Discriminated punishment: avoidable and unavoidable shock.

Pigeons pecked a key for food. Sometimes the key light turned red. In red, every peck also gave a mild foot-shock. The shock could be avoided by not pecking. In another condition the same red light meant shock would come no matter what. The bird could not escape it.

The team varied shock intensity across sessions. They counted pecks to see which setup stopped the birds more.

When the birds could avoid shock by stopping, they stopped fast. When shock was coming anyway, they still pecked more. Higher volts cut responding in both cases, but the drop was bigger when the bird had control.

Thomas et al. (1968) ran the same lab the next year. They showed the response that controls the shock is the one most suppressed. This supports the 1967 finding: avoidable shock bites harder.

Macdonald (1973) and Badia et al. (1972) added warning tones. A tone before unavoidable shock made it less nasty. Birds kept working when they knew what was coming. This extends the 1967 point: both contingency and signals matter.

Liberman et al. (1973) pushed further. Birds picked signalled shock even when it was three times stronger. Predictability beat intensity, a direct spin-off from the original discriminated-punishment setup.

For BCBAs the message is clear. If you use punishment, the learner must see a direct link between their action and the consequence. A warning stimulus helps if the consequence is truly unavoidable. Without these pieces, suppression will be weak and you will need higher intensity. Start with clear signals and tight contingency before you raise the magnitude.