FOUNDATIONAL EPISTEMOLOGIES IN CONSUMPTION THEORY

Alan Storkey

VU University Press is an imprint of:
VU Boekhandel/Uitgeverij bv
De Boelelaan 1105
1081 HV Amsterdam
The Netherlands

tel. (0)20-6444355
fax (0)20-6462719

ISBN

© Alan Storkey, London, 1993
All rights reserved. No part of this book may be repreduced, stored in a retrieval system, or transmitted, in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior permission from the copyright holder.

VRIJE UNIVERSITEIT TE AMSTERDAM

FOUNDATIONAL EPISTEMOLOGIES IN CONSUMPTION THEORY

ACADEMISCH PROEFSCHRIFT

ter verkrijging van de graad van doctor aan
de Vrije Universiteit te Amsterdam,
op gezag van de rector magnificus
dr. C. Datema,
hoogleraar aan de faculteit der letteren,
in het openbaar te verdedigen
ten overstaan van de promotiecommissie
van de faculteit der sociaal-culturele
Wetenschappen
op woensdag 17 november 1993 te 15.30 uur
in het hoofgebouw van de universiteit, De Boelaan 1105

door

Alan James Storkey

geboren te London

Promotoren : prof. dr. B Goudzwaard
prof. dr. S Griffioen
prof. dr. B Kee

Referent : prof. dr. D Th Kuiper

Preface and Acknowledgements
The genesis of this study was a period in 1966-7 when I was especially involved in epistemological issues, partly through attending seminars run by Popper, Lakatos and Winch and partly through discussions with Elaine on her work on Wittgenstein. One of the outcomes was a long paper on the different epistemologies embodied in the various movements of modern art circulated by Hans Rookmaaker. Elaine and I then continued to work through a variety of epistemological issues in relation to British and Dutch Christian philosophy, and to develop a Christian evaluation of theories of knowledge used in the social sciences. In the 1970s, partly in relation to Tony Cramp’s pathbreaking Cambridge lectures, it was articulated into a critique of some of the trends in economic and sociological thinking, expressed in embryo in chapter 3 of A Christian Social Perspective. This emerged as part of the background critique developed with the Calvin College Centre team of George Monsma, John Tiemstra, Karl Sinke and Fred Graham in 1980/1, and it was during this time that the ideas began to be shared with Bob Goudzwaard, whose thinking has deepened and opened the perspective considerably. At the end of that year Bob offered to supervise the study. The focus on consumption theory grew out of an interest at Cambridge, a later focus on the sociology of the family and the institutional reformulation which is present in the Calvin College team’s Reforming Economics. The study has developed over the last decade in work with Bob Goudzwaard, and also with Sander Griffioen and Bas Kee, who have also contributed much to its formulation. Although post-modernism has been slower emerging in economics, it is now beginning to be clear how deep the critical reflection will need to be, and this study is offered as part of that process.
Thanks for many of the resources of this study are due especially to the following libraries and those who work within them: Sheffield University, Calvin College, Gordon College, Oak Hill College, and especially the British Library of Political and Economic Science at the LSE and Cambridge University Library.
The privilege of working with Bob Goudzwaard, Sander Griffioen and Bas Kee on this topic goes far beyond the scope of the actual work itself and I thank them for what they have shared, for their insights, scholarship and friendship. I am also grateful to those mentioned earlier and to many others for discussions and insights, and to Calvin College and Oak Hill College for research time.
Finally, I give intimate thanks to my Father and Mother for long term encouragement, to Amos, Matthew and Caleb for the willingness to discuss issues and provide technical support, to Elaine for her love, faith, thought and commitment to this endeavour, and centrally to God as the source of knowledge, contaminated though it is by our weaknesses.

Contents
Preface and Acknowledgements I
Contents II
Introduction V
Chapter one: Foundationalism and Consumption Theory. 1
Theory as a Path. 1
The Earlier Crisis in Classical Economics. 2
The Epistemic Community of Later 19th Century Economists. 7
The Public Affairs Economists. 9
The Search for Academic Recognition. 10
The New Human Sciences: Their Definition and Coherence. 12
Locating the Problems of Knowledge. 15
Epistemological Foundationalism. 20
The Failure of Foundationalism. 27
The Structural Problems of Foundationalism. 33
1. Foundational Validity. 33
2. Foundational Dogmatism. 34
3. Theoretical and Methodological Prescription. 35
4. The Problem of Excluded Knowledge. 35
5. Nonscientific and Scientific Knowledge. 36
6. Otherworldliness. 36
7. Economic Boundary Disputes. 37
8. Disciplinary Autonomy and Imbalance. 38
9. The Subjectivity of the Economist. 38
The Early Development of Consumption Theory. 39
Foundationalism and Consumption Theory. 41
Chapter two: Rationalist Foundations in Consumption Theory 45
Introduction. 45
Rationalism as a Foundational Epistemology. 46
Section 1 The Logicist Tradition. 48
Jevons and the Changing View of Logic. 48
Jevons and Consumption Theory. 53
Pareto’s Consumption Theory. 56
Hicks’ Logical Foundationalism. 62
The Structural Problems of Logicist Consumption Theory. 67
Section 2 The Formal Rationalist Tradition. 70
Its Origins and Development. 70
Kant’s A Priori Foundational Vision. 71
Cournot and Auguste Walras. 74
Lèon Walras’ Formal Rationalism. 77
Schumpeter and the Transmission of Economic Culture. 80
Samuelson’s First Epistemological Base. 83
Evaluation of Formal Rationalism. 88
Section 3 Means-ends Rationalism. 91
Subjective Kantianism. 91
Kaufmann’s and Robbins’ Instrumental Activity. 97
Becker’s Means-ends Utility Function. 103
Conclusion. 106
Chapter three: Positivism and Consumption Theory. 109
Introduction and Background. 109
The Foundational Move and the Three Stages of Modification. 112
Section 1 Simple Positivism and Descriptivism. 114
Attempts at the Foundation. 114
Positivism, Otherworldliness and other problems. 118
Samuelson, Revealed Preference and Descriptivism. 122
The Cost and Standard of Living. 125
Section 2 Logical Positivism. 128
Logic and Positivism. 128
Samuelson Again. 132
Section 3 Hypothetical and Falsificationist Positivism. 134
Popper and the Break with Inductivism. 134
Friedman’s Hypothetical Positivism. 137
Section 4 The Frequency Inductivists. 141
Probable Certainty. 141
Consumption and Econometrics. 144
Conclusion. 149
Chapter four: Casual and Behavioural Foundationalism. 151
Background. 151
The Epistemological Base. 153
Section 1 Cambridge Causal Theory. 157
The Cambridge School. 157
Mill and the Foundation of Causal Epistemology. 159
Mill’s Epistemology in Economics and Consumption Theory. 162
Marshall and Causal Theory. 164
Marshall’s Consumption Theory. 168
John Neville Keynes and W E Johnson. 171
John Maynard Keynes and the Theory of Knowledge. 173
Keynes Economic Theory. 178
Keynes and Consumption. 181
Section 2 The Behaviourist Tradition. 183
Behaviourism. 183
Consumer Behaviourism. 188
Scitovsky’s Change of Basis. 189
Behavioural Methodology: Kantona and the Survey Research Centre. 191
Conclusion. 194
Chapter five: Conclusions and New Directions. 197
Review of Foundationalism. 197
A Christian Diagnosis of Foundationalism. 201
Christian Orientations to the Human Sciences. 204
The Theoretical Location of Consumption Studies. 205
The Culture of Consumption – The Reintroduction of Value. 207
1. Survival. 208
2. Tradition. 209
3. Puritan / Ascetic. 209
4. Hedonist. 210
5. Ecological. 210
6. Acquisitive. 211
7. The Personal Fulfilment Lifestyle. 212
8. Marriage and Family Identity. 212
9. Security. 213
10. The Meek. 213
An Interdisciplinary Framework. 216
The Institutional Context of Consumption. 220
Relational Consumption. 224
Investigating Thrift. 228
Consumption and Work. 231
Post-Foundationalist Theory and Methodologies. 233
Theory and Living. 235
ALAN STORKEY, The epistemological foundations of consumption theory., de kennistheoretische fundering van de consumptie-theorie. 239
Bibliographies 243
Epistemology and the Philosophy of Science. 243
Economic Epistemology and Consumption Theory Bibliography. 251
Index 265

