As a physiologist who was actively hostile to psychology, which he viewed as an intellectual dead end, believing that the study of physical processes was the proper path to a better understanding of human nature, Ivan Pavlov might be amused to see that he has had such a profound impact on the ﬁeld. Ivan Pavlov was the discoverer of classical conditioning, also known as respondent conditioning, one of the basic learning mechanisms. He won a Nobel Prize in 1904 for his work on the digestive processes of dogs, during which his discovery of classical conditioning occurred as a fortunate accident.
Pavlov was engaged in the study of salivation. He had surgically implanted a tube into his dogs’ salivary glands to measure the amount of saliva secreted in response to the presence of food. He noticed that sometimes, however, the dogs would salivate even when no food was present, such as upon the arrival of the assistant who ordinarily fed the dogs. This was interesting because salivation was seen as a reﬂexive response to the food. Barring a taste for human ﬂesh, there was no good reason for a dog to salivate at the sight of a lab assistant. Pavlov reasoned that the salivation reﬂex had somehow been transformed into a new response, by the frequent pairing of the stimulus that naturally produced the salivation response (food) with a neutral stimulus (the lab assistant), creating a brand-new learned response.
He conducted a series of famous experiments to test the learning process, which he called conditioning (“classical” was added later, to distinguish it from B. F. Skinner’s operant conditioning). He ﬁrst selected a new neutral stimulus that should produce no particular response in a dog, a musical tone sounded by a tuning fork, in this case, and made the sound repeatedly, so the dog would become habituated to the sound and would no longer look over when he heard it. At this point, before conditioning has occurred, there are two stimulus response pairs: the dog food is the unconditioned stimulus (UCS) and salivation is the unconditioned response (UCR), so-called because the connection between them is a natural reﬂex that requires no learning. The neutral stimulus (musical tone), henceforth to be referred to as the conditioned stimulus (CS), meanwhile, produces no salivation. The next step is to conduct a series of trials in which the CS is paired with the UCS. After a sufﬁcient number of such trials, the CS by itself will produce salivation, now called the conditioned response (CR), to reﬂect the fact that it is now a learned response to a new stimulus:
- Phase 1—Unconditioned reﬂex: UCS (food) produces UCR (salivation)
- Phase 2—Conditioning trials: CS (tone) paired with UCS (food) produces UCR (salivation)
- Phase 3—Conditioning has occurred: CS (tone) produces CR (salivation)
If the food is later phased out, however, the salivation response to the tone will gradually disappear. This is called extinction. If the tone is again paired with the food, however, the salivation response will be relearned far more quickly than the ﬁrst time around. Once extinction has occurred, moreover, the conditioned response may suddenly reappear if, after some time has passed, the conditioned stimulus occurs again unexpectedly. This is spontaneous recovery. Furthermore, after a conditioned response has been acquired, stimuli which are similar to the CS may also produce the response, through a process called stimulus generalization. A dog trained to salivate to a musical tone may also salivate in response to other tones, for example. This process is balanced by stimulus discrimination, however; as experience teaches the dog that those other tones don’t precede the arrival of food, he will come to only respond to the original one.
Not all UCS-CS connections are equally easy to learn—we seem to be biologically more “ready” to learn some associations than others, a factor known as biopreparedness. For example, people are far more likely to develop fears of snakes or dogs than fears of compact discs or air fresheners. Conditioned taste aversions, also known as the Garcia effect, after the person whose research documented them, are an especially vivid example of biopreparedness. Most classical conditioning requires multiple pairings of the two stimuli for the learning to occur, as was the case with Pavlov’s dogs, and the unconditioned stimulus has to occur right after the conditioned stimulus. Poisons may not produce their effects for several minutes, or possibly even several hours, after they are consumed, and yet people (and animals) who experience food poisoning may never wish to eat the food that made them sick again.
Unlike most classical conditioning, taste-aversion learning occurs in one trial, with a wide gap between the stimuli. A classic taste-aversion study involved cancer patients, chemotherapy, and ice cream. A group of cancer patients ate a particular ﬂavor of ice cream one hour before receiving chemotherapy, which is notorious for producing tremendous nausea. A second group ate the same distinctive ﬂavor on a day when they did not receive chemotherapy, and a third group ate no ice cream. A full ﬁve months later, the patients were asked to taste several ice cream ﬂavors. The ﬂavor that was eaten ﬁve months earlier was chosen as the favorite by the no-chemotherapy and the no ice-cream groups, but the chemotherapy group found that ﬂavor of ice cream repulsive. This occurred despite a single exposure to the CS, hours before the UCS.
Biopreparedness is also a factor in the development of phobias, which are now widely regarded as a result of classical conditioning. Developing a fear of dogs, for example, requires only a single frightening experience with a dog. Consider fear as a UCS and distress and anxiety as the UCR—when the fear is preceded by the presence of a dog, distress and anxiety can easily become a CR in the presence of dogs. Phobias are so difﬁcult to get rid of because the person with a phobia will avoid the stimulus that produces the fear. Extinction requires one to experience the CS while not paired with the UCS—as long as the CS is entirely avoided, the event that needs to happen to produce extinction doesn’t occur.
- Bouton, M., Mineka, S., and Barlow, D. “A Modern Learning Theory Perspective on the Etiology of Panic Disorder.” Psychological Review, 108(1) (2001): 4–32;
- Garcia, J., Lasiter, P. S., and Bermudez-Rattoni, F. “A General Theory of Aversion Learning.” Annals of the New York Academy of Sciences, 443 (1985): 8–21.