Divided control by past behavior, present stimuli, and future outcome value in a concurrent‐chains procedure

Gomes-Ng et al. (2026) let pigeons pick between two keys in a chained schedule. Each key led to a final link that gave either a small quick grain or a larger delayed grain.

The twist: the birds' own past hop to the key left a brief light on that key. That self-made cue had to compete with the steady key colors already there.

The birds mostly followed the light they had turned on by their own earlier hop. The steady outside colors mattered, but far less.

When the outside color pointed to the small quick grain, it pulled stronger choice than when it pointed to the big delayed grain.

Ramer et al. (1977) first showed that both stimulus-reinforcer and response-reinforcer links steer pecking. The new study keeps that joint control idea but adds that a cue the bird itself produces can win the tug-of-war.

Rutherford et al. (2003) used the same two-key chain method with token lights. They found choice tracks amount and delay. Gomes-Ng et al. echo that, then show the bird's own entry cue can outweigh those same amount/delay signals.

Donahoe et al. (2000) saw stimulus control fade fast while conditioned reinforcement stayed. That seems to clash with the new finding that a self-made cue dominates. The gap is timing: W tested memory for outside cues after seconds of delay, while Gomes-Ng tested the cue the bird just created. The two papers together tell us outside cues fade, but freshly made cues stick.

If you want a client to pick the better long-term option, first give them a brief signal tied to their own approach response. A colored sticker they place, a button they press, or a card they tap can out-shout the room's static cues. Keep the outside visuals simple and save the salient cue for the action you want repeated.