Chapter 4 Classical Conditioning: Learning to Predict Significant Events
Modified: 2024-09-17 9:53 AM CDST
Surely you have heard of Pavlov and classical conditioning by now. Remember the discussion of association from chapter 1? Well, Pavlov was the first to provide a mechanism linking two stimuli. In classical conditioning an existing stimulus (a puff of air to the eye) will be followed by an unlearned response (eye blink). But, when a second stimulus (a sound) precedes the puff of air then, eventually, blinking will occur to the sound, not the air. In other words, the sound predicts that the air will follow. That is the simple story; there is much more. The search for the physiological basis continues as do efforts to use classical conditioning in clinical situations.
I. Behavioral Processes
Pavlov's Contributions
For seven years he and his students (who included women and Jews, both unusual in Czarist Russia) worked out the details of the digestive process using living dogs. Unlike other researchers, Pavlov developed physiological techniques that allowed him to collect data on how dogs digested their food in real time.
His student, Yekaterina Shumova-Simonovskaya pioneered the surgical creation of gastric pouches, or flaps designed to catch food swallowed before it reached the stomach (sham feeding), and the successful insertion of tubes (fistulas) into the various glands of the digestive system (Hill, 2019).
Slowly but surely, he and his coworkers became internationally known, and in 1904 he won the Nobel Prize for Physiology and Medicine, the first Russian to receive a Nobel Prize.
Beginning in 1897, however, and following an interesting discovery, the direction of the laboratory’s research became behavioral. Windholz (1989) identified Pavlov’s students and their contributions. One of them, Ivan Tolochinov, first noticed that dogs began to salivate to non-food cues; that they were learning to anticipate food when it was paired with a neutral stimulus. The evidence was that the dogs began to salivate to stimuli such as color and odor alone. All that was required was the close pairing in time of such neutral stimuli followed by a food stimulus.
Very quickly, Pavlov shifted his laboratory’s research direction to study this new Pavlovian conditioning process.
The question of how to theoretically account for it caused dissention between him and one of his students. After some thought and many arguments Pavlov rejected any explanations for conditioning that involved introspective elements.
Pavlov explained his approach in his Nobel Prize speech:
We decided to take an entirely objective point of view also towards the psychical phenomena in our experiments with animals. Above all, we tried to discipline sternly our way of thinking and our words and ignored completely the mental state of the animal; we restricted our work to careful observation and exact formulation of the influence exerted by distant objects on the secretion of the salivary glands. The results were according to our expectations: the observable relations between external phenomena and variations in the activity of glands could be systematically analyzed; they appeared to be determined by laws, because they could be reproduced at will.
Basic Concepts of Classical Conditioning
Classical
Conditioning
CS -->UCS is "formula for classical conditioning (e.g., metronome followed by food in dogs)
UCR--unconditioned response (also not learned, "built in", salivating, eating, drinking)
CS--conditioned stimulus, learned, (could be nearly anything, tone, light...)
CR--conditioned response, learned, (performing a response similar to UCR, but BEFORE UCS happens)
In classical, organisms learn to ANTICIPATE the UCS and perform the CR in response to the CS, not the UCS
Note: the response now occurs before it normally would.
Think of Pavlov's dogs: they will naturally salivate when they detect food, but in classical they now salivate when the hear the CS (the metronome) and salivate to it.
See the video near the end of this section for more detail.
Ivan Pavlov was a Russian physiologist.
During the early 20th century
he worked on describing the process of digestion.
He developed
techniques of collecting saliva from dogs.
He did so because
salivation is the first step in digestion.
He presented meat powder
to his dogs, and then collected the saliva.
However, he soon noticed
that as the dogs became familiar with the experimental routine they
began to salivate before they received the meat powder.
That observation led Pavlov to investigate the process whereby
dogs learned to associate the experimental routine with the meat
powder.
That decision marked the time that Pavlov's research became
psychological instead of physiological.
He systematically presented
stimuli, e.g., metronomes, lights, buzzers, to his dogs and then
followed those with meat powder.
He soon found that the dogs easily
learned to salivate to the presentation of nearly any stimulus associated
with the meat powder.
He called the meat powder the unconditioned stimulus or UCS, because dogs did not need to learn that association.
It was already
"built-in" to each dog.
He called the stimulus he chose to present
first the conditioned stimulus or CS.
He called it the conditioned
stimulus because the dogs would not salivate to it at first.
They
would only salivate to it after it had been associated with the UCS.
When the dog salivated to the meat powder, Pavlov called that
response the unconditioned response or UCR.
He called it that because
it occurred to the UCS.
When the dog salivated to the stimulus Pavlov
chose and presented prior to the UCS, he called that the conditioned
response or CR.
Again, he called it that because it occurred to the
CS.
Although the UCR and CR are very similar, they are not identical.
So, Pavlov's discovery was that a new stimulus can lead to a new
response similar to a naturally occurring stimulus and response.
The
keys to classical conditioning are the temporal (time) relationships
of the CS and the UCS, and also, how well the CS predicts the
likelihood of the UCS.
