Classical conditioning usually depends on repeated contiguous pairings of the CS with the US. There exist, however, several examples involving classical conditioning that appear to violate these basic requirements: pseudoconditioning, one-trial learning, taste aversion, and imprinting. Another classical conditioning phenomenon not fitting neatly into the Pavlovian paradigm is what Solomon and Corbit (1974) has termed opponent processing, a special case of hedonic conditioning having pronounced (theoretical) effects on emotional learning and reactivity.
Pseudoconditioning is usually observed in classical conditioning situations, especially as a confounding influence that must be experimentally controlled against. As just discussed, the normal relationship between the CS and the US in Pavlov's paradigm depends on the forward and contiguous presentation of the CS followed by a US—a CS-US arrangement referred to as pairing. After repeated pairings (although sometimes one is enough), a conditional association between the two stimuli is established, so that the previously neutral stimulus (now a CS) is capable of eliciting a CR resembling the UR that had been originally elicited only by the US. In pseudocon-ditioning, the stimulus (neutral) elicits a response resembling a UR, even though it has never been paired with the US.
For example, if a dog receives an intense shock delivered by an electronic collar and then a few hours later a buzzer sound is delivered by the same collar, the dog may react to the sound as though it had been actually shocked rather than just buzzed. This behavioral change occurs even though the buzzer was never actually paired with shock in the past. The buzzer apparently acquires CS-like properties through sensitization and other association effects that do not strictly belong to the classical conditioning paradigm. in fact, any strong surprising or startling event may cause pseudoconditioning, that is, evoke responses to neutral stimuli that have never been paired with the eliciting US. Another important factor in the foregoing example is generalization—that is, the vibrating buzzer may seem similar in some particulars to shock, thus facilitating a connection between the two stimulus events. However, the similarities between the shock and buzzer sound are not the only factors involved. Although generalization is often present in pseudocon-ditioning, an even more important consideration is context. In the present instance, the buzzer occurs in close association with the source of electrical stimulation, with both stimuli being produced by a collar fastened around the dog's neck.
As the result of a particularly aversive event, the context in which the incident occurs may itself become conditionally linked to the aversive experience. When under the influence of a similar situation in the future, the dog may be more vigilant and alert for danger, exhibiting a much lower threshold for startle or escape behavior. Now an otherwise innocuous event may elicit a strong fear response (as occurs, by the way, in post-traumatic stress disorder and is commonly observed in abused dogs). For example, if a dog is attacked by another dog while out on a walk, on future occasions while walking in that same general vicinity at about the same time of day the dog may appear to be more cautious and defensive about its surroundings. Sounds and movements that might be ignored in other places now take on a new significance. A passing car with a loud muffler or a pile of leaves shifting abruptly in the wind might evoke a strong startle or panic reaction, even though there is no actual threat (past or present) associated with such occurrences themselves. For trainers working with emotionally motivated behavior problems (e.g., aggression, fears, or separation distress) behavioral effects associated with pseudocon-ditioning should be carefully assessed and taken into consideration when developing behavior modification programs.
Many chronic phobias can be traced to a single event. There appear to be hardwired neu-robehavioral mechanisms designed to facilitate rapid learning of information derived from particularly dangerous experiences or startling stimulus events. Under natural conditions, life-threatening or potentially injurious situations may not offer an animal the luxury of repeated exposures or close encounters in order for it to learn that the stimulus in question predicts danger. Consequently, some avoidance patterns appear to be innately programmed or prepared in advance so that they are easily learned with minimal exposure to the threatening situation (Seligman, 1971). Such preparedness is a natural safeguard, al lowing the animal to identify especially dangerous associations quickly and efficiently without depending on repeated exposure.
One-trial learning frequently results when a strongly startling or threatening US is paired with a novel CS. The operative word here is novel. Positive or neutral past experiences (latent learning) with the CS may interfere with one-trial learning. This interference effect stems from competitive safe expectancies that must first be disconfirmed before new learning can take place. Many arrangements provide sufficient conditions for one-trial learning, but it is optimally evoked in situations where the environment itself produces the desired effect. For example, a puppy that has developed the dangerous habit of chewing on electrical cords can be discouraged by preparing electrical cords so that an intense startle or aversive event occurs whenever they are disturbed. A common method employed for this purpose is to booby trap the forbidden item so that an intense startle is produced if the cord is disturbed. The resulting effect provides a lasting aversive association and avoidance of electrical cords.
