In the 1970s, researchers Paul Ekman, Wallace Friesen and Carroll Izard became interested in whether emotions differ across cultures, so they showed photographs of emotional expressions to people around the world to determine if a smile means the same thing in San Francisco as it does in Samoa. They found that everyone recognized an upturned mouth as the universal sign of happiness, and there was similar agreement about expressions of surprise, anger, disgust, sadness and fear. This impressive degree of accord among diverse cultures suggests that the basic emotions are automatic and preprogrammed, but the task of determining the underlying neural circuitry of emotions has been difficult. Fear has been a particularly attractive candidate for study because it has easily-measured physical correlates such as increased heart rate and release of stress hormones. From minor apprehension to stark terror, fear helps us make associations that keep us from harm. Fear learning is quick, powerful and long lasting. If you think back to your childhood, chances are good that within your earliest memory is an event colored by fear.
Interestingly, it appears that of all the emotions, the brain devotes the most space and energy to fear. Charles Darwin was one of the first scientists to suggest that fear has a biological basis, when he noted that nearly all animals exhibit fear in the same manner. From birds and rats to apes and humans, animals in peril display a stereotyped behavior pattern that includes freezing in place, increased respiration and heart rate, release of stress hormones, and increased tendency to startle. Because fear responses are so well conserved across species, it is possible to learn a lot about human fear from animal studies. Most of the research has focused on fear conditioning, which explores how an animal learns to fear specific stimuli within its environment. Researchers have been quite successful in pinpointing the specific brain areas that govern fear responses and fear learning.
Fear conditioning is a form of classical conditioning, the type of associative learning pioneered by Ivan Pavlov in the 1920s. It involves the repeated pairing of a non-threatening stimulus such as a light, called the conditioned stimulus, with a noxious stimulus such as a mild shock, called the unconditioned stimulus, until the animal shows a fear response not just to the shock but to the light alone, called a conditioned response. The most famous example of human fear conditioning is the case of Little Albert, an 11 month old infant used in John Watson and Rosalie Rayner’s 1920 study. Like most babies, Albert had a natural fear of extremely loud noises but no aversion to white rats. So Watson and Rayner presented him with a white rat, and when he reached to touch it, they struck a hammer against a steel bar just behind his head. After seven repetitions of seeing the rat and hearing the frightening noise, Albert burst into tears at the mere sight of the rat. In addition, Albert showed some generalization of his learned fear response–he would cry at the sight of objects that resembled the white rat, such as a white dog or a white coat. However, he also showed a lot of discrimination; he was not fearful of toys or objects that were dissimilar to the offending rat.
Thankfully, these days rats are more often the subjects of fear conditioning studies. They are taught to fear a tone or a light via repeated pairings with a moderate foot shook. Over the past two decades, researchers like Joseph LeDoux, Michael Fanselow, Jeansok Kim and Michael Davis have taken great advantage of this fast and reliable paradigm to elucidate the neural circuitry underlying fear learning.
To be a candidate for mediating fear learning, a brain area must satisfy two basic requirements. First, it must get input from the sensory systems that are processing the potentially dangerous stimuli, such as the visual system for a light or the tactile system for a shock. Second, it must project to the brain regions that are known to control fear responses, like the lateral hypothalamus and paraventricular nucleus. Using these criteria, researchers identified the amygdala as a critical region for fear conditioning.
The amygdala is a small, almond-shaped cluster of nuclei set deep in the temporal lobe that seems ideally positioned as the locus of fear learning. It receives input through its lateral nucleus from cortical areas and the thalamus, which is a key sensory relay station within the brain, and it sends output via its central nucleus to a variety of brain regions that are known to mediate fear responses, such as the hypothalamus. In fact, electrical stimulation of the amygdala can cause a previously calm animal to exhibit fearful and/or aggressive behaviors, and humans are not immune to this sort of manipulation. Autopsies of Charles Whitman, the formerly genteel man who carried out a sniper attack from the University Tower at Texas in 1966, showed he had a tumor pressing on his amygdala.
