Classical conditioning has become so closely associated with the Russian scientist Ivan Pavlov (1849-1936) that it is often called Pavlovian conditioning. Although Pavlov conducted much of this research with dogs, examples of classical conditioning can be found in many human behaviors. Classical conditioning is a form of learning that occurs when two stimuli that are "paired"—presented together—become associated with each other. After the word paint was paired with a pleasurable activity, Ruby associated it with that activity. Similarly, the sight of the golden arches at McDonald’s and the taste of a juicy burger have occurred together, and as a result many people associate the golden arches with tasty fast food.
Pavlov and the Elements of Classical Conditioning
As noted earlier, the procedure for establishing classical conditioning is to present two events—called stimuli—to a participant so that the pairing of these two events causes our participant or animal subject to make an association between them. At the start of conditioning, the first event, which in a laboratory setting may be the presentation of a light or a tone, is neutral with regard to the response to be established. When this neutral stimulus (NS) is presented, the participant may notice that it is there, but it does not cause any particular reaction. However, by presenting the second event, called an unconditioned stimulus (US), after the first event, we transform the NS into a conditioned stimulus (CS). The NS becomes a CS because it is paired with a US. This pairing causes the participant to establish an association between the two events.
As the term suggests, the US automatically produces a reaction; the participant does not have to be trained to react to it. The US never fails to produce the same reaction. Food in your mouth causes you to salivate; touching a hot stove causes you to jerk your hand away. In psychological terms, the US elicits, or calls forth, a response. The reaction that is elicited by the US is called the unconditioned response (UR). If you have a feeling that we have already discussed this type of response, you are correct. These URs are reflexes, just like those we described in Chapter 2. You do not have to learn a UR. For example, you do not learn to jerk your hand away when you touch a hot stove; you pull it away automatically. It is a built-in (unconditioned) response.
When a participant associates the NS (for example, a light or a tone) with the US, the NS is transformed into a CS that can elicit a response similar to the UR. The response elicited by the CS is known as the conditioned response (CR). When the CS elicits the CR, we say that classical conditioning has occurred. Pavlov (1927) used food as the US in his pioneering studies. While a metronome was ticking (CS), he placed a small amount of meat powder (US) into a hungry dog’s mouth. The meat powder caused the dog to begin salivating (UR). Later, when just the sound of the metronome ticking was presented, the dog salivated (CR).
Recall the demonstration with lemonade powder described at the beginning of this section. What was its purpose? The demonstration associating lemonade powder with the word now (Cogan & Cogan, 1984) is an easy way to experience classical conditioning firsthand. To understand why, complete the following sentence regarding that experiment: "The CS, _____, paired with the US, _____, results in the UR, _____." The CS is the word now. The lemonade powder is the US; it automatically elicits the unconditioned response (UR) of puckering and salivating. After the word now has been paired with the lemonade powder several times, it elicits the conditioned response (CR) of puckering and salivating.
Let’s put these elements together in another example. Suppose that your younger brother is just tall enough to reach a hot skillet on top of the stove. He grabs it and immediately drops it. His pain is obvious, and you try to comfort him. Finally his tears stop, and you put the incident out of your mind. Three days later, however, the same skillet is again on the stove. Your brother enters the kitchen, sees the skillet, and begins to cry. Clearly, the skillet has taken on a new meaning for him.
In classical conditioning terms, initially the skillet was an NS; after conditioning, it became the CS. The intense heat was the US, which always elicits pain and an avoidance response. Those responses—the pain and jerking your hand away or dropping the skillet—constituted the UR. Remember, the classical conditioning sequence involves first presenting the NS and then following it with the US. If these two events are associated, the NS becomes a CS that signals that the US is on its way.
After a conditioning experience of this type, when the CS is encountered alone, it produces a response, the CR, that is very similar to the UR. In our example, the sight of the skillet reminds your brother of the pain he experienced. The classical conditioning sequence is presented in diagrammatic form in Figure 6-1.
