convert a bodily warning into an emotion word, or decide which part of a complex pattern to explain first. Another person may think verbally but lose access to that language under stress. Neither case allows an observer to know the depth or accuracy of the underlying thought from response speed alone. Silence can indicate uncertainty, overload, reflection, disengagement, fear, word-finding difficulty, or simply a mind that does not narrate continuously.

Respect does not require pretending that every unspoken intuition is correct. Thoughts still need testing, evidence, and revision. The point is that evaluation should begin after a person has access to a workable form of expression, not before. When institutions reward only immediate speech, they may measure speed of translation instead of quality of reasoning. When relationships demand instant verbal proof of care or understanding, they may turn a communication preference into a character judgment. Quiet minds do not need romanticizing. They need the same evidentiary fairness we would want for any other mind.

What would change if we evaluated thinking by function instead of resemblance to our own?

8. Ask What Works

A useful approach to inner experience begins with curiosity about function. Does speaking a thought make it easier to organize? Does drawing reveal a relationship that sentences conceal? Does movement help sequence an idea? Does writing provide enough time to translate several simultaneous impressions? Does silence allow irrelevant language to fall away? These questions treat cognitive tools as adjustable supports rather than signs of greater or lesser intelligence.

Measures of inner speech reinforce this functional view. The revised VISQ separates positive and regulatory inner speech from critical and evaluative forms, showing that the presence of more internal language does not have one predictable meaning (Alderson-Day et al., 2018). Likewise, low reported inner speech was associated with differences on some verbal tasks but not others in the anendophasia study (Nedergaard & Lupyan, 2024). A strategy can help in one context, remain neutral in another, and create interference elsewhere.

I can experiment without turning the result into a diagnosis. If an intention keeps disappearing, I can say it aloud. If language is flattening a complex idea, I can sketch the relationships first. If a conversation demands an immediate answer, I can ask for time and return in writing. If an internal narrator is escalating shame, I can change its wording or shift attention to sensory information. The goal is not to discover the one authentic way my mind thinks. It is to notice which form makes a particular thought more available, testable, and usable. Cognitive diversity exists not only between people but within the same person across tasks, environments, and levels of stress.

If thought can change form, what becomes possible when we stop demanding that it always sound like us?

9. Thought Exceeds Language

Words are among the most powerful tools human beings possess. They allow us to coordinate across distance, preserve knowledge, negotiate meaning, and make private experience partially shareable. Inner speech can guide behavior, regulate emotion, and keep a fragile intention available. None of those functions require us to declare that every thought is made from language.

The power of language can make its limits difficult to see. Because words are how we explain an idea, defend it, teach it, and leave a record of it, the verbal version can begin to look like the thought itself. Yet expression and existence are different questions. A thought can be real but incomplete, available but difficult to translate, or carried in a form that language can only approximate.

The evidence supports a more spacious account. Inner speech varies widely in frequency and form. Visual and verbal thinking interact without becoming identical. Some complex cognition survives profound language impairment. Verbalizing an experience can sharpen it, reorganize it, or sometimes interfere with the representation that came first. These findings do not settle the philosophical question of what thought ultimately is. They make one practical conclusion difficult to avoid: the absence of a narrated sentence is not the absence of a mind at work.

I can therefore listen differently, both to myself and to other people. Hesitation may be translation. Silence may be processing. An image, gesture, diagram, or imperfect phrase may carry information that polished speech would miss. The responsibility is not to treat every expression as equally clear or every intuition as correct. It is to stop using resemblance to my own inner experience as the price of recognizing cognition in someone else.

“Language gives thought a voice, but thought was never required to live in words.”

Further Reading

Thank you for reading, sharing, commenting on, and ***supporting my work***. If you enjoy evidence-based explorations of autism, ADHD, AuDHD, inner speech, visual thinking, communication, cognition, and the hidden processes that shape experience, I hope you will consider following my work.