Taste aversion is another example of associative learning that does not fit neatly into the classical conditioning paradigm. A lasting taste aversion often results when an animal ingests a food item or flavor that is followed by a nausea-producing illness. As previously discussed in Chapter 5, Garcia and colleagues (1966) performed a series of experiments in which rats were presented with a compound stimulus involving flashing lights and noise while drinking saccharine-flavored water. While drinking the flavored water, the rats were simultaneously exposed to radiation. Such exposure to radiation causes nausea within an hour or so. Subsequent testing revealed that the exposed rats had developed an intense aversion toward the taste of saccharine but not toward the auditory and visual conditioned stimuli employed. A curious feature of taste aversion is that the effect can be produced even if the inducement of nausea is delayed for several hours. Also, taste aversions can be reliably established after only a single trial. The conditions under which taste aversions are established are inconsistent with the requirements normally present during classical conditioning, that is, repeated contiguous pairings of the CS and US. There appear to exist special learning sensitivities connected with taste and nausea, aiding some animals in differentiating safe from poisonous food items. Seligman (1970) has postulated an internal preparedness facilitating the learning of such associations. Taste-aversion techniques have been used effectively to discourage predation on sheep by coyotes (Gustavson et al., 1974; Garcia et al., 1977). It makes biological sense that a foraging animal would evolve strongly prepared sensitivities for the development of taste aversions. As in other examples of one-trial learning, the food item being conditioned must be novel, that is, lack a history of safe ingestions. Food safely ingested in the past may require nausea-producing exposures before it is avoided.
A taste aversion procedure may be useful for controlling refractory coprophagia. Eating feces is a common canine vice, and, though not usually harmful to the offending dog's health, it is aesthetically objectionable to many owners. In cases where everything else fails, a taste-aversion arrangement might prove helpful (Houpt, 1991), although such methods have not been demonstrated consistently effective in dogs (Hart and Hart, 1985). The procedure is simple: as soon as the dog ingests feces, the owner is instructed to induce vomiting with a chemical emetic. Sometimes the feces itself is contaminated with the emetic. Gustavson (1996) has noted, however, that emesis per se is not sufficient to establish taste aversion, suggesting that common emetics such as ipecac are inappropriate for establishing such learning. The critical factor involved here is that the chemical being used is capable of eliciting nausea, that is, producing sensations of physical illness. One of the most common compounds employed to achieve this end is lithium chloride. Gus-tavson and coworkers have recommended the use of taste-aversion procedures for controlling a variety of appetitive vices and excesses. In spite of the potential benefits of taste aversion, such treatment is not without potential risks, side effects, and discomfort for dogs. Consequently, the method should be performed only under close veterinary supervision and reserved as a last-resort treatment for serious and refractory appetitive behavior problems.
Konrad Lorenz, credited with the discovery of imprinting in birds, early expressed the opinion that imprinting was not a learning phenomenon but an instinctive process of attachment to a social object: "This process cannot be equated with learning—it is the acquisition of the object of instinctive behavior patterns oriented toward conspecifics" (1970:124). Subsequent study, however, has shown that a great many classical conditioning factors do play a role in the imprinting process (Sluckin, 1965). Imprinting may be interpreted as a variant form of classical learning in which several behavior patterns, attachments, and preferences are facilitated through brief exposures early in life. Imprinting is unique in that it takes place most efficiently (if at all) during narrowly defined sensitive periods occurring early in the animal's life. If this period of sensitivity passes without the occurrence of appropriate stimulation, then irreversible adjustment problems may develop (Scott and Fuller, 1965).
Imprinting is distinguished from most forms of classical conditioning along several different dimensions: speed of acquisition, permanence of associations, resistance to future learning effects, resistance to decay by disuse, and the reliance on sensitive periods early in life. Another significant difference between imprinting and other forms of classical conditioning is that imprinting often involves complex behavior patterns, whereas most other stimulus learning involves more simple and discrete units of behavior. Studies designed to determine whether imprinting and social attachment involve instrumental components have demonstrated that positive reinforcement (food rewards) does not play a significant role (Brodbeck, 1954; Scott, 1962). In fact, the effects of imprinting appear to be enhanced by conditions adverse to instrumental learning. Animals appear to be even more strongly attracted to the imprinted object when they are forced to endure obstacles and aversive stimulation during the imprinting process (Hess, 1964).
Was this article helpful?
Getting a new puppy is a fun and interesting time. You probably went to a breeder or pet store or maybe just saw an ad on the Internet or the newspaper, for puppies, and decided just to check it out. Before you knew it those little eyes and fluffy puppy fur had your heart melting and you were headed home with him or her in your arms.