The allocation of time to temporally defined behaviors: responding during stimulus generalization.

The researchers looked at how animals spread their responding across time.

They first taught pigeons to peck for food using several fixed durations.

Each duration was tied to a different colored light.

Later they tested the birds under new, in-between lights.

They recorded how long each peck lasted, not just how many pecks occurred.

The birds did not smoothly shift their peck length.

Instead, they jumped between the exact durations that had been paid earlier.

The smooth-looking gradient on a graph was really a mix of separate response classes.

Time, like color or pitch, is handled as discrete chunks.

Neuringer et al. (1967) saw smooth gradients in rats, but they only counted how fast the rats responded.

When you look at speed, the curve looks continuous.

When you look at duration, the same curve breaks into steps.

The two studies do not clash; they just aimed the ruler at different parts of behavior.

Brown et al. (1972) tried to get pigeons to match tone pitch with continuous key pressure.

The birds failed without extra training in the middle.

Crowley (1979) now explains why: animals store separate duration rules, not a sliding scale.

Together, the papers show that "generalization" is often a pile of tiny, well-learned habits.

If a client takes longer to start chores when you fade the prompt even slightly, do not assume the prompt is too different.

Check whether you accidentally reinforced two separate latencies during training.

Teach and probe one clear duration at a time, then chain them.

Your data will look cleaner and your learner will be less confused.