Knowing before doing: discrimination by rats of a brief interruption of a tone.
Latency drops before the first reinforced response, giving an early signal that discrimination has occurred.
01Research in Context
What this study did
Rats listened to a steady tone. Sometimes the tone skipped for half a second.
After the skip, a lever press produced food. The skip was the only cue.
The team measured how fast the rats pressed once the skip happened.
What they found
Before the rats pressed the lever, their pause time changed.
Shorter pauses showed up only after the skip, not during steady tone.
The rats "knew" food was ready before they actually moved.
How this fits with other research
Aragona et al. (1975) saw contrast effects when milk replaced pellets. Their rats also shifted pause times, hinting that latency hides learning.
Muller et al. (2016) later found rats chunk serial patterns using silent rules. Both papers show rats think before they act.
Galizio et al. (2023) taught rats to reverse olfactory cues fast. Quick reversals need the same pre-response discrimination Eisler (1984) caught with tones.
Why it matters
Watch latency, not just the final response. A child may sit still for two extra seconds before touching a picture card. That tiny pause can tell you the rule is already learned. Start measuring quiet time in your data sheets. It gives earlier news than correct answers.
Want CEUs on This Topic?
The ABA Clubhouse has 60+ free CEUs — live every Wednesday. Ethics, supervision & clinical topics.
Join Free →Start a stopwatch at the cue onset; stop at the first response—plot that latency across trials to see learning sooner.
02At a glance
03Original abstract
Eight rats' lever presses were reinforced after an interruption in a tone, provided the lever had not been pressed before the tone interruption. After a few sessions, long before the animals reliably refrained from lever pressing before the interruption, the latencies of postinterruption presses (time from the termination of the interruption to the moment of the lever press) dissociated into two classes: short ones for to-be-rewarded presses, and long ones for presses in the other trials, which contained no reward because one or more lever presses had occurred before the interruption. Thus discrimination of impending reinforcement in the current trial occurred before there was evidence of sensitivity to reinforcement in the reinforcement-producing aspect of behavior. This finding is related to Shimp's (1981) contention that the temporal properties of recent behavior are reinforceable, if remembered. The present finding shows that learning to discriminate whether one's behavior has met a contingency, and learning to carry out this behavior, need not go together, implying that memory of temporal properties is probably a necessary but not a sufficient condition for learning the latter.
Journal of the experimental analysis of behavior, 1984 · doi:10.1901/jeab.1984.41-329