Natural Science, pt. 2

In my last post, I lashed out at the "anti-science" folks within the social sciences. I decided to elaborate on that a little bit.

My comments need to be put in the context of how "science" has been used as a label within the social sciences. I have many, many problems with much of what's gone on under the banner of "scientific" social science. There's a big advantage to being a "science" in our culture - the label brings a lot of prestige and credibility to what you're doing. You need go no further than the countless advertisements that boast of "scientifically proven results" for everything from toothpaste to hair loss treatments for evidence of this. Much of "scientific sociology" has had all the trappings of a science (technical jargon, fancy statistics, big research grants) but has failed to go much beyond that. Besides resulting in obscenely boring journal articles, I don't think they've even really been all that "scientific." The picture of science that postmodernish critics of science paint makes more sense if you use the practice of social science as your model of science.

The history of sociology as a science is the history of a discipline obsessed with riding a fine line on the whole "science" issue. On the one hand, disciplinary entrepreneurs like Emile Durkheim insisted sociology was a science, as "rigorous" as any other and early sociologists borrowed the language of sciences like biology to talk about "the social organism." On the other hand, Durkheim had to insist that there was something about "the social" that made sociology distinct from all other sciences and, hence, requiring it's own discipline, methods, journals and, of course, professional departments and associations. These two goals had a paradoxical effect on the social sciences: they became isolated from the "natural" sciences at the same time that they were insisting they were "scientific." "Science," as it developed in sociology, became associated with statistical methods, abstract jargon and a belief that through the technical, "neutral" language of science, we could discover (or, at least, approximate) absolute truths and laws about the social world. Because "the social" was an entity in itself ("sui generis"), we didn't need to know anything about biology, physics or even psychology to do sociology. The field was founded on a particular strain of empiricism that believed that through observation and inductive reasoning, the truths of the world would become visible to us. So we go out and do a quantitative survey (or, for that matter, an ethnography) and gather all this "data" and then we look at it and uncover social reality and it's "laws" as they really are, "out there" in the world.

Critics of this approach correctly point out how naive it is. When you're doing a survey, the very categories that you choose to ask people about and the way you ask the question affects the way they respond. When you observe people, your presence impacts the way they act and, moreover, you always filter what you observe through your own preconceptions, biases, etc. Categories like race, class and gender aren't "out there" as absolute truths, but we "construct" them. Since we're always biased in these ways, "objective" science isn't possible in the social sciences (perhaps in the natural sciences, too), so we should give up the whole project in favor of an interpretive project of how people "socially construct reality."

The problem is that this view of science is ridiculous when looked at from the perspective of natural sciences. The subjectivist view is just incomprehensible, like I argued in the last posting, unless you believe in miracles. Obviously there's something real out there, or else how would any of the technology we take for granted ever work? But it's also true that categories like "race" are social constructed. Perhaps the social sciences really are just qualitatively different than the "natural sciences."

But this doesn't work either. Physics doesn't study "absolute truths" either. There also are no physicists who argue that we simply "socially construct," out of thin air, the content of physics. For example, electrons are theoretical entities. Do they really exist? We can't see them. But we know that if we act as if they're real entities, we can interact with them in consistent, predictable ways. So is the category of electron something we impose on the earth that, somehow, works or does it really correspond to some actual unit of stuff? Interesting question, but hardly a debate about whether or not electrons are absolute truths or just made up stuff. (Ian Hacking's the philosopher most associated with the electron example; he calls his position "experimental realism.")

So the natural sciences, in practice, reject both absolute truths as well as arbitrary "constructions." Of course, there's an inductive element, where you observe the world, but there's also a deductive element, where you devise theoretical entities with the purpose in mind of tested them and revising according to the extent that they do, or do not, correspond with your experience. No belief in "absolute truths" necessary. No defeatism about inescapable bias necessary either. Science is about experimentation and interaction - both processes which are not "neutral" (in the sense of an disembodied, transcendental view of "Truth") in either their aim, method or results. A science is "objective" to the extent that it is honest and transparent to others about it's experimentation. This method is not synonymous with a particular type of data, a particular institutional arrangement or a particular type of language. In fact, technical jargon can just as easily corrupt science, by rendering it less open to critical challenge. Similarly, institutional arrangements, such as the way academia is professionalized today, can corrupt science by giving people professional incentives to be dogmatic, parochial and nontransparent for fear of competition that may challenge their legitimacy and prestige. In these respects, "scientific sociology" has been profoundly anti-scientific in it's practice. The task, however, is to challenge it's claims to science and it's application of the scientific method, not to accept it's claims to science and reject science altogether.

So when I say that social science should be a natural science, I'm not saying that we should all become biologists. Nor am I standing with sociologists, past and present, who have used "science" as a cloak to cover up bad research. I'm saying that to be a science, whatever your subject matter, requires a naturalistic stance; a commitment to using the scientific method in the pursuit of natural, not supernatural, explanations of the world and practical, not absolute, beliefs about the world. Natural explanations require viewing human beings as part of the natural world. Rejecting absolute beliefs about the world means accepting that you might be wrong. This is the crux of the whole science/evolution debate because the fundamentalist Christians refuse to take these steps. They cannot accept, as necessary price of admission for anyone entering the world of science, that they might be wrong because, as believers in absolute truth, they cannot be wrong as this would then mean that God is wrong, which is incomprehensible to them. Science, far from being "just another discourse on absolute truth," is a method that requires, by definition, that we challenge notions of absolute truth.