when you happen to walk by a particular room in a particular lab, and see something a little

unnerving. Namely, you find a woman punching an inflatable clown named Bobo in the neck.

Over and over in its neck. This was the lab of legendary psychologist Albert Bandura,

and in 1961 he was studying one of the most important phenomena in psychology.

See, while the woman was throttling that big inflatable clown, there was a child watching

her. And after about ten minutes of observing this clown-beating display, the kid was taken

to a room full of fun toys, which were soon taken away, and then the frustrated kid was

left alone with Bobo, and Bandura watched what happened. And yeah, what happened was

kind of scary. Kids who watched the woman beating the clown were much more likely to

mimic her aggression -- kicking, punching, throwing, even attempting to maul Bobo with

a hammer. But other children who saw an adult playing nicely with the doll, or just ignoring

it, didn’t respond the same way in their frustration.

Bandura’s results may seem predictable now, but in those days, they challenged the dominant

behaviorist views that we talked about last week -- the views that learning is solely

about conditioning and association, rewards and punishments. Bandura’s research focused

on how learning can occur through observing and imitating someone else’s behavior. And

if that seems obvious to you, you have Bandura to thank for that. His research hastened the

evolution of 20th century experimental psychology from pure behaviorism into what we now know

as social-cognitive learning. While it was closely related to behaviorism, the social-cognitive

models added profoundly new dimensions to what Skinner and Watson and Pavlov had observed

in our feathered and furry friends. In other words, it showed us that -- just as there’s

more than one way to beat up a clown -- there’s way more than one way to learn.

[INTRO]

Last week we talked about the differences between classical and operant conditioning

in associative learning -- the kind of learning that comes from connecting different events

and stimuli. In classical conditioning, this means associating a stimulus with some kind

of involuntary response -- the whole dog slobbering at the sound of a bell phenomenon -- whereas

operant conditioning makes associations between stimulus and a voluntary behavior -- like

the rat pressing a lever to get delicious snacks, or jumping out of a cage to escape

an electrical shock.

And that’s all well and good, but if learning is the process of acquiring and retaining

new behavior and information, then Bandura’s experiments showed us that conditioning with

external rewards, punishments, or other stimuli isn’t the only way to do it. It’s hard

to deny that pretty much all animals are capable of learning certain things by association,

but critics of behaviorists like Pavlov, Watson, and Skinner had a problem with their assertions

that, when it came to learning, it didn’t matter much whether you’re training rats,

pigeons, or people--it’s all the same. Because, lots of research has demonstrated that an

animal’s capacity for conditioning is actually limited by its biology.

Consider this scenario: Say I get a raging case of food poisoning after eating my head-weight

in raw oysters with my friend Bernice. I’m probably not going to want to touch oysters

again for a long, long time, because I associate their smell and taste with the smell and taste

they made when I was — when they were coming back out, is what I’m trying to say. But,

that doesn’t mean that the sight of Bernice, or the sound of the sea shanties they were

playing at the restaurant would make me barf, because humans are, by our very nature, more

taste averse than we are sight or sound averse. On the other hand, sight-oriented animals,

like birds, may be biologically predisposed to avoid tainted food by sight, since that’s

how they hunt and forage. And presumably they go to restaurants that play better music.

Anyway, all of this tells us that species can more easily learn associations that help

them thrive or survive, and that not all associations are learned equally. It’s a lot easier to

teach a pigeon to peck an X on the ground to obtain a food reward than it is get it

to flap its wings to get that same reward, because pecking is a natural foraging behavior

for a pigeon. In the same way, it would be much harder for the bird to learn to peck

that X to avoid a shock, rather than to flap its wings to avoid the shock, because flying

away from danger is what pigeons naturally do.

Learned associations are even more complicated in humans of course, because what we learn

doesn’t only influence our behavior, it also shapes our attitudes. Our cognition -- that

is, our thoughts, perspectives, and expectations -- is important for learning, as is our social

context, as Bandura figured out. So, Pavlov-style conditioning experiments that ignore those

social-cognitive elements can really run into trouble. For example, someone under treatment

for an alcohol addiction may be given booze laced with a nauseating drug. According to

the pure classical conditioning model, that person would then equate booze with feeling

nasty.

But the brain can override this association if it’s aware that it’s the added drug,

and not the alcohol, is the thing that’s causing the illness. Sometimes we can think

our way out of intended associations. And by the same token, a person’s social context

- like, their friends, family traditions, or life stressors - can reinforce something

like alcohol consumption more than the nauseating pill could ever punish it.

