Further remarks on the role of cognition in the comparative analysis of behavior.

Nevin (1982) wrote a think-piece, not an experiment. The paper asks when it is okay for a behavior analyst to talk about hidden mental events. It sets one rule: bring in a cognitive idea only if it helps you predict or control behavior better than a pure behavior account can.

There is no new data. The author shows that many cognitive labels fail the test. They do not add useful predictions, so they can be dropped with no loss. The piece ends with a simple filter you can use: if the mental term does not improve your experimental power, leave it out.

The same year, Morris et al. (1982) issued the opposite call: never let cognition in, even if it seems useful. The two papers form an apparent contradiction, but the gap is methodological. Nevin (1982) sets a utility rule; K et al. stick to a purity rule. Both agree that most mental terms should go.

Later work keeps the debate alive. Leslie (2018) extends the question to modern animal-cognition findings. He says behavior analysis should widen its toolkit so complex animal data do not drift to cognitive camps. The stance keeps Nevin (1982)'s test—wider concepts must still earn their keep.

Branch (1977) is an earlier voice on the same path. It singles out memory as a construct that fails the utility test, foreshadowing Nevin (1982)'s broader rule.

Next time you write a program plan, run the Nevin (1982) filter. Swap every mental word for an environmental description. If the plan becomes harder to implement or teach, keep the term and collect data. If nothing changes, drop the fluff and stay behavioral. Your supervision notes, parent goals, and session graphs will be clearer—and you will stay consistent with the science that already works.