Scientific Realism and Human Emancipation

Roy Bhaskar

• Publisher: Routledge
• Year: 2009
• Print ISBN: 9780415454957
• eBook ISBN: 9781134009411
• Topic: Philosophy
• Subtopic: Epistemology in Philosophy

Following on from Roy Bhaskar's first two books, A Realist Theory of Science and The Possibility of Naturalism, Scientific Realism and Human Emancipation, establishes the conception of social science as explanatory—and thence emancipatory—critique.

Scientific Realism and Human Emancipation starts from an assessment of the impasse of contemporary accounts of science as stemming from an incomplete critique of positivism. It then proceeds to a systematic exposition of scientific realism in the form of transcendental realism, highlighting a conception of science as explanatory of a structured, differentiated and changing world.

Turning to the social domain, the book argues for a view of the social order as conditioned by, and emergent from, nature. Advocating a critical naturalism, the author shows how the transformational model of social activity together with the conception of social science as explanatory critique which it entails, resolves the divisions and dualisms besetting orthodox social and normative theory: between society and the individual, structure and agency, meaning and behavior, mind and body, reason and cause, fact and value, and theory and practice. The book then goes on to discuss the emancipatory implications of social science and sketches the nature of the depth investigation characteristically entailed.

In the highly innovative third part of the book Roy Bhaskar completes his critique of positivism by developing a theory of philosophical discourse and ideology, on the basis of the transcendental realism and critical naturalism already developed, showing how positivism functions as a restrictive ideology of and for science and other social practices.

1
Scientific Realism and the Aporias of Contemporary Philosophy

‘All social life is essentially practical. All mysteries which lead theory to mysticism find their rational solution in human practice and in the comprehension of this practice’.¹

1. Rifts in recent philosophy

Of late the philosophy of science has appeared a prisoner of paradox. The fundamental assumptions of the positivist worldview, dominant for over a century, that science is monistic in its development and deductive in its structure, lie shattered. But the ensuing accounts of science have been unable to sustain the coherence of precisely those features of science which they have placed to the fore of the philosophical agenda, namely scientific change and the non-deductive aspects of theory.

Consider first the anti-monistic tendency, exemplified most notably perhaps by the work of Bachelard, Canguilhem, Koyré, Popper, Kuhn, Lakatos, Feyerabend, Stegmüller and Laudan. Both Bachelard and Kuhn, in attempting to do justice to the phenomena of scientific change, veer very close to the position, which is in effect a collectivist variant of what I shall call ‘sub-

jective super-idealism’, that we create and change the world, along with our theories² —a position which renders change in either unintelligible. Neither Kuhn nor Feyerabend have managed to explain how there can be a clash between incommensurable descriptions, or to say over what such descriptions clash. Popper has not demonstrated how the refutation of a conjecture could be rational, unless nature were uniform; and he has apportioned no ground for assuming that it is, in the face of standard sceptical (Humean, Goodmanesque³ , etc.) scenarios. For what is there to prevent nature altering so that our most decisively rejected theories turn out true and our most cherished falsifiers, implicitly universal, false? This is of course nothing other than that old ‘scandal of philosophy’⁴ : the Humean problem of induction. Similarly Lakatos has not shown how, unless nature were uniform, it would be rational to work on progressive rather than degenerating programmes; or, for that matter, to pay any attention to the history of science. For almost all the philosophers within this corps there is a tension between (ontological) realism and (epistemological) relativism. And generally they have encountered some difficulty in squaring their emphasis on scientific discontinuities with the seemingly progressive character of scientific development, in which there is growth as well as—and in—change. Parallel problems beset the anti-deductivist camp, typified by the work of Kneale, Waismann, Hanson, Scriven, Polanyi, Toulmin, Hesse and Harré. Such philosophers have reacted sharply against the standard Humean and positivist account of science on which, in the last instance, there can be no ground for divaricating a necessary from a merely accidental sequence of events, for expecting emeralds to be green rather than grue⁵ or for supposing that water might not freeze rather than boil when it is heated. These philosophers, writing under the initial pull of Wittgenstein but also self-consciously recuperating the arguments of Kant, Whewell, Peirce and Campbell (and, to an extent, Aristotle and Locke), have sought to show how scientific practice yields cognitive items—whether dressed as models, paradigms, beuristics, conceptual schemata or regulative ideals—which are irreducible to syntactical operations upon sense-experience and yet indispensable for the intelligibility and empirical extension of theory. In this way such items function as it were, as social surrogates for natural necessity.⁶ However, so long as it retains an empiricist ontology, three stark problems straight away confront this school. First, to the extent that the surrogate can be empirically described, its independent cognitive role disappears (as the necessity of the connection, analogical character of the model, ideality of the order, etc. evaporate); conversely, to the extent that its cognitive role is preserved, its epistemic warrant crumbles (since it now ceases to designate real phenomena).⁷ Second, as the analysis presents natural necessity as a product or mediation of human mind, it scarcely seems adequate to explicate the sense in which science presumes to discover necessities in mind-independent things. Third, an evident tension exists between the principles (of coherence, intelligibility and structure) invoked to make sense of science and the properties (of atomicity, fragmentation and flatness) tacitly attributed to being. Together, these problems comprise a homologue of the old Kantian problem of the synthetic a priori in science: that is, the problem of sustaining the necessity and universality apparently accruing to scientific judgments, when such judgements are supposedly about a world of loose and disconnected events. The anti-deductivist faction then is immediately assailed by an antinomy between the cognitive function of the surrogate and its epistemic title, a discrepancy between philosophical analysis and its intended object and a rift between epistemological principles and ontological presuppositions. And these are as evident in their Kantian prototype as in its contemporary avatars. For patently there is no way in which, even if the Kantian principles were all demonstrably valid, they could be used to justify any particular inductive inference (any more than Aristotelian ‘nous’, or more precisely active intellect, could prohibit the inference to grue): that the world will ‘carry on’, under the descriptions in terms of which it is currently known to us, remains an article of faith.⁸ More generally, writers within this tradition have not always succeeded in blending their stress on the synthesising activity of the scientific imagination with the messy practicalities of science’s causal transactions with nature—the nuts and bolts, so to speak, of scientific life.

