Social Reinforcement

Abstract

Social reinforcement is a form of reinforcement — the strengthening of behavior by consequences supplied by other people, such as attention, approval, praise, or affection, rather than by primary rewards like food. Within operant conditioning these social consequences are understood as generalized conditioned reinforcers, acquiring their power by being paired across a lifetime with many primary reinforcers. This article explains why attention and approval function as reinforcers, obeying the same laws as any reinforcer: their effectiveness rises after social deprivation and falls after satiation. It distinguishes informational praise, which tends to support intrinsic motivation, from person-directed praise and tangible rewards, which can undermine it, and follows the concept into digital environments, where quantified peer approval recruits reward circuitry. Three demonstrations let the reader run a reversal design, vary a motivating operation, and compare kinds of praise.

Keywords: social reinforcement, attention, praise, generalized conditioned reinforcer, intrinsic motivation

The distinctive claim of social reinforcement is that a consequence need not satisfy a biological need to strengthen behavior. A smile, a word of praise, a nod of approval, or simply another person's attention can raise the future probability of the response it follows, and for humans these social consequences are among the most pervasive reinforcers of all. Because attention and approval are not themselves biologically primary, their reinforcing power has to be explained rather than assumed, and the operant account is that they are learned reinforcers, built up by a long history of accompanying food, comfort, and other primary rewards (Skinner, 1953; Gewirtz & Baer, 1958).

What Social Reinforcement Is

Social reinforcement is reinforcement in which the reinforcing consequence is the behavior of another person. When a response is followed by attention, a smile, praise, approval, or physical affection, and the response consequently becomes more likely in the future, that consequence is a social reinforcer. The category is defined by the source of the reinforcer, not by any special process: social reinforcement strengthens behavior through the ordinary operant contingency in which a consequence raises the future probability of the response it follows (Skinner, 1953). What sets social reinforcers apart is that they are mediated by another organism whose own behavior — a glance, a word, a touch — is the reinforcing event.

Because a smile or a word of praise meets no biological need directly, social reinforcement cannot rest on the reduction of hunger, thirst, or pain. The operant explanation is developmental: from infancy, the attention and approval of caregivers reliably accompany feeding, warmth, and relief from distress, and through this repeated pairing the social stimuli become conditioned reinforcers. Because they are paired with many different primary reinforcers rather than just one, they become generalized conditioned reinforcers, effective across a broad range of states and situations rather than only when one specific need is active (Skinner, 1953). This is the same logic by which money becomes a generalized reinforcer for humans, and attention is arguably its social counterpart.

Social reinforcement should be distinguished from the related process of observational learning, in which behavior changes not because the observer's own response is reinforced but because the observer watches a model's behavior being reinforced — vicarious reinforcement. Bandura's studies of imitation showed that children readily reproduce behaviors modeled by others, so that social influence on behavior operates through observation as well as through direct social consequences (Bandura, Ross & Ross, 1961). Social reinforcement proper concerns the latter, direct case: a consequence supplied by another person that follows, and strengthens, the learner's own response.

MeSH files the descriptor for social reinforcement directly under reinforcement (as 'Reinforcement, Psychology'), reflecting the taxonomic decision to treat social reinforcement as a subordinate kind of the reinforcement operation. The relation is genuinely one of kind: social reinforcement is reinforcement whose reinforcer is social, so it is a form of reinforcement studied within the same operant framework rather than a rival process.

Attention and Approval as Generalized Reinforcers

The strongest evidence that attention functions as a reinforcer comes from the reversal design of applied behavior analysis. If contingent adult attention is truly reinforcing a child's behavior, then making attention depend on that behavior should raise its rate, and removing the contingency should lower it again. The ABAB design tests exactly this: a baseline phase (A) in which attention is not contingent, a treatment phase (B) in which it follows the target behavior, a return to baseline (A), and a second treatment phase (B). When behavior rises and falls with the contingency across the four phases, the double reversal rules out coincidence and identifies attention as the operative reinforcer (Baer, Wolf & Risley, 1968).

This logic underwrote a large early literature showing that adult attention shapes children's behavior in classrooms and clinics, and it carried an important practical corollary: attention reinforces whatever it follows, including behavior an adult would prefer to reduce. Attending to a child only when the child misbehaves can inadvertently reinforce the misbehavior, which is why behavior-analytic interventions so often work by shifting the contingency — withdrawing attention from the unwanted response and delivering it for an incompatible desirable one (Baer, Wolf & Risley, 1968).

