What is taste aversion?

Introduction

When an animal eats a food, especially one with which it has had little experience, and then becomes ill, the food acquires a nauseating or aversive quality and will subsequently be avoided. This phenomenon is called bait shyness, food aversion, or, most commonly, taste aversion, although the odor and sometimes even the sight of the food also become aversive.

When confronted with a new food, rats will investigate it thoroughly by sniffing. If the odor is unfamiliar, and familiar food is available elsewhere, the rats may pass by the novel food without eating it. If sufficiently hungry, the rats may sample the new food by nibbling at it and then withdrawing to wait for adverse effects. If none occurs, the new food may be accepted, but if the rats become ill the food will be avoided; even the trails or runways where the new food is located may be abandoned. This cautiousness toward novel foods makes rats notoriously difficult to poison.

Stimulus and Response

Taste-aversion learning has been construed both as instrumental avoidance learning and as Pavlovian conditioning. Clearly, elements of both are involved. Development of the aversion itself, in which the food takes on a negative motivational quality, is seen by most learning theorists as Pavlovian conditioning. Subsequent avoidance of the food is learned by instrumental conditioning.

In development of the aversion, the smell or taste of the food clearly serves as the conditioned stimulus; however, there has been some confusion in the literature as to what constitutes the unconditioned stimulus
in taste-aversion conditioning. In a typical experiment, rats are presented with water that contains a distinctive flavor, such as saccharin or almond extract. After drinking the flavored water, the rats are treated in some way that makes them ill. Illness treatments have been as diverse as X-ray irradiation and spinning on a turntable, but the preferred method is injection of a toxic drug such as lithium chloride or apomorphine. As a result of the treatment-produced illness, the rats subsequently avoid the flavored water.

Most frequently, “illness” is cited as the unconditioned stimulus in these experiments, but one also sees references to “poisoning” or to the “illness treatment” in this regard. These latter references actually make more sense and are more consistent with other research in which a drug treatment (the unconditioned stimulus) is seen as producing an innate drug effect (the unconditioned response).

In 1927, Ivan Petrovich Pavlov described experiments with morphine and apomorphine in which the drug injection, the drug itself, or “changes in the internal environment due to alteration in the composition of the blood” were construed as the unconditioned stimulus. The drug effects, including salivation, nausea, vomiting, and sleep, were construed as unconditioned responses. Pavlov even described an experiment in which tying off the portal vein led to development of an aversion to meat in dogs because of buildup in the blood of toxic substances derived from the digestion of the meat. The implication was that the smell and taste of meat were conditioned stimuli and the toxins (or the alterations in blood chemistry) were unconditioned stimuli.

In taste-aversion conditioning, the smell or taste (and sometimes the sight) of food serves as the conditioned stimulus. This stimulus signals the presence of a toxin, which acts as the unconditioned stimulus by altering body chemistry, which in turn produces nausea, illness, or vomiting, the unconditioned response. Through conditioning, nausea or “aversion” develops as the conditioned response to presentation of the taste, smell, or sight of the food. This aversion then motivates an instrumental avoidance response; that is, because of the conditioned aversion, the animal does not eat the food.

Learning Aversions

Taste aversion plays an important adaptive role in the everyday life of animals, especially those that eat a diversity of foods. Food preferences are learned early in the lives of such animals—they eat what they see their mothers eating or, even earlier, they come to prefer foods with flavors encountered previously in mothers’ milk. To cope with a variable environment, however, animals must often adopt a new food. Animals with no mechanism for learning to accept safe foods while rejecting toxic ones would soon perish.

Nor is taste-aversion learning seen only in laboratory animals. Humans, too, learn food aversions quickly and convincingly. Martin E. P. Seligman, a prominent learning theorist, has supplied his own autobiographical account of taste aversion learning. Six hours after eating filet mignon flavored with béarnaise sauce, Seligman became violently ill with the stomach flu. “The next time I had sauce béarnaise, I couldn’t bear the taste of it,” he relates. He did not, however, develop an aversion to the steak, to the white plates from which it had been eaten, or to the opera that he attended during the six-hour interstimulus interval.

Seligman’s experience exemplifies several peculiarities of taste-aversion learning that have made it an important topic in the literature of learning theory: A strong conditioned response develops in a single learning trial, the conditioned response develops even when the conditioned and unconditioned stimuli are separated by long interstimulus intervals, the aversion develops selectively to some stimuli but not to others, and the conditioned response is irrational in the sense that it is not much affected by conscious knowledge that the food was not tainted or is not likely to be tainted in the future.

