This essay concerns the significance of ‘human sensuous activity’—what has become known variously as ‘praxis’ to many Marxists and ‘action’ to analytic and phenomenological philosophers.footnote footnote1 Put grandly, our thesis is that it is from labour, and not from language or thought, that the category of meaning arises. That is to say that a logically necessary foundation for agreement in what people say and mean is to be found in what they do—interpreted in a broad and not specifically economic sense which we shall clarify. Most of this essay will not be concerned with the more common problematics of Marxism, but we would argue that the topic we are addressing is crucial to the philosophical coherence of Marxist theory. Our subject matter can be described from a Marxist perspective as the division of manual from mental labour. However, we shall be discussing this division in ontological rather than historical terms. We will argue that labour possesses the property of being intelligible, prior to the intelligibility of language and thought, and that this must be so in order that the structures of meaning contained in the latter can arise. The view that such structures of meaning are created on the foundation of human labour is not one that most Marxists would dispute. What is dubious, however, is the significance and coherence of this historical thesis if labour is construed, ontologically, as a secondary construct: an amalgam of the logically and ontologically prior categories of physical activity and mental or linguistic meaning, respectively. While some Marxists might also reject the construal of labour as a secondary construct, it is noteworthy that others would not, quoting in support of their view the contrast Marx makes between the activity of the spider or the bee and that of ‘even the worst architect . . . (who) . . . raises his structure in imagination before he erects it in reality’.footnote2

Marx arrived at his notion of labour as a result of his critique of Hegel’s idealism. The arguments we wish to advance come to similar conclusions but by a different route. They arise from a critique of one of the most common dilemmas of modern non-Marxist thought, namely that associated with ‘linguistic’ idealism. This has made its most dramatic appearance in the guise of relativism which came to the fore in many fields in the sixties, all of which supposed that we inhabit the world we are able to describe rather than, as Berkeley in some ways similarly supposed, the world we are able to perceive. As an intellectual fad, relativism is now something of a spent force, but not we think because the underlying issue on which it capitalized has been squarely faced. Indeed, this was hardly recognized by most of the protagonists in the disputes that relativism engendered. In every case the substantive relativism that was proposed exploited the recalcitrance of the category of meaning in resisting objective determination.

It might not seem immediately clear that this is so. Although two trends can be discerned within linguistic idealism which appear to be sharply at odds, they are actually united on just the point in question. One trend, of which Winch and Feyerabend are notable exponents, indulges with relish in the apparently ‘subjective’ aspects of linguistic idealism, tilting at pompous proponents of ‘Western scientific rationality’. By contrast, it is because meaning does lack any objective external determination that ‘structuralist’ and ‘post-structuralist’ writers (e.g. Saussure, Levi-Strauss, Althusser, Foucault, Derrida, Kristeva, and their followers) rest their methods on the claim that structures of meaning in various ways provide their own objective determination from within. In order that such structures can be conceived as being capable of such a feat, it must be supposed that they exist independently of the subjectivity of any person. The individual who uses the code that the structure supplies—in order to say or mean something—is seen, therefore, as wholly secondary to the structure itself (whatever form it is supposed to take) and to have little or no epistemological significance.footnote3 Indeed, at one extreme, epistemology is itself dismissed, and it is claimed that ‘. . . the entities referred to in discourse are constituted solely in and through the forms of discourse in which they are specified’ (thus exactly fitting our parallel between linguistic idealism and Berkeley’s ‘Esse est percipi’).footnote4 That portion of the world that it is beyond the capacity of our present ‘discourse’ or ‘conceptual scheme’ to describe, is not, however, rendered unreal by our incapacity to describe it, as, effectively, the idealist claims when he savers that what is real depends upon what can be said or written!

