Instrumental Conditioning and Reinforcement Learning
Instrumental conditioning links a behavior to its consequence.
Start with the Contingency Question
Instrumental conditioning links a behavior to its consequence. The central question is not merely whether a consequence follows a behavior. It is whether what the animal does determines whether a reinforcing stimulus is delivered. If the answer is yes, the behavior is playing an instrumental role.
The defining idea is contingency: reinforcement depends on the animal's behavior rather than occurring independently of it.
Trace Behavior to Consequence
To analyze an instrumental-conditioning situation, trace the sequence in two parts. First, identify the behavior. Next, ask whether the consequence is delivered because that behavior occurred. The consequence is not simply occurring on its own; its delivery depends on the animal's action.
A Hypothetical Contingent Consequence
Suppose an animal performs a behavior, and a reinforcing stimulus is delivered only when that behavior occurs. Does this arrangement satisfy the defining criterion for instrumental conditioning?
Identify the behavior: Name the action performed by the animal. In this generated example, call it Behavior A.
Check the delivery rule: Ask whether the reinforcing stimulus is delivered when Behavior A occurs, rather than being delivered independently of what the animal does.
Determine contingency: Because the animal's behavior determines whether the reinforcing stimulus is delivered, reinforcement is contingent on behavior.
Yes. The arrangement illustrates instrumental conditioning because it links the behavior to its consequence through a behavior-dependent delivery rule.
What Reinforcement Changes
Reinforcement learning can be understood here as a behavior-consequence cycle. An animal performs a behavior, a consequence is delivered according to whether that behavior occurred, and the experience provides the basis for considering whether the behavior is likely to be repeated. The important learning relation is the link between the behavior and its consequence.
Instrumental and Classical Conditioning
| Question | Instrumental conditioning | Classical conditioning |
|---|---|---|
| What determines delivery of the reinforcing stimulus? | The animal's behavior | The reinforcing stimulus is independent of behavior |
| What is the defining relation? | A behavior is linked to its consequence | The consequence does not depend on what the animal does |
| What should you ask? | Does what the animal does determine delivery? | Is delivery independent of behavior? |
A consequence can be reinforcing in both discussions, but the defining difference here is whether its delivery depends on the animal's behavior.
Two Names for One Tradition
Operant conditioning is usually treated as another term for instrumental conditioning. In this context, the two terms refer to the same broad behavior-consequence relation: reinforcement is contingent on what the animal does.
Thorndike and the Earlier History
The roots of instrumental conditioning go back to experiments performed by the American psychologist Edward Thorndike. These experiments and Thorndike's law of effect belong to the earlier history of the field, about one hundred years before the first edition of the source book was published.
The historical connection places instrumental conditioning in a longer research tradition. It does not replace the defining criterion. Whether a situation qualifies still depends on whether reinforcement is contingent on behavior.
Mistakes in Identifying Contingency
Assuming that any consequence following a behavior is instrumental conditioning.
A following consequence is not enough. Instrumental conditioning requires reinforcement to be contingent on behavior.
Fix:
Ask whether the animal's behavior determines whether the reinforcing stimulus is delivered.Treating instrumental conditioning and classical conditioning as identical.
The source distinguishes them by the role of behavior in determining delivery.
Fix:
For instrumental conditioning, behavior determines delivery; in the stated contrast with classical conditioning, the reinforcing stimulus is independent of behavior.Treating operant conditioning as a completely separate process.
Operant conditioning is usually treated as another term for instrumental conditioning.
Fix:
Recognize the two terms as overlapping labels for the same broad relation.Using Thorndike's history as the definition.
Thorndike provides historical context, but the defining criterion is still contingency.
Fix:
Use the contingency question first, then use Thorndike to place the concept historically.
Apply the Contingency Test
For each situation, decide whether it satisfies the defining criterion for instrumental conditioning. Explain your answer by stating whether the animal's behavior determines delivery of the reinforcing stimulus.
Hints
- Do not begin by asking whether a consequence exists.
- First identify what the animal does.
- Then ask whether the reinforcing stimulus is delivered because of that behavior or independently of it.
Checking Your Reasoning
A situation includes a behavior and a reinforcing stimulus. The stimulus is delivered only when the animal performs the behavior. What conclusion follows?
Locate the behavior: There is an action performed by the animal.
Check dependence: The reinforcing stimulus is delivered only when that action occurs, so delivery depends on the behavior.
Classify the relation: The behavior is linked to its consequence through contingent reinforcement.
The situation meets the defining criterion for instrumental conditioning. It can also be described using the term operant conditioning.
Takeaway
- Instrumental conditioning links a behavior to its consequence.
- Reinforcement is contingent when the animal's behavior determines whether the reinforcing stimulus is delivered.
- The key contrast with classical conditioning is that, in the stated classical-conditioning contrast, the reinforcing stimulus is independent of behavior.
- Operant conditioning is usually treated as another term for instrumental conditioning.
- Edward Thorndike's experiments and law of effect are part of the earlier history of instrumental conditioning, but contingency remains its defining criterion.
Key Takeaways
- Instrumental conditioning is a behavior-consequence relation.
- The essential test is whether behavior determines delivery of reinforcement.
- This differs from the stated classical-conditioning contrast, in which the reinforcing stimulus is independent of behavior.
- Operant conditioning is usually another term for instrumental conditioning.
- Thorndike's experiments and law of effect provide an important historical connection.