Research figure: B. F. Skinner · Written by Artificial Sapiens
B. F. Skinner's account of reinforcement and punishment explains how the consequences of an action change the likelihood that the action will occur again under comparable conditions. In his experimental analysis of operant behavior, reinforcement names a demonstrable strengthening relation between a response and its consequence; punishment concerns response reduction through consequential events. Positive reinforcement involves contingent presentation of an event that strengthens responding, while negative reinforcement involves contingent withdrawal, termination, prevention, or reduction of an aversive event that strengthens responding. Punishment concerns a contingency that reduces the subsequent occurrence of the targeted response. These concepts describe functional relations established through observation and experiment, rather than the presumed kindness, unpleasantness, or intention of an adult, clinician, or institution. Skinner, 1953; Virues-Ortega et al., 2021.
This chapter examines Skinner's own formulation and experimental reasoning, the place of escape and avoidance in his account, the complex relation between reinforcement and aversive control, and how later investigators made the scientific treatment of punishment more systematic. The historical theory is considered alongside contemporary experimental, cognitive, neuroscientific, and applied evidence. Its central question is how an investigator can identify a behavioral consequence by its effect, while accounting for the environment, the individual's learning history, and the ethical consequences of arranging contingencies.
Historical foundations: From the law of effect to operant analysis
Edward L. Thorndike developed the law of effect before Skinner established his laboratory research program. Thorndike examined how consequences influenced the strengthening of stimulus–response connections; later summaries of his law have sometimes altered its original meaning. John B. Watson's early behaviorist program gave the study of publicly observable behavior a disciplinary platform. Skinner's distinctive achievement was an experimental framework in which the rate and pattern of emitted operant responses could be examined as functions of programmed consequences, deprivation conditions, discriminative stimuli, and reinforcement schedules. Domjan, 2026; Skinner Foundation bibliography.
In The Behavior of Organisms: An Experimental Analysis (1938), Skinner presented a systematic account based on the behavior of individual organisms and the experimental manipulation of environmental variables. He subsequently extended the framework to the analysis of human affairs in Science and Human Behavior (1953), where chapters on reinforcement, punishment, aversive control, and self-control address issues broader than the apparatus of the laboratory. These are two historical phases of a sustained research program, not successive claims that all human conduct can be reduced to one experimental procedure. Skinner Foundation books; Skinner, 1953.
The influential Schedules of Reinforcement (1957) was a joint work by Charles B. Ferster and B. F. Skinner. Its detailed study of temporally and numerically arranged contingencies further showed that a consequence's effect depends on how it is scheduled and on the organism's learning history. Punishment, however, was subjected to a particularly extensive independent experimental program by Nathan H. Azrin and colleagues in the following decades. Their findings are central to evaluating and refining claims associated with Skinner. Skinner Foundation books; Virues-Ortega et al., 2021.
What reinforcement means in Skinner's system
Reinforcement is an experimentally inferred relation: a response produces, terminates, or otherwise changes a consequential event, and the response subsequently becomes more frequent or probable in the relevant context. The presence of a pleasant event alone does not establish reinforcement. An adult may call praise a reward, but if praise leaves the measured target behavior unchanged, the evidence does not demonstrate that it functioned as a reinforcer in that arrangement. Likewise, a consequence experienced as unpleasant can support responding when the organism learns that performing an action brings relief or prevents an aversive event. Skinner, 1953.
Skinner's term operant emphasizes that an organism emits responses whose future probability is influenced by their consequences. Operant control is usually studied within a three-term relation: antecedent or discriminative stimulus, operant response, and consequence. A discriminative stimulus signals a history in which a response has been followed by a particular consequence; it need not cause the response as a mechanically eliciting reflex. Motivating operations such as deprivation and satiation, behavioral alternatives, and the availability of reinforcement modify what an experimenter will observe. Functional identification therefore requires specifying what response was counted, under what conditions, and across which observations.
The meaning of reinforcement is also conditional on the individual's history. Food can maintain a food-deprived organism's operant responding in a chamber. The same food may have much less effect following satiation. A classroom practice that sustains participation for one student may exert little control over another. General statements about reinforcement should therefore be read as statements about testable contingencies rather than fixed intrinsic properties of stimuli.
