Types of learning

Learning refers to a change in behavior as a result of experience. Learning is studied in a variety of species in an attempt to uncover basic principles. There are two major types of learning: classical (Pavlovian) conditioning and operant (instrumental) conditioning. Exposure to uncontrollable aversive events can have detrimental effects on learning. Consequences can be successfully used to develop a variety of behaviors, including even random, unpredictable performance. Learning produces lasting changes in the nervous system.

TYPE OF PSYCHOLOGY: Biological bases of behavior; learning; motivation

Introduction

Learning has been of central interest to psychologists since the beginning of the field in the late 1800s. Learning refers to changes in behavior that result from experiences. The term “behavior” includes all actions of an organism, both those that are directly observable, such as typing at a keyboard, and those that are unobservable, such as thinking about how to solve a problem. Psychologists who study learning work with a variety of species, including humans, rodents, and birds. Nonhuman species are studied for a variety of reasons. First, scientists are interested in fundamental principles of learning that have cross-species generality. Second, the degree of experimental control that can be obtained with nonhumans is much higher than with humans. These controlled conditions make it more likely that any effect that is found is due to the experimental manipulations rather than to some uncontrolled variable. Third, studying the learning of nonhumans can be helpful to animals. For example, a scientist might need to know the best way to raise an endangered giant condor so it is more likely to survive when introduced to the wild.

There are two major types of learning. Classical conditioning (also called Pavlovian conditioning, after Russian physiologist Ivan Petrovich Pavlov) involves the transfer of control of reflexes to new environmental stimuli. For example, when a person gets a glaucoma test at an optometrist’s office, a puff of air is delivered into the eyes, which elicits blinking. After this experience, putting one’s head into the machine elicits blinking. The glaucoma-testing machine now elicits the reflex of blinking before the air puff is delivered.

Operant conditioning, also called instrumental conditioning, involves the regulation of nonreflexive behavior by its consequences. American psychologist Edward L. Thorndike was a pioneer in the study of operant conditioning, publishing his work about cats escaping from puzzle boxes in 1898. Thorndike observed that over successive trials, movements that released a latch, allowing the animal to get out of the box and get some food, became more frequent. Movements not resulting in escape became less frequent. Thorndike called this the : responses followed by satisfaction would be strengthened, while responses followed by discomfort would be weakened. The study of operant conditioning was greatly extended by American behaviorist B. F. Skinner, starting in the 1930s.

In the 1960s, American psychologists Martin E. P. Seligman, Steven F. Maier, J. Bruce Overmier, and their colleagues discovered that the controllability of events has a large impact on future learning. Dogs exposed to inescapable electric shock became passive and failed to learn to escape shock in later situations in which escape was possible. Seligman and his colleagues called this phenomenon learned helplessness because the dogs had learned that escape was not possible and gave up. The laboratory phenomenon of learned helplessness has been applied to the understanding and treatment of human depression and related conditions.

In the 1970s, some psychologists thought the use of rewards (such as praise or tangible items) was harmful to motivation, interest, and creativity. Beginning in the 1990s, however, American Robert Eisenberger and Canadian Judy Cameron, conducting research and analyzing previous studies, found that rewards generally have beneficial impacts. Rewards appear to have detrimental effects only when they are given regardless of how the person or animal does. Furthermore, the work of Allen Neuringer and his colleagues has shown that, contrary to previous thinking, both people and animals can learn to behave in random, unpredictable ways.

The changes in behavior produced by learning are accompanied by changes in physiological makeup. Learning is associated with changes in the strength of connections between neurons (nerve cells in the brain), some quite long-lasting. Eric R. Kandel and his colleagues have documented the changes in physiology underlying relatively simple learning in giant sea snails, progressing to more complex behaviors in mammals. Similar physiological changes accompany learning in a variety of organisms, highlighting the continuity of learning across different species.