Condition Your Friends
The ease by which classical conditioning is demonstrated in class can be duplicated in real life. This exercise was developed by Mark Vernoy (1987) and rests on the premise that most of us have been conditioned to flinch when we see someone stick a balloon with a pin. (The pin is the CS, the bang is the US, and the startle response is the UR.) "In this way we have learned that needles always pop balloons" (Vernoy, 1987, p. 177). By using the following procedure, you may be able to surprise your friends and observe classical conditioning at the same time.
According to Vernoy (1987), "The equipment needed for this demonstration includes about 20 to 30 good quality, round balloons and a needle. Any sharp sewing needle will do, but for dramatic effect I use a foot-long needle that I borrow from a colleague who is an amateur magician. You can acquire these needles at any good magic shop". The procedure is simple. Start by blowing up several balloons; 15 or 20 should be sufficient. Then have your friends pop 5 or 6 of them with the needle you provide. Next, you should use the needle to pop 5 or 6 more balloons. Once you have popped several, stick the needle into an area of the balloon where there is less tension (such as the nipple or around the knot). Because there is less tension at these points, the rubber is relatively thick, and the balloon does not pop when it is stuck.
However, your friends still flinch. Why? This reaction occurs because your friends have been conditioned to expect a loud bang. If you are using a foot long needle, you can make the effect even more dramatic by passing the needle completely through the balloon (enter at the nipple and exit at the knot).
Phobias
When Scott was 3 years old, he was bitten by a poisonous snake and nearly died. Ever since, he has avoided snakes. Now, 30 years later, he is still deathly afraid of snakes and anything that reminds him of them. Seeing a piece of rope or a picture of a snake, or even hearing the word snake, makes him break out in a cold sweat. If you did not know Scott’s background, his intense fear of snakes and related items might seem rather strange. You may not fear snakes to the same degree as Scott, but the chances are good that you are afraid of some objects or situations that most other people do not fear. How do we acquire these apparently unrealistic, irrational fears?
Many of our fears and anxieties may have been classically conditioned, as in the case of Scott’s fear of snakes. Scott has a condition known as a phobia; more specifically, he is suffering from ophidiophobia (ophidio, "snake"; phobos, "fear"). A phobia is an irrational fear of an object, situation, or activity that is out of proportion to the actual danger it poses. Because phobias create so much anxiety that they interfere with normal functioning, they are classified as anxiety disorders (see Chapter 14).
We frequently hear about people who have claustrophobia, an intense fear of enclosed places. Some people may develop claustrophobia because they were locked in an abandoned refrigerator (US) when they were children and nearly died of suffocation (UR). Now they fear anything that remotely resembles a closed space (CS)--elevators, compact cars, small rooms, and so forth.
As you might expect, phobias can interfere with a person’s daily activities. For example, a business executive suffering from claustrophobia would not do very well in New York City, where meetings are frequently held on the top floors of tall buildings. Climbing the stairs to avoid the elevator could be time-consuming and very tiring. Psychologists have developed a procedure, known as systematic desensitization, to help eliminate phobias. Basically, systematic desensitization involves classically conditioning a desired response, relaxation, to the phobic stimuli. Thus a person with claustrophobia is conditioned to relax in enclosed spaces. We will discuss systematic desensitization in greater detail in Chapter 14.
John Watson; Little Albert, and the Ethics of Research
In 1913, John B. Watson proclaimed that psychologists should study only directly observable behaviors. As we saw in Chapter 1, Watson’s approach to psychology was called behaviorism. The main business of psychology, according to the behaviorists, is the study of behaviors such as jogging in the park, leaving the scene of an accident, running through an airport so as not to miss a plane, and even expressing emotion. Anything having to do with thinking, feeling, or consciousness was not an appropriate subject of psychological study because those processes could not be observed directly.
The behaviorists’ goal was to discover which stimuli elicit which responses (observable behaviors). In an experiment that applied this approach to human emotions, Watson and his assistant, Rosalie Rayner, classically conditioned a 9 -month-old infant, "Little Albert," to fear a white rat (Watson & Rayner, 1920). Initially Albert showed no fear of the rat and even allowed it to crawl on him. While Albert was playing with the rat, Watson hit a large steel rod with a heavy hammer, making a sudden, deafening noise. Not surprisingly, Albert was startled and scared. Each time the loud noise was paired with the presence of the rat, Albert cried in fear. After numerous pairings of the two stimuli, Albert started crying at the sight of the rat, even when there was no noise, and he eventually came to fear any object that resembled a rat, such as a white rabbit and even the white whiskers on a Santa Claus mask. Just as Scott developed a fear of snakes and things that reminded him of snakes, Albert developed a phobia for rats and ratlike objects.
