Kinds of reinforcers: unconditioned, conditioned, and generalized

Every reinforcer has a history, or a lack of one. Sorting reinforcers by where their power comes from tells you how to build new ones, how to keep them working, and why some stop working halfway through a session.

Why it matters

Most of the consequences you use in practice are learned. Praise, tokens, points, and a high five work only because they were connected to other reinforcers in the past. If you understand that connection, you can turn a neutral stimulus into a reinforcer for a learner who responds to very little, and you can avoid letting a token system quietly lose its value.

This task feeds the Concepts and Principles domain. Practice vignettes typically describe how a stimulus came to work and ask you to name the kind of reinforcer. The deciding facts are learning history and the number of backup reinforcers, not how pleasant or popular the item seems.

Key ideas

• Classify by origin, not by appeal. The question is whether the stimulus needed a learning history to function as a reinforcer.
• Unconditioned reinforcers need no learning history. Their effectiveness comes from the species' evolutionary history. Food, water, warmth, oxygen, sexual stimulation, and relief from pain are common examples. Their power still rises and falls with motivating operations such as deprivation and satiation.
• Conditioned reinforcers are learned. A once-neutral stimulus gains reinforcing value by being paired with an unconditioned reinforcer or with an already established conditioned reinforcer. In practice, the neutral stimulus is presented just before or along with the established reinforcer, many times.
• Generalized conditioned reinforcers are paired with many reinforcers. Money, tokens with a menu of backups, and social attention or approval are classic examples. Because they are linked to many backups, their value depends less on any single motivating operation, so they resist satiation better than a reinforcer tied to one item.
• Conditioned value can fade. If a conditioned reinforcer is delivered over and over without its backup reinforcers, it gradually loses effectiveness. Occasional pairing keeps it working.
• The kind of reinforcer and the type of contingency are separate questions. Unconditioned and conditioned reinforcers can both be positive or negative. Ending a painful stimulus is unconditioned negative reinforcement; ending a warning tone that has signaled an aversive event is conditioned negative reinforcement.

Definitions

• Unconditioned reinforcer (primary or unlearned reinforcer): a stimulus change that functions as a reinforcer without any prior learning history.
• Conditioned reinforcer (secondary or learned reinforcer): a previously neutral stimulus change that functions as a reinforcer because it has been paired with one or more other reinforcers.
• Generalized conditioned reinforcer: a conditioned reinforcer that has been paired with many unconditioned and conditioned reinforcers, so its effectiveness does not depend on one specific motivating operation.
• Stimulus-stimulus pairing: presenting a neutral stimulus right before or together with an established reinforcer so the neutral stimulus acquires reinforcing value.
• Backup reinforcer: the item or activity that a token or other conditioned reinforcer can be exchanged for.

Examples from a session

Adults are no different. A job coach's approval, a paycheck, and a supervisor's "That note was perfect" often function as generalized conditioned reinforcers, built over years of being paired with many other reinforcers. As always, you confirm the function by watching whether the behavior they follow increases.

Common mistakes / exam traps

• Equating "unconditioned" with "most powerful" or "natural." Unconditioned describes the absence of a learning history, not strength. A well-built token system can outcompete a snack for a learner who just ate.
• Assuming generalized reinforcers cannot satiate. They are less dependent on any single motivating operation, not immune to all of them.
• Labeling a stimulus a reinforcer before checking behavior. A preferred item, or praise that "should" work, is a reinforcer only if the behavior it follows increases.
• Confusing a conditioned reinforcer with a conditioned stimulus. A conditioned reinforcer is an operant consequence that strengthens the behavior it follows. A conditioned stimulus is a respondent term for an antecedent stimulus that elicits a conditioned response.
• Forgetting to maintain pairing. Praise or tokens delivered for weeks without backups can lose their value, and the plan seems to stop working for no obvious reason.

Quick check

Related topics

• Positive vs. negative reinforcement (B.4)
• Kinds of punishers: unconditioned, conditioned, and generalized (B.8)
• Motivating operations in action (B.16)
• Establishing conditioned reinforcers and token economies (G.4)
• Maintaining change after intervention: schedule thinning and natural reinforcers (G.16)