Self-Control is a Muscle

Experiments like this demonstrate why Puritanism is so psychologically unrealistic:

A paper in The Journal of Consumer Research looks at the effects of self-restraint on subsequent impulse purchases.

In one experiment, college students spent a few minutes free-associating and writing down their thoughts, under instructions not to think of a white bear. Given $10 afterward to save or spend on a small assortment of products, they spent much more money than students who had free-associated without having to avoid thoughts of bears.

This isn't the first time people have explored the impact of mental exertion on self-control. Stanford professor Baba Shiv invented an experiment where he manipulated the "cognitive load" of subjects. Shiv gave half of the subjects a two-digit number to memorize (low load), while the other half were given a seven-digit number (high load). Subjects were then instructed to walk to another room in the building. On the way they passed by a table at which they were presented with a choice between a caloric slice of chocolate cake or a bowl of fruit salad. Fifty-nine percent of the people trying to remember seven digits (high load) chose the cake, while sixty-three percent of the two-digit subjects (low load) chose the fruit salad. In other words, having people memorize an extra five digits made them exhibit significantly less self-control.

Why did the number of digits have such a strong effect? Shiv speculates that the effort required to memorize seven numbers drew cognitive resources away from our ability to control our urges. This makes anatomical sense, since working memory and self-control are both located in our pre-frontal cortex. Having to remember seven numbers occupied neurons that would otherwise help us decide what to eat, which causes us to become more reliant on our impulsive emotions. While we tend to think of self-control as being an innate trait, it is actually dependent on a range of extrinsic factors, all of which affect the way our brain responds to a given situation.

This model of limited "thinking resources" has now generated a large amount of supporting evidence. Our decisions really are swayed by the computational limits of our brain. For example, in 2003 neuroeconomists noticed that subjects on diets who resisted temptation in the morning (by foregoing the chance to grab snacks from a nearby basket) later ate significantly more ice cream in an ice-cream taste test than subjects who hadn't exercised self-control. They also quit 40 percent earlier when confronted with a difficult math problem. By resisting the morning snacks, they had temporarily "used up" their ability to resist further temptation. (Other variables that seem to exhaust our self-control are alcohol, stress, and sleep deprivation.)

The moral of this data is that we have to pick our battles. Everybody occasionally splurges on the slice of cake, or quickly gives up on a difficult problem. Instead of trying to never be bad, we should focus on being good when it matters. If you are trying to diet, then let yourself think about white bears. Self-control is a mental muscle, and we should always be conscious of not tiring it out.


More like this

My first reaction was to think of John Milton. Cranking out "Paradise Lost" and "Paradise Regained" didn't seem to inhibit his Puritanism. Nor, come to think of it, did Oliver Cromwell become noticeably more dissolute as a result of winning the English Civil War and having his own king beheaded.

By Tom Welsh (not verified) on 13 Mar 2007 #permalink

What Tom said. What is this Puritanism you're abusing? You probably don't mean the doctrine of the "priesthood of all believers". Are you talking about self-denial or religious rigor in general?

By Vance Maverick (not verified) on 13 Mar 2007 #permalink

I was using puritanism as a shorthand for self-denial and the desire to abstain from all types of worldly sin. It was a poor choice of words. Apologies to all the offended puritans out there.

Somehow, this reminds me of Mae West:

"I generally avoid temptation unless I can't resist it."

"I like restraint, if it doesn't go too far. "

By natural cynic (not verified) on 13 Mar 2007 #permalink

An interesting finding. I've noticed similar effects in myself, but I attributed it to my NLD. It occurs to me that ethanol is famous for both lowering inhibitions, and as a muscle relaxant.

By David Harmon (not verified) on 13 Mar 2007 #permalink

If the findings really reveal a real physiological phenomenon, then it ought to be true that people grappling with tough intellectual/philosophical problems should turn out to be obese or indulgent..!!?

What about students taking exams..?

I think we need a study on this too..

How about starting with Darwin, Einstein, Freud, Newton, Planck, Bohr and Sperry ?
Well, according to their biographies most scientist pro's led relatively disciplined lives ( of course, apart from the hairdo and absent-mindedness of some..!!)
Sure there are evidences for over-eating and limited self control in depression patients, but this new finding of Baba Shiv et al seems a bit indigestible...

