Independent Indian: Work & Life of Dr Subroto Roy

Philosophy of Economics

On the Scope of Reason in Economic Inquiry

Subroto Roy
© 1989, 1991, 2007 Subroto Roy

First published by Routledge of London & New York , 1989, in the International Library of Philosophy,

Library of Congress HB72.R69 1989

British Library 330’.01-dc19
Economics – philosophical perspectives
ISBN 0-415-03592-9

Reprinted in paperback, 1991
Library of Congress HB72.R69.1991
British Library 330’.01-dc20
ISBN0-415-06028-1

Preface to 2007 WordPress.com Republication

This book germinated when I was 18 or 19 years of age in Paris, Helsinki and London, and it was first published when I was 34 in Honolulu. I came to economics from natural science (biology, chemistry, physics), not mathematics. It was inevitable I would be drawn to the beauty of philosophy as a theoretical discipline while being driven, as a post-Independence Indian, to economics as the practical discipline that might unlock secrets to India’s prosperity and progress. I belonged to an ancient family of political men, and my father, who had joined India’s new foreign service the year before I was born, inculcated in me as a boy an idea that I had “a mission” (though he later forgot he had done so).

I was fortunate to fail to enter Oxford’s PPE and instead go to the London School of Economics. LSE was at an intellectual peak in the early 1970s. DHN Johnson in international law, ACL Day in international monetary economics, Brian Griffiths vs Marcus Miller in monetary economics with everyone still in awe of Harry Johnson’s graduate lectures in macroeconomics, Ken Wallis, Graham Mizon, JJ Thomas, David Hendry in econometrics with the odd lecture by Durbin himself – I was exposed to a fully grown up intellectual seriousness from the day I arrived as an 18 year old. Michio Morishima as my professorial tutor told me frankly that, as an Indian, I would face less prejudice in Western academia than in the private sector, and said he was speaking from experience as a fellow-Asian. He turned out to be wrong but it was wise advice nevertheless, just as wise as his requiring pupils to read Hicks’ Value and Capital (which, in our undergraduate mythology, he himself had read inside a Japanese gunboat during war).

What was relatively weak at LSE was general economic theory. We were good at deriving the Best Linear Unbiased Estimator but left unsatisfied with our grasp of the theory of value that constituted the roots of our discipline. I managed a First and was admitted to Cambridge as a Research Student in 1976, where fortune had Frank Hahn choose me as a student. That at the outset was protection from the communist cabal that ran “development economics” with whom almost all the Indians ended up. I was wholly impecunious in my first year as a Research Student, and had to, for example, proof-read Arrow and Hahn’s General Competitive Analysis for its second edition to receive 50 pounds sterling from Hahn which kept me going for a short time. My exposure to Hahn’s subtle, refined and depthless thought as an economist of the first rank led to fascination and wonderment, and I read and re-read his “On the notion of equilibrium in economics”, “On the foundations of monetary theory”, “Keynesian economics and general equilibrium theory” and other clear-headed attempts to integrate the theory of value with the theory of money — a project Wicksell and Marshall had (perhaps wisely) not attempted and Keynes, Hicks and Patinkin had failed at.

Hahn insisted a central question was to ask how money, which is intrinsically worthless, can have any value, why anyone should want to hold it. The practical relevance of this question is manifest. India today in 2007 has an inconvertible currency, vast and growing public debt financed by money-creation, and more than two dozen fiscally irresponsible State governments without money-creating powers. While pondering, over the last decade, whether India’s governance could be made more responsible if States were given money-creating powers, I have constantly had Hahn’s seemingly abstruse question from decades ago in mind, as to why anyone will want to hold State currencies in India, as to whether the equilibrium price of those monies would be positive. (Lerner in fact gave an answer in 1945 when he suggested that any money would have value if its issuer agreed to collect liabilities in it — as a State collects taxes – and that may be the simplest road that bridges the real/monetary divide.)

Though we were never personal friends and I did not ingratiate myself with Hahn as did many others, my respect for him only grew when I saw how he had protected my inchoate classical liberal arguments for India from the most vicious attacks that they were open to from the communists. My doctoral thesis, initially titled “A monetary theory for India”, had to be altered due to paucity of monetary data at the time, as well as the fact India’s problems of political economy and allocation of real resources were more pressing, and so the thesis became “On liberty and economic growth: preface to a philosophy for India”. When no internal examiner could be found, the University of Cambridge, at Hahn’s insistence, showed its greatness by appointing two externals: C. J. Bliss at Oxford and T. W. Hutchison at Birmingham, former students of Hahn and Joan Robinson respectively. My thesis received the most rigorous and fairest imaginable evaluation from them.

I had been attracted to Cambridge partly by its old reputation for philosophy, especially that of Wittgenstein. But I met no worthwhile philosophers there until a few months before I was to leave for the United States in 1980, when I chanced upon the work of Renford Bambrough. Hahn had challenged me with the question, “how are you so sure your value judgements promoting liberty blah-blah are better than those of Chenery and the development economists?” It was a question that led inevitably to ethics and its epistemology — when I chanced upon Bambrough’s work, and that of his philosophical master, John Wisdom, the immense expanse of metaphysics (or ontology) opened up as well. “Then felt I like some watcher of the skies, When a new planet swims into his ken; Or like stout Cortez when with eagle eyes, He star’d at the Pacific…”

It has taken me more than a quarter century to traverse some of that expanse; when I returned to Britain in 2004 as the Wincott Visiting Professor of Economics at the University of Buckingham, I was very kindly allowed to deliver a public lecture, “Science, Religion, Art and the Necessity of Freedom”, wherein I repaid a few of my debts to the forgotten work of Bambrough and Wisdom — whom I extravagantly compared with the Bodhisattvas of Mahayana Buddhism, also saying that the trio of Wittgenstein, Wisdom and Bambrough were reminiscent of what Socrates, Plato and Aristotle might have been like.

I had written to Bambrough from within Cambridge expressing my delight at finding his works and saying these were immensely important to economics; he had invited me to his weekly discussion groups at St John’s College but I could not attend. Between 1979 and 1989 we corresponded while I worked in America on my application of his and Wisdom’s work to problems in economics. We met only once when I returned to Cambridge from Blacksburg for my doctoral viva voce examination in January 1982. Six years later in 1988 he said of my Philosophy of Economics, “The work is altogether well-written and admirably clear”, and on another occasion he said he was “extremely pleased” at the interest I had taken in his work. The original preface of Philosophy of Economics said he was not responsible for the use I had made of his writings, which I reiterated in the 2004 lecture. At our meeting, he offered to introduce me to Wisdom who had returned to Cambridge from Oregon but I was too scared and declined, something I have always regretted since. It is only in the last few years that I have begun to grasp the immensity of Wisdom’s achievement in comprehending, explaining and extending the work of both Wittgenstein and Freud. His famous “Virginia Lectures” of 1957 were finally published by his admirers with his consent as Proof and Explanation just before his death in 1993. As for Bambrough, I believe he may have been or become the single greatest philosopher since Aristotle; he told me in correspondence there was an unfinished manuscript Principia Metaphysica (the prospectus of which appeared in Philosophy 1964), which unfortunately his family and successors knew nothing about; the fact he died almost in obscurity and was soon forgotten by his University speaks more about the contemporary state of academic philosophy than about him. (Similarly, the fact Hahn, Morishima and like others did not receive the so-called Economics “Nobel” says more about the award than it does about them.)

All I needed in 1980 was time and freedom to develop the contents of this book, and that I found in America — which I could not have done in either Britain or India. It would take eight or nine very strenuous years before the book could be written and published, mostly spent at Virginia Polytechnic Institute (1980-1985) and the University of Hawaii (1986-1990) Economics Departments, with short interludes at Cornell (Fall 1983) and Brigham Young (1985-86). I went to Virginia because James M. Buchanan was there, and he, along with FA Hayek, were whom Hahn decided to write on my behalf. Hayek said he was too old to accept me but wrote me kind and generous letters praising and hence encouraging my inchoate liberal thoughts and arguments. Buchanan was welcoming and I learnt much from him and his colleagues about the realities of public finance and democratic politics, which I quickly applied in my work on India, published in 1984 in London as Pricing, Planning & Politics: A Study of Economic Distortions in India and republished elsewhere here. The visit to the Cornell Economics Department was really so I could talk to Max Black the philosopher, who represented a different line of Wittgenstein’s students, and Max and I became friends until his death in 1988.

Buchanan’s departure from Blacksburg led to a gang of inert “game theorists” to arrive, and I was immediately under attack – one senior man telling me I was free to criticise the “social choice” work of Amartya Sen (since he was Indian too) but I was definitely unfree to do the same of Sen’s mentor, Kenneth Arrow, who was Jewish! (Arrow was infinitely more gracious when he himself responded to my criticism.) On top of that arose a matter of a woman, fresh off the aeroplane from India, being assaulted by a senior professor, and when I stood for her against her assailant, my time in Blacksburg was definitely up.

The manuscript of this book was at the time under contract with University of Chicago Press, and, thanks to Mrs Harry Johnson there, I had come in contact with that great American, Theodore W. Schultz. Schultz, at age 81, told me better to my face what the book was about than I had realised myself, namely, it was about economics as knowledge — its subject-matter was the epistemology of economics. Schultz wrote letters all over America on my behalf (as did Milton Friedman at Stanford and Sidney Alexander of MIT, whom I had also met and become friends with), and I was able to first spend a happy year among the Mormons at Brigham Young, and then end up at the University of Hawaii where I was given responsibility for the main graduate course in macroeconomics. I taught Harry Johnson-level IS-LM theory and Friedman-Tobin macroeconomics and then the new “rational expectations” vs Keynesian material.

I was also offered a large University grant to work on “South Asia”, which led to the books Foundations of India’s Political Economy: Towards an Agenda for the 1990s, and Foundations of Pakistan’s Political Economy: Towards an Agenda for the 1990s, both created by myself and WE James, and which led to the origins of India’s 1991 economic reform and the India-Pakistan peace process as told elsewhere. Also, this book came to be accepted for publication by Routledge, as the first economics book in its famed International Library of Philosophy.

Just as I was set to be evaluated for promotion and tenure at the University of Hawaii, I became the victim of a most vicious racist defamation (and there was some connection with Blacksburg). Quite fed up with the sordidness of American academia as I had experienced it, I sued in the federal court, which consumed much of the next half dozen years as the case worked its way through the United States Supreme Court twice. Milton Friedman and Theodore W. Schultz stood as expert witnesses on my behalf but you would not have known it from the judge’s ruling. There had been not only demonstrable perjury and suborning of perjury by the State of Hawaii’s officers, there was also “after-discovered” evidence of bribery of court-officers in the US District Court for the District of Hawaii, and I had to return to India in 1996 quite exhausted to recuperate from the experience. “Solicitation of counsel, clerks or judges” is “embracery curialis”, recognized as extrinsic fraud and subversion of justice since Jepps 72 E R 924 (1611), “firmly established in English practice long before the foundation” of the USA, Hazel Atlas, 322 US 238 (1943). “Embracery is an offense striking at the very foundation of civil society” says Corpus Juris 20, 496. A court of equity has inherent power to investigate if a judgement has been obtained by fraud, and that is a power to unearth it effectively, since no fraud is more odious than one to subvert justice. Cases include when “by reason of something done by the successful party… there was in fact no adversary trial or decision of the issue in the case. Where the unsuccessful party has been prevented from exhibiting fully his case, by fraud or deception practised on him by his opponent, as…where an attorney fraudulently or without authority assumes to represent a party and connives at his defeat; or where the attorney regularly employed corruptly sells out his client’s interest to the other side ~ these, and similar cases which show that there has never been a real contest in the trial or hearing of the case, are reasons for which a new suit may be sustained to set aside and annul the former judgment or decree, and open the case for a new and a fair hearing….” (Hazel Atlas). There is no time-limit in United States federal law for rectification of fraud on the court of this sort, and I remain fully hopeful today of the working of American justice in the case.

The practical result was that this book was never able to be properly publicized among economists as it would have been had I become Professor of Economics at the University of Hawaii by 1992 as expected. The hardback sold out quickly on its own steam and went into paperback by 1991, and a friend told me it was being used for a course at Yale Law School. The reviews were mostly intelligent. Upon returning to Britain as the Wincott Visiting Professor in 2004, I found times had changed and so had Routledge who would not keep it in print let alone permit a second revised edition. But I am now free to republish the book as I please, and today in 2007, with the Internet growing to a maturity which allows the young geeks at WordPress.com to want to encourage blogging worldwide, I can think of no more apt place to reproduce the first edition of this book than here at my own blog http://www.independentindian.com.

This is not a second or revised edition, and it is unchanged in content except for this lengthy new preface made necessary by the adventures and dramas the book’s author found himself unwittingly part of since its first publication. I am 52 now and happy to say I endorse the book just as I had published it at 34, though I do find it a little impatient and too terse in a few places. The 1991 paperback corrected a few slight errors in the 1989 hardback, and has been used. I am planning an entirely new book which shall have its roots in this one though it will be mostly in philosophy and not economics — the outlines it may take may be seen in the 2004 public lecture I gave on the work of Bambrough and Wisdom mentioned above and published elsewhere; its main aim will be to uncover for new generations the immense worth there is in their work which is in danger of being lost.

At least two names failed to appear in the original list of acknowledgements. G. Bruce Chapman, now of the University of Toronto, and I talked much of serious ethics and political philosophy when I first arrived at Cambridge in 1976. And in 1980 in Blacksburg, Anil Lal, then a graduate student and house-painter, borrowed my copy of Bambrough’s work, read it, and later made a comment on the metaphysics of John Wisdom which allowed me to see things more clearly.

Ballygunge, Kolkata,
April 7 2007

TO: R.A.R.

Contents

Preface

1. Introduction

Part I

2. Hume and the Economists

3. Understanding the Consensus

4. Difficulties with Moral Scepticism

Part II

5. Objectivity and Freedom

6. Expertise and Democracy

Part III

7. An Example from Microeconomics

8. A Dialogue in Macroeconomics

9. Mathematical Economics and Reality

10. Remarks on the Foundations of Welfare Economics

Envoi

Notes and References

Select Bibliography

Preface to First Edition
The publication of this work marks the end of an adventure of more than a decade and a half, most of the writing being done between 17 December 1980 and 22 May 1987. It has been quite perilous at times, especially as a foreigner in the West, and over the years many teachers, colleagues, friends and members of family have contributed to the author’s learning with their thoughts and actions. A number of senior scholars in economics and philosophy — especially Professor Frank Hahn, Professor James Buchanan, Professor Sidney Alexander, Professor Milton Friedman, Professor Max Black, Professor Sidney Alexander, Professor Amartya Sen, Professor Peter Bauer, Professor T. W. Hutchison and Dr C. J. Bliss, have lent their support to the work as it developed, even when they may have not known of its final form, or disagreed with its content, or been themselves a subject of its criticism. Most especially, the work has been honoured in the last six years with the unwavering encouragement of Professor T. W. Schultz of the University of Chicago. And Professor Ted Honderich of University College London has shown it the kindest consideration, without which publication would have been much delayed. Finally, a large philosophical debt will be seen to be owed to the work of Mr. Renford Bambrough of St. John’s College, Cambridge; however he should not be considered responsible for the use that has been made here of his writings.
HONOLULU
15 AUGUST 1988.

1. Introduction

1. IN this book, some of the central philosophical questions facing the modern economist will be raised. Most attention will be given to the question of the appropriate relationship between the positive and the normative, as well as to its parent question of the appropriate scope of objective reasoning in the making of evaluative judgements. Closely related is the question of the appropriate role of the economic expert in society, while slightly more distant questions have to do with the significance of interpersonal comparisons of utility, with the philosophical status of the concepts and theorems of mathematical economics, and with how judgements of probability should be understood. It is this family of questions which will be the concern of the present work.

Economics is a science with potentially important practical bearing upon the lives of men and nations. The state of the modern world may have been affected more profoundly and subtly by the use or misuse of economic knowledge than by many another science. Yet anyone familiar with the intellectual history of the field will know it to have seen more conflicts, and often conflicts of a more destructive kind, than may be reasonably expected or tolerated in the development of a scholarly discipline. The reader will be familiar with the many explicit and implicit divisions of opinion that have occurred upon theories and methods and evidence and policies, which have sometimes torn apart individual university departments and even threatened the integrity of the science itself. Indeed the modern economist in a despondent mood might be inclined to say of the state of his discipline as David Hume once said of philosophy: “There is nothing which is not the subject of debate, and in which men of learning are not of contrary opinions. The most trivial question escapes not our controversy, and in the most momentous we are not able to give any certain decision. Disputes are multiplied, as if everything was uncertain, and they are settled with the utmost warmth, as if everything was certain.”
At the same time as there have been deep and persistent divisions on substantive questions of economic theory and method and evidence and policy, there has been a deliberate or inadvertent consensus about the answer to an important question in the theory of knowledge. Modern economists happen to have been practically unanimous in their opinion on the possible scope of objective reasoning in the making of judgements, and thus in their opinion on the appropriate relationship between the positive and the normative. A broad consensus has developed to the effect that while common reasoning can have some scope in evaluative discussion, it is quite possible in practice and in principle for this scope to become exhausted. At such a point of the exhaustion of reason, only sheer and unadulterated subjective differences will be found to remain between people. Put another way, it has been believed possible for judgements ultimately to become immune to rational question and criticism.

Many of the pioneers of twentieth century economic thought, Kenneth J. Arrow, Milton Friedman, F. A. Hayek, Sir John Hicks, Oskar Lange, Gunnar Myrdal, Lionel Robbins, Joan Robinson, Paul Samuelson, Joseph Schumpeter, Jan Tinbergen, to name but a few, who between themselves would represent all of the main schools of contemporary economics, may be found to have shared such a thesis in the theory of knowledge, differing amongst themselves only upon the relatively minor question of the precise amount of room reasoning should be considered to have: some saying a great amount, others saying almost none, but all agreeing that whatever the exact amount it is a finite amount, both actually and potentially. The theory of demand, the theory of macroeconomic policy, the theory of welfare economics, the theory of social choice — each has in whole or in part rested upon an epistemological premise of this kind. If such a consensus can be shown to have existed, the reader may agree it to be something of a remarkable fact, since it would be difficult indeed to find a single substantive proposition of theory or method or evidence or policy to which a similar measure of consensus among modern economists might obtain.
One of the objects of the present work will be to argue that the fact there have been tremendous disharmonies on substantive economic questions, may not be independent of the fact there has been this kind of harmony in the theory of knowledge among many of the pioneers of twentieth century economics as well as the many more who have followed them. If the epistemological point hitherto accepted as true happens in fact to be false, it becomes possible that the scope of objective reasoning on substantive questions has been artificially prevented from being extended as far as it could have and should have been. Evaluative judgements are clearly of indefinite variety: attitudes towards goods or people, expectations of the future, recommendations to buy or sell, advice to a friend or a student or a government, etc. — roughly, all judgements taken by an individual or social agent about a right or optimal course of action in given circumstances. We shall find the consensus has been that it is possible for reasoning to come to a necessary halt in the process of coming to such judgements, whether the maker of the judgement is a public body or a private individual acting in the capacity of consumer or voter. A large amount (and possibly the whole amount) of what may deserve to be within the domain of common and objective reasoning comes to be placed instead under the rule of subjective will and caprice. Not only must we live with the fact that discussions between citizens or economists or politicians or spouses or states do frequently come to end without resolution, because there happens to be a lack of patience or tolerance or perseverance or good humour or whatever, but also that such outcomes may be written into the script from the start. In any normative discussion, we are to be permitted to call a unilateral halt merely by declaring “Well that is a value judgement of mine” or “That is a personal opinion of mine”, with the implication that any further questioning is out of bounds and unacceptable. Given a theory which allows us in this way to declare as we please what to call objective science and what to call subjective opinion, and given that it may be but human nature to be sceptical of the other fellow’s dogma while being oblivious to one’s own, we may have some explanation of how the consensus among economists in the theory of knowledge may have caused and preserved a state of affairs in which rival substantive dogmas can thrive — because the processes of common reasoning and even communication itself may have been allowed too often to come to a virtual standstill. (Or move at a snail’s pace.) “Disputes are multiplied, as if everything was uncertain, and they are settled with the utmost warmth, as if everything was certain.”