Introduction
This study has a shape which is unusual, and it is good at the beginning to share a strategic view of its scope and intentions. The subject matter is consumption theory as it has developed in economics from the 1870s to recent times, what could be called ‘modern consumption theory’. The focus of the study is on the epistemological construction of that theory, and as such forms part of the growing interest in the philosophy of economic science which has been fostered by Hutchinson, Blaug, Caldwell, Carabelli and others. It carries the concerns of these theorists further by articulating the epistemological issues in relation not to a single theorist or period but to a subdiscipline of economics. It also has something in common with post-modernist critiques in bringing into question long-term orthodox constructs which have had authority in areas of academic study. The central thesis is that much of consumption theory has been definitively shaped by a drive to establish epistemological security, what is here called a ‘foundational’ drive, which has distorted its development. As a result the basis of this subdiscipline is suspect in a way which damages its ability to address the issues of consumption which are thrown up by the domain of study. However, the aim of the critique is to open up the possibility of a reconstituted view of consumption theory itself, and this is the intention especially of the last chapter. Such an initial terse account needs to be filled out by a more discursive narrative of the book.
The first chapter sets the scene by reinterpreting the dominant way of seeing the growth of modern economic theory, that is as a progressive refinement of scientific knowledge. This view, assumed by Schumpeter, Samuelson and many others fails to allow the possibility that the discipline has, in part, been developing on the basis of progressively compounded errors. To investigate this possibility the study steps back into the era of classical economics and identifies a series of problems which theorists of the time faced which added up to a crisis for economics threatening its very existence as a discipline. The response to this first crisis, it is suggested, was to develop a conception of economic science which would allow the discipline to survive in the academic climate of its day. This involved establishing some kind of unassailable authority for the theory advanced as knowledge. The early neo-classical theorists tried to do this by adopting foundational theories of knowledge which would guarantee the validity of the theory they developed. These foundational positions have become the dominant orthodoxies within consumption theory over the last hundred years or so, when it has seemed as though the foundational move was a resounding success. Actually, however, it may have been a flawed academic response which has left the discipline, and especially consumption theory, with a series of internal problems. These have developed into a second crisis no less serious than the first, and like the former it has not been consciously addressed in a systematic way. Only gradually have the underlying problems become more evident and their ætiology is traced in detail in chapters two, three and four.
The first chapter, to provide a basis for telling this story, articulates in some detail what this foundational move entailed. It was not only an economic response but picked up on wider trends in academic culture which followed from similar crises of belief and knowledge in other areas. The idea of ‘foundationalism’ needs some clarification. It is a concept which has been employed by a number of philosophers of science and increasingly by post-modernists with various meanings which involve family resemblances but are not identical. The meaning employed here shows similar concerns, but has a slightly different intent. It is used to bring together and define what in retrospect turns out to be an astonishing modern preoccupation of thought seeking to establish an indubitable basis for knowledge in the human sciences. It is not limited to economics, but can also be seen as having influence in sociology, psychology, linguistics, history, politics and other social sciences. This study does no more than touch on these possibilities, but it does raise the likelihood that studies in other disciplines and subdisciplines could reveal a pattern which is similar. The first chapter identifies and defines the foundational response and examines some of the key structural problems which follow from adopting it as a basis for (economic) theory. The tenor of the critique of foundationalism can also be described. It suggests that the problem in theory formation is deeper than can be met by combining different approaches; there are certain generic problems, like the otherworldliness of theory, methodological dogmatism, the exclusivity of data, boundary disputes, and value-freedom which create insoluble dilemmas for the discipline. After exploring these possibilities in principle, the first chapter finally traces the early emergence of consumption theory in the second half of the 19th century and suggests how it was likely to be influenced by foundationalism.
Consumption theory is perhaps a favourable case. It emerged when foundationalism, as here described, was also taking shape, and its whole career has therefore been shaped within this kind of search for economic knowledge. In other areas of economics like production theory or international trade the effect might have been less decisive because of the earlier crystallisation of theoretical issues. To some extent this study remains agnostic about the weight of the influence of foundationalism in other disciplines, in wider economic theory and even in consumption theory. By focussing on the developments in consumption theory which are more obviously foundational, it tends to give less substantive treatment to other streams of thought which are not epistemologically based. In so doing it “conforms” to orthodoxy while offering a critique of it. Post-Keynesians and others have sought to study an alternative tradition of theorists who have different perspectives to offer. This task is not undertaken here; the emphasis is rather on uncovering the fundamental epistemological mistakes to see what can be learned through them. Although they may be particularly relevant to consumption theory, others working in the epistemology of economics are convinced that the extent to which theories of knowledge have shaped the development of the discipline is still widely underestimated. (Carabelli 1988 1-13) This study merely suggests foundational theories have had a debilitating effect on consumption theory, but the ramifications may be much wider.
The foundational move did not, however, result in one pattern of theoretical development. There were actually a variety of responses which theorists believed provided an infallible foundation for knowledge, and the next three chapters look at the main traditions exemplifying these responses – rationalism, positivism and causal-behavioural views of knowledge. Throughout each of these chapters a particular argument is mounted. The need for epistemological certainty is seen as pushing theorists to pursue their kind of foundationalism, often drawing on philosophers or philosophers of science who were available in their academic culture. This in turn shaped the kind of theory seen as necessary and valid, and each of these views of theory was then held dogmatically over against the other kinds which were also around. Thus, a picture emerges of a subdiscipline which fractured into more and more theoretical empires, each of which owed allegiance to its own foundation. Yet because the fundamental epistemological dogmatism of each of these positions was not recognized or admitted, the possibility of meaningful debate among consumption theorists was continually curtailed. Evidence is presented not just of noncomprehension across major epistemological traditions, but also between methodologies which seemed to be close neighbours.
However, the concern is not only the fragmentation of theoretical debate, but also the impoverished ability of foundationally-based theory to address its subject matter. Each prescribed theory excluded evidence, subject matter, modes of arguing and public issues and thereby reduced its kind of theory to a form which was empty of much significant content. The theory embodied its foundational weaknesses. Each of the positions, because it was self-referencing, cut itself off from a full engagement with the subject matter and even distorted the issues on which it did focus. The picture emerges, therefore, of a subdiscipline which has lost its usefulness. Indeed, within the literature we face a sharp fall off in relevant work alongside an expansion in the significance of consumption in economic life. It is, for example, astonishing that only 0.5% of members of the American Economic Association rate consumption theory as a major area of specialisation. (AER Dec 1989 573-5) Yet the analysis of these chapters is not all critique, for the detailed examination of particular theorists and positions allows a map to be drawn of some of the unaddressed areas and ignored modes of theorizing which can be developed in the final chapter. These are also available from some of the theories and traditions which have stayed outside a foundational approach. Thus, the analysis suggests, in varying degrees, that a radical deconstruction of consumption theory is needed, and in the final chapter a reconstruction is attempted using the insights gathered from the previous analysis.
This final chapter makes more explicit a christian critique of the idea of autonomous theoretical knowledge and shows a way of reintegrating theory with the wider issues of consumption. It picks up on some of the post-modernist consumption theorists to construct a different map of the domain and uses an alternative theory of knowledge to develop new ways of understanding. It acknowledges important contributions which are nonfoundational and give substantial insights to the discipline. It also recognizes the validity of epistemological issues in theory construction and honours those concerns as part of the growth of consumption theory, and above all it seeks in the final chapter to make some systematic suggestions for the reconstitution of this subdiscipline. When consumption seems to be the dominant concern and issue of contemporary economic life, this task is desperately needed.
The gulf between foundational consumption theory and substantive analysis is addressed in terms of a christian critique which touches and reunites them on a surer view of knowledge. It develops and exemplifies substantive areas which the consumption theorist should be able to investigate and harvest with appropriate methods of investigation and analysis. It especially suggests that post-foundationalist consumption theory needs to engage with the substantive and urgent issues raised by consumption – energy depletion, ecological damage, addiction, emulation and self-gratification to name but a few. It needs also to respond more directly to the way consumers live, both in gathering evidence from them and addressing the meanings which they give to it. Thus the study is not merely asking for a reaction to foundationalism into what happened before, or into a post-modernism which deconstructs the need for economic theory into various conflicting narratives. The aim is not reaction, but a reconstruction of consumption theory on a basis which addresses both the substantive and epistemological issues in an integrated way. This is a deep transition, because it involves moving understanding from an epistemological focus into relation with the lives of consumers. Yet without it modern consumption theory will largely remain, as it has in many respects become, defunct. With a fundamental reassessment, and with a new perspective, it is possible for consumption theory to address some of the most important issues of our lives with rigour and relevance.

