Behavioral contrast and the peak shift: effects of extended discrimination training.

Pigeons learned to peck one color for food and ignore another. The training went on for many sessions. Researchers tracked two side effects: contrast (extra pecks to the good color) and peak shift (choosing a color even farther from the bad one).

They wanted to know if these emotional spikes last or fade with more practice.

After long training, both contrast and peak shift disappeared. The birds settled into steady, accurate choices. The study calls these effects temporary emotional reactions, not lasting learning changes.

Reynolds (1968) set the stage by showing contrast and peak shift appear when responding to the non-reinforced stimulus is shut down. The current study adds the twist: keep training and those spikes go away.

Touchette (1971) and Dukhayyil et al. (1973) push back. They ran even longer training and still saw steep gradients and peak shifts. The difference: E used 64 sessions and watched generalization slopes, while A varied overtraining length and saw unpredictable peak-shift returns. Method choices, not contradiction, may explain the clash.

On the other hand, Locurto et al. (1980) and Richards (1974) support the fade idea. They show local contrast near the reward boundary dying out with extended practice, echoing the current "it’s temporary" claim.

If a client shows sudden bursts or odd stimulus choices after you introduce discrimination, don’t panic. Keep the program running. The emotional fireworks should quiet down as practice continues. Track data for a few more weeks before you tweak stimuli or schedules.