The gist of the present work will be that the present consensus in the theory of economic knowledge is logically inconsistent. It is therefore untenable and deserves to be abandoned. Men can aspire to, and in fact do attain and possess, certain and objective knowledge in an indefinite number of contexts. At the same time, there is no proposition of any kind held by anyone which must be thought of as necessarily closed to further question on grounds of reason or evidence. This simple maxim is something that may be found to hold in any field of human inquiry or endeavour one cares to mention — mathematics or medicine, ethics or physics, history or probability, logic or theology — and it will be our purpose in this work to examine its consequences in the context of economics in particular.

§2. Our study is one in what may be called theory of economic knowledge, and it may be worth a moment to consider what may be meant by this.

Bertrand Russell said of pure mathematics that it was a subject “in which we do not know what we are talking about” — meaning that the pure mathematician does not normally intend to refer in his theorems to substantive factual truths about the world. The epistemology or theory of knowledge of a discipline may be thought of similarly as being not concerned with either affirming or denying, corroborating or refuting the substantive propositions that happen to be made within the discipline. The study of the theory of economic knowledge may be thought of as not making any commitment one way or another to the substantive propositions which are to be found within the department of economics itself. Instead it is a more abstract undertaking, which seeks to examine certain kinds of questions from outside the department in the practical hope of dissolving or at least clarifying the character of substantive questions and controversies that may be occurring within. For example, to ask whether a criterion of truth and falsity can be applied to economic propositions, or whether objective knowledge is possible in the field, or how the kinds of propositions made in economics are to be justified, or how they compare and contrast with propositions made in other departments of inquiry — these would be the kinds of question we might see asked in the theory of economic knowledge; from which too the importance can be seen of generally abstaining from making substantive commitments in the process.

Much of the present work, especially Parts I and II, may be understood to be an attempt to provide a theory of economic knowledge of this kind. Thus the reader will not find in it commitments made to any substantive economic propositions. There is no theorem reported of the existence or efficiency of some new kind of economic equilibrium, no new model or evidence offered of the influence of the supply of money on prices, no new theory of how the expectations of economic agents may be formed or fulfilled or disappointed, no new evidence or explanation of why some country may be experiencing rapid growth or high inflation or increasing unemployment. No new result within economic science; one might almost say, nothing substantive! The present work will offer no more than “a machine to think with” on certain philosophical aspects of economics; it intends to leave economics as it is — and yet in so doing to have shown the way out of some of the philosophical difficulties that are encountered in its study. “For the clarity that we are aiming at is indeed complete clarity. But this simply means that the philosophical problems should completely disappear.”

Yet the practical purpose to making an investigation of this kind may be stated quite readily. For suppose, for sake of argument, we granted the truth of our simple maxim and assumed the epistemological concepts ‘knowledge’ and ‘doubt’, and their allied concepts ‘objectivity’ and ‘freedom’, should not be seen as incompatible in the project of inquiry. What consequences would follow from accepting such a viewpoint? Clearly first of all, we would be placed in a happy position of being able to say that no matter how deep or persistent the actual disagreements between economists or between citizens on economic questions happened to be, there is knowledge to be had in the study of economics. Not just high sophistry or rhetoric or political posturing or the opinions and prejudices of different people — but certain and objective knowledge about those actions, events, and phenomena that are part of the economic context. We would be able to say, in other words, there are at least some propositions in economics which are true, and which moreover can be known to be true.

An important ambiguity is possible here in asking whether there is knowledge about a given matter, insofar as such a question can be taken either as asking whether it is possible for there to be any knowledge about the matter, or as asking whether it is known that someone actually possesses such knowledge and how that has been determined. Defining as an expert someone who has the most reasonable and justifiable answer to give to a question, we need to distinguish, in other words, the relatively cool logical question of whether here can be any such thing as expert knowledge from the more heated political question of who is supposed to be such an expert and how we are supposed to know that. For instance, a question like “Is there a proof to Fermat’s last theorem?” can be understood either in the manner of the pure mathematician, as asking whether there can be a proof to the proposition it is impossible xn + yn = zn for positive integers x, y, z, n, and n > 2; or in the manner of the historian of mathematics, as asking whether any human being has come up with such a proof, as Fermat himself claimed to have done but of which no record exists. Among the great thinkers, Plato is the most influential to have crossed these wires in suggesting it possible not only for there to be objective knowledge about mathematics and ethics and statesmanship, but also for a special and closed set of experts to come to be identified to whom such knowledge should be thought of as being exclusively given. Plato’s theory can be and has been interpreted as giving license to elitism and dictatorship, yet the natural protest which the ideas of these would evoke in most of us may lead to an equal and opposite error of denying the very possibility of knowledge because we feared or wished to reject the idea of being ruled by a closed set of self-described experts. Once these wires are uncrossed, we may see it to be quite possible to maintain there can be objective knowledge and expertise in economics, without making any commitments toward specifying who should be considered an expert on some economic issue, or how we are supposed to determine that, or for that matter claiming any such knowledge or expertise for ourselves.

A second consequence of our simple maxim may seem more troubling. For by its second part, we should also have to say that even while there is objective knowledge in economics, there is nevertheless no proposition in the field which must be thought of as being necessarily closed to further question. Not the proposition that every human act is a rational act, nor the proposition that economic agents continually maximize utility, or are well modelled as doing so, nor the proposition that the market economy cannot be expected to reach full employment and needs to be and can be actively supplemented by macroeconomic policy, nor the proposition that the growth of money is necessary and sufficient for inflation, nor the proposition that free trade will maximize world output given factor immobility, nor the proposition that externalities imply a possible scope for taxes and subsidies, nor the proposition that the histories of nations is a history of class struggle.

By the second part of the maxim, there is no axiom or theorem of economic theory, no finding of economic history, no estimate of the value of an economic coefficient, no prediction of the course of an economic variable, no proposal of economic policy, which must be thought of as being closed to further question. None whatsoever. “No statement is immune to revision” (Quine).

Taken together, then, the net consequence of supposing objectivity and freedom, knowledge and doubt, to be compatible concepts deserving of equal respect, is that we shall be able to chart a course which steers us clear of two perennial and opposing hazards besetting all projects of human inquiry, viz., Scepticism and Dogmatism — the modern origins of which were traced by the American philosopher Charles Sanders Peirce to the cartesian proposal that philosophy “must begin with universal doubt, whereas scholasticism had never questioned fundamentals.” In the pages to follow, we will be denying universal doubt and we shall be free to question fundamentals. In an indefinite number of contexts, there is certain and objective knowledge to be had. Scepticism, understood technically as a logical thesis denying that we can possibly have or know that we have certain knowledge, is therefore a false thesis. At the same time, there is no proposition which is necessarily closed to question. Dogmatism, understood technically as a logical thesis implying there can be or must be some propositions which are absolutely and incorrigibly true, is therefore an equally false thesis. In place of a theory of knowledge restricting the scope of common reasoning to the finite or even the potentially finite, it is possible to have a theory of knowledge extending this scope to the potentially infinite. In particular, while normative proposals in economics or elsewhere may be supposed to be objectively better or worse depending on the soundness of the positive grounds given in their support, there are no unquestionable normative proposals — because there are no unquestionable positive grounds. The simple practical result of making the present investigation is that it will permit a sure and safe course to be found between Scepticism and Dogmatism for any project of economic inquiry.

§3. Would such a simple and straightforward thesis be new to economics in any way? To what extent would the argument which has been summarized above and which will be developed in the chapters to follow not been expressed before? The reader may wish an answer to such a question, and the author presently takes this to be as follows. With respect to the general debate which has occurred about knowledge and scepticism especially in moral philosophy, there will be little if anything in the present work which is a direct or novel contribution to it. While the philosophers have not been concerned with political economy at all, we shall be passive participants to their discussions, listening in to see what can be learned for our purposes and not intending to add to them directly. It may be remembered of course that it has not been long since economics formally broke away from philosophy to become a specialized discipline in its own right, in the belief the concerns of economics are of a more concrete and practical kind than those of philosophy. Since then we have made many highly abstract and theoretical claims, while also becoming scornful of philosophical thinking and believing ourselves to be exempt from its influences. Yet serious philosophical thought constitutes a mature and magnificent conversation which it would be foolish for any serious science to be deaf to. Moreover, it has been quite widely believed that there have been significant advances in philosophical understanding in the present century, and we are responsible to take such a claim seriously. It will be one of the aims of the present work to apply what may be learned from these discussions towards resolving, or at least clarifying, some of the main substantive disputations in modern economic science.

These are two broad traditions of moral philosophy relevant to our subject-matter, one deriving from Aristotle, the other from Hume (and a line of sceptics before him). Even though it would be unwise to expect agreement within either tradition, we may for convenience speak of an aristotelian and a humean tradition respectively. With respect to the discussions among economists on the relationship of the positive to the normative, we shall find an eminent consensus to have appeared on the humean side. This work will declare for the other side, and in so doing shall have to dissent from the humean consensus upon which all of the theory of social choice and much of the theory of welfare economics and theory of economic policy have appeared to rest. As far as is known by the author, there seem to have been but two published dissents on similar lines among economists in recent decades: those of Sidney Alexander and Amartya Sen. Of these, Professor Sen’s dissent has been very short and hesitant, and he would seem to have withdrawn it in other writings. Professor Alexander’s dissent has been clear and vigourous, but unlike his work on the balance of payments, his philosophical work has not received attention, and the present work was mostly developed in complete ignorance of its existence.

By the end of this work however, a clear choice should have been set out for the reader on the question of the relationship of the positive to the normative — between the consequences of accepting the humean consensus among economists and the consequences of the position of Professor Alexander and the author and possibly Professor Sen. The simple maxim “Objective knowledge is possible and yet there is no proposition which is closed to question” should not undermine its own content by being closed to question itself — instead it is supposed to refer and apply to itself as well. It may be true and deserving of our belief but it is not self evidently so, and will have to earn its credentials at the common bar of reason. Ultimately it will have to be the reader’s individual judgement whether it has been successfully shown that, contrary to what has been supposed by many of the pioneers of twentieth century economics, no conflict must arise between knowledge and doubt, objectivity and freedom. The history of the discussion may accord to our side the advantage J. S. Mill had seen to be enjoyed by all minority opinions: if the opinion of one or a few is false then not much will be lost by believing in it, while if it proves better able to stand the tests of time then much may be gained by allowing it to replace error. Put differently, it may seem quite risky that the pioneers of modern economic science have placed all their philosophical eggs in the humean basket — just in case it is Hume himself who happens to be mistaken.

§4. In Part I of the work will be found described the received theory of economic knowledge and its possible justification, as well as an account of the logical difficulties that arise with it. Chapter 2 has the task of documenting as fully as possible the existence of a humean consensus among economists in recent decades. Chapter 3 then examines the kinds of reasons that may incline us to be persuaded to such a view, and which may go to explaining how it has seemed to be an attractive theory to so many economists. These reasons appear to have been of two different but related sorts.

First the concept of value as used in ordinary life and ethics may have become confounded with the concept of economic value or scarcity or rareté in Walras’s term. Where economists have referred to a theory of value, they may have meant to refer more accurately to a theory of relative prices as determined by conditions of scarcity. The advance of the original neoclassicals in the late nineteenth century was to establish the importance of subjective estimations of economic agents to the determination of the relative prices of goods — as opposed to say how much labour went into different production processes as the classical economists might have said, or how much intrinsic value God had placed in the goods as the scholastics might have said. While it is clear by now that such an observation is broadly correct, it would be a mistake to go from a premise that market prices are determined in part by subjective estimations to a conclusion that the relative prices thus determined in any sense establish an order of how goods deserve to be valued or not. Goods are indeed valued the way they are because people happen to value them. Yet equally, in most cases, people seem to value goods in the way they do because the goods deserve to be thus valued — for example, because, like food or clothing or shelter, the goods are conducive to some valuable human purpose.
Secondly, it is possible the consensus has been motivated by a desire to find an effective shield against dogmatism and tyranny. For example, the context of an open parliamentary democracy presupposed by the modern theory of economic policy may have derived out of the experience of the great tyrannies of twentieth century history. There may have been a natural and understandable desire that the choices and decisions of citizens in the capacity of voters or consumers should be treated with the fullest due respect, and a humean scepticism may have been adopted because it has been believed to be something which is necessary and sufficient for this kind of respect to be shown. This would be an outstanding reason for adopting a humean point of view, and one which any critic must be required to account for. Yet it also places in relief the fatal self-contradiction that is present within the humean theory. For example, a theory of economic policy which has to rely upon an assumption of the polity being open and democratic would have to be silent about the conduct of economic policy in societies which were demonstrably not open or democratic, making it a theory very special and contingent in its range of application. Moreover, to give the defence of political or economic or religious freedom as a reason for holding a subjectivist epistemology would be to have left freedom entirely defenceless and toothless from those who would attack it from within precisely the same subjectivist framework. For example, if we conflated a general right to express an opinion freely with an idea that what such an opinion expresses is itself a matter of subjective opinion, then clearly, by the same token, an opinion that opinions should be freely expressed might also be considered merely subjective, and therefore no better or worse than its contrary. Within a subjectivist theory of knowledge, there ultimately can be nothing to choose between freedom and tyranny.

Chapter 4 is a survey of these kinds of logical difficulties with the humean position stated in Chapters 2 and 3. Its main result will be that the anti dogmatic campaign of the humean cannot succeed, and in fact comes to make the Sceptic resemble the Dogmatist more than anything else. It is possible this happens because both Sceptic and Dogmatist are sharing the same deductivist model of justification, to the effect that we cannot know a proposition to be true or right unless we have deduced it as the conclusion of a set of premises of whose truth or rightness we are certain. The Sceptic sees the threat of infinite regress that is implicit in such a model, and then denies we can be certain of anything. The Dogmatist sees the potential regress too, but responds to it by calling a halt at some arbitrary point, denying the need or possibility of going any further. In Part II a fresh picture will be given which attempts to preserve the truths the Sceptic and Dogmatist would each like us to take notice of, while correcting for the distortions both would force upon us by their unequivocal adoption of a deductivist model of justification. Chapter 5 reframes the main philosophical problems of Part I in the terms of the ancient dualism between Nominalism and Realism, and brings to light a possible resolution of this which has been advanced by a number of modern philosophers. Chapter 6 develops the argument further and applies it to the question of the appropriate role of expertise in a democracy. Taken together, Part II contains the main outlines of a fresh theory of economic knowledge with which to replace the flawed and inconsistent theory to which so many economists have thus far subscribed.

Part III of the work consists of a series of diverse illustrations and possible applications of the theory of knowledge developed in Part II. Chapters 7-10 all give examples of how inquiry and criticism can be seen to proceed in economics without sacrifice of either objectivity or freedom. Chapter 7 examines an actual debate on a concrete question of microeconomic policy, which may be compared and contrasted with the more academic examples of later chapters. Chapter 8 examines aspects of the division in macroeconomics and monetary theory since J. M. Keynes’s General Theory of Employment, Interest and Money. Chapter 9 considers a question with wide and general reference to economic theory: how the relationship between mathematical economics and real economic phenomena might be best understood. This has been the subject of long and bitter disputation, and some light is attempted to be shed on it from the vantage point of the philosophy of mathematics. It is possible that certain views in the philosophy of mathematics have been presupposed in modern mathematical economics; once these are exposed and aired, some of the conceptual problems which have been faced in this discussion may come to be dissolved. The theory of probability and expected utility and the theory of general equilibrium will be used as brief illustrations. Finally, in Chapter 10, the possible philosophical sources of the controversy surrounding the question of interpersonal comparisons of utility will be described, and a possible resolution suggested. This will be argued to have bearing on received understanding of the foundations of welfare economics.

§5. It will be found in the present work, then, that we shall be denying universal doubt on the one hand, while yet being free to question fundamentals on the other. Such a project will entail a critical examination of the philosophical premises and assumptions advanced by some of the most distinguished contemporary scholars in our field, and it is to be hoped the spirit in which the present criticism is offered will not be misunderstood. Every generation holds a peculiar advantage over preceding generations in having available to it what has gone before, while not being able to anticipate the criticisms of its own beliefs that will certainly come in the future. This kind of advantage that the present holds over the past may be thought of as being quite arbitrary, and we can expect it to carry with it a responsibility of taking what has gone before into serious account. Since no individual is able to do so on his own, we find every generation as a whole attempting to provide itself with critical discussions, which, when integrated over time, constitute the grand and unending conversation we call the history of human thought. It is with such a model in mind of a continuing and self-critical tradition of scholarship that we shall seek to address the questions raised at the beginning about the foundations of economic knowledge, while not making any pretence whatsoever to finality, and instead leaving the entire treatment as open as it can be made to the examination and criticism of others.

PART I

2. Hume and the Economists
THERE has been a broad and long standing consensus among economists about the character of the relationship between positive and normative propositions, as well as about the related question of the appropriate scope and limits of economic expertise in society. Joining in this consensus have been many of the pioneers of twentieth century economic thought: Kenneth J. Arrow, Milton Friedman, F. A. Hayek, Sir John Hicks, Oskar Lange, Gunnar Myrdal, Lionel Robbins, Joan Robinson, Paul A. Samuelson, Joseph Schumpeter, Jan Tinbergen, to name but a few. Many others are likely to be found in explicit or implicit agreement, while a survey by Professor T. W. Hutchison suggests that some of the most renowned figures of nineteenth century economics should probably be included as well. The main purpose of this chapter will be to provide enough documentary evidence to show that such a consensus has in fact existed. When we think of how many deep and wide differences there have been over the years in the field that was once called political economy and is now called economic science, differences on questions of method and theory and evidence and recommendations of policy, the existence of such a consensus may seem quite a remarkable fact.

Very briefly, what appears to have been accepted is that it is possible to identify a body of progressively changing knowledge called ‘positive economics’, which is the main contribution of economists to human knowledge and understanding in general. It consists of such things as the microeconomic and macroeconomic descriptions of present and past states of an economy, conditional predictions of such states in the future, hypothetical or substantive explanations of what economic causes may have what economic effects, the deduction and analysis of theorems of economic significance, and so on. That is to say, positive economics has been supposed to consist of the domain of propositions in an economic context which have to do in one way or another with questions of what is the case, or with what has been the case in the past or may be expected to be the case in the future. In contrast, evaluative or prescriptive or ‘normative’ propositions, having in one way or another to do with what ought to be done or not done by a government or a private economic agent, have been believed to fall into quite a different category. These have been believed to amount sooner or later to being expressions of subjective personal opinion, either on the part of the individual economist himself or of those whom he may happen to be advising.

Most economists who have considered the matter have allowed that there is usually at least some scope, and sometimes much scope, for common reasoning on logical and empirical grounds to be brought to bear in normative discussion; making it possible that at least some of the disagreements between economists or citizens or politicians on normative questions can come to be objectively resolved. But it has been believed possible also for the processes of common reasoning to become exhausted in discussions of normative questions like those of economic policy or ethics or jurisprudence, in a way they are not supposed to become exhausted in discussions of positive questions like those of economic theory or econometrics or natural science or mathematics. Once such a point of the exhaustion of reason has been reached, any residual conflict which remains is to be considered necessarily irreconcilable and of a sheer normative kind. And such sheer normative opinions, upon which it is not possible to bring to bear any further objective consideration, are to be supposed to express the purely subjective attitudes and feelings of the individual person, opinions which might happen to be shared by others too, but which are certainly closed to further argumentation, whether in public or in the person’s own mind. Put a little differently, the theory of knowledge and policy which we shall see to have been widely accepted by many economists in the twentieth century, has made an assumption that while all questions of analysis and evidence can have objectively true or false answers, only some and not all questions of evaluation and prescription can have objectively right or wrong answers.

§2. Underlying the consensus among economists has been a more general thesis in the theory of knowledge or epistemology. It is a thesis which may be called ‘moral scepticism’, and its most brilliant and influential exponent in the modern period has been David Hume (1711-1776). Among those to have advanced influential and persuasive points of view of a similar kind in twentieth century moral and political philosophy have been C. L. Stevenson, R. M. Hare, A. J. Ayer, and Karl Popper.

