The reason a habit survives the loss of its reward

A laboratory rat learns to press a lever for a sugar pellet. After several hundred repetitions the pellet is removed. The rat keeps pressing. It presses long after the pellet would have been consumed, long after a new rat placed in the same cage and given the same removal would never press at all. The lever has outlived its reason.

What everyone sees

Observers interpret persistent behaviour as either confusion or hope: the rat has not noticed the pellet is gone, or it expects the pellet to return. Both explanations assume the behaviour is still goal-directed. What they do not consider is that the behaviour has changed category — from instrumental (press to receive) to automatic (press because pressing is what happens here).

What is actually happening

Research on habit formation distinguishes between goal-directed and habitual action. Dickinson, Wood and Quinn, and others have shown that extensive repetition shifts behaviour from the goal-directed system (sensitive to outcome) to the habitual system (triggered by context, insensitive to outcome). Adams demonstrated that rats trained extensively still performed the lever press even after the food was devalued, whereas rats with less training stopped immediately. The habit had become stimulus-bound: the cue triggered the response without consulting the goal.

Why it stays hidden

The transition hides because the behaviour looks the same from the outside. A goal-directed lever press and a habitual lever press produce the same motion, so the observer — and often the performer — cannot tell which system is in control without running a devaluation test. In everyday life, people drive the same route, check the same app, and reach for the same snack long after the original reason has expired, and because the motion is familiar it feels chosen rather than automatic.