Recent research has further supported a crucial role for the amygdala in fear conditioning. If a rat has its amygdala destroyed, it will still show a fear response to the foot shock but fail to learn the association between the light and the foot shock. Even after many training sessions, the animal will not exhibit any fear to the light alone. Humans with amygdalar damage seem to have a similar problem. When Elizabeth Phelps and Joe LeDoux examined (mild) fear conditioning in people with localized damage to the amygdala, they found that their subjects could verbalize the connection between the light and the shock–“A light comes on, followed by a shock”–but they failed to show the usual conditioned fear responses, such as increased heart rate, when the light was presented alone. Interestingly, LeDoux has also shown that fear conditioning in rats induces long-term changes in the patterns of communication between neurons in the amygdala. The cellular response to the sound of the CS tone increases after pairing with the shock. There is no increased neurological response to other tones that were not associated with shock, and enhancement to the CS tone does not occur if there is no training. Collectively, these results suggest that the amygdala is a key structure for fear learning.
Context Conditioning and the Hippocampus
Scientists studying fear conditioning noticed a peculiar thing about animals receiving training. After a few days of tone and shock pairings, they began acting fearful the moment they were put in the conditioning chamber, even before the next phase of training had begun. Researchers wondered if the animals might have learned to associate both the specific tone and the general context of the chamber with the aversive foot shock. General rules of associative learning would suggest that this phenomenon is likely. An organism is always calculating how likely it is that event X will follow Y. If X happens every time Y occurs, then Y is an extremely good predictor of X. Alternatively, if X follows Y sometimes and other times appears after a different event (Z), Y is only half as good at predicting X. The rules work in the other direction as well. If cues A, B, and C are always present when X occurs, then A, B, and C are all equally good predictors of X. When a light is repeatedly paired with shock, the rat quickly learns the connection between the two stimuli, but he also learns to connect any other cues present in his environment at the time of the foot shock with the unpleasant experience. This is called context conditioning.
The fear response to the conditioned context is not as great as the response to the specific tone, and the learning rules help explain why. The context of the training (usually a box with a metal grid on the bottom) is certainly present at the time the animal receives the foot shock, so it is not surprising that the rat forms a connection between the two. However, unlike the light that directly precedes the shock, the chamber is also present before and after the shock. Thus, while the context is predictive of the shock, it is not the best predictor. This explains why fear-conditioned animals show no fear in their home cage (where they have never been shocked), some fear in the testing chamber alone, and great fear when the tone occurs. Interestingly, contextual fear conditioning seems to depend more on the hippocampus than the amygdala. Research from other experiments has suggested that the hippocampus is especially important for remembering locations, so scientists examined animals with hippocampal lesions to see if they displayed normal fear conditioning. Results showed that while the animals grasped the connection between the light and the shock, they failed to demonstrate any fear of the conditioning chamber alone. These studies indicate that normal fear conditioning requires both the amygdala and the hippocampus.
Real World Applications of Fear Conditioning
Much of our natural fear conditioning is to our benefit. We touch the hot stove once and learn not to try it again. However, sometimes fear becomes unmanageable, as in cases of agoraphobia, whose sufferers are afraid to venture out of their houses, or in post-traumatic stress disorder, which causes some rape victims and war veterans to experience terrible flashbacks. Because fear responses are so similar across species, it may be possible to glean insight from animal studies that would pave the way to better treatment of pathological human fear. Already some of the learning rules examined in fear conditioning are exploited to treat phobias. Systematic desensitization, a treatment that uses repeated presentations of the fearful stimulus (such as a snake) to show the person that they are actually in no real danger, takes advantage of the dwindling association between danger and the snake. After multiple experiences with the snake where nothing bad happens, the fearful person relaxes because the predictive value of the snake no longer seems valid. These therapies demonstrate the importance of understanding the basic principles of learning and behavior, and shed light on the complex interplay between emotion and reason that characterizes human life.