In the course of a critique of dogmatic religion and ethics, the young Hume was to attack with a sceptical scalpel what he took to be the illogic of trying to deduce evaluation and prescription from analysis and description: “In every system of morality, which I have hitherto met with… the author proceeds for some time in the ordinary way of reasoning… when of a sudden I am surpriz’d to find, that instead of the usual copulations of propositions, is, and is not, I meet with no proposition that is not connected with an ought or an ought not. This change is imperceptible; but is, however, of the last consequence. For as this ought, or ought not, expresses a new relation or affirmation, ’tis necessary that it shou’d be observ’d and explain’d; and at the same time that a reason should be given, for what seems altogether inconceivable, how this new relation can be a deduction from others, which are entirely different from it.” While the precise context and implications of this passage continue to divide philosophers, it will be adequate for our present purpose to follow the sympathetic and influential modern interpretation given by the Oxford moral philosopher R. M. Hare, and obtain for an economic context what may be called Hume’s First Law: No normative conclusion, for example, about what a private economic agent or a government ought to do or not do, can be validly deduced from a set of solely positive premises, i.e., from premises which only describe what is the case. No normative conclusion can be deduced without at least one normative premise having been made. A dualism of this kind between the ‘is’ and the ‘ought’ has been frequently supposed to separate science from ethics, the objective from the subjective, the rational from the irrational, public knowledge from private opinion.

Hume was to reinforce this opinion a decade later in a more recondite form of words: “[A]fter every circumstance and every relation is known, the understanding has no further room to operate, nor any object on which it could employ itself. The approbation or blame which then ensues cannot be the work of the judgement, but of the heart; and is not a speculative proposition or affirmation, but an active feeling or sentiment.” This passage too continues to divide philosophers, but for our present purpose R. M. Hare’s recent writing is once more helpful in obtaining a modern interpretation. Hare asks whether, in addition to logical questions and factual questions about how the world is, there can be “irreducibly evaluative or prescriptive questions” as well; once we have “done all we can” by way of reasoning and adducing evidence, “will there remain something to be done which is neither logic nor fact finding but pure evaluation or prescription?” Hare answers yes it is possible, and in the same vein we may restate the idea to obtain for an economic context what may be called Hume’s Second Law: After every empirical question and every logical and mathematical question has been answered in an economic problem, there is no further scope for common reasoning to work. If an evaluative statement is made at such a point, then it can express no more than a subjective attitude or feeling of the individual economist towards the subject.

This is a maxim which does grant that a measure of common reasoning and evidence can be brought to bear upon particular normative questions, and so some normative disagreements may come to be objectively resolved. But it also allows for the potential for such reasoning to become exhausted, leaving merely a subjective residue of personal sentiment or feeling which people might or might not happen to share with one another but which would be beyond further question and discussion. In the pages to follow, a position will be referred to as ‘humean’ if it implicitly or explicitly endorses one or both of Hume’s Laws as stated above. The small h is used to suggest that a close examination of Hume’s works may show him to have been not entirely clear in his own meaning, as well as to suggest that the question of what Hume himself may have actually or fully meant is not of as direct importance for the present purpose as the question of what he has been taken to mean by contemporary economists.

The remainder of this chapter is given to documenting at fair length the fact that a number of the pioneers of twentieth century economics have quite unambiguously seemed to endorse a humean point of view in the theory of knowledge. Chapter 3 will be given to placing this fact in an appropriate historical context. This needs to be done not only in order to understand the nature of the consensus as fully as possible, but also to realize how close economists have been to one another on a central question in the theory of knowledge, even while being engaged in any number of deep and well known and seemingly interminable disputes on substantive matters. The reader who may be impatient with a detailed record of this kind, or who is prepared for the present to take its existence for granted, may wish to move on directly to Chapter 3 without losing the main threads of the argument.

§3. Friedman. Following Neville Keynes, Professor Milton Friedman has clearly and emphatically argued the importance of extending the scope of common reasoning in economics: “Positive economics is in principle independent of any particular ethical position or normative judgments…. [it] is, or can be, an ‘objective’ science, in precisely the same sense as any of the physical sciences…. Normative economics and the art of economics, on the other hand, cannot be independent of positive economics…. differences about economic policy among disinterested citizens derive predominantly from different predictions about the economic consequences of taking action — differences that in principle can be eliminated by the progress of positive economics — rather than from fundamental differences in basic values, differences about which men can ultimately only fight.” It is well known that in this and other works, Friedman has argued for the extension of common reasoning and evidence, or positive economics, as the surest means to resolving normative disputations. Yet from the passage quoted, it is clear that Friedman has also accepted something like Hume’s Second Law, to the effect that while common reasoning can have some and indeed much scope, a point of ultimate and sheer normative disagreement can still be reached, distant though it might be, where reasoning must be considered to have become exhausted and “men can ultimately only fight”. In the same essay, Friedman added that it was the practical importance of economics which impeded objectivity and promoted confusion between “scientific analysis and normative judgment”, suggesting an endorsement of Hume’s First Law as well.

Myrdal. Gunnar Myrdal argued for many years that a number of economic concepts purporting to be analytical or descriptive in character in fact had evaluative or prescriptive overtones. Myrdal and his editor and translator, Professor Paul Streeten, argued that a view that there is no place for normative judgments in economic science has been a guise for the advocacy of a specifically liberal political economy, a thesis which might well be endorsed by many marxian and keynesian economists. While postponing an assessment of this claim to a later chapter, we may note that Myrdal also happened to endorse the extension of the scope of positive economics, with as much emphasis as Friedman would do after him: “By subjecting to impartial criticism those arguments in political controversies which concern the facts and the causal relations between them, economic science can make an important contribution to the political sphere. As often as not, conflicting political opinions spring not so much from divergent valuations about the best possible future state of society and the proper policy for securing it, as from subjectively coloured and therefore distorted beliefs regarding actual social conditions.” Myrdal went on to endorse Hume’s First Law in recommending that the economist leave the supply of evaluative premises to the politician. While the economist can provide descriptions, explanations and conditional predictions, “the scientist must not venture beyond this. If he wishes to go further he needs another set of premises, which is not available to science: an evaluation to guide him in his choice of the effects which are politically desirable and the means permissible for achieving them.” Finally, Myrdal reached the humean conclusion that the normative differences between economists are ultimately beyond objective resolution: “[E]conomic reasoning is often obscured by the fact that normative principles are not introduced explicitly, but in the shape of general ‘concepts’. The discussion is thus shifted from the normative to the logical plane. On the former there is either harmony or conflict; conflict can only be stated, not solved by discussion. On the logical plane we should define our concepts clearly and then operate with them in a logically correct manner. What is ‘correct’ and what ‘false’ can be discussed with the methods of logic, whereas conflicting interests can be recognized, never solved scientifically.”

Robbins. In his influential writings over many years, Lionel Robbins made a distinction between ‘economic science’, having to do with such questions as how best to allocate scarce resources between alternative ends, and ‘political economy’ or normative theories of economic policy, prescribing the ends themselves and the weights to be attached to them. In his well known methodological work we read as clear a statement of Hume’s First Law as might be found in economics: “Propositions involving ‘ought’ are on an entirely different plane from propositions involving ‘is’…. Economics is neutral as between ends. Economics cannot pronounce on the validity of ultimate judgements of value…. Economics deals with ascertainable facts; ethics with values and obligations. The two fields of inquiry are not on the same plane of discourse. Between the generalizations of positive and normative studies there is a logical gulf fixed which no ingenuity can disguise and no juxtaposition in space or time can bridge over.” Robbins’s endorsement of the Second Law was equally emphatic. While positive economics extends the scope of common reasoning, it is still possible to find normative differences which are rationally irresoluble: “If we disagree about ends it is a case of thy blood or mine — or live and let live according to the importance of the difference or the relative strength of our opponents. But if we disagree on means, then scientific analysis can often help us to resolve our differences. If we disagree about the morality of the taking of interest (and we understand what we are talking about), then there is no room for argument.”
Samuelson. Professor Paul Samuelson has seemed to feel a tension in the humean position, but also that its logic compelled him to follow closely in Robbins’s path: “It is fashionable for the modern economist to insist that ethical value judgments have no place in scientific analysis. Professor Robbins in particular has insisted upon this point, and today it is customary to make a distinction between the pure analysis of Robbins qua economist and his propaganda, condemnations and policy recommendations qua citizen. In practice, if pushed to extremes, this somewhat schizophrenic rule becomes difficult to adhere to, and it leads to rather tedious cicumlocutions. But in essence Robbins is undoubtedly correct. Wishful thinking is a powerful deterrent of good analysis and description, and ethical conclusions cannot be verified in the same way that scientific hypotheses are inferred or verified.”

Hicks. Like Samuelson, Professor Sir John Hicks has seemed to feel a tension in the humean position, yet he too must be considered as having endorsed at least an important version of it. On the one hand, Hicks has seemed critical of mid-century positivism and emotivism, and claimed the main rationale of the “new welfare economics” to be that it allowed a route of escape from them. “During the nineteenth century, it was generally considered to be the business of an economist, not only to explain the economic world as it is and as it has been, not only to make prognostications (so far as he was able) about the future course of economic events, but also to lay down principles of economic policy, to say what policies are likely to be conducive to social welfare, and what policies are likely to lead to waste and impoverishment.” Since then positivism had declared that explanation and only explanation may be part of scientific economics, and any move to prescribe “must depend upon the scale of social values held by the particular investigator. Such conclusions can possess no validity for anyone who lives outside the circle in which these values find acceptance. Positive economics can be, and ought to be, the same for all men; one’s welfare economics will inevitably be different according as one is a liberal or a socialist, a nationalist or an internationalist, a christian or a pagan.” But such a position is “rather a dreadful thing to have to accept”, one which might “become an excuse for the shirking of live issues, very conducive to the euthanasia of our science.” Fortunately we are not compelled to accept it, since the new welfare economics advanced by Kaldor, Hotelling and Hicks himself was a viable alternative, not open to the objections the positivists had raised to the utilitarianism of Pigou and others.

Yet we may ask, what had the new welfare economics been about? And did it in fact make a break with the positivism which seemed to be troubling Hicks, or had it not been prompted precisely by humean doubts? As is well known, the new welfare economics had to do with questions such as whether the potential gainers from a change in policy could possibly compensate the potential losers from the change by enough so as to get them to go along with it, or conversely for the losers from a change to compensate the gainers from the change by enough so as to get them to go along without it, and so on. As Hicks himself makes clear, it was a discussion very much motivated by the belief that while the Pareto criterion was not a wholly adequate substitute for the utilitarianism of Pigou, any emendation of the paretian theory must leave untouched its basic positivistic premise, viz., that interpersonal comparisons cannot be conceived of as anything but purely subjective judgements, outside the scope of objective reasoning. Hicks claimed it was because the new welfare economics avoided making interpersonal comparisons that it should be considered a positive advance, a scientific advance. And Hicks has emphasized that he, like Robbins, has not wanted any truck with interpersonal comparisons. The old welfare economics of Pigou required one “to admit the possibility of comparing the satisfactions derived from their wealth by different individuals. This is where Professor Robbins parts company; for my part, I go with him.” More recently: “A single individual… shows by his choices that he prefers one thing to another; we may put this, if we like, in the form of saying that he derives (or thinks he derives) greater satisfaction from the one than from the other. But there is no similar way in which we can see that the satisfaction derived by one individual from one good is greater than the satisfaction derived by another individual from another good; these satisfactions are not compared in any actual choice, so that for the comparison between them there is not the same evidence.”
While we shall be returning to these questions in Chapter 10, what we may note here is that since interpersonal comparisons certainly amount to being a particular species of evaluative judgement, Hicks’s scepticism with respect to the possibility of making them objectively must be considered to amount to an endorsement of at least a species of moral scepticism. If so, it would seem to sit uncomfortably with Hicks’s opinion that he had not cared much for the positivist dichotomy between explanatory science and subjective prescriptions, which was said to have prompted the search for the new welfare economics in the first place.

Robinson. Writing on the theory of employment, Joan Robinson was to give a superbly clear account of the humean position at its best, which requires no commentary: “[All economic] controversies should be capable of resolution. The rules of logic and the laws of evidence are the same for everyone, and in the nature of the case there can be nothing to dispute about. Controversies arise for five main reasons. First, they occur when the two parties fail to understand each other. Here patience and toleration should provide a cure. Second, controversies occur in which one (or both) of the parties have made an error of logic. Here the spectators at least should be able to decide on which side reason lies. Third, two parties may be making, unwittingly, different assumptions, and each maintaining something which is correct on the appropriate assumptions…. Here the remedy is to discover the assumptions and to set each argument out in a manner which makes clear that it is not inconsistent with the other. Fourth, there may not be sufficient evidence to settle a question of fact conclusively one way or the other. Here the remedy is for each party to preserve an open mind and to assist in the search for further evidence. Fifth, there may be differences of opinion as to what is a desirable state of affairs. Here no resolution is possible, since judgements of ultimate values cannot be settled by any purely intellectual process…. argument in the nature of the case can make no difference to ultimate judgements based on interest or moral feeling. The ideal is to set out all the arguments fairly on their merits, and agree to differ about ultimate values. On questions of policy, the differences can never be resolved.”

Hayek. Professor F. A. Hayek has stated an unambiguous commitment to Hume’s First Law, as when he wrote recently: “Our starting point must be the logical truism that from premises containing only statements about cause and effect, we can derive no conclusions about what ought to be.” In his earlier discussion of the economics of socialism, Hayek had hinted at the Second Law as well, saying that “problems of ethics, or rather of individual judgements of value… [are]… ones on which different people might agree or disagree, but on which no reasoned arguments would be possible.” If the questions about socialist planning are ethical by this definition then “no scientist, least of all the economist” would have anything to say about them. Positive argument presumes there to be some common values between the participants: “Meaningful discussion about public affairs is clearly possible only with persons with whom we share at least some values. I doubt if we could even fully understand what someone says if we had no values whatever in common with him. This means, however, that in practically any discussion it will be in principle possible to show that some of the policies one person advocates are inconsistent or irreconcilable with some other beliefs he holds.” In particular, the argument over socialist planning should be seen to be one on positive grounds: “[E]veryone desires, of course, that we should handle our common problems as rationally as possible and that, in so doing, we should use as much foresight as we can command. In this sense, everybody who is not a complete fatalist is a planner, every political act is (or ought to be) an act of planning, and there can be differences only between good and bad, between wise and foresighted and foolish and shortsighted planning. An economist, whose whole task is to study how men actually do and how they might plan their affairs is the last person who could object to planning in this general sense.” The dispute between socialists and their critics is “not a dispute about whether planning is to be done or not. It is a dispute as to whether planning is to be done centrally, by one authority for the whole economic system, or is to be divided among many individuals.”

Lange. Oskar Lange, the famous adversary of Hayek and Robbins on the question of socialist planning, was agreed with them that the only task within the scope of scientific economics was the determination of the best means, with economic ends having been decided politically. He gave this infelicitous analogy to the economist’s role: “The situation may be compared with that of two physicians treating a patient. There is no necessity of interpersonal agreement about the objective of the treatment. One physician may want to heal the patient, the other may want to kill him (e.g., the patient may be a Jew in a Nazi concentration camp; one physician may be a fellow prisoner who wants to help him, the other may be a Nazi acting under orders to exterminate Jews). But once the objective is set for the purposes under discussion (either of the two physicians may, of course, refuse to act upon it), their statements as to whether a given treatment is conducive to the end under consideration have interpersonal validity. Any disagreement between them can be settled by appeal to fact and to the rules of scientific procedure.”

Schumpeter. In discussing the wertfrei controversy between Carl Menger and the German historical school, Joseph Schumpeter was to suggest that the epistemological matters involved were neither difficult nor interesting and could be disposed of shortly. The distinction between ‘is’ and ‘ought’ had been correctly and adequately drawn already, so it only needed to be accepted that an ‘ought’ statement “that is to say, a precept or advice, can for our purpose be reduced to a statement about preference or ‘desirability’.” Schumpeter went on to endorse Hume’s First Law, saying that an acceptance of one value judgement always requires the acceptance of others. This “is of little moment when the ‘ultimate’ value judgments to which we are led up as we go on asking why an individual evaluates as he does, are common to all normal men in our cultural environment.” Unlike Lange, Schumpeter gave the physician as a negative analogy: “[T]here is no harm in the physician’s contention that the advice he gives follows from scientific premises, because the — strictly speaking extra-scientific — value judgment involved is common to all normal men in our cultural environment. We all mean pretty much the same thing when we speak of health and find it desirable to enjoy good health. But we do not mean the same thing when we speak of the Common Good, simply because we hopelessly differ in those cultural visions with reference to which the common good has to be defined in any particular case.” I.e., common reasoning can proceed in normative discussion but only so long as we find common values among “all normal men in our cultural environment”, which is to suggest reasoning may be helpless with abnormal men or those who are outside our cultural environment. Further, siding with Menger, Schumpeter suggested that the bitterness of the wertfrei controversy could be explained because it had been not so much a logical dispute as one between those who were practising and those who were protesting a kind of scholarly deceit, viz., the propagation of personal dogmas within an ostensible pursuit of objective knowledge: “Those who profess to be engaged in the task of widening, deepening, and ‘tooling’ humanity’s stock of knowledge and who claim the privilege that civilized societies are in the habit of granting to the votaries of this particular pursuit, fail to fulfil their contract if, in the sheltering garb of the scientist, they devote themselves to what really is a kind of political propaganda.”

Arrow. In opening his famous paper on the theory of social choice, Professor Kenneth J. Arrow was to refer explicitly to the ancient ontological dualism between Nominalism and Realism. To take aggregate rankings of “social states” as independent of individual rankings “is to assume, with traditional social philosophy of the Platonic realist variety, that there exists an objective social good defined independently of individual desires. This social good, it was frequently held, could be best apprehended by the methods of philosophic inquiry. Such a philosophy could be and was used to justify government by elite, secular or religious, although the connection is not a necessary one. To the nominalist temperament of the modern period the assumption of the existence of the social ideal in some Platonic realm of being was meaningless.” Nineteenth century utilitarianism had “sought instead to ground the social good on the good of individuals”, which, when combined with a hedonistic psychology, implied “each individual’s good was identical with his desires” and “the social good was in some sense to be a composite of the desires of individuals.” Such a view “serves as a justification of both political democracy and laissez faire economics, or at least an economic system involving free choice of goods by consumers and of occupations by workers.”
While Arrow found it necessary to remark that a connection between elitist rule and a Realist ontology was “not a necessary one”, he did not also remark upon whether he took a connection between democratic rule and a Nominalist ontology to be logically necessary. If not, then we might of course entertain other cases equally well, such as Nominalism being associated with elitist rule, or Realism with democratic rule, or perhaps more subtle cases which may arise from a denial of the dualism altogether — matters to which we shall return more explicitly in Part II. In any case, it would seem evident Arrow’s sympathy has been with the humean thesis, which he endorses strongly in suggesting, like Schumpeter, that no distinction can be made between a personal preference and a judgement of value: “One might want to reserve the term ‘values’ for a specially elevated or noble set of choices. Perhaps choices in general might be referred to as ‘tastes’. We do not ordinarily think of the preference for additional bread over additional beer as being a value worthy of philosophical inquiry. I believe, though, that the distinction cannot be made logically, and certainly not in dealing with the single isolated individual. If there is any distinction between values and tastes it must lie in the realm of interpersonal relations.” That Arrow believes normative questions to be only personally and subjectively answerable is further suggested by his remarks that “[t]he only rational defense of what may be termed a liberal position… is that it is itself a value judgment”; that his own values are such he is willing “to go very far indeed in the direction of respect for the means by which others choose to derive their satisfactions”; that he personally shares “a strongly affirmed egalitarianism, to be departed from only when it is in the interest of all to do so”; that he is personally “in favor of very wide toleration”; and so on. In Chapters 9 and 10, we shall return to examine certain aspects of the theories of general equilibrium and social choice which Professor Arrow has helped pioneer.

