Punishment of autoshaped key-peck responses of pigeons.

Corfield-Sumner et al. (1977) tested how electric shock changes pigeons' key pecks. The birds first learned to peck a key when food appeared. Then the team gave shocks in two ways: only after a peck, or on a fixed clock no matter what the bird did.

They varied shock strength across birds. Each condition ran until pecking stabilized. They counted pecks per minute to see which shock style and strength cut responding the most.

Stronger shocks produced bigger drops in pecking. Response-dependent shocks worked best. When the shock came only after a peck, birds stopped faster and stayed quiet longer.

Low-voltage shocks let pecking bounce back. High-voltage, response-dependent shocks kept rates near zero for the whole session.

The result lines up with Sailor (1971), a direct replication that also mapped shock intensity in pigeons. Both papers show the same downward curve: more volts, fewer pecks.

SHETTLEWORTCHARNEY et al. (1965) had already shown that contingent shock beats non-contingent shock. K et al. confirm this rule under autoshaping, so the older finding still holds.

Dunham et al. (1969) seems to disagree. They found that shocking key pecks made off-key pecks go up. The studies differ because they measured different responses. K et al. counted only the target key; J et al. watched nearby keys and saw contrast effects. Both can be true: punished topographies drop, adjacent ones may rise.

For BCBAs, the paper is a lab reminder that contingency and intensity drive punishment effects. If you use mild reprimands or delayed consequences, don't expect lasting suppression. Save punishment for high-risk behaviors, deliver it right after the response, and monitor for contrast: other behaviors might spike when the target one falls.