Behaviorism and the
Cognitive Revolution
BCS 153, Spring 2016
Wilhelm Wundt
●
Wilhelm Wundt:
●
●
First real psychology lab (1879), first
journal
Pre-Wundt: There is not really a
science of psychology
–
–
●
Work is more philosophical – why
are were here?
No clinical or abnormal
psychology
Main tool: introspection
Introspection
●
Introspection – Study mental states by reporting your own
experience, states.
●
●
Psychology is the study of conscious states
“stomach ache” – called a stimulus error (instead, you want
a description of the pain)
An Experimental Study of Fear
(Conkin & Dimmick 1925)
An Experimental Study of Fear
(Conkin & Dimmick 1925)
The overall goal
●
Structuralism – Wundt & Titchener (Wundt's student)
●
●
Break down the components of consciousness
Analog of chemistry – find the basic elements and how they
combine
Flaws of introspection
●
No way to settle contradictory reports
●
Limited to humans (=no biology)
●
Not obviously useful for anything
●
Not exactly clear what happens in introspection
–
Might be more about our inferences/interpretations of our
own behavior
“It could take twenty minutes to describe the conscious
content of a second and a half and at the end of that
period the observe was cudgeling his brain to recall what
had actually happened more than a thousand seconds
ago, relying, of course, on inference” [Boring, 1953]
Choice blindness
(Johansson, et al. 2005)
Choice blindness
(Johansson, et al. 2005)
Telling More than We can Know: Verbal
Reports on Mental Processes
(Nisbett & Wilson 1977)
●
●
Highlighted limitations of verbal reports:
●
Subjects don't always know what caused a stimulus
●
Subjects don't always know that they've responded
●
Subjects don't always know that the stimulus caused a response
Propose that people don't directly observe their cognitive
processes.
●
●
Reports are based on a priori causal theories
“People use the contents of consciousness to construct a
personal narrative that may or may not correspond to their
nonconscious states”
Behaviorism
●
Behaviorism: Psychology is about behavior, not mental
states or processes.
●
●
●
Psychological “logical positivism” – statements in science
should be understood through the observables that make
them true or false.
Behaviorism was a response against introspection and
structuralism.
Two types:
●
Methodological – our science should be about behavior
●
Radical – internal states are not explanations of behavior
Watson
●
Originator of behaviorism
“Psychology as the behaviorist views it is a
purely objective experimental branch of
natural science. Its theoretical goal is the
prediction and control of behavior.
Introspection forms no essential part of its
methods, nor is the scientific value of its
data dependent upon the readiness with
which they lend themselves to interpretation
in terms of consciousness. The behaviorist,
in his efforts to get a unitary scheme of
animal response, recognizes no dividing
line between man and brute. The behavior
of man, with all of its refinement and
complexity, forms only a part of the
behaviorist's total scheme of investigation.”
Psychology as the Behaviorist Views it
Little Albert
Little Albert
●
Classical conditioning of a phobia in a human child.
https://www.youtube.com/watch?v=FMnhyGozLyE
The big question about human nature
●
Two major positions in cognitive psychology
●
●
Nativism – innate cognitive resources, structures,
algorithms, etc.
Empiricism – “tabula rasa” – everything comes from learning
Watson: “Give me a dozen healthy infants, well-informed, and
my own specified world to bring them up in and I’ll guarantee to
take any one at random and train him to become any type of
specialist I might select – doctor, lawyer, artist, merchant-chief
and, yes, even beggar-man and thief, regardless of his talents,
penchants, tendencies, abilities, vocations, and race of his
ancestors.”
Mary Cover Jones
●
Saw Watson lecture on Little Albert,
wanted to use behaviorist
techniques in clinical work
●
Desensitization – diminished
response with repeated exposure
to a negative stimulus.
“Peter was put in a crib in a play room and immediately became
absorbed in his toys. A white rat was introduced into the crib
from behind. (The experimenter was behind a screen). At sight
of the rat, Peter screamed and fell flat on his back in a
paroxysm of fear. The stimulus was removed, and Peter was
taken out of the crib and put into a chair. ... Invited to get down
from the chair, he shook his head, fear not yet subsided.”
“Progressive degrees of toleration”
A. Rabbit anywhere in the room in a cage causes fear reactions.
B. "
12 feet away in cage tolerated.
C. "
4 "
" " "
"
D. "
3 "
" " "
"
E. "
close
" "
"
F. "
free in room tolerated.
G. "
touched when experimenter holds it.
H. "
touched when free in room.
I. "
defied by spitting at it, throwing things at it, imitating it. [p. 311]
J. Rabbit allowed on tray of high chair.
K. Squats in defenseless position beside rabbit.
L. Helps experimenter to carry rabbit to its cage.
M. Holds rabbit on lap.
N. Stays alone in room with rabbit.
O. Allows rabbit in play pen with him.
P. Fondles rabbit affectionately.
Q. Lets rabbit nibble his fingers.
Pavlov
Conditioning
●
●
Unconditioned stimulus – elicits a response without any
training
Conditioned Stimulus – effect is through its association with
food (the bell)
Pavlov
●
https://www.youtube.com/watch?v=hhqumfpxuzI
Conditioning - Questions
●
What are some things you might want to know about
conditioning? Some experiments?
