Many theorists now believe that classical conditioning often involves the formation of associations not between two stimuli but between internal mental representations of those stimuli (e.g., Bouton,
1994; Forsyth & Eifert, 1998; McDannald & Schoenbaum, 2009). Furthermore, the conditioned stimulus may enable an organism to predict (in a decidedly mental fashion) that the unconditioned
stimulus is coming (Hollis, 1997; Jara, Vila, & Maldonado, 2006; Mineka & Zinbarg, 2006;
Step 4: CS2 (tests) CR (anxiety) UCR (anxiety) Step 2: NS1 (school) UCS (traumatic event) (no response) Step 1: NS1 (school) Step 3: CS1 (school) NS2 (tests) CR (anxiety) Figure 3.3
Rescorla, 1988). As you can see, then, behaviorists are now beginning to talk about the thinking processes that they so deliberately steered clear of in earlier years.
Classical conditioning doesn’t always involve cognition, however. To be more precise, it
doesn’t necessarily involve conscious awareness (Baccus, Baldwin, & Packer, 2004; Campanella & Rovee-Collier, 2005; Papka, Ivry, & Woodruff-Pak, 1997). 4 For instance, when organisms
receive certain drugs (e.g., morphine or insulin), the drugs, as unconditioned stimuli, naturally lead to certain physiological responses (e.g., reduced pain sensitivity or hypoglycemia). Curiously, stimuli presented prior to these drugs—perhaps a light, a tone, or the environmental context more generally—begin to elicit an opposite response (e.g., increased pain sensitivity or hyperglycemia),
presumably to prepare for—in this case, to counteract—the pharmaceutical stimuli that will soon follow (e.g., Flaherty et al., 1980; S. Siegel, 1975, 1979).
Such physiological responses, which are almost certainly not consciously controlled by the
recipients of the drugs, are one likely explanation for people’s addictions to nicotine, alcohol, and street drugs. When habitual smokers and abusers of other substances return to environ- ments in which they’ve previously used an addictive substance, their bodies respond in counter- active ways that make the substance even more desirable or seemingly necessary. Meanwhile, they develop greater tolerance to their substance of choice and need increasing quantities to gain the same “high” or other sought-after physiological state (C. A. Conklin, 2006; McDonald & Siegel, 2004; S. Siegel, 2005; S. Siegel, Baptista, Kim, McDonald, & Weise-Kelly, 2000).
Changing Undesirable Conditioned Responses
Conditioned responses are often difficult to eliminate because they’re involuntary: People have little or no control over them. Yet some classically conditioned responses (e.g., some irrational fears) can seriously impair everyday functioning. Two possible strategies for reducing counter- productive conditioned responses are extinction and counterconditioning.
Extinguishing Undesirable Responses
One obvious way to eliminate a conditioned response is through the process of extinction. If the conditioned stimulus is presented in the absence of the unconditioned stimulus frequently enough, the conditioned response should disappear. Often this is exactly what happens.
Unfortunately, however, extinction is notoriously undependable as a means of eliminating conditioned responses: It simply doesn’t always work. There are several possible reasons why:
• The speed at which extinction occurs is unpredictable. If, during the conditioning process, the conditioned stimulus was sometimes presented in conjunction with the unconditioned stimulus but sometimes alone (i.e., the stimulus pairings were inconsistent), extinction is apt to be especially slow (Humphreys, 1939).
• People (and many other species as well) tend to avoid a stimulus they’ve learned to fear, thus reducing the chances that they might eventually encounter the conditioned stimulus in the absence of the unconditioned stimulus. (We’ll look more closely at avoidance learning in Chapter 4 .)
4 Some theorists have suggested that the amygdala plays a key role in “thoughtless” forms of classical conditioning (Byrnes, 2001; LeDoux, 2003). As you may recall from Chapter 2 , this structure within the limbic system is actively involved in connecting particular emotional reactions to particular stimuli.
• Through second-order conditioning, a wide variety of stimuli may have been associated with the original conditioned stimulus and thus have begun to elicit unproductive responses (e.g., extreme fear or anxiety). It can sometimes be quite difficult to extinguish all of these second-order conditioned responses—a problem that has been reported in people suffering from posttraumatic stress disorder (Wessa & Flor, 2007).
