STIMULUS ASPECTS OF AVERSIVE CONTROLS: THE EFFECTS OF RESPONSE CONTINGENT SHOCK.
Shock delivered only for the target response stops behavior faster but also returns faster than shock that is not tied to the response.
01Research in Context
What this study did
SHETTLEWORTCHARNEY et al. (1965) worked with pigeons that pecked a key for food.
They gave two kinds of shock during a tone. One group got shock only if they pecked. The other group got shock no matter what they did.
The team watched how fast each group stopped pecking and how fast the behavior came back when shock ended.
What they found
Birds that had to peck to get shocked stopped pecking almost right away.
When the shock ended, their key-pecking also came back more slowly.
Birds that got shock no matter what kept pecking a little during the tone and bounced back faster later.
How this fits with other research
HAKMCMILLAN et al. (1965) ran the same year with the same birds and setup. They showed that even a light that had been paired with shock could punish pecking. Together the two papers prove that the contingency, not just the pain, does the work.
Last et al. (1984) later flipped the story. They gave shock for long pauses and saw responding speed up. This seems opposite, but the difference is timing: shock for a pause makes short, fast bursts pay off. The 1965 study shocked every peck; the 1984 study shocked the lack of pecks.
Rachlin (1966) used the same tone-shock cue but showed the tone could later speed up pecking if it signaled time-out instead of shock. This extends the 1965 finding: the same stimulus can suppress or accelerate depending on what it predicts.
Why it matters
For BCBAs, the lesson is clear: punishment works best when it is tied to the exact response you want to stop. Non-contingent aversives, like random reprimands, give weaker and less lasting effects. If you must use punishment, deliver it right after the target behavior and be ready for the behavior to return quickly once you stop. Better yet, pair brief punishment with reinforcement for good behavior to lock in safer, lasting change.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →If you use a punisher, deliver it right after the problem response every time, and track how fast the behavior comes back when you fade it out.
02At a glance
03Original abstract
A tone ending with electrical shock was periodically presented to pigeons while they pecked a key for food. Pairs of birds were run simultaneously under a yoked program which insured that both birds received the same number and temporal distribution of shocks. For one of the birds, shock was always initiated by a peck; for the other, shock was unavoidable. Both procedures led to reduced rates of pecking in the presence of the tone, and gradients of stimulus generalization were obtained. But the effects of response contingent shock extinguished more rapidly than the effects of unavoidable shock. In general, birds exposed to unavoidable shock tended to respond at intermediate rates throughout tone, whereas those exposed to response contingent shock ceased to peck for part or all of the tone period.
Journal of the experimental analysis of behavior, 1965 · doi:10.1901/jeab.1965.8-89