Latent Learning In Psychology and How It Works

Latent learning is a type of learning that isn’t apparent in the learner’s behavior at the time it happens. It only becomes visible later, once a suitable motivation and circumstance appear, showing that learning can occur without any reinforcement.

Edward Tolman is widely credited with discovering and popularizing latent learning. Although the broader idea that learning can happen without immediate reinforcement existed earlier, in various forms, Tolman’s systematic rat-maze experiments in the 1930s were the first to demonstrate it clearly.

Tolman argued that humans do this every day. Walking or driving a familiar route, we absorb the locations of buildings and landmarks without trying. That learning only becomes obvious when we actually need to find one of them.

Examples of Latent Learning

An example of latent learning in children is when a child watches their parents drive multiple times without actively trying to learn the process.

Later, playing with a toy car, the child copies the whole routine: checking mirrors, using indicators, starting the engine. No one ever taught or rewarded them for learning these steps.

The child’s earlier observations became evident when there was a relevant situation to apply the knowledge.

Cognitive maps as an example of latent learning in rats

Tolman coined the term cognitive map: an internal representation, or image, of the physical layout of an environment.

He believed individuals pick up large numbers of cues, or signals, from the environment. They use these cues to build a mental image of that environment, a cognitive map.

This internal map lets an animal reach a goal by knowing its location among many environmental features. It also makes shortcuts and route changes possible.

In his experiments with rats in mazes, Tolman observed that even without direct rewards, rats seemed to develop a “mental map” of the maze.

Tolman’s clearest evidence for the cognitive map came from a “sunburst” maze experiment with colleagues Benjamin Ritchie and Daniel Kalish.

Aim: Tolman, Ritchie, and Kalish (1946) tested whether rats had learned a fixed sequence of turns, or the true location of their goal in space.

Method: Rats first learned one circuitous path across a table and through a curved corridor to reach food. Researchers then replaced the corridor with a “sunburst”: straight alleys fanning out in different directions, with only one pointing at the food’s true location.

Results: Most rats ignored paths resembling their original route, instead picking the alley that pointed most directly at the food, even though they had never walked it before.

Conclusion: A rat that only learned a fixed chain of turns would have no reason to choose a brand-new path toward the goal. Choosing correctly showed the rats held a cognitive map of where the food was, not a memorized route.

Aim

Tolman and Honzik (1930) tested whether rats need a reward to learn a maze. Or can they learn its layout even with no reward at all?

Procedure

In their study, 3 groups of rats had to find their way around a complex maze. At the end of the maze, there was a food box. Some groups of rats got to eat the food, some did not, and for some rats the food was only available after 10 days.

In their famous experiments Tolman and Honzik (1930) built a maze to investigate latent learning in rats. The study also shows that rats actively process information rather than operating on a stimulus response relationship.

Results

The delayed reward group learned the route on days 1 to 10 and formed a cognitive map of the maze. They took longer to reach the end of the maze because there was no motivation for them to perform.

From day 11 onwards, they had the motivation to perform (i.e. food) and reached the end before the reward group.

Between the stimulus (the maze) and the response (reaching its end), a mediational process was taking place. The rats were actively processing information, using the cognitive map they had built.

This is the theoretical heart of latent learning. Reinforcement did not create the rats’ knowledge of the maze; it only gave them a reason to act on knowledge they already had. Learning and performance, Tolman argued, are separate processes.

Tolman’s Purposive Behaviorism

Tolman called his approach purposive behaviorism, later also called cognitive behaviorism. He accepted behaviorism’s core rule: psychology must rely on objective, observable evidence. But he rejected the claim that this rules out any reference to internal mental states.

He positioned himself between two extremes: introspectionism, which relied on unverifiable conscious content, and Watson’s and Skinner’s radical behaviorism, which banned any reference to internal states.

Intervening Variables and Cognitive Maps

Tolman treated ideas like the cognitive map as intervening variables. These are internal processes we cannot observe directly, but infer from behavior. A rat’s maze layout and its choice of turn are both observable; the cognitive map linking them is not.

This intervening-variable move was itself a methodological innovation, not just a theoretical one.

Tolman explored this idea in his own 1938 paper, The determiners of behavior at a choice point. He examined what happens at the exact moment a rat pauses at a maze intersection, weighing alternative expectancies before choosing a path.

This let Tolman keep behaviorism’s rigor while restoring room for purpose and cognition. Where strict stimulus-response theory said reinforcement simply strengthens a habit, Tolman argued something different. Rats learn information about their environment, then act on it to reach a goal.

Sign Learning vs. Hull’s S-R Theory

Tolman’s biggest rival was Clark Hull, whose drive-reduction theory said all learning strengthens stimulus-response habits by reducing a physiological drive.

Tolman called his own account sign learning. On his view, an animal learns that one cue, a “sign,” leads to a particular outcome, not a fixed muscular habit.

Through the 1930s and 1940s, Tolman and Hull’s students designed rival maze experiments to test the two theories. Neither side landed one knockout result.

But the sign-learning side aged better. Modern research on spatial navigation and reinforcement learning still separates habitual, “model-free” responding from “model-based” learning of an internal map. That split echoes Tolman’s argument from eighty years ago.

Tolman spent his career working under the behaviorist banner. But his tools, inferred mental representations pursued flexibly toward a goal, are recognizably the tools of a cognitive psychologist. The formal cognitive revolution did not arrive until the 1950s and 1960s, decades after Tolman’s data made the case for it.