Positive reinforcement: Response-produced presentation
Positive reinforcement occurs when a consequence is added after a response, and the contingency increases the future occurrence of that response. In an operant chamber, food delivered contingent on a rat pressing a lever can maintain lever pressing, provided the experimental conditions give the food that function. For a student, access to a preferred activity after completing a defined learning step might increase completion of comparable steps. The operational issue is always whether the future response changes because of the contingent delivery, relative to suitable comparison conditions.
The word positive designates presentation or addition, not psychological benefit or moral approval. Positive reinforcement can maintain adaptive skills, perseverative rituals, risky acts, or coercive interaction patterns. Approval, attention, tangible items, sensations, and opportunities to engage in activities can each serve as consequences, although none qualifies automatically as an effective reinforcer. Researchers may compare contingent and noncontingent conditions, observe behavior across reversals, or examine response allocation when alternatives are available to establish what maintains the behavior. Richman et al., 2015.
Several questions matter when interpreting apparent positive reinforcement. Did the behavior increase only after a preferred object was made available, or was it already increasing through practice? Did another person provide attention simultaneously? Was the supposed reinforcer delivered after the specific target response, or after an entire chain of behaviors? Does the effect persist after the immediate setting changes? These questions are integral to the experimental method and protect the concept from becoming a label for any encouraging gesture.
Negative reinforcement: Relief and prevention that strengthen behavior
Negative reinforcement occurs when a response becomes more likely because it terminates, decreases, postpones, or prevents an aversive condition. The classic escape arrangement begins with an unpleasant event already occurring: the organism performs an action, the event ends, and the action is subsequently strengthened. Avoidance arrangements differ because a response prevents or postpones a predicted aversive event, sometimes after a warning signal and sometimes without an explicit discrete warning cue. Negative reinforcement strengthens the escaping or avoiding response; punishment, as a behavioral process, decreases the punished response. Skinner, 1953; Krypotos et al., 2015.
Imagine an alarm that continues sounding until a person fastens a seat belt. When fastening reliably stops the alarm and fastening becomes more likely on later occasions, the termination of the sound can function as negative reinforcement. For someone who worries about being judged, avoiding an interaction may bring immediate relief and maintain avoidance even when the avoidance creates longer-term problems. The immediate decrease in distress and the broader life consequences are distinct outcomes. In neither case does the word negative imply that the learner is being punished.
Escape versus avoidance
In escape learning, an ongoing aversive condition is terminated by a response. In avoidance learning, behavior prevents an expected or scheduled aversive condition, and the investigator must determine what features of the experimental setting support responding when the event is absent. Early explanatory approaches included Mowrer's two-factor theory, which combined Pavlovian fear learning and instrumentally reinforced fear reduction. Later research on avoidance, including unsignaled avoidance, challenged the sufficiency of a simple fear-reduction account and encouraged more complex behavioral and cognitive models. Krypotos et al., 2015.
Skinner's conceptual framework foregrounded operant contingencies and histories of aversive stimulation. It did not resolve every later theoretical question about what is learned during avoidance. Modern research examines expectations, threat appraisal, action–outcome knowledge, stimulus generalization, and neural systems in addition to measurable avoidance behavior. These developments elaborate a historical issue rather than transferring their authorship to Skinner.
Why the immediate relief trap matters
A negatively reinforced action can solve an urgent problem while also becoming inflexible. Repeatedly leaving a distressing situation may reliably remove discomfort, thereby increasing escape in similar situations. Whether that pattern is adaptive depends on the reality of the threat, the available choices, context, and longer-term outcomes. In clinical contexts, this mechanism is one hypothesis to assess, not a diagnosis or a blanket instruction to prevent escape. Safety, autonomy, trauma history, and collaboratively identified goals remain essential.
Punishment: The reduction of future responding
In contemporary functional usage, punishment refers to an event or contingency that decreases the future probability of a specified response under relevant conditions. Presentation of an event following a response may produce positive punishment; removal of an available reinforcer following the response may produce negative punishment. The words positive and negative describe additions and withdrawals, while the term punishment describes the observed response-reducing effect. A severe reprimand, an unpleasant sensation, loss of an opportunity, or temporary removal of access is therefore not scientifically classified by intention alone. The effect must be examined. Ayllon and Azrin, 1966; Virues-Ortega et al., 2021.