Blaug. In his influential writings in the history and methodology of economics, Professor Mark Blaug has appealed directly to Hume, declaring that the “orthodox Weberian position on wertfrei social science is essentially a matter of logic: as David Hume taught us, ‘you can’t deduce ought from is’.” Blaug grants that scientific practice does continually call for the exercise of judgement, but he wishes to distinguish “methodological” judgements, having to do with such questions as “the levels of statistical significance, selection of data, assessment of their reliability, and adherence to the canons of formal logic”, from “normative” or “appraising” judgements, which “refer to evaluative assertions about states of the world, including the desirability of certain kinds of behavior and the social outcomes that are produced by that behavior; thus all statements of the ‘good society’ are appraising value judgments.” It is judgements of this latter sort which are “incapable of being eliminated in positive science”. In support of such a dualism Blaug claims “there are long established, well tried methods for reconciling different methodological judgments” but none “for reconciling different normative value judgments — other than political elections and shooting it out at the barricades.” Blaug’s acceptance of Hume’s Second Law is as explicit as may be found in contemporary economics. There sometimes can be rational discussion over normative differences “and that is all to the good because there is a firmer tradition for settling disputes about facts than for settling disputes about values. It is only when we distill a pure value judgment… that we have exhausted the possibilities of rational analysis and discussion.” Echoing Robbins, Blaug suggests that at such a terminal point we are left with “factual statements and pure value judgments between which there is indeed an irreconcilable gulf on anyone’s interpretation.” Like Arrow, Blaug also makes reference to an ontological division between Realism (or “essentialism”) and Nominalism, and hints at a necessary link between a Realist ontology and dogmatism and tyranny. From Plato and Aristotle up through the nineteenth century, Western thought had been under the malign and mistaken impression that “it is the aim of science to discover the true nature or essence of things”. Such a view “raises its ugly head” even today, and Blaug charges the authors of a recent marxian thesis as being one such recent manifestation: “Adherents of essentialism are inclined to settle substantive questions by reaching for a dictionary of their own making, and Hollis and Nell exemplify this tendency to perfection: reproduction is the ‘essence’ of economic systems because we tell you so!”

Hahn. Professor Frank Hahn reports that contemporary economists “in keeping with the Positivist perspective” make “a thorough distinction of ‘is’ from ‘ought’ (positive from normative).” While Hahn has been mostly guarded in his own opinion as to the precise relationship between positive and normative, he has suggested recently that while normative questions are subject to reasonable argument, and economic theory is intended to widen this scope of common reasoning, “the intention is to take a small step in distilling what are genuinely questions of values.” Such a remark would seem to place Hahn among the moderate humeans like Joan Robinson and Milton Friedman — which in turn would make it an interesting fact that while Hahn has had long and well known disputes on substantive matters with both Friedman and Robinson, he would appear closely agreed with them on a point in the theory of knowledge, viz., that while there is much room for objective discussion to take place, it is possible for sheer differences of a normative kind to exist and come to be identified.

A few others. To take some final examples, Professor Robert Sugden affirms “Hume’s Law reflects a liberal view of the universe”; Professor William Baumol and Professor Allan Blinder write in their textbook that the economist defines rational decisions as those “that are most effective in helping the decision maker achieve his own objectives, whatever they may be”; Professor James Quirk writes in his textbook that “normative economics is based on a system of axioms, but these axioms concern ethics” and because these and any propositions derived from them are not “verifiable through empirical observation”, a person is “free to accept or reject the conclusions of normative economics as he wishes, simply by accepting or rejecting the axiom system — there are no scientific issues involved.” And Professor Jack Hirschleifer wrote in his textbook that “if one economist prefers Maoism and another capitalism, or if one prefers to exterminate and the other to tolerate an inconvenient minority group, the fundamental sources of contention are almost surely divergences in ethical values… [which] will not be eliminated by advances in scientific economics.”

3. Understanding the Consensus
THE great German philosopher and mathematician Gottlob Frege suggested at one place that we should not “ask for the meaning of a word in isolation, but only in the context of a proposition.” In the same vein, it may be said the meaning of a proposition or a hypothesis should not be asked for except in relation to the particular context in which it has been advanced. And we can maintain this without requiring the description of such a context to be fully explicit or even one which can be easily expressed in words. A proposition needs to be understood in relation to the fullest possible description of its implicit and explicit context — which may be a good sense too in which to understand the reference by Wittgenstein to the concept of a “language game” .

In the previous chapter, we have marshalled considerable evidence for our initial thesis that there has been a broad measure of consensus among many of the pioneers of modern economics about the appropriate relationship of the positive to the normative. Irrespective of their many and well known substantive differences, they have seemed all to share an affinity with a humean thesis of moral scepticism, whether in a radical way like Schumpeter and Professor Arrow when they say there can be no difference in kind between personal preferences and value judgements, or in a more moderate way like Joan Robinson and Professor Friedman and Professor Hahn, when they say there can be a great amount of room for objective argumentation to take place about normative questions before a naked and irreconcilable difference will be found to appear. The first question that needs now to be addressed is how this consensus should be understood, and this will require as full a description as can be attempted in this work of the context in which it has occurred. The second question would be whether or not the consensus is correct and justified — whether or not there are firm and adequate grounds for us to think we should join it, and so take the is ought dualism to be a barrier which it is neither possible nor necessary to surmount. The reader will have known from the Introduction that it is a main purpose of this study to make the argument that such grounds are not in fact available, that a humean position is ultimately untenable and misleading, and deserves to give way to a theory of economic knowledge and policy which treated objectivity and freedom as compatible concepts deserving of equal respect. Nevertheless we are first obliged to identify the strengths and motivations of a humean point of view, if only so that we might explain how it has come to command the kind of assent it has done among many of the most eminent of twentieth century economists as well as the many more who have followed them. When expressed as thoroughly as it has been by some, a humean point of view is certainly a respectable and recondite one to hold in the theory of knowledge; there seems nothing obvious that is wrong with it; to the contrary, it may seem foolhardy to try to refute it or even place its merits under scrutiny. In other words, a well thought-out moral scepticism deserves the respect of its critics, and any difficulties with it may be expected to be of a relatively subtle and not self evident kind.
The purpose of this chapter will be then to give as full a description as possible of the historical and political context — of the “language game” or the civilization — within which it is possible for the humean consensus in modern economics to be understood. The economists quoted in Chapter 2 do not appear to have attempted such descriptions themselves, and may even have assumed a humean point of view on the positive and normative to be self-evidently justified, for little thought seems to have been given as to why we should want to endorse it. Thus it will be fair to caution the reader that while a possible justification and explanation of a humean point of view will be given here, it will be one which has been constructed by a critic. Furthermore, the discussion will refer first to a more distant and then a more proximate context, and the discussion of the former will have to be speculative and greatly simplified — a mere thumbnail sketch of an actual drama of indefinite proportions.

§2. The adoption of moral scepticism in twentieth century economics may be most briefly explained as having been motivated by a genuine desire to shield against dogmatism and tyranny, whether in political, economic, scientific, or religious contexts. As scientist and scholar, the economist has been naturally concerned to extend the scope of common reasoning, as well as to protect the objectivity of the findings of his science from the imposition of personal or political dogma. Equally, it has been felt that the choices of the individual agent who is studied by economists, whether as consumer or voter, deserve to be treated with the fullest respect. A humean scepticism may have been adopted because it has been believed to be necessary and possibly sufficient for this kind of respect to be shown to the results of popular choice, whether in parliament, the market place, or in private life. This is summarized in for instance Sugden’s remark “Hume’s Law reflects a liberal view of the universe”, as well as in Schumpeter’s suggestion that the wertfrei controversy had been merely one between those who practised and those who protested a kind of scholarly deceit, namely, the propagation of personal dogma in the guise of a pursuit of knowledge. In other words, someone might become a moral sceptic because he wishes to defend, and wishes perhaps to be seen as defending, the freedom of the individual person to form and hold his or her own normative beliefs, as well as the objectivity of science from being compromised by the forced imposition of the beliefs of any one or a few people. In particular, the modern humean economist is likely to wish to contrast his theory as sharply as possible with the famous theory given by Plato, both directly with the political philosophy which is to be found in Plato’s writings, as well as indirectly, with the medieval scholasticism which came to be deeply influenced by the rediscovered works of Plato and Aristotle and to which the origins of modern economic and political thought can be traced.

Now the question of whether there is any objective knowledge in a field of inquiry is open to be understood either as asking whether there possibly can be any knowledge in the field, or as asking who should be thought of as possessing such knowledge and how they may have been identified. The first of these senses can be thought of as epistemological and the second as political in character. In Republic, Plato offered answers to both questions with respect to the knowledge of the statesman, and the answers he gave were yes — not only is it logically possible for there to be objective knowledge of use to the statesman, but it is practically possible to identify certain men and women in society as actually possessing or being considered fit to possess such knowledge. It is these special people who are the only true lovers of wisdom in society, and since we surely should want the policies of a state in which we lived to be the wisest and most prudent possible, informed by the best available knowledge, it appears to follow at once that what needs to be done is unite knowledge with authority and make these special people our guardians and rulers.

Plato’s ideal city-state is a place where individual freedom is conspicuous by its absence. Its rulers are to be imagined as being about as perfect rulers as there can be: the single and genuine source of all true wisdom and justice, and deserving therefore to be granted absolute authority on all significant questions of private and political conduct, including the right to suppress dissent, since any dissent would be misguided by definition. This is not to say the philosopher-kings would be entitled to a life of luxury or even ordinary comforts. To the contrary, since those who deserve to be philosopher-kings may well be disinclined to seek power and privilege for themselves in the normal course of politics, they may have to be first discovered and then forcibly drafted to take the office which rightfully should be theirs. In preparation for the serious business of piloting the ship of state, they will be placed in seclusion and rigourously educated in such disciplines as aesthetics and gymnastics and mathematics and music, their lives certainly without any of the signs of corruption that we would frequently associate with the exercise of power. At the end of the tenure of one generation of such rulers, they will be retired and replaced by a new generation, bred and educated through a similar and careful programme of eugenics and training in the arts and sciences of statesmanship. Finding actual examples of such extraordinary beings may be quite impossible; perhaps some appropriate mixture of the Dalai Lama, Gandhi, Attaturk and Mozart’s Sarastro might help our modern imagination.
A number of modern political thinkers have roundly condemned Plato for having written a theory hostile to democratic political institutions, and even for having provided the blueprints for the tyrannies of modern history. Yet while there is no question that Plato was no friend of democracy, or at least of the kind of democracy which had brought about the judicial murder of his friend and teacher Socrates, a fair-minded reader of Republic is unlikely to find in it any justification of tyranny at all. If we were to define tyranny in the way Plato and his contemporaries would have done as the rule of the ignorant and capricious, it would be a state of affairs Plato found abhorrent, the complete antithesis of his own ideal of a full union between knowledge and authority, of rule by the genuinely wise and the genuinely good; even the faulted system of democracy would be preferable to it. Moreover, Plato was to discuss at length the dynamics of how even his ideal city-state would be likely to degenerate into a tyranny; and besides, his single attempt to put theory into practice ended in pathetic failure, when he accepted an invitation to train a fatuous prince, who was incapable of and soon became bored with the rigorous education Plato had in mind for him, and who eventually became the worst of tyrants, much to Plato’s disgust. In fact Kant, the modern lover of freedom, was led to come to the defence of Plato, the ancient authoritarian, precisely because the logical possibility of a utopia is suggested to the reader of Republic — a state of affairs in which everyone is a genuine lover of wisdom, everyone a philosopher-king, and therefore all external government made redundant. Republic is a masterpiece of philosophy and mathematics and literature and political economy as well, and it would be a mistake to suppose its author to have been so inexperienced of human nature and society as to provide it as a textbook for grand or petty tyrannies, whether of his own time or of ours.

What is true however what is true is that the theological culture of medieval Europe would come to be deeply influenced by the rediscovered works of Plato and Aristotle, with which a synthesis of medieval Christianity was sought to be made. And it may also be fair to say that regardless of Plato’s intentions, Republic came to provide something of a model for the tyrannies to be experienced in subsequent European history.
Social and economic life in medieval Europe is marked by a four-fold division of society into the nobility, the clergy, free artisans and tradesmen self governed within a system of guilds and corporations, and the peasantry. The medieval church is seen as an eternal institution representing divine will on earth, deserving to be endowed with final and absolute authority on all significant questions of right conduct, somewhat perhaps in the manner of Plato’s philosopher-kings. Specific duties and rights belong to the members of different occupations, and it is within one’s calling that one is expected to lead one’s life in accordance with the divine law as interpreted by the church and the natural law as discovered by the temporal authorities. In particular, there is a notion that economic activities may be licit or illicit in nature, and since the general moral question of what ought to be done is closely identified with whether there is the sanction of the church for it to be done, whether a particular economic activity is to be approved of or not comes to depend on whether or not it has such a sanction. There is an idea too of economic goods having a ‘true’ or ‘intrinsic’ or ‘natural’ value endowed in them by God — an idea which will become perhaps a precursor of the labour theory of value of classical economics in the eighteenth and nineteenth centuries. Determining this intrinsic value establishes the ‘just’ price of a good or service, i.e., the price at which it ought to be traded, even if the actual market price as determined by the subjective estimates and actions of traders happens to contingently differ from this. There is a related concept of ‘equivalence’ in transactions, with a suggestion that one party to a trade can gain from it only at the expense of the other. Merchants and middlemen thus come to be treated with some disdain, since it does not seem apparent they are adding anything to the intrinsic values of goods, making the just price of their services seem hard to determine. Indeed the unabashed pursuit of wealth by anyone is probably the object of some considerable social and religious disapproval. Similar thinking may underlie the condemnation of usury, since, given a premise of money having no intrinsic worth, what is perceived to be the lending out of money should seem to have a just price of nought.

The common medieval culture and economy was to be transformed drastically though differently across Europe between the fourteenth and eighteenth centuries. The sea routes are discovered, nation states emerge competing with one another in trade and war, the age of modern science begins, a long and rapid succession of scientific discoveries and technological inventions takes place, there is a vast expansion of commerce and population and the settlement of European colonies in other continents. Accompanying these transformations in some places are intellectual rebellions against the medieval church, and almost everywhere in Europe a decline in the influence of formal faith. The assertion of individual will and conscience as the principal guides of human conduct is a challenge directed at church doctrine and dogma; but given that the medieval concept of reasoning is one of reason ultimately bounded by the doctrines and dogmas of faith, the assertion of a subjective individual will may have been assumed to amount to being a challenge to the full possibilities of objective reasoning itself.

In this new mercantilist age, the pursuit of material gain must come to be freed of the sanction of the church, and once more, since right and wrong are closely identified with such sanction and prohibition, a declaration of the independence of economic activity from the sanction of the church amounts virtually to a declaration of its independence from ethics as well. In particular, the medieval notion of ‘equivalence’ in the intrinsic value of goods in a transaction is transformed with the aid of mechanistic analogies at hand into a concept of ‘equilibrium’ in trade, such that each party to a trade is conceived of as gaining from it as an individual and continuing to transact until the prospect of such gain has come to be exhausted. It is understandable perhaps that England and Holland will be in the vanguard of the mercantilist revolution, given their theological distance from Rome as well as their growing commercial interests and naval power. Nor does it seem obviously foolish, at least in the early mercantilist years, for the wealth of a nation to be identified with its ability to export and its holdings of precious metals, when the circumstances of the time make it a first priority of the business of government to have liquid payment available for navies and armies. In France there comes to be the liberal protest of the physiocrats against the iniquities upon the peasantry, a protest which serves to rehabilitate a more secular version of the natural law of the scholastics. But the calls of men like Quesnay and Turgot for reform are too late, and the system of physiocracy is itself swept away with the onset of the French Revolution.

Adam Smith however has admired and learned from the physiocrats, while observing at first hand the dismal effects of a staling British mercantilism. This he rises to condemn in The Wealth of Nations, thereby starting an intellectual revolution of his own, ringing in a new century of free enterprise and imperial expansion, and establishing the concern of the economist with the workings of individual interest and the market economy which continues to this day. Forty years later it is David Ricardo who introduces to political economy the practice of an abstract hypothetical method, by which it is a body of abstract and general principles that the economist’s speculations and ratiocinations are intended to discover, detached from the rush of concrete economic realities. And Ricardo and his immediate followers exemplify the application of the new method to a main subject of Smith’s preoccupation, namely, the workings of individual self interest and the market economy.

In the musty passage-ways of Victorian thought, the new methods of abstraction in political economy must have been felt to be as invigorating as fresh air. Jevons, Walras, Menger and the other original neoclassicals firmly insist upon making the plain and simple observation that in the case of many and perhaps most goods, the prime determinant of relative value is not how much labour went into the different production processes, nor how much intrinsic value God might have placed in the goods, but rather the subjective estimations of economic agents in the market place. The victory seems complete. Out of the medieval notion of the scope of reasoning being limited by the dictates of doctrine and dogma, is eventually born the neoclassicals’ notion of the concept of value as fully and exactly synonymous with the concept of scarcity or market value, or rareté in Walras’s term. Economists are seemingly freed to speak of ‘a theory of value’ when meaning to refer more specifically to a theory of scarcity-determined relative prices, determined by conditions of supply and demand in the marketplace. From an idea that something is or is not a good only and merely because the church happens to say so, the wheel comes full circle to an idea that something is or is not a good only and merely because of the price it happens to command in the marketplace. The moral absolutism of the platonist and the scholastic gives way to the moral scepticism of the humean, and we reach the threshold of the modern period of economics in the later nineteenth and early twentieth centuries.

§3. Briefly then, the development of the kind of sceptical and subjectivist point of view represented by Hume and the humean economists may be seen as the democratic reaction which occurs to medieval and platonist authoritarianism. And in parallel with these democratic developments occurring in the marketplace and economic thought, there occurs between the medieval and the modern period an emancipation of the political mind as well. No more will it be for clergy and aristocracy to dictate divine and temporal laws respectively. Men are born equal — which is to say there are not grounds ex ante why one human being should be supposed to deserve more or less authority or dignity than another merely in virtue of his or her humanity. The political process must reflect this new emancipation, and displace the hierarchies of the past with the equalitarian notion that every man’s vote should count the same, and the most popular choice be established to rule.

The modern institutional context of a parliamentary democracy, bound by formal or informal constitutional principles and precedents, may be roughly sketched somewhat as follows. From among the body of citizens, some will choose to run for elected office. While reasonable restrictions may be placed on who can so choose (e.g., they must be adult nationals) any citizen normally will be free to be a candidate. Before a vote is conducted, a reasonable time will be allowed for candidates to put their respective cases to the public. There will be some constitutional rule, like first-past-the-post or proportional representation, agreed upon more or less unanimously in advance of the vote, which will map how the actual balloting will induce particular outcomes as to the composition of the parliament. The individual voter casts his or her ballot, reflecting some private mixture of interest, prejudice, caprice or good sense about the common welfare. The rule is applied, and the largest coalition of winning candidates come to constitute the new government, with smaller coalitions constituting the loyal opposition. Once elected, a government will be expected prima facie to carry out the agenda it had proposed to the public before the election and not something different. What it actually does will be the subject of constant scrutiny and criticism by the opposition, the press, and the public at large, but the laws finally enacted will have jurisdiction over all. After a certain maximum time, elections must be held again and the process repeated, with an incoming government either maintaining or changing the policies of its predecessor in large or small measure. The system may be considered indirectly democratic insofar as that at any given time citizens shall have given themselves, via their elected representatives, the policies and laws under which they are themselves to live.

