Classical conditioning, made famous by Pavlov's dogs, links a new trigger to an existing, involuntary reflex — the trigger comes before the reflex, and the learning is about which stimulus now sets that automatic response off. Operant conditioning, developed and named primarily through B.F. Skinner's work, describes a genuinely different learning process: it shapes voluntary behaviour, not involuntary reflexes, and it works through consequences that follow a behaviour, not a stimulus that precedes it — an organism becomes more or less likely to repeat a given voluntary action based specifically on what happened immediately after that action occurred.
Consequences, not preceding triggers, do the shaping
In operant conditioning, a behaviour that's followed by a rewarding consequence — described formally as reinforcement — becomes more likely to be repeated in the future; a behaviour followed by an unpleasant or punishing consequence becomes less likely to be repeated. Skinner demonstrated this systematically using an apparatus, now commonly called a Skinner box, in which an animal's voluntary action, like pressing a lever, could be reliably reinforced with a reward such as food, reliably increasing how often the animal performed that action going forward. Crucially, the lever-pressing itself is a voluntary behaviour the animal chooses to perform, not an automatic reflex triggered involuntarily by some preceding stimulus the way classical conditioning's responses are — the learning in operant conditioning is entirely about how consequences following a voluntary action reshape how likely that action is to happen again.
Reinforcement and punishment can each work in two different ways
Operant conditioning distinguishes reinforcement, which increases a behaviour's future likelihood, from punishment, which decreases it, and further distinguishes positive forms (adding something to the situation) from negative forms (removing something from it) — positive reinforcement adds a reward following a behaviour, negative reinforcement removes something unpleasant following a behaviour, positive punishment adds something unpleasant following a behaviour, and negative punishment removes something desirable following a behaviour. All four combinations reliably shape how likely a given voluntary behaviour is to recur, but they work through genuinely different mechanisms and often produce meaningfully different side effects on an organism's broader emotional state and behaviour, which is part of why operant conditioning theory treats these four categories as importantly distinct tools, not interchangeable variations on a single underlying idea.
What we're still unsure about
The basic mechanics of operant conditioning, and the distinction between reinforcement and punishment and their positive and negative forms, are extremely well established, extensively replicated findings in behavioural psychology, forming a foundational part of the field. What remains more genuinely debated, particularly in applied contexts like education and parenting, is exactly which specific combination of reinforcement and punishment produces the best outcomes for a given goal and a given individual over the longer term, since the field's own research shows that punishment-based approaches, while often producing a faster short-term reduction in an unwanted behaviour, can carry meaningfully different longer-term side effects than reinforcement-based approaches — a genuinely practical, still actively studied question, not one basic operant conditioning theory alone settles on its own.
This sits inside Operant Conditioning (Skinner), one of seven topics in Behavioral Psychology, one of four domains in Psychology, one of seventeen subjects the app can quiz you on.