Thanksgiving is a time when many Americans experience a somewhat catastrophic failure of willpower. With vast volumes of tasty food set before us, we often fail to resist temptation, eating more than we ought or, perhaps, more than we want to.
Or do we? When we serve ourselves more from the turkey plate, aren’t we, somehow, eating exactly the amount of turkey we want to eat? If we didn’t really want to eat it, we wouldn’t.
Looking at our decisions this way suggests that the notion of what we “want” is a bit more complicated than it appears at first. We seem to want to eat the extra helping of turkey at the same time that we don’t want to eat the extra helping. Both can’t be true.
Psychologists and economists have worked their way out of this puzzle using the notion of multiple selves. One self – the short-sighted, easily-tempted-by-gravy self – does indeed want the turkey, while the other – long-sighted, concerned about health and weight –wants to pass on the second helping. These two selves somehow both live in one’s head.
One way to think about willpower, therefore, is to think about the conflict between the short-sighted-self and the long-sighted self. In this conflict, it is possible to tip the scales. To do so, a key point to bear in mind is that the short-sighted-self is, in some sense, local.
Many readers will be familiar with the somewhat famous “marshmallow task” developed by Walter Mischel. Young children were put in a room with a marshmallow and told that if they could resist eating the marshmallow for some time, they would be given a second marshmallow to eat. This pits, using the notion of selves, the short-term self – which wants to eat the marshmallow now – against the long-term self – who would prefer more than one marshmallow.
One finding that came out of this line of research is that children were able to resist eating the marshmallow longer if the marshmallow was hidden from view. This is, in some sense, puzzling, insofar as the choice the child faces hasn’t changed: one now versus one later. Why should the short-term self lose its edge simply because the marshmallow is out of sight.
This result illustrates, roughly, that the short-term self is local. It seems to be affected by the senses: it’s not the availability of the marshmallow in the abstract that puts the short-term self in charge. It’s the concrete sensory experience of the marshmallow, its shape, aroma, and general gooiness.
From an evolutionary perspective, it’s not hard to understand the way that the short-term self is local. We all have the intuition that this is, more or less, the way that most non-human animals interact with their world. When they see an opportunity that might contribute to survival or reproduction, such as the opportunity to consume food, pursue a mate, and so on, they seem to take advantage of it. Indeed, we find it surprising and, often, entertaining, when pets and other animals don’t quickly eat the marshmallow, as it were. Natural selection has shaped organisms’ psychological systems to identify opportunities in their local world, and take advantage of them, lest they disappear. The short-term self is analogous in at least some sense to these familiar systems.
An important additional detail to add is that it’s not only the external environment that matters: the internal environment does to. We expect a very hungry rodent to jump at the chance to eat a morsel of cheese. The same rodent might not bother if it’s full from morsels it has recently eaten. The rodent’s psychology is, then, local both in the sense that what’s going on nearby matters, and what’s going on inside does too.
We’re similar. When we’re very hungry – perhaps before our parents set out the appetizers for Thanksgiving dinner – we might feel as if our short-term self is in charge, ready to leap at the morsel of cheese, no different from our friend above. And there is some justice to this. People’s behavior does change when they’re hungry, as depicted in the Snickers ad campaign that says that “you’re not you when you’re hungry.” In my language here, they might be saying, “You’re not your long-term self when you’re hungry; your short-term self is in charge.”
Not surprisingly, it’s not quite that simple. I recently was lucky enough to work with Jacob Orquin, currently at the University of Aarhus in Denmark, who wanted to know what the scientific literature could tell us about the effect of being hungry on various kinds of decision making. To look at this, he gathered together the results of a large number of studies – 42, as it turned out – that investigated if people made different decisions depending on whether their blood sugar was low – i.e., they were hungry – or normal.
For instance, one branch of studies asks the following question. When people are hungry, is their short-term self in charge when it comes to making decisions about getting something now as opposed to more of that something later: in essence, these studies investigate if being hungry affects performance on the sorts of tasks that resemble the one with the marshmallows I referred to above. Importantly, some measurements of impatience were very much like the marshmallow task: they were all about food. However, some of the studies were about money: do people who are hungry choose impatiently, taking less money sooner than more money later?
Now, some theories in psychology would predict exactly this effect. A set of popular ideas in psychology suggests that when one’s willpower goes down, one becomes impatient on any task that requires willpower.
Compare this, however, to what you might expect from an evolutionary point of view. As we saw above, an evolutionary account might reasonably imply that a hungry mouse would be more impatient when it comes to food; but this effect should be specific to the domain of food. A hungry mouse shouldn’t be more willing than usual to take a risk for, say, access to water, or a potential mate. The mouse’s state should change its decision making specifically to food.
And that is what Orquin and I found. Being hungry does make people more impatient, but only when food is at stake. We found similar patterns of decision making for the other tasks we reviewed.
Now let us return to the Thanksgiving Day table. There, the problem seems to be that people choose to eat even when they are quite far from hungry. The explanation for this probably resides in the fact that these dinners – if they are like the ones that my mother cooks for my family – are delicious. You can think of decisions to eat as depending on, as I’ve said, one’s state, but also on what is available. If you’re full from a nice meal and someone offers you radishes, well, it’s easy to resist. Compare that to the case in which someone presents you with a dense calorie package with very low handling costs: say, a piece of apple pie. Even satiated, one might choose to indulge.
In the end, how much one overeats is likely to depend on what is around. The more tempting the morsels, the harder it is to resist. My goal here has been to explain, rather than advise, and, in the end, the only advice I might offer is to remember that the short-term self is very local. If you want to avoid eating desert, best to let the chef know in advance that your long-term self would rather not give the short-term self the opportunity to be tempted by it at all.