Operant conditioning, also called instrumental conditioning, is a learning process where voluntary behaviors are modified by association with the addition (or removal) of reward or aversive stimuli. The frequency or duration of the behavior may increase through reinforcement or decrease through punishment or extinction.
Operant conditioning originated in the work of Edward Thorndike, whose law of effect theorised that behaviors arise as a result of whether their consequences are satisfying or discomforting. In the 20th century, operant conditioning was studied by behavioral psychologists, who believed that much, if not all, of mind and behaviour can be explained as a result of environmental conditioning. Reinforcements are environmental stimuli that increase behaviors, whereas punishments are stimuli that decrease behaviors. Both kinds of stimuli can be further categorised into positive and negative stimuli, which respectively involve the addition or removal of environmental stimuli.
Operant conditioning differs from classical conditioning, which is a process where stimuli are paired with biologically significant events to produce involuntary and reflexive behaviors. In contrast, operant conditioning is voluntary and depends on the consequences of a behavior.
The study of animal learning in the 20th century was dominated by the analysis of these two sorts of learning,[1] and they are still at the core of behavior analysis. They have also been applied to the study of social psychology, helping to clarify certain phenomena such as the false consensus effect.[2]
See main article: Law of effect. Operant conditioning, sometimes called instrumental learning, was first extensively studied by Edward L. Thorndike (1874–1949), who observed the behavior of cats trying to escape from home-made puzzle boxes.[3] A cat could escape from the box by a simple response such as pulling a cord or pushing a pole, but when first constrained, the cats took a long time to get out. With repeated trials ineffective responses occurred less frequently and successful responses occurred more frequently, so the cats escaped more and more quickly. Thorndike generalized this finding in his law of effect, which states that behaviors followed by satisfying consequences tend to be repeated and those that produce unpleasant consequences are less likely to be repeated. In short, some consequences strengthen behavior and some consequences weaken behavior. By plotting escape time against trial number Thorndike produced the first known animal learning curves through this procedure.[4]
Humans appear to learn many simple behaviors through the sort of process studied by Thorndike, now called operant conditioning. That is, responses are retained when they lead to a successful outcome and discarded when they do not, or when they produce aversive effects. This usually happens without being planned by any "teacher", but operant conditioning has been used by parents in teaching their children for thousands of years.[5]
See main article: B. F. Skinner. B.F. Skinner (1904–1990) is referred to as the Father of operant conditioning, and his work is frequently cited in connection with this topic. His 1938 book "The Behavior of Organisms: An Experimental Analysis",[6] initiated his lifelong study of operant conditioning and its application to human and animal behavior. Following the ideas of Ernst Mach, Skinner rejected Thorndike's reference to unobservable mental states such as satisfaction, building his analysis on observable behavior and its equally observable consequences.[7]
Skinner believed that classical conditioning was too simplistic to be used to describe something as complex as human behavior. Operant conditioning, in his opinion, better described human behavior as it examined causes and effects of intentional behavior.
To implement his empirical approach, Skinner invented the operant conditioning chamber, or "Skinner Box", in which subjects such as pigeons and rats were isolated and could be exposed to carefully controlled stimuli. Unlike Thorndike's puzzle box, this arrangement allowed the subject to make one or two simple, repeatable responses, and the rate of such responses became Skinner's primary behavioral measure.[8] Another invention, the cumulative recorder, produced a graphical record from which these response rates could be estimated. These records were the primary data that Skinner and his colleagues used to explore the effects on response rate of various reinforcement schedules.[9] A reinforcement schedule may be defined as "any procedure that delivers reinforcement to an organism according to some well-defined rule".[10] The effects of schedules became, in turn, the basic findings from which Skinner developed his account of operant conditioning. He also drew on many less formal observations of human and animal behavior.[11]
Many of Skinner's writings are devoted to the application of operant conditioning to human behavior.[12] In 1948 he published Walden Two, a fictional account of a peaceful, happy, productive community organized around his conditioning principles.[13] In 1957, Skinner published Verbal Behavior,[14] which extended the principles of operant conditioning to language, a form of human behavior that had previously been analyzed quite differently by linguists and others. Skinner defined new functional relationships such as "mands" and "tacts" to capture some essentials of language, but he introduced no new principles, treating verbal behavior like any other behavior controlled by its consequences, which included the reactions of the speaker's audience.