Basic phenomena in conditioning
●
Extinction – Decrease in effect with CS with no US
Basic phenomena in conditioning
●
●
●
Temporal: food every N minutes. Dogs start to salivate before
it's presented. What does this indicate?
Second order:
●
Metronome → food
●
Light → metronome
●
Results in: light → food
Zero contingency – CS occurs without US sometimes. Then
no/weaker CS → US link. Why? So what?
Basic phenomena in conditioning
●
●
Blocking
●
Metronome → Food
●
Metronome + Light → Food
●
NOT light → Food
Overshadowing
●
Bright light + Dim light → food
●
Bright light → Strong response
●
Dim light → Low response
Rescorla-Wagner model
Learning
●
Key variables
●
– associative strength to x
●
– change in associative strength
●
– salience
●
– learning rate
●
– maximum associability (0 for no US present)
Rescorla-Wagner model
Learning
●
Extinction
Key variables
●
– associative strength to x
●
– change in associative strength
●
– salience
●
– learning rate
●
– maximum associability (0 for no US present)
Rescorla-Wagner model
●
Key assumptions:
●
Associative strength depends on the sum of all cues
●
Constant salience/associability of each stimulus (alpha)
●
New learning is not history dependent
●
Monotonic increase with training
Rescorla-Wagner model
Extinction
Learning
Measure of the
prediction error
Rescorla-Wagner model
+light
+food
Amount added to
the initial
Rescorla-Wagner
●
What does the shape of the Rescorla-Wagner curve look
like?
Rescorla-Wagner
●
Extinction: what does the curve look like?
Phenomena captured
by Rescorla-Wagner
●
Acquisition/extinction curves
●
Overshadowing
●
●
Via manipulation of the salience parameter
Blocking
●
Via the source of change in V is prediction error; with little
prediction error, there is little learning
Rescorla-Wagner model
+light
+metronome
+food
Rescorla-Wagner model
+light
-food
Amount added to
the initial
Challenges to Rescorla-Wagner
●
●
●
Pre-exposure to CS – less conditioning
History of extinction can affect re-learning – extinct stimuli
are learned faster
What's one of the phenomena before that it can't get?
BF Skinner
Operant conditioning
●
Operant conditioning – Conditioning via reward and
punishment of arbitrary behaviors
(Contrast: Classical conditioning works with reflexive
behaviors.)
Superstition in the pigeon
●
What are the results?
●
https://www.youtube.com/watch?v=I_ctJqjlrHA
●
https://www.youtube.com/watch?v=vGazyH6fQQ4
Complex behavior in behaviorism
●
Associative chaining – chaining together of simple stimulusresponse pairs into a complex behavior
(e.g. walking from here to Wilson Commons by associating
each location with a next-location; as opposed to a mental map)
Project Pigeon
●
You can get very complex behavior in operant
conditioning: in WWII, Skinner tried to train Pigeons to
guide bombs.
Verbal Behavior
●
●
Alfred Whitehead – how could behaviorism ever
explain language and language learning?
Skinner writes Verbal Behavior
●
●
Emphasizes use of language in context, feedback
from parents as conditioners
Chomsky's Critique:
●
●
●
Skinner doesn't address “real” linguistic behavior;
doing so requires a theory of internal mental
representation
Children's knowledge appears to radically surpass
their input
Skinner's ideas are untested hypotheses
Downfall of behaviorism
(see Gallistel 1998)
●
Critically relied on a “blank slate.”
●
●
Is a blank slate obviously false?
Critically supposed that animals respond during learning.
●
This is false. Birds learn their parents song without explicit behavior.
Baby indigo buntings learn the geometry of the sky in order to direct
migration.
https://www.youtube.com/watch?feature=player_embedded&x-ytcl=84359240&v=sWZWonlJzSQ&x-yt-ts=1421782837
Downfall of behaviorism
(see Gallistel 1998)
●
Critically relied on a “blank slate.”
●
●
Critically supposed that animals respond during learning.
●
●
●
Is a blank slate obviously false?
This is false. Birds learn their parents song without explicit behavior. Baby
nestling buntings learn the geometry of the sky in order to direct migration.
Too constraining – assumes that we already know the relevant biology.
Example: we are lucky that early geneticists did not constrain their theory
to known biology at the time!
Many kinds of learning/information processing aren't associations.
●
Path integration (aka “dead reckoning”)
Path integration
Path integration
(Wehner 2003)
Downfall of behaviorism
●
Consider how hard it would be to explain the operation of an
ordinary computer with only references to input/output relations
Another troubling example for
behaviorism: mental rotation
Mental Rotation
Shepard and Metzler 1971
Rise of cognitive psychology
●
Really what we want is an integrative account of behavior.
●
This almost certainly requires statements about unseen
processes and representations.
Question due next Thursday
Think about Skinner's claim (at the end his first video link)
that if we knew all of the causes of behavior, there would
be no place for our intuitive conception of free will.
Is he right? Why or why not?
(Note: This is not a question about behaviorism. Even
cognitive theories try to work out the causes for behavior.)
Turn in: email to bcs153cognition@gmail.com