Feyerabend

Philosophy, Science and Society

John Preston

• Publisher: Polity
• Year: 2013
• Print ISBN:
  9780745616766
  9780745616759
• eBook ISBN:
  9780745668321
• Edition: 1
• Topic: Philosophy
• Subtopic: Philosophy & Ethics in Science

This book is the first comprehensive critical study of the work of Paul Feyerabend, one of the foremost twentieth-century philosophers of science.

The book traces the evolution of Feyerabend's thought, beginning with his early attempt to graft insights from Wittgenstein's conception of meaning onto Popper's falsificationist philosophy. The key elements of Feyerabend's model of the acquisition of knowledge are identified and critically evaluated. Feyerabend's early work emerges as a continuation of Popper's philosophy of science, rather than as a contribution to the historical approach to science with which he is usually associated.
In his more notorious later work, Feyerabend claimed that there was, and should be, no such thing as the scientific method. The roots of Feyerabend's 'epistemological anarchism' are exposed and the weaknesses of his cultural relativism are brought out.

Throughout the book, Preston discusses the influence of Feyerabend's thought on contemporary philosophers and traces his stimulating but divided legacy. The book will be of interest to students of philosophy, methodology, and the social sciences.

1

Philosophy and the Aim of Science

1.1 Scientific and Analytical Philosophy

It has not gone unnoticed that philosophy has a deeply self-reflexive quality which sets it apart from other activities, for the nature of philosophy itself has often been an important philosophical issue. In this century the self-conception of Anglo-American philosophy has been shaped mainly by the notions of scientific philosophy, and analytical philosophy.

Paul Feyerabend set out one conception of the subject in one of his earliest papers, where he argued as follows. Philosophy cannot be both scientific and analytical at the same time. If a discipline is to be scientific it must have a certain subject matter, and it must be progressive, in that it will involve coming to know more about the objects which comprise this subject matter. But if we assume that philosophy is scientific in this sense and that it consists of analyses (of language, for example), none of its propositions could express discoveries. This is because of the ‘paradox of analysis’,¹ that any correct philosophical analysis of a concept must be uninformative, and any informative one must be incorrect. Feyerabend concluded that ‘philosophy cannot be analytic and scientific, i.e. interesting, progressive, about a certain subject matter, informative at the same time’ ([1956a], p. 95). Philosophers must choose between analytical and scientific philosophy.

This conclusion he compared with Wittgenstein’s dictum that ‘If one tried to advance theses in philosophy, it would never be possible to debate them, because everyone would agree to them’ (Wittgenstein [1953], § 128), and he criticized philosophers who want a purely descriptive philosophy which nevertheless leads to discoveries and extends our knowledge. But Feyerabend found the Wittgensteinian ideal of pure philosophical analysis unsatisfactory because he thought it unfruitful. So he decided to pursue ‘scientific’ philosophy, philosophy which, in yielding scientific knowledge, makes progress.

Such a conception, on which philosophy is distinguished from empirical science only by its greater generality, is problematic. The ideal of philosophy not just as contributing to scientific inquiry, but as scientific inquiry, is barely intelligible when applied to parts of philosophy other than epistemology (the theory of knowledge) and metaphysics (the theory of what really exists). It threatens to make nonsense of the scope and the history of the subject, and may incur the wrath of scientists, who will accuse philosophers of overstepping their proper domain.

Although at odds with analytical philosophy, this early Feyerabendian conception does not coincide with recent conceptions of scientific philosophy either. Many contemporary philosophers see philosophy as continuous with science, but they do so on the basis of their rejection of any distinction between the a priori and the empirical or between the analytic and the synthetic. The analytic/synthetic distinction purports to divide propositions true solely in virtue of their meaning from propositions whose truth or falsity depends on how things are. Analytic or ‘conceptual’ propositions, the territory of the analytical philosopher, are factually uninformative; synthetic propositions, with which the empirical scientist is concerned, are not. Philosophers who reject such distinctions thereby repudiate the idea of a choice between scientific and analytical philosophy. But Feyerabend, at this early stage, accepts some such distinction in order to choose scientific as opposed to analytical philosophy. One cannot opt for scientific philosophy (in this sense) and yet deny the existence of analytical philosophy, or of conceptual truths, even though one might ignore them for being uninformative.

