If a dog can learn to salivate to the ringing of a bell, so can you. In fact, you probably have learned to salivate to the sound of a lunch bell, the phrase hot fudge sundae, and “mouth-watering” pictures of food. But classical conditioning affects us every day in many other ways.

One of the first psychologists to recognize the real-life implications of Pavlovian theory was John B. Watson, who founded American behaviorism and enthusiastically promoted Pavlov's ideas. Watson believed that the whole rich array of human emotion and behavior could be accounted for by conditioning principles. He even went so far as to claim that we learn to love another person when that person is paired with stroking and cuddling. Most psychologists, and nonpsychologists too, think Watson was wrong about love, which is a lot more complicated than he thought. But he was right about the power of classical conditioning to affect our emotions, preferences, and tastes.

Classical conditioning plays a big role in our emotional responses to objects, people, symbols, events, and places. It can explain why sentimental feelings sweep over us when we see a school mascot, a national flag, or the logo of the Olympic Games. These objects have been associated in the past with positive feelings:

Many advertising techniques take advantage of classical conditioning's role in emotional responses. When you see ads, notice how many of them pair a product with music the advertiser thinks you'll like, with good-looking people, with idyllic scenery, or with celebrities you admire or think are funny. In classical-conditioning terms, the music, scenery, attractive person, or celebrity is an unconditioned stimulus for internal responses associated with pleasure, and the advertiser hopes that the product in the ad will become a conditioned stimulus, evoking similar responses in you.

Positive emotions are not the only ones that can be classically conditioned; so can dislikes and fears. A person can learn to fear just about anything if it is paired with something that elicits pain, surprise, or embarrassment. Human beings, however, are biologically primed or “prepared” to acquire some kinds of fears more readily than others. It is far easier to establish a conditioned fear of spiders, snakes, and heights than of butterflies, flowers, and toasters. The former can be dangerous to your health, so in the process of evolution, human beings acquired a tendency to learn quickly to be wary of them and to retain this fear (LoBue & DeLoache, 2008, 2011; Öhman & Mineka, 2001). Evolution may also have instilled in humans a readiness to learn to fear unfamiliar members of groups other than their own, a tendency that could contribute to the emotional underpinnings of prejudice (Maia, 2009; Navarrete et al., 2009; Olsson et al., 2005).

When fear of an object or situation becomes irrational and interferes with normal activities, it qualifies as a phobia. To demonstrate how a phobia might be learned, John Watson and Rosalie Rayner (1920/2000) deliberately established a rat phobia in an 11-month-old boy named Albert. Their goal was to demonstrate how an inborn reaction of fear could transfer to a wide range of stimuli; today, we call this stimulus generalization. They also wanted to demonstrate that adult emotional responses, such as specific fears, could originate in early childhood. The research procedures used by Watson and Rayner had some flaws, and for ethical reasons, no psychologist today would attempt to do such a thing to a child. Nevertheless, the study's main conclusion, that fears can be conditioned, is still well accepted.

“Little Albert” was a placid child who rarely cried. (Watson and Rayner deliberately chose such a child because they thought their demonstration would do him relatively little harm.) When Watson and Rayner gave Albert a live, furry rat to play with, he showed no fear; in fact, he was delighted. The same was true when they showed him a variety of other objects, including a rabbit and some cotton wool. However, like most children, Albert was innately afraid of loud noises. When the researchers made a loud noise behind his head by striking a steel bar with a hammer, he would jump and fall sideways onto the mattress where he was sitting. The noise made by the hammer was an unconditioned stimulus for the unconditioned response of fear.

Having established that Albert liked rats, Watson and Rayner set about teaching him to fear them. Again they offered him a rat, but this time, as he reached for it, one of the researchers struck the steel bar. Startled, Albert fell onto the mattress. A week later, they repeated this procedure several times. Albert began to whimper and tremble. Finally, they held out the rat to him without making the noise. Albert fell over, cried, and crawled away so quickly that he almost reached the edge of the table he was sitting on before an adult caught him; the rat had become a conditioned stimulus for fear:

Tests done a few days later showed that Albert's fear had generalized to other hairy or furry objects, including a white rabbit, cotton wool, a Santa Claus mask, and even John Watson's hair.