Chapter one: Foundationalism and Consumption Theory.
Theory as a Path.
We take for granted the progress of economic science, but what if the discipline made some crucial mistakes in its early stages and is perpetuating and magnifying the errors? How would it know? The business of retracing to the wrong turning would be a long journey during which the landscape would be reinterpreted, and only at the end would it be possible to confidently break out in a new direction. This book makes such a journey, but because it is a book, it can take liberties with which bits of the journey it does first. This chapter goes straight back to the initial turning and tries to chart why it was the wrong direction and what induced us to take it. Like the Irishman asking the way, it says we should not start from here.
The journey is the development of consumption theory and it has taken something more than a hundred years. The following chapters will suggest there are serious flaws in the way consumption theory has been approached within the major orthodox traditions which cannot be rectified by technical adjustments. Rather it requires a reconsideration of the whole historical development of this subdiscipline, and a rethinking of one of the basic questions of its development, which is described in this study as Foundationalism. The term, as used here, has a different but related meaning to the other uses it has been given. It signifies a way of approaching knowledge which requires theory to be constructed according to principles which guarantee its certainty and validity. It has not resulted in one body of theory, because there have been varieties of foundational views of knowledge which have channelled theoretical work, including consumption theory, in their own particular directions.
This kind of exercise can only be undertaken in stages, and before we examine in more detail what Foundationalism is and how it has influenced economic theory, especially consumption theory, it may be valuable to go back to the second half of the 19th century to examine the preconditions for this journey and try to see why the direction which is now taken so much for granted was the chosen path. There are three levels of this sitzproblem which we shall examine.
We begin by considering a crisis which developed within classical economics after the middle of the 19th century. What had been taken as a stable orthodoxy within the discipline for much of the previous era became questionable and fragmented. This generated a need, when consumption theory was embroynic, for a new way of doing economics. It was the ferment of this revolution which was part of the context of our concern.
Second, the situation also changed as the practioners of economics moved from a business, church and public affairs milieu to an academic one. This change in social context tended to demand a different kind of validation in relation to peers, in particular the need to have recognition of economic science alongside the other scientific disciplines. Thus, behind the early development of consumption theory lies the emergence of this new kind of academic economist. Normally this change is taken for granted, but here it is argued that it has had more profound and less beneficial consequences than are normally assumed.
Third, interwoven with these, was the question of the development of the philosophy of the special sciences. Previously philosophy had tended to be seen as the substantive core within which specialist thought developed. Social, political, economic and natural philosophy were tendencies within a broad cultural movement with antecedents in classical philosophy and an established tradition. After the middle of the 19th century this tradition of social and economic philosophizing became confused and inconsistent; it was out of touch and dogmatic and seemed not to relate to much of the more detailed knowledge which was being gathered. Many of the newer philosophers of science looked for different ways of approaching the issue. This revolution in approach is central to the emergence of Foundationalism and had a far wider cultural significance.
These streams helped to create the turbulence out of which the response of Foundationalism developed and in which consumption theory took shape. Still, however, the actual shape of this response needs to be examined in more detail. It converged on the difficulty of holding any believe about the economy or society with certainty. A widespread response was to be agnostic about the substantive beliefs which contemporaries held and to seek surety in a certain mode of understanding. Foundationalism was the response to this epistemological agenda. It was by no means the inevitable path and was opposed by many more at the time than have been recognized in many standard histories of economic thought. The path which struck across new country is now worn with generations of docile tourists. It is so much part of the scenery of the human sciences in the 20th century that it will take some time to escape from it. As part of this process of disengagement from these preconceptions we consider some major changes in the ways of approaching the question of how knowledge is to be understood.
We are then in a position to look at Foundationalism as a key movement in the definition of modern disciplinary knowledge and to see something of its pervasive influence within consumption theory. Its internal problems as a theory of knowledge are examined with the help of two philosophers, Nelson and Dooyeweerd, and the structure of its weaknesses are examined in systematic terms. In three subsequent chapters the articulation of this misdirected theoretical approach within consumption theory is explored in more detail.

