Structuralism, like positivism, is no longer of immediate interest in its own right, and I will not be attempting to provide anything like a comprehensive account of it. Where it is of interest is in a historical perspective, in that it represents one of the most thorough-going versions of modernist thinking in sociology, and a series of ideas which remain current in, for example, much neo-Marxist as well as much "post-structuralist" thought.
Thus structuralism imposes itself on us as a stage in social theory whose effects are still widely felt; its period of dominance is now long-since past, but what remains is very often a social theory which has developed from structuralist approaches, which has defined itself against them, and which bears the marks of this encounter.
Structuralism's claim to be considered as a form of critical modernism, however, is rather more tenuous. While much structuralism claims to be "Marxist", very often it appears rather more as an incorporation of Marxism into a rather more affirmative form of modernism; and this is particularly evident in the difficulties structuralist thought faces in coming to terms with reflexivity, as well as its consequent explicit or implicit flirtation with positivism.
I'll mostly be talking about Lévi-Strauss and Althusser, who are the best-known strictly structuralist theorists and can certainly be said to be critical in terms of their political positions and the implications of some of their work, if not always in terms of reflexivity. This structuralism intersects, however, with the positivist and functionalist school deriving from Comte and Durkheim, which leads into Parsons' "structural-functionalism"; and we shall see that there are good reasons for this convergence.
I won't be attempting to give a remotely comprehensive account of either Lévi-Strauss or Althusser, who are in any case opposed on a very wide variety of issues; I'm interested here only with those elements of their arguments which bear on our themes in this course. Incidentally, some of the best accounts of structuralism are critiques rather than sympathetic expositions. Beyond the material on the reading list, you might be interested in the important critique of Althusser by EP Thompson in The poverty of theory, which I'll be discussing later in the course.
I've made a distinction in previous weeks between "methodological individualist" approaches which take the individual as the starting-point for social theory and "relational" approaches which focus on the relations between individuals. We've seen that in Marx the relational emphasis derives from a conception of the individual as essentially social in nature and that in Weber what is relevant to the sociologist is action which is oriented towards the behaviour of others. In both cases, structure arises out of social interaction, geared particularly towards labour in Marx and towards meaning in Weber In structuralism, relationship takes off and becomes fully independent: it is no longer human beings who relate with each other, but the fact of relationship which first creates the social and cultural individual out of an amorphous biological mass. This is sometimes turned into a statement that we can only know the social, in other words the relational, and that "the individual" or "human nature" are therefore metaphysical concepts in the strict sense that we cannot know them. We cannot know Mary Murphy in her unique, individual internal experience (even if we believe it exists), because all we have available to us is our social interaction with her: what she says and what she does. This ultra-relationalism, in other words, leads to what is known by the slogan of "the death of the subject": either the individual literally does not exist because they are only created by social interaction, and form simply an intersection between different social relations; or the individual is methodologically unknowable because we can only know the social.
The argument that the individual literally does not exist, that they are only the intersection of social relations or the "bearers" of social structure, is argued very strongly by Althusser, who sees our belief that we are individuals to be a psychological illusion. Instead, he argues, "the category of the subject ... is the constitutive category of all ideology". Our (illusory) subjectivity generates ideology, and ideology reproduces our illusions of subjectivity.
All that we can know, then, or all that exists, is the relational. If all that we can know about is relations, then we can think about the way in which those relations interact with one another in a very detached, and often very formalistic approach. We can also try and categorise the different types of relation which are possible; Weber's four types of meaningful action are a move in this direction, and we've seen that despite his methodological individualism the concept of instrumental rationality in particular has a tendency to become dominant in his thinking.
What relationalism is likely to lead us to, in other words, is a categorisation of different types of relation and different levels of relation, and an account of society in terms of the interrelation of these different relations. This can clearly become very abstracted very rapidly (and Mills' discussion of the problems of "Grand Theory" is very relevant here). So relational approaches tend towards this kind of categorisation, but they also tend to privilege intellectual consistency over empirical usefulness. In other words, because our description of different types of social relation is likely to be quite an abstracted one if it is to be much use in telling us things we don't already know, it will be a highly intellectual account.
As we generate more of these concepts describing types and levels of relations, we are going to want to make them as consistent as possible with each other, for very valid intellectual reasons. For equally valid reasons, we are likely to want to be able to generate all of them from as restricted a number of basic concepts as possible, in other words, to generate typologies of possible variations and interrelations of particular types of relations. The net effect of all of this is that relational approaches have a tendency towards what we can properly describe as structuralist accounts, that is accounts which derive all of social reality from the operation and permutation of a limited number of basic concepts; ideally this number can be reduced to one.
