How Do People Actually Think? Inside the Mental Process
Thinking is so constant and so automatic that it rarely feels like something that requires explanation. You ask yourself a question, an answer appears. You face a problem, ideas start forming. The process seems transparent — thoughts just arise. But what's actually happening in the brain between a question and an answer, between a problem and an insight, is considerably more complex and more interesting than the subjective experience suggests. Understanding it changes how you approach learning, problem-solving, and the development of your own thinking.
Thinking Is Not One Process
The first thing cognitive science makes clear is that "thinking" is not a single activity. It's a label applied to a family of different mental processes that happen to share the feature of operating on internal representations rather than immediate sensory experience. Reasoning, problem-solving, planning, mental imagery, inner speech, daydreaming, and creative insight all get called "thinking," but they involve different cognitive systems, different brain regions, and different computational mechanisms.
What they share is a dependence on working memory — the system that holds currently active mental representations and manipulates them. Almost all thinking, in the ordinary sense, happens in working memory. When you deliberate about a decision, the options and their implications are held in working memory while you compare them. When you follow an argument, each premise is maintained in working memory while the next is processed. When you plan a route, a mental representation of the space is held and traversed. Working memory is the stage on which conscious thought plays out.
This is also why thinking capacity is so closely linked to working memory capacity. People with larger working memory can hold more information active simultaneously, which directly expands the complexity of problems they can reason through without losing track of earlier steps. The N-Back test measures and trains exactly this — the ability to hold and update information in working memory under continuous cognitive load.
Inner Speech: The Voice Inside Your Head
One of the most distinctive features of human thinking is inner speech — the experience of talking to yourself in your own head. For most people, much of deliberate thinking happens in a verbal or quasi-verbal form: you formulate questions, talk through problems, argue with yourself, rehearse what you're going to say. This internal monologue is so pervasive that it can be easy to mistake it for thinking itself, rather than one form thinking takes.
Research by Alderson-Day and Fernyhough (2015), reviewing decades of work on inner speech, established that inner speech serves several distinct cognitive functions: it supports working memory maintenance (the phonological loop in Baddeley's model), it helps regulate and organize thinking by providing a running self-commentary, it aids in planning by allowing you to run through scenarios verbally before committing to action, and it plays a role in metacognition — the monitoring of your own thinking process.
Not everyone's inner speech is equally verbal, however. People vary significantly in how verbally or visually they think. Some people think primarily in words; others think primarily in images, spatial representations, or something harder to categorize. There's no evidence that one format is superior for all purposes — verbal thinking has advantages for reasoning through logical structure, while visual and spatial thinking has advantages for problems involving spatial relationships, pattern recognition, and creative synthesis.
Two Kinds of Thinking: Fast and Slow
Perhaps the most practically important distinction in cognitive science is between automatic and deliberate thinking — what Kahneman famously described as System 1 and System 2. Most of what goes on in the mind at any moment is automatic: pattern recognition, familiar associations, habitual responses, background monitoring of the environment. This processing happens rapidly, without effort, and largely outside of conscious awareness. It's the basis of intuition, expertise, and skilled performance.
Deliberate thinking — the kind that feels like thinking — is slower, effortful, and serial. It operates one step at a time, requires attentional resources, and is easy to interrupt or derail. This is the thinking you do when solving an unfamiliar problem, carefully evaluating evidence, or working through a complex argument. Because it's cognitively expensive, the brain tends to avoid it when automatic processing can produce a good-enough answer — which is most of the time.
The interplay between these two modes is central to understanding why thinking goes wrong. Automatic processing is fast and efficient but captures the familiar even when the familiar is wrong. Deliberate processing is accurate but slow, resource-limited, and easily hijacked by the emotional and motivational states that bias what it attends to. Most thinking errors arise at the interface — when automatic processing produces an answer that feels compelling enough to suppress deliberate checking, or when deliberate processing is performed with insufficient attention to question its own assumptions.