Skinner's discussion in Chapter 12 of Science and Human Behavior (1953) distinguished withdrawal of a positive reinforcer and presentation of an aversive or negative reinforcer as candidate operations relevant to punishment. The Foundation preserves a passage in which he emphasizes that both positive and negative reinforcers strengthen a response under their respective reinforcement contingencies. This historical statement is essential because a negative reinforcer in Skinner's terminology is an aversive stimulus whose removal strengthens responding; the modern instructional table of four consequence classes is a later standardized way of expressing related functional distinctions and should not be presented as a verbatim quotation from his book. B. F. Skinner Foundation, Chapter 12.
Skinner argued that punishment often produces suppression while leaving the conditions supporting the original behavior inadequately addressed. His skepticism toward punitive control rested partly on the analysis of unwanted emotional, escape, and avoidance effects. Such arguments were important to the historical discussion of control and human welfare. They are separate from the empirical assertion that every punishing procedure has identical effects. Later studies clarified how punishment varies with parameters, behavioral alternatives, and the environment. Skinner, 1953; Virues-Ortega et al., 2021.
Positive and negative punishment in functional examples
If a response is followed by a newly introduced aversive noise and the response decreases in otherwise comparable situations, the contingency fits positive punishment. If a response is followed by removal of access to a valued activity and responding subsequently decreases, the contingency fits negative punishment. The same stimulus can have different functions at different times. In experiments by Theodore Ayllon and Nathan Azrin, a noise suppressed a response under one contingency, then acquired discriminative or reinforcing properties under different arrangements. Its physical identity was insufficient to predict its behavioral function. Ayllon and Azrin, 1966.
A useful analytic contrast is therefore between the operation the researcher arranges and the function the event acquires. Calling an act a punishment expresses neither a causal demonstration nor an ethical justification. A response may temporarily stop because it becomes impossible, because the organism is distracted, because a competing action is reinforced, or because the physical context changes. Only appropriate measurement and controls help separate these possibilities from punishment as a learning process.
The four-contingency framework and its limits
Contemporary introductory teaching frequently summarizes operant consequences in four cells. Positive reinforcement adds a consequence and increases future responding. Negative reinforcement removes or avoids an aversive condition and increases responding. Positive punishment adds a consequence and decreases responding. Negative punishment removes a consequence and decreases responding. This matrix provides an operational vocabulary, provided each cell is assigned after an observed behavioral effect rather than solely from the observer's expectations.
An individual action can enter multiple contingency relations. A child who asks for a break may receive attention (potential positive reinforcement) and obtain a brief respite from a difficult task (potential negative reinforcement). The fact that two consequences follow the same behavior does not specify which maintains it. Likewise, a teacher's reprimand may briefly interrupt disruption but also deliver attention that increases later disruption in a specific classroom. An adequate analysis measures the pattern of responding before and after the contingency and assesses plausible alternative functions.
In ordinary language, reward often means a pleasing incentive, while punishment often means blame, penalty, or retribution. The behavioral definitions remain anchored to measured changes. A prize may fail to increase the intended action; a disliked event may maintain an escape response; and withholding a previously available reinforcer can reduce one activity while accidentally increasing another. Applied decisions therefore require person-specific evidence and attention to unintended effects.
How Skinner's experiments established causal relations
Skinner's experimental method studied individual organisms repeatedly under carefully controlled conditions. An operant chamber provided a measurable manipulandum such as a lever or response key, a device capable of arranging consequences, and recording equipment for observing changes in responding. When a contingency was introduced, withdrawn, or altered, researchers could assess changes in response rate and temporal pattern. Cumulative records transformed sequences of discrete responses into visible functions over time. The experimental question concerned relations between arranged contingencies and behavior, rather than whether the animal appeared to understand a reward or punishment in everyday terms. Skinner Foundation bibliography.
A convincing demonstration calls for more than a single pre–post difference. Behavior can change with time, hunger, fatigue, stimulation from the apparatus, earlier exposures, or shifts in competing reinforcement. Repeated conditions, stable baselines, counterbalancing where possible, and direct replication improve causal interpretation. In single-organism designs, replicated changes under systematically varied contingencies can offer strong within-subject evidence, though the number of organisms, experimental selection, and naturalistic generalizability limit the scope of inference.
Reinforcement schedules and response patterns
Continuous reinforcement, in which each eligible response produces the consequence, is analytically distinct from intermittent reinforcement, in which only some responses are followed by consequences or reinforcement depends on elapsed time. Fixed and variable interval or ratio arrangements produce different response patterns and resistance profiles. Ferster and Skinner's 1957 joint monograph documented these relationships in detail. The schedule changes the temporal structure of learning and matters when assessing whether apparent reinforcement or punishment effects are stable or transient. Ferster and Skinner, 1957.
