New Page 1

	II. Theory and Terminology
	Learning Theory The main focus of this web site is program development and application. The following review of terminology and theory is just a brief sampling of what is in the literature to help you to understand some very basic information. This information is not meant to stand alone. We hope that this will stimulate you to go off and read more! This information will hopefully be a supplement on learning theory to what you can check out and read at your local library. As mentioned earlier, training can be defined as a type of learning. Learning is broadly defined as a change in behavior resulting from practice or experience; when that practice is dictated by humans, the process is called training [Mellen and Ellis, 1996]. Given this very broad definition, virtually every direct and indirect interaction that human caretakers have with their animals can be viewed as a form of training. Four types of learning have been described for animals: habituation, classical conditioning, operant (instrumental) conditioning, and complex learning.
	Habituation is the declining or waning of a behavior as the result of repeatedly presenting a stimulus. An example of this is that an animal initially is startled by a loud noise, but if that loud noise is repeated many times, the animal's reaction to the loud noise becomes imperceptible; in this example, the stimulus (loud noise) went from being aversive to the animal to negligible.	Classical conditioning is the process through which a stimulus that formerly had no effect on a particular reflex (e.g., blinking or salivating) acquires the power to elicit that reflex.
	Operant or instrumental conditioning shapes or modifies the behavior of the animal. (The animal "operates" on the environment that leads to a reward; the animal’s behavior is "instrumental" in acquiring a reward.) In operant conditioning, the animal receives a reward or avoids an aversive stimulus by exhibiting a particular behavior. For example, the animal enters a holding area, and the animal receives a food reward; after making this association, the animal is more likely to enter the holding area at its next opportunity.	Complex learning is basically a "catch-all" category. Scientists initially thought that all learning could be categorized as habituation, classical or operant conditioning. However, these three types of learning did not explain observational learning (one animal learns how to perform a particular behavior simply by watching another animal) or latent learning where rats ran through a maze faster if they had been allowed first to explore the maze.
	The concepts of reinforcement and punishment are integral to an understanding of learning and training theories. Reinforcement (both positive and negative) refers to an event that occurs in conjunction with a behavior and increases the likelihood that a behavior will occur again. A positive reinforcer is a stimulus that is sought and is something that the animal presumably finds "pleasant." A negative reinforcer is a stimulus that is avoided and that the animal finds "unpleasant." If an animal comes into a holding area (the behavior requested) and is provided with a desired food item, we are using positive reinforcement to train the animal to enter the holding area. If an animal comes into a holding area because its outdoor area is being hosed (the animal is avoiding the water spray), we are using negative reinforcement. Both methods increase the rate of responding (entering the holding area). Punishment differs from reinforcement in two ways: first, punishment occurs after the undesirable behavior, and second, punishment decreases/suppresses the frequency of the undesired behavior. In our example of asking an animal to come into a holding area, we may want the animal to always come into stall A and never into stall B. With punishment, we might squirt an animal with water if it entered the wrong stall. The animal has already "made the mistake," i.e., entered the wrong stall; presumably by then squirting the animal with water, it learns not to enter the wrong stall. It is common to hear people say that they only use "positive reinforcement" techniques when they train. However, it’s unlikely that this training technique is the only one used. Negative reinforcement and punishment, when used properly, also are effective tools. For example, keepers often "walk" an animal into the barn, i.e., uses the animal’s flight distance to encourage the animal to move away from the keeper (and into the barn). This is an example of using negative reinforcement. The animal is moving away from a mildly aversive stimulus (keeper entering within the animal’s flight space) and increasing the rate of responding (entering the barn).
	Similarly, keepers often use a "time out" (e.g., keeper walks away) when an animal shows aggression during a training session. This is an example of punishment. The animal has directed aggression at the keeper (undesirable behavior); after the fact (the "mis-deed" has occurred), the keeper responds by walking away. Presumably, this time out decreases the occurrence of keeper-directed aggression during a training session. When using positive reinforcement, negative reinforcement, or punishment, it is important to track the progress you are making. It is possible that you could be using the technique in an ineffective manner. For example, if a keeper is using a time out (punishment) to decrease aggression, and s/he finds the need to use time outs frequently over an extended period of time, the use of time outs should be re-evaluated. If the animal is demonstrating the same level of aggression (when you are using time outs), there is a good chance that the time outs are not being used in a manner that is effective. New techniques to decrease the aggressive behavior or the approach of training incompatible behaviors may need to be investigated.
	back home next