Copyright © 2007-2018 Russ Dewey
Now we come to the labels for elements of classical conditioning. These are the terms students sometimes find difficult to remember.
The best way to learn these terms is to notice what they mean. For example, think of conditional as meaning conditional upon learning or dependent upon prior experience.
Pavlov labeled the signal that occurs before a biologically significant event a conditional stimulus (CS). Its ability to trigger salivation is conditional, that is, "depending on conditions." In other words, the conditional stimulus does not automatically trigger salivation in an untrained dog.
Another way to remember what CS stands for is to think of the word CUES. The conditional stimulus is what cues the conditional response. Pavlov sounded a tone (the CS) before giving the dog meat powder.
What do the abbreviations CS and CR stand for?
How did Pavlov initiate the biological reflex of salivation? He squirted meat powder in the dog's mouth.
Even before learning, salivation is triggered by meat powder. So this response is said to be unconditional (not dependent upon special learning conditions).
The meat powder is called the unconditional stimulus (US). It is the stimulus that provokes a response automatically and unconditionally.
Because the salivation response to meat powder is born-in and unconditional, it is called the unconditional response (UR). Every reflex listed on the previous page (gagging, knee jerk, etc.) consists of an unconditional stimulus that triggers an unconditional response.
Why is the word "unconditional" used?
To repeat: classical conditioning occurs when a signal (CS) is put before the reflex. Many textbooks will say the CS is "paired" with the UCS, but conditioning is fastest if the CS comes slightly before the reflex.
Conditioning occurs most quickly if the signal comes about a half second before the reflex is activated, in the case of movements like the fingertip withdrawal. With glandular responses like salivation, conditioning occurs fastest if the signal comes about two seconds before the reflex is activated.
In Pavlov's procedure, the tone was sounded before the meat powder was squirted into the dog's mouth. Then the dog anticipated the meat powder when it heard the tone. Therefore it salivated to the tone.
The learned response to a signal is the conditional response (CR). It depends on a prior act of learning, hence the name "conditional." After training, the CR occurs in response to the signal or conditional stimulus (CS).
The unconditional response and conditional response obviously resemble each other, but they are not identical. They have different causes (one is a reaction to the biologically natural stimulus; the other is an anticipatory reaction).
They also occur at a different strength. The conditional response is typically somewhat weaker than the unconditional stimulus.
How is the CR different from the UCR?
A dog does not salivate as much to the signal as it does to real food. Watson's subjects probably did not withdraw their fingertips quite as vigorously when they heard a bell compared to when they actually received a shock.
Therefore the two have different names. The original, reflex response is the unconditional response. The learned response is the conditional response.
Due to a mistranslation of Pavlov's early works, most psychologists refer to "conditioned responses" rather than "conditional responses." Conditional is the correct translation of the Russian word uslovnye (Fitzpatrick, 1990).
What mistranslation of Pavlov's terms occurred?
The word conditional is also more meaningful than conditioned. In the classic Pavlovian setup, the dog salivates because of previous experience. Therefore the CR is literally conditional upon the association of the CS with the reflex.
The word conditioned carries no such meaning, unless one make a comparison between the conditional response and conditioning a piece of metal by pounding it into shape–a metaphor Pavlov did not intend.
Why is "conditional" arguably more meaningful than "conditioned"?
Cyril Franks, a noted behavior therapist, wrote the following in an essay titled, "Behavior Therapy and Its Pavlovian Origins":
Pavlov distrusted absolutism in any form, and it was no accident that he employed the term "conditional" in his writings rather than "conditioned."
His intent was to convey the essentially temporary nature of the connections thus formed, connections that lacked the certainty and regularity of innate or "unconditional" reflexes. For Pavlov...the conditional reflex was a creative, emergent activity of the organism, not a stereotyped or unchanging process. (Franks, 1973)
What did Cyril Franks say about Pavlov's view of CRs?
Franks may have written "Pavlov distrusted absolutism in any form" to counteract a widespread belief that Pavlov sold out to the communist regime after the Bolshevik revolution in 1917. Lenin, leader of the Russian revolution, had the highest respect for Pavlov, one of his country's greatest scientists.
Lenin's endorsement made Pavlovian psychology the state-approved psychology of the Soviet Union. For many years, Russian psychologists believed classical conditioning could explain a large percentage of human behavior. Some American psychologists felt the same way about operant conditioning, which we will cover in the second half of the chapter.
