Poulantzas, N.  (1975)  'The Capitalist State: A Reply To Miliband and Laclau', in New Left Review 95

The debate is worth renewing after the publication of Fascism and Dictatorship [Fasicsm...] , and Classes in Contemporary Capitalism [Classes...]. Other people have joined in as well. The debate is not best seen as instrumentalism versus structuralism, however. It is still difficult to reply to Miliband because he has no problematic and therefore no real concrete alternative. Laclau is right to insist that facts do need an alternative theoretical problematic if they are to confront theoretical work. Poulantzas has also rectified some of the errors in his earlier work, but after a self criticism of his theoretical position. In particular, he has referred to concrete issues, but after developing a proper understanding of the concrete. In replying to his critics it is important to avoid  'demagoguery of "common-sense"... [and]... the "illusion of the evident"' (66, apparently citing Durkheim!).

He admits to a certain theoreticism as a result of being influenced by Althusser. This arises because of Althusser's emphasis on the process of theorising  [the  'Generalities' model, which sees theory as a production process at level two, operating on existing facts and opinions at level one, and producing adequate theoretical object at level 3]. Obviously the real, including practices, also affects theory  prior to level one and after level 3, but this is underemphasized in Althusser, producing a misleading view that therefore it is only discourse that creates validity  [Laclau's view]. The whole process is glossed in the Althusserian notion of  'theoretical practice', which has been now abandoned anyway. There is a need to restore the concrete real/practice. This theoreticism lingered in Poulantzas's early work, exacerbated by his failure to clarify that there was a difference between the order of exposition of his work and the order of research itself  [a distinction also rooted in Marx's own remarks about his work]. The order of exposition made it seem as if his work was apparently focused on the theoretical rather than the real -- hence Miliband's view that there was no research of the concrete and the real, and no role for them in the development of theory.

Poulantzas also admits a certain formalism, as in Laclau's attack. He is now fully repentant, and his book on fascism offers a detailed analysis of German and Italian fascism, while Classes... does refer to a range of empirical material. It is still necessary to insist that the theoretical treatment of this material is essential, however, and that the concrete is produced as a  'unity of a multiplicity of determinations' (67, quoting Marx). In this sense, there is no real concrete analysis in Miliband, only a narrative description of what is the case.

The emphases in his early work should be seen as deriving from the context of a struggle with positivist marxism, with mechanism, empiricism, and economism, and especially, the lack of marxist analysis of the state. His work therefore  'bent the stick' too far in the other direction  (68 -- NB the phrase is Lenin's, not Stuart Hall's at all!). There was also a need to react to official marxist accounts of the events of 1968, which tended to be dominated by reformism. The changes since 1968 are very important for Poulantzas, but not for Miliband it seems. The changes show how real history had a personal impact on Poulantzas's theories, and this is how they actually work. Miliband is unable to grasp this because he sees a role for [social classes as] 'positivist empirical positions', unaffected by facts. In this way, his apparent empiricism conceals a dogma.

Poulantzas admits that he has used difficult language, although occasionally a theorist must break from ordinary descriptive discourse. Nevertheless, his work has been academicist. The necessary relation to political problems as he saw them was not discussed. As an example, Miliband's criticism of the circularity of his argument on the 'pertinent effects' of class really referred to the combination of distinctiveness, autonomy andpolitical ineefectualness of a particular class. This was meant to refer to the actual example of the peasantry under Napoleon III. The discussion is mostly about the non-fundamental classes, but when applied to the working-class there were political objectives behind the account as well: (a) to deny that the working class had ever been fully incorporated, even if there was no evidence of revolutionary class consciousness. The working-class could still agitate for important reforms. Indeed, social democracy itself is the result of working class 'pertinent effects' (69); (b) but social reformism is ineffectual, not in the long term strategic interest of the working classes. At least Poulantzas attempts to discuss the standpoint of the working-class, unlike Miliband.

Miliband does not define exactly what he means by 'structuralism'. The term seems to operate within a bourgeois humanist problematic about the dangers of ignoring concrete individuals and subjectivity. Poulantzas has nothing to add to the debate here, because he simply rejects this problematic. The issue for him is really choosing between idealism or materialism. Idealism is always at the heart of bourgeois objections to marxism. 'Structuralism' sometimes refers to an approach that neglects class struggle as well [as in the British gramscian criticisms of Althusser. Here, the alternative is sometimes rendered as  'culturalism']. However, Poulantzas has focused clearly on class struggle. Finally, the accusation of  'structuralism' has taken on a polemical force among some American critics, again relying on some  'factual and empirical critique' (76), and has led to some propositions for a compromise between structualism and instrumentalism, some 'Third Way' [!].

