Preface for the Dictionnaire de la Sociologie (Universalis: Paris, 2007)

Sociology has no agreed-on birthday or founder. Some begin its history with Herodotus or Machiavelli. I prefer Comte, largely because his Système de politique positive ou traité de sociologié appeared in 1839, when Daguerre first described how to fix a photographic image (another way to describe society).

Sociology has always been disorderly, engaged in interminable conceptual arguments. The many great theoretical and empirical monuments to its usefulness exist because, though sociologists often disagree, their differences are smaller than their large areas of agreement.

The core of agreement in sociology is this: people everywhere and always live in groups. This axiom distinguishes sociology from disciplines which take this point for granted or ignore it. Several things then follow.

Because people live in groups, they do what they do together. When we investigate something sociologically, we look for all the people involved and all the ways they are involved with with one another. We investigate the togetherness of their collective action. This focus on doing things together distinguishes the sociological point of view from others.

Some ways of doing things together aren't "social" in the usual sense. The people involved in these interactions may not have social norms, not even be aware of each others' existence. But they are still affected by and respond to others' actions. Demography studies populations: the relative ages of a group's members, its gender ratio, the flow of immigration and emigration, which all influence how things get done, though the people involved never communicate with one another. Studies of the global economy describe how people who, though they share no norms and don't communicate with each other, are strongly affected by each others' activities.

But the largest area of sociological work deals with those human activities that people carry on collectively, through communication. People who talk to each other eventually come to share ways of thinking about what they do together, usually disagreeing on many matters but agreeing on other, more general topics, as sociologists might disagree on how to do research but agree that it should be done and that it should be aimed at discovering something more general than the specifics of a situation. Not just the name and coordinates of someone who may have committed a crime and the details of the alleged crime, but the pattern found in many activities in which people cooperate to create the label of criminal and apply it to specific actors. Not the specific details of how your family or mine lives, but the patterns of family life that characterize a society, which its members accept as "correct."

How to think about these patterns is what divides sociologists. Some want to understand observed patterns of social interaction as the result of the existence of a culture or collective consciousness. Others prefer to talk about social organization, embodied in classes, ethnic and racial groups, occupations, local communities, fields, or worlds. In every case, they describe how people take account of the way others impinge on their activities and the accomodations everyone makes to what is undeniable and easily observable: that they co-exist with and through others.

The two possibilities--highly specific descriptions of historically unique situations and abstract generalizations about theoretically defined classes of events--define the poles, the extremes, of what sociologists do. Some sociologists explain specific phenomena: the causes of the French Revolution, the culture of an occupational group or ethnic community, the specific features of a work of art. Others think that the work of science requires us to discover universally applicable laws, like the alleged "laws of physics," statements of regularities in social life which hold true for all such situations anywhere, any time.

I reveal my own prejudices and allegiances when I say that both are admirable aims, and that many great sociological workers have shared them, but that in the end what we all achieve is something less than the one and more than the other: statements which hold true provisionally for some class of phenomena under conditions which we cannot fully specify.

That may seem unfortunate, but it is exactly the provisional character of sociological knowledge which identifies it as a true science. Because all scientific knowledge is provisional. Our discoveries and laws all depend on conditions which may seem to us universal, but which further research will inevitably show to be true in the places we know about but not in some other places we have yet to discover. Taking these further possibilities into account as they arise is what keeps sociological science moving, learning from our mistakes, adding more and more to what we know even as we learn, more and more, how little we know.