What 'play' technically means

Gordon Burghardt's framework, the most-cited definition of play in animal behavior research, requires play behavior to meet five criteria: (1) the behavior is not fully functional in the immediate context (it doesn't directly serve survival or reproduction), (2) it is voluntary and spontaneous, (3) it differs from non-play versions of similar behaviors in form or timing, (4) it is repeated, and (5) it occurs in low-stress conditions when the animal's basic needs are met. These criteria distinguish play from other behaviors that might superficially resemble it — practice, accidental movement, displacement behavior. Behaviors that meet all five criteria are properly called play. Behaviors that meet only some are something else.

Several documented examples of American crow and common raven behavior that satisfy Burghardt's criteria.

What crows do that meets the criteria

Several documented examples of American crow and common raven behavior that satisfy Burghardt's criteria. Sliding down snowy or sloped surfaces repeatedly, climbing back up, sliding again — meets non-functional, voluntary, distinct, repeated, low-stress criteria. Object manipulation play: picking up sticks or stones, dropping them from height, catching them on the way down, repeating. Aerial play: flying in non-functional aerobatic patterns, particularly in wind, apparently for the experience rather than to reach any destination. Social play: pulling at the tails of unsuspecting other animals (cats, dogs, sometimes humans), which combines investigation with what looks like deliberate provocation. Each of these has been reported in the corvid behavioral literature across multiple researchers and contexts.

Why play matters for cognition

Play behavior is correlated across animal species with cognitive complexity. Mammals with more elaborated brains show more play behavior, and the kinds of play they show track cognitive sophistication: simple play in less cognitively-complex species, complex play (object manipulation, social interaction, rule-following) in more cognitively-complex species. Corvid play patterns include the cognitively-complex types: object play with apparent goal-directed structure, social play with apparent partner-selection, and exploratory play with apparent novelty-seeking. The play repertoire is consistent with the cognitive profile that other lines of evidence (face recognition, episodic-like memory, social learning) already establish for corvid species.

The 'snowboarding crow' viral video

A widely-circulated video from Russia in 2012 showed a hooded crow apparently sliding down a snowy rooftop on a lid or jar lid. The video went viral in part because it appeared to show such a clearly-play-like behavior. Subsequent analysis confirmed the behavior met play criteria, but more importantly, similar behaviors had been previously reported in the academic literature for multiple corvid species — the video wasn't a new observation, it was a vivid documentation of behavior that researchers were already aware of. Heinrich[1]'s books include similar accounts; Marzluff[2] lab observations include similar accounts. The viral video brought to popular attention something the research literature had been describing for decades.

Why this matters for communication research

Two implications. First, the play behavior is additional evidence for the cognitive complexity that AI bioacoustic interpretations of corvid communication need to be premised on. Species capable of object play, social play, and exploratory play have the cognitive substrate for communication systems that exploit similar flexibility. Second, play behavior is itself sometimes accompanied by vocalizations; the social play of corvids includes vocal exchanges that aren't well-categorized by the standard call-type taxonomy and may represent a poorly-understood part of the species's vocal repertoire. CrowLingo's atlas has no verified play-specific group; the play-specific vocal patterns probably warrant more attention than current research has given them.

What the play research is still missing

Despite multiple reported observations, systematic experimental study of corvid play behavior is limited. The reasons are partly practical (play is hard to elicit experimentally because it's voluntary), partly conceptual (the field lacks consensus on what specific cognitive capacities play behavior would demonstrate beyond the cognitive-complexity correlation), and partly funding-driven (play behavior is less obviously application-relevant than predator-detection or social learning research). The play behavior research is a normal area of corvid cognition that deserves more attention than it has gotten. CrowLingo's atlas can usefully include play-related vocalizations as a topic for future expansion as the academic field produces more systematic work in this area.