Acquisition of the Conditional Response
Acquisition is the process of learning in conditioning experiments. In acquisition, a creature acquires a conditional response.
One lecturer, Thomas Landauer, then of Bell Laboratories in New Jersey, used a simple technique to demonstrate acquisition of a conditional response in a large introductory psychology class. While lecturing about Pavlov, he periodically raised a starter pistol, shouted "NOW!" and fired a shot.
Conditioning of a motor response like jumping occurs most quickly when the signal comes about half a second before the reflex is activated. Landauer's timing was close to optimum. He shouted "NOW!" about half a second before creating a loud noise that caused a startle reflex (made everybody jump).
How did Landauer quickly establish classical conditioning in a large crowd?
It was natural that everybody jumped in reaction to the gun report. A loud noise is an unconditional stimulus (US) that leads to an unlearned startle response (UR).
Five more times in the next 20 minutes of his lecture about classical conditioning Landauer raised his arm, shouted "NOW!" and fired his pistol. "NOW!" was the conditional stimulus (CS).
Then Landauer raised his hand and shouted "NOW!" but did nothing. Everybody jumped. He had conditioned the entire crowd in less than half an hour.
Why did the students jump?
Why did this happen? The essence of classical conditioning is putting a signal before a reflex so that an organism can get a "jump" or head start on the reflex.
Remember we described classical conditioning as a primitive form of prediction. Landauer's students were making an anticipatory response.
After hearing "NOW!" five times, followed immediately by a gunshot, their nervous systems learned the predictive relationship between the sound and the noise. When they heard the signal again, they responded (they jumped).
Of course, it was prediction of a simple sort. They were not planning to jump. They just did it. Classical conditioning does not require conscious intention.
The Acquisition Curve
Acquisition of a conditional response often occurs in just a few trials or repetitions of the CS-reflex pair. The Landauer example shows this.
Landauer only had to shout "NOW!" and fire the pistol five times before everybody was jumping to the word "NOW!" Experiments with animals show the same thing: a conditional response is set up quickly.
A typical acquisition curve involving 15 trials
The figure shows a learning curve. On the y-axis is the amount of saliva secreted during each training session.
Notice there is evidence of conditioning by the 7th trial. (Each trial is one pairing of the signal with activation of the reflex.)
How quickly can a conditional response be learned?
Unlike many textbook diagrams, which are smoothed, this particular example shows true-to-life variability. It goes up and down, not just up.
What "true-to-life variability" is shown in the curve?
Conditioning does not always proceed smoothly from no response to a large response, and there may be setbacks during training when the organism does not respond as much as before.
Extinction and Spontaneous Recovery
A classically conditioned response can be eliminated or extinguished by eliminating the predictive relationship between the signal and the reflex. This is accomplished by presenting the signal (CS) while preventing the reflex.
How can the reflex be prevented from occurring? One technique is to activate a behavior incompatible with the reflex. For example, a cat will be fearful of a box in which it receives an electric shock. The cat can be given "therapy" later by feeding it in the box.
Feeding prevents the anxiety response. So the cat loses its fear of the box.
This discovery led to a therapy called desensitization designed to eliminate fears and phobias. Desensitization was very successful. It marked the beginning of behavior therapy as a discipline.
If a CS occurs many times but the reflex is never activated, the organism learns that the signal no longer has the same meaning as before. The signal no longer predicts the activation of the reflex, and the conditional response disappears. This overall process is called extinction.
Landauer demonstrated extinction in his lecture. After establishing the conditional response (students jumping when he said, "NOW!") he extinguished it. He stuck up his arm periodically and shouted "NOW!" without firing the starter pistol.
At first people continued to jump. However, as he repeated the action, shouting "NOW!" every few minutes without firing the starter pistol, the response weakened.
Eventually nobody jumped. By repeatedly giving the signal without firing the gun, he eliminated the predictive relationship between the conditional stimulus ("NOW!") and the CR or conditional response (the jump). Soon there was no conditional response. Extinction had occurred.
How does one extinguish a classical conditioned response?
As time passes after extinction, a classically conditioned response typically recovers a bit and comes back. This recovery takes place spontaneously (without any additional training) so it is called spontaneous recovery.