While a government would be expected to implement the agenda chosen indirectly in this way by the public, it will be expected also to elicit expert advice upon the best means to be employed towards achieving the chosen ends. Yet the expert must be appropriately humbled, brought down from the high altar where Plato had placed him to being the modest and self-effacing servant of the popular will. The scientist in government is to take as given the ends of his political masters, under a presumption that these reflect the democratic choice and any interference or criticism would be impertinent. More generally, the competence of the expert in a democratic society is not to extend to questioning the uses to which his expertise may be put. Thus Popper was to write: “No amount of physics will tell a scientist that it is the right thing for him to construct a plough, or an aeroplane, or an atomic bomb. Ends must be adopted by him, or given to him; and what he does qua scientist is only to construct means by which these ends can be realised.” Or as Myrdal put it in the passage quoted in the previous chapter, the expert must not go beyond advising on the means, for he would otherwise require premises of a normative kind which have not been given to science, but which are to be presumed available instead to the elected politican. And Robbins wrote of how economists ought not to judge the ends to which economics is put, indeed that ultimately “there is no room for argument” about ends, but rather how the quintessence of economics is the study of the optimal allocation of scarce resources between competing ends. It is only the question of the best or optimal means towards such an allocation that is within the scope of rational inquiry, and therefore within the competence of the economist qua scientist; it is not for the economist to question the ends given to him by the representatives of the public.

Now the widespread view since that there is a unique and quintessential economic problem, and that in particular it is the problem of the allocation of scarce resources between competing ends, is of course one initially advanced in the course of the neoclassical revolution. As Marshall put it: “if a person has a thing which he can put to several uses, he will distribute it among these uses in such a way that it has the same marginal utility in all. For if it had a greater marginal utility in one use than another, he would gain by taking some of it from the second use, and applying it to the first.” The housewife must decide how much yarn should be put to making socks and how much to making vests so “as to contribute as much as possible to family well-being”; she will have allocated the yarn efficiently if the marginal increase in family well-being is the same whether she puts the last ball of yarn to making an extra pair of socks or to making an extra vest. In modern terms, the problem is one of constrained maximization in which a concave objective function is to be maximized subject to a number of linear or non-linear constraints. We might imagine, for example, a hospital administrator who must allocate fixed quantities of various resources at his disposal like medical staff, beds, dressings, and so on, between a number of alternative outputs which have to be produced in different hospital wards, with the aim of maximizing an objective function containing these outputs as concave arguments. The objective function itself, that is, the relative weights which should be given to the various outputs, is not ultimately for the administrator to decide, but rather to be taken by him as a parameter from an appropriate authority. If the necessary conditions for a maximum are met, an optimal allocation would be one in which (a) the ratio of marginal increases in the objective function from marginal increases in the output of any two goods equalled the implicit shadow prices of their technologies; and (b) the marginal increase in the objective function from increased use of a resource in any two production activities would be the same and equalled the shadow price of the particular resource. Thus the marginal hour of a nurse’s skills would be equally well applied whether in assisting mothers in labour or in providing aid in the Emergency Room. Similarly, a humean view of the expertise of economists would be one in which the economist did not question the social objective function but rather takes as his task the statement and solution of the formal problem of the allocation of scarce resources between the defined ends.

With the necessary change of detail, the same has been required in the influential theory of macroeconomic policy advanced by Professor Jan Tinbergen and his principal expounder, Professor Henri Theil. In this theory, normative premises are seen as being given to the expert economist by a representative of the political process, for instance “the Minister of Finance or Economic Affairs, who is interested in the employment level of his country and its balance of payments”. Such a person is assumed to know the set of variables relevant to determining the present state of the economy, which are divided into those whose values can be changed (“instruments”) and those whose values cannot be changed (“targets”), with a change in the value of an instrument being defined as a “policy measure”. The expert economist is called upon to specify as best as possible the structural relations between targets, instruments, and exogenous disturbances, and predict as best as possible the future course of the targets under alternative assumptions about the instruments. As Theil put it, the policy-maker is to receive from his forecasters “conditional expectations about the time-patterns of non-controlled variables, the conditions being alternative measures to be taken by himself in the present and the future.” Alternative futures of the economic model are then to be evaluated one against the other by means of a social utility function decided upon by the policy-maker. Its arguments could be a pair of macroeconomic ills such as inflation and unemployment implying the function should be minimized, or a pair of microeconomic goods like efficiency and equity implying the function should be maximized subject to the relevant constraints, with the relative weights given to the ends presumed to be reflecting the democratic mandate. An optimal vector of targets is determined which yields the least possible social disutility or the highest possible social utility; the values of the instruments which would result in this optimal vector are calculated, and changes from the present values of these instruments to these optimal values define the optimal set of policy measures to be taken.
Such briefly was the kind of theory of economic policy Tinbergen put forward in the early years after the Second World War. It was soon to have much influence among macroeconomists, especially in the United States. Fairly or not to both Keynes and Tinbergen, the models themselves came to be called “Keynesian”, yet their influence has been significant enough that contemporary critics of Keynes and Tinbergen have described their method and purpose in similar terms. For keynesians and their critics, the macroeconomist principally has a positive role, extending the scope of reasoning and discussion on logical and empirical grounds as far as he is able to. He assumes a constitutional democracy, and takes for granted that the normative premises of the policy-maker reflect the popular will.

§4. Drawing together, then, the main threads of this highly simplified and summary discussion, it may be possible to explain the adoption by twentieth century economists of a humean theory of knowledge by the widespread belief that such a theory provides a necessary and even a sufficient defence against dogmatism and tyranny. It is part of the democratic reaction to medieval authoritarianism. The modern civilization which has adopted the moral scepticism of Hume is one born out of the great medieval civilizations which had been influenced by the authoritarianism of Plato. And just as Plato’s theory was affected by his disgust with the doings of the democracy of his time, so it may be the theory of knowledge which has come to be adopted by as eminent and diverse economists as Robbins and Friedman and Samuelson and Hicks and Robinson and Myrdal and Arrow and Hayek and Lange and Tinbergen and Hahn and Schumpeter, and the many others who have followed them, has been conditioned in part by their disgust with the tyrannies and ideologies of twentieth century history, and their desire to protect from these both the objectivity of economic science as well as the individual in his capacity of consumer and voter.

The question arises however, whether, in making their escape from Plato, the pioneers of twentieth century economic thought have not become entranced by Hume.

4. Difficulties with Moral Scepticism

We have now a description of some of the main features of the theory of economic knowledge most widely accepted in the twentieth century, and we have seen also how its plausibility and influence may be explained by placing it in appropriate historical and political context. In this chapter we shall examine some of the main difficulties and paradoxes which happen to arise with this theory. These have been serious in their implications, and the more general problems from which they derive have been well known to many contemporary philosophers, yet they do not appear to have been given adequate notice by modern economists.
Briefly, the difficulties are two-fold.
First, if the justification of adopting a humean theory of knowledge by contemporary economists is to be what we have taken it to be, viz., that such a theory and only such a theory can provide an adequate bulwark for science and the individual against tyranny and dogmatism, then we clearly have the makings of an internal contradiction on our hands — since what is patently a moral purpose would have been advanced within a theory of knowledge whose ostensible aim was to deny the possibility of moral knowledge! In a theory in which all moral propositions are taken ultimately to be statements of mere personal opinion, the defence of the freedom of the individual or of the integrity of science must also be taken ultimately to be matters of mere personal opinion, and the declared or undeclared purpose of protecting freedom by adopting moral scepticism would have been internally defeated by that very scepticism itself.
Secondly, we shall find that sceptical attacks just as powerful as Hume’s attack on the possibility of moral knowledge can be made upon the possibility of knowledge in a number of non-moral contexts as well. Hume himself is responsible for one such attack when he raised his famous doubts about the possibility of induction, and analogous attacks can be made in diverse other contexts such as those of science, history, mathematics, or psychology. The result of recognizing these new possibilities for scepticism is to make evident that an acceptance of moral scepticism on its own may force a choice between either sliding into total scepticism, the position of believing there is ultimately nothing whatsoever that can be objectively known, or forsaking parity of reasoning, and denying that what may be sauce for the goose is also sauce for the gander. Either the possibilities of mathematical knowledge and scientific knowledge and historical knowledge all come to be denied ultimately because we wish in a consistent way to deny the possibility of moral knowledge, or one sort of knowledge is accepted and another sort rejected when there are reasons to think they must stand or fall together. Either all of positive economics is attacked with just as much scepticism as anything in normative economics, or we accept one and reject the other when instead there are reasons to think they share the same ultimate grounds and must be accepted or rejected together.
Such will be the main hazards we shall find on the humean course taken in the theory of knowledge by the economists quoted in Chapter 2. Their precise locations however are subtle and quite well hidden, so if we are to avoid them we must move here as carefully and precisely as possible.

§2. Let us recall at the outset Hume’s First Law as saying to the effect that a normative conclusion cannot be validly deduced from solely positive premises; that a normative conclusion cannot be deduced without at least one normative premise being made. Faced with a normative proposition then, a moral sceptic will ask to see the set of prior positive and normative premises from which it is to derive. To take a simple example, if you were to say “I think the government should reduce the rate of growth of the money supplym from 6% to 3%”, a moral sceptic may ask “Could you say why you think so, since your proposition is plainly normative and cannot have derived from a set of solely positive premises?” (We can suppose this not to be meant rhetorically, that some opinion like “What a stupid idea!” is not being surreptitiously introduced in the guise of asking a question, but rather that a genuine inquiry is being made to be told the grounds that may go to support the proposal.) If you were to reply “Well the government should try to reduce the rate of inflationp , it is necessary and/or sufficient to reducem in order to reducep , that is why I think the government should reducem ,” it would remain open for the sceptic to respond “Certainly I can agree if your premises are true then your conclusion follows. But your premises once more are not solely positive ones, including as they do one that is plainly normative. Could you now say why you think the government should try to reducep in the first place?”
It is not difficult to imagine a fair reply being given to this as well, such as perhaps “Well inflation has been rampant and the election was fought and won on a promise inflation would be curbed, election promises should be attempted to be kept, that is why the government should make a determined attempt to reducep .” But in practice the economist would typically and rightly allow such discussion to fade into the background — since an important and difficult task would already have been defined for him, which is to ask whether it is likely a reduction inm by the stated amount will succeed in reducingp , assuming that the government should be trying to do this in the first place. Trying to answer it will require abiding by the practices of language and logic and scientific method; but the question itself is a positive and not a normative one insofar as it asks what is the case, or what has been the case or is likely to be the case, and the desire to keep it distinct for analytical convenience from the explicitly normative may be understandable. The modern economist is one of many kinds of expert in civil society, and as such is expected to have some special theoretical or practical knowledge not possessed by the non-economist. And economists everywhere are in fact being called upon to evaluate whether or not a dam or a highway should be built, a budget balanced or unbalanced, a bond released or redeemed, a tax or a tariff levied or lifted; to judge whether the argument of a government or a colleague or a student or a critic is valid, substantiated, compelling, sound, cogent. In any such investigation, it may well be useful for purposes of clarity and analytical convenience to work with a dualism between the ‘is’ and the ‘ought’, the descriptive and the prescriptive — just as it is commonly useful to work with a dualism between an analytical sense of ‘is’ as in “two plus two is four”, and a descriptive sense of ‘is’ as in “the cat is on the mat”.
Yet from saying it may be useful to make working dualisms between what is possible and what is actual or between what is the case and what ought to be done, it does not follow there are any absolute or ineradicable lines to be drawn. Taking a set of normative premises as given and from there proceeding to extend the scope of positive reasoning would not imply the normative premises are unquestionable — only that they are not now in question, not presently in question. It is as if they have been temporarily taken out of the game while we attempted to see how far we may proceed without them. They can still be brought back and others taken out — indeed, in the game of inquiry, we might even wonder if there needs to be any proposition which must be so privileged as never to be benched, so indispensable that we must fear the whole project will collapse without it.

§3. We may recall next Hume’s Second Law to the effect that while it may be possible to bring to bear objective reasoning in some normative discussions, a point of sheer and unadulterated difference over ‘basic’ or ‘ultimate’ values can nevertheless come to be reached. The moderate humean may allow for much room for common reasoning to take place, but he takes the further step of supposing such reasoning to have a limit, a finite limit. In any normative discussion, it is eventually possible for the scope of objective reasoning to become exhausted and a difference of a sheer normative kind to come to be identified. While it is clear the economists quoted in Chapter 2 have meant to refer to a limit of this sort being reached, it is strictly speaking not clear if they have meant to refer to such a limit being reached just as a contingent matter of fact — in actual arguments and discussions — or whether they have meant to refer to such a limit being possible in principle as well. In other words, whether it is merely intended to be an empirical possibility that a disagreement will come to end without resolution, or whether it is also intended for this to be the logically necessary outcome. If a residue of disagreement remains after the processes of common reasoning have been allowed to work, is this residue to consist of differences which just happen to be closed to further discussion in a particular case, say because the discussants lack patience or good humour or tolerance or perseverance or whatever, or is it supposed to consist of sheer and naked differences over ‘basic’ values which must be thought of as necessarily beyond the scope of further discussion?
If it is the first interpretation alone which has been intended, then only a fairly small claim would have been made, which may need to be clarified and fully set out but which would not need to be disputed by someone wishing to attribute a greater scope to reason than does the moral sceptic. For it is quite evident that actual arguments and discussions frequently do come to end without full resolution — those between physicists, mathematicians, biologists, doctors or engineers no less perhaps than those between politicans, economists, writers, historians, spouses, or nation states. Yet an observation of this sort of the frequency or intensity of disagreement would not be directly relevant to the theory of knowledge, insofar as the fact an argument happens to stop where it does, does not bear upon whether a question in dispute is capable of having a true or a right answer. It is possible for the true or right answer to a question not to be available to those who happen to be discussing it, or even to others in their generation or those in later generations; that there can be an objectively true or right answer to a question is a different question from whether it has been found or will be found today or tomorrow or next year. What the answers happen to be to the questions raised by Darwin or Freud or Keynes is a different question from what they themselves might have thought the answers to be, or what their contemporary state of opinion happened to think the answers to be, or what the state of opinion in our own time or in some future time happens to think the answers to be. It is of course natural to want to know the true or right answer to a question, to know whether the answer which we think is true or right is true or right, and certainly we should be surprised and find it incongruent if someone said he or she believed something even while knowing it was not true, or approved of something even while knowing it was not right — we normally want to know what is true and what is right and make our beliefs congruent with it. In other words, we may distinguish the actual and contingent history of inquiry and conflict from the logic of inquiry and conflict.
Moreover, some concepts and propositions will be found to form a context or a background in any disagreement, being understood by both sides and being unnecessary to be made explicit. If we were discussing the monetary history of the United States in the 1980s for example, we would take for granted such facts as that the United States was not at war or civil war or in the throes of any major social convulsion during this time; assumptions which may not have formed the implicit background if we were instead discussing the monetary history of the 1960s or the 1860s. Not every feature of a description may be relevant to a particular question at hand nor must it be made explicit. And an observation of this kind may be made of any dispute in economics, once it has been carefully and thoroughly characterized, whether on method or theory or evidence or policy, in microeconomics or macroeconomics, whether between mathematical economist and applied economist, or keynesian and quantity theorist, or marxian and mainstream. Some aspects of any description will be implicitly understood or taken for granted by the participants in a discussion.
More strictly, it has been argued by the Cambridge philosopher Renford Bambrough that it is necessary for the participants in a discussion to be in at least some agreement before they can be even said to be in any disagreement at all: “You and I cannot be known to be in conflict unless it is possible to identify a proposition that I assert with a proposition that you deny; no such proposition can be identified unless there is some expression that you and I use in the same way; if we use an expression in the same way then we regard the same steps as relevant to determining the truth or falsehood of what is expressed by it; for a disagreement about what is relevant is or involves a disagreement about what the dispute is that we are engaged in, and when such a case of cross-purposes is resolved it resolves itself either into agreement or into a disagreement to which all these conditions again apply.” In other words, it must be either that the participants in a dispute are giving different answers to the same question or that they are giving answers to different questions. If the first, we have identified a genuine case of disagreement; if the second, we have what is strictly speaking not a genuine disagreement at all but a case of cross-purposes, where each is giving a different answer to the question as to what the question they are disagreeing over happens to be. The English literary critic F. R. Leavis suggested at one place that critical inquiry proceeds as if one person declares to another “This is so, isn’t it?”, and the other replies “Yes, but…”. When A declares “This is so, isn’t it?” he has invited both the challenge and collaboration of others. B’s yes in reply would indicate a certain agreement, while his “but…” would indicate the agreement was not total, that there perhaps is some case or circumstance to which what A has said will be found not to apply. In effect, the “but…” amounts to being a fresh “This is so, isn’t it?”, inviting in turn the collaboration and challenge of A, and so on. Applying such a scheme to our example of a simple debate over economic policy, we would obtain an abstract form of the following sort:
A : n1.
B : Why n1?
A : Given n2, p1 implies n1.
B : Granted (p1), but why n2?
A : Given n3, p2 implies n2.
B : Granted (p1, p2), but why n3?
A : Given n4, p3 implies n3.
B : Granted (p1, p2, p3), but why n4?
A can think B to be stupid or stubborn or self-seeking, and B can think the same of A, and neither or one or both of them may be partly or wholly correct in thinking so, and all these may be facts which go to explaining how their dispute actually happens to proceed or fail to proceed over time — yet the correct answer, the most reasonable and justifiable answer, to the question to which different answers may be given at any stage will be independent of all this. We should want to distinguish, in short, questions of the logic of thought from questions in the history of thought.
Thus if someone becomes persuaded to a moderate moral scepticism only through observing that as a matter of fact many normative disputations seem heated or interminable, then we need only to demonstrate that such an observation does not and should not be allowed to bear upon the theory of knowledge or epistemology we come to hold. Certainly the scope of objective reasoning may be found to be finite in practice in actual disagreements and disputations between people, because there happens to be a lack of patience or good humour or tolerance or perseverance or whatever. But from that it does not follow at all that there is no further room for discussion, or indeed that reasoning cannot be thought of as being of potentially indefinite scope.
If however, as seems equally likely, the economists who have endorsed a humean theory of knowledge have meant it to be possible not only in practice but also in principle for the scope of objective reasoning to become exhausted, then a much more serious claim would have been made, which deserves appropriately more rigorous scrutiny. It would then have been claimed that it is logically possible for A and B to be in total and justifiable agreement about all the empirical evidence and about every logical relation, and still for each to declare in favour of a sheer and contradictory ‘ultimate’ value.
B : Granted (p1, p2, p3,…, pω-2); but why nω-1?
A : Given nω, pω-1 implies nω-1.
B : Granted (p1, p2, p3,…, pω-2, pω-1); but why nω ?
A : nω that’s why! (Go jump in the lake if you don’t accept it too.)
B : I deny nω that’s all! (And it’s you who can jump in the lake.)
Not only in practice but also in principle the scope of common reasoning would be supposed to have a finite limit. Not only is it a handicap we have to live with that many disputes between economists or scientists or citizens or spouses or nation-states do come to halt without full and justifiable resolution, through lack of patience or tolerance or good humour or whatever, but it is inevitable that common reasoning will become exhausted and only sheer and unadulterated differences remain over ‘basic’ or ‘ultimate’ values over which only the irrational holds sway. Hume and Hare among philosophers certainly may be interpreted to have taken such a view, and, on the basis of the writings quoted in Chapter 2, it would not be unfair to interpret at least some of the economists to have meant the same. However no proof or example of the existence of a sheer dispute over ‘basic’ or ‘ultimate’ values between people who are in justifiable agreement over everything else, has ever been offered by Hume or any philosopher or economist after him. It seems merely to have been asserted or taken for granted that a point can come where the scope of reason must have become exhausted and nothing further could remain to be said or done.