The Earlier Crisis in Classical Economics.
The development of foundationalist economic theory did not arise from an automatic progression of thought, but was conceived in reaction to problems which occurred in mid 19th century Classical Economics – what could be called the First Crisis. It is to this period we must therefore initially turn. Earlier, despite the many different positions and conclusions within Classical Economics, it had a coherence which was widely acknowledged and identified. A self-confident vision grew out of the Enlightenment which was formative in the unfolding of economics, biology, sociology, psychology, politics and other disciplines. It involved facing Nature, and grasping by reason and representation its modus operandi. Confronting Nature was the motive of Enlightenment thought, and derivatively late 18th century British thinking was continually asking what the nature of the state, human understanding, man, society, the universe and law was. (Hazard 1946, Gay 1969 II 126-87) The driving faith was that, as for Newton, Nature would give up its secrets to triumphant human understanding. Automatically, therefore, the focus was on understanding the Nature of the economy. (Clark 1992) The first approximation was a vision of a natural system which produced wealth. (Appleby 1978 242-79) Part of the genius of Smithian economics was the detachment which resulted from stepping back and looking at the economy as a mechanism exhibiting a natural order. The Physiocrats had a similar naturalistic conception, and although debate and variation occurred in the later developments of classical theory, still there remained a belief in a system which ran itself according to certain natural laws to be understood by rational reflection. This optimistic hope for the future of economic reasoning ran strongly through the earlier part of the 19th century.
This study will not examine the internal debates of Classical Economics of these years but will identify a pattern of breakdown which occurred after its statement in Mill’s Principles, when its hegemony, especially in English-speaking thought, was nearly complete. By the 1860s the dominant Classical model of the economy began to encounter severe problems, and by the centenary of the publication of the Wealth of Nations in 1876 there was concern about why the standing of the discipline had fallen so much in public estimation and was viewed with suspicion. (Smyth 1962 41-72, Political Economy Club 1876) Jevons and Ingram were markedly hostile towards the orthodoxy, but this unease was shared by a much wider fraternity. The economic expansion and euphoria of the period from 1850-1872 had given way to agricultural depression, unemployment, money and trade problems which led people to ask more searching questions of the economic orthodoxy. Classical theory became internally weak, patching itself up against a series of attacks which wounded its assumed natural provenance. What had brought about this change?
The criticisms came from a number of directions. The first was the Comtean argument that Sociology, or more accurately, an integrated study of Humanity, was the crowning science under which economics was to be subsumed. Comte’s work was available in English from the mid-fifties onwards, and he had corresponded with Mill from 1844. If economics was part of a larger framework, it had to be prepared to adapt its categories to the greater science. Mill resisted this possibility, but it appealed to others. Perhaps the most influential economist was Ingram, who was later important for the American Institutionalist School. In his critique it was clear that the Positivist creed of an organic and systemic society was seen as superceding the System of Natural Liberty of the classical model. (Ingram 1888 53-190) Ingram emphasised the mutuality and trust reflected in a complex economy over against the classical emphasis on the efficacy of self-interest, a point made in different ways by Harrison, Spencer, Carlisle, Ruskin and Cliffe Leslie. This theme threatened the autonomy of economics which rested on self-interest as the self-subsistent economic motive, especially as nationalism provided a potent form of integration for economics.
The second critique had gathered round the free-trade/protection issue. The classical system’s main metaphysical assumption was a natural economic order, of which (since the time of Smith) free trade was a part. The arguments of Hamilton, Raymond, List, Carey and others that protection might be a better policy at certain historical stages was not only a challenge to British national policy, but also to the presumed universality of the classical free trade system. (List 1841, Pribam 1983 206-7) By the mid 1850s List had been translated and was popular in the States and his ideas gathered enough weight to challenge the classical orthodoxy, especially as soon as agricultural protection became an issue, adding to the powerful German non-acceptance of free-trade.
Another key element of the classical tradition, indeed, its very definitional basis, was the concept, “wealth”. The tradition saw the discipline in terms of “The Nature and Causes of the Wealth of Nations”. Mill kept this concern central to the Principles, and defined wealth as things which possess exchangeable value. (Mill 1848 9, 436-40, but Hollander 1985 I 251-7) Ruskin seized on this definition and shredded it in Unto This Last. (1862 38-66) He showed that an important part of wealth is command over the labour of others, because things are useless without it. The supposed natural order of things therefore gave way to the question of whether wealth was fairly or unjustly acquired and used. Ruskin showed that this “wealth” could be malignant when it involved the power to control other people’s lives and needed to be evaluated by more fundamental criteria than liberalism and utilitarianism could offer. Further, wealth was not mere possession, but required that articles be used, which in turn involved the question of who could use them and to what purpose. Thus Ruskin linked the theory of production and value indissolubly with distribution and the central purposes of humanity. The key concept of the classical tradition was left in some disarray, and Munera Pulvis and Fors Clavigera kept Ruskin’s critique alive throughout the 1870s, although by then Mill had already admitted the limitations of his definition of “material wealth”, and espoused a mild, pre-marxist, ideal Socialism. (Mill 1909 752-94, Harrison 1902 91-108)
A further plank in the classical system was the Ricardian wages fund theory, as developed by Senior, Longfield and Mill. This also collapsed. Thornton in On Labour had broken down the idea that there is some inherent distinction between funds for profits and for wages, which had been used to drive distributional issues from the classical model. He showed that because employers paid below the price they would be prepared to pay for labour when supply exceeded demand, there had to be a range of possible distributional effects. (Thornton 1868 43-110) Mill’s retraction brings out the framework within which the old model worked.
There is no law of nature making it inherently possible for wages to rise to the point of absorbing not only the funds which he had intended to devote to carrying on his business, but the whole of what he allows for his private expenses beyond the necessities of life. The real limit of the rise is the practical consideration of how much would ruin him or drive him to abandon the business; not the inexorable limits of the wages fund. (Mill 1909 993 [1869])
Marx pushed this point further by arguing for an exploitative distributional effect which led to the wrong appropriation of wages, and there was no way back to Ricardian automaticity.
The failure of the natural order paradigm was also evident in its lack of a significant theory of choice. Senior and Lloyd made initial statements of the concept of diminishing marginal utility, but they did not draw the choice conclusions which came later with marginal theory. When the dominant political ideology was Liberalism, an economic model which gave choice or freedom so little scope was out of touch. Mill, who had no shortage of Liberal credentials, allowed utility only a temporary role within the more determinate concept of Natural Value. Mill, strong on Liberty, could not break through his naturalistic concept of “economic man” and “real laws of nature, dependent on the properties of subjects” (Mill 1873 234) to a more explicit recognition of economic freedom of choice. The failure of consumption theory to develop within the classical paradigm was a symptom of this problem which will be examined more fully later.
A further attack on natural order came from those who refused to see labour as merely a factor of production subject to ineluctable laws. They included the Christian Socialist Movement, the Chartist movement and the Evangelical philanthropists. The natural laissez-faire order did not work, argued members of this tradition, because it produced so many human casualties and degraded people. On moral grounds it was an unacceptable and inhumane economic philosophy. Maurice argued economic activity came within the scope of the Kingdom of God, rather than being a “natural” activity, Kingsley that rental income was often unjust, Shaftesbury that the terms of labouring could and should be redefined by law. This principle even applied to donkeys; as the costermongers learned, “With twenty-four hours’ rest on Sunday, they would do thirty miles a day without exhaustion; whereas without it, they do not do an average of more than fifteen.” (Hodder 1887 647) Although this critique was not fully developed theoretically, it had a pervasive influence and required a justification outside naturalistic explanations. By the 1860s the social improvements of the working classes, aided by legislation and social action, was shown to have improved the contribution of the workers; the French envied the level of art education evidenced in British pottery and fabrics. (Ludlow 1867) A commitment to workers and the quality of work, evidenced also in Guild Socialism, replaced the fatalistic views of subsistence labour present in Malthus and others.