Feyerabend must, therefore, have had an early change of mind about the analytic/synthetic distinction, for he never subsequently deployed it, and his (very few) relevant published comments consistently deny its existence, as well as the existence of any class of non-empirical statements. Not only did he choose to pursue ‘scientific’ philosophy, he also denied that analytical philosophy was of any value.

There is no small irony in Feyerabend plumping for ‘scientific philosophy’ at this point. His initial conception of philosophy represents a deeply scientistic starting-point for someone whose later work is seen as a searing critique of the claims of science and the very idea of scientific progress.² In subsequent early writings, after he had more or less explicitly scouted any distinction between a priori and empirical truths, Feyerabend’s conception of philosophy settled down into one that was still scientistic, but less vibrantly so, and thus moved closer to that of mainstream contemporary philosophers. But throughout his career Feyerabend railed against analytical and ‘linguistic’ philosophy, insisting that philosophy should help science and humanity progress, rather than hinder them or simply try to clarify science from the sidelines.

1.2 The Third-Person Approach to Epistemology

Although Feyerabend is thought of primarily as a philosopher of science, his real interest lay in the more general subject of human knowledge, and his approach to that subject was heavily influenced by Popper. To investigate Feyerabend’s conception of philosophy of science, we must begin with their intellectual relationship.

Such a suggestion will, however, meet with resistance. The later Feyerabend chastised those who regarded him as a former Popperian. The many acknowledgements to Popper in his earlier papers were, he said, ‘friendly gestures, not historical statements’ (SFS, p. 144n). Complaints about Popper became an oft-repeated theme in Feyerabend’s later work.³ But however strong the current of invective, it is insufficient to dislodge the verdict that Feyerabend was, until the late 1960s, a Popperian.⁴ His own work cannot be seen in clear perspective unless we realize that he is tackling the questions ‘Is knowledge possible?’ or ‘How is knowledge possible?’, and ‘What is the best way to attain knowledge?’ These questions comprise just what Popper called the central problem of epistemology, the ‘problem of the growth of knowledge’ (see Popper [1959], pp. 15–18).

Feyerabend tells us that he discusses scientific theories not out of a wish to restrict himself to philosophy of science (a discipline he often refers to in scathing terms), but because he regards them as ‘excellent examples of actual knowledge’ ([1965a], p. 217). In his early work he is engaged on the same philosophical project as Popper, the epistemology of science, because during this time he accepted that ‘the growth of knowledge can be studied best by studying the growth of scientific knowledge’ (Popper [1959], p. 15). Popper was not alone in his conviction that scientific knowledge is an extension of common-sense knowledge (Russell and Einstein agreed), and that common-sense knowledge grows principally by turning into scientific knowledge. But most importantly he also believed that

scientific knowledge can be more easily studied than common sense knowledge. For it is common sense knowledge writ large, as it were. Its very problems are enlargements of the problems of common sense knowledge. (Popper [1959], p. 22)

To say that Feyerabend is enlisted in pursuing the epistemology of science is not yet to say that he was a Popperian. The project outlined is common to thinkers of many divergent standpoints, few of whom could be described thus. But it does help to explain why, although Feyerabend claims to be searching for a model for the acquisition of knowledge, his abiding concern was with scientific knowledge.

This starting-point, however, must be probed. By assimilating common-sense knowledge to theoretical science, it presupposes that general epistemic rationality and scientific rationality coincide. We have to take seriously the possibility that they do not, that they are attuned to different goals. If this is the case, one cannot do general epistemology by doing epistemology of science. If there is no single all-purpose rational method of inquiry, usable in any domain (a later Feyerabendian conclusion which would undercut the rationale for his early work), if science is not ‘common sense writ large’, then the epistemology of science is no more than a corner of epistemology as a whole, and may be of no greater intrinsic importance than any other part, as many of the ‘linguistic philosophers’, whom Popper and Feyerabend despised, always suspected.