Natural Aversions

In nature, taste-aversion learning is a common event. Animals that do not specialize on one or a few foods must be able to reject toxic foods. Rats especially have a problem in this regard, since they do not vomit and therefore cannot expel poisons once they have ingested them. When rats have access to many foods, their behavior is marvelously adapted to detecting toxins. They eat only one or two different food types at a time and may eat these exclusively for days. Then they shift to concentration on another food type. If illness develops, the rats know immediately which type of food is probably to blame and subsequently avoid it. If the rats had eaten a variety of foods all the time, such discrimination would not be possible.

Human infants may adopt a similar strategy when allowed to eat without supervision. In the 1920s, Clara Davis gave infants the opportunity to eat any of a variety of nutritious foods, none of which alone supplied a balanced diet. The infants specialized on one or two foods for days at a time before shifting to another food. Although daily diets were certainly not nutritionally balanced, the infants did, over the long run, eat a balanced, healthful diet. The behavior of one infant was particularly interesting. This child voluntarily consumed cod-liver oil, a vile-tasting fluid usually rejected by children. This child, however, had a vitamin D deficiency, and the cod-liver oil supplied the necessary vitamin. After the deficiency was eliminated, the infant stopped eating cod-liver oil and never went back to it.

The idea that the infant’s behavior may be related to taste learning was shown by Paul Rozin, who found that rats fed a thiamine-deficient diet subsequently chose a food laced with thiamine supplements even though thiamine itself is tasteless. The rats apparently were able to use the taste of the food as a discriminative stimulus for its nutritive properties. Thus, the phenomenon is the opposite of taste-aversion learning—the development of specific hungers for foods with nutritive qualities, foods that promote health or recovery from illness. Anecdotal reports suggest that humans sometimes also suddenly develop tastes for foods that contain needed nutrients.

Taste aversion is apparently only one side of the story of food selection and rejection in nature. Both appetitive and avoidance behaviors can be predicated on taste cues. In some cases, these behaviors are innate responses to the taste. Bitter tastes usually indicate the presence of toxic alkaloids and are often rejected by young animals that have had no prior experience with them. Human infants do the same. In other cases, the response to taste cues must be learned. Thus, specific hungers and taste aversions both represent examples of appropriate behaviors that are cued by discriminative taste stimuli.

Lincoln Brower has described a classic example of taste-aversion learning in nature. Blue jays, he noted, typically avoid preying on monarch butterflies. If hungry enough, however, jays will take and eat monarchs. The caterpillars of these butterflies eat milkweed, which contains a poison to which the butterflies are immune but birds are not. Enough of the poison remains concentrated in the tissues of adult monarchs to make a bird that eats one quite sick. The jays subsequently reject monarchs after a brief taste, and eventually the distinctive orange and black insects are rejected on sight.

Induced Aversions

In more applied settings, Carl Gustavson and John Garcia have described the use of taste-aversion conditioning in wildlife management. On the western ranges where large flocks of sheep are left relatively unprotected, ranchers often face the threat of predation by coyotes, wolves, and mountain lions. One response has been wholesale shooting and poisoning of these wild predators, but this is a less-than-ideal solution. Gustavson and Garcia found that predators, such as coyotes, that scavenge a lamb carcass laced with a sublethal dose of lithium chloride will subsequently develop a strong aversion to lamb and may even avoid areas where lamb and sheep are grazing. The authors proposed a scheme for reducing predation on sheep using taste-aversion conditioning that would drastically reduce the need for shooting and the use of indiscriminate lethal poisons.

In humans, many medical conditions are accompanied by loss of appetite and weight loss. Although this is often attributable to chemical changes within the body, it can also be caused by taste-aversion learning. Ilene Bernstein investigated the loss of appetite, or anorexia, that frequently accompanies cancer chemotherapy and found that, in all likelihood, it was attributable to aversive conditioning caused by the cancer medications, which often induce nausea and vomiting. Bernstein and her colleague, Soo Borson, investigated other anorectic syndromes and found the same possibility. In an important review article published in 1999, Bernstein proposed that taste-aversion learning may play a significant role in such conditions as cancer anorexia, tumor anorexia, anorexia nervosa, and the anorexias that accompany clinical depression and intestinal surgery.

On the other hand, taste-aversion learning is intentionally induced in some types of aversion therapy for maladaptive behaviors. Alcoholics are sometimes given a drug called disulfiram (Antabuse) that interferes with alcohol metabolism in the liver. Drinking alcohol after taking this drug results in a very unpleasant illness that conditions an aversion to alcohol. Subsequently, the taste, smell, or even the thought of alcohol can induce nausea. Cigarette smoking has been treated similarly.

Experimental Aversions

Before 1966, psychologists believed that learning obeyed the law of equipotentiality. In Pavlovian conditioning, the nature of conditioned and unconditioned stimuli was seen as unimportant—if they were paired appropriately, learning would occur with equal facility for any stimulus pair. In instrumental conditioning, psychologists believed that any reinforcer would reinforce any behavior.