Operant behavior is said to be "emitted"; that is, initially it is not elicited by any particular stimulus. Thus one may ask why it happens in the first place. The answer to this question is like Darwin's answer to the question of the origin of a "new" bodily structure, namely, variation and selection. Similarly, the behavior of an individual varies from moment to moment, in such aspects as the specific motions involved, the amount of force applied, or the timing of the response. Variations that lead to reinforcement are strengthened, and if reinforcement is consistent, the behavior tends to remain stable. However, behavioral variability can itself be altered through the manipulation of certain variables.[15]
See main article: Reinforcement and Punishment (psychology). Reinforcement and punishment are the core tools through which operant behavior is modified. These terms are defined by their effect on behavior. Either may be positive or negative.
Another procedure is called "extinction".
There are a total of five consequences.
Actors (e.g. a rat) are not spoken of as being reinforced, punished, or extinguished; it is the actions that are reinforced, punished, or extinguished. Reinforcement, punishment, and extinction are not terms whose use is restricted to the laboratory. Naturally-occurring consequences can also reinforce, punish, or extinguish behavior and are not always planned or delivered on purpose.
Schedules of reinforcement are rules that control the delivery of reinforcement. The rules specify either the time that reinforcement is to be made available, or the number of responses to be made, or both. Many rules are possible, but the following are the most basic and commonly used[18] [9]
The effectiveness of reinforcement and punishment can be changed.
Most of these factors serve biological functions. For example, the process of satiation helps the organism maintain a stable internal environment (homeostasis). When an organism has been deprived of sugar, for example, the taste of sugar is an effective reinforcer. When the organism's blood sugar reaches or exceeds an optimum level the taste of sugar becomes less effective or even aversive.
See main article: Shaping (psychology). Shaping is a conditioning method often used in animal training and in teaching nonverbal humans. It depends on operant variability and reinforcement, as described above. The trainer starts by identifying the desired final (or "target") behavior. Next, the trainer chooses a behavior that the animal or person already emits with some probability. The form of this behavior is then gradually changed across successive trials by reinforcing behaviors that approximate the target behavior more and more closely. When the target behavior is finally emitted, it may be strengthened and maintained by the use of a schedule of reinforcement.
Noncontingent reinforcement is the delivery of reinforcing stimuli regardless of the organism's behavior. Noncontingent reinforcement may be used in an attempt to reduce an undesired target behavior by reinforcing multiple alternative responses while extinguishing the target response.[21] As no measured behavior is identified as being strengthened, there is controversy surrounding the use of the term noncontingent "reinforcement".[22]
See main article: Stimulus control. Though initially operant behavior is emitted without an identified reference to a particular stimulus, during operant conditioning operants come under the control of stimuli that are present when behavior is reinforced. Such stimuli are called "discriminative stimuli." A so-called "three-term contingency" is the result. That is, discriminative stimuli set the occasion for responses that produce reward or punishment. Example: a rat may be trained to press a lever only when a light comes on; a dog rushes to the kitchen when it hears the rattle of his/her food bag; a child reaches for candy when s/he sees it on a table.
Most behavior is under stimulus control. Several aspects of this may be distinguished:
Most behavior cannot easily be described in terms of individual responses reinforced one by one. The scope of operant analysis is expanded through the idea of behavioral chains, which are sequences of responses bound together by the three-term contingencies defined above. Chaining is based on the fact, experimentally demonstrated, that a discriminative stimulus not only sets the occasion for subsequent behavior, but it can also reinforce a behavior that precedes it. That is, a discriminative stimulus is also a "conditioned reinforcer". For example, the light that sets the occasion for lever pressing may be used to reinforce "turning around" in the presence of a noise. This results in the sequence "noise – turn-around – light – press lever – food". Much longer chains can be built by adding more stimuli and responses.
In escape learning, a behavior terminates an (aversive) stimulus. For example, shielding one's eyes from sunlight terminates the (aversive) stimulation of bright light in one's eyes. (This is an example of negative reinforcement, defined above.) Behavior that is maintained by preventing a stimulus is called "avoidance," as, for example, putting on sun glasses before going outdoors. Avoidance behavior raises the so-called "avoidance paradox", for, it may be asked, how can the non-occurrence of a stimulus serve as a reinforcer? This question is addressed by several theories of avoidance (see below).