Many of my articles examine what becomes visible when we stop treating one cognitive style as the template for every mind. In this article, I explore why inner speech varies, how visual and verbal thought interact, and why silence cannot tell us how much cognition is present. I also examine how language can organize thought without creating all of it, why verbal descriptions sometimes reshape what they attempt to report, and how neurodivergent experience complicates simple claims about the way people think. My goal is to replace reflexive judgment with better questions about function, access, context, and evidence, without romanticizing nonverbal thought or treating every private intuition as accurate.

If this article resonated with you, you may also enjoy my books:

📘 ***Evolution in Progress: Rethinking Neurodivergence*** explores autism, ADHD, AuDHD, and cognitive diversity through neuroscience, evolutionary biology, genetics, predictive processing, and complexity science. It provides a broader context for understanding why different minds organize perception, language, attention, and memory in different ways, and how assumptions about normal cognition shape identity, accommodation, and self-understanding.

📗 ***Anti-Anxiety Method: A Grounded 30-Day Path to Peace*** explores how the brain responds to uncertainty, perceived threat, and expectations shaped through earlier experience. It offers practical, evidence-informed ways to create space between activation and response, examine thoughts without treating every thought as a command, and develop greater flexibility when anxiety narrows attention.

📙 ***Ubuntu: The Power of Connection in a Divided World*** examines how empathy, cooperation, trust, and shared humanity shape individual identity and the communities we create. It asks what becomes possible when understanding another person does not require their inner world to resemble our own.

Together, these books continue the conversation beyond this article. They explore how minds develop, how language and culture influence the meaning of difference, and why better explanations can create more room for agency, compassion, and practical support. Thought becomes easier to respect when we stop demanding that every mind narrate itself in the same way.

References

• Alderson-Day, B., & Fernyhough, C. (2015). Inner speech: Development, cognitive functions, phenomenology, and neurobiology. Psychological Bulletin, 141(5), 931–965. **https://doi.org/10.1037/bul0000021**
• Alderson-Day, B., Mitrenga, K., Wilkinson, S., McCarthy-Jones, S., & Fernyhough, C. (2018). The varieties of inner speech questionnaire–Revised (VISQ-R): Replicating and refining links between inner speech and psychopathology. Consciousness and Cognition, 65, 48–58. **https://doi.org/10.1016/j.concog.2018.07.001**
• Alderson-Day, B., & Pearson, A. (2023). What can neurodiversity tell us about inner speech, and vice versa? A theoretical perspective. Cortex, 168, 193–202. **https://doi.org/10.1016/j.cortex.2023.08.008**
• Amit, E., Hoeflin, C., Hamzah, N., & Fedorenko, E. (2017). An asymmetrical relationship between verbal and visual thinking: Converging evidence from behavior and fMRI. NeuroImage, 152, 619–627. **https://doi.org/10.1016/j.neuroimage.2017.03.029**
• Fedorenko, E., & Varley, R. (2016). Language and thought are not the same thing: Evidence from neuroimaging and neurological patients. Annals of the New York Academy of Sciences, 1369(1), 132–153. **https://doi.org/10.1111/nyas.13046**
• Hurlburt, R. T., Heavey, C. L., & Kelsey, J. M. (2013). Toward a phenomenology of inner speaking. Consciousness and Cognition, 22(4), 1477–1494. **https://doi.org/10.1016/j.concog.2013.10.003**
• Lind, A. (2025). Are there really people with no inner voice? Commentary on Nedergaard and Lupyan (2024). Psychological Science, 36(9), 765–767. **https://doi.org/10.1177/09567976251335583**
• Lupyan, G., & Bergen, B. (2016). How language programs the mind. Topics in Cognitive Science, 8(2), 408–424. **https://doi.org/10.1111/tops.12155**
• Nedergaard, J. S. K., & Lupyan, G. (2024). Not everybody has an inner voice: Behavioral consequences of anendophasia. Psychological Science, 35(7), 780–797. **https://doi.org/10.1177/09567976241243004**
• Schooler, J. W., & Engstler-Schooler, T. Y. (1990). Verbal overshadowing of visual memories: Some things are better left unsaid. Cognitive Psychology, 22(1), 36–71. **https://doi.org/10.1016/0010-0285(90)90003-M**