There seem to be a couple of fundamentally different approaches to the problem of "understanding society." I'm not entirely happy with these labels, but perhaps "empiricist" and "critical" will suffice to characterize them. We might think of these as styles of sociological thinking. One emphasizes the ordinariness of the phenomena, and looks at the chief challenges of sociology as embracing the tasks of description, classification, and explanation. The other highlights the inherent obscurity of the social world, and conceives of sociology as an exercise in philosophical theory, involving the work of presenting, clarifying and critiquing texts and abstract philosophical ideas as well as specific social circumstances.
The first approach looks at the task of social knowing as a fairly straightforward intellectual problem. It could be labeled "empiricist", or it could simply be called an application of ordinary common sense to the challenge of understanding the social world. It is grounded in the idea that the social world is fundamentally accessible to observation and causal discovery. The elements of the social world are ordinary and visible. There are puzzles, to be sure; but there are no mysteries. The social world is given as an object of study; it is partially orderly; and the challenge of sociology is to discover the causal processes that give rise to specific observed features of the social world.
This approach begins in the ordinariness of the objects of social knowledge. We are interested in other people and how and why they behave, we are interested in the relationships and interactions they create, and we are interested in institutions and populations that individuals constitute. We have formulated a range of social concepts in terms of which we analyze and describe the social world and social behavior -- for example, "motive," "interest," "emotion," "aggressive," "cooperative," "patriotic," "state," "group," "ethnicity," "mobilization," "profession," "city," "religion." We know pretty much what we mean by these concepts; we can define them and relate them to ordinary observable behaviors and social formations. And when our attention shifts to larger-scale social entities (states, uprisings, empires, occupational groups), we find that we can observe many characteristics of each of these kinds of social phenomena. We also observe various patterns and regularities in behavior, institution, and entity that we would like to understand -- the ways in which people from different groups behave towards each other, the patterns of diffusion of information that exist along a transportation system, the features of conflicts among groups in various social settings. There are myriad interesting and visible social patterns which we would like to understand, and sociologists develop a descriptive and theoretical vocabulary in terms of which to describe and explain various kinds of social phenomena.
In short, on this first approach, the social world is visible, and the task of the social scientist is simply to discover some of the observable and causal relations that obtain among social actors, actions, and composites. To be sure, there are hypothetical or theoretical beliefs we have about less observable features of the social world -- but we can relate these beliefs to expectations about more visible forms of social behavior and organization. If we refer to "social class" in an explanation, we can give a definition of what we mean ("position in the property system"), and we can give some open-ended statements about how "class" is expected to relate to observable social and political behavior. And concepts and theories for which we cannot give clear explication should be jettisoned; obscurity is a fatal defect in a theory. In short, the task of social science research on this approach is to discover some of the visible and observable characteristics of social behavior and entities, and to attempt to answer causal questions about these characteristics.
This is a rough-and-ready empiricism about the social world. But there is another family of approaches to social understanding that looks quite different from this "empiricist" or commonsensical approach: critical theory, Marxist theory, feminist theory, Deleuzian sociology, Foucault's approach to history, the theory of dialectics, and post-modern social theory. These are each highly distinctive programs of understanding, and they are certainly different from each other in multiple ways. But they share a feature in common: they reject the idea that social facts are visible and unambiguous. Instead, they lead the theorist to try to uncover the hidden forces, meanings, and structures that are at work in the social world and that need to be brought to light through critical inquiry. Paul Ricoeur's phrase "the hermeneutics of suspicion" captures the flavor of the approach. (See Alison Scott-Baumann's Ricoeur and the Hermeneutics of Suspicion for discussion.) Neither our concepts nor our ordinary social observations are unproblematic. There is a deep and sometimes impenetrable difference between appearance and reality in the social realm, and it is the task of the social theorist (and social critic) to lay bare the underlying social realities. The social realities of power and deception help to explain the divergence between appearance and reality: a given set of social relations -- patriarchy, racism, homophobism, class exploitation -- give rise to systematically misleading social concepts and theories in ordinary observers.
Marx's idea of the fetishism of commodities (link) illustrates the point of view taken by many of the theorists in this critical vein: what looks like a very ordinary social fact -- objects have use values and exchange values -- is revealed to mystify or conceal a more complex reality -- a set of relations of domination and control between bosses, workers, and consumers. With a very different background, a book like Gaston Bachelard's The Psychoanalysis of Fire makes a similar point: the appearance represented by behavior systematically conceals the underlying human reality or meaning. The word "critique" enters into most of Marx's titles -- for example, "Contribution to a Critique of Political Economy." And for Marx, the idea of critique is intended to bring forward a methodology of critical reading, unmasking the assumptions about the social world that are implicit in the theorizing of a particular author (Smith, Ricardo, Say, Quesnay). So Capital: Volume 1: A Critique of Political Economy is a book about the visible realities of capitalism, to be sure; but it is also a book intended to unmask both the deceptive appearances that capitalism presents and the erroneous assumptions that prior theorists have brought into their accounts.