Two kinds of experimental settings are commonly used: discriminated and free-operant avoidance learning.
A discriminated avoidance experiment involves a series of trials in which a neutral stimulus such as a light is followed by an aversive stimulus such as a shock. After the neutral stimulus appears an operant response such as a lever press prevents or terminate the aversive stimulus. In early trials, the subject does not make the response until the aversive stimulus has come on, so these early trials are called "escape" trials. As learning progresses, the subject begins to respond during the neutral stimulus and thus prevents the aversive stimulus from occurring. Such trials are called "avoidance trials." This experiment is said to involve classical conditioning because a neutral CS (conditioned stimulus) is paired with the aversive US (unconditioned stimulus); this idea underlies the two-factor theory of avoidance learning described below.
In free-operant avoidance a subject periodically receives an aversive stimulus (often an electric shock) unless an operant response is made; the response delays the onset of the shock. In this situation, unlike discriminated avoidance, no prior stimulus signals the shock. Two crucial time intervals determine the rate of avoidance learning. This first is the S-S (shock-shock) interval. This is time between successive shocks in the absence of a response. The second interval is the R-S (response-shock) interval. This specifies the time by which an operant response delays the onset of the next shock. Each time the subject performs the operant response, the R-S interval without shock begins anew.
This theory was originally proposed in order to explain discriminated avoidance learning, in which an organism learns to avoid an aversive stimulus by escaping from a signal for that stimulus. Two processes are involved: classical conditioning of the signal followed by operant conditioning of the escape response:
a) Classical conditioning of fear. Initially the organism experiences the pairing of a CS with an aversive US. The theory assumes that this pairing creates an association between the CS and the US through classical conditioning and, because of the aversive nature of the US, the CS comes to elicit a conditioned emotional reaction (CER) – "fear." b) Reinforcement of the operant response by fear-reduction. As a result of the first process, the CS now signals fear; this unpleasant emotional reaction serves to motivate operant responses, and responses that terminate the CS are reinforced by fear termination. The theory does not say that the organism "avoids" the US in the sense of anticipating it, but rather that the organism "escapes" an aversive internal state that is caused by the CS.Several experimental findings seem to run counter to two-factor theory. For example, avoidance behavior often extinguishes very slowly even when the initial CS-US pairing never occurs again, so the fear response might be expected to extinguish (see Classical conditioning). Further, animals that have learned to avoid often show little evidence of fear, suggesting that escape from fear is not necessary to maintain avoidance behavior.[23]
Some theorists suggest that avoidance behavior may simply be a special case of operant behavior maintained by its consequences. In this view the idea of "consequences" is expanded to include sensitivity to a pattern of events. Thus, in avoidance, the consequence of a response is a reduction in the rate of aversive stimulation. Indeed, experimental evidence suggests that a "missed shock" is detected as a stimulus, and can act as a reinforcer. Cognitive theories of avoidance take this idea a step farther. For example, a rat comes to "expect" shock if it fails to press a lever and to "expect no shock" if it presses it, and avoidance behavior is strengthened if these expectancies are confirmed.[23]
Operant hoarding refers to the observation that rats reinforced in a certain way may allow food pellets to accumulate in a food tray instead of retrieving those pellets. In this procedure, retrieval of the pellets always instituted a one-minute period of extinction during which no additional food pellets were available but those that had been accumulated earlier could be consumed. This finding appears to contradict the usual finding that rats behave impulsively in situations in which there is a choice between a smaller food object right away and a larger food object after some delay. See schedules of reinforcement.[24]
The first scientific studies identifying neurons that responded in ways that suggested they encode for conditioned stimuli came from work by Mahlon deLong[25] [26] and by R.T. Richardson.[26] They showed that nucleus basalis neurons, which release acetylcholine broadly throughout the cerebral cortex, are activated shortly after a conditioned stimulus, or after a primary reward if no conditioned stimulus exists. These neurons are equally active for positive and negative reinforcers, and have been shown to be related to neuroplasticity in many cortical regions.[27] Evidence also exists that dopamine is activated at similar times. There is considerable evidence that dopamine participates in both reinforcement and aversive learning.[28] Dopamine pathways project much more densely onto frontal cortex regions. Cholinergic projections, in contrast, are dense even in the posterior cortical regions like the primary visual cortex. A study of patients with Parkinson's disease, a condition attributed to the insufficient action of dopamine, further illustrates the role of dopamine in positive reinforcement.[29] It showed that while off their medication, patients learned more readily with aversive consequences than with positive reinforcement. Patients who were on their medication showed the opposite to be the case, positive reinforcement proving to be the more effective form of learning when dopamine activity is high.
