How Many Self-Controls?

Just when I had decided that my little psychology idea was too half-arsed to publish on even this blog, along comes one of the world's most famous psychologists - Martin Seligman - to support the idea with a personal anecdote:

Some theorists, like my friend Roy Baumeister, believe that self-control is a general trait. My experience with weight-loss versus exercise belies this. I have weighed 95 kg for the last twenty years, and I have dieted a dozen times only to return to 95 kg each time, usually after losing about 5 kg. No self-control? Hardly. Eighteen months ago I took up walking, knowing that 10,000 steps per day halves cardiac risk for someone my age and with my profile of risk. I have walked an average of 14,000 steps per day ever since and my New Year's resolution is 5,000,000 steps in 2009. I am well on track to my goal. So self-control is for me highly domain specific. For you?

The half-arsed idea was not that self-control is highly domain-specific, but that what we call self-control seems to actually be (only) two abilities: (i) The ability not to do something pleasant that you think you shouldn't do and (ii) the ability to do something unpleasant that you think you should do. If there is some truth to that, Seligman seems to be good at doing something (walking) but not at not doing something (eating); I think I'm the opposite.

While I'm at it, let me suggest that which one people are good at shows a decent correlation with how prone they are to be active more generally. Having said that, it seems entirely possible that people are good or bad at both; I would think the two are in fact somewhat positively correlated.

I know, it sounds like there is a contradiction in the above paragraph, but I don't think there is.