Why doubt is a scientific virtue worth supporting



On May 28, the House Committee on Science, Space, and Technology passed the First Act. Among other things, the legislation would cut some $50-million in funds to the National Science Foundation for research in the social sciences.

Elected officials might have more than one reason to oppose NSF support for the social sciences.

First, social scientists study humans, and politicians govern humans. For that reason, the social sciences cannot avoid producing political claims. “We don’t like your study” has been the most prominent and least persuasive argument against NSF funding; the committee has issued news releases singling out the NSF-supported studies they liked the least.

What’s more, all people have beliefs about how the social world works, which may create a strong temptation to meddle with or contradict the conclusions drawn by researchers. Although major projects in the natural sciences, such as Fermilab, have lost federal funds, the problem is particularly significant for the social sciences. Members of Congress may argue that public funds should not be used for a particle accelerator, but they probably do not have their own, competing theories about the nature of elementary particles.

It’s a third objection that deserves the most consideration, although I have not heard it voiced by any politician: The social sciences attempt to explain why things happen, and these efforts at explanation have had mixed results. There is substantial disagreement within and across disciplines about the proper way to explain things and the reasons to do so. In other words, doubt is all too visible. This weakness is also, in my view, an important virtue.

To make a sweeping distinction, the social sciences employ two basic styles of explanation. A colleague of mine, borrowing from a colleague of his, refers to one as “arrow salad.” Here a relevant outcome is brought about by the interaction of some large number of real or notional entities or variables, represented visually with causal arrows.

The other style of explanation points to a primary cause to explain all effects—everything is social class, or rational self-interest, or brain chemicals, or whatever. Call these “baked potato” accounts: hearty, to be sure, but bland if consumed daily for an entire career.

What is the purpose of explaining? Sometimes an explanation can allow us to turn something complex into something simpler and, for that reason, more useful. Ask a distraught friend why a romantic relationship ended last week, and one will hear the preliminary attempt to make sense of things—a long and disorderly rehearsal of occurrences that were the proximate cause of the breakup. Ask that same friend about the breakup five years on, and the explanation may be simplified to a line or two: We were different people, we wanted different things. “We were different people” may be, in its way, no less true, and a useful guide for the future.

Simplification is not always the goal. Explanation can also be a way of making things more complex, or keeping them alive in social memory. Many Americans, academics and others, continue attempting to explain why the Civil War happened or why Reconstruction failed. However strong the explanations, these events never stay explained. This is for the best, if the restlessness keeps us alert to their enduring importance .

To those who are unsympathetic to the enterprise, this divergence in approach may already sound plenty unscientific. Now consider that there is persistent, well-informed disagreement about the generally appropriate logical bases, technical means, and social ends of explaining human behavior. Further, some intrinsically interesting phenomena are too complex, rare, or historically tangled to afford researchers a clear view.

In my own discipline, sociology, revolutions and the emergence of the modern nation-state are two notable examples. They are far too important to ignore, but too messy to admit of scholarly consensus. Some topics are so freighted with meaning that explaining “why,” in a technical sense, may appear to miss the point. (Ascertaining the situational predictors of interpersonal violence, for instance, is not the same as making sense of murder.) With all that in mind, there is hardly any causal claim a social scientist can make that will not prompt another social scientist to disagree.

In my own research, I mostly avoid asking why things happen. I spend much of my time in the humanistic corner of my discipline, which is somewhat more concerned with describing how than explaining why. I no longer receive my daily serving of arrow salad, and I read my favored baked-potato prophet, Karl Marx, much less often than I did at the start of graduate school. Part of this is a personal disposition. I have also been persuaded by thoughtful social-scientific criticisms of social-scientific explanation, which I have crudely caricatured here.

This may appear to be a wishy-washy and unscientific attitude for a purported scientist in training. I am dragging my doubts and the doubts of others into the open because doubt is a scientific virtue that has been scarce in the present debate. The defense of federal funding for social research, led by the various professional associations, has been based on the claim that social research is an unambiguous social good. In my view, it is too soon to tell. The social sciences are relatively new and small. The empirical object—individual and collective human behavior, past, present, and future—is extraordinarily complex and unpredictable. The claims produced by the social sciences truly are political, which leaves ample room for mischief and misuse. I am never certain that we are correct or justified.

Yet doubt is the reason that it is important to support the social sciences. Everybody may have a theory about the world, but most people do not submit their theories to formal scrutiny, and the people in charge may have strong reasons to avoid scrutiny. A responsibly governed society therefore needs doubters. Doubters agree or disagree in accordance with the available evidence. They are prepared to publicly contradict others and, in turn, to submit to the risk of being publicly wrong. An openness to evidence and contradiction is particularly desirable in a political climate like ours, in which far too many people are certain that they have the answers, and show contempt for those who disagree. Public support for the social sciences is a ready way to promote constructive doubt.

Author Bio:Ben Merriman is a doctoral candidate in sociology at the University of Chicago.