What is a ‘fact’? This deceptively simple question provides a route into the most pressing issues concerning the philosophy of science. In a short essay, “Philosophies as Ideologies of Science: A Contribution to the Critique of Positivism”, Roy Bhaskar offers a compelling answer to this question which impressively incorporates an epistemic account of knowledge alongside a sociological account of theories of knowledge. It is directed towards a critique of positivism which, he writes, “is, in the first instance, a theory of the nature, limits and unity of knowledge”. On the positivist account, knowledge is of events which are revealed in perception. The generalisation of such knowledge involves identifying the patterning of such events, in space and over time, with sense perception being understood to exhaust the possible objects of knowledge (Bhaskar 1989: 49). Underlying this is an (often implicit) ontology of closed systems and atomistic events. However Bhaskar also claims that any theory of knowledge also presupposes a sociology “in the sense that it must be assumed, implicitly if not explicitly, that the nature of human beings and the institutions they reproduce or transform is such that such knowledge could be produced” (Bhaskar 1989: 49). This conception of “people as passive sensors of given facts and recorders of their given constant conjunctions” has important implications (Bhaskar 1989: 51). Given that constant conjunctures are worked for in laboratories rather than discovered in nature, it follows that the intelligibility of the objects of scientific investigation must be understood as ontologically independent of the activities of human beings but conversely that the concepts under which we encounter them must be seen as part of an irreducibly social process of science.
Thus experiences (and the facts they ground), and the constant conjunctions of events that form the empirical grounds for causal laws, are social products. But the objects to which they afford us access, such as causal laws, exist and act quite independently of us.
Now positivism can sustain neither the idea of an independent reality nor the idea of a socially produced science. Rather what happens is in a way quite extraordinary – for, as in the interests of a particular conception of philosophy, it allows a particular conception of our knowledge of reality to inform and implicitly define the concept of the reality known by science, these ideas (absolutely minimally necessary conditions for an adequate account of science) become crossed, so that we have a naturalised science purchased at the expense of a humanised nature. (Bhaskar 1989: 51)
It is in this ‘transference’ that we begin to locate the ideological implications of positivism. Positivism bequeaths us an alienated understanding of science, repudiating its social nature, as well as a mystified account of nature, cleaved to our putative sense perceptions of it. This is what leads to the reification of facts. If we view “what is apprehended in immediate sense-experience as a fact constituting an atomistic event or state of affairs, existing independently of the human activity necessary for it” then we come to see ‘facts’ as self-subsistent and independent of human activity. Furthermore, as Bhaskar puts it, “when the conjunctions of such facts are reified and identified with causal laws, science becomes an epiphenomenon of nature” (Bhaskar 1989: 52). In coming to such a view, where ‘facts’ can be read from ‘nature’, we deny the social character of science, particularly “its character as work involving the transformation of antecedent objects, both material and ideational” (Bhaskar 1989: 52). The point is not to repudiate ‘facts’ but rather to recognise the grounding of facts on things and constant conjunctions on causal laws. Human beings produce facts but they do so through collective labour of a particular sort, orientated towards both material and ideational objects, which seeks to intervene in and accumulate knowledge of a natural world irreducible to this labour.
Underpinning this is a fallacious image of a “moment of subjectivity which is free from the effects of all pre-formed or extraneous, including theoretical content” (Bhaskar 1989: 52). For positivism this “autonomized sense-experience constitutes the form in which knowledge is acquired and the reified fact the content that is expressed” (Bhaskar 1989: 54). The atomicity of events necessitated by this account of sense experience in turn entails the constancy of conjunctions as a condition for general knowledge. I’m not 100% certain I follow Bhaskar’s point here but I think he is suggesting that positivism offers no other ‘building blocks’ out of which general knowledge can be constructed, such that it becomes a matter of establishing logical relations been atomistic impressions of events rather than the achievement of deeper knowledge of natural processes. For positivism “facts and their conjunctions both exhaust the real content of science and determinate the knowable nature of the world, or fix science in its ontology” (Bhaskar 1989: 54). In rejecting the ‘reification of facts’, Bhaskar’s point is not that ‘facts’ are not ‘things’. Rather he is arguing that “facts are things, but they are social not natural things, belonging to the transitive world of science, not the intransitive world of nature” (Bhaskar 1989: 55). There are non-reified accounts of facts which he believes to be problematic, as correctly understanding the relation between facts and human activity poses a conceptual and methodological challenge extending beyond the simple recognition that we do not ‘discover’ facts in the natural world.
One might take ‘a fact’ as a synonym for a ‘true assertion’ or alternatively, to consider Strawson’s well-known formula, to attempt to explicate it as what a true statement states. But this will not do – for we make statements, but we do not make facts (rather it is as if they were made for us); and it is unclear, to say the least, how statements can do anything. Why do we not make facts? The answer is because, as the etymology of the word suggests, they are already made. In stating facts we are acknowledging results already achieved, the results being achieved (in the domain of empirical discourse) by readings, of which the varieties of sense experience constitute special kinds of skill. But of course we discover as well as state results: the facts pre-exist their discovery as results to be achieved (just as they pre-exist their statement as achievements). They must thus be conceived as potentialities of the conceptual schemes or paradigms that govern our enquiries, which when actualised constitute discoveries. A fact, then, is a potentiality actualised in discovery, sustained in practice (both discursive and on-discursive) and objectified in sense-perception (Bhaskar 1989: 60)
The cognitive structures in virtue of which ‘facts’ become possible are reproduced or transformed by human activity. However we do not as such create them because we are always positioned in relation to them and, in many senses, dependent upon them. This is why the notion of scientific labour is so important for Bhaskar’s project to critique positivism while also avoiding conventionalism. It’s in this sense that our tendency to “read the world as if it were constituted by facts” becomes important, as “what is involved in the categories in terms of which positivistic philosophical consciousness understands the concept of a fact is not therefore so much a crass mistake or straightforward error, as a superficiality which merely reflects the spontaneous consciousness of science” (Bhaskar 1989: 61). What Bhaskar is affirming is the need to become reflectively conscious of this ‘spontaneous consciousness’ in order to achieve distance from it. Facts are real but they are “historically specific social realities” and it is recognition of this which is essential for understanding how scientific progress becomes possible. The fetishism of facts also dehistroicizes them and it is in this sense that Bhaskar sees the category of ‘ideology’ as appropriate to these discussions. In other words, positivism has a social function which is “to conceal the historically specific structures and relations constituting sense-experience in science” (Bhaskar 1989: 61). More specifically, the social function of the constant conjunction is to “conceal the reality of structures irreducible to events” (Bhaskar 1989: 62).