Popper and Feyerabend also shared a conception of how epistemology should be done, a conception Popper came to call ‘epistemology without a knowing subject’,⁵ and which he contrasted with a more traditional conception of epistemology. Adherents of this more traditional conception, such as the British Empiricists, Kant, Mill, Russell, and Husserl focus on ‘the knowing subject’ and try to figure out its relation to ‘the external world’. Third-person epistemology, by contrast, focuses on the products of knowledge: the statements, laws, and theories which cognitive activity issues in. There is nothing exclusively Popperian about third-person epistemology either, since other major twentieth-century philosophers such as Wittgenstein, Quine and Davidson, also take a third-person approach to epistemology. Feyerabend’s full Popperian credentials only become clear when we identify him as a contributor to the project of giving a rational model of science, and as an adherent of three crucial views: normative epistemology, falsificationism, and inductive scepticism.

1.3 Feyerabend’s Project: A ‘Model for the Acquisition of Knowledge’

What we might call⁶ a rational model of science specifies both an aim for science, and a methodology for approaching that aim. I shall eventually concur with Feyerabend’s later suggestion that there is no single such model to be had. But in his early work Feyerabend was as much a contributor to the project of developing a rational model of science as Popper. He characterized the aim of his early work as being ‘to present an abstract model for the acquisition of knowledge, to develop its consequences, and to compare these consequences with science’ ([1965c]: PP1, p. 104). The papers which he wrote between 1957 and the late 1960s, and which will be the major focus of my attention in this book, are most profitably understood as a contribution to this project. A study of Feyerabend’s attempt to develop this model is an essential prerequisite for understanding his infamous 1970s work.

We can learn an important thing about Feyerabend’s philosophy from his decision to pursue the epistemology of science and from his aim to construct a model of scientific knowledge. This is the overriding importance he attached to the attainment of knowledge. Feyerabend took knowledge of scientific statements, rather than understanding, insight, wisdom, conceptual clarity, or enhanced experience, to be the aim of science and therefore of scientific philosophy, and this sets the tone for his early work.

We might also enter some reservations about this project. Science can be said to aim at understanding, explanation, the discovery of natural laws, prediction, and technological control, among other things. Not all of these are profitably characterized as forms of knowledge. Unless we understand ‘knowledge’ in a loose sense, Feyerabend’s position here involves a distortion. Scientific knowledge, narrowly understood as knowledge of scientific propositions, cannot be said to be the exclusive aim of science itself, let alone of philosophy. Just because the epistemologist is primarily interested in science under its aspect as knowledge-gathering does not guarantee that the aims of such a multifarious activity can be characterized in exclusively epistemological terms. The question ‘What is the aim of science?’ may well be misconceived if it is supposed to have a single answer. The truth may be that science affords us a variety of things, not all of them covered by the concept of knowledge. The idea of a rational model of science is already put in some jeopardy, therefore, by the suspicion that science has no single aim. What is more, as we are about to see, Feyerabend’s way of characterizing the methods of science is flawed.

1.4 Normative Epistemology, and Falsificationism

When epistemology is conceived of as a description of the conditions under which knowledge exists, or is attainable, it is said to be descriptive or naturalistic. Two of Feyerabend’s most important teachers, Viktor Kraft and Karl Popper, both rejected such a conception. Instead, they conceived of epistemology as a wholly normative discipline, a discipline which lays down rationally grounded rules or norms which, if followed, would produce good science. Feyerabend initially took this conception on board more deeply than any of his contemporaries.

