Philosophy of Social Science: Lecture 3, Interpretative Social Science 1. Basic Doctrines The standard model, well explicated by Rosenberg, sees explanation in the physical sciences as causal, and analyses causality in terms of constant conjunction that is, in terms of events being brought under a covering law that states if an event of type A occurs, it is invariably followed by an event of type B. [ Cyanide kills is, if it is a causal law, analysed as if someone drinks cyanide that person dies. ] Believers in this view explain the gap between a full explanation and what we ordinarily proffer as an explanation in terms of everyday speech involving explanation sketches: we know that we could but needn t fill out the explanation. Whence the first week s question whether there are social laws and if so, what they are. We ve seen the immediate response to the search for a real science of society so construed: the denial that social life is or can be amenable to causal analysis: whence the cause vs meaning, controversy. It is difficult to provide a genuine historical pedigree for the view that the Geisteswissenschaften [the sciences of spirit or mind or culture ] are logically, conceptually, or categorically different from the Naturwissenschaften because that lures us back to Vico and his Scienza Nuova and then to Hobbes and Descartes. The simplest thought is that until there were some nascent social sciences to argue about 18 th and 19 th century political economy for instance the contrast is not easy to explicate. One important point to remember is that the contrast between the cultural and the natural is not Mill s contrast between the physical and the moral. I evade the history of ideas by saying a very little about Weber and verstehen only to contrast him with both Marx and Durkheim then leap headlong to Winch s The Idea of a Social Science a. Weber held that the object of social science was to explain actions to which a social meaning is attributed by the actors under investigation; Weber thought this implied methodological individualism and meant that there was no clear boundary between history and sociology. Weber distinguishes between explanations adequate on the level of meaning and explanations adequate on the level of causality. (As shall I at the end of this lecture.) The contrasts between meaningless and meaningful, and between causal adequacy and intelligibility are not difficult to draw. We could (though it is hard to say why) give an account of metal tokens passing from hand to hand, but only when we know that the pieces of metal count as coins can we know that someone is paying someone for a good they have bought. Again, you may see me insult a colleague and ask why I did it; I say he insulted my wife and thereby achieve meaningadequacy; but if I have no wife it can t really be the cause and fails the causal-adequacy test. Weber s emphasis on meaning-adequacy is what is much-discussed as the doctrine of Verstehen. And his emphasis on causal adequacy is exceedingly good sense. b. There are many puzzles about Verstehen; is it a form of empathetic understanding, and if not, what; does it play a basic role, or is it only a decorative supplement to real explanations. Weber was eager to avoid what he thought of as Idealism (the thought that history is driven by thought) see The Protestant Ethic but it remains unclear quite where the verstehende level fits. 1
c. What is clear is that Weber thought the Durkheimian determination to produce causal accounts that did not invoke social meanings or individual intentions was a non-starter. (Not that Durkheim really manages to do it: contrast the methodological prescriptions of Suicide with the text of the explanations.) And Weber thought that vulgar Marxism, with its insistence on deterministic iron laws was a non-starter. (Not that this is the only interpretation of which Marx is susceptible.) So, one interesting question is whether a Winchian/Taylorian emphasis on the role of interpretation in social science picks up the practice of real sociology as many of us think or is wholly at odds with sociological practice as Gellner, for instance, thought. This requires us to look for the most satisfactory interpretation of what interpretivism maintains. d. Winch, recapitulated, holds the following views: that social science explanation rests on the concept of meaning not that of cause, or in the diluted form, that the social sciences rely more on meaning than cause; what an action means is given by the local way of life; there is no higher court of appeal beyond that of the local way of life in assigning a meaning to an action; that is, practitioners say what counts as a move in chess, a part of the mass, a dishonourable advance to a lady, and so on. What this entails for the practice of the social sciences is importantly and on Winch s account, nothing. It s important to see why nothing follows by distinguishing two concepts of methodology: in the first, we offer cookbook recipes to people doing research, and their practice is the immediate focus; in the second, we discuss the meaning of the results. Winch s views affect the second, not the first. On Winch s view, Durkheim s belief that social facts are things and operate coercively upon individuals must be false because it is an ontological claim in the second category; but the injunction in the Rules of Sociological Method to treat social facts as things may be less obviously bad advice, because it may be that we ought to collect and analyse data in that spirit. Still, it must follow that Durkheim s ambition in Suicide to explain differential suicide rates in different countries by means of suicidogenic currents is misguided. e. What follows? That the identification of action in the first instance is given by the conceptual schema of the way of life investigated. Less follows about the explanation of the action so identified. A particular step in the celebration of Mass is identified by reference to the prescribed way of celebrating Mass; at one level, the explanation of that step is also explained by reference to the Mass, and the Mass is what the worshippers say it is. But nothing follows about the answer to such questions as: why do people celebrate Mass at all; why are there religions at all, in any society if indeed there is a religion in some particular society? Why do some societies have no sense of the numinous and others a strong sense of it? Now, it s clear enough that the say-so of the local priests etc is decisive about what follows step by step in the Mass; but equally clear that they have no authority over comparative anthropology while Mary Douglas, in Natural Symbols, might tell one a bit. 2. Sophisticated Versions a. Winch paints a broad brush picture, dependent rather heavily on setting up a Mill-like account of a causal naturalism as the stalking horse. It s clear 2
