An experimental study of the effects of play upon the dog-human relationships

It has often been suggested that intraspecific dominance: relationships are established through play. By analogy, it is also claimed that the outcome of competitive games can affect dog-human relationships. This paper experimentally tests the latter idea. Fourteen Golden, Retrievers were each subjected to two treatments; 20 sessions of a tug-of-war game with the experimenter which they were allowed to win, and 20 sessions which they lost. Their relationship with the experimenter was assessed, via a composite behavioural test, once at the outset and once after each treatment. Principal components analysis allowed the 52 behavioural measures to be combined into nine underlying factors. Confidence (the factor most closely corresponding to conventional dominance) was unaffected by the treatments. Dogs scored higher for obedient attentiveness after play treatments, irrespective of whether they won or lost, and demandingness scores increased with familiarity of the test person. The 10 most playful dogs scored significantly higher for playful attention seeking after winning than after losing. We conclude that, in this population, dominance dimensions of the dog-human relationship are unaffected by the outcome of repetitive tug-of-war games. However, we suggest that the effects of games may be modified by the presence of play signals, and when these signals are absent or misinterpreted the outcome of games may have more serious consequences. Games may also assume greater significance for a minority of "potentially dominant" dogs. (C) 2002 Elsevier Science B.V. All fights reserved.