Critical Evaluation

Strengths

1. Beyond Behaviorism: Latent learning challenged the dominant behaviorist idea that reinforcement is necessary for learning. It broadened the understanding of how and when learning occurs to include cognitive factors, such as information processing.

   

2. Support from Experiments: Tolman’s evidence did not rest on one result. Delayed-reward, maze-rotation, and swimming studies all converged on the same conclusion (Restle, 1957).

3. Explains Everyday Learning: The theory explains how we learn from our surroundings even without a clear reward, like noticing a shortcut only once we need it.

4. Cognitive Emphasis: Latent learning paved the way for a more cognitive approach to understanding learning, emphasizing mental processes and internal representations.

Weaknesses

1. Measurement Challenges: Since latent learning is not immediately observable, it can be difficult to measure and quantify.

2. Over-reliance on Animal Studies: While Tolman’s experiments were groundbreaking, they were primarily conducted with rats. Extrapolating findings from animal studies to human behavior can be problematic.

3. Limited in Scope: Latent learning mainly addresses situations where learning is not immediately apparent. It doesn’t provide a comprehensive view of all learning processes or explain why some learned behaviors might manifest immediately while others don’t.

4. Ambiguity in Mechanism: Latent learning describes what a cognitive map contains but not how animals turn it into action; critic Edwin Guthrie (1935) argued rats are left “buried in thought” at each choice point.

Contemporary Research

Tolman’s cognitive map earned some of the strongest confirmation of any early-20th-century construct: a literal neural version of it. The hippocampus contains place cells, neurons that fire when an animal is in a specific location. This work won the 2014 Nobel Prize in Physiology or Medicine.

Aim: Wikenheiser and Redish (2015) tested whether hippocampal activity does more than track location, specifically whether “theta sequences” of place cells project ahead toward the animal’s current goal.

Method: Rats chose between routes to goals that differed in distance and value. Researchers recorded place-cell activity and measured how far these forward-projecting sequences reached ahead of the rat’s true position.

Results: The look-ahead distance was not fixed: it stretched further on trips to distant goals than nearby ones, and predicted which goal the rat would choose.

Conclusion: Hippocampal activity does not just mark where an animal is. It represents where the animal intends to go, consulted flexibly in the service of a current goal. That is close to a direct neural confirmation of what Tolman could only infer from behavior.

A related study, Ólafsdóttir et al. (2015), found that place-cell sequences can “pre-play” a route through space the rat has never explored. This happens right after the rat sees a reward placed there, an echo of Tolman and Honzik’s original latent-learning logic.

A broader synthesis, Behrens et al. (2018), argues in Neuron that this same hippocampal system may organize knowledge generally, not only physical space.

Who Was Edward Tolman?

Edward Chace Tolman (1886–1959) was an American psychologist. He spent almost his entire career studying how rats learn to run mazes, yet used that work to challenge behaviorism from the inside.

He trained first in the physical sciences, earning a bachelor’s degree in electrochemistry from MIT in 1911. He then switched to psychology, completing his PhD at Harvard in 1915.

After a brief spell at Northwestern University, he moved to Berkeley in 1918. He stayed at the University of California for the rest of his working life.

In 1950, the University of California imposed a loyalty oath during the McCarthy era. Tolman led the Berkeley faculty who refused, on principle, to sign an oath disavowing Communist Party membership.

He was dismissed despite tenure and became the named plaintiff in the resulting suit. The California Supreme Court ruled the oath unconstitutional in Tolman v. Underhill (1952), reinstating him.

Tolman’s standing in American psychology was considerable. He was elected to the National Academy of Sciences in 1937 and served as President of the American Psychological Association that same year. He was also Vice President of the American Association for the Advancement of Science in 1942.

He died in Berkeley on 19 November 1959.

References

Behrens, T. E. J., Muller, T. H., Whittington, J. C. R., Mark, S., Baram, A. B., Stachenfeld, K. L., & Kurth-Nelson, Z. (2018). What is a cognitive map? Organizing knowledge for flexible behavior. Neuron, 100(2), 490–509. https://doi.org/10.1016/j.neuron.2018.10.002

Guthrie, E. R. (1935). The psychology of learning. Harper.

Ólafsdóttir, H. F., Barry, C., Saleem, A. B., Hassabis, D., & Spiers, H. J. (2015). Hippocampal place cells construct reward related sequences through unexplored space. eLife, 4, e06063. https://doi.org/10.7554/eLife.06063

Restle, F. (1957). Discrimination of cues in mazes: A resolution of the “place-vs.-response” question. Psychological Review, 64(4), 217–228. https://doi.org/10.1037/h0040678

Tolman, E. C. (1938). The determiners of behavior at a choice point. Psychological Review, 45(1), 1–41. https://doi.org/10.1037/h0062733

Tolman, E. C. (1948). Cognitive maps in rats and men. Psychological Review, 55(4), 189–208. https://doi.org/10.1037/h0061626

Tolman, E. C., & Honzik, C. H. (1930). Introduction and removal of reward, and maze performance in rats. University of California Publications in Psychology, 4, 257–275.

Tolman, E. C., Ritchie, B. F., & Kalish, D. (1946). Studies in spatial learning. I. Orientation and the short-cut. Journal of Experimental Psychology, 36(1), 13–24. https://doi.org/10.1037/h0053944

Wikenheiser, A. M., & Redish, A. D. (2015). Hippocampal theta sequences reflect current goals. Nature Neuroscience, 18(2), 289–294. https://doi.org/10.1038/nn.3909

FAQs