§4. We are in position to have a clear sighting at last of the first major hazard which is present on the humean course: It is possible that the declared purpose of the humean economist of extending objectivity and thwarting dogmatism will be contradicted by an ultimate adoption of irrationality and personal dogmatism. Huge and invaluable edifices of inquiry and argument can crumble to the ground because the scope of reasoning must sooner or later become exhausted, and mere personal prejudice take its place. The presence of a single ‘ought’ would signal the presence of another, and then another, and another… until some set of private moral primes or absolutes or supreme principles are supposed to be reached, which others might or might not share but which are in any event beyond further question. According to the received theory of knowledge, the economist is ultimately able only to persuade or coax or cajole or perhaps bribe others into accepting the absolutes he may himself wish to endorse, but common reasoning is of no further avail. Sooner or later the advice of the expert economist cannot but express the personal dogmas and prejudices of the adviser (or those of his employer).
It was a tension of this kind in the humean doctrine that Professor Samuelson may have felt when he called it a “somewhat schizophrenic rule” even as he endorsed it in the passage quoted in Chapter 2. Yet while Samuelson was not afraid to describe the role of the economist in society that follows from the humean thesis, he did not see the paradox to which it leads. Following Robbins and in keeping with the modern theory of economic policy, Samuelson said we should keep distinct the economist qua scientist from the economist qua citizen. The former expresses objective knowledge (“pure analysis”), the latter expresses subjective opinions (“propaganda, condemnations and policy recommendations”). Thus when Professor Samuelson himself writes from his offices at the Massachusetts Institute of Technology, we must take him to be doing so qua rational, objective, scientific economist, while if the very same person writes from his home qua citizen of the United States, we must take him to be expressing a subjective and possibly irrational personal point of view. Or must Samuelson expect himself to sign and stamp everything he writes either as being a claim to objective knowledge made by the eminent economist which he is and deserving the world’s attention, or as being a subjective and possibly irrational opinion expressed by the ordinary citizen and human being which he also is, and perhaps not deserving nearly as much of the world’s attention? What would happen if the same human being came to say the same thing in both scientific and civic capacities? Clearly we would be in a quandary of having to decide whether it should be considered objective or subjective, public knowledge or private opinion, rational or irrational, economic science or personal prejudice. In the previous chapter we have seen that the humean economist is likely to want to sharply contrast his theory of the role of economic expertise from the famous theory given by Plato in Republic. Now we are able to see that there seems to be a less well known similarity too between the moral scepticism of the humean and the moral absolutism of the platonist. For just as in Plato’s theory so in the modern humean theory, there is evidently no way of telling from within the theory who is supposed to be the expert. Either the humean has to join the platonist whom he takes to be his enemy and declare there to be some arbitrary and unspecified way of distinguishing expert from layman, philosopher from commoner. Or the humean has to part company with Plato and the scholastics, and say that there is ultimately no objective distinction possible between knowledge and opinion, expert and layman, science and prejudice. What appears to be at stake when the merits of the humean epistemology are brought under critical scrutiny in this way, therefore, is nothing less than whether there ultimately can be objective knowledge in economics; and so, whether or not the economist can rightly consider himself to be a seeker after such knowledge — or whether we are all involved merely in some highly evolved and sophisticated branch of rhetoric, having “the semblance of wisdom without the reality” whose teacher and practitioner is just “one who makes money from an apparent but unreal wisdom.”

§5. The problem we are observing here with the received theory of economic knowledge can be placed in relief by comparing the moderate moral sceptic with his more radical cousin, the emotivist. For the emotivist is one who flatly denies there to be any scope at all for common reasoning to occur upon normative questions, maintaining instead that normative propositions amount only to being the expressions of personal feeling or emotive attitude. Thus a statement like “the government should reducem from 6% to 3%” would be taken by the emotivist to express merely the personal feelings or preferences of the individual, its full meaning and implications being equally well described if the speaker had said “I wish the government would reducem from 6% to 3%”, just as someone might say “I wish to have my coffee black” or “I do not like boiled vegetables” or “I like to wear colourful shirts”.
Now the feelings and emotions and attitudes of a speaker or author may be naturally and normally involved in the making of evaluative or prescriptive statements, in a way they may not be in the making of logical or empirical statements. When I propose something should be done I must mean what I say, or I would not be being sincere, what I outwardly expressed would be incongruent with what I inwardly felt, I would be engaged in a kind of self-contradiction or inner dissonance. Yet this sort of involvement of matters of personal sincerity and authenticity in the making of normative judgements does not imply these are all that is involved, or even the most important of what is involved, or that common reasoning cannot make headway in normative discussion. The emotivist correctly observes the involvement of the emotions in normative discussion but exaggerates its significance, perhaps by the confounding of simple and literal uses of concepts like “taste” and “preference” as in “I have a taste for ice-cream” or “I prefer my vegetables lightly cooked” with looser and more metaphorical and so more complex uses of the same concepts like “I prefer Truman to Dewey” or “I have no taste for public executions”. Where the moderate moral sceptic supposes a residue of irrational difference to remain after every relevant empirical and logical question has been answered, the emotivist wants to call a halt the instant a normative proposition is sighted. The difference is one of degree and not of kind. If a moderate moral sceptic like R. M. Hare or Milton Friedman or Joan Robinson remonstrated with the emotivist saying “Look you really should try to bring to bear as much logic and evidence as you possibly can in a normative dispute”, the emotivist has only to coolly reply “Sorry, but what you have just said is patently normative. Since, as you know, I take all normative propositions to amount to being expressions of personal taste or emotive attitude, I cannot take what you have said to be anything more than that either. That does not mean I cannot share the same emotive attitude as you, but that is no reason to think we can construct an objective justification for it.” The humean can bang his head in frustration at the emotivist’s behaviour, but he may not without circularity argue against it.
A more dramatic illustration of this sort of difficulty with the humean doctrine may be found in the writings of Hare and Popper, suggesting that even the most tough-minded and critical of moral sceptics may have allowed themselves to admit an ultimate irrationalism. Hare considers a fanatic who so fervently believes some group of innocent people should be put to death that he is prepared to be made such a victim himself if his own ancestors transpired to be of the same group. And the fanatic is closed to all further discussion of the matter. This, Hare takes it, would be a case of an ultimate value judgement, impervious both in practice and in principle to further question. Hare says that “fortunately” there are few fanatics who would be found to hold such an “extreme” position, leaving unsaid that if they were found then they should be just as entitled to their opinion as anyone else — not merely in the sense of having a legal right to hold such an opinion but in the more significant sense that such an opinion ultimately must be considered to be just as good, just as reasonable, just as cogent, just as sound, as its contrary. We could try to persuade or cajole or bribe our fanatic to give up his opinion and to hold ours, but there is no way for us to say he is simply wrong in his belief. If it turned out there were more fanatics than there were of us, it could of course become their turn to persuade or cajole or bribe us away from our opinions, yet none of their acts could be condemned, since, in the last analysis, there cannot be any such thing as moral knowledge.
Popper has written frankly that he knows of no rational grounds for recommending a rational temperament: “It is impossible to determine ends scientifically. There is no scientific way of choosing between two ends. Some people, for example, love and venerate violence. For them a life without violence would be shallow and trivial. Many others, of whom I am one, hate violence. This is a quarrel about ends. It cannot be decided by science…. you cannot, by means of argument, convert those who suspect all argument, and who prefer violent decisions to rational decisions. You cannot prove to them that they are wrong….” “I frankly confess that I choose rationalism because I hate violence, and I do not deceive myself into believing that this hatred has any rational grounds. Or to put it another way, my rationalism is not self-contained, but rests on an irrational faith in the attitude of reasonableness. I do not see that we can go beyond this.” But if Popper is entitled to have an irrational faith in being reasonable, then the fanatic is surely entitled as well to have an irrational faith in being unreasonable. Thus Professor Max Black responds on behalf of the fanatic who engages Popper thus: “Bravo! You hate violence, but I hate argument (a sneaking use of force by other means). You call me irrational, but I glory in that title. Like you, I hold that there are no ultimate reasons for my irrationality (for that would detract from the purity of my position). The difference between us is like that between a Protestant and a Catholic: your faith is my heresy; my faith is your heresy. That’s all there is to say.” (Yet Black himself does not say why differences between protestant and catholic must be supposed beyond discussion!)

§6. This kind of internal contradiction we are observing here to be associated with moral scepticism can be seen in a slightly more positive light as well. For we may ask, what does the moral sceptic’s recognition that dogma and tyranny should not be imposed upon science or the individual amount to being except a manifest example of a moral recognition? Or a proposal that the integrity of science as well as the freedom of the individual as consumer and voter should be preserved, except a manifest example of a moral proposal? All the economists quoted in Chapter 2 have recommended and practised the extension of the scope of common reasoning in economic science; what sort of recommendation would that be except a patently moral recommendation? When the theory of economic policy requires the economist to respect the ends of the elected politician, what sort of a premise does that rest upon except a moral premise that the institutions of constitutional democracy should be respected and not abused? It would presuppose in turn such things as that parliamentary elections do take place periodically and are in fact genuine and not fraudulent elections, that citizens will be judicious and well enough informed in their voting so that a good indication of what things are conducive to the common welfare will come to be determined as closely as possible given the size and diversity of the electorate, that the policies of a resulting administration are sincere attempts to reflect the ends chosen by the voters, that candidates for elected office and private citizens and scientists and scholars and others are not subject to being shot or jailed or persecuted for saying publicly what they think these ends should or should not be, and so on. It is implicitly or explicitly within the context of a free and open society, and one which probably has working democratic institutions, that the modern theory of economic policy makes sense at all, that positive questions like “Does the evidence support the hypothesis that reducingm from 6% to 3% is necessary and/or sufficient to reducep ?” are supposed to be discussed in the first place. Regardless of what the humean economist happens to say or suppose himself to be doing or not doing by adopting the theory of knowledge which he does, we are entitled to conclude that he is in fact far from asserting there cannot be any such thing as objective moral knowledge — since he himself may have advanced his moral scepticism precisely upon substantive moral grounds. Put differently, it does not seem possible without contradiction to start with a set of moral premises and arrive at a conclusion that there cannot be moral knowledge.
Equally, if the received theory of economic policy must presuppose a context of a free and open society and working democratic institutions, then it would seem it must be silent where such a context cannot be presumed. When we consider that most societies most of the time probably have not been very open or very democratic (and in such a count we must consider societies not only on the scale of nation-states but also families and clubs and corporations and university departments and armies and religions, and so on) this would at once make the received theory one of quite special and contingent application. Indeed it is a theory which must be silent about the appropriate role of the expert not only under conditions of tyranny (Solzhenitsyn: “The prison doctor was the interrogator’s and executioner’s right-hand man. The beaten prisoner would come to on the floor only to hear the doctor’s voice: ‘You can continue, the pulse is normal’” ); but also where the duly elected government of an open and democratic society proceeded to do things patently wrong or tyrannical (the imprisonment of the Japanese Americans). Hence Popper’s “paradox of democracy” and “tyranny of the majority”. It is ironic that the economist who may have adopted a humean epistemology as a reaction to dogmatism and tyranny in the first place, will come to be prevented by his own moral scepticism from condemning an act of tyranny whether it is committed in the name of the popular will or by an outright despotism. A theory of economic policy which both assumes a free and open society and bases itself upon a moral scepticism cannot have anything to say ultimately about the objective reasons why a free and open society may be preferred to an unfree or closed society, or about the good or bad outcomes that may be produced by the working of democratic processes.
A parallel difficulty arises for the humean economist with respect to market institutions and their possible outcomes. Ultimately, the received theory of economic knowledge cannot allow that there may be objective reasons why market institutions may be preferable (or not preferable) to non-market ones, whether one is speaking roughly and generally in a theory of political economy or more precisely and specifically about some actual set of concrete circumstances. Just as the medieval scholastics might have said that a good was a good only because the church said it was a good, so the modern humeans may have to say that a good is a good only because market forces have made it a good — i.e., because it happens to have a positive price in an equilibrium of supply and demand. And just as the church may have said a lot of things were goods which were indeed good, so market forces make a lot of things goods which indeed are good — for instance, like food, clothing and shelter, because they are conducive to some valuable human purpose. But also, just as there could have been things which the church said were good but were not, and things which were good but which the church said were not, so it is not at all hard for any of us to find in experience things which the market may have put a high value on but which were not in fact valuable, as well as things which the market did not value but which were indeed valuable.

§7. Drawing these simple threads together then, a first set of reasons why the modern economist may think himself poorly served by a subjectivist theory of knowledge has to do with the fact that it is a theory which falters and fails even in its own declared purpose of being an adequate shield against dogmatism and tyranny. In a theory in which nothing, ultimately, can be considered objectively right, it cannot be objectively right to extend the scope of reasoning in economics, or to preserve the integrity of science, or to protect the individual from dogmatism or tyranny. In a theory in which nothing, ultimately, can be considered objectively wrong, it cannot be wrong to block or subvert reason or to force dogma and tyranny upon science or the individual. If all moral propositions are ultimately taken to be matters of mere personal opinion, then the defence of individual freedom or the integrity of science also must be taken ultimately to be matters of mere personal opinion. Professor Arrow remarks: “The only rational defense of a liberal position… is that it is itself a value judgment.” Combine with this the idea that judgements are subjective, and you would have the result that no objective justification can be given ultimately for a liberal position, or for any other position either for that matter. When all has been said and done, protecting individual freedom is no better or worse than attacking it, preserving the integrity of science is no better or worse than destroying it. “Nothing is good or bad, but thinking makes it so.” Such fragile things as the preservation of human freedom and the integrity of science would seem to have been left exposed by the accepted epistemology in twentieth century economics to the shifting whims of popular opinion. The purposes that many eminent economists may have had in adopting the humean thesis, and these may have been invaluable purposes, would seem to be able to be fulfilled only in a theory which denied the humean thesis that nothing can be right or wrong but thinking makes it so.

§8. We have now sketched the first important set of dangers that are present on the humean course which has been adopted by modern economists. There happens also to be a second set with equally serious implications, calling for us to continue to move as carefully and precisely as possible. The reader who may have been unconvinced by the argument so far will therefore have a fresh set of challenges to consider, while the author will have to ask for the patience of the reader who may have agreed that there does happen to be something wrong at the foundations of the received theory of economic knowledge.
In short, there is the problem that an adoption of moral scepticism on its own may lead by parity of reasoning to total scepticism, to the ‘pyrrhonism’ which Hume himself had drawn back from. For what will come to be noticed by the truly serious and tough-minded sceptic is that the general logic employed in Hume’s First Law is in fact extremely powerful, more powerful than Hume or the modern humean economist may wish or intend it to be. For the tough-minded sceptic will look at Hume’s First Law and say: Why stop at ethics? Why so half-hearted? That it is not legitimate to deduce one kind of statement from another kind of statement is surely an argument of more general application. Just as a sceptical attack can be launched upon the possibility of ethics, so why not launch sceptical attacks everywhere: on the possibilities of science and history and induction and deduction and everything? In particular, the tough-minded sceptic will say to the humean economist: Why do you stop with normative economics? — Surely you can and you must destroy all of positive economics as well!
It was shown some years ago by the English philosopher John Wisdom how sceptical attacks analogous to Hume’s attack on ethics in fact can be made in a number of other contexts as well. Let us consider an example similar to one given by Wisdom to show how easily it may be possible to proceed to be sceptical of something so obvious as our knowledge of the past. A sceptic says “Do we really know anything about what has happened in the past? Can we be certain about anything that has happened at all before this very instant?” You say to him “What do you mean? Surely you don’t mean that while we know some things for certain such as that we are now having this conversation, we don’t know for certain other things such as that we did get up from bed this morning or that Nazi Germany did invade Poland on September 1 1939?” The sceptic says “Yes that’s the kind of thing I mean.” You reply “Well that’s crazy. I for one am just as confident of knowing that here I am talking with you now, as I am that I got up this morning, as indeed I am that Nazi Germany invaded Poland on September 1 1939.” The sceptic says “Please tell me how you can be so certain you got up this morning.” Staring at him in disbelief, you reply “Look I usually get up to the alarm clock at 7 am; this morning was no different; I remember the clock going at 7 am as usual, and I got up. That’s all there’s to it.” The sceptic makes a flanking movement. “If you remember something taking place you would of course imply the event did take place?” You are now perhaps quite irritated by this odd fellow — “Obviously; I could not have remembered the alarm clock going off if it had not in fact gone off.” But in fact the sceptic has got you exactly in his sights and can move in for the kill. “In that case it appears to me you have missed the point of my original question completely. I wished to know how we can know anything about the past. You gave me an example that you knew you had gotten up this morning, and that you knew this for certain because the alarm clock had gone off as usual and that you remembered getting up when it did. I can agree of course that if you knew this premise to be true then you are entitled to deduce that you know you did get up this morning. But you will have to grant that this is a premise which itself refers to the past. So all you would have done in supporting one statement about the past is to have given me another statement about the past, when the point of my question was to ask how we can know anything at all about the past for certain.”
Just as the fact we cannot deduce a normative conclusion without a normative premise having been made might lead someone to a moral scepticism, so the fact we cannot deduce a conclusion about the past without a premise about the past being made might lead someone to a historical scepticism. That Nazi Germany did invade Poland on September 1 1939, cannot be deduced except by reference to other historical premises — films and photographs of the dive-bombers going in against the Polish Cavalry, government documents, the testimony of eye-witnesses, reports in the newspapers of September 2 1939, etc. The sceptic agrees that if the premises were known to be true then the conclusion would be true as well, but he says that that would be to miss his point. Like the moral sceptic, he is challenging the possibility of our knowledge of all propositions of a particular kind, and it is no use giving him for his scepticism what amounts to merely a another proposition of the same kind. Bambrough has put the matter clearly thus: “So long as the premises used in support of a proposition include any propositions of the same type as itself, a philosophical sceptic, or any other enquirer who is determined to seek the ultimate grounds, is properly dissatisfied, since his question is about how propositions of that whole type are to be validated, and he cannot consistently permit any such proposition to be unproblematic when it occurs among the premises of an argument whose conclusion is of the same type…. the grounds offered for a proposition of kind k will necessarily be either of kind k or not of kind k; if they are of kind k they may be logically sufficient for the proposition that they are intended to support, but a further question will arise about the validation of the premises themselves; if on the other hand they are not of kind k then they necessarily cannot be logically sufficient for the truth of the proposition that they are intended to support.”
Yet once this box has been opened, we are obliged to examine all its contents, and there are quite a number. For one thing we may now join with the sceptic of the senses and cast doubt on all the knowledge the natural sciences purport to provide of the physical world; since, surely, no conclusion about the physical world can be deduced without a premise about the physical world having been made. Next we might join with the solipsist and question the possibility of knowledge in psychology, doubting whether one can ever know what someone else thinks or feels; since, surely, no conclusion about a mind other than one’s own can be deduced without a premise of the same sort having been made. It is this species of scepticism which forms the basis of the widespread belief in modern economics of the impossibility of interpersonal comparisons of utility, which we observed in discussing the views of Professor Hicks in Chapter 2 and to which we shall be returning in Chapter 10. Then of course there is Hume himself being just as famous for his sceptical attack on the possibility of induction as he is for his attack on the possibility of ethics: “there can be no demonstrative arguments to prove, that those instances of which we have had no experience resemble those of which we have had experience.” “Nay, I will go farther, and assert, that [reason alone] could not so much as prove by any probable arguments, that the future must be conformable to the past. All probable arguments are built on the supposition, that there is this conformity betwixt the future and the past, and therefore can never prove it. This conformity is a matter of fact, and if it must be proved, will admit of no proof but from experience. But our experience of the past can be a proof of nothing for the future, but upon a supposition, that there is a resemblance betwixt them. This therefore is a point, which can admit of no proof at all, and which we can take for granted without any proof.” In short, no conclusion about the future can be deduced without at least one premise about the future having been made.
And then again, the full force of the sceptical onslaught can be felt when we direct its method against that of which we might seem most certain of all: the procedure of deduction itself in logic and mathematics. Adapting an example given by Wisdom and Bambrough, we can see how it may not be possible without circularity to use deductive reasoning to justify deduction itself.
For consider the propositions All firms maximize profits and GM is a firm. We would be normally inclined to think GM maximizes profits is something which follows from these. But the serious sceptic can once more ask how we may justify such a conclusion. We might be inclined to take such a challenge lightly, and try to dismiss it by stating a general rule of the form of modus ponnens: “If all S is P, and x is S then x is P.” But that would be a mistake and we would have fallen directly for the trap set for us, since the sceptic would need only to make the following decisive response: “A rule of this sort must necessarily either exclude or include the particular case at hand. If it is intended to exclude this particular case but is intended to apply to every other case, then clearly I need not accept in this case that the conclusion GM maximizes profits follows from the premises All firms maximize profits and GM is a firm. On the other hand, if the rule is intended to include this case as well, then you are asking me to reason as follows: ‘In all syllogisms, deduction proceeds like this; this is a syllogism; therefore, deduction proceeds like this here as well.’ All you would have done in trying to justify the deduction at hand is to have given me yet another deduction against which all my arguments would apply with equal force once more. You may not mind arguing in a circle but I am not going to join you.” If making an is ought dualism is sufficient ground for us to doubt the possibility of moral knowledge, then we seem now to have just as good grounds to doubt we can know anything at all. The upshot of these kinds of sceptical attacks on the practice of modern economics may be seen quite readily. For consider the fact that it would be difficult to overestimate the significance to the practice of modern economic science of (i) the elementary mathematical concept of a function, mapping all the values taken by one variable X upon a range of values taken by another variable Y, and (ii) the formal and informal procedures of statistical inference. Yet at their foundations, all procedures of statistical inference must rest upon the possibility of a rational induction. Suppose there was some economic variable Y which has been found to take a particular value in each of the last 100 or 200 or 300 or 500 periods. Or suppose it is found in each of a large number of observations that Y happens to be systematically related by some identifiable functional form to another economic variable X. It will be seldom if ever that we shall be obliged with such neat data, but it will be readily agreed the study of such relationships whether in economic theory or in economic history or in applied economics or in econometrics constitutes the very stuff of the modern science. The variable Y might be the quantity traded of a good where X is the market price, or Y the long-term interest rate and X the state of expectations, or Y the change in the price and X the difference between quantity demanded and quantity supplied, or Y the rate of inflation and X the money supply, and so on indefinitely in hundreds of different contexts. If we are genuinely serious about adopting a humean scepticism — that is, adopting it consistently, without contradiction — then we must lead ourselves to conclude that even with a thousand observations of Y taking a certain value after X had taken a certain value, we would still have no grounds, no deductive grounds, for predicting the value of Y given the 1001st observation of X. From no amount of past evidence can any proposition about the present or the future be deduced. Equally, if we were to prevent ourselves out of a debilitating scepticism of this kind from employing the modus ponnens of deductive reasoning — if all S is P, and x is S then x is P — then all reasoning in economic theory would immediately come to a standstill. Without induction and deduction, we cannot proceed in economics or elsewhere: it would be not only normative economics but all of economics which would come to be lost in the whirlpools of scepticism.
The point the sceptic wishes to make is that we cannot deduce one kind of proposition from an altogether different kind of proposition — the is ought dualism may be a useful reminder that we cannot deduce a normative conclusion from any number of positive premises. Every normative conclusion must have had at least one normative premise, and it is the attempt to justify one normative proposition by offering another as a premise that allows the moral sceptic to keep repeating his challenge indefinitely. But that does not prevent us from asking whether the sceptic has not skewed the rules of the game in such a way that he must always win, and if he has done so, we can certainly decline to play. For what the sceptic seems to require is that the grounds for any kind of justification specifically be deductive grounds. We are to deduce every proposition as the descendant of other higher or more primitive propositions, which might explain how the sceptic is able to raise the threat of an infinite regress in every field in which he attacks. “Everything we offer and everything we could conceivably offer is either too little or too much…. Nothing will ever do to meet the sceptic’s requirement. But that is different from saying nothing will ever do.” Perhaps it is not necessary to meet the sceptic’s requirement. Perhaps it is not even possible to do so. Perhaps we do not have to have a deductive proof to justify that we can and we do know some things in science, in history, in ethics, in psychology, in economics, or that we can and do frequently and reliably use inductive reasoning in these and a hundred other contexts. In Part II we shall be making an argument on these lines more fully to show how scepticism can be avoided even as we steer well clear of the opposite dangers of dogmatism. What is important here is only to notice the slide into total scepticism that may be entailed by adopting moral scepticism on its own. The economist who accepts an is-ought dualism as an adequate reason for adopting a subjectivist theory of knowledge comes to face an unhappy choice between either becoming in the interest of consistency a sceptic of all of economics — theory, history, econometrics, everything, not to mention everything else outside economics as well like natural science and mathematics and history; or denying the parity of reasoning, and not having adequate grounds for believing objectivity is possible in one context but not another. Either accept the propositions of positive economics and natural science and mathematics and history etc. to be, in the final analysis, just as subjective as normative propositions. The infinite regress threatens everywhere, what is sauce for the goose is sauce for the gander, so there cannot be objective knowledge of any kind anywhere. The economist slides into a scepticism about everything — into the pyrrhonism which Hume himself had rejected. Or become a partial and prejudiced sceptic like the positivist — led to the inconsistency of threatening only normative propositions with infinite regress when analogous sceptical attacks can be made with equal force in any number of non-normative contexts as well, and therefore not having adequate reason to maintain objective knowledge to be possible in contexts other than ethics. When asked “Can there be objective knowledge in economics?” if we answer “No, truth is defined merely by agreement of opinions; we know a proposition in economics to be true only insofar as economists happened to agree it to be true; if such agreement fails to hold in the future the proposition would no longer be true”, we next may be asked “Can there be objective knowledge in physics?”, to which we can only reply yes or no. If yes, we shall have said that there is merely rhetoric in economics, perhaps a highly evolved and sophisticated rhetoric but mere rhetoric nevertheless, certainly not objective knowledge. We would justify the cynic and the cartoonist who mocks economists as the most querulous of breeds, for every one who says this there is another who says that, how it is entirely a matter of caprice or fashion or pecuniary interest which side one happens to take, whose “paradigm” one happens to accept. We should have to frankly admit to the scholarly commmunity that since there is nothing which may be properly called objective knowledge in economics, the Department of Economics in every university should be closed down, or why there might just as well be a Department of Astrology on campus too, teaching and researching the reading of palms, the writing of horoscopes, and so on. On the other hand, if we denied there to be objective knowledge possible of the physical world as well, if we said we cannot be certain of such things as that there is a table in this room or that the window is open and there is a tree outside it, then we would have to do battle not only with every scientist in history but also with the man on the street, whose commonsense like our own tells us the opposite.
It is said that Hume thought himself leaving his scepticism behind when he left his study. Yet “[his] scepticism is at odds with his actions even when he is at his most deliberately and consciously philosophical. His pen goes confidently to the ink-pot, he turns the pages of Sextus Empiricus with the well grounded expectation that Book II will be found between Books I and III…. it is shown by his life that he believes what he is trying to doubt.” Just as surely as the scholastics fell under the Spell of Plato, so modern economists may have fallen under the Spell of Hume. The time has come at last to see how both spells may be broken.
PART II