A neurochemical process involving dopamine has been suggested to underlie reinforcement. When an organism experiences a reinforcing stimulus, dopamine pathways in the brain are activated. This network of pathways "releases a short pulse of dopamine onto many dendrites, thus broadcasting a global reinforcement signal to postsynaptic neurons."[30] This allows recently activated synapses to increase their sensitivity to efferent (conducting outward) signals, thus increasing the probability of occurrence for the recent responses that preceded the reinforcement. These responses are, statistically, the most likely to have been the behavior responsible for successfully achieving reinforcement. But when the application of reinforcement is either less immediate or less contingent (less consistent), the ability of dopamine to act upon the appropriate synapses is reduced.
A number of observations seem to show that operant behavior can be established without reinforcement in the sense defined above. Most cited is the phenomenon of autoshaping (sometimes called "sign tracking"), in which a stimulus is repeatedly followed by reinforcement, and in consequence the animal begins to respond to the stimulus. For example, a response key is lighted and then food is presented. When this is repeated a few times a pigeon subject begins to peck the key even though food comes whether the bird pecks or not. Similarly, rats begin to handle small objects, such as a lever, when food is presented nearby.[31] [32] Strikingly, pigeons and rats persist in this behavior even when pecking the key or pressing the lever leads to less food (omission training).[33] [34] Another apparent operant behavior that appears without reinforcement is contrafreeloading.
These observations and others appear to contradict the law of effect, and they have prompted some researchers to propose new conceptualizations of operant reinforcement (e.g.[35] [36] [37]) A more general view is that autoshaping is an instance of classical conditioning; the autoshaping procedure has, in fact, become one of the most common ways to measure classical conditioning. In this view, many behaviors can be influenced by both classical contingencies (stimulus-response) and operant contingencies (response-reinforcement), and the experimenter's task is to work out how these interact.[38]
Reinforcement and punishment are ubiquitous in human social interactions, and a great many applications of operant principles have been suggested and implemented. The following are some examples.
Positive and negative reinforcement play central roles in the development and maintenance of addiction and drug dependence. An addictive drug is intrinsically rewarding; that is, it functions as a primary positive reinforcer of drug use. The brain's reward system assigns it incentive salience (i.e., it is "wanted" or "desired"),[39] [40] [41] so as an addiction develops, deprivation of the drug leads to craving. In addition, stimuli associated with drug use – e.g., the sight of a syringe, and the location of use – become associated with the intense reinforcement induced by the drug. These previously neutral stimuli acquire several properties: their appearance can induce craving, and they can become conditioned positive reinforcers of continued use. Thus, if an addicted individual encounters one of these drug cues, a craving for the associated drug may reappear. For example, anti-drug agencies previously used posters with images of drug paraphernalia as an attempt to show the dangers of drug use. However, such posters are no longer used because of the effects of incentive salience in causing relapse upon sight of the stimuli illustrated in the posters.
In drug dependent individuals, negative reinforcement occurs when a drug is self-administered in order to alleviate or "escape" the symptoms of physical dependence (e.g., tremors and sweating) and/or psychological dependence (e.g., anhedonia, restlessness, irritability, and anxiety) that arise during the state of drug withdrawal.