Unfortunately, Watson and Rayner lost access to Little Albert, so we do not know how long the child's fears lasted. (Who Albert actually was has also been a matter of some debate, leading to a mix of speculation and investigation regarding his identity and ultimate fate [Beck & Irons, 2011; Fridlund et al., 2012; Powell et al., 2014].) Furthermore, because the study ended early, Watson and Rayner had no opportunity to reverse the conditioning.

However, Watson and Mary Cover Jones did reverse another child's conditioned fear—one that was, as Watson put it, “home grown” rather than psychologist-induced (Jones, 1924). A 3-year-old named Peter was deathly afraid of rabbits. To reverse Peter's fear, Watson and Jones used a method called counterconditioning, in which a conditioned stimulus is paired with some other stimulus that elicits a response incompatible with the unwanted response. Counterconditioning probably does not eliminate prior learning, but instead leads to new learning that supplants or overrides old learning. In this case, the rabbit (the CS) was paired with a snack of milk and crackers, and the snack produced pleasant feelings that were incompatible with the conditioned response of fear. At first, Watson and Jones kept the rabbit some distance from Peter, so that his fear would remain at a low level. Otherwise, Peter might have learned to fear milk and crackers! Then gradually, over several days, they brought the rabbit closer and closer. Eventually, Peter learned to like rabbits:

Counterconditioning

Peter was even able to sit with the rabbit in his lap, playing with it with one hand while he ate with the other. A variation of this procedure, called systematic desensitization, was later devised for treating phobias in adults.

Back when John Watson was using counterconditioning to help young Peter overcome his fear of rabbits, he focused on the observable associations between unconditioned and conditioned stimuli and responses. Techniques for understanding the biological basis of classical conditioning were limited. But today, technology and a better understanding of the brain are letting scientists explore methods for helping people overcome debilitating conditioned fears in ways that Watson could scarcely have imagined.

The amygdala plays a central role in the conditioning of fear, in part because of a receptor for the neurotransmitter glutamate. Giving rats a drug that blocks this receptor prevents extinction of a conditioned fear, whereas giving a drug that enhances the receptor's activity speeds up extinction (Schmidt et al., 2015; Walker et al., 2002). Inspired by these results, scientists set out to discover whether the same receptor-enhancing drug, which is safe in humans, could help people with a phobic fear of heights (Davis et al., 2005). Using a double-blind procedure, they gave the drug to 15 people with that phobia and a placebo to 15 others. All participants then underwent two therapy sessions in which they donned virtual-reality goggles and “rode” a glass elevator to progressively higher floors in a virtual hotel—an incredibly scary thing to do if you're terrified of heights. They could also “walk” out on a bridge and look down on a fountain in the hotel lobby. During each session, and again at 1-week and 3-month follow-up sessions, the participants rated their discomfort at each “floor.” Combining the therapy with the drug reduced symptoms far more than combining it with the placebo. Furthermore, in their everyday lives, people who got the drug were less likely than the controls to avoid heights.

Such research helps us to understand the biological mechanisms that underlie our fears, and the principles of behaviorism that may help us control and even overcome them. Learn more about these important topics by watching the video Learning to Overcome Phobias.

Classical conditioning can also explain learned reactions to many foods and odors. In the laboratory, behavioral scientists have taught animals to dislike particular foods or odors by pairing them with drugs that cause nausea or other unpleasant symptoms. One team trained slugs to associate the smell of carrots, which slugs normally like, with a bitter-tasting chemical they detest. Soon the slugs were avoiding the smell of carrots. The researchers then demonstrated higher-order conditioning by pairing the smell of carrots with the smell of potato. Sure enough, the slugs began to avoid the smell of potato as well (Sahley, Rudy, & Gelperin, 1981).