Classical Conditioning

Classical conditioning was first systematically investigated by Pavlov beginning in the late 1800s and into the 1900s. Classical conditioning involves the transfer of control of an elicited response from one to another, previously neutral, stimulus. Pavlov discovered classical conditioning accidentally while investigating digestion in dogs. A dog was given meat powder in its mouth to elicit salivation. After this process had been repeated a number of times, the dog would start salivating before the meat powder was put in its mouth. When it saw the laboratory assistant, it would start to salivate, although it had not initially salivated at the sight. Pavlov devoted the rest of his long career to the phenomenon of classical conditioning.

In classical conditioning, a response is initially elicited by an (US). The US is a stimulus that elicits a response without any prior experience. For example, the loud sound of a balloon bursting naturally causes people to blink their eyes and withdraw from the noise. The response that is naturally elicited is called the (UR). If some stimulus reliably precedes the US, then over time it, too, will come to elicit a response. For example, the sight of an overfull balloon initially does not elicit blinking of the eyes. Because the sight of the balloon predicts the loud noise to come when it bursts, however, eventually people come to blink and recoil at the sight of an overfull balloon. The stimulus with the new power to elicit the response is called the conditioned stimulus (CS) and the response elicited by the CS is called the conditioned response (CR).

Classical conditioning occurs with a variety of behaviors and situations. For example, a person who was stung by a wasp in a woodshed may now experience fear on approaching the building. In this case, the woodshed becomes a CS eliciting the CR of fear because the wasp’s sting (the US) elicited pain and fear (the UR) in that place. To overcome the classical conditioning, the person would need to enter the woodshed repeatedly without incident. If the woodshed was no longer paired with the painful sting of the wasp, over time the CR would extinguish.

Many are thought to arise through classical conditioning. One common successful treatment is systematic desensitization, in which the person, through progressive steps, gradually faces the feared object or situation until the fear CR extinguishes. Classical conditioning has also been recognized as the culprit in food aversions developed by people receiving chemotherapy treatments for cancer. In this case, the food becomes a CS for illness (the CR) by being paired with the chemotherapy treatment (the US) that later elicits illness (the UR). Using more advanced principles of classical conditioning learned through research with nonhumans, people are now able to reduce the degree of aversion that occurs to regular meals, thus preventing the person from developing revulsions to food, which would further complicate the treatment of the cancer by introducing potential nutritional problems.

Operant Conditioning

Operant conditioning (also called instrumental conditioning) involves the regulation of voluntary behavior by its consequences. Thorndike first systemically studied operant conditioning in the late 1800’s. He placed cats in puzzle boxes and measured the amount of time they took to escape to a waiting bowl of food. He found that with increasing experience, the cats escaped more quickly. Movements that resulted in being released from the box, such as stepping on a panel or clawing a loop in a string, became more frequent, whereas movements that were not followed by release became less frequent. This type of operant learning is called trial-and-error learning, because there is no system to teach the behavior. Instead, the organism makes many mistakes, which become less likely over time, and sometimes hits on the solution, which then becomes more likely over time.

Beginning in the 1930s, Skinner greatly extended and systematized the study of operant conditioning. One of his major contributions was to invent an apparatus called the operant chamber, which provided a controlled environment in which behavior was automatically recorded. In the operant chamber, an animal, such as a rat, would be able to make an arbitrary response, such as pressing a small lever on the side of the chamber with its paws. The apparatus could be programmed to record the response automatically and provide a consequence, such as a bit of food, to the animal. There are several advantages to this technique. First, the chamber filters out unplanned sights and sounds that could disturb the animal and affect ongoing behavior. Second, the animal is free to make the response at any time, and so response rate can vary over a wide range as a result of any experimental manipulations. This range means that response rate is a sensitive measure to detect the effects of changes the experimenter makes. Third, the automatic control and recording means that the procedure can be repeated exactly the same way in every experimental session and that the experimenter’s ideas about what should happen cannot influence the outcome. The operant conditioning chamber is used extensively today in experiments investigating the learning of a variety of species from different perspectives.