5. Objectivity and Freedom

SUPPOSE there was a philosopher who addressed modern economists in a strange way as follows
Consider the entities that we call ‘firms’. I mean banks, manufacturers, airlines, law partnerships, farms, grocery-stores, and so on. What is common to them all? — Don’t say: “There must be something in common, or they would not be called ‘firms’” — but look and see whether there is anything common to all. — For if you look at them you will not see something that is common to all, but similarities, relationships, and a whole series of them at that. To repeat: don’t think, but look! — Look for example at banks with their multifarious relationships. Now pass to savings and loans associations; here you find many correspondences with the first group, but many common features drop out, and others appear. When we pass next to manufacturers and transporters, much that is common is retained, but much is lost. — Are they all ‘profit-maximizing’? Compare the taxi company with the electricity company. Or is there always a separation of ownership from management? Think of the tailor’s shop at the corner. With corporations there is the buying and selling of shares; but when a farmer is offered a price for his homestead this feature too may have disappeared. Look at the part played by entrepreneurship; and at the difference between the entrepreneurship of a mom-and-pop shop and the entrepreneurship of a firm of lobbyists. Think now of firms like General Motors; here is the element of giant size, but how many other characteristic features have disappeared! And we can go through the many, many other groups of firms in the same way; can see how similarities crop up and disappear.
What should we think of such a strange philosopher? And what answer is to be made to him by the economist?
The philosopher is Ludwig Wittgenstein, and the passage which has been paraphrased here, odd though it may seem, is among the most famous in twentieth century philosophy, from his posthumous work Philosophical Investigations. The problem that can be found to be raised in it is the ancient problem of universals, the problem of the One and the Many, of Unity and Diversity: Must all instances of a general term or concept have anything in common, over and above the fact they are all instances of the same concept? Must all firms have anything in common, over and above the fact they are all firms? Must all red things have anything in common, over and above the fact they are all red things? Certainly we know there to be individual red things like red poppies and red roses and red corpuscles and redheads and Red Square, and we know there to be individual firms like General Motors and Mitsubishi and Kodak and the corner grocery-store. But how is each individual red thing related to the general concept ‘Red’? How are General Motors, Mitsubishi, the corner-store etc. each related to the general concept ‘Firm’? Should we think of red poppies and red corpuscles and redheads as each sharing or partaking of some transcendental property, a universal, called ‘Redness’? Should we think of General Motors and Mitsubishi and the corner-store as each sharing or partaking of some universal called ‘Firmhood’? Would it be because they do that we call a red thing red or a firm a firm?
Interpreting Wittgenstein’s passage in this way, one response that might be made to it would be this: “What you seem to be doing is to test whether there is any property common to all firms. However, as your example suggests, individual firms are actually indefinitely varied — in their goals, constraints, size, type of ownership, operating characteristics, and so on. (Even if they were not indefinitely varied as a matter of fact, we can certainly imagine them being indefinitely varied in principle.) Indeed so much do individual firms vary that, in my opinion, we should not think there to be anything at all in common to all of them, besides of course our arbitrary decision to call them all ‘firms’.” Let us call such a reply the reply of the Nominalist.
But another response to the same passage could go like this: “I agree that what you are trying to suggest is that there is no common property between all the things we call firms. But surely in applying the concept ‘firm’ we must have an objective justification. For instance, while we do and we may apply the concept to General Motors and to Mitsubishi and to the corner-store, we do not and may not apply it just arbitrarily to any old thing at all — such as to my umbrella or to the number 16 or to Harry Truman or to the characters in a Dickens novel. Even when people refer to modern Japan as ‘Japan Inc.’, what they mean is that some analogy can be drawn between the way a firm works and the way political and economic arrangements in Japan seem to work, not that Japan is literally a firm, for that would be absurd since Japan is not a firm but a sovereign nation-state, a parliamentary democracy, a former Axis power, etc., and to call her a firm would be an objective misuse of language. It is likely that a property common to all individual firms does exist, and indeed it seems to me it is precisely because it does exist, whether or not we have been able to identify it, that we are entitled to call all firms ‘firms’, and so distinguish what are firms, such as General Motors and Mitsubishi and the corner-store, from what are not firms, such as Harry Truman or my umbrella or the nation-state of Japan.” Let us call such a reply the reply of the Realist.
The Nominalist stresses the Many — he is the lover of Freedom and Diversity, and the enemy of all Dogmatism and Conformity. He looks and insists that we look at the vast differences there are or can be — between firms, in the uses of words and concepts, across ways of life and culture, in the histories of nations, in the circumstances and personalities of individuals. The Realist worries about the indiscipline and caprice that can result from the exaggeration or corruption of freedom. He recognizes and insists that we recognize the vast areas of commonality there are or can be. We use words and language only because there are objective or “intersubjective” (Popper) ways of speaking and understanding. No matter how diverse individual personalities or circumstances or ways of life may be, the fact is we belong to one species (or one genus etc.), which implies something different from if we had not. The Realist stresses the One; he is the lover of Objectivity and Reason, and the enemy of all Scepticism.
A similar division may be made to obtain with any of a number of other concepts in economics as well — ‘capital’, ‘money’, ‘utility’, ‘competitive market’, ‘unemployment’, ‘development’, ‘mixed economy’, ‘socialist economy’, or any of a hundred others. In each case, the plea of the Nominalist would be that we observe the differences between the individual instances, the plea of the Realist that we respect the similarities. Indeed what should be supposed to be in common between individual economists themselves? From Aaron, Abramovitz and Ackley, through Bagehot, Baran and Bauer, and Cantillon, Cassell and Cournot, all the way to Zeckhauser, Zellner and Zeuthen, what is there in common except that each happens to be listed in a recent bibliographic dictionary of economists? The Nominalist would say “Nothing. Ultimately there is nothing in common to all economists except that we have chosen to call them all economists. That these people happen to be in the dictionary and other people like Picasso or Jesse Owens or Greta Garbo are not is, ultimately, just a matter of arbitrary choice.” The Realist would say “Surely there must be something in common to all economists, otherwise we would not call them economists. We wouldn’t in our right minds consider Picasso, Jesse Owens, or Greta Garbo to be economists, just as we wouldn’t consider Wicksell, Keynes, or Milton Friedman to be famous artists, athletes, or cinema stars. There must be an objective justification to calling someone an economist — it must be that economists are economists because they all believe in Q”; where Q would refer to some criterion like the practice of mathematical modelling, or an attribution of utility-maximization, or an attendance to statistical data, or a concern with the distribution of wealth and income. If someone did not believe in Q, did not fall under a specific definition of this kind, the Realist would be inclined to say such a person was not really an economist at all but something of an imposter or a charlatan who did not rightfully belong in the dictionary. And of course if one man chooses one Q and another chooses another then we may begin to explain how each might think himself to fall under his own definition of economist while it was the other fellow who was the charlatan.
A similar division can be made to obtain upon the larger concept of science itself. The Nominalist would observe the rich and indefinite variety there is in the methods and subject-matter of the individual sciences, and indeed that there can be within any of the individual sciences as well — certainly within physics, chemistry, biology, and engineering but also within mathematics, law, medicine, economics, history, and philosophy itself. Dazzled by all the different colours and the different shades of different colours, the Nominalist would tend to conclude there to be no unifying characteristic between the sciences, nothing except that we have chosen to name them all sciences. The Realist for his part would observe and be impressed by the many points of comparison there are between and within the individual sciences. And being especially concerned to protect the concept of science from being hijacked and employed arbitrarily to just anything at all, the Realist will be in search of the common ingredient which he thinks must be present in each individual science to warrant our calling it a science at all. The Realist will be inclined to say that all scientific statements have this in common — where his this would now refer to something like “hypothetico-deductive methodology”, or the use of mathematics or deductive proof, or the empirical testability or falsifiability of propositions, or knowing the means of verification. The Realist searches for the criterion or set of criteria which he believes to be necessary to demarcate science from non-science (Popper), public knowledge from private opinion. And again, if one man chooses one criterion to demarcate science from non-science and another chooses another and contrary criterion, we can imagine the merry possibility of how each might think himself to fall under his own definition of scientist while really it is the other fellow who is the charlatan and the fraud.
Parallel to this kind of a division between Nominalism and Realism in the theory of existence occurs the division between Scepticism and Dogmatism in the theory of knowledge which we have met with in previous chapters. A Nominalist in ontology is likely also to be a Sceptic in epistemology, and a Realist in ontology is likely also to be a Dogmatist in epistemology and vice versa. C. S. Peirce had remarked that two points of contrast between scholastic and modern thought lay in the modern opinions that thought “must begin with universal doubt, whereas scholasticism had never questioned fundamentals” and that “the ultimate test of certainty is to be found in the individual consciousness; whereas scholasticism had relied on the testimony of sages and of the Catholic Church.” The Dogmatist finds there are at least some things which are certainly known. Therefore, he concludes, it must be that we cannot question everything, it must be that there are at least some propositions which should be supposed to be closed to further inquiry and discussion. Thus the medieval schoolmen would have supposed the Christian Scriptures to contain at least some propositions of this sort. Certainly there is scope to reason but it is a scope necessarily limited by the doctrines and dogmas of the faith. It would be precisely against this kind of a barrier being placed on the road of inquiry that the Sceptic protests. And finding there to be no human belief which must be thought of as closed to further question, the Sceptic concludes that it must be we cannot know anything for certain. Each side seems to have a compelling reason in its favour yet to be in direct contradiction of the other. One asks for belief and conviction, the other for doubt and question. The feeling of an antinomy arises because we feel we must choose between them.
It was suggested in Chapter 3 that medieval political thinking was platonistic and absolutist in important respects, and evidence has been given in Chapter 2 that modern economists have adopted the sceptical humean epistemology which may be seen as a reaction to the medieval dogmatism. As Peirce’s remarks make clear, this would not be a new thesis, though it is perhaps something which has not been adequately noticed before by modern economists and it has now been plainly set out. It is also a thesis which amounts to being a generalization, and suffers, as all generalizations must, from a lack of truth in its details, especially in not doing nearly enough justice to the depth and diversity of medieval thought. Yet every generation must be concerned with identifying and correcting the errors of its own time, and the purpose of trying to establish even such a generalized thesis as this has been to correct contemporary errors: to argue that the humean foundations of the modern theory of economic knowledge entail serious difficulties, that it is these and not the is-ought dualism which turn out to be insurmountable, that the broad and long standing consensus on the central question of the relationship between economic knowledge and economic advice, the positive and the normative, cannot be held consistently and deserves to be abandoned.
Nevertheless the reader who may have agreed with the drift of these arguments may wish to ask whether, in an attempt to correct contemporary errors, we shall not be led to commit the errors of an earlier time. Will we become Dogmatists if we renounce Scepticism? Are we forced to choose between Realist and Nominalist, Dogmatist and Sceptic, Plato and Hume? Must we either admit objectivity and reality and knowledge and expertise and common reasoning and commonsense, and suppress diversity and individuality and creativity and freedom and question and criticism; or embrace diversity and individuality and creativity and freedom and question and criticism, and abandon objectivity and reality and knowledge and expertise and common reasoning and commonsense? Can we lead our thinking lives coherently enough without making a choice, or would we find ourselves inevitably being shuttled between the rival parties, one moment in the Nominalist’s camp the next moment in the Realist’s, one moment with the Sceptics the next moment with the Dogmatists? If we decide to abandon Hume, is there no choice but Plato? If we find Plato’s embrace too close and claustrophobic, is there no alternative but to continue to live in doubt with Hume? Are we caught between the Spell of Plato and the Spell of Hume? Is the choice: Either Objectivity or Freedom?

