Field Guide · Glossary
Reinforcement history
Learning scienceThe accumulated record of how often, and how well, a behavior has been rewarded in the past.
How it works
A behavior with a long, rich reinforcement history holds up under pressure, because the dog has been paid for it many times in many places. A behavior with a short or thin history breaks down easily, even if the dog appears to know it. The idea explains several things that otherwise look like stubbornness. An unwanted behavior that has been rewarded for years, such as pulling toward things on a leash, competes with a new behavior that has been rewarded for a few weeks, and the older history usually wins at first. It also explains why behaviors that work occasionally are so persistent, and why the places a behavior has been practiced matter as much as how often.
What it looks like
A dog has been rewarded for checking in with their handler several hundred times on quiet walks and a handful of times near other dogs. On a quiet street the check-in is fast and reliable. Near another dog it is slow or absent, because the history in that context is thin.
Commonly confused with
- Rate of reinforcement
- Rate of reinforcement is how often reward arrives during a given session. Reinforcement history is the total built up over all past sessions and situations, which is what a high rate today eventually adds to.
Related terms