One major technique to teach new behavior is called . Shaping refers to providing a consequence for successive approximations to a desired response. For example, to teach a child to tie shoelaces, a parent might start by crossing the laces, forming the loops and crossing them, and having the child do the last part of pulling the loops tight. The parent would then praise the child. The parent could then gradually have the child do more and more of the task, until the whole task is successfully completed from the start. This type of approach ensures that the task is never too far out of reach of the child’s current capabilities. Shaping takes place when young children are learning language, too. At first, parents and other caregivers are overjoyed at any approximation of basic words. Over time, however, they require the sounds to be closer and closer to the final, precisely spoken performance. Shaping can be used to teach a wide variety of behaviors in humans and nonhumans. The critical feature is that the requirement for the reward is gradually increased, in pace with the developing skill. If for some reason the behavior deteriorates, then the requirement can be lowered until the person is once again successful, then proceed again through increasing levels of difficulty. In order for any consequence to be effective, it should occur immediately after the behavior and every time the behavior occurs.

Reinforcers and Punishers

In operant conditioning, there are four basic contingencies that can be used to modify the frequency of occurrence of nonreflexive behavior. A contingency refers to the relation between the situation, a behavior, and the consequence of the behavior. A reinforcement is a consequence that makes a behavior more likely in the future, whereas a is a consequence that makes a behavior less likely in the future. Reinforcements and punishments come in positive and negative forms. A positive consequence is the presentation of a stimulus or event as a result of the behavior, and a negative consequence is the removal of a stimulus or event as a result of the behavior. Correctly used, the terms “positive” and “negative” refer only to whether the event is presented or removed, not whether the action is judged good or bad.

A positive reinforcement is a consequence that increases the future likelihood of the behavior that produced it. For example, if a parent were to praise a child at dinner for eating properly with a fork, and as a result the child used the fork properly more often, then praise would have served as a . The vast majority of scientists studying learning recommend positive reinforcement as the best technique to promote learning. One can attempt to increase the desired appropriate behavior through positive reinforcement, rather than focusing on the undesired or inappropriate behavior. If the appropriate behavior becomes more frequent, then chances are that the inappropriate behavior will have become less frequent as well, because there are only so many things that a person can do at one time.

A is a consequence that increases the future likelihood of the behavior that removed it. For example, in many cars, a buzzer or bell sounds until the driver puts on the seat belt. In this case, putting on the seat belt is negatively reinforced by the removal of the noise. Another example of negative reinforcement occurs when a child is having a tantrum in a grocery store until given candy. The removal of the screaming would serve as a negative reinforcement for the parent’s behavior: In the future when the child was screaming, the parent would probably be more likely to give the child candy. Furthermore, the parent is providing positive reinforcement for screaming by presenting a consequence (candy) for a behavior (screaming) that makes the behavior more likely to occur in similar situations in the future. This example should make clear that reinforcement is defined in terms of the presentation or removal of an event increasing the likelihood of a behavior in the future, not in terms of intentions or opinions. Most parents would not consider the behavior inadvertently created and maintained in this way to be “positive.”

Positive punishment refers to the presentation of an event that decreases the likelihood of the behavior that produced it. For example, if a person touches a hot stove, the pain that ensues makes it much less likely that the person will touch the stove under those conditions in the future. In this case, the behavior (touching the stove) produces a stimulus (pain) that makes the behavior less frequent. Negative punishment, on the other hand, refers to the removal of an event that decreases the likelihood of the behavior that produced it. For example, if a birdwatcher walking through the woods makes a loud move that causes all of the birds to fly away, then the watcher would be less likely to move like that in the future. In this way, watchers learn to move quietly to avoid disturbing the birds they are trying to observe.