§2. The simple answer that may be offered is that it is not. When objectivity and freedom, knowledge and doubt, have been carefully and adequately characterized, there is no conflict which must arise between them, whether in natural science, mathematics, ethics, history, economics, medicine, law, literature, or any other context of inquiry. There may be good reasons to be a Nominalist and also good reasons to be a Realist and yet better reasons to be neither. There may be good reasons to adopt a sceptical theory of knowledge and also good reasons to adopt a dogmatic theory of knowledge and yet better reasons to adopt neither. A course can be found which will allow us to steer clear of the hazards of Dogmatism on the one side while avoiding the whirlpools of Scepticism on the other.
How we may proceed to chart such a course is by airing and exposing a hidden and questionable assumption which may be being shared by both Nominalist and Realist. Namely, an assumption that for a general term or concept like ‘firm’ or ‘game’ or ‘science’ to be objectively employed, there must also correspond some sort of object. Just as alcohol is common to whiskey and beer and gin, so some common ingredient must be present in General Motors and Mitsubishi and the corner grocery-store in order to make them all firms. If such an assumption does happen to be at the source of the division between Nominalist and Realist, we might readily explain how it is that each seems plausible in part yet neither seems satisfactory as a whole. The Nominalist finds he cannot distill out any single common ingredient from all the particular instances of firms that there are or can be. But because he may be committed to an assumption that such an ingredient is necessary for the concept ‘firm’ to be objectively employed, he concludes it cannot be objectively employed. The Realist is certain the concept ‘firm’ can be objectively employed, and very certain it should not be arbitrarily employed, but because he too may be committed to the same assumption, he concludes there must be a common ingredient, a common “essence” which every particular firm must share, prompting him to make a search for it or merely declare his faith in it being “there”, somewhere, “out there”.
Wittgenstein in his later works (as well as others before and after him such as H. A. Price) may be understood to have offered a suggestion that to make this kind of dualism between Nominalism and Realism is ultimately mistaken and misleading. After careful and detailed examination of a variety of the individual entities or institutions or activities which fall under a general concept like ‘firm’ or ‘game’ or ‘competitive market’ or ‘mixed economy’ or ‘economist’ or ‘science’, it may well be that we shall wish to make an entry in our notebooks of the following sort: “We see a complicated network of similarities overlapping and criss-crossing: sometimes overall similarities, sometimes similarities in detail. I can think of no better expression to characterize these similarities than ‘family resemblances’; for the various resemblances between members of a family: build, features, colour of eyes, gait, temperament etc. etc. overlap and criss-cross in the same way.” An alternative to a common ingredient model of the structure of concepts would be a family resemblances model, and an example constructed by Bambrough may easily illustrate its working. Suppose there to be five objects, A, B, C, D, E, each of which has four out of five possible properties, a, b, c, d, e. A pattern may be produced like
object A B C D E
properties bcde acde abde abce abcd
in which each object would evidently share 75% of its properties with every other yet there would no single property or set of properties common to all the objects. “But if someone wished to say: ‘There is something in common to all these constructions — namely the disjunction of all their common properties’ — I should reply: Now you are playing with words. One might as well say: ‘Something runs through the whole thread — namely the continuous overlapping of those fibres.”
Many concepts, perhaps even most concepts, may be family resemblance concepts, their instances constituting “a ‘family’ of diverse things bundled together by virtue of shifting similarities”. While there may be no single or constant similarity between all the individual instances of firms or games or economists or sciences or competitive markets or mixed economies, there may be diverse and shifting similarities between the different instances. It is these shifting similarities which can provide an adequate justification for supposing the different instances to fall under the same concept; while the recognition that there is no need for them to be anything but shifting in kind would equally justify not making a search for some mysterious essence which must be common to the individual instances. (We might even “throw away the ladder” after we have climbed with it — for armed with such a model of the structure of concepts, we might even take Nominalism and Realism as family resemblance concepts themselves!)
A parallel observation is suggested about the division between Sceptic and Dogmatist, and a parallel resolution may be offered as well. Perhaps there too the problem occurs because the Sceptic and the Dogmatist have been united in sharing a hidden and questionable assumption, viz., that if knowledge is to be considered objective, it must also be considered absolute, not admitting any error or exception. The Sceptic correctly sees error to be possible, indeed error to be ubiquitous, and so an absolute or exceptionless knowledge to be impossible; from which he mistakenly concludes objective knowledge to be impossible. The Dogmatist correctly sees many things indeed to be known, but mistakes the character of what is known or at least some of what is known as incorrigible and unexceptionable, and goes on to deny error and exception to be possible. An equal and opposite error would be to confound the notion of something being personal or subjective with respect to an individual and the notion of something being relative to a given individual case or context or circumstance. That something can be true or right in a given case, context, or circumstance does not imply it must be true or right in all cases or contexts or circumstances. Nor does it have to mean that such knowledge must have been derived by applying an absolute and unexceptionable law or theory to a particular case. What may be true or right simply may be true or right relative to the particular case or context or circumstance, while the fact it is relative to the case or circumstance would not imply that it is a matter of subjective choice whether it is true or right.
An example can illustrate. If a child asked us whether Chicago is to the left or the right of New York, we might say that this is an incomplete question with no definite answer. Relative to someone looking north in Washington, Chicago is certainly to the left of New York, while relative to someone looking south in Montreal, it is to the right of New York. In each case, there is an objectively right answer to the question relative to the situation of the observer. And the significant fact would be the situation of the observer, not what his subjective beliefs might happen to be. If a man in Montreal said Chicago was to the left of New York he would be making an objective mistake in the sense that anyone in his situation should be reasonably expected to conclude the opposite. Or consider that while the West is due West and the East is due East of Istanbul, the West is due East and the East is due West — of Honolulu. The Sceptic would take the fact different and conflicting answers are possible to the same question as evidence for the conclusion that it is ultimately arbitrary what we call West or East, or whether Chicago is to the left or right of New York. The Dogmatist would take one or the other answer and conclude it must hold absolutely true everywhere, without possibility of exception or error. The division has been expressed clearly by Bambrough like this: “Both the sceptic and his dogmatist opponent assume that the absoluteness of logical space is necessary for the objectivity of enquiry; that in seeking knowledge and understanding we orient ourselves, if at all, by fixed landmarks whose own positions neither can be nor need to be the subject of investigation. Sceptics become sceptical because they recognise that what they believe to be necessary is nevertheless not possible. Dogmatists become dogmatic because they rebel against the paradoxes of scepticism but still agree with the sceptic on what is necessary for the validity of our knowledge. One party denies the possibility of knowledge because it sees that logical space is relative and the other denies that logical space is relative because it sees that knowledge is possible.” Both Sceptic and Dogmatist may be seen as united in their belief as to what will be allowed to count as knowledge — in what must be supposed to be the appropriate model of the justification of knowledge. In answering the question “How do we know this?” both may be assuming that we have to deduce our answer from some previous and more general law, rule, or theory; the answers we seek or arrive at must always be a particular application or exemplification of some more general thesis. (Wittgenstein wrote of a “craving for generality” and a “contemptuous attitude towards the particular case”.) The Sceptic becomes sceptical because he finds the process of deduction to be one without end. Deduction cannot be done without a remainder of unproven premises — a conclusion is deduced from a set of premises, each of which is the conclusion of other sets of premises, each member of which is the conclusion of yet other sets of premises, and so on. For every proposition there seems to be a genealogical tree consisting of all the lines from which the proposition deductively descends. The fact these lines can be indefinitely extended to unknown reaches leads the Sceptic to think the pedigree of every proposition to be questionable, that every argument ultimately must be inconclusive, that there really can be no such thing as certain knowledge. The Dogmatist shares the same kind of idea that the only justification of knowledge is a deductive justification, and also observing the same kind of threat of infinite regress in argument, decides to call a halt at some or other point; the precise point where to halt either being determined ex cathedra (the medieval schoolmen) or being chosen arbitrarily (the humean economist!). At such a point the Dogmatist is ready to stand and fight, and of course if different people choose different and contrary points, we may expect some mighty rows indeed to develop between rival dogmas. Indeed it is possible that economists who have subscribed to the received theory of knowledge have been both sceptical about the possibility of moral knowledge and dogmatic about the existence of supreme unquestionable normative primes and principles. The widespread adoption of moral scepticism may be itself a relevant fact in explaining how it is that numerous divisions of opinion have been so persistent in modern economics, whether with respect to the methods or the substance of inquiry in the subject. Thus it is possible to find eminent economists being in deep and seemingly irreconcilable conflict with one another on questions of method or theory or evidence or policy, being members or even founders of rival schools of thought, yet being completely agreed that the logical status of economic advice is equivalent ultimately to that of personal bias or prejudice. As Peirce remarked: “When society is broken into bands, now warring, now allied, now for a time subordinated one to another, man loses his conceptions of truth and of reason. If he sees one man assert what another denies, he will, if he is concerned, choose his side and set to work by all means in his power to silence his adversaries. The truth for him is that for which he fights.”

§3. It is possible that this parallelism between the Nominalist/Realist divide in the theory of existence and the Sceptic/Dogmatist divide in the theory of knowledge is not accidental. There is a possible connection which goes back to Plato. For it was part of Plato’s thinking that the things we find in the world are merely distorted and defective versions of ideal entities not actually given to human experience. In mathematics for example, a platonist would say that the dot we make on a piece of paper and call “a point” is but a defective image of the ideal point which has no parts or magnitude; the chalk mark on the blackboard which we call “a line” is but a defective version of the ideal line which has no breadth or width, and so on. It is these kinds of ideal points, lines, planes, etc. which are the true objects of mathematics; while they do not have location in the world in which we live that does not mean they are any less real. Rather mathematical objects should be thought of as inhabiting a kind of transcendental universe, a domain not directly observable yet which is reachable through the reasonings of the mathematician and philosopher, whose task it would be to discover and chart this unobservable terrain much as the geographer and astronomer discover and chart the observable earth and universe in which we live. As the English mathematician G. H. Hardy put it: “For me, and I suppose for most mathematicians, there is another reality, which I will call ‘mathematical reality’; and there is no sort of agreement about the nature of mathematical reality among either mathematicians or philosophers. Some hold that it is ‘mental’ and that in some sense we construct it, others that it is outside and independent of us…. I believe that mathematical reality lies outside us, that our function is to discover and observe it, and that the theorems which we prove, and which we describe grandiloquently as our ‘creations’, are simply notes of our observations.” Professor Michael Dummett has put it recently like this: “[Platonism is] the thesis that there really do exist such structures of abstract objects, and that we are capable of apprehending them by a faculty of intuition which is to abstract entities as our powers of perception are to physical objects.”
And ideal mathematical objects need not be the only inhabitants of Plato’s heaven. So could be ideal men and ideal women, ideal marriages and ideal families, ideal languages and ideal cultures, ideal economic agents trading at ideal prices in ideal markets, ideal societies and ideal polities. In fact there is some evidence to think modern economic theorists may have subscribed to such a view. For example, Professor Arrow remarked in his Nobel Lecture: “In my own thinking, the model of general equilibrium under uncertainty is as much a normative ideal as an empirical description. It is the way the actual world differs from the criteria of the model which suggests social policy to improve the efficiency with which risk bearing is allocated.” And Professor Hahn in his Political Economy Lecture at Harvard University and elsewhere has argued that the model of general equilibrium “serves a function similar to that which an ideal and perfectly healthy body might serve a clinical diagnostician when he looks at an actual body”, that even though the model “is known to conflict with the facts” and “is not a description of an actual economy” it nevertheless tells us “what the world would have to look like” if a neoclassical view of the economy is to be considered plausible. What is it possible to understand Arrow and Hahn to mean by such remarks except to be endorsing a platonist ontology? If so, it would of course sit oddly with their subjectivism elsewhere; we shall return to these matters in Chapters 9 and 10.
The platonist seeks to mentally grasp the ideal entities by his “mind’s hand” as it were, to use a phrase of Professor Morton White . And once he believes himself to have done so, the expression of his understanding would amount to being not only an expression of objective knowledge but an expression of absolute knowledge as well — something which is necessarily free of error or exception since it would have been the ideal which had been understood and expressed. The Realist becomes the Dogmatist. The Nominalist for his part wants nothing whatever to do with tales of airy fairy entities in transcendental heavens. As Professor W. V. O. Quine might have put it, what needs to be done instead is to make a clean shave of Plato’s Beard with Occam’s Razor. But in rejecting a picture of transcendental entities and the theory of absolute knowledge that goes with it, if the Nominalist cuts too thickly, he ends up rejecting the possibility of objective knowledge as well; the Nominalist becomes the Sceptic.
The theory of knowledge suggested by the writings of Peirce and Wittgenstein independently, suggests a third route. Reject Plato’s theory of a transcendental universe, as being unnecessary to the resolution of any question in the theory of knowledge. With it therefore is rejected the idea that to know something certainly and objectively we must have deduced it from some absolute and general law, theory, rule or principle; that when we say we know something we must be in fact expressing the discovery of some ideal transcendental “form”. Gone at once would be the possibility of an error-free and exceptionless knowledge which forms the basis of the Dogmatist’s dogmatism. Error and folly are ubiquitous: Let freedom ring! At the same time, once we unshackle ourselves from the cramped idea that every claim to genuine knowledge must be deduced from some previous and higher claim to knowledge and ultimately from some set of unquestionable supreme principles or axioms, we may reject Hume just as decisively as we reject Plato. The antidote to Hume’s debilitating and self-contradictory scepticism is commonsense. — We know some things are true and other things are false, we know some things to be right and other things to be wrong. And we can know these things without having to be haunted by an idea that we do not truly know them unless we have deduced them from some “higher” or more general proposition. The general rule or principle or theory may serve perfectly well as the unquestioned premise of one argument only to be the questionable conclusion of another. The inductive and the deductive may alternate in the activity of reasoning, as we proceed from one set of particular cases and questions to another set of particular cases and questions via as many general rules, principles, and theories that we need. As John Wisdom put it: “Examples are the final food of thought. Principles and laws may serve us well. They can help us to bring to bear on what is now in question what is not now in question. They help us to connect one thing with another and another and another. But at the bar of reason, always the final appeal is to cases.”
Furthermore, there may be a third and alternative mode of reasoning too, namely, reasoning by analogy. When faced with a question to which we do not have an answer, what may be required of us may involve neither induction nor deduction but comparison and contrast. The most reasonable way to proceed in a given situation may be to take the question at hand to which we do not presently have an answer and compare and contrast it with questions on either side of it to which we do have true or right answers. Here is a question L to which we do not presently have an answer. But we do know the answer to a question K which is close to L on one side, as well as the answer to another question M which is close to L on the other side. Now our question is, is L more like K or more like M? The reader may agree that that is how much reasoning does in fact proceed — in mathematics as much as in medicine, in science as much as in literature, in engineering as much as in ethics. It may turn out that on a particular question L the present state of our knowledge happens to be so poor that we require an answer not only to K but also to I, H, G, F, E, on the one side of it, as well as an answer not only to M but also to N, O, P, Q, R, on the other side of it, as well as perhaps to questions above and below and all around it. Will that mean our project is hopeless or that common reasoning can be of no avail in answering L? Not at all — it would only mean there is that much work to be done. For inquiry to be inchoate does not have to be cause for despair.
This kind of a notion that in the actual process of inquiry we always do start somewhere, and indeed that that is the only place to start, is to be found being expressed in the writings of Peirce: “We cannot begin with complete doubt. We must begin with all the prejudices which we actually have when we enter upon the study of philosophy. These prejudices are not to be dispelled by (the Cartesian maxim that philosophy must begin with universal doubt) for they are things which it does not occur to us can be questioned. Hence this initial skepticism will be a mere self-deception, and not real doubt; and no one who follows the Cartesian method will ever be satisfied until he has formally recovered all those beliefs which in form he has given up…. A person may, it is true, in the course of his studies, find reason to doubt what he began by believing; but in that case he doubts because he has a positive reason for it, and not on account of the Cartesian maxim. Let us not pretend to doubt in philosophy what we do not doubt in our hearts.” Then again: “Philosophers of very diverse stripes propose that philosophy shall take its start from one or another state of mind in which no man, least of all the beginner in philosophy, actually is. One proposes that you shall begin by doubting everything, and says that there is only one thing that you cannot doubt, as if doubting were ‘as easy as lying’. Another proposes that we should begin by observing ‘the first impressions of sense’, forgetting that our very precepts are the results of cognitive elaboration. But in truth, there is but one state of mind from which you can ‘set out’, namely, the very state of mind in which you actually find yourself at the time you do ‘set out’ — a state in which you are laden with an immense mass of cognition already formed, of which you cannot divest yourself if you would; and who knows whether, if you could, you would not have made all knowledge impossible to yourself? Do you call it doubting to write down on a piece of paper that you doubt? If so, doubt has nothing to do with any serious business.” A remarkable resemblance to this line of thought is to be found in the later writing of Wittgenstein: “If you tried to doubt everything you would not get as far as doubting anything. The game of doubting itself presupposes certainty.”
No theory of knowledge can compel us to think of the activity of reasoning to be starting all of a sudden out of nothing and nowhere, nor are we obliged to suppose it must have any necessary end. We always start somewhere — there are always cases to which we do have answers with which to compare and contrast the particular case presently in question. And there are always unexamined cases and unasked questions remaining, which we may bring to test the validity and soundness of any general law or theory or definition or principle in which we may have come to believe on the basis of the known and settled cases. Thus reasoning can be thought of as a certain and objective activity without having to be thought of as an exhaustive activity. Argument can be potentially endless, but it is not thereby inconclusive. It is conclusive, but it is not thereby absolute or final. There need not be either any canonical points from which we have to begin our reasonings, or any ultimate destination at which we have to stop. Reasoning can be objective without being thought of as having to have either an absolute beginning or an absolute end. We can be objective without being platonist, we can admit a rich and indefinite variety and diversity without being subjectivist.
In the next chapter this line of argument is continued in more detail and concluded.

6. Expertise and Democracy

In this chapter we shall consider in more detail the thesis introduced in the last, with the intent of together providing the main outlines of a theory of economic knowledge with which to replace the received humean theory.

§2. Our first task is to try to provide a more formal refutation of scepticism, i.e., to formally prove the existence of knowledge, a task which is in fact quite readily accomplished.
We have noted in previous chapters the important difference between the question of whether it is possible for an objective answer to be given to a question, and the question of whether someone should be thought of as possessing such an answer and how we are supposed to identify him or her. The question of whether there can be any expertise about a given matter is independent of (and prior to) the question of who if anyone should be thought of as an expert about it. Scepticism, considered technically as a thesis in the theory of knowledge, needs to be concerned with the former question alone; the consistent and universal sceptic being someone who takes each and every concept like ‘scientific knowledge’, ‘historical knowledge’, ‘moral knowledge’, ‘mathematical knowledge’, ‘probable knowledge’, ‘economic knowledge’ etc. and argues it to be empty, devoid of content, ultimately extending to no instances, in the way concepts like ‘unicorn’ or ‘reigning Czar of Russia’ would be said to have no instances. Equally a refutation of scepticism may proceed as a logical exercise as well, amounting to showing the existence of just one instance of knowledge. And to argue the possible existence of knowledge in this way would not be to commit oneself to any claim of knowing who should be thought of as an expert or indeed to any claim of knowledge for oneself. The heated political problem of who is supposed to be an expert and how we are supposed to identify him or her deserves to be kept separate from the cooler logical problem of whether there can be any knowledge on a question in the first place.
It is in such a light that we may view the proof of the existence of an external world given by the English philosopher G. E. Moore. Moore raised his hands one at a time before the British Academy and declared to the effect “Here is one hand and here is another. Therefore we know there are at least two objects in the external world.” Or Moore might have taken a pencil from his pocket and said: “Here is a pen; therefore we know there to be a world outside our minds.” The sceptic who protested that Moore was holding a pencil and not a pen would have helped Moore to prove his point, in that an attempt to deny Moore was holding an object in his hand could not be more certain than Moore’s claim itself. A single such example may suffice to show the concept ‘knowledge of the external world’ to be not empty and scepticism of the senses to be false and misleading. Moore wanted to show that we can and we do know some things for certain, and that we know them neither by induction or deduction necessarily, nor by fiat or dogma or mysticism, but simply by commonsense. Furthermore, if a theory of knowledge came to imply we did not know such things to be true when we did know them to be true then it was likely that it was the theory and not commonsense which was in error. Thus Moore declared that he most definitely knew that there was a living human body which was his body; that this body had been born at a certain time in the past and had existed continuously since then though not without changes; that it had come into contact with and been at various distances from many other things also having shape and size in three dimensions; that the earth had existed for many years before he had been born; that his body had been always in contact with or not far from the surface of the earth, and so on. Moore said that not only did he know these things to be certainly true, but that all of us know such things to be certainly true as well. In short, the problem of proving the existence of knowledge of an external world had a simple and yet rigorous solution.
An analogous proof of the existence of moral knowledge has been given recently by Bambrough by way of the following example: “We know that this child, who is about to undergo what would otherwise be painful surgery, should be given an anaesthetic before the operation. Therefore, we know at least one moral proposition to be true.” Bambrough claims there can be no argument to refute this proposition which does not accept the logical existence of moral knowledge. For suppose we tried to disagree on whether the child should be given the anaesthetic; there might be any of a number of grounds for doing so — such as the parents forbidding it, or because it went against the religion of the child and the child refused it, or because it was wartime and there was a shortage of anaesthetic and the child needed only a stitch on the hand when there were more serious cases needing the same scarce anaesthetic, or because the child was a premature and underweight newborn and there was danger it would not survive an operation under anaesthetic, and so on. That is, because there were other values besides that of avoiding unnecessary pain which were considered relevant to the problem at hand. We would have entered into a substantive moral debate with Bambrough, and pari passu we would have implied that whether it was he who was right to say the child should be given the anaesthetic or we who were right to say the child should not be given the anaesthetic, there was a right answer to the question whether the child should or should not be given the anaesthetic in the circumstances. A logical thesis of the objectivity of moral knowledge needs to establish only that there is, in principle, a right answer to every question as to what ought to be done. And this can be maintained without having to make any claim of either