Slightly later appeared the argument of the German Historical School that all economic forms were historically conditioned and shaped by the era and area in which they developed. The classical analysis of the economy was thus culturally relativised. What could be the correct policy for the British nation might not be for the German. Cliffe Leslie was the British economist who most strongly represented this position (1970 and 1888) and Mill was happy to some extent to go along with this relativism on the Irish Land reform question. (Hollander 1985 II 922-8) During the 1870s and 80s it gained more weight through the work of Ashley and challenged the classical idea of one natural model for the economy and its supposed universal conclusions. There were already enough differences in European economies for the point to be forcefully made, and its impact can be seen in the framework which Marshall largely adopted in Industry and Trade. (1919)
Yet another attack had its origins in the Common Sense School. Price, for example, questioned the superiority of the “scientific” ideas of economists over the common sense practicality of businessmen. In mounting this attack Price was undermining the classical economists’ authority to make superior statements about what was happening in the economy. He showed that many of the arguments were circular and many of the definitions suspect. (Price 1882 1-31) The economic version of the Common Sense school continued the belief in the superiority of businessmen over economists, and the inferiority of economic reasoning to the understanding of the Manchester men.
A more weighty issue was the change in the relationship between economics and politics. The minimalist State view of late mid-century Liberalism had left the classical economy largely intact, but the nationalist state involved a different view of economic relationships, whether in Germany, Britain, Italy or the United States. Later when idealist philosophy came to dominate British political thought in the work of Green, Maine, Maitland, Bradley and Bosanquet, the conception of the State tended to claim priority in the organisation of economic affairs. The Socialists, Trade Unionists and Fabians similarly saw political issues as a prior concern in the economic debate. Consequently, the discipline faced a clear choice, or so it seemed. Either it must break with the political element of its nomenclature and establish itself as independent of political theory, or some overall philosophy which articulated politics and economics as a total framework had to be established. This was reflected in the increasing influence of Fabian and ILP Socialist economics on the one hand and economists who asserted their purity, but not in terms of laissez-faire naturalism, on the other. By the turn of the century it was accepted that the polity necessarily influenced the economy, even if only by default, so that laissez-faire was a political choice. Those who wanted to maintain the purity of economics therefore had to find an extra-political ground for doing so.
An even deeper challenge was the ideological one. At the middle of the century the perceived issues had been the Anti-Corn Law League, the right to unionisation, commercial crises, the response to the Irish famine and the protection of the British farmer. With the elapse of another twenty or thirty years far deeper ideological differences were being expressed among the articulate public. Christian Socialists questioned industrial organisation. The Co-operative movement attacked orthodox marketing. Marxists attacked capitalism. Disraeli Toryism attacked liberal laissez faire policies on trade. Imperialism promoted a preferential alternative to free trade. Education and welfare developments undermined political laissez faire. These were not just differences of opinion but of conscience and principle. The commitments challenged at root the benign idea of a natural understanding rooted in the middle classes; the prescient knew that another basis for human knowledge in these areas had to be established which would hold against all ideological attack, especially for political economy.
The awareness of these attacks and criticisms grew piecemeal. The kind of unease expressed at the centenary of The Wealth of Nations poorly articulated only some of these points. But gradually the perception grew that the conception of the economy as a natural order which could be studied by detailing its anatomy and understanding how the components of the system worked was no longer tenable. No given natural order was discernible and no one rational understanding of it became evident. Although the consciousness of the situation developed very slowly, the end had come for the Enlightenment epistemological response which characterized the work of Smith and Ricardo. Their aim had been to examine the nature of the economy, which was seen in an analogous way to the solar system, through rational understanding uncovering the order of a mechanical causal system. The hope that this mode of understanding would issue in orthodoxy was now dead, because there was now such fundamental disagreement about the nature of the economy. The Historicist, the Socialist and the Laissez-faire Liberal each claimed to have rationally uncovered the nature of the economy, but with contradictory and irreconcilable claims as to what it was. Each and none was rational in the Enlightenment sense and the basic orientation of Classical understanding was in crisis.
If this analysis is correct, it suggests that the Neoclassical interpretation of the later response is less than accurate and rather disingenuous. Schumpeter and others focussed on the Marginal Revolution, happening more or less simultaneously in the Austrian, English and French/Lausanne schools, as providing a new development in Classical economics which has continued more or less uninterrupted to the present. (Schumpeter 1954 825-9, 909-20) Yet the marginal revolution addressed so few of the criticisms levelled at the classical system. The challenge was far bigger and deeper, requiring academic economists to move away from an ideational conception of the nature of the economy towards an epistemological consideration of how economic understanding is obtained. Mill, Jevons, Pareto, Walras, Weber and Marshall all had important early concerns with methodology and the theory of knowledge. Marginalism, it will be argued, was therefore part of this larger epistemological transition which attempted to remedy the deficiencies of the classical approach and establish a new orthodoxy.
The extent of its assumed success can be judged by the way the work of many of the economists who did not adopt an epistemological focus was sidelined in the “great tradition” of the history of economic thought. The contemporary position of Ruskin, Morris, the Webbs, Wallas, Beveridge, Hawtrey, Pigou, Cole, McFie, Tawney, Hobson, Veblen, Commons, Mitchell, Myrdal, Eucken, Hayek, Tugwell, Knight, Rueff and many others was far greater than the reduced significance they have been given in Neoclassical histories of economic thought. Sadly, this study cannot pursue the public, political and economic significance of these thinkers and the alternative traditions which they generated. We merely note the editing process has understated the extent to which neoclassical economists had to fight for their definitions of economic science to prevail.
In summary, the sitzproblem demanded a fundamental transition to a different way of approaching the issue: if the architecture of the discipline could no longer be constructed in terms of natural laws, there had to be a reliable way of establishing economic knowledge which could be defended against the charge of ideological bias, partiality and subjectivity. Crucial were the criteria for judging what well-formed or scientific understanding was. The groundwork for this approach had already been done by growing traditions of theorists, but now it was a movement come of age. If a certain kind of knowledge could be held definitively, or foundationally, correct then economic understanding was well formed and could allow a cumulative growth of economic knowledge. This shift allowed a relaxation of the need to have a view of the economy as a whole in favour of systemic agnosticism. However, it required confidence in the method of obtaining knowledge before one gathered the fruits of that method. This approach in principle therefore allowed a break from the dogmatic classical belief that the inner nature of the economy had been infallibly grasped by the mind; it allowed the failure of classical economics to be tenable, but it replaced it with the hope that correct methods of constructing theory would generate a new orthodoxy which would be beyond challenge.

