extensive. In medieval political thought, kings had two bodies reflecting their political and spiritual status. By contrast, corporations had legal personalities, but only fictive bodies, while slaves had bodies but not persons.

Leaving aside the question of how ideology constitutes collective agents, and adopting Therborn's frame of reference that the theory of ideology is concerned with the human subject, one may accept the logic of what he sets out to do in his classification of ideologies of the subject and still find the account somewhat incomplete and ambiguous. Because Therborn appears to take for granted the unity of body and subject, he does not consider, for example, how disease theories as medical ideologies fit into his model of interpellation. As Foucault has reminded us, medical classificatory schemas have enormous political significance. But are these addressed to diseases, bodies or persons? The debate about disease, illness behaviour and deviance comes eventually to the problem of the moral responsibility of the individual, and thus to the 'cause' and 'motives' of behaviour. However, it would be difficult to know where to locate, for example, the sociological notion of 'vocabularies of motive' within Therborn's categorization. Such vocabularies are not precisely elements of 'inclusive-existential ideologies', since they do not locate persons as members of the world; they simply specify what is to count as acceptable behaviour. This raises another issue concerning the classification of ideologies of the subject: there appears to be considerable and unclear overlap between boxes 1 and 4, and 2 and 3 in his table. It is not obvious, for example, why membership of a tribe (inclusive-historical) should be significantly different from membership of a system of tribes (positional-historical).

Therborn's approach to ideology represents a decisive move away from the problem of the falsity of ideological beliefs to the problem of possibility -- what are the possibilities of subject construction? Therborn's work, like our The Dominant Ideology Thesis, is thus less concerned with questions of legitimation and incorporation and more concerned with the question of possibility. However, what he does not ask is: what are the variations in the effectivity of ideological systems, given differences in their apparatus, in establishing the possible? Such an omission is odd given the title of the work, and as a result it is never made explicit what the power of ideology actually is. What is clear is that, for Therborn, ideology is a very important social force. As he himself indicates, there is a definite Althusserian imprint here. Indeed, his conception could almost be described in Althusser's words: 'Human societies secrete ideology as the very element and atmosphere indispensable to their historical respiration and life', ⁶ and, more specifically, 'ideology (as a system of mass representations) is indispensable in any

-159-