its title—and manifesto—to the volume. For ‘The Poverty of Theory—or an Orrery of Errors’ is not only a polemic against Althusser: it is also the most sustained exposition of Thompson’s own credo, as a historian and as a socialist, that he has given us to date. The purpose of the essay here, then, will be threefold. It will look at Thompson’s criticisms of Althusser, and try to determine their justice. Simultaneously, and more importantly, it will seek to elicit some of the cruces of Thompson’s substantive work, through the grill of the principles and procedures he recommends in The Poverty of Theory.3 The treatment of Althusser, starting moderately and ending in a gale of fury, is unconventional in organization. Discussion of it will be facilitated by some regroupment of its themes, for a more compact comment on them. The Poverty of Theory is, in effect, dominated by four main problems: the character of historical inquiry, the role of human agency in history, the nature and fate of Marxism, and the phenomenon of Stalinism. I will consider each of these in turn, as they present themselves in Thompson’s criticisms of Althusser and in his own practice as a historian; and in conclusion will try to put Thompson’s work into a comparative context, capable of clarifying in some degree the differences which have arisen between him and New Left Review, a journal he took a central part in creating. Whatever our view of specific arguments in The Poverty of Theory, the enterprise itself must be welcomed. It represents the first full-scale engagement by an English historian with a major philosophical system from the continent, over the terrain of Marxism. A direct encounter between the two different discursive traditions represented by Thompson and Althusser has for some time now been much needed, for the development of historical materialism as a whole.4 It is to Thompson’s credit that he has undertaken the task, initiating a process of exchange that we must hope will eventually be many-sided.
1.Past and Present, No 38, 1967, pp. 56-97.
2.London, 1978.
3.References to the latter will henceforward be abbreviated to PT; The Making of the English Working Class (Penguin edition: 1963) to MEWC; Whigs and Hunters (1973) to WH; William Morris—Romantic to Revolutionary (1977 re-edition) to WM.
4.See the remarks in Considerations on Western Marxism, London 1976, pp. 111-112.
The opening sections of The Poverty of Theory are addressed to certain general issues of historiography as a discipline. Three distinct problems are explored by Thompson, which can be formulated as follows: (i) what is the particular nature and place of evidence in any historical inquiry? (ii) what are the appropriate concepts for the understanding of historical processes? (iii) what is the distinctive object of historical knowledge? In each case, Thompson evokes and rejects what he takes to be Althusser’s answer, and proposes his own solution. He begins his case with the charge that Althusser’s epistemology exhibits a radical indifference towards the primary data which make up what it terms Generalities I: no explanation or attention is ever given to either the character of these data, or their origins—chief among which is ‘experience’. Althusser’s cavalier attitude towards empirical facts is confirmed by his account of Generalities II, or the process of cognition itself, which in effect assumes that any scientific theory can define and produce its own facts by self-validating protocols, without recourse to external appeals. Thompson argues that this is an abusive extension of the very limited and exceptional procedures of mathematics or logic, that is wholly illegitimate if applied to either the social or physical sciences, where the controls of evidence are always central. The result is that no genuine new knowledge can emerge in Althusser’s Generalities III (its ostensible site), since Generalities II has already pre-packaged the data of Generalities I anyway—there is an epistemological circle. The result is ‘exactly what has commonly been designated, in the Marxist tradition, as idealism’1—that is, ‘a self-generating conceptual universe which imposes its own identity upon the phenomena of material and social existence, rather than engaging in continual dialogue with them’.2
What is the justice of these charges? In my view, a great deal. Althusser’s theory of knowledge—both of science and of ideology—is, as I have argued elsewhere, directly tributary to that of Spinoza.3 It is not surprising that an epistemology with this metaphysical background should be incompatible with the canons of modern science. Lucio Colletti once remarked: ‘One could say that there are two main traditions in Western philosophy in this respect: one that descends from Spinoza and Hegel, and the other from Hume and Kant. These two lines of development are profoundly divergent. For any theory that takes science as the sole form of real knowledge, there can be no question that the tradition of Hume-Kant must be given priority and preference over that of Spinoza-Hegel’.4 The broad truth of this claim is incontrovertible. In the case in hand, there is no doubt whatever that Althusser displays no interest in the (diverse) origin and nature of Generalities I, within his schema. In one respect Thompson even goes too far towards him, when he casually supposes that ‘sense-perception’ is not ‘knowledge’.5 In fact, certain kinds of perceptual experience—the sense-data with which radical empiricism from Hume onwards has always been so preoccupied—do not need transformation by any Generalities II to yield knowledge: they constitute an elementary form of knowledge in themselves, without further ado (for example, what is the weather like?). Althusser’s system wrongly assimilates knowledge to science tout court—an inaugural slip far from trivial in its consequences: the ultimate sources of his insensibility towards evidence lie here. Thompson is certainly right to indict this. On the other hand, his bracing attack on the notion that primary historical facts are in some sense typically ‘rigged’ or ‘pre-selected’ by the intention of those who left them behind6 is germane to Popper, who has advanced this absurd contention, but not to Althusser, who has never done so. An argument salutary in itself is here misused to suggest guilt by association. Similarly, Thompson condemns with every justification two English sociologists, Hirst and Hindess,7 for their dictum that ‘facts are never given, they are always produced’, but fails to note that the work from which he quotes precisely attacks Althusser for ‘empiricism’, and hence can scarcely be regarded as a stand-in for the latter.
In constructing an eloquent and necessary general defence of the historian’s craft, Thompson in effect too often proceeds to an amalgamation of individual positions, each of them deficient, but in significantly different degrees and ways. Thus Althusser does indeed reply improperly on logico-mathematical protocols of proof as models of scientific procedure. His theory of knowledge, dissociated from the controls of evidence, is untenably internalist: above all, it lacks any concept of falsification. Vice versa, however, the strength of Popper’s philosophy of science—one is not sure whether Thompson realizes how strong it is—has always lain precisely in its insistence on falsifiability, a principle crucially qualified by Lakatos and others, but uncompromised by Popper’s egregious illusions about historical records. The hostility which Thompson senses in the two philosophers to the