The various proponents of the relativism that ensues employ arguments which turn, in one way or another, on the linguistic co-variability of meaning, belief and reference. It is because of this co-variability that any conceptually or culturally neutral determination of meaning seems such an insurmountable problem. Purely on an analytical level, Quine has most rigorously explored these arguments and their consequences in his thesis of the ‘indeterminacy of translation’.footnote5 He argues that rival conceptual schemes expressible in different languages are always underdetermined by experience. This ‘under-determination’ is not just inductive, in the sense that conceptual schemes are always more general and therefore more fallible than the particular experiences which confirm them. It is also the case that our capacity to refer to what we believe the world to contain, and our capacity to state what we believe to be true, are both necessarily mediated by the concepts in the language we employ. It is this which sustains the powerful image of our language as a conceptual prison. Since no description of experience can avoid being stated in some language, experience itself cannot arbitrate between the rival conceptual schemes that different languages may contain. Any given domain of experience will be consistent with different and conflicting conceptual schemes. Schemes of translation and, therefore, of communication, can do no more than note the coincidence of utterances with particular circumstances of sensory stimulation. Beyond this, identity of meaning will be indeterminate and, most especially, what speakers of divergent languages suppose they are referring to (and what they believe to exist) will be wholly relative to the conceptual scheme their language embodies. This Quine calls his thesis of ‘ontological relativity’.footnote6

Quine articulates his views on indeterminacy and relativity in the somewhat artificial context of ‘radical translation’—a situation where a translator tries to understand an alien culture knowing nothing about it or its language. It is in this context that the idea of a conceptual scheme takes on its clearest meaning. Here Quine’s arguments about communication and language lend their most powerful support to the derivative and less sophisticated forms of relativism of recent years. Discourse about debates between ‘paradigms’ in the history of science, communication between cultures in anthropology, understanding the divergent ‘ideologies’ of sub-cultures in sociology, and opposing the ‘myth of mental illness’ in psychology would all fall into this category. Yet despite its explicit or implicit influence, when we examine some of the consequences of Quine’s position, it is clear that it—and related formulations of relativism—face serious problems. Two of these in particular will concern us.

First, it seems to follow from Quine’s views that it is impossible to distinguish in principle between disagreements concerning matters of fact and misunderstandings of the meaning of what has been said. In a situation of radical translation, the question is raised: ‘Do the aliens really mean what they are apparently claiming about the world or themselves?’ We may either regard their beliefs as false or we may doubt that we have properly understood them. Kuhn, for example, poses an analogous problem with his famous reference to ‘gestalt switches’ in discussing the difficulties of communicating between paradigms’.footnote7 Yet the problem of disagreement/misunderstanding cannot be restricted only to radical translation, since it is just as possible to misunderstand a neighbour as it is an alien. Were the indeterminacy thesis therefore linked to a theory of communication, it would follow that we could never be sure we were communicating with anybody, thus undermining any distinction—qua communication—between ordinary understanding and radical translation.footnote8 This is the case because, again, for Quine ‘communication’ amounts to no more than different individuals uttering the same sounds under the same conditions of sensory stimulation. He has, indeed, shown more effectively than anyone that there are no objectively determinable ‘meanings’ to utterances that are shared by two or more speakers of different (or the same) language, over and above their shared circumstances of sensory stimulation. It is with respect to such circumstances that the sense of their utterances is indeterminate. But what, then, is conveyed when we speak, such that something might turn on agreement or disagreement, in contrast to understanding or misunderstanding what is said? That is to say that, since some understanding is necessary for anything to be conveyed about which there might be agreement or disagreement, it is unclear from Quine’s account what the object of this agreement might be. Our agreement must amount to more than the mere fact that we all happen to make the same sounds under the same stimulus conditions, like a flock of identically trained parrots who all ‘agree’ on the proposition ‘Pretty Polly’.

So, and this is our second point, two consequences flow from this for the possibility of our understanding cultural and conceptual change. Such changes ought to be both unintelligible in their content and inexplicable in their motivation. Suppose that Quine is right, and the sole criterion for agreement/understanding is no more than saying the same under the same stimulus conditions. It then follows that any change in what is said is ex hypothesi unintelligible, because we now no longer do say the same under the same stimulus conditions. More strictly, by his criterion we could make no distinction between our doubting the truth of our former opinions and our ceasing to understand what we used to say—precisely the dilemma that Kuhn suggests with regard to scientific revolutions. Moreover, were this an adequate account of what is real and rational, wholly determined by existing culture, what could motivate intellectual or social change, much less allow the all important identification of what might or might not be progressive in any such change? If one were confined within a ‘conceptual prison’ but had no way of knowing it, then ‘getting out’ would have no meaning and no one would ever want to do so.footnote9