Cognitive simulation denotes the process of modeling or reproducing mental operations—such as reasoning, problem‑solving, decision‑making, and imagination—either within the human mind or through artificial computational systems. In psychology it is studied as a fundamental cognitive mechanism that enables individuals to mentally rehearse scenarios, predict outcomes, and evaluate alternatives without physical action. The concept also overlaps with cognitive science through computational models that simulate human thought, providing a testbed for theories of mind and intelligence.

Cognitive simulation refers to the mental or computational replication of cognitive processes. Its scope encompasses both the conscious, deliberate use of imagination (e.g., planning a route) and the automatic, unconscious generation of mental models (e.g., predicting how a ball will bounce). The term is used across psychology, artificial intelligence, and philosophy of mind.

1.1 Historical background

The roots of cognitive simulation trace to early introspectionist psychology (e.g., Wilhelm Wundt) and later to the cognitive revolution of the 1950s–1960s, which emphasized mental representations and processes. Kenneth Craik (1943) proposed that the mind constructs “small‑scale models” of reality to test alternatives. In artificial intelligence, Allen Newell and Herbert Simon (1970s) developed production‑system models that simulated human problem‑solving, laying the groundwork for cognitive architectures.

Cognitive simulation is broader than mental imagery, which is specifically the perceptual‑like re‑creation of sensory experience (e.g., visualizing a scene). Simulation includes non‑sensory processes such as logical deduction and probabilistic inference. It also differs from pure reasoning: simulation often involves running a mental model step by step, whereas reasoning may rely on formal rules or heuristics without a dynamic “run‑through.”

Cognitive simulation can be classified by its implementation (mental vs. computational) and by its temporal orientation or purpose.

2.1 Mental simulation

Mental simulation occurs entirely within the human mind, often as a conscious or semi‑conscious activity.

2.1.1 Prospective simulation (future planning, prediction)

Prospective simulation involves constructing possible future scenarios. For example, mentally rehearsing a job interview or predicting the outcome of a chess move. It relies on episodic future thinking, supported by the hippocampus and prefrontal cortex.

2.1.2 Retrospective simulation (counterfactual thinking, memory reconstruction)

Retrospective simulation imagines alternative past events (“what if” scenarios) or reconstructs memories by filling in missing details. Counterfactual thinking helps people learn from mistakes and regulate emotions.

2.1.3 Simulative reasoning (hypothetical scenarios)

Simulative reasoning evaluates “if‑then” propositions by mentally running a model of the situation. It underpins deductive and inductive reasoning, especially when combining multiple premises, and is a core component of mental model theory (Johnson‑Laird, 1983).

2.2 Computational simulation

Computational simulation implements cognitive processes in software or hardware, aiming to replicate or shed light on human cognition.

2.2.1 Symbolic models (e.g., production systems, logic‑based)

Symbolic models represent knowledge as explicit symbols and rules. Production systems (e.g., ACT‑R’s production rules) apply “if‑condition, then‑action” to simulate step‑by‑step reasoning. Logic‑based models (e.g., SOAR’s operator selection) manipulate formal statements.

2.2.2 Connectionist models (e.g., neural networks)

Connectionist models use distributed, subsymbolic representations and learning through weight adjustments. Artificial neural networks can simulate pattern recognition, memory retrieval, and decision‑making, often capturing human cognitive biases.

2.2.3 Hybrid and Bayesian models

Hybrid systems combine symbolic and connectionist elements. Bayesian models treat cognition as probabilistic inference, updating beliefs based on evidence. These approaches are popular for modeling perception, language processing, and reasoning under uncertainty.

Cognitive simulation serves several adaptive functions across diverse domains.

3.1 Problem‑solving and decision‑making

By mentally “running” alternative solutions, individuals evaluate consequences without real‑world risk. Studies show that mental simulation reduces decision errors and improves planning, especially when combined with explicit reasoning.

