Animal Intelligence: An Experimental Study of the Associative Processes in Animals

Thorndike argues that animal learning is not the product of reasoning, imitation, or insight but of the gradual strengthening of specific stimulus-response connections through trial and error. He rejects the anecdotal method that had dominated earlier comparative psychology, in which stories of clever animals were used to infer mental faculties, and he rejects the interpretation of animal behavior in terms of human-like reasoning. His experiments with cats in puzzle boxes and chicks in mazes showed that learning curves are continuous and gradual, not discontinuous as would be expected if animals suddenly 'solved' problems. The monograph introduces the law of effect—responses followed by satisfying consequences are strengthened—as the mechanism of learning.

Thorndike constructed puzzle boxes of various designs—some requiring the animal to press a lever, pull a string, or step on a platform to escape and obtain food. A cat was placed inside, and the time taken to escape was recorded across successive trials. The learning curves showed a gradual decrease in escape time, with many errors along the way. Thorndike found no evidence of sudden insight: the cats did not appear to understand the mechanism of escape but instead repeated responses that had previously been followed by escape. He also used mazes with chicks and dogs to study the formation of associations. The procedures established the experimental method for studying animal learning under controlled conditions.

The law of effect states that responses followed by satisfying consequences are strengthened, while responses followed by discomforting consequences are weakened. Thorndike formulated it in Animal Intelligence based on his observations that cats repeated the responses that led to escape and food and gradually eliminated responses that did not. The law was the first systematic statement of the principle that consequences select and strengthen behavior, and it became foundational for behaviorism. In later work, Thorndike revised the law to exclude the weakening effect of punishment, concluding that reward strengthens connections but punishment does not weaken them directly. The original 1898 formulation was more symmetrical than the revised version.

Thorndike argued that the anecdotal method—collecting stories of clever animals from pet owners, hunters, and observers—was unreliable because it selected for dramatic and unusual events and relied on anthropomorphic interpretation. He demonstrated that controlled experiments could yield quantitative learning curves and reproducible findings, and he insisted that claims about animal intelligence be based on observable behavior under standardized conditions. This methodological critique was as important as his substantive findings: it established that comparative psychology must be an experimental science. His approach contrasted sharply with that of George Romanes, whose anecdotal collection Mental Evolution in Animals (1883) had been the standard reference. Thorndike's insistence on controlled observation shaped the subsequent development of both comparative psychology and behaviorism.

Darwin, in The Descent of Man (1871) and The Expression of the Emotions in Man and Animals (1872), had emphasized continuity between human and animal mental faculties, arguing that animals possess reason, imagination, and emotions in rudimentary form. Thorndike's findings seemed to challenge this continuity thesis by showing that animal learning lacks the rational, insightful character of human problem solving. However, Thorndike did not deny continuity altogether: he argued that the associative processes he studied are the foundation on which more complex human intelligence is built. The difference is one of degree and complexity, not of kind. The debate over whether animals reason or merely associate became a central controversy in comparative psychology, with Thorndike's position representing the more conservative, experimental approach.

The monograph was recognized immediately as a major contribution and established Thorndike's reputation as a leading experimental psychologist. William James praised it, and it influenced a generation of researchers in animal learning. Its long-term influence was enormous: it provided the empirical foundation for the law of effect, which became a central principle of behaviorism; it established the experimental method in comparative psychology; and it shaped the study of learning in both animals and humans. The learning curve and the puzzle box became standard tools and concepts. Later research refined and in some cases challenged Thorndike's conclusions—particularly his claim that animals show no insight—but his work remained the starting point for the experimental analysis of learning for decades. It is one of the founding documents of behaviorist psychology.