Wednesday, November 29, 2006

Two Kinds of Habit?

Last week, Tim Schroeder spoke to the philosophy of action workshop here at UCR on habit and addiction. He gave a unitary account of "strength of habit", according to which a habit is strong to the extent it's difficult to extinguish.

It struck me, however, that there might be two different ways for a habit to be strong. Some habits are difficult to extinguish simply because we thoughtlessly repeat them, but if we pause to reflect they have no pull on us. For example: My new car has a foot parking brake and my old car (which I still sometimes drive) has a hand brake. Now that I've grown used to my new car, I find myself pushing on the floor with my foot even in my old car when I park. Or: I used to absently chew my fingernails, though I recognized this made my fingernails look unappealing. Though when I stopped to think about it, I found no particular impulse or craving to chew the nails, it was a hard habit to break simply because I would find myself chewing them without thinking about it.

Other bad habits -- perhaps these more closely resemble addictions -- are hard to break in a different way: Even when I reflect I find myself drawn to continue them. For example: Eating junk food at night when I'm stressed. It's not unusual for me to stop and reflect as I'm reaching toward the cookies -- but then I rationalize: It's just a little bit, or I'm already eating it, or just this one cookie, or it's a special occasion, or even "how interesting that I'm doing this -- an occasion to reflect on weakness of will!"... In contrast, there's no impulse to rationalize pressing my foot on the floor or chewing my nails!

If a habit is a repetitive pattern of behavior that proceeds somewhat independently of rational guidance (this is close to but not exactly Schroeder's definition), then there are at least two neurologically distinct ways to form a habit. One is mere repetition. Doing something over and over again tends to strengthen the neural pathways that generate that behavior and weaken the pathways to other behavior, independent of any particular reward. The other is reward: If a behavior is rewarded, especially if it is strongly rewarded, that tends to increase the likelihood of its occuring again. Maybe, then, these two different mechanisms of habit formation underwrite two different types of habit with different types of habit strength and different conditions of extinction -- and in particular, the first type of habit may be more thoughtless and more easily defeated by thought than the second.

1 comment:

  1. I’m not sure that habits formed by mere repetition are completely without reward. The reward for using your parking breaks is that your car doesn’t roll downhill. Even if your car rolling downhill isn’t a real possibility at a given time, the habit of using your parking breaks, wherever they are placed, is nevertheless useful for those rare occasions when you’re parking on a slope. If you value your vehicle and detest having to deal with insurance claims, the payoff for having the habit will be relatively large compared to the amount of effort you put into creating the habit in the first place. You might dispute this use of ‘reward,’ but then one might also dispute whether it’s actually rewarding to eat junk food. So maybe the real difference is not in the types of habit but in the types of reward. Here you could still say that different types of reward, as two different types of causes, by that very fact create different types of habit. Even so, it’s completely arbitrary how you want to draw distinctions or lump things together. You could distinguish between “types” of car accidents according to which hill a car rolls down, each hill representing a different “type” of cause. Or you could say that car accidents involving cars rolling downhill are of the same “type” because they each involve a slope as cause. Which cause is more important? Likewise, you could distinguish between “types” of habit according to which “type” of reward causes it, or you could say that all habits are of the same “type” because in each case a reward is involved as cause, regardless of the specifics of each reward.

    ReplyDelete