Play Jeff Schank ANB 218a 2013. What is Play? Criteria for play 1.Incompletely functional; 2.spontaneous, pleasurable, rewarding, or voluntary; 3.differs.

Play

Jeff SchankANB 218a 2013

What is Play?

• Criteria for play1. Incompletely functional; 2. spontaneous, pleasurable, rewarding, or

voluntary; 3. differs from other more serious behaviors in form

(e.g., exaggerated) or timing (e.g., occurring early in life before the more serious version is needed);

4. is repeated, but not in abnormal and unvarying stereotypic form (e.g., rocking or pacing); and

5. is initiated in the absence of severe stress.

Types of Play• Solitary locomotor-rotational play

– Vigorous motor acts, typically performed alone (e.g., playful running and twisting in ungulates, and somersaulting in monkeys)

• Object play– Involves the playful use or manipulation of inanimate objects (e.g., a

dog retrieving a stick or a cat batting a ball) • Social Play

– Social play involves two or more players that are usually, but not always conspecifics.

– Typical movement patterns involve chasing, wrestling, and tail-pulling, and even a form of peek-a-boo

– Of all social play patterns, rough-and-tumble play (R&T), or play-fighting, is most frequently studied in animals

Theories of Play: Historical View

• Surplus Energy– Animals play when the have surplus energy and they

are in good health—Why play?• Instinct-Practice– Instinctive behavior requires practice to become

optimal (assumption) and play is a form of practice—Why and how is play practice?

• Recapitulation– Play is a vestigial relic from past evolution—Why is

play so common?

Theories of Play: Modern Functional View

• One problem with the older theories is that each was viewed as an independent theory

• However, aspects of each are relevant to the modern functional view

• Functional view– Play likely has multiple functions: motor training,

practice, and socialization– Very difficult to test the functions of play because

presumably the fitness benefits are delayed till adulthood

Functional Theories

• Motor training– Physical exercise: No evidence– Cerebellar synaptogenesis: play occurs after most of

cerebellar synaptogenesis is complete• Training for unexpected events– Play allows an animal to acquire kinematic and emotional

skills required for dealing with unexpected events– Predicts play should be more frequent in changing

environments– Can’t explain why play is less frequent is poor (food

shortages, stressful, challenging) environments

Functional Theories• Practice

– Little evidence that play facilitates learning corresponding adult behaviors• Social benefits

– Enhancing social skills– Strengthening social bonds– Reducing aggression– Refining social assessment– Learning and promoting cooperative behavior

• sharing• Reciprocity• Altruism• Fairness

• Little evidence to support any of these theories since some animals deprived of social play still develop adult social competencies

• Playing animals seem to learn fairness and cooperation, and may even punish cheaters by not playing with them

• I’ll come back to social play later

Costs and Benefits of Play

• Costs– Injury from falls or aggressive retaliation– Reduced time spent in survival behaviors (e.g., foraging)– Expenditure of energy– Increased predation risk as a result of reduced vigilance

and the conspicuousness of play to predators• Delayed Benefits– Nunes (2004) found that social play in female Belding’s

ground squirrels increased reproductive success during their first breeding season

Interspecies Play• Examples of interspecies play

– Lion and Tiger– Wolf, Bear, Human– Cat and Owl– Rat and Cat– Kangaroo and Lemur– Dog and Deer– Polar Bears and Dogs– Cat and Dolfins– Monkeys and Dogs– Humans playing with Stingray?

• How do we explain interspecies play?• Does it tell us anything about social play?• Does it tell us anything about the evolution of social play?

Why is Social Play Beneficial?• It seems plausible if not obvious that social play could derive its

fitness benefits by learning the skills necessary for adult cooperation

• The evolution of cooperation, however, is hard to explain, especially at low frequencies

• Suppose a social play gene is introduced into a population of non-cooperators at low frequencies– How do juveniles with the play gene find others to engage in social play?– If they do, and learn to cooperate, adults would be at a disadvantage in

a population of mostly non-cooperators• It would appear that the evolution of social play by facilitating

adult cooperation is improbable at best

Fitness Landscape

A Model of the Evolution of Social Play

• Sometimes intuitions can be misleading!• Let’s consider agents with generic biological properties

characteristic of animals that engage in social play and adult cooperation

• Agents can– Movement and Aggregation– Development (juvenile adult)– Learning– Reproduction– Parental investment– Life span– Foraging for resources

Movement and Aggregation

Cooperation

Prisoner’s Dilemma: T > R > D > S (e.g., T = 5 > R = 3 > D = 1 > S = 0)

Stag Hunt: R > T ≥ D > S (e.g., R = 2 > T = 1 ≥ D = 1 > S = 0)

Learning to Cooperate

Rescorla-Wagner learning model

Rescorla-Wagner learning model

Swapping: Controlling Group Selection

Parameter Sweeps

Fixe

d Pa

ram

eter

s

Coop

erati

on

Soci

al G

ene

Fine

-Gra

ined

Ana

lysi

s

Example of Evolution

Drift Phase

Natural Selection Phase

Migration, Group Selection Phase

Conclusions

• Social play could evolve facilitating the learning of adult cooperative skills

• How does it work?– Sewall Wright’s shifting balance theory– Effectively neutral mutations in low frequency– Parental investment: inheriting wealth

• There is much to be done to understand play– None of Tinbergen’s 4 questions have been

adequately developed