Equipotentiality had been challenged. Ethologists insisted that each species of animal is unique in what it learns, that learning is an evolutionary adaptation, and that species are not interchangeable in learning studies. Nikolaas Tinbergen, in The Study of Instinct (1951), wrote of the innate disposition to learn. Keller and Marian Breland, who trained animals for commercial purposes, discovered that animals drifted toward species-specific food-related behaviors when their arbitrary instrumental responses were reinforced with food.

In 1966, Garcia, Robert Koelling, and Frank Ervin published their research on taste-aversion learning. In an article called “Relation of Cue to Consequence in Avoidance Learning,” they described an experiment in which rats received aversive consequences for licking water from a drinking spout. In the “tasty water” condition of the experiment, the water was flavored with saccharin, while in the “bright-noisy water” condition, licking the spout activated a flashing lamp and a clicking relay. Half the animals from each condition were made sick after drinking. The other half received a mild but disruptive electric shock after licking the spout. In the tasty water condition, animals that were made sick, but not those that were shocked, avoided drinking. In the bright-noisy water condition, animals that were shocked, but not those that were made sick, avoided drinking. Thus, light and noise were easily associated with shock, and taste was easily associated with illness, but the contrary associations were much more difficult to establish.

In a second article, called “Learning with Prolonged Delay of Reinforcement,” Garcia, Ervin, and Koelling demonstrated that taste aversion developed even when the taste and illness treatment were separated by seventy-five minutes. Learning with such prolonged delays had been regarded as impossible, and it could not be reproduced in shock-avoidance experiments. These results were quickly replicated in other laboratories. Similar effects were demonstrated in other types of learning experiments, including traditional avoidance paradigms and even in mazes and Skinner boxes. The fact that something was wrong with traditional learning theory and equipotentiality was soon evident.

The doctrine of prepared learning replaced equipotentiality. Preparedness is the idea that evolution equips animals to learn things that are important to their survival. Examples of prepared learning already existed in the literature, but until 1966 their significance was not widely recognized among psychologists. Ethologists, however, pointed to studies of imprinting, food recognition, song learning, and place learning in a variety of animals, all illustrating prepared learning. Psychologists quickly included language learning and the learning of some phobias under the umbrella of preparedness. It was even proposed that human cognition evolved to cope with widely divergent situations that require unprepared learning. Taste-aversion learning, however, which is strongly prepared apparently even in humans, seems relatively immune to such ratiocination.

Bibliography

Bernstein, Ilene L., and Soo Borson. “Learning Food Aversion: A Component of Anorexia Syndromes.” Psychological Review 93, no. 4 (1986): 462–72. Print.

Bernstein, Ilene L. "Taste Aversion Learning: A Contemporary Perspective." Nutrition 15.3 (1999): 229–34. Print.

Bolles, Robert C. Learning Theory. 2d ed. New York: Holt, Rinehart and Winston, 1979. Print.

Braveman, Norman S., and Paul Bronstein, eds. Experimental Assessments and Clinical Applications of Conditioned Food Aversions. New York: New York Academy of Sciences, 1985. Print.

Brower, Lincoln Pierson. “Ecological Chemistry.” Scientific American 220 (February, 1969): 22–29. Print.

Bures, Jan, Federico Bermudez-Rattoni, and Takashi Yamamoto. Conditioned Taste Aversion: Learning of a Special Kind. New York: Oxford University Press, 1998. Print.

Gustavson, Carl R., and John Garcia. “Pulling a Gag on the Wily Coyote.” Psychology Today 8 (August, 1974): 68–72. Print.

Holstein, Sarah E., et al. "Adolescent C57BL/6J Mice Show Elevated Alcohol Intake but Reduced Taste Aversion as Compared to Adult Mice: A Potential Behavioral Mechanism for Binge Drinking." Alcoholism: Clinical and Experimental Research 35.10 (2011): 1842–51. Print.

Pierce, W. David, and Carl D. Cheney. Behavior Analysis and Learning. 5th ed. New York: Taylor & Francis, 2013. Print.

Reilly, Steve, and Todd R. Schachtman, eds. Conditioned Taste Aversion: Neural and Behavioral Processes. New York: Oxford University Press, 2009. Print.

Seligman, Martin E. P., and Joanne L. Hager, eds. Biological Boundaries of Learning. New York: Appleton-Century-Crofts, 1973. Print.

Verendeev, Andrey. "Conditioned Taste Aversion and Drugs of Abuse: History and Interpretation." Neuroscience & Biobehavioral Reviews 36.10 (2012): 2193–205. Print.