Punishment likewise cannot be assessed in isolation from the schedule and magnitude of reinforcement sustaining the target behavior. A strongly reinforced action may continue despite a weak competing punishing contingency; a seemingly effective punishment may lose its effect if conditions change. Contemporary studies examine relative immediacy, intensity, reinforcement alternatives, response effort, and generalization. None of these findings licenses transferring a laboratory parameter directly into clinical practice without ethical and person-specific assessment.
What later punishment research actually demonstrated
Nathan H. Azrin and colleagues made decisive empirical contributions to the analysis of punishment and aversive control. His program systematically varied punishing stimuli, their intensity, immediacy, and scheduling, and the simultaneous availability of reinforcement. A historical synthesis documents experimental work with Donald Hake, William Holz, Ronald Hutchinson, Roger Ulrich, Theodore Ayllon, and others. Attributing this body of punishment research entirely to Skinner obscures a major scientific achievement of subsequent investigators. Virues-Ortega et al., 2021.
One experiment by Ayllon and Azrin (1966) involved human psychiatric patients responding in a two-alternative operant situation. Noise initially used as a punishing event suppressed responding on a targeted option; under a changed arrangement the noise acquired another behavioral function. The experimental result showed that a consequence's function depends on contingencies, and that aversive properties and reinforcing or discriminative properties can coexist or change across contexts. The findings were established in a specific institutional setting, with important restrictions on generalizing to all clinical populations. Ayllon and Azrin, 1966.
Azrin, Hake, Holz, and Hutchinson (1965) also examined escape from a situation where responses were punished into a situation where responses were not punished. Their experiment illustrates how punishment can induce escape-related behavior and how one contingency creates another set of consequences. In a separate series, Azrin, Hutchinson, and Hake (1966) demonstrated attack behavior by pigeons under extinction conditions. Extinction-induced aggression must be attributed to that experimentally arranged removal of reinforcement rather than generalized indiscriminately to punishment as such. Azrin et al., 1965; Azrin, Hutchinson, and Hake, 1966.
These studies support a nuanced picture: punishment may suppress behavior under specified conditions, yet aversive control can also alter avoidance, response allocation, emotional behavior, and other interactions. The specific consequences depend on history and procedure. Neither unconditional efficacy nor unconditional inefficacy follows from the name of the procedure.
Reinforcement, stimulus control, extinction, and behavioral alternatives
A response maintained by reinforcement may decline when the previously contingent reinforcer is withheld; in operant analysis this is extinction. Withdrawal of reinforcement during extinction and removal of access contingent on a response are analytically different arrangements. They can have overlapping practical appearances but distinct procedural definitions. Extinction can produce transient increases in responding and context-dependent return; it does not simply erase an acquired learning history. Researchers must examine the original maintaining variables, the conditions under which extinction operates, and the circumstances in which responding recurs.
Differential reinforcement strengthens a selected alternative, incompatible, or otherwise specified class of behavior while arranging the environment so that the targeted behavior receives less support. This offers a different practical strategy from merely trying to suppress the target. For example, teaching an effective request for assistance may alter a difficult behavior maintained by task escape or attention when the request reliably produces appropriate support. Whether the intervention helps a particular person depends on functional assessment, accessibility of the alternative, and the consequences actually delivered. Richman et al., 2015; Ghaemmaghami et al., 2021.
A systematic review of differential reinforcement in skill acquisition published in 2024 summarized a small and heterogeneous group of studies, finding generally favorable results for reinforcing more independent responses in the studied populations. Such findings support selective contemporary applications rather than a sweeping claim about every classroom, disability group, or life situation. The quality of a behavioral intervention includes acquisition, maintenance, generalization, acceptability, and the individual's experienced outcomes. Cividini-Motta et al., 2024.
Contemporary research on avoidance: Learning what does not happen
Avoidance makes an important theoretical demand on models of negative reinforcement: a successful response often prevents an event that the investigator consequently never observes. An adequate account may need information about warning signals, learned expectations, reductions in conditioned fear, action–outcome relations, and contextual learning. Reviews of avoidance research describe evidence for several mechanisms and the limits of theories that explain every avoidance response exclusively as the relief of fear. Krypotos et al., 2015.