I'm curious as to how far the muscle analogy goes. Are there ways "to exercise" one's self-control?

By Thomas Cartwright (not verified) on 15 Mar 2007 #permalink

suraj: your argument doesn't follow. people grappling with tough intellectual/philosophical problems, students, scientists etc. are self-selected and are likely to have higher "stores of self-control" to begin with. you need to hold that constant to assess the effect of cognitive load - either by assigning tasks randomly, or looking at the same people under different loads.

Suraj, Cartwright, conchis: I think the answer to Cartwright's question (personally) is "yes," it can be exercised and built like a muscle. So the alternative to conchis's claim is not just having more (necessarily) as a purely inate ability, but having developed it.

I was thinking why the experimenter concluded only on the correlation between "lack of self control and the cognitive load"..
Of course, there could be other possibilities, like for example the brain seeking out more calories to refill after the "work-out" with tougher cognitive issues. And any experienced brain will naturally choose chocolates over fruit salad as a better source of calories.!
Every time you do a PET scan, you actually are looking at the differential glucose uptake of brain areas; its the same thing, i guess...

Thomas: yes, there are ways to exercise this "mental muscle", as any other muscle. I have seen recently a paper where computer-based working memory training improves Stroop Test scores, for example (Klingberg, 2005). You can check an interview with Torkel Klingberg in our blog.

There are also common-sense methods: to build a muscle, you need to practice it in a variety of contexts for better cross-training and transfer.

wow i want to do you

If that's all there is to this study.... cake or fruit after
a taxing mental task, then maybe it's just measuring the body's response to mental exasustion/stress:
Cake looks better to a tired mind than does fruit.

suraj: "...for example the brain seeking out more calories to refill after the "work-out" with tougher cognitive issues."

I was thinking the same thing. It's not necessarily "using up" the ability to resist temptation, but actually trying to increase glucose levels so as to resist tempation more effectively. The researchers' choice to use cake or ice-cream to assess self-control is rather unfortunate, since there is already a correlation between the level of glucose in blood and self-control, and choosing these foods after or before a demanding task is probably the right thing to do for the brain.

I think the experimental procedure should attempt to invoke cultural situations in which self-control is more and less salient. I couldn't get the paper to load so I couldn't look at their procedure directly, but it occurs to me that if they are assuming that the choice reflects self control then it could simply be chalked up to the different explanations offered above about implicit energy demands. But if there is a way to ensure that self-control is an active frame of reference for the choice, then those explanations might not hold water.

I found another way of looking at self control that puts it in the frame of evaluations over time. In that sense, an act of self-control demonstrates that someone prioritizes future rewards over present rewards. I find this particularly compelling considering that humans consistently mis-evaluate future risk (in insurance-related tests - eg pay five now and don't bear the brunt of a flood vs. save the five and gamble that a flood won't happen).

I posit that what we think of as "self control" is an artifact of benefit evaluations between present and future that require lots of processing in the pre-frontal cortex. When the cortex is otherwise occupied (say with number memorization), we're less able to run those computations of present vs. future benefit.

I've found that if I explicitly ask myself to do a future/present benefit comparison for this sort of tempting situation, I rationalize the future benefit and deny myself the present pleasure. For me, "practicing" with neural functions largely falls into a set of actions I would describe as:

1) articulating (or speaking in my mind) the problem
2) generating criteria for a successful outcome from the scenario
3) consider the actions available to me in light of (2)
4) execute the action that will be most beneficial over the long term*

In short, I suggest that if you're interested in practicing, or "exercising" any mental muscle like this, you put the whole problem into a rational framework and make a conscious decision instead of a non-conscious decision.

It's anecdotal, but (from my conversations with friends who've recovered from anorexia) therapy for early-detection eating disorders follows the same pattern: create a framework in which the patient can rationally understand the reasons why they're engaged in self-destructive behavior, and then train them to engage the pre-frontal cortex when evaluating decisions related to food.

It's amazing how breaking a mental bug down into its component parts and committing to thinking (consciously) every decision related to the bug all the way through can override even really deeply wired patterns.

*naturally, this is a gross over-simplification. sometimes the short-term sugar kick from a piece of cake is just too biologically necessary to ignore. like when my brain's exhausted from remembering all those digits...

What if the choice between the cake or fruit salad is not one of self control but of personal preference instead?

Could this perhaps be a case of false correlation?

- satat