E.G.

It is often said that overdetermination is not or cannot be "widespread."  Whether this is true or false depends on just how common something needs to be in order to qualify as "widespread." 

Nobody should deny that overdetermination of events can happen, and probably has happened.  There probably has been at least one occasion on which someone has been shot in the heart by two simultaneous bullets, for instance.  After all, I think I’ve heard that it was at one time common practice to execute criminals by having a dozen or so men simultaneously shoot the criminal.  This kind of practice provides a lot of opportunities for overdetermination to occur.  So even if, in each such case, it is highly unlikely for overdetermination to occur, the fact that there have been so many such cases makes it fairly reasonable to suppose that overdetermination has occurred at least once.

But how many times has overdetermination occurred?  Does the claim that overdetermination is not "widespread" amount to the claim that less than 1% of all events are overdetermined?  Or to the claim that less than .00001% are?  Or even less than that?

I don’t know the answer to this question, but I think we can put some tentative limits on the answer by considering some examples.  Suppose I am unsure why, exactly, rocks fall to the ground.  But I have two theories in mind.  The first theory is that there is a force called "gravity," and this force acts on rocks to make them fall.  The second theory is that rocks have some kind of intrinsic tendency to move toward the center of the Earth.  Suppose I am torn between these two theories; I think there are reasons to believe each one.

These two theories are not mutually exclusive; it is logically possible that rocks both tend to move toward the center of the Earth "intrinsically," and are acted on by an "extrinsic" force which compels them to do the same.  In such a case, the event in which a rock falls would be overdetermined: there would be two sufficient causes for that event.

If I accept the view that overdetermination is not "widespread," I should think that there is a good chance that one theory is correct, and a good chance that the other theory is correct, but not very much chance that both theories are correct.  Perhaps I should believe that there is almost no chance that both theories are correct.  But does this commit me to say that overdetermination almost never occurs?  Does it commit me to say, for instance, that it is extremely difficult to find any cases of overdetermination?  I do not think so.  In fact, I think that in the case just described, one could think that overdetermination is fairly common, and fairly easy to find, even if one thinks that that overdetermination is highly unlikely in any given case.

To illustrate this, consider another case.  I’ve just gotten home.  My roommate is usually home from work by now, but he is nowhere to be found.  Two explanations spring to mind: 1. My roommate was delayed at work; 2. My roommate was hit by a drunk driver on the way home from work.  I probably have very little reason to believe 2; I should probably believe 1.  After all, it is highly unlikely that my roommate, in particular, was hit by a drunk driver.  But this does not commit me to say that nobody is ever hit by a drunk driver; nor does it commit me to say that it is hard to find a case in which someone is hit by a drunk driver.  Indeed, if I turn on the evening news, I am fairly likely to find that someone was hit by a drunk driver.  So I can believe that "getting hit by a drink driver" is a fairly common event even if I think it is not widespread enough to support the hypothesis that someone in particular is missing because he was hit by a drunk driver.

I think that those who hold the view that "overdetermination is not widespread" can take a similar line.  Perhaps overdetermination happens all the time; perhaps it is extremely easy to find cases of overdetermination, if you know where to look.  But this is compatible with the view that overdetermination is quite rare — rare enough, anyway, that one should usually assume, when trying to explain a given event, that overdetermination has not occurred in this particular case.  And this more modest result is, I think, the one at which people are usually aiming when they make the claim that overdetermination is not "widespread."