Contemporary models distinguish defensive reactions from actions that are directed at controlling or preventing outcomes. Some avoidance behaviors are sensitive to the current value of an expected event, while others become relatively automatic after extensive repetition. This distinction helps connect Skinner's historical attention to contingencies with the modern study of goal-directed action and habits. The shared question concerns the influence of consequences; modern models introduce additional explanatory variables and neuroscience methods and should receive their own intellectual attribution.
In anxiety research, avoidance can restrict opportunities for corrective learning, but experiments and treatment research also show that successful regulation and context-appropriate avoidance can be protective. Clinical application requires understanding the person's real safety and the functional circumstances of the behavior, not applying a simplistic rule that all avoidance should be extinguished.
Reinforcement learning and neuroscience: Continuities and differences
Behavioral reinforcement is identified through the consequences of action on later responding. Computational reinforcement learning often models how an agent updates expected values and chooses actions, sometimes through prediction-error signals. Neuroscientific studies examine neural populations, including dopamine systems, that contribute to aspects of learning, motivation, and action selection. These are intersecting research programs with different levels of explanation, measurement, and modeling commitments. Skinner did not originate contemporary temporal-difference algorithms or the modern dopamine reward-prediction-error framework. Amo, 2024; Gershman et al., 2024.
The prediction-error approach addresses discrepancies between expected and actual outcomes. Recent reviews argue that dopamine signaling exhibits complexity involving sensory, motor, belief-state, and planning variables beyond a single simple reward-value signal. Consequently, an observed rise in response frequency is not automatically evidence for a particular neurotransmitter release pattern, and a neural reward signal does not by itself identify which experimental operant contingency controls behavior. The historical functional account and contemporary neural accounts can be tested together without conflating their concepts. Lerner et al., 2021; Gershman et al., 2024.
Ethical significance: From Skinner's critique to contemporary safeguards
Skinner's preference for constructive, reinforcement-based arrangements was embedded in a broad argument about the costs of punitive social control. In clinical and educational practice today, the ethical question is more exacting: what outcome matters to the learner, what evidence establishes the function of the behavior, what less intrusive options are available, and how will dignity, consent or assent, safety, and the person's access to meaningful alternatives be respected? Experimental effectiveness does not settle these questions by itself.
Research on contemporary punishment-related procedures is heterogeneous. A review of 59 studies published from 2014 through 2023 reported increased attention to social validity, generalization, and maintenance in some research practices alongside reduced reliance on physical forms of punishment; it also noted problems, including reduced use of functional assessment in the surveyed literature. These are observations about a selected research corpus, not proof that all contemporary services operate safely. Ferrier and Cengher, 2025.
A 2024 systematic review of stimulus-avoidance assessments examined assessment methods used when severe challenging behavior and restrictive procedures were being considered, emphasizing procedural and quality limitations. A 2025 survey of 534 behavior analysts described reported contemporary use of restrictive and punishment procedures, while also recognizing sustained ethical debate and disability-rights concerns. Such work underscores the importance of direct measurement, independent oversight, person-centered outcomes, and restraint in interpreting intervention efficacy. Hurd et al., 2024; Flowers et al., 2025.
An ethical educational response ordinarily begins by clarifying the learner's needs, improving task accessibility, reinforcing meaningful skills, and reducing environmental sources of distress. For dangerous behavior, clinical decisions require qualified assessment, legal and professional safeguards, and individualized monitoring. Readers should not treat this historical analysis as a protocol for punishment or coercive treatment. Contemporary effective-care research is necessary for clinical questions and cannot be inferred from Skinner's laboratory evidence alone.
Evidence status and methodological limits
The functional distinction between reinforcement and punishment is an established part of experimental behavior analysis, supported by controlled operant demonstrations across settings and species. Specific effects remain conditional on the response class, deprivation or motivation, timing, schedule, alternatives, history, and contextual cues. Classification is strongest when behavior change is observed across appropriate comparisons and replicated, rather than inferred from what the experimenter intended to deliver.
Historically, Skinner's 1938 and 1953 works document his conceptual and experimental program. Ferster and Skinner's schedule research extends that program through joint authorship. Azrin and colleagues provided much of the direct experimental analysis of punishment intensity, immediacy, side effects, and context. Reviews of contemporary treatment research address different questions about populations, service settings, outcomes, and ethics. The credibility of an argument depends on using the appropriate kind of evidence for its claim. Skinner Foundation bibliography; Virues-Ortega et al., 2021.