The Epistemic Community of Later 19th Century Economists.
At the same time as Classical Economics was undergoing this crisis another transition was taking place which had its effect on the development of the discipline. The community of economists moved from a focus in public affairs to one which was academic. This change in subculture deeply affected the way economics was conceived, yet because the subsequent community of economic scholars has been largely rooted in academe the effects of this changed cultural milieu have scarcely been recognized.
This is partly because of a commitment to the idea of the purity of theoretical thought; ideas are to be examined in and of themselves, uncorrupted by the consideration of social, psychological, economic or political motives, which cast a slur on their integrity as ideas. We are examining an era when this idea of purity was given strong assent in the form of the assumed rational orthodoxy of neoclassical economics. Yet this approach fails to recognize the subcultural limitations of the views which are defined by the groups concerned as universal and necessary ones. The preoccupations and priorities of all groups are limited and focussed, and the subculture out of which those concerns grow is therefore an important aid to understanding what was occurring which the sociology of knowledge provides. Such an approach does not require a reductionist attitude towards the ideas involved; it merely enlarges the scope of the truth criteria to which they are subject and may explain their limitations and biasses.
There is a tendency for the history of economic thought to be conceived in terms of a progressive academic march towards the present. The critical tools of this position are those which belong to the present, which of course cannot criticize the growth of proto-orthodoxy; it detects a “pure filiation of ideas”. (Blaug 1985 300) It also assumes that ideas are sifted purely on the grounds of their present “truth” value, rather than interest, self-validation and patterns of group loyalty and cohesion. If the present is wrong in some sense, this orientation is especially weak. Yet the idea of the progress of economic thought remains tenaciously strong. Samuelson has seen the recent past as progress to a golden age in the development of analytical economics with yet new frontiers of knowledge to conquer, (Samuelson 1965 xiii) and many other theorists automatically adopt this stance. Because economics is now strongly rooted in an academic community, it becomes difficult with this view to see its social location as problematic. In this section we shall consider how changes in the socio-cultural context of economists put pressures on them which could not longer be solved by the old liberal patterns of debate. Their response of fitting in with the academic ethos of similar disciplinary groups is by no means the automatic route to progress in economic thought which it is often supposed to be. Indeed, it may have been, and be now, deeply problematic for the discipline.
In the sociology of knowledge there is an awareness of the located and subcultural attitudes which have generated claims to science, knowledge and theoretical orthodoxy. Science in one sense is a cultural idea promulgated by various interest groups who may use it as a means of enhancing their own status, pay and authority. (Mulkay 1979, Wallis 1979 49-66) The importance of the group in deciding what issues receive currency, priority and acceptance is also considerable. Orthodoxy is a powerful social force in science; it shortcircuits other forms of examination by arguing that of course this view must be accepted. The groups who espoused the idea of economic science had their own agenda and occurred in a situation where many other views had as strong a claim to orthodoxy. By identifying the epistemic communities in historical context it is possible to see the competition of ideas, reference groups, interests and claims which were present in the generation of new lines of thought. This need not discount belief, or veracity, but it can reveal special pleading. (Code 1987 166-97, Barnes 1974, 1977, Weiller and Dupuigrenet-Désroussilles 1974 40-47) Finally, the different levels at which epistemic communities operate need to be recognised; changes in theory, methods, paradigm, epistemology, communication and recruitment can be quite dissimilar. (Butts and Hintikka 1977 31-49) What, then, were the 19th century economists like as a community?
The Public Affairs Economists.
At first economists in Britain were a small group without an epistemic focus. The Political Economy Club, founded in 1821 was limited to 35 members, many of whom were passive and only 4 of whom were ex officio academics. (Political Economy Club 1 1860) A laissez-faire, free trade approach meant economists were superfluous as advisors, planners and executives; if the economy ran itself, they were redundant. James Mill and the other Board of Trade Free Traders had done their kind out of a job by their doctrines. Thus the Civil Service was no strong source of employment and usefulness for economists until the Imperial emphasis grew; even then Keynes found it full of dotardry. (Barber 1975, Skidelsky 1986 178) The result was to leave economics as a part-time, even leisure occupation for many members of the classical tradition and to shift the emphasis to the practitioners of economic life. The major full-time occupations of these economists were Parliament, Banking and Commerce, Country Gentleman, Merchant and Manufacturer, the Civil Service and the Church. (Fetter 1980 9-13, 243-59) Even J.S.Mill, dominating mid-century economics, was not a government advisor and was heavily preoccupied between 1851 and 1858 with his East India Company work. Of those who could be called professional economists, Cairnes and M’Culloch depended on chairs endowed by Whately and Ricardo, and Malthus held a Civil Service teaching post. Senior at Oxford was, like Cairnes and M’Culloch, an avid supporter of the scientific neutrality of economics, which suggests that the professionals were already more concerned about the scientific status of their discipline than those who merely traded opinions.
This was further reflected in the journals and forums of debate that economists used. The main places for economic discussion were journals like The Edinburgh Review, The Fortnightly Review, The London and Westminster Review, Blackwoods, The Nineteenth Century, The Journal of the Royal Statistical Society and other, more popular magazines and pamphlets. Until the Economic Journal was established in 1890 the locus of debate was among the literate public, not specifically among economists. The patterns of contact were similarly diffuse. No university had a community of economic students and teachers before 1870. Jevons accounts of his early experience in London show that such a community did not exist. Walras and Pareto evidence a similar isolation. Marshall created such a community through teaching generations of students. The accounts of the early meetings of the Royal Economic Society show how slowly and painfully the community developed. The early pattern therefore was of isolated scholars in contact through letters, articles and book reviews, but in a context of wider social intercourse. Many of the forms of contact – Parliament, Committees and Commissions, Banking and Commercial gatherings, involved a more open pattern of discourse. In addition few of the participants in the debate depended in any way financially on their work as economists. In this situation they could disagree without it being of much moment to them professionally.
Economic issues were joined by scholars of other disciplines. Some of the most significant economic debate around 1850 came from the stimulus of Comte and Spencer (Sociology), Carlyle, Ruskin and Morris (Aesthetics), Maurice, Kingsley, Whately and Chalmers (Theology) and Sidgwick (Ethics). All of these were able to invade the domain of economics quite substantially, and expected to do so. Mill fended off Comte, but Ruskin examined Mill’s technical terms with far more penetration. In a similar way Croce (Philosophy and Aesthetics) engaged in a detailed discussion with Pareto on his theory of value. (Int. Ec Papers No 3 172-207) Maurice was encouraged to apply for the Oxford chair, but in the end did not, and Sidgwick moved easily into the domain of economics. Thus, throughout this period the surrounding disciplines, some established and some struggling for identity, refused to accept the boundary which more academic economists tried to draw round their discipline and readily invaded its territory.
The boundaries were loosely and differently drawn, as we have seen, but that was because economists were unclear as to their court of reference. Was it the public, business, merchants, bankers, students, academics, scientists, government or the church? Manchester School industrialists, merchants and bankers knew their job better than economists and seemed to have no direct need of theorists, who obviously had an ambiguous role. Walras tried his hand at being a political advisor, but failed. Jevons was thrilled at being noticed by Gladstone over the Coal Question. (1865) By the late 1870s economists were expected to explain to the public what had gone wrong with the British economy and began to develop some kind of professional identity. But it was tenuous; in the 1880s public economic debate was being conducted along lines drawn up by the Socialists, (Pelling 1965 13-61) and by 1886 Beatrice Webb had articulated objections to a “self-contained, separate, abstract political economy”. (Webb 1926 290, 437-44) In good time Marshall would make his worthy appearances before commissions, but the public and the government had a limited conception of the value and expertise of this group, especially as the old middle-class organs of debate declined with more populist politics. Socialism had effectively sidelined neoclassical economic contributions, as Pareto, Marshall and Walras frequently complained from differing contexts. They were ignored and seemed to speak without public authority.
The Search for Academic Recognition.
Nor did the economists fare much better with their fellow academics. Inclusion in the scientific community was difficult. In 1877 Galton proposed the abolition of Section F (Economic Science and Statistics) of the British Association for the Advancement of Science because the scientific content of the papers was so small. (Stone EJ 1980 720) In 1863 Owens College, Manchester could scarcely offer an economist a viable job, and the situation changed only slowly as the provincial universities began to take up economics. At Cambridge even in 1902 Marshall wrote to Neville Keynes as part of his agitation to set up an Economics and Politics Tripos which would not be swamped by the Moral Sciences Board:
Put yourself in my position. I am an old man…. I have no time to wait. Economics is drifting under the control of people like Sidney Webb and Arthur Chamberlain. [LSE and Birmingham School of Commerce] And all the while, through causes for which no-one is in the main responsible, the curriculum to which I am officially attached has not provided me with one single high class man devoting himself to economics during the sixteen years of my Professorship (quoted in Skidelsky 1986 45)
Thus the struggle to establish economics as an autonomous discipline with stature was a difficult one which daunted even the greatest figures until into the 20th century. (Hutchinson 1953 1-31)
The triumphalist history of the progress of economic thought which comes from some neo-classical scholars ignores how crucial this issue of establishing a domain was for the new professionals. For unlike the gentry, bankers, MPs and merchants, these new economists had to earn their living by writing and teaching economics. Walras searched for ten years for an economics professorship in France between 1860 and 1870, during which time he was a journalist, administrator and bank employee, and then he had to go to Lausanne to find a post. (Jaffé 1983 125) The practitioners required an area of expertise and authority which they could claim as secure professional knowledge, and this meant different points of reference for the development of the discipline than had obtained hitherto. They had to throw their lot in with the university as an epistemic community and break, to some extent, the old ties with banking, politics and trade. Thus the autonomous bounded discipline which tended to emerge may have done so in part because of the professional problems which the early economists faced.
The conception of the university varied. The personal authority and following which professors in Germany and Austria commanded tended to encourage the kind of schools of thought exemplified in the Methodenstreit, so that not being swayed by student popularity was an issue of conscience for Weber. (Weber 1948 129-56) Oxford and Cambridge faced less direct English language competition, and this kind of warfare was not marked. The governing bodies tended to be establishments of church, local government and older academics who exercised influence, controlled appointments and wanted to further their institution. (McAlister 1976 10-47) Disciplines vied for status, staff and funds. Academic debates tended to take place in their own terms through processes of acclaim and acceptance which were independent of the daily issues of political economy. This retreat from engagement with money, trade, banking, commerce, government data gathering and policy formation was considerable, because the daily agenda presented by lectures, tutorials and articles was the validation of one’s academic position. Government was left to professional and popular politicians, trade to businessmen and detailed knowledge of banking was lost to many of the new economists; Maynard Keynes could find no treatise in any language dealing with money in his modern world. (Keynes 1930 I vii) The distance which economics travelled in moving from London to Cambridge was far more than 60 miles, and the same kind of journey was being undertaken by the discipline elsewhere as academic economics developed.
The change in priority which this move generated was considerable. Economists moved from being concerned with public affairs and the conduct of business, work and banking to establishing a corpus of knowledge which could be taught with authority. Immediately, it is clear that the new generation found themselves alienated from the old political economy; it could not work for them, partly because they did not have access to the responsibilities, decision-making and informal contacts through which the system operated. They could easily feel they needed a new kind of knowledge which gave them authority and status in academe. Again the sitzproblem for those who need to earn a living from economics is evident. They needed an established professional knowledge which would command respect and income, and this required a way of validating the status of this knowledge against all comers.
Yet, of course, these socio-economic needs neither validate nor invalidate the new kinds of discourse, debate and theory formation which were generated by the academic context of economic thinking. The central question is how the new academic community of economists redefined theory and whether their way of securing it against attack is justifiable. Nor does the fact that the old way was defended by such weak protagonists as Bonamy Price (1882) automatically validate the new view. The creation of a separate epistemic community for academic economists may have been one of the biggest problems of the emerging subdiscipline of consumption theory. Behind all these issues is the bigger problem of how the wider academic epistemic community defined knowledge and whether the widespread claims to the scientific status of various disciplines were justified.

