Can Individual Differences in History of Dominance Explain the Development of Linear Dominance Hierarchies?

Abstract Observed dominance hierarchies are often more linear than expected from randomly‐formed dominance relationships, and in triads of animals attacks are distributed non‐randomly. I hypothesize that an individual's history of dominance affects its probability of initiating aggressive interactions in the future and that individuals with winning records are more likely to initiate (winning begets initiating). Consistent with this hypothesis, evidence is presented that dominant individuals are more likely to attack than subordinate individuals. The winning begets initiating hypothesis may also explain why correlations between predicted dominance ranks (based on size, age etc.) and observed dominance ranks can be low: If the cost of engaging in and losing an interaction is high relative to the potential benefits of winning, then a large individual conditioned to be subordinate may refrain from contesting smaller, dominant individuals despite its actual competitive superiority.