Let us review. Pavlov's discovery is now called classical or
respondent conditioning.
The UCS-->UCR relationship is naturally
occurring; it is not learned.
Salivation in dogs occurs to the
presentation of meat powder because that is part of the biology of
being a dog.
A similar example in humans is an air puff to the eye
following the presentation of a light.
Humans (and bunnies, see text) will blink if air is blown gently into their
eye. (UCS->UCR)
The CS (the light) at first leads to no observable behavior.
But, as the CS
is followed by the UCS over and over, do humans start to blink when the
CS is presented? Do they blink to a light when
that light (a CS) is presented prior to the puff of air.
So, one can think of classical conditioning as the pairing of a
CS and a UCS in time, with the CS occurring first, or, CS-->UCS.
What role, if any, does the time between the CS and the UCS play? It
plays a very significant role, it turns out.
Delayed conditioning is easily established. Delays of
around .5 sec. between the CS and the UCS lead to the strongest conditioning.
In trace conditioning a period of time is allowed to lapse
between the CS and the UCS.
It is called trace conditioning, because
a memory trace is presumably responsible for the conditioning.
Traces
of up to 12 sec. can be established in pigeons without a great deal
of effort.
Longer traces are not likely to lead to conditioning.
That
is why punishing children hours after they commit some infraction
does not lead to learning.
Later in the chapter, however, we will see that traces of
up to 75 minutes have been established with some pairs of stimuli. That is called taste aversion. See below.
In backwards conditioning, the UCS precedes the CS.
Typically, little or no conditioning results, although some have
succeeded in demonstrating backwards conditioning when
life-threatening shock was used.
Simultaneous conditioningleads to little or no
conditioning as well.
Simultaneous conditioning provides us with a
glimpse into the logic of conditioning.
A CS that does not predict
the UCS is of no value.
Temporal conditioning does not use an explicit CS.
Rather,
the UCS repeats itself on a regular basis.
Some animals may use that
repetition to achieve conditioning by timing the interval between UCS
onset and offset.
Example: Think of a curvy road with no signs telling you a curves is coming.
The Highway Analogy graphic below provides you with a useful analogy,
I think.
I saw this analogy presented at an APA meeting some years
ago, and I created the graphic for it later. I cannot recall who
presented it originally.
The analogy is useful because it shows how
the CS is a signal for the UCS.
In this analogy, the sign is the CS, the curve is the UCS.
Notice that putting the sign far away from the curve does not lead to conditioning, nor does putting the sign in the curve, or after the curve.
Summary
Pavlov was a Nobel Prize-winning physiologist who discovered
classical conditioning in the course of his physiological research.
Ivan PavlovHe eventually and begrudingly became a psychologist and in the
process helped change the direction of psychological research.
Classical conditioning became a major tool in the theoretical
formulations of later behaviorists. It allowed them to explain
behavior without having to consider consciousness.
Pavlov's discovery
of classical conditioning provided the first mechanism for explaining
learning without reference to the mind.
The long-standing doctrine of
associationism had its first explanatory mechanism.
Later research
demonstrated the importance of timing and correlation of the CS--UCS
pairing.
(See Rescorla-Wagner below)
Classical conditioning was the first salvo of the behaviorist
revolution.
Watson used Pavlov's methods to demonstrate the
Conditioned Emotional Response with Little Albert, for
instance.
American behaviorist theorizing took off with the
combination of Watson's manifesto and Pavlov's classical
conditioning.
John B. Watson's Little Albert Study
Albert showed fear of the white rat that had not scared him previously.
Five days later, they showed him a rabbit. Again, he showed fear and burst into tears.
He also showed fear (but not as much) to a white seal fur coat and to cotton wool.
The conditioned emotional response transferred.
Watson and Rayner waited a month to see if Albert’s fears disappeared over time. They did not.
Counterconditioning
Used to eliminate already conditioned behaviors
Aversive Conditioning is one type
For problem drinkers.
People will volunteer to take a nausea-inducing drug that will make them throw up shortly after consuming an alcoholic beverage
Drink->Nausea, Drink is the CS and Nausea is the CR
Taste Aversion
Garcia was studying classical conditioning in rats.
Nausea was the
UCR, and it was induced by two UCSs, either X-rays or LiCl (a salt).
Both of the UCSs will cause nausea in the appropriate dosage.
Other
groups of rats were given electric shock as a UCS.
The CSs were
either bitter water or plain water associated with light and sound.
So, some rats drank the bitter water and were made sick.
Others drank
the bitter water and received a shock. Still others drank the plain
water and were made sick.
Finally, some drank the plain water and
received an electric shock.
Then all were allowed to choose between the bitter water and the
plain water associated with the light and sound.
Those that had been
made sick would not drink the bitter water.
One group would not drink
even when the interval between the bitter water and the nausea had
been 75 minutes!
Talk about trace conditioning.
The rats that had
received the foot shock did not avoid the water in either form.
