Beyond Positivism

Bruce Caldwell

• Publisher: Routledge
• Year: 2015
• ISBN:
  9781134838639
• Edition: 1
• Topic: Betriebswirtschaft
• Subtopic: Business allgemein

Since its publication in 1982, Beyond Positivism has become established as one of the definitive statements on economic methodology. The book's rejection of positivism and its advocacy of pluralism were to have a profound influence in the flowering of work methodology that has taken place in economics in the decade since its publication. This editi

1

---

Introduction

---

 

When you are criticizing the philosophy of an epoch, do not chiefly direct your attention to those intellectual positions which its exponents feel it necessary explicitly to defend. There will be some fundamental assumptions which adherents of all the variant systems within the epoch unconsciously presuppose. Such assumptions appear so obvious that people do not know what they are assuming because no other way of putting things has ever occurred to them. With these assumptions a certain limited number of types of philosophical systems are possible, and this group of systems constitutes the philosophy of the epoch.

Alfred North Whitehead—Science and the Modern World (1925)

The study of methodology is an agonizing task; writing a book on the subject requires the skills of an individual who is at once presumptuous and masochistic. By the very nature of methodological work, solutions to important problems seldom seem to exist. One’s thinking on a particular subject is never complete; indeed, it is more likely that one’s opinion will change often through time, and sometimes change dramatically. Even more troublesome, the prolonged study of methodology forces a person to examine his or her own preconceptions, to see why certain ideas make sense, and why others seem so patently absurd. Nor is that self-examination a simple task, since preconceptions are not truly prior to experience, but invariably reflect both the material studied and the process involved in its study.

One preconception of mine that is admittedly unoriginal is that there is no single infallible method: there is no ‘best way’ waiting out there to be discovered, neither in the form of some Platonic ideal, nor by the careful objective study of the history of method. Rather, I am a methodological pluralist, by which I mean that, just as there exist many tasks for theories to perform, there are also many methods for the evaluation and criticism of theories. Most methodologies are prescriptive; they limit the range of acceptable theories and explanations in science. Such normative methodologies also place restrictions on what constitutes legitimate criticism; they prohibit not only certain types of theories, but certain types of methodologies as well. I would argue against both theoretical and methodological monism. Like any methodological view, methodological pluralism has problems associated with it, and, in light of my earlier comments, I would certainly like to retain the right to change my mind. But for now, methodological pluralism seems to be a reasonable, and potentially fruitful, approach to economic methodology.

One disquieting implication of methodological pluralism is that the present study and others like it are ineffective, that they are little more than a form of twentieth century scholasticism. What sense is there in studying methodology if one asserts at the outset that there is no hope of finding the best method? My particular approach must seem especially superfluous, since there is almost no discussion of the practice of economists, the focus being a discussion of the interface between the philosophy of science and the writings of economists on methodology.

The claim that the independent study of methodology divorced from the practice of economists is ineffectual is a serious one that must be considered. It is most plausible if one believes that the goal of methodological study is the discovery of an infallible method. If that is the goal, such study is truly a Sisyphean task that is best left not started. But there are more modest, and I think more attainable, goals of methodological study. In particular, I hope to make progress towards three of these.

First, methodological study aids one in understanding the essence (that word is chosen carefully, and not without trepidation) of various methodological views. Methodological discussion often involves the advocacy or critique of a particular theoretical formulation, and as such seems little more than the partisan rationalization of a preferred worldview. Methodology in such cases is perceived as simply another instrument of persuasion. But there is another aspect of methodological debate that must not be ignored. For there is imbedded in every methodological position a unique perspective on the question of how to gain knowledge, of how to most fruitfully investigate a given phenomenon. Methodology systematizes man’s curiosity; each methodological view directs the scientist to seek knowledge differently. By ‘getting inside’ a variety of such views, one gains new ways of perceiving the subject under investigation. Perhaps most essential, one may avoid the chains of a narrow perspective. This is especially important given that one’s methodological views are rarely consciously held: methodology is nowhere explicitly taught in modern curricula; rather, the modern scientist learns his methodology by plying his scientific trade.

A second goal of this study is to assess the cogency of various arguments made by economic methodologists. We can assess the cogency, though not, I think, the ultimate validity of the positions we investigate. My particular approach is to examine the writings of economic methodologists from the vantage point of twentieth century philosophy of science.