I have tried to show elsewhere⁹ how the antinomic constitution of contemporary philosophy of science, betraying its historical station as an incomplete critique of positivism and reflected in these and a plethora of related aporias, can be traced to the supervention of new epistemological insights upon old (empiricist or idealist) ontologies, more or less materially incompatible with them. My primary aim in this chapter is to demonstrate why and how, if the rational gains of both the anti-monistic and anti-deductivist movements are to be saved, a new ontology, and corresponding account of science, must be cultivated to accommodate them. This new ontology comprises a Copernican revolution in the strict sense (of Copernicus, not Kant) of a de-anthropocentric shift in our metaphysical conception of the place of (wo)man in nature, in which the umbilical cord uniquely tying thought to things in traditional philosophy is snapped (the significance of discontinuism) and ontological structure, diversity and change emerge as conditions of the practical cognitive activity of science (the significance of non-deductivism). My secondary aim is to locate the impasse in contemporary philosophy of science in terms of its foundational structural, historical and aporetic parameters. For we shall see that positivism, discredited but not dissolved, is merely the dominant historical attractor position in a plate with deeper alethic roots. I want also to prepare the ground here for the specification of the logic of distinctively social scientific enquiry and so to set the scene for an ‘explanatory critique’ of positivim and its current progeny. Such a critique aims to demonstrate not just why an idea or system is false, but how it comes to be believed and acted upon, i.e. reproduced, or more or less transformed, in some or other historically determinate society.¹⁰

2. Forms of realism

The account of science I wish to commend to the reader is a realist one. In its broadest sense in philosophy any position can be nominated ‘realist’ which asserts the existence of some disputed kind of entity (universals, material objects, causal laws, numbers, probabilities, propositions, etc.). But for my story the most historically significant types of realism are: predicative realism, asserting the existence of universals independently (Plato) or as the properties (Aristotle) of particular material things; perceptual realism, asserting the existence of material objects in space and time independently of their perception; and scientific realism, asserting the existence and activity of the objects of scientific enquiry absolutely or relatively independently of the enquiry of which they are the objects or more generally of all human activity. (The different grades of strength of scientific realism need not detain us at present.) It is scientific realism with which I am directly concerned here, although of course it reduces to predicative or perceptual realism if the objects of scientific knowledge just are Platonic (or Aristotelian) forms or material objects.

Scientific realism, then, is the theory that the objects of scientific enquiry exist and act, for the most part, quite independently of scientists and their activity. So defined, it might be supposed that the question of whether or not natural science is ‘realist’ can only be settled empirically, viz. by determining whether or not scientists believe, or behave as if, the theoretical terms they employ possess real referents independently of their theorising.¹¹ This issue is obviously important. But I intend to argue the case for a metaphysical realism, consisting in an elaboration of what the world must be like prior to any empirical investigation of it and for any scientific attitudes or activities to be possible—a realism which neither endorses nor presupposes a realistic interpretation of any particular theory.

It is clear that such a metaphysical realism, in contrast to a merely first order or ‘internal’¹² realism, depends upon the feasibility of a philosophy, as distinct from a sociology or history, of science; and, within philosophy, of an ontology as well as an epistemology. For realism is not a theory of knowledge or truth, but of being—although as such it is bound to posses epistemological implications. Accordingly, a realist position in the philosophy of (natural) science will consist, first and foremost, of a theory about the nature of the being, rather than the knowledge, of the objects investigated by the sciences—to the effect that they endure and operate independently of human activity, and hence of both sense-experience and thought. So realism is immediately opposed to both empiricism and rationalism, wherein being is defined in terms of the human attributes of experience and reason. And it repudiates, from the beginning, that dogmatic canon of post-Humean philosophy (which I have styled the ‘epistemic fallacy’¹³ ) decreeing that ontological questions can always be transposed into an epistemological key, i.e. that statements about being either just are or may always be parsed as statements about knowledge.