Deprivation and Satiation

If social reinforcers are genuine reinforcers, their effectiveness should vary with the same motivating operations that govern food or water: recent deprivation should heighten their power, and recent satiation should blunt it. Gewirtz and Baer tested this directly. Children who had spent a brief period in social isolation before a task — socially deprived — worked harder for an adult's approval than children who had not, and children given ample adult attention beforehand — socially satiated — worked less for it. Adult attention thus behaves as a reinforcer whose value is set by a motivating operation, exactly as a food reinforcer's value depends on hours of food deprivation (Gewirtz & Baer, 1958; Gewirtz & Baer, 1958).

The finding matters because it distinguishes a true reinforcer from a stimulus that merely elicits a response. A motivating operation changes how hard an organism will work for a consequence, and social reinforcers show this signature clearly, which is one of the cleaner demonstrations that attention and approval are reinforcers in the technical sense rather than loose metaphors (Gewirtz & Baer, 1958).

Praise, Rewards, and Intrinsic Motivation

Not all social reinforcement is alike, and the most consequential modern qualification concerns praise. Praise is the most common deliberate social reinforcer, yet its effect on later, unprompted engagement depends on what it communicates. Praise that conveys information about competence and effort — process praise, which credits how hard the learner worked — tends to support subsequent intrinsic motivation, whereas praise directed at the person, such as calling a child smart, can leave motivation fragile, because it offers no guidance about what to do and makes the child's standing contingent on continued success (Henderlong & Lepper, 2002).

The sharpest version of the concern involves tangible rewards. A large meta-analysis found that expected, tangible rewards offered for engaging in an already-interesting task reliably reduce free-choice engagement once the reward is withdrawn — the undermining effect — whereas verbal reinforcement (praise) generally does not, and can enhance intrinsic motivation (Deci, Koestner & Ryan, 1999). This is not a refutation of reinforcement but a refinement of it: the same delivery that raises behavior in the short run can, if it reframes an intrinsically interesting activity as a means to an external end, lower it in the long run. In practice this is why teacher praise is most effective when it is contingent, specific, and credible — behavior-specific praise that names what the student did well — rather than global and effusive (Brophy, 1981; Royer et al., 2019).

Worked Example

Consider a classroom reversal test of whether teacher attention is reinforcing a child's on-task behavior. In the baseline (A) phase the child is on task about 8 intervals per session while the teacher attends non-contingently. In the treatment (B) phase the teacher delivers brief attention each time the child is on task. If this child finds attention strongly reinforcing, on-task behavior climbs toward an asymptote of about 8 + 34 × 0.7 ≈ 32 intervals per session — roughly a fourfold increase over baseline (32 / 8 = 4.0). Returning to baseline (A) withdraws the contingency and on-task behavior falls back toward 8; reinstating it (B) drives the rate up again. The two reversals, up-down-up, are what license the causal claim that attention, not the passage of time or a change in mood, is the reinforcer.

The same child's responsiveness depends on a motivating operation. Model the reinforcing value of attention as decaying with recent social contact, E = 100 × e^−m/22, where m is minutes of contact just before the session. After a brief deprivation of m = 5 minutes, attention retains E = 100 × e^−5/22 ≈ 80% of its maximum value, and the reversal effect is large. After m = 40 minutes of lavished attention the child is relatively satiated, E = 100 × e^−40/22 ≈ 16%, and the same contingency buys far less behavior — the effectiveness has fallen roughly fivefold (80 / 16 ≈ 4.9). The lesson for practice is that a social reinforcer is not a fixed quantity: its value is set by the recent history of access, exactly as for any reinforcer.

Current Directions

The most active contemporary front for social reinforcement is the digital one. Social media platforms deliver social approval in a quantified, intermittent form — 'Likes,' reactions, follows — and this has let researchers study social reinforcement at a scale and precision the early laboratory could not reach. Neuroimaging of adolescents viewing photographs that had received many versus few Likes showed that quantified peer approval recruits reward-related circuitry, including the nucleus accumbens, and that the number of Likes a photograph carried influenced whether adolescents endorsed it — a peer-influence effect operating through social reinforcement (Sherman et al., 2016). Providing approval to others engages overlapping reward and social-cognition systems, suggesting that both giving and receiving social reinforcement online is intrinsically rewarding (Sherman et al., 2018).

