Concurrent schedules: discriminating reinforcer-ratio reversals at a fixed time after the previous reinforcer.

Researchers watched pigeons peck two keys. Each key paid off on its own variable-interval schedule. After every food delivery, the payoff ratio flipped. The team tracked which key the bird hit next and how that choice changed as time passed.

They wanted to see if the birds would switch keys right away or ease into the new ratio.

Choice did not jump the instant the ratio reversed. Instead, it drifted downhill like a ball on a gentle slope. The birds’ next few pecks still followed the old ratio, then slowly bent toward the new one.

A math model fit the curve by spreading the old and new ratios across the seconds since the last piece of food.

Hastings et al. (2001) ran a similar two-key pigeon setup and saw clean matching to the new ratio almost at once. The difference: they watched long-run totals, not the first pecks after food. The slow drift only shows up when you zoom in on the local moment.

Davison et al. (2010) added flashing lights that told the birds which key would pay next. With those cues, choice flipped faster. Take the cues away and the gradual curve returns—same birds, same box, just less information.

Baer (1974) framed reinforcement as a ‘situation change.’ That idea fits the new data: the food delivery marks a new ‘situation,’ but the bird needs a few seconds to feel the new payoff winds.

If you run concurrent schedules in the clinic, do not expect an immediate jump when you flip the token chart. Watch the first few responses after each reinforcer; they still echo the old rate. Give the learner a short window—extra prompts, richer schedule, or salient cues—to help the shift happen faster. The matching law still rules, but it rides a gentle slope, not a cliff.