Acceptance of Hume’s canon characterises the analytical and dialectical traditions in modern philosophy alike. For it was ratified by Kant, objectified by Schelling and beatified by Hegel as the criterion of philosophy, understood as the unfolding consummation in thought of the primal Parmenidean postulate of the identity of being and thought. This is a postulate from which contemporary scientific realism registers a complete, but reasoned, break.

Now any theory of the knowledge of objects entails some theory of the objects of knowledge; that is, every theory of scientific knowledge logically presupposes a theory of what the world must be like for knowledge, under the descriptions given it by the theory, to be possible. Thus suppose a philosopher analyses scientific laws as, or as dependent upon, constant conjunctions of events—as, for instance, Kant did in the Second Analogy, in maintaining that for every event there is a condition upon which it invariably (and necessarily) follows.¹⁴ S/he is then committed to the view that there are such conjunctions; that, in Mill’s words, ‘there are such things in nature as parallel cases: that what happens once will, under a sufficient degree of similarity of circumstance, happen again’.¹⁵ In this way, then, as Bachelard recognised, ‘all philosophy, explicitly or tacitly, honestly or surreptitiously…deposits, projects or presupposes a reality’.¹⁶ Hence, to invert and circumscribe Hegel’s famous dictum about idealism: every philosophy, inasmuch as it takes science for its topic, is essentially a realism, or at least has realism for its principle, the pertinent questions being only how far and in what form this principle is actually implemented.¹⁷

Thus the mainstream in the philosophy of science, in both its classical empiricist (Humean) and transcendental idealist (Kantian) currents, presupposes an implicit empirical realism according to which the real objects of scientific investigation are defined in terms of actual or possible experience. More recently, the super-idealist party has secreted an implicit subjective conceptual realism, according to which the real objects are envisaged as the product of thought, i.e. of the spontaneous play of the human mind (or its socialised form, the scientific community), unconstrained by sense-experience. The remoter ancestry of this position includes Vico’s Facimus and Fichte’s Tathandlung as well as Nietzsche’s perspectivism. But it is nowadays usually approached by the simple shedding (e.g. in a historically relativised neo-Kantianism¹⁸ ), rather than by the reconstruction of a practical or subjective genealogy, of the idea of an independent reality. Both these kinds of implicit ontology are explicitly anthropomorphic. In contrast, the older precritical rationalism of Platonic and Cartesian provenance, positing an explicit objective conceptual realism, is implicitly anthropomorphic. For the real objects it entertains, although presented as independent of finite, human minds, are divined as quintessentially rational, as, in effect, reason in the (dis)guise of being: that is, as constituted by and/or causally or teleologically dependent upon what is known to us only as an attribute of human being, namely thought or reason. The philosophy of science also contains a persistent romantic strain, nurturing an intuitional realism, incubating real objects identified wholly or partially in terms of human intuition, sensibility or affect. In each case we have an anthropomorphic definition of being, based on some anthropocentric (which need not always be an epistemo-centric) view of the world, with the human attribute in question (e.g. knowledge) being in turn reciprocally—in the momentous anthroporealist exchange of subject-object identity theory—defined by the presumed character of the world. In rejecting this anthroporealist array, whose constituents may be combined in more or less subtle ways, I am going to contend that only a realism fully consonant with the principle of scientific realism enunciated above—a transcendental realism—can uphold and display the intelligibility and rationale of science.

But in addition to those anthroporealisms, transcendental realism is also and equally opposed to any non-immanent or transcendent realism, which posits a sphere of pure uncognisable other-being, defined in terms of its inaccessibility to human being (as are Kantian noumena). A realm of ‘ineffibilia’, e.g. held in contrast or conjunction with the sensibilia or intelligibilia of empirical or conceptual realism, is certainly conceivable, but it is something about and for which nothing (or at any rate very little) can be said. In other words, transcendental realism is occupied, at least in the first place, with the being of the objects of science. Now I am going to argue here, and show in more detail in Chapter 3, that any non-transcendental realism, i.e. any incomplete, inexplicit or ineffable realism, any ‘irrealism’¹⁹ —such as an empirical, conceptual, intuitional and/or transcendent realism—must in pr...

Table of contents

• Cover Page
• Title Page
• Copyright Page
• Preface
• Introduction
• Chapter One —Scientific Realism and the Aporias of Contemporary Philosophy
• Chapter Two —Critical Naturalism and the Dialectic of Human Emancipation
• Chapter Three —The Positivist Illusion: Sketch of a Philosophical Ideology at Work