Let’s analyze the elements of Little Albert’s fear. What US was used? What was the UR? What were the CS and the CR? While you think about these questions, do not forget that the rat was also exposed to a frightening situation. For both Little Albert and the rat, the US was the loud noise. The UR was being startled and scared. For Albert, the CS was the rat; for the rat, the CS was Albert. The CR for both of them was fear of an object that signaled that a loud noise might follow.
Ethical Principles of Psychologists and Code of Conduct, a manual published by the American Psychological Association, would probably say no to all of our questions (American Psychological Association, 1992). If Watson were conducting his research with Little Albert today, he would have difficulty meeting the ethical standards you read about in Chapter 1.
Pleasant Unconditioned Stimuli
Up to this point we have focused on USs that produce disagreeable or unpleasant effects. Thankfully, the US does not have to be a painful event like heat, electric shock, or hitting your finger with a hammer (Capaldi & Sheffer, 1992;Owens, Capaldi, & Sheffer, 1993). A bite of your favorite food when you are hungry automatically causes you to salivate. Food in your mouth is the US; salivation is the UR.
Obviously, psychologists do not hide in restaurants to present a tone while you are eating, nor do most instructors bring lemonade powder to class with them. Yet people become classically conditioned in much the same way that Pavlov’s dogs did. For example, the sights and sounds that accompany our meals can become CSs. For many people the unique decor of a restaurant or even a menu may act as a CS. Have you ever found yourself salivating as you looked at the tempting pictures on a menu or browsed up and down the aisles of a grocery store?
Examples of classical conditioning of pleasant CRs abound in everyday life. Is there a particular song (CS) that prompts you to recall a happy moment (CR)? Do you know someone who purchases only cars of a certain color because that color is associated with a favorite car?
Other Aspects of Classical Conditioning
Pavlov’s research revealed several additional characteristics of classical conditioning besides those discussed so far. These findings fall into two categories: acquisition, or how we acquire CRs, and extinction, or how we eliminate those responses.
Acquisition.
Acquisition is the training stage during which a particular response is learned. Several factors influence the acquisition of CRs. Among them are the sequence of CS-US presentation, the intensity of the US, and the number of times the CS and US are paired. We now take a closer look at each of these factors.
Sequence of CS-US presentation. The sequence in which the CS and US are presented influences the strength of conditioning (Keith-Lucas & Guttman, 1975; Sherman, 1978). Up to this point we have simply indicated that the CS should precede the US. This statement is rather imprecise. Several sequences for presenting the CS and US are possible (see Figure 6-3):
1. The CS comes on and goes off before the US is presented. This arrangement is called trace conditioning because the US is associated with a memory trace of the CS, not with the CS itself. It produces weaker conditioning, but not as weak as simultaneous or backward conditioning .
2. The CS comes on and stays on, and then the US is presented and the CS and US occur together. This arrangement is called delayed conditioning because the presentation of the US is delayed for a specified interval after the CS has been presented . Researchers have found that delayed conditioning in which the CS precedes the US by a short interval produces very strong conditioning, whereas longer delays between the CS and US produce weak conditioning. The optimal delay varies according to the type of response being conditioned; 450 milliseconds (one millisecond is one one-thousandth of a second) is the optimal interval for conditioning the eyelid closure reflex (a puff of air on the eye reflexively causes an eye blink; Smith, Coleman, & Gormezano, 1969).
3. The CS comes on at exactly the same time as the US. For example, the sound of the metronome and the food powder would be presented to a dog simultaneously. This arrangement, known as simultaneous conditioning, results in weak conditioning .
4. The CS comes on after the US. For example, the dog would be given food powder and then would hear the sound of the metronome. This sequence, known as backward conditioning, produces very weak, if any, conditioning
Strength of the US.
