Instinctive drift is the tendency of an animal’s trained behavior to gradually revert toward its innate, species-typical behavior, even when that shift delays or prevents a reward. The term was coined in 1961 by two former students of B.F. Skinner who discovered, through years of commercial animal training, that conditioning has hard biological limits. No matter how well an animal learned a task, its inherited instincts could override that learning over time.
Where the Idea Came From
Keller and Marian Breland worked in B.F. Skinner’s lab at the University of Minnesota during the 1930s and 1940s, helping develop many of the foundational ideas in operant psychology. After leaving academia, they built a career training animals for commercial shows, advertisements, and tourist attractions. They were skilled trainers who understood reinforcement principles deeply. But across species after species, they kept running into the same problem: animals that had been successfully conditioned to perform a behavior would slowly stop doing it correctly and start doing something else entirely.
In 1961, the Brelands published a paper with a pointed title: “The Misbehavior of Organisms” (a deliberate play on Skinner’s landmark book, “The Behavior of Organisms”). In it, they described case after case of trained animals drifting away from learned behaviors and toward instinctive ones. They called this pattern instinctive drift and argued it represented “a clear and utter failure of conditioning theory” as it was understood at the time.
The Classic Experiments
Raccoons and Coins
The Brelands trained a raccoon to pick up coins and drop them into a container, a simple task that should have been straightforward to reinforce. The raccoon learned to pick up the coins easily enough, but then something unexpected happened: it couldn’t seem to let go. It would rub the coin against the inside of the container, pull it back out, and clutch it tightly for several seconds. When given two coins, the raccoon would rub them together repeatedly before (reluctantly) depositing them. This “dipping” and rubbing behavior looked exactly like the way raccoons wash food in the wild. The coins had become, in the raccoon’s behavioral world, something like food items to be handled and cleaned.
Pigs and Piggy Banks
Pigs were trained to pick up large wooden coins and carry them to a piggy bank. Early in training, they performed beautifully, running the coins over and depositing them quickly. But over weeks of practice, the pigs began to slow down. They would drop the coin on the ground, push it with their snouts, pick it up, drop it again, and root at it. This is rooting behavior, the same nose-to-ground foraging action pigs use to find food in soil. The pigs knew the reward was waiting at the piggy bank, yet they couldn’t stop treating the coin like buried food. The behavior got worse, not better, with continued training.
Chickens on the Playing Field
Several chicken experiments showed the same pattern. Chickens trained to stand still on a platform for 12 to 15 seconds would instead begin scratching at the platform, and over 50% developed a strong, persistent scratching pattern that worsened over time. In another setup, chickens conditioned to peck capsules off a slide began grabbing the capsules, dragging them into their cage, and pounding them on the floor. Chickens trained in a “baseball” demonstration became wildly excited when a ball started moving, chasing it around the field and pecking at it in every direction, completely abandoning the trained routine.
Why This Challenged Behaviorism
At the time, the dominant view in psychology was that the laws of learning were universal. A principle called equipotentiality held that any stimulus could be paired with any response in any species, and the same rules of reinforcement would apply. If you rewarded a behavior consistently, the animal would keep doing it. Species differences were considered mostly irrelevant to the mechanics of learning.
Instinctive drift directly contradicted this. The animals weren’t failing to learn. They had learned the task and performed it correctly at first. The problem was that over time, the learned behavior eroded and was replaced by innate food-related behaviors specific to each species. Raccoons washed. Pigs rooted. Chickens scratched and pecked. Each species drifted toward its own evolutionary programming, not toward some random alternative behavior.
What made these findings especially striking was that the instinctive behaviors actually worked against the animals’ interests. The drifting behaviors required more physical effort than the trained response and delayed the food reward, sometimes preventing it entirely. This violated the “law of least effort,” which predicted animals would always gravitate toward the easiest path to reinforcement. As the Brelands put it, the animals were “trapped by strong instinctive behaviors,” and those instincts proved more powerful than the conditioned ones.
Biological Constraints on Learning
Instinctive drift belongs to a broader category now called biological constraints on learning. This framework recognizes that an animal’s evolutionary history shapes what it can and cannot easily learn. Some associations form almost instantly because they align with a species’ natural tendencies (a rat learning to avoid a food that once made it sick, for example). Others resist conditioning because they conflict with innate behavioral patterns.
These constraints don’t mean conditioning fails entirely. They mean it has limits that vary by species and by how closely the trained behavior maps onto the animal’s natural repertoire. A behavior that works with an animal’s instincts can be reinforced rapidly and reliably. A behavior that runs against them may hold up in the short term but erode over time as instinctive patterns reassert themselves.
What This Means for Animal Training
For anyone training animals, instinctive drift is a practical reality, not just a theoretical concept. It means that the best training strategies work with an animal’s natural tendencies rather than against them. A behavior that resembles something the animal already does instinctively will be far more stable than one that conflicts with its hardwired responses, especially when food is involved as a reward.
Trainers who understand instinctive drift know to watch for the early signs: a gradual increase in the time the animal takes to complete a task, or the appearance of small, species-typical behaviors that weren’t part of the original training. These are signals that the trained behavior is beginning to break down. Choosing behaviors and reinforcement strategies that align with the animal’s biology, rather than fighting it, produces more reliable and longer-lasting results.
The concept also applies more broadly to understanding the limits of reinforcement-based learning in any species. Innate biological factors, from simple reflexes like an infant’s suckling response to complex instincts like a spider spinning a web, set boundaries on what conditioning can accomplish. Learning is powerful, but it operates within a framework shaped by millions of years of evolution.