Evidence is limited when broad claims are extrapolated from restricted laboratory organisms to complex human relationships, from short-term behavioral suppression to long-term welfare, or from narrowly selected participants to the public at large. Cognitive theories of avoidance and goal-directed control, neurobiological theories of learning, and social and developmental models introduce factors that functional descriptions alone do not settle. Research into their integration continues. Careful evidence interpretation requires describing exactly what was manipulated, measured, and found.
Frequently asked questions
Did Skinner say negative reinforcement means punishment?
No. In Skinner's account, withdrawing an aversive stimulus can strengthen responding. That is negative reinforcement. Punishment addresses contingencies that suppress or reduce later responding. His 1953 discussion explicitly treated presentation of an aversive event and withdrawal of a positive reinforcer as issues for the analysis of punishment. The terms should be interpreted from their stated functional context. Skinner Foundation, Chapter 12.
Can something be a reward without being a reinforcer?
Yes. In ordinary speech a reward is something intended to be appreciated; scientifically, a reinforcer is identified through its effect on subsequent responding. A gift, compliment, or bonus that fails to increase the target behavior in the measured setting may be called a reward without functioning as a demonstrated reinforcer. Conversely, escape from an aversive event can reinforce behavior even when the outcome is mainly relief.
Did Skinner discover the law of effect or invent every reinforcement schedule?
The law of effect originated with Thorndike. Skinner established his own experimental operant analysis and made major contributions to its methods and conceptual structure. Schedules of Reinforcement (1957) was coauthored with Charles B. Ferster. Later investigators, including Nathan Azrin, generated independent foundational findings on punishment. Domjan, 2026; Skinner Foundation books.
Is punishment always effective, and is it the same as extinction?
Punishment can reduce responding under specified contingencies, with effects depending on its parameters and learning history. Extinction concerns discontinuation of a previously operative reinforcement relation. Different experiments can produce suppression, escape, aggression, and behavioral substitution, so the terms describe different procedures and effects. Clinical and educational acceptability requires separate ethical and outcome assessment. Virues-Ortega et al., 2021; Ferrier and Cengher, 2025.
What is Skinner's continuing contribution to these concepts?
Skinner's enduring contribution is the demand that behavioral concepts be specified as observable, experimentally testable relations between acts, antecedent conditions, and consequences. Modern behavior analysis preserves this methodological clarity while incorporating independent developments in social context, human choice, cognitive modeling, neuroscience, treatment evaluation, and ethical practice.
Scientific conclusion
The essential insight of Skinner's account is that the meaning of a consequence is established by the behavioral relation it enters. Presenting a desirable event can strengthen responding; terminating or preventing an aversive event can also strengthen responding; and appropriately arranged contingencies can reduce responding. These classifications require evidence of changed behavior. They gain explanatory power when a response is studied in its full history of antecedents, alternatives, motivating conditions, schedules, and social consequences. Skinner's historical arguments remain important to experimental psychology, while later work by Ferster, Azrin, Ayllon, and numerous contemporary investigators has expanded and corrected the empirical record. The resulting scientific framework demands accurate attribution and gives practical inquiry an equally demanding ethical obligation.
Related Articles
B. F. Skinner — the central scientific chapter on Skinner's life, experimental analysis, radical behaviorism, writings, and intellectual legacy.
Learning Theories in Education: Behaviorism, Cognitivism, Constructivism, and Sociocultural Learning — the wider educational context in which behavioral reinforcement is compared with other theories of learning.
Behavior Change Techniques: What Research Supports — contemporary evidence and limitations for applied behavior-change methods beyond Skinner's historical theory.
Brain Reward System: Circuits, Reinforcement, and Motivation — complementary neuroscience questions about prediction, incentive learning, and motivated behavior.
B. F. Skinner on Extinction and Resistance to Extinction — Extinction, renewal, and persistence of learned responses.
B. F. Skinner and C. B. Ferster's Schedules of Reinforcement — Ferster and Skinner's 1957 reinforcement-schedule studies.
