You are on page 1of 27

CHAPTER 11

Social Learning and


Cognition
PY 211 Learning
Mounia Ziat
Email: mziat@nmu.edu

Social Learning (aka Observational Learning)


Social learning is learning in which the learner actively

monitors events involving other individuals and then chooses


later actions based on their observations.
Informally, often called copying or imitating

Powerful form of learning in humans:


Learn from watching others, watching video, reading books, etc.
Difficult to study, especially in non-humans:
Usually, no reward/punishment specifically given during training.
Depends on the learners attention to and perception of the situation
they observe.
Up to the learner to decide when/how actually perform behaviors that
have been observed.
Hard to predict and measure what is learned.

Social Learning (aka Observational Learning)


Social learning is learning in which the learner actively

monitors events involving other individuals and then chooses


later actions based on their observations.
Does it exist in nonhuman animals?
Some dolphins can learn from both other dolphins and humans.
Some primates can learn to imitate other primates.
It seems, however, that the human species is the only one that makes
extensive use of social learning by
copying

Copying what is seen


The experimental study of social learning started with a set of

seminal experiments by Albert Bandura and colleagues. The


intent of these experiments was to see if aggressive behavior
in adults would be copied by children.

Exposure:
Children watched an adult modeling aggressive behavior (beating on an

inflatable Bobo doll) or an adult modeling quiet play.


Test:
Children were watched while playing with toys, including the Bobo doll.

Copying what is seen


Bandura found a complex pattern of results.

Despite their complexity, clear demonstration that just

observing adult behavior influences child behavior.


Moreover, the ability to learn from observation seems to reflect
complex cognitive operations, including perspective taking.
Note that these points, if true, reflect poorly on Behaviorism

Copying what is seen


Bandura found a complex pattern of results:

Some children were provoked by taking away their first toy of

choice.
For children provoked, those who had viewed adult
aggression were more likely to be aggressive.
Often copied specific actions and words from the adult, demonstrating

clear use of the adult as a model.

For children unprovoked, those who had viewed adult

aggression were less likely to be aggressive themselves.

Social Learning
Banduras results challenge Behaviorist approaches in at

least two ways:


Strong learning in the absence of any specific reinforcement (self-

motivated learning)
Apparent need for complex cognitive processing (couldnt be
understood in simple S-R terms)

Bandura used the term modeling to describe all kinds of

copying
True-imitation: copying specific motor actions copying actions
Emulation: replicate the outcome without replicating specific motor

actions copying goals

Social Learning: More than one way to skin a cat


True Imitation: Copying specific actions learned from

another.
e.g., not just hitting Bobo but hitting it the same way the adult did and yelling

the same phrases


Demonstrates that the learner has encoded the specific actions of the model
and selected those precise memories to guide their ongoing behavior. Without
a doubt, cognitively complex.

Emulation: Accomplishing the same overall goal as the model but

in a different way.
e.g., A child might puncture Bobo, which is aggressive, but not in the

exact same way as the adult modeled.


Although fascinating, psychologically ambiguous. The learner could
have just wanted to achieve the same end and learned their own way
of accomplishing it. Doesnt demonstrate full cognitive representation of
the model.

Studies of True Imitation


Two-action test detects true imitation and is passed by

young humans and chimps.


Procedure:
Adult human models opened a box by

a) poking pins out of its latch or


b) twisting and pulling the pins out of the latch.
Next, young chimps and humans observed
to see if they would copy the precise
style of opening.

Studies of True Imitation


Results:
Human children precisely copied the
actions of the human model they
observed, demonstrating true imitation.
Young chimps exhibited a mix of true
imitation and emulation.
Interestingly, adult humans also show a
mix of imitation and emulation.

Young humans and chimps can exhibit true imitation in this two-

action test (though chimps less so than humans).


In other two-action tests, as least some imitation in birds and
rodents as wellindicating true cognitive representation of their
models.

Studies of True Imitation


Do-as-I-do task:
Animals trained on a do this command, where reward is given only if
the animal repeats the next behavior of the trainer.
e.g., Trainer signs do this, then claps her hands; monkey must clap
hands to get reward.
First, animal is trained on the concept of do this by using the a
standard set of actions.
Next, trainer demonstrates a novel action, never before rewarded.
Results:
Two chimps could complete 30 novel actions (Custance, Whiten, &
Bard, 1995).
Dolphins can also learn this game (Video of Educat).
Indicates an ability to cognitively represent the models arbitrary actions
and repeat them! Definitely beyond the bounds of behaviorism.

Social Learning: Reproducing actions without copying


Contagion: the observation of a given response increases the

likelihood that the observer will produce a similar response.


Not as a result of imitation; the matching reaction is instead often
an unconditioned response (UR)
e.g., A friend yawns, and then you do too.
e.g., Someone throws up, and you may too!
Not really learning, and definitely not true imitation.
Just a complex reflex. Doesnt require complex cognitive operations.

