The analysis of ideologies and forms of knowledge and belief is in a state of disorder. In contemporary Marxism, the autonomy and independent importance of ideology have been stressed at the expense of a discredited economic reductionism. In many ways this is a desirable development, although, as we have pointed out elsewhere,footnote1 it also carries with it some very misleading consequences. However, the critical problem that contemporary Marxist theories of ideology have to face is: how is one to reconcile materialism with the autonomy of ideology? This implies a second difficulty: namely, how is one to reconcile the notion of ideology as critique with a general theory of ideology? In terms of disciplinary definitions, there is a parallel question about the relationship of the Marxist theory of ideology to the sociology of knowledge which developed in opposition to classical Marxism.

The significance of these problems is nicely illustrated by Therborn’s The Ideology of Power and the Power of Ideology,footnote2 in which he attempts to clarify a variety of theoretical issues in contemporary Marxism and sociology. He conceives his project as taking ‘Marx’s insights as a point of departure for an attempt at a more systematic theory’ (p. 41). Elsewhere he suggests that Marxism has a great deal to learn from the empirical findings of sociology, and in our view his own attempt to generate a new theory of ideology can also be seen as an attempt to synthesize a sociological perspective with Marxism. This is a most interesting project. Nevertheless, there is clearly a wide variety of possible destinations even if one takes Marx as one’s point of departure, since one can as easily end outside the Marxist tradition as within it, nor need the terminus be a theory that is systematic or general.

Therborn rejects the notion that ideology involves beliefs in people’s heads, specifically beliefs that are false or mystified or misconstrued. He further denies that ideology is the opposite of science. Ideologies are defined as all social (in distinction to psychological) phenomena of a discursive (in distinction to non-discursive) nature. They include ‘both everyday notions and “experience” and elaborate intellectual doctrines, both the “consciousness” of social actors and the institutionalized thought-systems and discourses of a given society’ (p. 2). This is deliberately a broad definition, and one that in our view effectively reproduces the sociological notion of ‘culture’. Following Althusser, Therborn suggests: ‘The operation of ideology in human life basically involves the constitution and patterning of how human beings live their lives as conscious, reflecting initiators of acts in a structured, meaningful world. Ideology operates as discourse, addressing or, as Althusser puts it, interpellating human beings as subjects’ (p. 15). This operation of ideology involves two processes: the constitution and subjection of human, conscious agents and their qualification to fulfil their positions in society. Therborn recognizes that an analysis of ideology in terms of inserting agents in their places is analogous to the traditional sociological analysis of social roles, but he maintains that traditional role analysis is too subjectivist. The main burden of ideology is to construct human subjectivity, so that ‘to search for the structure of the ideological universe is to seek the dimensions of human subjectivity’ (p. 17). These dimensions form ‘a property space’:

Ideologies thus situate individuals in time and space by reference to personal, positional and social characteristics.

Therborn sees ideologies as being materially determined, and the definition of materialism is deliberately and unusually broad to encompass ‘the structure of a given society and . . . its relationship to its natural environment and to other societies’ (p. 43). Materialism in the classical Marxist usage of the economic structure, is used to explain the determination of one specific ideological set which appears to comprise those class ideologies required for the subjection and qualification of economic agents, though Therborn’s presentation is not clear on this point. He states explicitly, however: ‘Any given combination of forces and relations of production of course requires a particular form of ideological subjection of the economic subjects . . .’ (p. 47).

It is noteworthy that Therborn does not accept the contention, familiar from many classical Marxist accounts of ideology, that the principal function of ideology is to incorporate subordinates, to act as ‘social cement’. He argues, by contrast, that subordinates will adhere to alterideologies which are oppositional, and he attempts to specify the conditions under which those alter-ideologies may arise. There are three possible explanations. The first and most general explanation, which Therborn emphasizes, is that, by its very nature, every positional ideology must generate an alter-ideology in the process of generating differences between self and other, us and them. These ideologies have thus ‘an intrinsically dual character’ (p. 27) and the implication is that any ideology of domination must generate resistance in the very act of setting up a Self/Other opposition. Such an argument links Therborn’s position directly to that of current structural linguistics in that language subsists on the play of differences. A difficulty with the notion that the imposition of knowledge/ideology produces resistance is to show exactly how this comes about, and, more importantly, under what conditions resistance prevails—a difficulty manifest also in Foucault. Secondly, Therborn refers to the fact that class ideologies ‘are inscribed in the relations of production’ (p. 61). For example, feudalism involved a hierarchy of rights and obligations between peasant and landlord, and these were the foci of class struggle. Curtailment of peasant rights created alter-ideological conceptions of injustice that were the basis of peasant oppositions to the illegality of landlords’ activities. In one place he also talks of ‘the irreducibility of psychodynamic processes to complete social control’, which creates ‘a small margin of individual “misfits”’ (p. 43). Thus it would seem that interpellation can never really be effective, as ideologies have an inherently dialectical character, while complex social processes mean that ‘ideologies overlap, compete and clash, drown or reinforce each other’ (p. vii). Indeed, ideologies actually operate ‘in a state of disorder’ (p. 77), so it is not surprising that ideological theory is itself disordered.

On the subject of class ideologies and alter-ideologies, which have mainly concerned Marxists and sociologists alike, Therborn has a number of comments. He suggests that class ideologies are typically core themes rather than elaborated forms of discourse; that they can only be derived theoretically, seemingly on the basis of the imputed functional requirements of a mode of production; that non-class ideologies are not reducible to class but are class patterned or over-determined; and that class ideologies have to compete with and relate to non-class positional ideologies such as nationalism and religion. His brief analysis of nationalism and religion shows that the former is class patterned in different ways in different societies, while the latter seems scarcely patterned at all. The two-by-two matrix of the universe of ideological interpellations given above makes clear that class ideologies fall mainly into cell 4, with some dimensions in cell 2, and that they constitute a small part of the total population with which Therborn’s theory is concerned.