Negative reinforcement, positive punishment, and negative punishment all involve what is called aversive control. An aversive stimulus is anything that an organism will attempt to escape from or try to avoid if possible. Aversive control refers to learning produced through the use of an aversive stimulus. For example, parents sometimes use spanking or hitting in an attempt to teach their child not to do something, such as hitting another child. This type of approach has been shown to have a number of undesirable outcomes, however. One problem is that the appropriate or desired alternative behavior is not taught. In other words, the child does not learn what should be done instead of what was done. Another problem is that the use of aversive stimuli can produce aggression. Humans and nonhumans alike often respond to painful stimuli with an increased likelihood of aggression. The aggression may or may not be directed toward the person or thing that hurt them. Additionally, the use of aversive control can produce avoidance—children who have been spanked or hit may try to stay away from the person who hurt them. Furthermore, through observation, children who have been spanked may be more likely to use physical harm to others as an attempted solution when they encounter conflict. Indeed, corporal punishment (the use of spanking or other physical force intended to cause a child to experience pain, but not injury, for the purpose of correction) has been linked to many undesirable outcomes for children, some of which extend well into adulthood. Beginning in the 1970s, American psychologist Murray Straus and his colleagues investigated the impact of corporal punishment on children. Their findings indicated that the use of corporal punishment is associated with an increase in later antisocial behavior as a child, a decrease in cognitive development relative to children who are not spanked, and an increased likelihood of spousal abuse as an adult, in addition to several other detrimental outcomes.

Learned Helplessness

As Seligman, Maier, and Overmier discovered, exposure to uncontrollable aversive events can have profound impacts on future learning, a phenomenon called learned helplessness. In learned helplessness, an organism that has been exposed to uncontrollable aversive events later has an impaired ability to learn to escape from aversive situations and even to learn new, unrelated behaviors. The phenomenon was accidentally discovered in laboratory research with dogs. Seligman and his colleagues found that dogs that were exposed to electrical shocks in a harness, with no possibility of escape, later could not learn to escape shocks in a shuttle box in which they had only to jump to the other side. Disturbingly, they would lie down and whimper, not even trying to get away from the completely avoidable shocks. Dogs that had not been exposed to the uncontrollable shocks learned to escape in the shuttle box rapidly. More important, dogs exposed to the same number and pattern of shocks, but with the ability to turn them off, also had no trouble learning to escape in the shuttle box. In other words, it was the exposure to uncontrollable shocks, not just shocks, that produced the later deficit in escape learning. Moreover, the dogs that had been exposed to uncontrollable aversive events also had difficulties learning other, unrelated tasks. This basic result has since been found many times with many different types of situations, species, and types of aversive events. For example, learned helplessness has been shown to occur in dogs, cats, mice, rats, gerbils, goldfish, cockroaches, and even slugs. Humans show the learned helplessness phenomenon in laboratory studies as well. For example, people exposed to an uncontrollable loud static noise later solved fewer anagrams (word puzzles) than people exposed to the same amount and pattern of noise but who could turn it off.

Learned helplessness has major implications for the understanding and treatment of human depression. Although certainly the case with people is more complex, animals that have developed learned helplessness in the laboratory show similarities to depressed people. For example, they have generalized reduced behavioral output. Similarly, early on researchers discovered that learned helplessness in rats could be prevented by treatment with antidepressant medication. Furthermore, exposure to uncontrollable aversive events produces deficiencies in immune system function, resulting in greater physical ailments, in both animals and people. In people, serial combinations of uncontrollable aversive events, such as sudden and unexpected loss of a spouse or child, being laid off from a job, or losing a home to fire, can result in the feeling that one is powerless and doomed. These feelings of helplessness can then produce changes, such as decreased interest in life and increased illness, which further compound the situation. Fortunately, there are effective treatments for learned helplessness. One solution already mentioned is antidepressant medication, which may work in part because it overcomes the physiological changes produced by the helpless experience. Additionally, therapy to teach effective coping and successful learning experiences can reverse learned helplessness in people and laboratory animals.

Learned Creativity and Variability

Beginning in the 1970s, some psychologists began to criticize the use of rewards to promote learning. Tangible rewards as well as praise and attention, they argued, could interfere with creativity, problem-solving ability, motivation, and enjoyment. Fortunately, these concerns were allayed in the 1990’s by careful research and examination of previous research, most notably that of Eisenberger and Cameron. Together, they analyzed the results of more than one hundred published studies on the effects of rewards and found that, in general, rewards increase interest, motivation, and performance. The only situation in which rewards had detrimental effects was when they were offered independently of performance. In other words, giving “rewards” regardless of how the person does is bad for morale and interest.

