Scientists warn: awards are harmful

Scientists have discovered that the expectation of a reward for work prevents the brain from adequately approaching learning and remembering information.

The study of certain abilities of animals is almost always associated with some kind of treat – a kind of reward for efforts and work. For example, rodents that demonstrate their skills are subsequently given a sweet or other delicacy. Some animals themselves “get” a treat if they are resourceful and fulfill the requirements of the researchers.

However, as noted by experts from Johns Hopkins University and New York University, the expectation of a reward has a direct impact on the behavior of animals: this means that the subjects behave completely differently if they know that there will be no reward. What does this look like? For example, a rodent crawled to a special window in which there was an outlet with water; after the sound of pouring water was reproduced, the animal began to lick the outlet, trying to drink. If another sound followed, the animals were confused, reacting to both reproductions in different ways, in random order. In this case, the probability of a logical hit was 50%. In order to remember which of the sounds means the flow of water, the animal needs a certain amount of time. However, when the water outlet was removed on the first day of the experiment, the rodents suddenly "switched on" their intelligence and showed a reaction to the necessary sound with a 90% hit rate. Thus, although there was no water output, they tried to "lap" as if they were drinking water, and this reaction followed precisely in those cases when it was necessary.

The study was repeated in two variations and with other animals. For example, rodents were asked to press a button to get water – and again they showed a better response when no reward was provided. Or other rodents were trained to look into a food bowl after a certain sound – but here food only appeared when there was no flash of light before the sound. In addition to rodents, experiments were conducted on ferrets. And even in these cases, the animals were trained more productively when no reward was involved.

According to the authors of the study, the learning process consists of two mechanisms: first, animals learn certain content, information (for example, a behavioral rule - to press a button after a sound signal), and the information context, which includes the subsequent reward, is also learned. At the same time, the expectation of a treat in many cases prevents the brain from showing its already acquired knowledge. In this situation, we have to look at a less savvy rodent, and a more intelligent one, although in fact there is a completely different difference between them: the degree of sensitivity to the expected reward.

Since the experiments were conducted with the participation of different animals, we can assume the presence of a cognitive pattern that is also applied to people. Of course, we would like to obtain the results of research on humans too, using different types of rewards (material, non-material, etc.). However, scientists are already advising everyone to think less about any kind of reward - only in this case can we be sure of high mental efficiency.

Details of the experimental project are described at www.nature.com/articles/s41467-019-10089-0