Alexander's model-cricket experiment

Alexander’s model-cricket experiment was an experimental demonstration that a cricket’s recent fighting record can alter its later behavior: real crickets beaten by a model cricket became more likely to lose fights against other real crickets. The result supports an account of fighting strategy in which experience updates an animal’s estimate of its relative ability, with consequences for its behavior and for the emergence of rank within a group.

Experimental Demonstration

Alexander used a model cricket to beat up real crickets. After this treatment, the real crickets became more likely to lose fights against other real crickets. The experiment thus connected an imposed history of defeat to a subsequent behavioral outcome, rather than relying only on an illustrative story about how animals might respond to past contests.

General Memory and Fighting Strategy

The experiment appears in a discussion distinguishing general memory of fight outcomes from recognition of particular opponents. Crickets are described as retaining a general record: a cricket that has recently won many fights becomes more hawkish, while one on a losing streak becomes more dovish. Alexander’s model supplied the losing experience, and the real crickets’ later behavior was consistent with defeat shifting their fighting tendency.

The proposed mechanism is an ongoing adjustment in each animal’s estimate of its fighting ability relative to the average individual in its population. On this account, a cricket need not recognize the particular opponent it lost to. Its strategy changes with its accumulated record, making past outcomes relevant to encounters with other crickets.

From Individual Outcomes to Group Rank

When animals that respond to a general memory of fights remain together in a closed group, these individual adjustments can produce a kind of dominance hierarchy. Individuals can be ranked by their outcomes: those lower in the order tend to give in to those higher in it. The account does not require individual recognition; repeated winning makes an individual increasingly likely to win, while repeated losing makes another increasingly likely to lose.

The resulting rank order can emerge even if individuals initially win or lose at random. As the pattern settles, the number of serious fights in the group gradually declines. Alexander’s experiment therefore illustrates a possible link between a manipulated contest history and a wider social consequence: accumulated outcomes can organize interactions and reduce the need for repeated serious conflict.