Because this core concept, from which our description of society is generated, is a highly intellectual one, this is very likely to produce a form of philosophical idealism: a theory which treats the (social) world as generated from ideas, and in this case from a single idea. While there are dramatic differences in the content, the structure of our account of society is likely to be very similar whatever idea we start from; in some ways Althusser's account, not of actual modes of production but of the idea of modes of production, and Lévi-Strauss' account of culture oriented around difference, produce quite similar ways of thinking.
To finish with this general account, it has to be said that Lévi-Strauss' account privileges this approach rather more visibly, and it's worth saying something briefly about what it is. Essentially Lévi-Strauss performs two operations in his account of human culture: on the one hand he employs a linguistic analogy to treat culture, not just as a system of relations, but as a system of symbolic relations, such as myths; on the other hand, using the same analogy, he aims at a purely formal description of the various elements involved in particular myths: in other words, he sets out to describe structure but not content.
What this leads to is an argument that there is an objective meaning in human culture (revealed by structure) which is other than the subjective meaning (revealed by content). Since, however, this objective meaning cannot be straightforwardly shown to be present in a particular myth once we bracket any question of the way people say they understand it or the contexts they tell it in, it has to be located within the unconscious. In other words, from a description of social relations we move to a description of the nature of the human psyche. To complete this account, what Lévi-Strauss claims to be the central feature of the human unconscious - a claim which he believes to be backed up by linguistics - is naturally enough identical with the concept he uses to analyse the objective meaning of the form of myths; this concept is that of difference or distinction. For Lévi-Strauss, then, the end of the intellectual journey is a description of the social and in particular cultural world as a reflection of the supposed tendency of the human brain to divide things up.
There is a problem with this approach, and it is not just a difficulty with Lévi-Strauss. If we assume that the social world can be derived from an idea - the idea of the capitalist mode of production, the idea of difference - then in principle there is no possible explanation of how social change arises. The world is divided up like this because it is identical with the way the idea is organised; there is no reason why it should change. Now of course one can develop ad hoc explanations of any changes in this structure, and in practice this is very often done. Another possibility is to develop a typology of different possible types of society, so that change is simply change from one way of expressing the idea to another one. A more interesting and widely-used approach, however, is what is known as functional explanations.
Functional explanations are explanations of events in terms, not of their causes, but of their effects. For example, we might explain the fall of a government, not in terms of the events which led up to it, but in terms of what it led to. On the face of it this is simply unacceptable. The rules of logic do not allow us to reverse the flow of causality and say that an event A can be caused by an event B which has not yet happened. This form of explanation, which is known as teleological, can only make sense in one of two contexts.
One is if event A is caused by a prior event alpha, which is somebody's intention with regard to the future. We can certainly argue that a government fell because someone wanted to form a different government and thus forced the collapse of the current government. However, intention and effect are two different things: the intention to bring about event B may not in fact be realised, and our action in causing event A may have completely different results. This is generally characterised as "unintended consequences", and we saw last week that it is important in Weber's account of instrumental rationality as a method we adopt as a means to a particular goal, but which then becomes an end in itself. So an intentional explanation can only work where the person with the intention is in fact not just all-powerful but has total knowledge of the context of their action; in other words, where they are God. Much mediaeval thought is teleological in this sense: events are explained in terms of God's plans for the future of the world.
Apart from intentional explanations, there is one other form of potentially valid explanation in terms of effects, which is the argument known as functionalism, which you are no doubt familiar with in practice; it is represented, for example, by the claim that such-and-such a thing happens "because the economy needs it" or "because of the interests of capital". Note that this is quite different from an explanation in terms of the perceived needs of the economy (as seen by the government, by the electors or by individual managers). It is also different from a simple explanation in terms of compatibility: by definition, if a form of state is incompatible with a form of economic organisation, they will not coexist - but this says nothing about the reasons for their incompatibility or the mechanism which prevents their coexistence.
Strict functionalist explanations are based on an analogy to Darwinian evolutionary theory. This argues in terms of competition for survival in a situation of relative scarcity. Over immense periods of time, genetic mutations and variations will occur. Some of these will be functional for survival, in the sense that they will either enable the new individual to survive more effectively or to breed more effectively. From the point of view of genetic reproduction, of course, what matters is that a plant or animal survives long enough to reproduce itself; the better its statistical chances of survival to this point, or the more successful it is at reproduction, the more individuals with this different genetic structure there will be. Over time, in other words, functional mutations will tend to reproduce themselves and spread; less functional mutations will survive less frequently (given competition for the same food etc.) or will be outclassed in terms of reproduction.