Furthermore, several aspects of performance previously thought to be beyond the domain of learning, such as creativity and even randomlike behavior, have been demonstrated to be sensitive to consequences. Children can learn to be creative in their drawing, in terms of the number of novel pictures drawn, using rewards for novelty. Similarly, as shown by the work of American psychologist Allen Neuringer and his colleagues, people and animals alike can learn to engage in strings of unpredictable behavior that cannot be distinguished from the random sort of outcomes generated by a random number generator. This finding is particularly interesting given that this novel behavior has been found to generalize to new situations, beyond the situation in which the learning originally occurred. Learned variability has been demonstrated in dolphins, rats, pigeons, and humans, including children with autism. Learning to be creative and to try new approaches has important implications for many aspects of daily life and problem solving.

Biological Bases of Learning

The features of learning do not occur in a vacuum: They often produce lasting, physiological changes in the organism. The search for the physical underpinnings of learning has progressed from relatively basic reflexes in relatively simple organisms to more complex behaviors in mammals. Beginning in the 1960s, Kandel and his colleagues started to examine simple learning in the large sea snail Aplysia. This snail was chosen as a model to study physiological changes in learning because its nervous system is relatively simple, containing several thousand neurons (nerve cells) compared to the billions of neurons in mammals. The neurons are large, so researchers can identify individual cells and monitor them for changes as learning progresses. In this Nobel Prize–winning work, Kandel and his colleagues outlined many of the changes in the degree of responsiveness in connections between neurons that underlie classical conditioning processes. The same processes have been observed in other species, including mammals, and the work continues to expand to more complex behavior. This research shows the commonality in learning processes across species and emphasizes the progress in understanding the physical basis that underlies learning.

Bibliography

Branch, Marc N., and Timothy D. Hackenberg. “Humans Are Animals, Too: Connecting Animal Research to Human Behavior and Cognition.” In Learning and Behavior Therapy, edited by William O’Donohue. Boston, Mass.: Allyn & Bacon, 1998. Print.

Carroll, Marilyn E., and J. Bruce Overmier, eds. Animal Research and Human Health: Advancing Human Welfare through Behavioral Science. Washington, D.C.: American Psychological Association, 2001. Print.

Cherry, Kendra. "What Is Classical Conditioning in Psychology?" Reviewed by Steven Gans. Verywell Mind, 1 May 2023, www.verywellmind.com/classical-conditioning-2794859. Accessed 17 Jan. 2025.

Eisenberger, Robert, and Judy Cameron. “The Detrimental Effects of Reward: Myth or Reality?” American Psychologist 51, no. 11 (1996): 1153–66. Print.

Mazur, James E. Learning and Behavior. 6th ed. Upper Saddle River, N.J.: Prentice Hall, 2006. Print.

Menzel, Randolf, and Paul R. Benjamin. Invertebrate Learning and Memory. London: Elsevier, 2013. Print.

Overmier, J. Bruce, and V. M. LoLordo. “Learned Helplessness.” In Learning and Behavior Therapy, edited by William O’Donohue. Boston: Allyn & Bacon, 1998. Print.

Schachtman, Todd R., and Steve Reilly. Associative Learning and Conditioning Theory: Human and Non-Human Applications. New York: Oxford UP, 2011. Print.

Seligman, Martin E. P. Learned Optimism. New York: Random House, 2006. Print.

Skinner, B. F. Science and Human Behavior. 1953. Reprint. Delray, N.J.: Classics of Medicine Library, 2000. Print.

Straus, Murray A., and Denise A. Donnelly. Beating the Devil out of Them: Corporal Punishment in American Families and Its Effects on Children. New Brunswick, N.J.: Transaction, 2001. Print.

Taylor, Edward W., and Patricia Cranton. The Handbook of Transformative Learning: Theory, Research, and Practice. San Francisco: Jossey-Bass, 2012. Print.