The New Human Sciences: Their Definition and Coherence.
A third issue which was current at this time was the formation of the newly emerging human sciences. Sociology, Psychology, Aesthetics, Pedagogy and other disciplines were just forging a distinct identity. Some, like History, Philosophy and Ethics had a longer pedigree, but were challenged by the changes brought about by new kinds of knowledge. At the same time the relationship among these different bodies of knowledge was ambiguous: were they part of a larger corpus, or autonomous or hierarchical? Again, they all faced the relationship which they had with the natural sciences; were they geisteswissenschaften, arts subjects or social science? Each of these disciplines therefore faced three major challenges: their identity, their relationship to the others and their status as science. Economics, as a relatively established discipline, yet with questions beside its identity, had its own path through these issues.
As “Political Economy”, Economics had its identity within public affairs, but gradually, except at Oxford, this meaning waned with the 19th century. It also had links with Theology, History, Geography, Ethics, Sociology, Psychology, Anthropology and Linguistics, and was engaged in debates with many of them as to the relationship between them. This issue was really a philosophical one, because it contained questions which went beyond the competence of the discipline itself. The argument of this section is that by the end of the century the philosophical problems posed by these relationships had become seemingly insoluble; the impasse required another kind of solution. Such a solution was offered by Foundationalism.
During the 18th century Enlightenment humanism had promised that a direct confrontation of Reason with the Nature of man, society, morals, the state, the economy, the natural world and even of the mind would bring unequivocal understanding free from the trammels of revelation. Underlying this movement was the kind of hope which had generated L’Encyclopédie, namely that a comprehensive rational understanding of humankind would emerge. The universe of knowledge could be brought together in one coherent corpus or place, the university. Although a shadowy idea, for a while it seemed possible; there was a mode of thinking which emerged from the landed, leisured aristocracy which expressed general principles of rights, contract, freedom and self-interest as a basis for interpreting areas of human life, but by the time of the Revolutionary Wars these had disintegrated into conservative, nationalist, whig and radical modes of thought which were deeply at odds with one another. When the perspectives of romanticism and the evangelical revivals were added, there was no possibility of consensus about humankind. There were conservative historians, economists, political theorists and sociologists, but conservatism could not hold these disciplines together because it was a divisive view of humankind around which wars had taken place.
After this stage a number of systematic thinkers emerged who were committed to a comprehensive vision of human study, often articulated from one area onto a more general canvas. Hegel, Bentham, Owen, Comte, Spencer and Marx all had this kind of vision, but they were resisted by many of the practioners of each of the new “social” sciences. Ricardo ignored Bentham, Mill backed away from Comte, Marx overturned Hegel and everybody resisted Marx. The problem is exemplified in Comte, who while claiming to have developed the crowning positive science of sociology within a general religion of humanity, failed to engage economists and historians because of the sweeping claims made about their areas of study. (Comte 1875-7 557-9 [1822], Ashley in Mill 1909 xi-xxiv, Lenzer 1975 Introduction.) Similarly, outside the historical school no economic author gave much attention to Hegel’s economic views expressed in his Rechtsphilosophie with the possible exception of Marshall. (but see Schumpeter 1954 780) Each perspective tended to view the other sciences from a vision which grew out of its own founding matrix: the State for Hegel, Society for Comte and Production for Marx. The competitive pervasiveness of each of these positions is well known, and in the end their extravagant claims foundered on one another.
Perhaps the most dominant of all these comprehensive visions grew in the German Historical School. The formative thinking of Herder in Auch eine Philosophie der Geschichte (1774) and Ideen zur Philosophie der Geschichte der Menschen (1784-91) initiated a pattern of German thought which focussed on historical causal explanation as basic to all areas of study. It was further developed by Hegel’s Philosophie der Geschichte and many other works. Judgements of progress and regress, evolution and regression, radicalism and conservatism were built into analysis. (Goudzwaard 1979 151-161, Manicas 1987 53-72, Bottomore and Nisbet 1979 39-117) Historical study became a way of approaching other disciplines by demanding of them the study of developmental processes as the founding meaning of those sciences. Historical stages were central in economic, social, political, geographical, anthropological and psychological thought. Origins became of the utmost significance, not just in biology, and temporal direction shaped all analysis. Retrospective essentialist causal explanations and utopian future ones were often part of historicist analysis. (Popper 1957) Historical modes of analysis were seen as methodologically central in other disciplines. Higher Criticism and the Quest for the Historical Jesus gave a decisive slant to theology. This development was close to the growth of Foundationalism, for the Historical movement developed into Historicism, a view which claimed that the methods of history were a universal method for all the disciplines of Geisteswissenschaften. These views were developed by Ranke, Dilthey, Burckhardt, Schmoller, Weber and others into perspectives which addressed the question of how historical knowledge was to be constructed and sought to apply them throughout the human sciences. This transition, from a comprehensive historical vision to a belief in historical method as basic to the human sciences, was part of the Foundationalist move which we shall be later examining, and it was, of course, the crucial component in the Methodenstreit which then followed. For our immediate purposes, however, we only have to note that Schmoller’s attempt to make the historical method of establishing temporal relationships among particulars a general mode of scholarship applicable to economics was on the whole firmly repudiated by the discipline. History also was not going to be the unifying principle of the human sciences. (Pribam 1983 209-24, Manicas 1987 216-21) Although Social Darwinism continued until Nazism, it and other forms of historicism were being resisted strongly by other disciplines from the 1870s onwards.
Nor was metaphysical philosophy generally able to provide this framework of coherence. Under the Third Republic in France philosophy was encouraged as a potential source of moral order and authority to replace Catholic teaching. It assumed a pivotal role in the educational system, from the École Normale downwards, on the assumption that rational inquiry could develop a framework of cosmic and social order around which the disciplines could develop. (Bottomore and Nisbet 1979 192-203) A similar agenda occurred in other countries of Europe. However, much philosophy was shaped by various forms of post-Hegelian metaphysics which made competing claims about the nature of reality, and thereby failed to offer any coherent base for disciplinary development. Before long the metaphysics of Renouvier, de Biran, Lachelier, Guyau and Bergson within the French tradition, Bradley, Green, Alexander, Bosanquet, McTaggart, Royce and others in the English-speaking tradition and Haeckel, Brunschvicg and Bauch were attacked as an obsolete mode of philosophy which was merely making assertive claims about the world. (Walsh 1985 352-71, 408-44, 517-72, Copleston 9II) There was also a more popular development in the work of Croce, Gentile, Langbehn, von List and others which combined romantic ideas of soul, sentiment, intuition and consciousness with the Absolute, realism and evolution to create all kinds of self-serving and quasi-fascist philosophies. Indeed, by this time the metaphysical philosophers tended to reveal their distance from and lack of grasp of the disciplinary areas which they hoped to address, and the question than arose as to which areas of knowledge philosophers could genuinely lay claim. They were no longer the source of basic principles, views of the society, state and history. Nor could they claim any kind of superior intellectual impartiality for rational reflection, since in its name all kinds of dogmatic assertions were being made.
By the end of the century the misunderstandings and conflicts generated by Philosophy and these quasi-disciplinary empires – the Historicists, the integrating Science of Sociology people, the political Idealists and the Marxists, had created enough confusion to generate their own reaction. This took the form of insisting on the sui generis form of knowledge of each discipline. It is conveyed well by James.
Most thinkers have a faith that at bottom there is but one Science of all things, and that until all is known, no one thing can be completely known. Such a science, if realized, would be Philosophy. Meanwhile it is far from being realized; and instead of it, we have a lot of beginnings of knowledge made in different places, and kept separate from each other merely for practical convenience’ sake, until with later growth they may run into one body of Truth. These provisional beginnings of learning we call “the sciences” in the plural. In order not to be unwieldy, every such science has to stick to its own arbitrarily-selected problems, and to ignore all others.. (James 1892 1)
James and other psychologists were concerned during this period with defining psychology as a science of mental states, of consciousness or experience. (Ward 1918 ch1) At the same time Durkheim was distinguishing the social and others the economic, political, historical and theological in more discrete terms. Although this approach defused some of the conflict and confusion, it tended to produce the conclusion that knowledge in each of the disciplines was sui generis and had nothing in common. As a statement about the university, and the universality of knowledge, this was extremely damaging. What, then, did hold human knowledge together?
This created a dilemma which grew more acute each year; the specialised disciplines were expanding their study base, were seeking to be more professional, were being located in academies and universities to constitute universes of knowledge, yet did not cohere in their understanding. Philosophy and Theology no longer claimed to provide the broader vision. Those whose awareness was wider than their own specialism thus faced acutely the question of whether the knowledge which claimed to be emerging from the special sciences had any coherence at all or was merely ad hoc. If they looked for coherence, they could not find it either in a philosophy of the nature of things, or in the subject matter of the particular sciences. Things fell apart; there was no centre. There had to be another kind of answer, and one possibility was to find coherence not in a substantive view of humankind, but in a common methodology of the human sciences. This Foundationalism purported to do.