See main article: Animal training. Animal trainers and pet owners were applying the principles and practices of operant conditioning long before these ideas were named and studied, and animal training still provides one of the clearest and most convincing examples of operant control. Of the concepts and procedures described in this article, a few of the most salient are the following: (a) availability of primary reinforcement (e.g. a bag of dog yummies); (b) the use of secondary reinforcement, (e.g. sounding a clicker immediately after a desired response, then giving yummy); (c) contingency, assuring that reinforcement (e.g. the clicker) follows the desired behavior and not something else; (d) shaping, as in gradually getting a dog to jump higher and higher; (e) intermittent reinforcement, as in gradually reducing the frequency of reinforcement to induce persistent behavior without satiation; (f) chaining, where a complex behavior is gradually constructed from smaller units.[42]
See main article: Applied behavior analysis. Applied behavior analysis is the discipline initiated by B. F. Skinner that applies the principles of conditioning to the modification of socially significant human behavior. It uses the basic concepts of conditioning theory, including conditioned stimulus (SC), discriminative stimulus (Sd), response (R), and reinforcing stimulus (Srein or Sr for reinforcers, sometimes Save for aversive stimuli).[23]
Practitioners of applied behavior analysis (ABA) bring these procedures, and many variations and developments of them, to bear on a variety of socially significant behaviors and issues. In many cases, practitioners use operant techniques to develop constructive, socially acceptable behaviors to replace aberrant behaviors. The techniques of ABA have been effectively applied in to such things as early intensive behavioral interventions for children with an autism spectrum disorder (ASD)[43] research on the principles influencing criminal behavior, HIV prevention,[44] conservation of natural resources,[45] education,[46] gerontology,[47] health and exercise,[48] industrial safety,[49] language acquisition,[50] littering,[51] medical procedures,[52] parenting,[53] psychotherapy, seatbelt use,[54] severe mental disorders,[55] sports,[56] substance abuse, phobias, pediatric feeding disorders, and zoo management and care of animals.[57] Some of these applications are among those described below.
See main article: Parent management training. Providing positive reinforcement for appropriate child behaviors is a major focus of parent management training. Typically, parents learn to reward appropriate behavior through social rewards (such as praise, smiles, and hugs) as well as concrete rewards (such as stickers or points towards a larger reward as part of an incentive system created collaboratively with the child).[58] In addition, parents learn to select simple behaviors as an initial focus and reward each of the small steps that their child achieves towards reaching a larger goal (this concept is called "successive approximations").[58] [59]
See main article: Behavioral economics. Both psychologists and economists have become interested in applying operant concepts and findings to the behavior of humans in the marketplace. An example is the analysis of consumer demand, as indexed by the amount of a commodity that is purchased. In economics, the degree to which price influences consumption is called "the price elasticity of demand." Certain commodities are more elastic than others; for example, a change in price of certain foods may have a large effect on the amount bought, while gasoline and other everyday consumables may be less affected by price changes. In terms of operant analysis, such effects may be interpreted in terms of motivations of consumers and the relative value of the commodities as reinforcers.[60]
See main article: Gambling. As stated earlier in this article, a variable ratio schedule yields reinforcement after the emission of an unpredictable number of responses. This schedule typically generates rapid, persistent responding. Slot machines pay off on a variable ratio schedule, and they produce just this sort of persistent lever-pulling behavior in gamblers. The variable ratio payoff from slot machines and other forms of gambling has often been cited as a factor underlying gambling addiction.[61]
See main article: Military psychology.
Human beings have an innate resistance to killing and are reluctant to act in a direct, aggressive way towards members of their own species, even to save life. This resistance to killing has caused infantry to be remarkably inefficient throughout the history of military warfare.[62]
This phenomenon was not understood until S.L.A. Marshall (Brigadier General and military historian) undertook interview studies of WWII infantry immediately following combat engagement. Marshall's well-known and controversial book, Men Against Fire, revealed that only 15% of soldiers fired their rifles with the purpose of killing in combat.[63] Following acceptance of Marshall's research by the US Army in 1946, the Human Resources Research Office of the US Army began implementing new training protocols which resemble operant conditioning methods. Subsequent applications of such methods increased the percentage of soldiers able to kill to around 50% in Korea and over 90% in Vietnam. Revolutions in training included replacing traditional pop-up firing ranges with three-dimensional, man-shaped, pop-up targets which collapsed when hit. This provided immediate feedback and acted as positive reinforcement for a soldier's behavior.[64] Other improvements to military training methods have included the timed firing course; more realistic training; high repetitions; praise from superiors; marksmanship rewards; and group recognition. Negative reinforcement includes peer accountability or the requirement to retake courses. Modern military training conditions mid-brain response to combat pressure by closely simulating actual combat, using mainly Pavlovian classical conditioning and Skinnerian operant conditioning (both forms of behaviorism).
