Steven Pinker relates this joke, I think that its in How the Mind Works.
OK, calling it a joke in its current form is charitable.

While watching a network legal drama, we see a suspect break down crying in the witness stand. Why did this person break down? If one were to answer, because of activity in the tear glands, anyone who was looking to get the plot would no doubt find as little humor in your response as you do in this joke. But why should you find any humor at in this vignette?

Well, if you take a look at any formal models of causation, you see that the activity in the tear glands "d-seperates" the crying from just about any other cause you might want to explore. That is, it makes them informationally irrelevant. Once you know the tear glands are active, knowing any causes prior to that will not give you any information about wether or not the person on the stand was crying. Also, intervening to change any of those prior causes will not stop or start the crying unless it also changes the states of the tear ducts.

Thus, on a formal model of causation, given the glandular activity, the asked for plot points not only don't cause the crying, they're not even relevant.

This makes sense, given that the person is crying and their tear glands are secreting tears then it doesn't really matter wether the crying is caused by the revelation of duplicity on the cryer's part, an allergy attack, or the instruction's of the show's director. This is highlighted by the fact that somone asking, "why is the witness crying?" would probably accept the first of these three possibilities as the most legitimate answer to their question, but the third is most likely the actual reason.

Recently, I used Judea Pearl's text Causality in a seminar I taught on the subject. If you're interested in formal models of causality, that would be a good place to start.  ¶ #

0 comments Post a Comment