Another Enlightening Trip into Theory Issues

Problem: what’s a cause?

Despite probably sounding like I am veering into banal high school philosophy here, it’s an honest question. The search for causal factors in social science is rarely questioned. Working assumptions in much research are that we will hypothesize a relationship (informed by some kind of theory) then use statistical techniques to evaluate a sample and determine whether we can prove our initial hypotheses for the entire population under consideration.

Sorry, not prove — that’s fairly well explained in methods courses, that we can’t “prove” anything. We can be 95%, 99% sure, 99.99% sure, and so forth. (By the way, 95% is the usual standard for us. Higher standards aren’t necessary since lives rarely hang in the balance in this line of work.) Despite the repeated statements about the standard scientific idea that we can only disprove hypotheses, there doesn’t seem to be a lot of doubt about causes.

The tricky thing about causes is that they generally have causes themselves. I don’t mean to get into any kind of metaphysical “first cause” argument here, because that’s not the issue. The problem is that the social world is very complicated, and causes impact upon other causes in a terribly convoluted string of events which can be traced back potentially forever.

I don’t think this comes across very well in sociology courses, or else it is not something that many people are interested in talking about. And so, we seem to push the problem of regress out of our heads, forgetting our own mantras of unprovability. Is the problem solved by practicalities such as running out of resources to trace lines back? Perhaps, but this doesn’t seem to be a very satisfying answer.