• Even when a response has been extinguished, it may reappear through spontaneous recovery. We can never be totally sure when a response will spontaneously recover and when it won’t. Spontaneous recovery is especially likely to occur if extinction has occurred in only one context; the conditioned response is apt to reappear in contexts in which extinction hasn’t taken place (Bouton, 1994). 5
Counterconditioning More Desirable Responses
At its best, extinction merely eliminates a conditioned response, whereas counterconditioning replaces the response with a new, more productive one and, accordingly, tends to be more effec- tive. Mary Cover Jones’s (1924) classic work with “Little Peter” provides a good example. Peter was a 2-year-old boy who had somehow acquired a fear of rabbits. To rid Peter of his fear, Jones placed him in a high chair and gave him some candy. As he ate, she brought a rabbit into the far side of the same room. Under different circumstances the rabbit might have elicited anxiety; however, the pleasure Peter felt as he ate the candy was a stronger response and essentially over- powered any anxiety he might have felt about the rabbit’s presence. Jones repeated the same procedure every day over a two-month period, each time putting Peter in a high chair with candy and bringing the rabbit slightly closer than she had on the preceding occasion, and Peter’s anxiety about rabbits eventually disappeared. More recently, researchers have used a similar pro- cedure in helping an 8-year-old boy lose his fear of electronically animated toys and holiday decorations (Ricciardi, Luiselli, & Camare, 2006).
In general, counterconditioning involves the following steps:
1. A new response that is incompatible with the existing conditioned response is cho- sen. Two responses are incompatible with each other when they cannot be performed at the same time. Because classically conditioned responses are often emotional in nature, an incompatible response is often some sort of opposite emotional reaction. For example, in the case of Little Peter, happiness was used as an incompatible response for fear. An alternative would be any response involving relaxation, because fear and anxiety involve bodily tension.
2. A stimulus that elicits the incompatible response must be identified; for example, candy elicited a “happy” response for Peter. If we want to help someone develop a happy response to a stimulus that has previously elicited displeasure, we need to find a stimulus that already elicits pleasure—perhaps a friend, a party, or a favorite food. If we want someone instead to acquire a relaxation response, we might ask that person to imagine lying in a cool, fragrant meadow or on a lounge chair by a swimming pool.
5 Findings of one recent research study indicate that when a phobia is the result of a traumatic event, extinction can be quite effective if the extinction procedure is conducted within six hours after the event. Apparently such a rapid follow-up extinction procedure affects the brain’s consolidation of the stimulus–response association (Schiller et al., 2010). (The role of consolidation in learning and memory is discussed in Chapters 2 and 9 .)
3. The stimulus that elicits the new response is presented to the individual, and the conditioned stimulus eliciting the undesirable conditioned response is gradually
introduced into the situation. In treating Peter’s fear of rabbits, Jones first gave Peter some candy; she then presented the rabbit at some distance from Peter, only gradually bringing it closer and closer in successive sessions. The trick in counterconditioning is to ensure that the stimulus eliciting the desirable response always has a stronger
effect than the stimulus eliciting the undesirable response; otherwise, the latter response might prevail.
A technique I recommend to many graduate students who dread their required statistics courses is essentially a self-administered version of counterconditioning. In particular, I suggest that math-phobic students find a math textbook that begins well below their own skill level—at the level of basic number facts, if necessary—so that the problems aren’t anxiety arousing. As they work through the text, the students begin to associate mathematics with success rather than failure. Programmed instruction (described in Chapter 5 ) is another technique that can be useful in reducing anxiety about a given subject matter, because it allows a student to progress through potentially difficult material in small, easy steps.
Counterconditioning provides a means for decreasing or eliminating many conditioned anx- iety responses. For instance, in systematic desensitization , people who are excessively anxious in the presence of certain stimuli are asked to relax while imagining themselves in increasingly stressful situations involving those stimuli; in doing so, they gradually replace anxiety with a relaxation response (Head & Gross, 2009; Wolpe, 1969; Wolpe & Plaud, 1997). Alternatively, people might virtually “experience” a series of stressful situations through the use of goggles and computer-generated images, all the while making a concerted effort to relax (P. Anderson, Rothbaum, & Hodges, 2003; Garcia-Palacios, Hoffman, Carlin, Furness, & Botella, 2002).
Systematic desensitization has been widely used as a means of treating such problems as test anxiety and fear of public speaking (Head & Gross, 2009; Hopf & Ayres, 1992; W. K. Silverman & Kearney, 1991). I should point out, however, that treating test anxiety alone, without remedi- ating possible academic sources of a student’s poor test performance as well, may reduce test anxiety without any concurrent improvement in test scores (Cassady, 2010b; Naveh-Benjamin, 1991; Tryon, 1980).