What it is
Operant conditioning is the process by which the consequences of a behavior determine how likely you are to do that behavior again. Where classical conditioning is about reflexes and pairings, operant conditioning is about voluntary actions and the outcomes they produce. If an action leads to something good, it becomes more frequent. If it leads to something bad, it becomes less frequent. Simple in the abstract, surprisingly deep in practice.
The framework is usually summarized as a two-by-two matrix. Something can be added or removed, and the addition or removal can be pleasant or unpleasant. That gives four kinds of consequence: positive reinforcement (adding something good), negative reinforcement (removing something bad), positive punishment (adding something bad), and negative punishment (removing something good).
The research, carefully
B. F. Skinner (1904–1990) introduced the operant framework in The Behavior of Organisms (1938), building on Edward Thorndike’s earlier work on the “law of effect.” Skinner’s contribution was methodological as well as theoretical: he invented the operant chamber, informally the “Skinner box,” in which an animal could press a lever or peck a key and receive an automated consequence. The rate of responding became the core dependent measure of his lab.
Skinner discovered the power of intermittent reinforcement almost by accident. His food dispensers sometimes jammed, and the rats went on pressing the lever anyway, often more vigorously than before. That observation opened decades of work on schedules of reinforcement — how the timing and pattern of reward changes behavior. Fixed-ratio, variable-ratio, fixed-interval, and variable-interval schedules each produce characteristic response patterns. Variable-ratio— reward after an unpredictable number of responses — produces the highest, steadiest, hardest-to-extinguish rates of behavior ever measured in a laboratory.
This is the result that explains an enormous amount of human life. Slot machines pay out on a variable-ratio schedule. So do social media feeds. So, often, does a partner whose approval is inconsistent. So does a boss whose praise is rare and unpredictable. The behavior around these schedules looks compulsive, because in a specific technical sense, it is.
The four boxes, with human examples
Positive reinforcement is adding something good after a behavior: praise after a presentation, a paycheck after a week of work, a smile after a joke. It’s the most reliable way to increase a behavior you want more of.
Negative reinforcement is removing something bad after a behavior: the headache that goes away after you take the pill, the anxiety that goes away when you check your phone, the tension that drops when you finally send the email you’ve been avoiding. Negative reinforcement is under-recognized and quietly runs a huge amount of adult life. Most avoidance habits are maintained by negative reinforcement — relief of an unpleasant feeling — not by any positive reward.
Positive punishment is adding something bad: a burn when you touch the stove, criticism after a mistake, a fine after a violation. It suppresses behavior in the moment but tends to produce avoidance and resentment rather than durable change.
Negative punishment is removing something good: losing privileges, a partner going cold after an argument, being ignored after you interrupt. It tends to be more effective than positive punishment for shaping behavior, because it doesn’t introduce new aversive stimuli into the situation — it simply withdraws the reward.
How to recognize it in yourself
The honest question is: what behavior do I do that I claim I don’t want to, and what does it get me? If the answer is “nothing,” look again — behavior that produces nothing tends to extinguish. If a pattern has stuck around for months or years, there is a reinforcer somewhere. Often it’s negative — relief from discomfort. Sometimes it’s social — attention, approval, connection. Sometimes it’s biological — a dopamine hit from novelty.
Once you’ve named the reinforcer, change becomes tractable. You can remove or interrupt the reinforcer, substitute a different behavior that earns the same reinforcer, or accept that some low-cost behaviors are genuinely worth their reward and stop fighting them.
Related patterns
- Previous: classical conditioning.
- Next: reinforcement and reward — the neuroscience under the matrix.
- Anxiously attached people are often on a variable-ratio schedule with their partner — see anxious attachment.
- High-conscientiousness tends to make reinforcement stick faster — conscientiousness.
- The symbolic parallel: the tarot card Seven of Cups — the scattered chase after unreliable rewards.