Reading focus
Syllabus: Why Does Learned Behavior Persist—and When Does It Change?
Add a measurement question to the review's map: how quickly does responding decline relative to its baseline, and under what schedule? Notice where an apparently simple reinforcement rule depends on the comparison being made.
A question to consider. Two training schedules produce different initial response rates. What would you plot before deciding which produces more persistent behavior?
Why this reading
A frequently performed response is not necessarily a persistent one. Nevin separates response rate from resistance to disruption and examines how reinforcement history, the measurement of decline, and the arrangement of schedules affect that distinction.
Before you read
Extinction suspends the contingency that previously produced reinforcement. Resistance is assessed relative to responding before that change. A multiple schedule alternates components signaled by different stimuli; a single schedule does not provide that within-session contrast. Behavioral momentum is a theoretical analogy, not a claim that a response has physical mass. Follow what each term in the equations represents; the argument does not require reproducing every calculation.
Read for
- Compare absolute response rates with proportions of baseline. What changes when persistence is measured against elapsed time, trials, or omitted reinforcers?
- Identify the problem partial reinforcement poses for a simple strengthening account. Track the additional assumptions used to accommodate it.
- Examine the differences between single and multiple schedules. Ask when a fitted parameter explains variation and when it leaves the source of that variation unresolved.
Consider afterwards
- One response falls from 100 to 40 occurrences per minute; another falls from 20 to 15. Which is more resistant to change, and what further information would you need before comparing their learning histories?
Editorial context
The account combines strengthening by reinforcement with disruption produced by its removal. The article itself identifies limits: schedule arrangements can yield opposite relations between reinforcement rate and decline, and fitting different data requires different parameters. It does not support the general rule that more frequent rewards always produce greater persistence. Shahan and Craig (2017) later challenge the momentum account's extension to resurgence and develop a choice-based alternative; that dispute concerns a particular explanatory extension, rather than invalidating every finding about resistance to change.