A molecular analysis of choice on concurrent-chains schedules.

Calamari et al. (1987) watched pigeons peck on concurrent-chain schedules. Some schedules linked the two initial keys so both had to finish before either terminal link opened. Others let each key finish on its own.

The team counted every peck to see how the birds timed their responses within each initial link.

When the initial links were tied together, birds started fast and then slowed. When the links were separate, birds pecked more at the end of the initial link than at the start.

In other words, the microscopic pattern of pecks flipped depending on the interdependency rule.

Garcia et al. (1973) showed that, on simple concurrent schedules, overall peck proportions match overall food proportions. The 1987 study zooms in closer and shows that local peck timing can still differ even when molar matching holds.

DeRoma et al. (2004) later tracked rapid choice shifts within sessions and also found that local reinforcers drive quick changes. Both papers agree that moment-to-moment behavior is dynamic, not static.

Lloyd (2002) looked at whether keeping a fixed 10-s delay difference keeps preference constant. It does not. Taken together, these studies tell us that both local contingencies and absolute delays reshape choice, so neither response rate nor delay rules stay fixed.

If you run concurrent schedules in practice, do not assume steady responding across time. Watch for response bursts early or late in a component and check whether your links are interdependent. Shaping or timing prompts may need to shift as the within-component pattern changes.