The main result of Garcia's research was to demonstrate that not
all CS-UCS pairings are equivalent.
This was a blow to established
learning theory (More recent research has shown that nearly any CS
can be made to act like the bitter water.
However, the procedures to
do so are different than Garcia's, so the effect of Garcia's research
remains.)
In a practical sense, a new field emerged, one that studies taste
aversions.
For example, suppose you have never eaten oysters.
You
arrive in New Orleans, are persuaded to try a dozen raw ones, but one
was bad.
Your next three days are spent retching.
What will you do
if, three months later, someone offers you an oyster? Will you recall your bad experience in New Orleans? Will you eat raw oysters again?
This
phenomenon is related to food neophobia, the fear of new foods.
We
are much more likely to become conditioned to the UCSs and UCRs
associated with new foods than to old foods.
So, if you were a longtime raw oyster eater
and ate a bad one, you will probably eat oysters again.
Finally, a very real problem is the issue of pest control.
Coyotes
have thus far resisted all efforts at control, including poisoning.
Poison baited carcasses have been set out in an attempt to reduce
their numbers.
The coyotes very quickly learn to avoid such baits.
Why? If they only eat enough to become sick, they learn to avoid such
new food sources.
The same is true with rats. They will learn to
avoid poison baits for the same reason.
Classical Conditioning Video
Video shows how Pavlov discovered classical conditioning
Jaws movie: Music is CS
At first, music plays, no reaction: it's a neutral stimulus
That means they are automatic and they will happen.
When classical condition occurs the CS will trigger the CR
CS properties
Nearly any stimulus in all modalities can become a CS.
Examples:
Dog sees owner's car leads to tail wagging
High school class ending bell leads to students leaving
Coffee smell in the morning leads to making own cup
Refining the Basic Principles
Tolerance and Homeostasis
Drugs such as adrenaline used in classical conditioning led to lower and lower heart rate
Larger doses were required to increase heart rate
See pp. 131-132
Can a CS be a UCS (pp. 132-133)
Yes
For Nathalie, sex is the CS
For Sharon, sex is the UCS; but her ex-boyfriend's voice is the CS
Extinction
In classical conditioning, quit delivering the UCS
Metronome-->Food to Metronome-->no food
Not just "unlearning" but "unlearning" plus "new learning"
Metronome used to signal food, now it signals absence of food
Spontaneous Recovery
One extinction episode is usually not enough to undo completely a
conditioned response.
So, when an animal is returned to the original
conditioning situation and allowed to make the conditioned response,
it will.
In order to obtain complete extinction, many extinction
trials will be required.
Think of how many trials were required for
the original conditioning.
It only makes sense that a similar number
of trials will be required for complete extinction.
Think about someone trying to quit drinking alcohol.
When they
return to the same situations where they used to drink, they must
battle spontaneous recovery.
That is why programs like Alcoholics
Anonymous try to prevent their members from returning to their old
haunts.
Another good example is when a recently divorced couple visit
each other.
They may, through spontaneous recovery, engage in an old
conditioned response (e.g., lovemaking, AKA 'sex with ex'), much to the chagrin of their
new partners.
Compound Conditioning (p. 135)
Two CSs at same time (e.g., light and sound)
Overshadowing
One of the CSs in a compound conditioning situation becomes more salient
Error Correction and the Modulation of US Processing
Informational Value of Cues
Redundant cues are ignored
Kamin's Blocking Effect (p. 137)
Rats learned that light predicts shock
Later, light plus sound used, followed by shock
Sound not conditioned, redundant
Rescorla-Wagner Model of Conditioning
The graphic above shows 10 experimental conditions
The number in each is the percent freezing response (freezing response explained in text)
Rats were shocked in silence (e.g., no warning) or after a tone (warning) 0, 10, 20, or 40 percent of the time
Notice how the percent freezing is maximal when rats were shocked only with warning and never without warning
Notice how rats just as likely to be shocked with or without warning hardly ever froze
Notice that it is the predictability of the shock, its correlation with the warning that determined the freezing response
It is NOT simply contiguity
Below is the original study.
Rescorla, R. A. (1968). Probability of shock in the presence and absence of cs in fear conditioning. Journal of Comparative and Physiological Psychology, 66(1), 1–5. https://doi.org/10.1037/h0025984
Error-Correction Learning
Errors lead to changes
Tennis:
Too far-->hit softer
Too short--> hit harder
In serve box near rear line--> make no more changes
Brain Substrates
Conditioning of Motor Reflexes in the Mammalian Brain
Cerebellum involved
Cellebellar Cortex
Contains Purkinje cells
Cellebellar Deep Nuclei
Interpositus nucleus is one (see Figure 4.17)
Electrophysiological Recording in the Cerebellum
Strong correlation between activity of the interpositus nucleus and eyeblink in trained rabbits (see Figure 14.8 A))
Purkinje cell (single-cell record) decreases during air puff CS
Brain Stimulation as a Substitute for Behavioral Training
Direct stimulation of the inferior olive causes rabbit to blink (with no CS)