I shall concentrate here on relating the views of Kraft, whom Feyerabend saw as anticipating some ideas now associated with Popper.⁷ Kraft investigated the nature of knowledge and the way in which its essential characteristics can be determined, arguing that ‘the idea of knowledge can be arrived at only on the basis of stipulations and its validity is a matter of agreement (with these stipulations)’.⁸ The theory of knowledge, he thought, must be ‘very different from the factual sciences; it does not deal with something existing in reality but puts forth aims and norms for our intellectual activity’ (Kraft [1960], p. 32); it suggests a set of ideals which we use to criticize or praise knowledge claims. This bears comparison with Popper’s suggestion about how proposed aims of science should be suggested and evaluated. It is the same conception of epistemology Feyerabend committed himself to when he insisted that ‘the issue between positivism and realism is not a factual issue which can be decided by pointing to certain actually existing things, procedures, forms of language, etc., it is an issue between different ideals of knowledge’ ([1958a]: PP1, pp. 33–4). In the same breath, Feyerabend explicitly acknowledged that this thesis is an extension of Popper’s views on scientific method, and that the normative character of epistemology had been stressed by Kraft.⁹

Feyerabend praised Kraft’s procedure as being the kind of bold and optimistic attempt to change and criticize existing theories, and to develop new theories, which led to modern science in the first place. He followed Popper and Kraft in thinking that the epistemology of science is not a descriptive discipline, but has a normative cast: it tells us what good science should be. His most explicit statement to this effect is as follows:

[S]cientific method, as well as the rules for reduction and explanation connected with it, is not supposed to describe what scientists are actually doing. Rather, it is supposed to provide us with normative rules which should be followed, and to which actual scientific practice will correspond only more or less closely. It is very important nowadays to defend such a normative interpretation of scientific method and to uphold reasonable demands even if actual scientific practice should proceed along completely different lines. It is important because many contemporary philosophers of science seem to see their task in a very different light. For them actual scientific practice is the material from which they start, and a methodology is considered reasonable only to the extent to which it mirrors such practice. ([1962a], p. 60)¹⁰

But Feyerabend criticized Kraft for not following up this conception of epistemology with an equally clear conception of the norms to be adopted, and of the reasons for adopting these norms. This, he said, ‘would have been a revolutionary undertaking indeed, the first construction of a purely normative epistemology’ ([1963c], p. 320). This is what Feyerabend undertook to provide.

Kraft, like Popper, held that almost all the procedures of science and of common sense involve hypotheses, such as the hypothesis of the existence of material things and the hypothesis of the existence of other minds. He thought of these as ‘unjustified conjectures for which no foundation can be given’ ([1963c], p. 322), but argued that the only alternative to their use is solipsism, the view that the only knowable things are one’s own momentary psychological states. We are therefore forced into an uncomfortable choice between solipsism and a body of doctrine which makes essential use of unjustifiable conjectures. Kraft refused to embrace solipsism, and hence resolutely denied that our hypotheses can be justified. But, like Moses, he could not enter the promised land he had seen: he failed to make the transition to falsificationtsm.

A rational model of science, recall, specifies an aim and a methodology for science. If knowledge is the aim of science according to Feyerabend, its methodology, the means of achieving that aim (and the way out of Kraft’s dilemma) is falsificationism. Feyerabend suggested a scientific methodology consisting of two principles: the demand that our theories be testable, and the demand that they explain known phenomena. The former ensures that we connect our theories with experience, the latter specifies the connection with experience in detail: our theories must not be ad hoc, and they must be richer in content than what they are to explain. It also results in their transcending experience, for ‘if the entities postulated for explanation completely coincide with laws in the domain of experience, then they are ad hoc with respect to these laws and therefore no longer capable of giving a satisfactory expla...

Table of contents

• Cover Page
• Title Page
• Copyright
• Dedication
• Contents
• Preface
• Acknowledgements
• Note on References
• Introduction: Feyerabend’s Life and Work
• 1 Philosophy and the Aim of Science
• 2 Meaning: The Attack on Positivism
• 3 Theories of Observation
• 4 Scientific Realism and Instrumentalism
• 5 Theoretical Monism
• 6 Incommensurability
• 7 Theoretical Pluralism
• 8 Materialism
• 9 Science without Method
• 10 Relativism, Rationalism and a Free Society
• Notes
• Bibliography
• Index