The stronger the US, the stronger the conditioning (Prokasy, Grant, & Myers, 1958; Holloway & Domjan, 1993). When Pavlov gave his dogs a small amount of food powder, they did not salivate as much as they did when he gave them a large amount of food powder. Stronger USs elicit stronger URs; weaker USs elicit weaker URs.
Number of CS-US pairings.
The more times the CS and US are presented together, the stronger the CR becomes. It is easy to conduct research on the relationship between the CS and the US when we can use a laboratory setup like Pavlov’s. Under such conditions, the number of times the CS and US are presented can be determined precisely. However, the effects of varying numbers of CS-US pairings can also be demonstrated in real life. If your little brother grabs the hot skillet more than once, the chances are good that his CR to the skillet will be stronger. If you have eaten at an exceptionally fine restaurant several times, your CRs to the sight of the restaurant and its menu will be stronger than they would be if you had eaten there only once. Figure 6-4 displays several patterns of classical conditioning acquisition; as you can see, as the percentage of pairings increases, acquisition of the CR becomes stronger.
Extinction.
Once a CR has been acquired, what can be done to extinguish, or eliminate, that response? The easiest procedure is to present the CS without the US and record how strong the CR is and how many times or how long we can present the CS alone before the CR disappears. The number of times the CS is presented without the US is a very important factor in eliminating the CR (Monti & Smith, 1976; Shipley, 1974). When Pavlov repeatedly sounded the tone (CS) without giving the dog any meat powder (US), the number of drops of saliva that were produced gradually decreased each time the tone was sounded. Similarly, if your brother were to grab the skillet several times when it was cold, his fear would decrease a little each time.
The process through which the strength of the CR is decreased is called extinction (Schreurs, 1993). The stronger the CR, the longer extinction takes. What takes place during acquisition influences the process of extinction. For example, the stronger the US and the more frequently it is presented during acquisition, the longer it will take to extinguish the CR (Hull, 1943).
Spontaneous Recovery.
At times a classical conditioning participant seems to "forget" that extinction has occurred. Consider Pavlov’s dog once again. Only the CS (a bell in this case) has been presented to the dog several times during its daily extinction session. The CR (salivation) has decreased until it appears that the dog is not salivating. The dog is returned to its cage, and we might conclude that the extinction process was complete. On the following day the CS is presented, and the dog begins to salivate once more. Pavlov (1927) called this phenomenon spontaneous recovery because the CR recovers some of the strength it lost during the previous extinction session. This process is diagramed in Figure 6-5. Notice that the amount of spontaneous recovery decreases from day to day until the CR finally does not occur at the start of a session. Extinction is probably complete at this point.
Spontaneous recovery is not limited to laboratory experiments; it occurs in real-life situations as well. For example, assume that you went skiing last winter and on the second day you had a bad fall. After summoning much courage and determination, you put your skis on and went back out on the slopes for the next 2 days. Thankfully, nothing adverse happened on those days; your fear seemed to be extinguished. Now, a year later, you have returned to the ski resort. You experience some apprehension as you pull on your ski boots. You thought that the fear had gone—that it was extinguished—by the time you were finished skiing last year, but a bit seems to have returned (spontaneous recovery) this year.
Generalization and Discrimination.
Suppose that several days after your little brother made the mistake of grabbing the hot skillet, you and he are walking through the housewares section of a local department store when he starts crying and refuses to walk any farther. You look up and see a display of skillets. Is he afraid of them also? Yes, he is, and the more those skillets resemble the one at home, the greater is his fear.
Psychologists (such as Hovland, 1937; Razran, 1949) call this phenomenon generalization. By this they mean that the effects of classical conditioning may be generalized to other stimuli that are similar to the original CS; they "spread" from the original stimulus to others. For example, although Pavlov’s dogs were conditioned to salivate in response to a specific tone (CS), they also salivated when other tones were presented. Likewise, we saw that Little Albert’s fear of the white rat generalized to other objects that were white and furry. An early study using humans as subjects (Hovland, 1937) involved pairing a tone (CS) with a mild electric shock. As you can see in Figure 6-6, when different tones were tested, the CR still occurred but was weaker than the response to the original tone.