At first I supposed that more could be done. I thought that, since the philosophy of science analyzes the methods of all the sciences, and the methodology of economics is a particular instance of that larger study, the philosophy of science could be used to judge various pronouncements in the methodology of economics. But it cannot, for there is no guarantee that the propositions gleaned from the philosophy of science have either prescriptive force or descriptive accuracy when applied to economics. Moreover, most of the philosophy of science with which economists are familiar was written with the natural sciences, and particularly physics, in mind; its application in a social science like economics can and should be questioned. (There is a philosophy of social science, but few economists have expressed any interest in it.)

On the other hand, the study of philosophy of science can be useful as an aid to clear thinking on matters methodological. Philosophers of science, after all, have thought about the issues more than have most scientists. More to the point, many economic methodologists borrow phrases, terms, and concepts from the philosophy of science; yet often such expropriation has only brought further confusion, that is, economists who disagree about the meanings of those phrases, terms, and concepts. A study of economic methodology from a philosophy of science perspective may help one to clarify, unify, categorize, and explicate debates in the former field. But it will not provide ultimate grounds for arbitrating among well-developed and well-argued alternative positions. Fortunately, there is a sufficient number of poorly-developed and badly-argued cases to keep us busy. In addition, I will on occasion pause to offer suggestions of what forms potential criticisms of certain positions might take. There is no malicious intent; methodological pluralism insists that there are many roads to criticism, and these efforts are undertaken with that sentiment in mind.

A final goal of this study is to shake up some preconceptions that may presently exist in the economics profession. The chapter begins with a pertinent quote from Alfred North Whitehead—it is no easy job, but it is a vital one, to enumerate the assumptions of the epoch in which we live. I submit that one operative assumption of our time is the almost unquestioned authority of science. Its particular manifestation within our profession had its origins many years ago, when the notion first blossomed that economics could be, and should try to be, a scientific discipline. In the twentieth century the dream seemed realized with the emergence of positivism, a philosophical doctrine that appeared to offer a solid epistemological foundation for those sciences willing and capable of adhering to the rigors of the scientific method. Positivist exhortations soon dominated the methodological rhetoric of economics, even if they did not always inform the actual practice of working economists.

Positivism in its many variations has been in decline within the philosophy of science for the last twenty years or so, and that knowledge is now filtering down into the special sciences, especially since the works of the ‘growth of knowledge’ philosophers (Thomas Kuhn, Imre Lakatos, J.A.Agassi, and others) have gained prominence. Few economists keep up with developments in the philosophy of science, and as such it is understandable that many may still labor under the illusion that economics is, or can be, a positivist discipline. Part of the purpose of this study is to dispel that illusion by carefully documenting the demise of positivist thought within the philosophy of science.

Of course, positivism may not be dead, it may only be temporarily in eclipse. If the growth of knowledge approach which seems to be its successor leads nowhere or to speculative excess, we may witness a return to the rigorous and prescriptive models that characterize the positivist contribution. Whether or not that occurs does not concern us here; what is needed is a solid understanding of the present situation.

Thus we find that the study of economic methodology from a philosophy of science perspective has at least some chance of bearing some fruit. But it should also be emphasized that this is only part of the story. A complete study would include the practice of economists in various research traditions as a topic to be explored along with the writings of philosophers of science and of methodologists of economics. Though the scope of this work is restricted to the last two areas, that focus should not be interpreted as indicating that I believe the first area to be of less significance.

The format is straightforward. I begin with a review of some of the major themes in twentieth century philosophy of science from the logical positivism of the Vienna Circle to the present. Next I review in a collection of critical essays some of the major methodological writings in the positivist era in economics. In four concluding chapters I address the question: What form should methodology take in the post-positivist environment?