A second front is computational. Treating a Like as a reinforcer, reinforcement-learning models predict the timing of people's social-media posting from the reward history of their prior posts: users post more when recent posts were more rewarded and space their posting as a reward-maximizing agent would, so that engagement follows the same reward-learning equations that describe responding for any reinforcer (Lindström et al., 2021). Together these lines recast a mid-century construct — attention and approval as generalized reinforcers — as a quantitative account of behavior in the attention economy.

Key Researchers

Edward L. Thorndike (1874-1949). Teachers College, Columbia University; his law of effect established that consequences select behavior, the empirical foundation on which the reinforcing power of social consequences such as attention and approval later rested. Wikipedia

B. F. Skinner (1904-1990). Harvard University; he identified attention, approval, and affection as generalized conditioned reinforcers maintained by their pairing with many primary reinforcers, giving social reinforcement its place within the operant analysis of behavior. Wikipedia

Jacob L. Gewirtz (1924-2021). Florida International University; with Baer he showed experimentally that social reinforcers obey motivating operations — brief social deprivation raises, and satiation lowers, the effectiveness of adult attention as a reinforcer for children. Wikidata

Donald M. Baer (1931-2002). University of Kansas; a founder of applied behavior analysis who, with Gewirtz, established the deprivation-satiation dynamics of social reinforcement and set the framework within which social reinforcers are applied in practice. Wikipedia

Richard M. Ryan. Australian Catholic University and University of Rochester; with Deci and Koestner he showed by meta-analysis that tangible rewards can undermine intrinsic motivation while verbal reinforcement tends to enhance it, the boundary that distinguishes social from material reinforcement. ORCID

Patricia M. Greenfield. University of California, Los Angeles; she extends the study of social reinforcement into digital environments, showing with neuroimaging how quantified peer approval recruits reward circuitry and shapes adolescent behavior online. ORCID

Björn Lindström. Vrije Universiteit Amsterdam; he provides a computational reinforcement-learning account of social reinforcement, showing that patterns of social approval on digital platforms follow the same reward-learning equations that govern responding to any reinforcer. ORCID

Commonly Confused With

Reinforcement

Reinforcement is the general operation by which any consequence strengthens the behavior it follows; social reinforcement is the subclass in which that consequence is supplied by another person — attention, approval, praise. The distinction is the source of the reinforcer, not the process: social reinforcers strengthen behavior through the same contingency as food or water, but they are learned, generalized conditioned reinforcers rather than primary ones.

Reinforcement Schedule

A reinforcement schedule is the rule specifying when and how often a reinforcer is delivered; social reinforcement concerns what the reinforcer is. The two are orthogonal: social reinforcers such as attention can be delivered on any schedule, and much of their real-world potency — and the persistence of behavior maintained by intermittent attention — follows from the schedule on which they happen to be delivered.

Discussion

Social reinforcement is the concept that extended operant analysis from the biology of primary rewards to the fabric of everyday human interaction. Thorndike's law of effect established that consequences select behavior (Thorndike, 1927), and Skinner's account of generalized conditioned reinforcers explained how consequences that meet no biological need — a word, a glance, a nod — come to be among the most powerful reinforcers in human life (Skinner, 1953). Gewirtz and Baer's demonstration that these social reinforcers rise and fall with deprivation and satiation gave the idea its experimental spine, showing that attention behaves as a reinforcer in the full technical sense rather than by analogy (Gewirtz & Baer, 1958; Gewirtz & Baer, 1958).

The concept's later history is a story of qualification and reach. Applied behavior analysis turned it into a technology for shaping behavior through contingent attention (Baer, Wolf & Risley, 1968), while the intrinsic-motivation literature established the crucial boundary condition that the manner of social reinforcement, and its confusion with tangible reward, can undermine the very behavior it aims to encourage (Deci, Koestner & Ryan, 1999; Henderlong & Lepper, 2002). The current move into digital environments, where social approval is quantified and delivered intermittently at scale, has returned social reinforcement to the center of behavioral science, now as a quantitative account of engagement in the attention economy (Sherman et al., 2016; Lindström et al., 2021).

Glossary

Approval.