What is spontaneous recovery?
After Landauer first extinguished the startle response, he lectured on a different topic for 20 minutes. Then, without warning, he pushed his arm in the air and shouted "NOW!" (still without firing the pistol).
Most of the crowd jumped again. Landauer had demonstrated spontaneous recovery.
Spontaneous recovery after the first extinction period
In order to completely extinguish a classically conditioned response, one must go through the extinction procedure repeatedly. Spontaneous recovery is extinguished each time it occurs.
Incidentally, this diagram is idealized: unrealistically smoothed. A genuine record of behavior would never be this clean and simple.
How does one completely extinguish a CR?
For the animal, spontaneous recovery is like testing the hypothesis that the predictive relationship will return. No sea slug would use those exact words, but in effect spontaneous recovery allows animals to avoid giving up too soon on a predictive relationship.
This would be adaptive tendency if (for example) a conditional response related to feeding did not work for one day. Instead of abandoning knowledge previously gained (linking odor to a food source) animals will respond to the cue again the next day.
That is spontaneous recovery. Maybe the predictive relationship will return, and if so, spontaneous recovery allows the animal to take advantage.
Pavlovian Conditioning: "It's not what you think it is"
Robert A. Rescorla of the University of Pennsylvania was probably the most influential classical conditioning researcher in the late 20th Century. He wrote a 1988 American Psychologist article titled, "Pavlovian Conditioning: It's not what you think it is."
Rescorla complained that psychologists had ignored developments in classical conditioning research occurring over the previous 20 years. Rescorla cited an out-of-date description of classical conditioning in an introductory psychology textbook.
[Quoting the textbook] "The essential operation in conditioning is pairing of two stimuli. One, initially neutral in that it elicits no response, is called the conditioned stimulus ( CS) ; the other, which is one that consistently elicits a response, is called the unconditioned stimulus (US).
The response elicited by the unconditioned stimulus is the unconditioned response (UR). As a result of the pairing of the conditioned stimulus (CS) and the unconditioned stimulus (US), the previously neutral conditioned stimulus comes to elicit the response. Then it is called the conditioned response (CR)."
Rescorla noted, "This description is typical of those found in both introductory and advanced textbooks 20 years ago." The problem was that the same sort of description was continuing to appear in 1988, when Rescorla wrote his article, despite "dramatic conceptual changes that had taken place."
In fact, it is still typical of how classical conditioning is described in many textbooks, an additional 30 years later. What is wrong with the old-fashioned description of conditioning? Rescorla cited these problems:
1. The textbook description emphasized contiguity [closeness in time] of the conditional stimulus and unconditional stimulus, while modern studies emphasize the informative or predictive nature of the conditional stimulus.
2. The textbook description gave the impression any two stimuli could be associated as conditional stimulus and unconditional stimulus. Modern research showed that some stimuli were much easier to associate with a particular biological response than others.
3. The old textbook description gave the impression conditioning was slow and gradual, requiring many repetitions or trials. Rescorla wrote, "Although conditioning can sometimes be slow, in fact most modern conditioning preparations routinely show rapid learning" requiring from 1 to 8 trials.
What were problems with many textbook descriptions of classical conditioning, in Rescorla's view?
Rescorla also wanted to change the image of classical conditioning. He said his professors back in the 60s conveyed the impression classical conditioning was "all spit and twitches," because famous experiments relied on salivation, eyeblinks, and finger withdrawal.
Rescorla pointed out that classical conditioning in the modern era is relevant to much more than spit and twitches. Classical conditioning "is intimately involved in the control of central psychological processes, such as emotions and motivation."
The following section of the chapter contains many examples that support Rescorla's arguments. Classical conditioning is now known to be involved with the immune system, sexual anticipation, tolerance to addictive drugs, and much more.
Fitzpatrick, S. (1990). Russians on the psyche: A review. Science, 248, 881-883.
Franks, C. M. (1966). Clinical application of conditioning and other behavioral techniques: Conceptual and professional considerations. Conditional Reflex : A Pavlovian Journal of Research & Therapy, 1, 36-50.
Rescorla, R. A. (1988). Pavlovian conditioning: It's not what you think it is. American Psychologist, 43, 151-160.
Write to Dr. Dewey at firstname.lastname@example.org.