Many people have learned to dislike a food after eating it and then falling ill, even when the two events were unrelated. The food, previously a neutral stimulus, becomes a conditioned stimulus for nausea or other symptoms produced by the illness. Psychologist Martin Seligman once told how he himself was conditioned to hate béarnaise sauce. One night, shortly after he and his wife ate a delicious filet mignon with béarnaise sauce, he came down with the flu. Naturally, he felt wretched. His misery had nothing to do with the béarnaise sauce, of course, yet the next time he tried it, he found to his annoyance that he disliked the taste (Seligman & Hager, 1972).

Notice that unlike conditioning in the laboratory, Seligman's aversion to the sauce occurred after only one pairing of the sauce with illness and with a considerable delay between the conditioned and unconditioned stimuli. Moreover, Seligman's wife did not become a conditioned stimulus for nausea, and neither did his dinner plate or the waiter, even though they also had been paired with illness. Why? In earlier work with rats, John Garcia and Robert Koelling (1966) had provided the answer: the existence of a greater biological readiness to associate sickness with taste than with sights or sounds (the “Garcia effect”). Like the tendency to acquire certain fears, this biological tendency probably evolved because it enhanced survival: Eating bad food is more likely to be followed by illness and death than are particular sights or sounds.

Psychologists have taken advantage of this phenomenon to develop humane ways of discouraging predators from preying on livestock, using conditioned taste aversions instead of traps and poisons. In one classic approach, researchers laced sheep meat with a nausea-inducing chemical; after eating it just one or two times, coyotes and wolves still ran up to lambs, but instead of attacking, they retreated, hid, and vomited. They had developed a conditioned aversion to sheep (Dingfelder, 2010; Gustavson et al., 1974). Similar techniques have been used to control other predators—for example, to deter raccoons from killing chickens, and ravens and crows from eating crane eggs (Garcia & Gustavson, 1997).

Because of classical conditioning, medical treatments can create unexpected misery or relief from symptoms for reasons that are entirely unrelated to the treatment itself. Many patients know that unpleasant reactions to a treatment can generalize to a wide range of other stimuli. This is a particular problem for cancer patients. The nausea and vomiting resulting from chemotherapy often generalize to the place where the therapy takes place, the waiting room, the sound of a nurse's voice, or the smell of rubbing alcohol. The drug treatment is an unconditioned stimulus for nausea and vomiting, and through association, the other previously neutral stimuli become conditioned stimuli for these responses. Even mental images of the sights and smells of the clinic can become conditioned stimuli for nausea (Dadds et al., 1997; Redd et al., 1993):

Some cancer patients also acquire a classically conditioned anxiety response to anything associated with their chemotherapy (Jacobsen et al., 1995). However, patients may have reduced pain and anxiety when they receive placebos, pills and injections that have no active ingredients or treatments that have no direct physical effect on the problem. Placebos can be amazingly powerful, especially when they take the form of an injection, a large pill, or a pill with a brand name (Benedetti & Levi-Montalcini, 2001). In fact, placebos can act on the same neural pathways in the brain that real medicines do (Price, Finniss, & Benedetti, 2008).

Why do placebos work? Cognitive psychologists emphasize the role of expectations, at least in humans; expectations of getting better may reduce anxiety, and that reduction may have a positive effect on the immune system. Such expectations may also cause us to act in ways that produce the outcome we hope for (Michael, Garry, & Kirsch, 2012). But behaviorists emphasize conditioning: The doctor's white coat, the doctor's office, and pills or injections all become conditioned stimuli for relief from symptoms because these stimuli have been associated in the past with real drugs (Ader, 2000). The real drugs are the unconditioned stimuli, and the relief they bring is the unconditioned response. Placebos acquire the ability to elicit similar reactions, thereby becoming conditioned stimuli.

The expectancy explanation of placebo effects and the classical-conditioning explanation are not mutually exclusive (Kirsch, 2004; Stewart-Williams & Podd, 2004). As we saw earlier, many behaviorists now accept the view that classical conditioning itself involves the expectation that the conditioned stimulus will be followed by the unconditioned stimulus. Thus, at least some classically conditioned placebo effects may involve the patient's expectations. In fact, the patient's previous conditioning history may be what created those expectations to begin with.