10Characterizing behavior in any organism as play, including in humans, has often been controversial. Intuitive understandings of what constitutes ‘play’ are often difficult to describe in words so that other researchers can use them. This leads to problems in comparing studies, formulating and testing research hypotheses, and even in having a shared conversation. To alleviate this problem, many attempts have been made to define or identify play in all its guises. Unfortunately, these attempts have generally failed due to the narrow context in which they have been developed or the conceptual language in which they are expressed (e.g., cognitive, behavioristic, physiological). A careful perusal of these attempts has led to the identification of a set of five criteria, each of which need to be satisfied in at least one respect, in order to identify a behavior as play in whatever context or species being studied. In addition, play can be viewed as operating, at a functional level, in three processes (primary, secondary, tertiary) spanning the continuum from the seemingly atavistic to the developmentally valuable if not essential. Where specific instances of play fall along this continuum can only be answered by empirical research and not by inference from the labels we give to them. In any event, conceptual clarity appears necessary for scientific progress. © 2011 by Oxford University Press, Inc. All rights reserved.