Mental Models: How We Represent the World to Think About It
To think about anything — a problem, a situation, a plan — you need a mental representation of it. Cognitive psychologists call the structured representations used for reasoning mental models: internal simulations of how things work that can be mentally manipulated and interrogated to generate predictions and conclusions.
When you try to figure out why your computer isn't working, you're running a mental model of the system and testing hypotheses against it. When you plan how to rearrange furniture in a room, you're running a spatial mental model that lets you mentally rotate and reposition objects without physically touching them. When you reason about someone else's beliefs and intentions, you're running what cognitive scientists call a "theory of mind" — a model of another person's mental state that allows you to predict their behavior.
The quality of thinking depends heavily on the quality of the mental models it operates on. Expert thinkers in any domain have richer, more accurate mental models of their domain than novices — which is why expert problem-solving looks qualitatively different from novice problem-solving even when both are working on the same problem. The novice applies general reasoning to surface features; the expert operates on a deep structural model that makes the relevant relationships immediately visible. This is what the Matrix Reasoning test measures — the ability to build and operate on abstract relational models under novel conditions, stripped of any domain-specific knowledge.
Why Some Thinking Feels Effortless and Other Thinking Stalls
The subjective experience of thinking varies enormously: sometimes ideas flow easily and rapidly, sometimes progress is slow and grinding, and sometimes you hit a complete block. These variations reflect real differences in the cognitive conditions operating at the time.
Effortless thinking typically occurs when automatic processing has a strong, accurate answer ready — when the problem is familiar, when expertise provides a fast pattern match, or when the solution is retrieved from memory rather than derived from first principles. The subjective experience of ease doesn't guarantee correctness (familiar automatic answers can be wrong), but it is informative about how well the current situation maps onto stored knowledge and patterns.
Effortful thinking occurs when automatic processing produces no confident answer and deliberate reasoning must take over. The difficulty of the thinking is partly determined by the inherent complexity of the problem, and partly by the available cognitive resources — working memory capacity, current attentional state, and the quality of the mental models available to reason with. When thinking stalls, it's usually because working memory is overloaded (too many elements to hold simultaneously), because the mental model being used is missing crucial information, or because the problem is being framed in a way that makes the solution invisible.
Insight — the sudden "aha" experience where a solution appears fully formed — is a distinctive mode of thinking that cognitive neuroscience has studied extensively. It typically occurs when a problem has been framed in a way that makes it unsolvable, then reframed in a way that makes the solution obvious. During the period before insight, the brain is often processing the problem implicitly in the background; the "aha" moment reflects the point at which that implicit processing produces a representation that breaks through into conscious awareness. This is why relaxed, diffuse attention states sometimes facilitate insight while intense focused effort blocks it — the focused state locks in the current framing, while the relaxed state allows alternative framings to emerge.
The Limits of Introspection
One of the more counterintuitive findings from cognitive science is that people have poor insight into their own thinking. You can typically report the conclusion of a reasoning process but not the process itself. You can report that a decision felt right but not accurately reconstruct the factors that produced that feeling. You can report that a memory feels familiar but not reliably distinguish between remembering something and merely believing you remember it.
This limited access to our own cognitive processes has practical consequences. It means that confidence in a judgment is a weak signal of its accuracy. It means that post-hoc explanations of decisions — what we tell ourselves about why we chose something — are often reconstructions that don't accurately reflect the actual causal process. And it means that improving thinking requires external feedback and measurement rather than just introspective reflection, because introspection systematically underestimates the extent to which thinking is automatic, biased, and resource-limited.
This is precisely where objective cognitive assessment becomes useful. Knowing your actual working memory span, your genuine processing speed, and your real pattern recognition capacity — rather than your intuitive sense of these things — gives you an accurate map of the cognitive resources available to your thinking. The Short-Term Memory test and Reaction Time test both provide concrete measurements that bypass the inaccuracies of self-assessment.
For a broader look at the cognitive systems that thinking draws on, the What Does Cognition Actually Consist Of article covers the full architecture of the cognitive system. For a closer look at how the decision-making component of thinking works — and where it systematically fails — Why Your Brain Makes Decisions Before You Do covers the dual-process dynamics in detail.