Multiple punishment schedule.

DeArmond (1966) tested two ways to punish pigeons that were working on a fixed-ratio food schedule. One group got shocked for every peck during the punishment period. The other group got shocked only on the first peck after a red light came on.

The birds still had to finish the ratio to earn grain. The question was: which punishment rule would slow them down the most?

Continuous shock almost stopped the pigeons. They pecked much less and took longer to finish the ratio.

First-peck-only shock did little. Birds ran through the ratio, took the single shock, then kept going. Suppression only showed up after the run ended.

HENDRY et al. (1964) saw the opposite pattern. Their pigeons first paused after punishment, then returned to full speed. DeArmond (1966) shows that when the shock keeps coming, the pause never ends. The two studies differ only in how long the aversive stimulus stays on.

Kelley et al. (2023) revived the same multiple-schedule trick sixty years later. They used signaled periods to thin reinforcement instead of punishment, proving the method still saves clinical time.

Hearst (1960) built the tool D used. He showed that birds can learn separate rules for separate lights. D simply swapped the rules from reinforcement to punishment.

If you punish problem behavior, decide whether you can apply the consequence every time. Continuous punishment kills the response; intermittent punishment often fails. When safety limits force you to punish only once, expect the behavior to reappear after a brief pause. Use extra signals or schedule thinning, like Kelley et al., to keep the effect while reducing the intensity.