The relative autonomy of the state is addressed, but discussed in terms of the relative autonomy of the different levels  [e/p/i -- see diagram in notes on Laclau]. The state's autonomy is also described in terms of the specificity of class struggle, and how classes are not the same as power blocs, leading to a role for the state as an organiser and unifier. This role is itself a function of the separation of  the economic and political, and that separation in turn is the result of class struggle. The issue really is to decide how relative is the autonomy of the state. It is impossible to get a general answer, since although there are general negative limits to state autonomy, the actual conjuncture itself is the most important factor.

State apparatuses are not the same state power, argues Miliband, and the state must have some independent power if it is to be relatively autonomous. This is another  'appeal to common-sense' (73), however. It seems based on standard sociological or Weberian notions that the state holds power and groups compete for it. This is also a view taken of many organisations in sociology. In reality, class struggles are prior. This only limits the autonomy of the state if power is conceived in Weberian terms [that is the ability to compel another to act against their will]. The practice of the state arises from its institutional specificity rather than its autonomy in this bourgeois sense. The state cannot simply directly express economic corporate interests  [Gramsci is cited here]. It must represent the unity of a bloc under the hegemony of a class or class fraction. Specific post holders and their backgrounds are irrelevant to this argument.

In Classes..., class struggle is given its proper place. The state is neither a thing/instrument nor a subject. It does not confront classes, nor does it resist class-action nor referee between classes. Its basis lies in the separation of the economic and political. It is necessarily contradictory, and one requires precise analysis of the actual mechanisms of the state, its bureaucracy and elites and how they work, which is only just begun in marxism. The overall role of the state is to reproduce class divisions so it cannot be a simple monolith. It is contradictory, divided, for example between parliament, the administration, and the army. The chaos and negotiation between these elements is a result of  'structural selectivity'-- that process decides the contradictions which have to be managed, through compensation or through filtering different interests (75). Poulantzas has indeed specified the differences between different types of bourgeois state, and has not blurred differences between fascism and social democracy. His work discusses the typology of exceptional States, including fascism as a special case.

Laclau has accused him of formalism, and a tendency towards taxonomy. The e/p/i model features undiscussed and a priori notions of three levels, which must then combine to produce specific cases [this  'combinatory' model has been likened to functionalism -- certain invariant 'functional prerequisites' simply and always take different specific forms. As with functionalist analysis, this leaves almost no need for any kind of concrete investigation at all]. Laclau argues that  'the economic' is described ambiguously, and that relative autonomy is specific to capitalism and not a universal theory of the state. There is some strength in these criticisms, but they are generalised across the work of Poulantzas, Althusser, and Balibar. Balibar in particular has recognised the force of the critique and substantially changed his views. However, Balibar was the most formalist and economist, and the most likely to underestimate class struggle. In his work, the economy remained as some invariant and self regulating process, and this was indeed generalised to the other levels as well: all alike were seen as possessing some invariant core, with some specific combinations and aspects. The role of class struggle was minimised, partly because there was not enough emphasis placed on the difference between a social formation and a mode of production. However, Poulantzas always saw the mode of production as a matter of specific combinations, with no invariants, as a complex unity of determinations, deeply affected by class struggle.

Nevertheless, the early Poulantzas work was formalist to some extent. The mode of production does determine ultimately, and articulates the other levels. However this was not meant to be a general theory, but was always only about the capitalist state. The social formation was never equated with the mode of production, and Poulantzas advocated the investigation of specific conjunctures rather than the development of formal taxonomies of combinations. However, the early works still saw a general theory as possible and legitimate: he now realises that that is impossible because there are only ever very different conjunctures produced. He admits that he also sometimes did see social forms as concretisations of the mode of production, especially with absolutist states.

The notion that relative autonomy of the levels is general and invariable was argued by Althusser and Balibar, and appears even in the ISAs essay. Poulantzas has made his own criticisms of this essay  in Fascism... Relative autonomy is specific and conjunctural for him in capitalism, and not just another combinatory. However, it was a mistake to generalise back to pre-capitalist societies --  'a classic error of historical hindsight' (81). Are the levels necessarily partitioned? Poulantzas has always recognised that the state does intervene in the economy, and takes a special form of intervention now in monopoly capitalism  (although the two are not fused as in stamocap). Similarly, the classes are determined by political and ideological elements as well as economic ones, but these are found 'within' the economic classes.

The latest work on classes as a series of practices is intended to be anti-structuralist. Social classes are objectively determined, but they are not  'ontological and nominalist entities, but... only exist within and through the class struggle (practices)' (82). There must be class struggle, not just a system of classes as in bourgeois sociology. Structures and practices are not to be seen as two distinct domains here, though. In Classes... class practices are seen as occupying a single field, but structural class determinations take place within that field. So, for example, the so-called aristocracy of labour is still structurally a layer of the working class, with its own distinctive class practices underneath the bourgeois veneer. The same goes for the petty bourgeoisie [see file on Classes...].

Overall, the debate with Miliband and Laclau has been useful, but it has also been falsely schematized, its complexity has been reduced, and it has been far too personalised. His own position has evolved as a result of debates with these and other colleagues, however.

back to main menu