Social Learning: Reproducing actions without copying


Observational Conditioning: When cues in the environment

become associated with contagion reactions.


e.g., A nave monkey sees a snake. Other monkeys, who have

experienced snakes, freak out. The nave monkey freaks out too, due
to contagion. The snake now becomes linked to freaking out, so on the
next sight of the snake, the monkey now freaks out.
This allows social transmission of learned associations, but it fits all the
basic principles of classical conditioning.
Moreover, no complex cognitive processing is required. (Monkey
doesnt need to understand why the other monkeys are afraid; it simply
reacts to their fear and associates this with stimuli around at the time.)
This form of learning can really seem like true imitation, but clever
experimental design can show that it does not involve any complex
cognition

Social Learning: More than one way to skin a cat

Observational conditioning is dumb, not cognitively complex:


Model blackbird has learned to attack a stuffed owl, which preys on
blackbirds.
A nave blackbird observes attack behavior from the model, but is shown only
a plastic bottle, not the owl. Contagion sparks its attack behavior, but it pairs it
with whatever stimulus is around, in this case the bottle.
The nave blackbird learns to attack the bottle!
Obviously, not true imitation, just a complex form of classical conditioning.

Social Learning: Reproducing actions without copying


Stimulus Enhancement: When a teacher directs the attention of

the learner to particular parts of the environment.


e.g., Parents point out wet floor and tell child to take care. The child

may then be faster to learn that when the floor is wet, running will
cause falls.
The learner still undergoes traditional instrumental or classical
conditioning; the stimulus enhancement just makes associations more
salient (e.g., highlighting a discriminative stimulus).
Again, doesnt require complex cognitive operations.

Copying what is heard


Vocal learning: Learning to produce particular sound

patterns.
Again, there are many ways for an animal to learn to produce

a particular sound:
Instrumental conditioning: trial and error
Innate/fixed sound patterns (the ribbit of a frog)
True vocal imitation listening to sounds in the environment and

trying specifically to copy them; a form of true social learning.

Question: Does true vocal imitation occur? If so, in what

species?

Copying what is heard


Vocal Imitation is rare in the animal kingdom:
Seems to be limited: humans, some birds, some marine mammals
This is a rather strange set of species, with only distant relationships.
Especially strange is that our close primate relatives show now ability to
imitate sounds. Even traditional operant conditioning produces very
limited ability to train specific sounds.
It may be, though, that most animals lack the physical structures
necessary to control vocalization with the precision required for true
imitation.
Songbirds, however, provide an excellent system for studying

vocal imitation as a form of vocal learning

Copying what is heard


Songbirds learn socially:
If raised in isolation, correct
singing doesnt develop.
Song dialects are based on
region of the species.
Cross-fostering can lead to
adoption of foster parent song
style, but only if the chick
interacts with the parent.
Template model:
Memorize songs similar to
genetic template
Refine own production to
template
Learn when/why to sing

Social Transmission of Information


Social transmission of information: A process in which an

observer learns something through experiences involving


other agents.
e.g., If you see someone lose his or her money at a soda machine, you

probably wont try that machine yourself.

Social transmission of information is ubiquitous in human

society (TV, Internet, books, college classes).


Experiments also demonstrate social transmission in other

species, including rats

Social Transmission of Information


Social transmission of information: A process in which an

observer learns something through experiences involving


other agents.
e.g., If you see someone lose his or her money at a soda machine, you

probably wont try that machine yourself.

Social transmission of information can lead to social

conformity, a tendency to adopt the behaviors of the group.


Enables adaptive behaviors to spread rapidly through a set of

conspecifics (e.g., this cheese is a good food source).


Similar to blocking effects, though, social conformity can also impair
learning of novel solutions

Social Transmission of Information


Social conformity can block learning of novel solutions.
Procedure:
Demonstrator guppies trained to swim to one of two open holes in a net
to escape an aversive stimulus.
Nave guppies then tested with the presence of
the demonstrators and an additional, better
escape route.
Results:
Nave fish tended to follow the
demonstrators to the less optimal
opening.
Decreasing the number of
demonstrators deceased conformity.
Note: this also occurs without social
conformity: demonstrators also stick with their
initial escape route.

Social Transmission of Information

Centerwall (1992) found that TV ownership in the U.S.

increased in tandem with homicide rates (no change in South


Africa, where TV was banned).

Social Transmission of Information


Research has often shown correlations between exposure to

television violence and aggressive behavior.

How does social learning and transmission inform our use of

media?
Violent movies and video games
Sexualized behavior in the media
Images of health and beauty

Social Transmission of Information


Children exposed to videos of aggression often behave more

aggressively when provoked.


e.g., 7-9 year old boys who already scored high in aggression were more likely

to assault other boys during a hockey game after watching a violent video
(Josephson, 1987)
Unfortunately,
Most of this research examines only short-term changes.
Relatively little research into possible positive benefits (e.g., deciding to

become a medic after watching a war movie).


Still controversial whether or not there is a true causal relationship.
The National Institute of Mental Health, though, cautions that repeated

exposure to media violence may decrease sensitivity and increase


aggressive behavior.

Social Learning Theory


Early social learning theory adapted Behaviorist principles:
Focused on the incentives for a behavior: punishments and rewards
Saw social learning as a form of vicarious instrumental learning,
enabling learners to observe about the consequences of a behavior
without having to try it first hand:
Jill observes Bill crying, and sees his mom pick him up to sooth him.
Jill now knows that crying is reinforced and is more likely to try it.

Modern approaches take a more cognitive approach, seeing

social learning as self-motivated (not needing explicit


punishment/reward) and guided by cognitive appraisals.
Bandura has proposed 4 basic processes for true imitation

Social Learning Theory


Banduras account of social learning:
Presence of a model: increases/focuses attention to the situation
Encoding of models actions: learner must store memories of the
actions in an accessible format for later recall.
Action reproduction: learner must be able to reproduce the actions
encoded into memory (e.g., must be able to execute the dance move
observed).
Motivation: learner must have a reason to select the observed behavior.