4Restraint is thought to be essential to enable the reciprocity needed for play ï¬ghting to remain playful. Descriptions of playing in pigs suggest that they do not exhibit restraint. Analysis of videotaped sequences of play ï¬ghting in captive family groups of warty pigs was used to test three hypotheses about restraint and reciprocity. Hypothesis 1 asserts that the lack of restraint arises from neither participant handicapping its actions in favour of its opponent: this was supported. Hypothesis 2 asserts that the winner of a contest will show restraint by not prosecuting further attack: this was not supported. However, the winner did refrain from attacking if the loser signalled submission. Hypothesis 3 asserts that restraint by the winner will allow reciprocal attacks by the loser — this was supported. The dissociation of restraint and reciprocity evident in the pigs offers some new insights into the evolution of play ï¬ghting.