This argument does not hold for social explanations, however, for three very important reasons. Firstly, it assumes fixed units such as individual animals; in other words, its natural affinity is with a radical methodological individualism which takes the individual (or some other unit, perhaps the family or the enterprise) not just as the starting-point but effectively as the only reality: which does not examine, for example, the social origins of the individual's ways of thinking and definition of needs, and which does not consider the possibility of interaction between, for example, the individual and the family. Secondly, it assumes that, whatever the unit is, it has a means of self-reproduction which is as exact and as stable as genetic transmission. Obviously enough, however, even when firms copy successful firms, they do not reproduce all features of the successful firms, and they cannot; all they do is import what they perceive to be the important features. So we can think of a general diffusion, for example of instrumental rationality, which is intentional in nature: people think that it is likely to be effective, and it happens that they are right. But we cannot say that this is a functional process. The continuing history of Anglo-American interest in Japanese management methods is a sufficient example of this: "Japanese management" is not a single fixed entity like a collection of genes, but is transmitted as a series of what may be very differing assumptions about its key elements; just as importantly, Anglo-American workers and managers and Japanese workers and managers have different cultural backgrounds, so that the assumption that the firm is a unit which is not influenced by other social realities falls.
Thirdly, and perhaps most importantly, not only do we not have straightforward units, and not only can they not reproduce themselves in a simple fashion, but we have to say that the Darwinian argument of the "survival of the fittest" can only be a metaphoric one when it is applied to society. We could not live like that, even if we felt it was desirable. This can be seen very clearly at the level of "societies": contemporary societies are not disputing a common living space; in fact the economically dominant societies are experiencing a population decline. More generally, conflict between contemporary societies is only very rarely expansionist; even where it is, it is generally a matter of the imposition of a new form of government, but not of the obliteration of the previously-existing society. Even where this is the case, as for example in the population movements of the Migration period around the fall of the Roman Empire, "functionality" is a fairly ambiguous concept. The societies which expanded into the declining Roman empire were not in general technologically superior to the Romans, or even necessarily economically superior: indeed, their need to migrate may be seen as evidence of the problems they experienced in maintaining their way of life in the regions they originated from. Their superiority was partly demographic, and partly military. In other words, "functionality" in these terms is almost entirely destructive, and tells us very little about features of economic or social organisation.
I've shown how radical relationalism leads to structuralism as a holistic account of society, and also indicated the well-known difficulty that structuralism has with explaining change. The last feature of holism that is worth mentioning here is the concept of modernity expressed in structuralism. This will be fairly brief, because - while structuralism is strongly modernist in its approach, it does not treat modernity as a key term: it is itself modern, but it is not very interested in the specificity of the modern. There are obvious reasons for this: if society consists of a structure of relations deriving from a single key concept, it is hard to see how we can have dramatically different types of society.
This is a problem for Lévi-Strauss, who derives the organisation of culture from the biological structure of the unconscious brain, in other words from something which, if it changes at all, does so over enormously long periods of time. Unsurprisingly, Lévi-Strauss' work, as was at the time the general practice among anthropologists, was largely devoted to the study of what were seen as "traditional" societies, and his concept of the modern is largely defined against these. To an extent, it seems that he treats the modern as an aberration, an unnatural separation of culture and nature, and doomed to destruction for that reason. This may be appealing as a political position, but it does not really deal with the problem, and later structuralists have tried to show that modern culture can also be analysed in the terms Lévi-Strauss uses for "traditional" culture.
Althusser, by contrast, fits modernity into a static typology in which it is effectively simply one variant on a pattern. This derives from his version of Marxism, which replaces the crude version of economic determinism found in vulgar Marxism - the argument that everything else can simply be reduced to the economic - with a more sophisticated analysis of different levels of social life, including the economic, the political and the ideological. Each of these, for Althusser, can be described as "relatively autonomous": in other words, it has a logic of its own, and cannot simply be reduced to the economic. Thus Althusser's model of the social totality is that of a "decentred whole". However, the economic is "determinant in the last instance", in other words, it has the final say. Since "the last instance never comes", though, it is the interaction which is most important. Incidentally, this tension between "determination in the last instance" and the insistence that "the last instance never comes" is one of the major theoretical problems of Althusser's holism.