that any idea of explaining the Mass by looking for correlations between drinking glasses of wine and then kneeling down would be foolish, and that Winch is quite right that i) we need to identify actions before we can think about generalisations, and ii) we have to do that using a lot of background knowledge that amounts to interpreting the local cultural setting. b. But this leaves two questions: first whether there are ways of explaining how interpretive explanation works, and second what backing to interpretive explanations we can arrive at. The first question leads towards the discussion of the fussion of horizons, a notion associated with Gadamer s Truth and Method. No discussion of Gadamer can do justice to the subtlety of the way he thought, but the elementary suggestion is this: when we look at a painting from another epoch, a Giotto fresco for instance, we try to put ourselves in a situation in which we think we can see the painting as the artist saw it. To understand it from the inside we must fuse the horizon of our understanding of the world with his. We can t know that we have done this successfully this is the sociological version of the hermeneutic circle since there is no third place to stand. The difficulty with Gadamer s idea in practice is that it suggests what one might call a quietist methodology; we are to learn from them, not to think they were irrational or superstitious or whatever. It is this Habermas, for instance, will not accept, and that leads him to reject a Gadamerian interpretivism. c. As to backings, we might take one of three views. The first is to deny the need for them. If we today understand Giotto as he understood his own work, we have done all we can we may ask different questions about the paintings and about him, but within that circle of explanation we know what there is to be known. The second is to suggest that we can change the subject; within the conventions of art, the religious beliefs of the day, and the techniques that a painter had available, we can ask why the subject matter, style, audience and patrons were chosen as they were. We don t in a narrow sense improve our understanding of the painting, but we improve our understanding of the act of painting it. The third is to suggest a more radical change of subject; we can ask large questions about the economic, political, and other causes of the evolution of styles of painting and of these paintings in particular. Habermas, for one, seems to argue that Gadamer restricts us in such a way that we cannot ask these third sorts of question. 3. Complaints a. Gellner objects that Winch s account of what social science can do simply restricts the activities of social scientists in a way they couldn t tolerate. This may or may not be true. b. A common objection from Gellner and Hollis is that prioritising the say-so of the folk on the ground does not have the consequences Winch thinks. Thus, if the natives come to say they were mistaken but we may not second guess them, we either have the view that in following their view we deny their view or in denying their view we follow their view. (Imagine a member of the Azande announcing that they used to believe in the witch oracle but now they don t; if Winch is right and the meaning of what they 3
believed is unchallengeable, then the present Azande are wrong; but the theory says that the native informants are right ) There are in fact some elegant examples of native informants holding views that are much more sophisticated than the simple form of life account can cope with. c. Much else is hard to pin down. Habermas objects that Gadamer forbids a critical social science; but it is not obvious that this is true. One may think with Charles Taylor mode that the interpretative exploration of a way of life shows up its flaws, so that a critically empathetic stance is not only possible but so commonplace as to be almost inevitable. There are large questions about whether the explanatory stance from which we judge a practice as irrational is to come from inside or outside the practice itself, and another large question whether the inside/outside distinction is ultimately viable or only an initial shorthand device. Of course, sometimes there is a simple jurisdictional answer Catholicism possesses a means of deciding what is to believed by Catholics de fide but can there be a jurisdictional answer to the question of what counts as a religion at all? d. But the widest objection is simply that it shortcuts the possibilities of explanation. Gellner always held that a loose functionalism was inescapable in social science, and thought that Winch had ruled it out a priori. Winch thought he had not done so; but then it becomes hard to see what we are arguing about. Or, again, we may agree that some sort of interpretation is needed even if we go on to invoke rational-choice in our explanation of why people are doing what they do thus, we know we aren t to count sneezes and sniffles as actions, we know that this is a game of Monopoly and not the real thing, we know that it s a church service not a party but we then go on to explain the activities in terms of utilitymaximisation. e. And it can shortcut the full sophistication of actors on the ground; we talk about the beliefs of the Azande, but do we have a clear view of what we are after? To infer that the Azande believe in witches because they employ a witch-detector to discover whose fault it is that someone is ill may be a bit quick. Quesalid the shaman in Levi-Strauss s account, has a wonderfully complicated atttude towards his own success as a healer. To engage in a simple in the practice/outside the practice dichotomisation loses that subtlety. 4. Compromises a. The object is to explain why things do and don t happen; interpretaivism tells us how to identify what happens rather than why. This means that dualists like Lessnoff have to say something about how the interpretation gets into the causality of the happening. Many do not. b. The answer is to rely on Popper s concept of the thing to do. When explaining human behavior we show that what an agent does is the thing to do (or a botched attempt at it.) There is no a priori limit on the ways in which something can be the thing to do, and the concept itself is a matter of modal logic not ethics. (The thing to do if you want to be an effective cut-throat is to start by getting a really sharp knife ) This is situational logic and covers the various sorts of Weberian rationality. c. But, there is a superfluity of things to do in any given situation and too many ways of explaining conduct as the thing to do. So the final thing to 4
insist on is that the interpretation is the one with the causal purchase on events. In which case, it isn t a conflict between cause and meaning but an insistence that cause runs through meaning this produces dualism in one sense because it invokes an emphasis on perspectives, but monism in another because it insists that explanations adduce causes. d. The advantage of the compromise position is that it leaves open the great subject of adduced causes versus real causes, and with it issues of deception, self-deception, ideology and much else; only if we avoid drawing too sharp a contrast between reasons and causes can we even raise such issues which are evidently central to the study of politics. Take any piece of political science research and you can dissect it for thing to do explanations embedded in other accounts of why something was, or seemed to be, the thing to do, how people acquire their beliefs and how they acquire their values. 5