Responding maintained under intermittent schedules of electric-shock presentation: "Safety" or schedule effects?
Electric shock can reinforce operant behavior when it follows an intermittent schedule, so schedule design matters more than the pleasantness of the consequence.
01Research in Context
What this study did
Kendrick et al. (1981) asked a simple question. Can electric shock keep a monkey pressing a lever? They ran squirrel monkeys on several intermittent shock schedules. The shocks were unavoidable and never cancelled by the lever press.
The team varied the timing pattern. Some schedules gave shocks every few minutes. Others used random gaps. They wanted to see if the lever press would keep going even when it did not produce safety.
What they found
The monkeys kept pressing. Response rates stayed steady across every shock schedule. The pattern looked like food-reinforced behavior, not like frightened avoidance.
Because shocks were unavoidable, the results challenged the old idea that animals press to feel safe. Instead, the schedule itself seemed to reinforce the lever press.
How this fits with other research
Henton (1972) showed the same thing earlier. Monkeys pressed under VI shock and stopped under FR-1 shock. The new study widens the test and rules out safety explanations more cleanly.
KELLEHER et al. (1963) saw high pressing before unavoidable shocks. Their preshock stimulus may have acted like a schedule cue. F et al. remove the cue and still get steady pressing, proving the schedule alone can maintain behavior.
James et al. (1981) ran second-order sequences with shock that year. Both labs find shock can reinforce just like food. Together they bury the claim that shock only punishes.
Why it matters
If you use punishment or sensory consequences, remember: the delivery schedule can turn even an aversive event into a reinforcer. Watch for accidental reinforcement when you space warnings, reprimands, or sensory feedback. Track response rate, not just topography. A steady rate under intermittent aversive input may signal reinforcement, not tolerance.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Graph response rate when you give intermittent warnings or reprimands; a flat or rising line may mean the schedule is reinforcing the behavior you want to stop.
02At a glance
03Original abstract
Four experiments were conducted in which lever pressing by squirrel monkeys was maintained under multiple, mixed, or chained schedules of electric-shock presentation. In the first two experiments, a multiple schedule was employed in which a fixed-interval schedule of shock presentation alternated with a signaled two-minute component. Initially, no events were scheduled during the two-minute component (a safety period). In the first experiment, the safety period was "degraded" by introducing and systematically increasing the frequency of periodic shocks presented during that component. In the second experiment, the proportion of overall safe time to unsafe time was decreased by decreasing the value of the fixed-interval schedule while holding constant shock frequency during the two-minute component. In the third experiment, the overall arrangement was changed from a multiple to a mixed schedule in an attempt to determine whether fixed-interval responding would be maintained when a single exteroceptive stimulus was associated with both components. In the fourth experiment, the overall arrangement was changed from a multiple to a chained schedule in an effort to determine whether fixed-interval responding would be maintained when its consequence was presentation of a signaled "unsafe" period. Fixed-interval responding was well maintained under all experimental conditions; the varied relationships obtained lend more support to conceptualizations of shock-maintained behavior as exemplifying schedule-controlled behavior than to suggestions that such behavior may be readily accounted for by "safety theory."
Journal of the experimental analysis of behavior, 1981 · doi:10.1901/jeab.1981.36-171