There’s been a great deal of research on appetite and satiation, both on animals and humans. For humans, of course, the motivation is often focused on how we can lose weight. Almost everyone believes they would look better if they could just lose a few pounds. Most of the research has focused on the taste of food and the physical sensation of fullness, and the results—as you might have suspected—have been inconclusive.

There is some evidence that if you leave the remnants of a meal around (used candy wrappers, for example), then people will eat less than if the evidence of the food is discarded immediately. A study on a single amnesic patient (“H.M.”) mentions that he ate a second full dinner only one minute after finishing the first one.

Inspired by this evidence, a team of researchers led by Paul Rozin suspected that perhaps memory is the key (Paul Rozin and Sara Dow, University of Pennsylvania; Morris Moscovitch, University of Toronto; and Suparna Rajaram, SUNY-Stony Brook, “What Causes Humans to Begin and End a Meal? A Role for Memory for What Has Been Eaten, As Evidenced by a Study of Multiple Meal Eating in Amnesic Patients,” Psychological Science, 1998).

The experiment they devised to explore the role of memory was simple. They found two severely amnesic patients who were otherwise normal. “R.H.” was an intelligent man who managed to live alone with his condition for 20 years. Yet without reminders, he could not remember any new information for more than about a minute. After over 12 hours with the experimenters, he was unable to recognize them. “B.R.” was a man of average intelligence who was hospitalized by his amnesia after it occurred in 1992. Even after spending a year in the hospital, he was completely unfamiliar with its environment, and like R.H., could not remember the experimenters after working with them for many hours.

The experimenters began the session as if they were merely interviewing the patient. Then, without any fanfare, they presented a meal that they knew the patient liked and said “here’s lunch.” B.R. ate the entire meal and R.H. consumed part of his. After the meal, patients were asked to rate their hunger level on a scale of 1 (extremely full) to 9 (extremely hungry). Ten minutes later, each patient was offered another meal, and each patient proceeded to eat exactly as before. B.R. finally rejected his third meal, but R.H. consumed that as well. There were no differences in R.H.’s hunger ratings for the duration of the experiment, which was repeated on three separate occasions. While B.R., a small man weighing only 120 pounds, did report becoming slightly more full, he did not reject the second meal, and on two occasions he began his third meal before being stopped by the experimenter.

For comparison, the experiment was repeated with two people who were not amnesic. Not surprisingly, both participants reported being completely full after one meal, and rejected offers for a second meal.

The obvious conclusion is that memory of what we’ve eaten actually accounts for a significant portion of our hunger. Rather than being solely a physical sensation, being “full” is largely a matter of recalling that we’ve eaten a meal appropriate for the occasion. Rozin et al. point out that most people are very reluctant to eat dinner foods like spaghetti or lamb chops for breakfast. The reason we feel full after meals may be similar: it’s what we expect to happen.

One potential objection to the study’s conclusion is that the same condition that led to the three patients’ amnesia may have also impaired their ability to detect fullness. Rozin and his colleagues reason that this is highly unlikely, because each patient had a different condition, each of which affected slightly different areas of the brain. The only brain area which had suffered damage in all patients was the amygdala, which has been shown in animals not to be related to hunger/satiety.

If hunger and satiety are predominantly social phenomena, then this explains why some of the tricks dieters use are sometimes successful. For example, a common piece of dieting advice is to use small plates to “fool” yourself into believing you’ve eaten a lot. Given the results of Rozin et al.’s study, it’s easy to see why this strategy could work.