What You Can Change _. And What You Can't - Martin E. Seligman [41]
This is one drug effect that makes Pavlovians happy. Pavlovian theory tells us that agoraphobia starts when the CS of being in the marketplace, the agora, coincides with the UR of the panic attack. This conditions the agora to terror. When a patient is administered the combination of drug and extinction therapy, she ventures out and does not experience another panic attack, because she is drugged. She is now exposed quite effectively to the CS of the agora in the absence of the UR of panic, so Pavlovian extinction occurs.
Thus the case for phobias resulting from Pavlovian conditioning looks quite strong: Some innocent object conjoined to a terrifying trauma imbues that object with terror. As predicted, extinction therapies work quite well.
Phobias and evolution. There are too many loose ends, however: three, to be exact. Each of these makes phobias look more like the sauce béarnaise phenomenon and taste aversions than like ordinary Pavlovian conditioning.
First, ordinary Pavlovian conditioning is not selective. Any CS that happens to coincide with any trauma gets conditioned. But phobias are, in fact, highly selective:
A seven-year-old on a picnic sees a snake crawling through the grass. She is interested but not disturbed by it. An hour later she returns to the family car and has her hand smashed in the car door. She develops a lifelong phobia—not of car doors, but of snakes!12
There are in fact only about two dozen common phobic objects, most prominently open spaces, crowds, animals, insects, closed spaces, heights, illness, and storms. By and large these are all objects that were occasionally dangerous to our ancestors during the progress of evolution. Even the exceptions, like airplanes, are usually traceable to more primitive fear objects, like falling, suffocating, or being trapped.13
Arne Ohman, one of Sweden’s leading psychologists, decided to find out if phobias were more like taste aversions than like ordinary Pavlovian conditioning. He performed the Garcia experiment with human fear, giving student volunteers Pavlovian conditioning with electric shock as the unconditional stimulus (US). The conditional stimulus (CS) was either an evolutionarily prepared object, a picture of a spider, or an evolutionarily unprepared object, a picture of a house. A prepared phobic object is one that has actually been dangerous to humans over evolutionary time; an unprepared object is one that has not. The students were not afraid of the spider to begin with, but after just one pairing of the spider and shock, they broke into a sweat when the spider was shown. The picture of the house induced fear—and only mildly so—only after many pairings. Unlike ordinary conditioning, then, phobic conditioning in the laboratory, the sauce béarnaise phenomenon, and human phobias in nature are all selective.14
The second problem with the Pavlovian view is that conditioning requires short delays and explicit pairing of CS and US to work. Phobias don’t. There was an hour delay between the snake and the hand smashed in the car door; yet the phobia bridged this gap. Even more telling, while about 60 percent of phobics relate a story of explicit pairing of the object with a trauma to explain how it started, 40 percent do not. Rather, these phobics relate a much vaguer, more social story of the origin of their phobia: for example, they had heard that their best friend was bitten by a dog, and after that dogs always terrified them.15
The Swedes again forged the link. Ohman showed his students the prepared object, a slide of a scorpion, and they looked at it without any fear. No explicit trauma followed. No electric shock came on. Rather, one of the students, a confederate, jumped up and raced out of the room, shrieking “I can’t stand it!” After that, the subjects sweated when they saw the slide of the scorpion.
With another group, Ohman then repeated the scenario, but with an unprepared object, a