Resurgence of integrated behavioral units.

Gustavo's team worked with 12 rats. Each rat first learned a four-step lever routine: press left, then right, then left, then right. Food dropped only when the whole chain finished.

Next they removed the food. No reward came no matter what the rat did. The scientists watched to see if the full chain would pop back up later.

Every rat brought the whole four-press routine back. The sequence returned even though it had not been rewarded for a long time.

The result shows resurgence can revive entire behavioral chains, not just single button presses.

Thrailkill et al. (2018) later showed more training rewards cause bigger relapse later. Their pigeons and one child had stronger spontaneous recovery when the original teaching was rich. Gustavo's rats echo the warning: the better you reinforce a skill, the more it can surge back during extinction.

Stancliffe et al. (2007) argued that aggression during extinction is really old behavior returning. Gustavo gives a live demo—whole lever chains reappear—supporting the idea that 'side effects' are often just recovered responses.

Rogers-Warren et al. (1976) found that reinforcing a clear alternative response beats DRO or plain extinction. If you must put a behavior on hold, pair extinction with a strong replacement skill to cut the chance of future resurgence.

When you fade reinforcement, think in sequences, not single responses. A client who stops a three-step hand-washing chain may suddenly run the full routine tomorrow. Build sturdy replacement chains and plan booster sessions so the old one does not sneak back.