It is easy to see how generalization occurs. If you have ever been stung by a wasp, you probably have a healthy respect for all flying insects, especially those that resemble wasps. Your response has generalized from one stimulus to many others. But if you are not stung every time you encounter a flying insect, many of these generalized responses will be extinguished. Thus as children we quickly discover that butterflies and moths do not sting, whereas hornets and bees do. We therefore come to fear the stinging insects but not the others. In other words, we learn to distinguish or discriminate between CSs that accurately predict the occurrence of the US and those that do not (Bouton & Brooks, 1993; Nakajima, 1993). Through the process of discrimination, we have extinguished our fear of insects that do not sting but have retained our fear of insects that do.
If you think about it, generalization and discrimination work in opposite ways. Generalization makes you more likely to respond to a number of similar stimuli; discrimination narrows your responding to the appropriate stimulus and no other. Discrimination thus requires stimuli that are clearly distinguishable (Brown, 1942; Gantt, 1971). If you could not easily distinguish between insects that sting and those that do not, for example, imagine how apprehensive you would be whenever you went outdoors. As we will see in the next section, such apprehension can influence our motivated behavior.
Classical Conditioning and Our Motives
Even though Jim did not have cats as pets when he was a child, his friends persuaded him to adopt the cute stray kitten he found last week. Jim is now convinced that this was a very bad idea. Each evening, when Jim has settled into his favorite chair to watch the evening news, the kitten launches a sneak attack. After a week of this behavior, Jim becomes tense and anxious as soon as he sits in his favorite chair, whether the kitten is in the room or not. The sound of the evening news increases his anxiety. When Jim leaves the room, he immediately feels better.
What is the relationship between the kitten’s attacks, Jim’s tension and anxiety, and motivation? Parts of the story about Jim and his cat may sound familiar. Can we describe the cat’s sneak attack in psychological terms that you have already learned? Would certain responses automatically follow such an attack? If you said pain, fear, or anxiety, we would agree. What kind of stimuli automatically elicit a response? If you are thinking that unconditioned stimuli (USs) elicit unconditioned responses (URs), you are right again. What about Jim’s favorite chair and the evening news program? What role do they serve? After these stimuli have been associated with cat attacks several times, they cause Jim to become tense and anxious. These are conditioned stimuli (CSs), and the tension and anxiety they produce are conditioned responses (CRs).
How is Jim’s conditioned anxiety related to motivation? Through the process of classical conditioning, Jim’s favorite chair and the evening news have become CSs that elicit tension and anxiety. Jim finds these feelings of tension and anxiety unpleasant and is motivated to reduce them when they occur. Motives that are acquired through the process of classical conditioning are called learned motives. Many other motives are acquired in this manner. Phobias are excellent examples of learned motives. Classical conditioning appears to be at the core of many of these unusual fears.
The same procedure and logic can be applied to the learning of goals and incentives. The importance of many of the goals and incentives that motivate our behavior is also learned through classical conditioning; hence they are termed learned goals or incentives. Consider money, diamonds, gold, and concert tickets. An infant’s response to these objects will quickly convince you that they do not possess intrinsic value; we must learn their value before their acquisition is reinforcing.
New Directions in Classical Conditioning
Our conception of classical conditioning has changed dramatically since Pavlov’s time. We now know that conditioning is not an automatic process that simply links CSs and USs. Today psychologists place more emphasis on what information the CS conveys to the subject—that is, what the CS seems to tell the participant.
Contingency Theory.
One principle that has emerged from recent research is that the better the CS is able to predict the occurrence of the US, the stronger the conditioning will be (Bolles, 1979; Rescorla, 1968). A study by Robert Rescorla (1968) clearly illustrates this point. Rescorla studied classical conditioning in two groups of rats. For one group, the CS (tone) was always followed by the US (a shock). These animals were called the contingent subjects because the occurrence of shock always followed (was contingent on) the tone. The animals in the second group heard the tone before or after the shock was presented. These animals were called the noncontingent subjects because the occurrence of shock did not depend on (always follow) the tone. Because the tone perfectly predicted the US for the contingent animals, classical conditioning was strong. Classical conditioning was weaker for the noncontingent animals because the CS did not always precede the US.