The amount of space devoted to the philosophy of science may seem too large in a book written by an economist for the use of economists. The rationale, again, is straightforward: so many errors and misrepresentations regarding ideas taken from philosophy have been made by economic methodologists, both past and present, that a lengthy summary is absolutely necessary if clarity is to be brought to the field. No such summary yet exists in the field, though a first step is contained in the first section of Mark Blaug’s admirable recent study, The Methodology of Economics (1980). As I will argue presently, however, Blaug’s presentation is less than complete because he is a falsificationist, and as such he sees the world of the philosophy of science through the glasses of a convinced Popperian. The treatment presented here differs from others most dramatically in its emphasis on the evolution of positivist thought from the dogmatism of the Vienna Circle to its most sophisticated form, logical empiricism. Neither proponents nor adversaries of positivist thinking in economics have previously devoted sufficient attention to the mature forms of positivist thought, and that needs remedying. In addition, it is my intention to show how the analyses of men like Kuhn, Lakatos, and Feyerabend are primarily a (as yet unassessed) response to a presumably failed positivism, a point that is too often overlooked by economists enamored with the particular models of historical change that are found in the writings of these growth of knowledge theorists. Each of the four chapters in the section on positivism and economic methodology begins with a review of either a famous debate between two economists, or a position statement by a prominent positivist economist. The four major topics treated are the Robbins-Hutchison debates on the status of the fundamental postulates of economics and the proper method of economic analysis; the Hutchison-Machlup debate on the necessity of testing assumptions; Friedman’s unique contribution, which I label methodological instrumentalism; and Samuelson’s espousal of operationalism and descriptivism. The debates and positions are examined in some detail because these original contributions are too often caricatured in the massive secondary literature that has sprung up in response to them. My assessment of their arguments is from the vantage point of philosophy of science, and I find that sometimes that vantage point provides worthwhile insights, and sometimes it is useless. (It will be shown, for example, that though many of Terence Hutchison’s contributions in his 1938 tract, The Significance and Basic Postulates of Economic Theory, were technically incorrect from a philosophical perspective, his larger contribution—the introduction of positivist language and thought into economic methodology—is of great importance.) These studies will aid us in our assessment of the strengths and limitations of using the philosophy of science in understanding economic methodology.

In the latter half of each of the chapters on economic methodology, related themes are treated. The content of these varies widely. In Chapter 6, I argue that the usual criticisms of Austrian economics (a branch of economics which has certain affinities with Robbins’s methodological views) have thus far proven unsuccessful, primarily because they proceed by assuming the validity of a rival epistemological or methodological system (e.g. positivism or falsificationism) then ‘proving’ that the Austrian system does not meet the qualifications of their presumptively true system. That chapter concludes with some suggestions of the form that a legitimate critique of Austrian methodology might take. In Chapter 7, I review some recent attempts to test the rationality assumption, and conclude that it is indeed untestable, at least as currently stated. In Chapters 8 and 9, some recent contributions to the secondary literature, which seem destined someday to be accorded the status of ‘original contributions’, are reviewed and evaluated: Lawrence Boland’s defense of Friedman’s instrumentalism, Stanley Wong’s critique via the method of rational reconstruction of Samuelson’s revealed preference theory, and Wilber and Harrison’s introduction of the pattern model of scientific explanation. These chapters conclude with a potpourri of opinions on current issues and alternative methodological schemas, as well as an interaction with some of whom I think are the best contemporary contributors to the methodological literature. What is not attempted is a comprehensive survey of current thinking in methodology. And again, the unifying theme is the vantage point from which I began—the chapters on philosophy of science which provide a common framework for interpretation and assessment.

An apt, if cynical, characterization of methodological study is that it is the systematic categorization of unanswerable questions. There are a number of questions that are referred to obliquely throughout this work that are best placed in the category of, if not unanswerable, then certainly unsettled. I mention them now because they are omnipresent, and because a person’s response to them is perhaps the crucial determinant of his or her beliefs concerning the importance and function of methodology, and perhaps one’s vision of the nature of science itself.

First, what is the best way to go about the study of methodology? Is there a single approach, or a plurality?

Next, is methodology essentially a prescriptive discipline, a descriptive one, or both? Note that this question itself can be posed either descriptively (What has been the role of methodology?) or prescriptively (What should it be?).

Finally, what is the best response to the perennial problem of theory choice? The problem arises because, in many significant instances in science, there are no objective criteria according to which competing theories may be compared, ranked, and evaluated; in a phrase, there is no algorithm of choice. Part of our task is to show that the theory choice problem is a real one in economics; the real question is, however: How do we respond to it?

All of these questions have been broached by the growth of knowledge theorists. I offer some views on each of them in the last chapters of the book.

PART ONE

---

TWENTIETH CENTURY PHILOSOPHY OF SCIENCE

---

2

---

Logical Positivism

---

The Vienna Circle

In 1922, the physicist and philosopher Moritz Schlick was appointed professor of the philosophy of inductive science at the University of Vienna. Three years later he organized a Thursday evening discussion group of philosophically-minded mathematicians and scientists. Though its membership varied over time, the group met regularly for the next eleven years, and through the efforts of its members a new philosop...