Some implications of a relational principle of reinforcement.

Atnip (1977) wrote a theory paper. No kids. No rats. Just ideas.

The paper says reinforcement is relational. Value comes from the ratio of response rate to reinforcer rate. Not from the reinforcer itself.

The main idea: a high rate of payoff feels good only if you also respond fast.

Slow payoff can feel fine if you respond even slower. The relation sets the value.

Angle (1970) showed the same thing with data. Pigeons only changed inter-response times when the reinforcer was tied to the prior response. The 1977 paper gave that fact a name: relational.

Nangle et al. (1993) later moved the idea inside the brain. They said neural networks get selected the same relational way. Same rule, new hardware.

Green et al. (1993) added a twist: reinforcers can swap in for each other like Coke for Pepsi. The 1977 rule still works, but now you must check if the items are substitutes first.

Stop asking “How big is the candy?” Ask “How does getting candy compare with how fast the kid works?” If the ratio feels unfair, behavior will drop even with the same candy. Check the relation first, then adjust rate of reinforcement or response requirement.