Behavioral contrast and inhibitory stimulus control as related to extended training.

Pigeons learned a color discrimination. One key color paid off with grain. A second color never paid. The birds pecked for many daily sessions. The team tracked how response rates changed as training piled up.

The goal was to see if behavioral contrast keeps growing. Contrast means the rate on the paid key rises above baseline when the other key is unpaid.

Contrast showed up fast. After that, extra sessions did not push the rate higher. The elevated responding stayed flat for the rest of the study. More training did not make the effect bigger.

In plain words, the jump in pecking reached a ceiling. It neither faded away nor kept climbing.

Reynolds (1966) ran a near-copy of this setup but saw the opposite. In that study, contrast and peak shift both disappeared after long training. The two papers seem to clash. The gap likely comes from how each team defined extended. S stopped once the effect vanished; W kept going and saw stability.

Touchette (1971) adds another layer. That lab ran 64 sessions and still found steep generalization gradients. The birds kept tight stimulus control, matching the flat contrast seen here. Together the papers suggest emotional spikes calm down, but the learned difference remains sharp.

Locurto et al. (1980) zoomed in on local contrast shoulders next to the positive stimulus. Those mini-peaks did fade with more training. Their data refine the current story: broad contrast stays, narrow edge effects wash out.

When you shape a new discrimination, expect an early burst of extra responding to the reinforced cue. Do not assume it will keep growing if you just keep drilling. Once the rate plateaus, move on to the next skill or add a new stimulus. Extended practice maintains the gain without further boost, so use your session time for fresh targets instead of endless repetition.