The concepts of ideology and false consciousness have a key role to play in this discussion about the visibility of social reality. And it turns out to be an ambiguous role. Here is a paragraph from Slavoj Zizek on the concept of ideology from Mapping Ideology:
These same examples of the actuality of the notion of ideology, however, also render clear the reasons why today one hastens to renounce the notion of ideology: does not the critique of ideology involve a privileged place, somehow exempted from the turmoils of social life, which enables some subject-agent to perceive the very hidden mechanism that regulates social visibility and non-visibility? Is not the claim that we can accede to this place the most obvious case of ideology? Consequently, with reference to today's state of epistemological reflection, is not the notion of ideology self-defeating? So why should we cling to a notion with such obviously outdated epistemological implications (the relationship of 'representation' between thought and reality, etc.)? Is not its utterly ambiguous and elusive character in itself a sufficient reason to abandon it? 'Ideology' can designate anything from a contemplative attitude that misrecognizes its dependence on social reality to an action-orientated set of beliefs, from the indispensable medium in which individuals live out their relations to a social structure to false ideas which legitimate a dominant political power. It seems to pop up precisely when we attempt to avoid it, while it fails to appear where one would clearly expect it to dwell.Zizek is essentially going a step beyond either of the two positions mentioned above. The empiricist position says that we can perceive social reality. The critical position says that we have to discover reality through critical theorizing. And Zizek's position in this passage is essentially that there is no social reality; there are only a variety of texts.
So we have one style that begins in ordinary observation, hypothesis-formation, deductive explanation, and an insistence on clarity of exposition; and another style that begins in a critical stance, a hermeneutic sensibility, and a confidence in purely philosophical reasoning. Jurgen Habermas draws attention to something like this distinction in his important text, On the Logic of the Social Sciences (1967), where he contrasts approaches to the social sciences originating in analytical philosophy of science with those originating in philosophical hermeneutics: "The analytic school dismisses the hermeneutic disciplines as prescientific, while the hermeneutic school considers the nomological sciences as characterized by a limited preunderstanding." Habermas wants to help to overcome the gap between the two perspectives, and his own work actually illustrates the value of doing so. His exposition of abstract theoretical ideas is generally rigorous and intelligible, and he makes strenuous efforts to bring his theorizing into relationship to actual social observation and experience.
A contemporary writer (philosopher? historian? sociologist of science?) is Bruno Latour, who falls generally in the critical zone of the distinction I've drawn here. An important recent work is Reassembling the Social: An Introduction to Actor-Network-Theory, in which he argues for a deep and critical re-reading of the ways we think the social -- the ways in which we attempt to create a social science. The book is deeply enmeshed in philosophical traditions, including especially Giles Deleuze's writings. The book describes "Actor-Network-Theory" and the theory of assemblages; and Latour argues that these theories provide a much better way of conceptualizing and knowing the social world. Here is an intriguing passage that invokes both themes of visibility and invisibility marking the way I've drawn the distinction between the two styles:
Like all sciences, sociology begins in wonder. The commotion might be registered in many different ways but it's always the paradoxical presence of something at once invisible yet tangible, taken for granted yet surprising, mundane but of baffling subtlety that triggers a passionate attempt to tame the wild beast of the social. 'We live in groups that seem firmly entrenched, and yet how is it that they transform so rapidly?' ... 'There is something invisible that weights on all of us that is more solid than steel and yet so incredibly labile.' ... It would be hard to find a social scientist not shaken by one or more of these bewildering statements. Are not these conundrums the source of our libido scindi? What pushes us to devote so much energy into unraveling them? (21)What intrigues many readers of Latour's works is that he too seems to be working towards a coming-together of critical theory with empirical and historical testing of beliefs. He seems to have a genuine interest in the concrete empirical details of the workings of the sciences or the organization of a city; so he brings both the philosophical-theoretic perspective of the critical style along with the empirical-analytical goal of observational rigor of the analytic style.
Also interesting, from a more "analytic-empiricist" perspective, are Andrew Abbott, Methods of Discovery: Heuristics for the Social Sciences, and Ian Shapiro, The Flight from Reality in the Human Sciences. Abbott directly addresses some of the contrasts mentioned here (chapter two); he puts the central assumption of my first style of thought in the formula, "social reality is measurable". And Shapiro argues for reconnecting the social sciences to practical, observable problems in the contemporary world; his book is a critique of the excessive formalism and model-building of some wings of contemporary political science.
My own sympathies are with the "analytic-empirical" approach. Positivism brings some additional assumptions that deserve fundamental criticism -- in particular, the idea that all phenomena are governed by nomothetic regularities, or the idea that the social sciences must strive for the same features of abstraction and generality that are characteristic of physics. But the central empiricist commitments -- fidelity to observation, rigorous reasoning, clear and logical exposition of concepts and theories, and subjection of hypotheses to the test of observation -- are fundamental requirements if we are to arrive at useful and justified social knowledge. What is intriguing is to pose the question: is there a productive way of bringing insights from both approaches together into a more adequate basis for understanding society?