Determination by the economic level expresses itself primarily in the creation of these separate levels and the prioritising of one or the other at different historical periods. In other words, within a given mode of production, it is the economic level which determines which level is dominant in a more immediate sense. In feudalism, the political and ideological levels are dominant; in capitalism, it is the economic level itself which is dominant. In both cases, however, the economic level is ultimately determinant; in other words, it determines whether it will itself be dominant or whether some other level will be.
This makes a certain kind of sense: the economic (for these purposes) can be thought of in terms of the relations of ownership and control. In feudalism, the lord owns the land, but the peasant controls their agricultural production; so the appropriation of the peasants' surplus production by the nobility does not take place within the actual process of production, but as an effect of political or ideological structures which guarantee this transfer. In capitalism, on the other hand, the means of production are both owned and controlled by the capitalist; thus the appropriation of surplus value takes place within the process of production; the society is therefore said to be dominated by the economic. For Althusser, in other words, the difference between modern and other societies is that they represent different possible arrangements of the ownership - control situation.
Thus structuralism is unable to do anything very interesting with the idea of modernity, or indeed of social change more generally; it tends to reduce history either to contingent change without any real meaning or to variations on a theme.
As with social change, so with social movements structuralism has remarkably little to contribute. This derives partly from the "death of the subject": if human agency is simply an illusion, then social movements can be explained either in terms of a functional contribution to social change (particularly in the case of the workers' movement) or, more commonly, as an ideological reaction against social change (particularly in the case of the new social movements).
This weakness derives partly from the interaction of theory with social movements themselves: a good example here is Althusser. Althusser was a member of the French Communist Party (PCF), which was perhaps the most immobile of the major Communist Parties of western Europe, and an organisation which could perhaps be described less as the political wing of the workers' movement than as the congealed wing of the workers' movement. French workers, throughout the twentieth century, participated in a number of extremely radical actions - general strikes, mass occupations of factories, the French Resistance. At the same time, the Communists were by far the most important working-class party, and indeed controlled the main trade union federation and the greater part of the French Resistance. However, the Party was not just a particularly orthodox and dogmatic one - by contrast with the Italian Communist Party, for example - it was also committed to the view, for most of its history, that radical change was simply not on the cards in France, partly for reasons of economic organisation but particularly because of the post-war situation where, on the one hand, it saw that a take-over of power would be likely to be crushed militarily by the Western Allies, and where, on the other hand, de Gaulle's semi-independent foreign policy, which led to the French armed forces, for example, not being under NATO control, was felt to be the best that could be hoped for. The net result was that, in 1944 - 45, instead of turning the predominantly Communist Résistance into an attempt at taking power, as happened in countries like Yugoslavia, the Party accepted the political realities reflected by the Western Allies' support for de Gaulle. Similarly, in 1968, the Communist-dominated CGT was instrumental in keeping the general strike under control, physically excluding student radicals from the factories where they might have undermined the Party's position, and deradicalising the demands of the strike. In France in particular, 1968 was almost as much a rebellion of the libertarian Left against the PCF as it was a challenge to the state. It is therefore not very surprising to find a Party philosopher keen to exclude any possibility that human agency could actually make a significant difference.
More generally, Althusser's structuralist Marxism can be seen as the logical development of one strand, but only one strand, of Marx's thinking. The other approach, which we will be looking at next week, has come to be described as "Western Marxism", and is more closely associated with social movements and activist parties; the kind of static Marxism practiced by Althusser is associated with parties who are either in power or who, for other reasons, are keen to minimise the possibility of large-scale social action which is not entirely under their control.
Structuralism is also of interest in terms of its notion of rationality, or, as it is more usually phrased, its claim to scientific status. Sometimes, this represents a pure positivism in terms of its research methods: the "social facts" are assumed to be out there, to be amenable to pure observation, and analysed on the model of natural science. This kind of thing happens to any theory, and it's not a fault peculiar to structuralist practice. What is rather more interesting is the rationalist version of science represented in much structuralist thought.
We normally assume, when we hear the word "science" in English, that it refers to the natural sciences, or to methods which are based on those of the natural sciences. What generally lies behind this is what we can loosely call an empiricist model of science: science as taking its starting-point from what is believed to be empirical reality, which literally means the reality available to the senses. We can observe and experiment with this reality and attempt to build up valid generalisations about its behaviour. In sociology, this is what is normally meant by arguments about "sociology as a science"; what is commonly argued against it is that the reality we experience is already structured by ideas, such as the idea of time, and that social reality is already mediated by the forms of social interaction, such as language. In each case, it is said, we cannot have a "pure" or unproblematic knowledge of reality.