A social consequence — verbal or gestural endorsement by another person — that commonly functions as a generalized conditioned reinforcer.

Attention.

Another person's orienting and responding to one's behavior; among the most pervasive social reinforcers, and one that can inadvertently strengthen unwanted behavior.

Behavior-specific praise.

Praise that names the particular behavior being reinforced (noting that a child lined up quietly); more effective than global praise because it is contingent and informative.

Generalized conditioned reinforcer.

A learned reinforcer paired with many different primary reinforcers, so that it strengthens behavior across a broad range of states; attention and money are examples.

Intrinsic motivation.

Engagement in an activity for its own sake; can be supported by informational praise but undermined by expected tangible rewards.

Motivating operation.

An event, such as deprivation or satiation, that changes the momentary effectiveness of a reinforcer and the frequency of behavior it maintains.

Person praise.

Praise directed at a stable trait (calling a child smart) rather than at effort or strategy; associated with fragile motivation after failure.

Process praise.

Praise directed at effort, strategy, or the process of working (crediting how hard a child worked); tends to support later intrinsic motivation.

Reversal (ABAB) design.

A single-case method that alternates baseline and treatment phases to show that a behavior changes with, and only with, the reinforcement contingency.

Satiation.

Recent, ample access to a reinforcer, which lowers its momentary effectiveness; for social reinforcers, ample recent attention blunts the power of further attention.

Social deprivation.

A period with little social contact, which raises the momentary effectiveness of social reinforcers such as adult attention.

Social reinforcer.

A reinforcing consequence supplied by another person — attention, approval, praise, affection — rather than by a primary reward.

Undermining effect.

The reduction in intrinsic motivation that can follow an expected tangible reward offered for an already-interesting activity.

Vicarious reinforcement.

A change in an observer's behavior that follows from watching a model's behavior be reinforced, rather than from reinforcement of the observer's own response; the mechanism of observational learning, distinct from direct social reinforcement.

Frequently Asked Questions

What is social reinforcement?
Social reinforcement is reinforcement in which the consequence that strengthens a behavior is supplied by another person — attention, approval, praise, or affection — rather than by a primary reward like food. It works through the ordinary operant contingency, differing only in the social source of the reinforcer (Skinner, 1953).

Why do attention and approval work as reinforcers if they meet no biological need?
Because they are generalized conditioned reinforcers. From infancy, the attention and approval of others reliably accompany feeding, warmth, and comfort, and through this repeated pairing they acquire reinforcing power in their own right across a broad range of situations (Skinner, 1953).

How do we know attention is really the reinforcer and not a coincidence?
The reversal (ABAB) design demonstrates it: making attention contingent on a behavior raises the behavior, and withdrawing the contingency lowers it, across repeated phases. The double reversal rules out coincidence and identifies attention as the operative reinforcer (Baer, Wolf & Risley, 1968).

Does the effectiveness of social reinforcement change over time?
Yes. Like any reinforcer, its power depends on a motivating operation. Brief social deprivation raises the effectiveness of adult attention, and recent satiation lowers it, just as food deprivation and satiation change the value of food (Gewirtz & Baer, 1958).

Is praise always beneficial?
No. Praise that conveys information about effort or strategy tends to support intrinsic motivation, but praise directed at the person, such as calling a child smart, can leave motivation fragile after failure. The content of praise, not merely its presence, determines its effect (Henderlong & Lepper, 2002).

Can rewards undermine motivation?
Expected, tangible rewards offered for an already-interesting activity can reduce free-choice engagement once the reward is withdrawn — the undermining effect. Verbal reinforcement, or praise, generally does not have this cost and can enhance intrinsic motivation (Deci, Koestner & Ryan, 1999).

How can teachers use social reinforcement well?
Teacher praise is most effective when it is contingent on a specific behavior, credible, and informative — behavior-specific praise that names what the student did well — rather than global and effusive praise delivered indiscriminately (Brophy, 1981; Royer et al., 2019).

How does social media relate to social reinforcement?
Social media delivers social approval in a quantified, intermittent form. Neuroimaging shows that quantified peer approval such as 'Likes' recruits reward circuitry, and reinforcement-learning models predict posting behavior from the reward history of prior posts, so online engagement follows the same laws as any reinforcer (Sherman et al., 2016; Lindström et al., 2021).

References