Modern marksmanship training is such an excellent example of behaviorism that it has been used for years in the introductory psychology course taught to all cadets at the US Military Academy at West Point as a classic example of operant conditioning. In the 1980s, during a visit to West Point, B.F. Skinner identified modern military marksmanship training as a near-perfect application of operant conditioning.[64]Lt. Col. Dave Grossman states about operant conditioning and US Military training that:
It is entirely possible that no one intentionally sat down to use operant conditioning or behavior modification techniques to train soldiers in this area…But from the standpoint of a psychologist who is also a historian and a career soldier, it has become increasingly obvious to me that this is exactly what has been achieved.
See main article: Nudge theory. Nudge theory (or nudge) is a concept in behavioural science, political theory and economics which argues that indirect suggestions to try to achieve non-forced compliance can influence the motives, incentives and decision making of groups and individuals, at least as effectively – if not more effectively – than direct instruction, legislation, or enforcement.
See main article: Praise. The concept of praise as a means of behavioral reinforcement is rooted in B.F. Skinner's model of operant conditioning. Through this lens, praise has been viewed as a means of positive reinforcement, wherein an observed behavior is made more likely to occur by contingently praising said behavior.[65] Hundreds of studies have demonstrated the effectiveness of praise in promoting positive behaviors, notably in the study of teacher and parent use of praise on child in promoting improved behavior and academic performance,[66] but also in the study of work performance.[67] Praise has also been demonstrated to reinforce positive behaviors in non-praised adjacent individuals (such as a classmate of the praise recipient) through vicarious reinforcement.[68] Praise may be more or less effective in changing behavior depending on its form, content and delivery. In order for praise to effect positive behavior change, it must be contingent on the positive behavior (i.e., only administered after the targeted behavior is enacted), must specify the particulars of the behavior that is to be reinforced, and must be delivered sincerely and credibly.[69]
Acknowledging the effect of praise as a positive reinforcement strategy, numerous behavioral and cognitive behavioral interventions have incorporated the use of praise in their protocols.[70] [71] The strategic use of praise is recognized as an evidence-based practice in both classroom management and parenting training interventions,[72] though praise is often subsumed in intervention research into a larger category of positive reinforcement, which includes strategies such as strategic attention and behavioral rewards.
Several studies have been done on the effect cognitive-behavioral therapy and operant-behavioral therapy have on different medical conditions. When patients developed cognitive and behavioral techniques that changed their behaviors, attitudes, and emotions; their pain severity decreased. The results of these studies showed an influence of cognitions on pain perception and impact presented explained the general efficacy of Cognitive-Behavioral therapy (CBT) and Operant-Behavioral therapy (OBT).
See main article: Compulsion loop. The majority of video games are designed around a compulsion loop, adding a type of positive reinforcement through a variable rate schedule to keep the player playing. This can lead to the pathology of video game addiction.[73]
See main article: Loot box. As part of a trend in the monetization of video games during the 2010s, some games offered loot boxes as rewards or as items purchasable by real world funds. Boxes contains a random selection of in-game items. The practice has been tied to the same methods that slot machines and other gambling devices dole out rewards, as it follows a variable rate schedule. While the general perception that loot boxes are a form of gambling, the practice is only classified as such in a few countries. However, methods to use those items as virtual currency for online gambling or trading for real world money has created a skin gambling market that is under legal evaluation.[74]
One of the many reasons proposed for the dramatic costs associated with healthcare is the practice of defensive medicine. Prabhu reviews the article by Cole and discusses how the responses of two groups of neurosurgeons are classic operant behavior. One group practice in a state with restrictions on medical lawsuits and the other group with no restrictions. The group of neurosurgeons were queried anonymously on their practice patterns. The physicians changed their practice in response to a negative feedback (fear from lawsuit) in the group that practiced in a state with no restrictions on medical lawsuits.[75]