Plus, we also do a lot of latent learning, like without even knowing it. Have you ever

been walking around a new city, someone stops you to ask directions, and you surprise yourself

by actually being able to tell that tourist how to get to the park? That’s because we’re

constantly developing cognitive maps, or mental representations of our surroundings, without

explicitly telling ourselves to do it. We’ve all seen the experiments with in mazes: Well,

those show us that even rats develop these cognitive maps, figuring out how to get around,

even if there’s no reward at the end. And days later, when they finally do get food

at the end of the maze, they quickly demonstrate all that earlier latent learning by scuttling

through the maze as fast, or faster, than rats that had been rewarded all along.

So, learning isn’t just about associating a response with a consequence. There’s thinking

happening, too. And this kind of thinking is also a big part of observational learning,

which is basically learning by watching other people, or being influenced by them in other

ways. Because, you don’t need direct experience to learn. You can just pick up stuff up through

modeling -- not like modeling on the catwalk, I just mean observing and imitating specific

behaviors.

Rats, crows, pigeons, primates, and other animals learn through imitation. Chimps learn

how to use sticks to fish ants out of a nest this way. One study found rhesus macaques

were usually slow to make up after a fight unless they grew up watching more forgiving

older macaques, in which case they tended to make up more quickly. Of course we humans

learn A LOT from modeling -- I mean, most of our popular culture is based on it: new

slang, skinny jeans, foodie trends, pixie cuts -- they’re all racing around the globe

through observation and imitation. So it makes a lot of sense that social observation shapes

behavior, especially in children.

Which brings us back to Bobo. Again, the fact that we learn by imitating, even when we don’t

mean to, seems pretty intuitive, but until Bandura’s famous experiment, it hadn’t

been studied in a scientific way. I mean, these kids started abusing Bobo not just with

little toddler punches, but with hostile language and even using things, like toy guns, that

they previously had no interest in -- and all because they saw aggressive modeling in

action. And since Bandura’s time, technology has allowed us to peer even deeper into this

dynamic.

Neuroimaging in humans, for instance, has shown that when an individual watches someone

else, especially someone whom they relate to, receive an award or score a goal or something,

their own brain’s reward systems light up vicariously. Italian researchers found this

out pretty much by accident in the early 1990s: They were studying signals from key regions

in a lab monkey’s brain that were associated with planning and doing. Their brain-monitoring

device buzzed softly when the monkey did something like pick up a piece of fruit and eat it.

But one hot day, a researcher came back from lunch licking an ice cream cone, and suddenly

heard the animal’s brain monitor buzz -- the monkey was watching him, and his brain worked

as if it was actually doing the licking.

Many scientists suspect that this is the work of a previously unknown type of brain cell

called mirror neurons, which fire when a subject both performs an action, and when they observe

someone else doing it. Mirror neuron research is still relatively new, and we’re still

figuring them out, but combined with Bandura’s earlier work, it’s revealing a strong connection

between observation, imitation, and learning. So the takeaway here is: Models are important!

And not just Gisele and Antonio Sabato Junior. You can, if you want, observe and imitate

them; I’m just saying that observational, social learning starts really early, and parental

figures are powerful role models.

Positive, supportive, and loving models usually prompt similar behavior in others, just as

negative, aggressive modeling can spark antisocial effects. And, as we’ll talk about later,

what we see and feel and learn as children is not easily displaced when we’re adults.

Literary giant George Bernard Shaw wrote, “Imitation is not just the sincerest form

of flattery - it's the sincerest form of learning.” And British statesman Lord Chesterfield once

said, “We are, in truth, more than half what we are by imitation.” Even if these

ideas were only half-true, they’d still be a powerful lesson on who you choose to

spend your time with, and how you choose to act.

If you learned anything by watching me today, hopefully it involved the limitations of classical

and operant conditioning, the basics of cognitive, observational, and social learning, a look

at mirror neurons, and how to beat up a Bobo doll.

Thanks for watching, especially to all of our Subbable subscribers, who make this whole

channel possible. If you’d like to sponsor an episode of Crash Course, get a special

decal, or even be animated into an upcoming episode, just go to subbable.com. This episode

was written by Kathleen Yale, edited by Blake de Pastino, and our consultant is Dr. Ranjit

Bhagwat. Our director and editor is Nicholas Jenkins, and the script supervisor is Michael

Aranda, who is also our sound designer, and the graphics team is Thought Café.