Some effects of punishment shock intensity upon discriminative responding.

The team worked with hungry pigeons that pecked a key for food. Every tenth peck also gave a brief electric shock. The scientists raised the shock level across days to see how punished and later unpunished pecking would change.

After punishment training, food was removed. They watched how fast the birds pecked during extinction with no food or shock.

Stronger shocks quickly cut the punished pecks. During the later extinction phase, birds that had felt medium shock pecked the most. Birds that had felt light or very strong shock pecked less, making an upside-down U shape.

The result shows punishment intensity can have odd side-effects: mid-level shock later bumped up unpunished responding.

McKearney (1970) ran a near-copy of the same set-up and also found that higher shock cut responding. The 1971 paper extends that work by showing the extra twist: mid-level shock can later inflate unpunished extinction pecks.

SIDMAELLIOTT et al. (1962) first saw punished pecks drop while unpunished pecks rose. The 1971 data fit that pattern, but only for the mid shock group, giving the first inverted-U curve during extinction.

Kaufman (1965) looks like a contradiction. In that study adult humans given high shock showed big recovery by session two. The pigeons here stayed suppressed. The gap is likely species plus task: people understood instructions and could predict shock, birds could not.

When you use punishment, watch for hidden rebounds. A medium-strength consequence might later make other behaviors pop up faster once the contingency ends. Track unpunished responses, not just the target, and be ready to reinforce safe alternatives if post-punishment bursts appear.