This relationship should not be surprising. As we have seen, a strong and predictable CS is the goal of the acquisition or training process during which an association between the CS and US is formed. A particular CS predicts that a particular US is about to occur. For Pavlov’s dogs, for example, the sound of the ticking metronome reliably predicted the delivery of food. Conversely, when we undertake extinction, we try to convince the participant that the CS will no longer be followed by—will no longer predict—the US (Delemater, 1995). The more reliable the CS is in predicting the US, the harder it will be to extinguish. However, as we will see in the next section, CSs do not automatically become associated with USs, even if they are predictable.
Blocking.
If classical conditioning simply involves the pairing of a CS and US, the time at which the pairing occurs should not make any difference. But it does. Animal researcher Leon Kamin (1969) demonstrated the importance of when the pairing occurs. In Kamin’s study, one group of rats was classically conditioned by presenting a tone (CS) and following it with an electric shock (US). Once the tone had been conditioned, a second CS (a light) also was presented before the shock. A second group of animals received only pairings of the light and the shock. Kamin then tested the strength of conditioning to the light. Conditioning was weaker for the animals that received tone-shock pairings before tone and light were paired with shock. Stronger conditioning was shown by the animals that received only the light-shock pairing. Kamin reasoned that conditioning of the tone before presenting the light had blocked or reduced the conditioning of the light.
If you consider predictability, this interpretation makes perfect sense. If the tone had already been established as a predictable CS, another predictable CS was not needed (see Williams, 1994). The prior condition blocked the light -and-shock association for the animals that had already received tone-and-shock pairings. Because light was the only CS presented to the other group of animals, it was conditioned strongly.
Suppose that you have just moved to a large metropolitan area, which is vastly different from the small town you left. For example, the variety of restaurants is amazing. Last night some friends took you to a seafood restaurant. The decor and atmosphere of the restaurant were intriguing, and the taste of the food was unlike that of any you had ever had. Unfortunately, however, during the night you came down with the stomach flu that had been going around. For the rest of the night you were nauseated or worse—not a pleasant experience. These events may not seem to involve learning, but as you will see, classical conditioning took place: You were conditioned to avoid seafood.
Taste Aversion.
Does becoming nauseated after eating an unusual food have anything to do with learning? The notion of predictability is useful here. Whenever a person or animal becomes ill after consuming a food with a novel taste (taste-aversion conditioning), that taste (CS) may become an excellent predictor of illness.
In the mid-1960s, John Garcia and his colleagues demonstrated that when a novel flavor was used as the CS and illness or nausea was the UR, rats developed an intense aversion to (dislike of) the flavor (Garcia & Koelling, 1966). Taste-aversion learning involves the development of an aversion to a flavor that has been associated with illness (Batsell & Best, 1992, 1993). Classical conditioning of a tone and food occurs best when the tone is sounded one-half second before the food is presented, but strong taste aversions can be conditioned when illness occurs more than an hour after the taste is experienced (Garcia, Ervin, & Koelling, 1966).
There are two points of interest in these taste-aversion results. First, the flavor had to be novel for it to become associated with the illness. A flavor that has been consumed many times does not predict illness. Second, the time between the onset of the CS (taste) and the onset of the US (illness) can be quite lengthy, yet strong conditioning still occurs.
Preparedness.
Certain stimuli, such as flavors, can be associated with certain URs, such as illness, more easily than they can be associated with other URs, such as electric shock. According to Martin Seligman (1970), animals seem to be biologically ready or prepared to associate certain CSs with certain USs. Some events seem to go together naturally, whereas others do not. For humans and many animal species, taste and illness form one such natural pairing; presenting a tone or light CS with an electric shock forms another natural pair. If instead we try to pair the tone or light CS with illness or the taste CS with an electric shock, we usually get very weak conditioning (Garcia & Koelling, 1966). For this reason, humans, as well as many household pets, seem prepared to associate the sight of lightning (CS) with the sound of thunder. However, we do not seem prepared to associate loud noises with nausea or illness.