B. F. Skinner's Operant Conditioning Theory — Operant contingencies and their experimental foundations.
References
Amo, R. (2024). Prediction error in dopamine neurons during associative learning. Neuroscience Research, 199, 12–20. https://doi.org/10.1016/j.neures.2023.07.003
Ayllon, T., and Azrin, N. H. (1966). Punishment as a discriminative stimulus and conditioned reinforcer with humans. Journal of the Experimental Analysis of Behavior, 9(4), 411–419. https://doi.org/10.1901/jeab.1966.9-411
Azrin, N. H., Hake, D. F., Holz, W. C., and Hutchinson, R. R. (1965). Motivational aspects of escape from punishment. Journal of the Experimental Analysis of Behavior, 8(1), 31–44. https://doi.org/10.1901/jeab.1965.8-31
Azrin, N. H., Hutchinson, R. R., and Hake, D. F. (1966). Extinction-induced aggression. Journal of the Experimental Analysis of Behavior, 9(3), 191–204. https://doi.org/10.1901/jeab.1966.9-191
B. F. Skinner Foundation. (2016). Science and Human Behavior, Chapter 12: Punishment, Quote 6. https://www.bfskinner.org/science-and-human-behavior-chapter-12-punishment-quote-6/
B. F. Skinner Foundation. (n.d.). Books. https://www.bfskinner.org/publications/books
B. F. Skinner Foundation. (n.d.). Full bibliography. https://www.bfskinner.org/publications/full-bibliography/
Cividini-Motta, C., Livingston, C., and Efaw, H. (2024). Systematic review of differential reinforcement in skill acquisition. Behavior Analysis in Practice, 17(2), 401–416. https://doi.org/10.1007/s40617-023-00903-z
Domjan, M. (2026). Thorndike's law of effect and its inconsistent description over the years. Journal of the Experimental Analysis of Behavior, 125(1), e70073. https://doi.org/10.1002/jeab.70073
Ferrier, S., and Cengher, M. (2025). Punishment happens, but the state of behavior analysis is changing for the better. Behavioral Interventions. https://doi.org/10.1002/bin.2064
Ferster, C. B., and Skinner, B. F. (1957). Schedules of Reinforcement. Appleton-Century-Crofts. https://www.bfskinner.org/publications/books
Flowers, J., Dawes, J., Lund, E., and Georgio, T. (2025). Use of restrictive and punishment procedures: A survey of behavior analysts. Neurodiversity, 3. https://doi.org/10.1177/27546330251367846
Gershman, S. J., et al. (2024). Explaining dopamine through prediction errors and beyond. Nature Neuroscience. https://doi.org/10.1038/s41593-024-01705-4
Ghaemmaghami, M., Hanley, G. P., and Jessel, J. (2021). Functional communication training: From efficacy to effectiveness. Journal of Applied Behavior Analysis, 54(1), 122–143. https://doi.org/10.1002/jaba.762
Hurd, A. M., Brown, K. R., and Randall, K. R. (2024). Stimulus avoidance assessment: A systematic literature review. Perspectives on Behavior Science, 47(1), 139–166. https://doi.org/10.1007/s40614-024-00398-1
Krypotos, A. M., Effting, M., Kindt, M., and Beckers, T. (2015). Avoidance learning: A review of theoretical models and recent developments. Frontiers in Behavioral Neuroscience, 9, 189. https://doi.org/10.3389/fnbeh.2015.00189
Lerner, T. N., Holloway, A. L., and Seiler, J. L. (2021). Dopamine, updated: Reward prediction error and beyond. Current Opinion in Neurobiology, 67, 123–130. https://doi.org/10.1016/j.conb.2020.10.012
Richman, D. M., Barnard-Brak, L., Grubb, L., Bosch, A., and Abby, L. (2015). Meta-analysis of noncontingent reinforcement effects on problem behavior. Journal of Applied Behavior Analysis, 48(1), 131–152. https://doi.org/10.1002/jaba.189
Skinner, B. F. (1938). The Behavior of Organisms: An Experimental Analysis. Appleton-Century. https://www.bfskinner.org/publications/books
Skinner, B. F. (1953). Science and Human Behavior. Macmillan. https://books.google.com/books/about/Science_and_human_behavior.html?id=f6QZAAAAMAAJ
Virues-Ortega, J., Arias-Higuera, M., Hurtado-Parrado, C., and Iwata, B. A. (2021). Nathan H. Azrin: A case study in research translation in behavior science. Perspectives on Behavior Science, 44, 41–67. https://doi.org/10.1007/s40614-020-00278-4