This "empiricist" model of science can, however, be contrasted with a "rationalist" model of science, which argues that our knowledge of the world is, at least initially, a mental one rather than a real one; the implication being that, in one way or another, we can know reality through thought alone. This programme takes an enormous variety of forms, but two elements are fairly constant. The first is that, at the end of the day, the most important thing is to think systematically and consistently. The second is that, in general, we will tend to look for a hidden reality underlying and explaining the observable world. In explanation, these two emphases tend to take precedence over what we might call faithfulness to the world as observed or experienced. The latter is pressed into consistency, or the elements which do not fit are discarded. It is summarily "explained" in terms of what are claimed to be the "real", underlying truths of the situation.
Many authors in practice combine elements of both these approaches, and it may be difficult not to: a fairly common-sense understanding of social theory, after all, would say that it aims both at internal consistency and at being an adequate account or explanation of the observed world. But if the two of these are pulling in different directions - if we claim, as does Lévi-Strauss, that the real world is unobservable, because it is unconscious, for example - then we will have to come down on one side or the other. Structuralism's claim to be scientific generally comes down on the side of rationalism, in other words of aiming at being systematic and aiming to uncover a hidden reality. Incidentally, this sense of the word "science" is rather more widespread in Continental languages, which are capable of describing literary criticism, theology and so on as "sciences": what is meant is not that they represent an equivalent to physics or chemistry, but that they are systematic in approach. If we add that the hidden reality which is aimed at or discovered is likely to be a mental one (given that the rationalist is explicitly taking their own thought as the starting-point or indeed the totality of all that is known), we can see the fit between this model of scientific rationalism and structuralism as a systematic ordering of mental categories.
Althusser's scientific rationalism is in some ways even more thorough-going than Lévi-Strauss's: while he claims that there is a real world out there to which theory in some sense corresponds, scientific method has absolutely no need of empirical verification. Martin Jay summarises Althusser's conception of science very well:
"Science, he claimed, operates on the level of conceptual production in which experimental verification plays no role; it is nonetheless materialist because it posits an ultimate congruence between thought objects and a real world. The raw material for scientific activity is provided by ideological conceptions of the world, the 'facts' that positivists innocently take as the givens of existence." (Marxism and totality, p.401). In other words, scientific activity consists of the progressive refining, rethinking and systematising of everyday ("ideological") knowledge of the world; in Althusser's own practice this takes the form of a scholastic project in which an ever-decreasing selection of Marx's work is examined and rethought in order to produce what is presumably an ever-purer form of scientific knowledge.
The last point that I want to mention in this lecture is, perhaps, a minor one, but it is unusual and perhaps worth bearing in mind. Lévi-Strauss, like many subsequent authors, argues for a close analogy between culture and language. Obviously this can mean virtually anything, depending on what we understand the nature of language to be, and Simon Clarke has argued that Lévi-Strauss's concept of language does not correspond closely with how linguists either then or now thought about it. One element of Lévi-Strauss' linguistic analogy is the argument that we can distinguish between the form and the content of a culture, or of a language. Just as a language (for Lévi-Strauss) exists as a number of elements related in particular ways (form) which we can then use to express particular meanings (content), so culture is fundamentally a form, within which different contents can be expressed. In other words, while myths (for example) may express a particular meaning to the people who actually tell them or hear them, this meaning is expressed within and determined by the broader form of myths structured around difference. What Lévi-Strauss deduces from this is a form of uncertainty principle. Whereas in physics we may be in a position where we can measure light as a particle, or as a wave, but not as both simultaneously, in anthropology or sociology, for Lévi-Strauss, we can know the content of a culture, or its form, but not both at the same time. In other words, we can think about the way a culture is structured, and indeed about the way culture as a whole is structured; this is like thinking about the syntactical structure of language. And like thinking about the syntax of a language, it cannot be done at the same time as thinking about the content or meaning. This means that we can study the structure of a culture with one method, or we can study the actual cultural meanings which are expressed with another method, but we cannot study both at the same time, because a study of actual meanings presupposes a knowledge of the structure with which those meanings are expressed. We can think about current meanings within our culture; we can stand back and think about our culture as a whole. We can even think about a foreign culture. But we cannot think about the meanings expressed within a foreign culture, because we lack the necessary knowledge of its structure.
There are a number of problems with this claim, which I won't disentangle for you; but it is worth thinking about as an unusual approach to the problem of reflexivity.
Next lecture - Previous lecture