3.2 Social cognition and theory of mind

People simulate others’ mental states (beliefs, desires, intentions) to predict behavior. This “simulation theory” of mindreading suggests that we use our own cognitive architecture to model others, supported by mirror‑neuron systems and default‑mode network activity.

3.3 Creative thinking and insight

Creative insight often arises from combining existing knowledge in novel ways through mental simulation. For example, inventors may visualize and tweak a design before building a prototype. Simulation allows exploration of many possibilities without material costs.

3.4 Learning and skill acquisition

Mental rehearsal (simulation of motor or cognitive tasks) accelerates skill learning. Professional athletes, musicians, and surgeons use mental practice to refine techniques. Neuroimaging shows that the same brain regions are activated during overt and imagined performance.

Researchers study cognitive simulation through behavioral, neuroscientific, and computational approaches.

4.1 Behavioral experiments (e.g., mental chronometry, self‑report)

Mental chronometry measures reaction times to infer simulation processes. Participants may be asked to mentally rotate an object (Shepard & Metzler, 1971) or imagine walking a route; longer times for harder transformations suggest a simulation cost. Self‑report scales assess how often or vividly individuals engage in simulation.

4.2 Neuroimaging studies (e.g., mental simulation in prefrontal cortex, hippocampus)

fMRI and EEG reveal that mental simulation activates the prefrontal cortex (executive control), hippocampus (episodic memory), and posterior parietal cortex (spatial imagery). Prospective simulation engages the default‑mode network, while counterfactual thinking involves the orbitofrontal cortex.

4.3 Computational modeling as a tool for theory testing

Computational models (e.g., ACT‑R, LEABRA) instantiate theories of simulation. Researchers can compare model outputs with human data to validate or refine hypotheses. Bayesian models are used to formalize how people update predictions during simulation.

Cognitive simulation has practical uses in technology, education, and therapy.

5.1 Artificial intelligence and cognitive architectures (e.g., ACT‑R, SOAR)

AI systems incorporate simulation for planning and reasoning. ACT‑R (Adaptive Control of Thought—Rational) simulates human cognition in tasks like air‑traffic control or learning. SOAR uses a universal subgoaling mechanism to simulate problem‑solving in games and robotics.

5.2 Education and training (e.g., mental rehearsal)

Mental rehearsal improves performance in fields from sports to surgery. Educational programs teach students to simulate historical events or scientific processes (e.g., visualizing molecular interactions). Virtual labs and serious games extend simulation to interactive contexts.

5.3 Clinical psychology (e.g., simulation of anxiety scenarios in therapy)

Cognitive‑behavioral therapy uses guided simulation to help patients confront feared situations in a safe imagined environment. Imaginal exposure reduces anxiety and trauma‑related distress. Simulation also aids in social‑skills training for individuals with autism spectrum disorders.

Despite its widespread use, cognitive simulation faces conceptual and practical challenges.

6.1 Validity of mental simulation as a construct

Critics question whether mental simulation is a distinct process or merely a by‑product of other cognitive functions. Some argue that “simulation” may be too broad, conflating imagery, reasoning, and memory retrieval. Disentangling these components remains difficult.

6.2 Computational tractability and ecological validity

Full simulation of a realistic situation would require enormous computational resources. Current models must rely on simplified representations, which may not capture the richness of human cognition. Ecological validity is also a concern: laboratory simulation tasks often lack the complexity and motivation of real‑world behavior.

Research continues to refine the concept and expand its applications.

7.1 Integration with embodied and situated cognition

Emerging theories emphasize that cognition is grounded in body and environment. Future work will explore how physical actions (gestures, eye movements) and environmental feedback shape mental simulation. Embodied cognitive architectures (e.g., using robots) may provide more realistic models.

7.2 Advances in brain–computer interfaces and virtual reality

BCIs and VR offer new ways to study and enhance simulation. VR creates immersive, controllable environments for mental rehearsal. BCIs can decode imagined actions (e.g., motor imagery) and provide real‑time feedback, opening possibilities for neurorehabilitation and skill training.