Thus, after the mid century three levels of dilemma grew within economics calling for a solution. With the crisis of the Classical system there was no substantive orthodox view of the nature of economics which could claim wide assent, nor did any such view seem likely to emerge. At the same time the social locus of economic analysis was moving from public affairs to academic institutions, with the consequence that the audience for economic ideas was expecting a different kind of thinking centred on scientific analysis. Third, Philosophy as traditionally conceived did not seem able to provide the key to understanding each of the human sciences and their relationships with one another. These dilemmas converged on the issue of what kind of knowledge economists (and others) should be seeking, and it is to this broader issue that we now turn.

Locating the Problems of Knowledge.
The sitzproblem of economic theory in the late 19th century was therefore: What kind of knowledge can be treated as reliable economic science? This would now be treated as a classical epistemological issue. Yet this term is suspect in assuming a steady investment in the theory of knowledge down the centuries. It ignores the dramatic waxing and waning of epistemological concerns. At other periods, as we shall now see, alternative routes to knowledge have taken precedence. Nor has “the problem” of knowledge always been the same problem. In different eras the agenda which is addressed in epistemological terms has varied. As the title of this section suggests the question, How do we know? has been approached in substantially different ways, and it is necessary to have some grasp of the transformations in the approach to knowledge between the 18th and 20th centuries to see how different were the dilemmas addressed by the epistemological responses and the hopes they generated.
The epistemological stance is largely a modern nonchristian mode of thought. It involves trying to make the ego secure in what it understands and to establish a mode of knowing the world which is an infallible route to knowledge. This contrasts with a christian mode which identifies the ego as itself problematic, especially when it tries to establish knowledge independently of the God on whom all knowledge inevitably depends, and sees faith as a pattern of understanding with inherent limitations and patterns of fallibility. We shall explore this difference later.
The first great modern leap into this mode was made by Descartes’ attempt to doubt everything until some indubitable basis for knowing could be established through rational proof. (1637) This made scepticism and proof for the knowing ego the epistemological focus with the key hope of the human mind taming the world through indubitable chains of deductive reasoning. (33) Already some themes emerge: certainty beyond any doubt is possible for the ego; understanding is dissociated from persons and from the world and is reified into a method; agnosticism is transferred to the meaning of the subject and the world and away from the process of relating one to the other; understanding and action are dissociated; and the dialectic between the subject controlling understanding and the object giving it begins to take shape. With Descartes the controlling deductive subject and the indubitability of its conclusions is the focus.
In 17th century Britain a different view was held by Newton, Boyle and other Puritan philosophers of science. They began much more directly from faith in God and an understanding of what the creation was like as God’s handiwork. The knowing subject’s response of faith did not have the primary role which it was given within the Cartesian framework. (Merton 1966, Hall 1963 316, Thayer 1953 65-6) The later eclipse of Puritanism resulted in an alternative approach to the nature of understanding. Locke and others believed that what was known could be elucidated by mirroring the natural processes by which knowledge was acquired and imprinted on the unformed mind. (1700 bk 2) Already the contrast with addressing knowledge in terms of the deductive ego is evident; here the “objective” sense impressions are dominant, setting up the objective/subjective dialectic which is so central to humanist epistemology. (Coffey 1917 135-67, Woolhouse 1971 25-32) The conception was of complex ideas being built from simple elements generated by sense impressions which the subject received from the external world. This process described the formation of knowledge seen in a wide sense, the kind that human beings do and should gather from the cradle onwards. The concern, especially with Hume, was to replace superstitious belief with an alternative. He saw dogma or unjustified belief as that which did not properly grow out of experience and sense impressions. At this stage in Britain and throughout the rest of Europe establishing forms of knowledge which were independent of the Christian faith was important, as Hume discovered when dining with other atheists over the Channel. The apologetic orientation to the theory of knowledge was thus central and has been developed by other philosophers later, although it did not remain the focus of later epistemology. (Butler 1736, Berkeley 1713, Hume 1779)
Hume developed the theme of epistemological doubt by questioning the legitimacy of conclusions which the perceiving subject could draw from sense impressions. (1738 vol 1) It involved a radical deconstruction of much which previous rational thought had taken for granted, especially the cause-effect analysis of much materialist thought. This emphasis was a presentiment of the foundationalist emphasis on constructing an infallible route to knowledge. It not only waked Kant from his slumbers, but also transmuted rationality into logic, a “passive” mode of understanding which draws necessary conclusions completely without reference to the ego. (This is the origin of the distinction between rationalism and logicism which is basic to the later analysis in this study.) At its deepest level this brought into question the whole epistemological enterprise; the construction of a theory of knowledge seemed to issue not in an infallible route to knowledge, but a degree of scepticism which stifled the growth of knowledge. Hume asked men to be fully perswaded “That there is nothing in any object, consider’d in itself, which can afford us a reason for drawing a conclusion beyond it; and That even after the observation of the frequent or constant conjunction of objects, we have no reason to draw any inference concerning any object beyond those of which we have experience.” (Hume 1738 139) Accepting Hume’s conclusions meant abandoning all but descriptive scholarship and playing backgammon. Russell called it the “bankruptcy of 18th century reasonableness”. (Russell 1961 65) Most scholars sidestepped the issue and proceeded in other terms. Indeed, in Hume himself there is a discontinuity between his epistemological work and the History of England, the essays on National Character and the Original Contract and the Natural History of Religion. In the latter he adopts a conventional, sociological causal framework which is not consistent with his epistemological conclusions, and it is this latter framework which also fits the mode of analysis of Smith and the Scottish Enlightenment School. The waning of the epistemological mode in Britain was therefore quite rapid and complete, and the work of Hutcheson, Ferguson, Millar, Ricardo, Malthus and Bentham had less concern with the process of validating knowledge and a more optimistic belief in its direct accessibility.
There were other thinkers who challenged this way of proceeding. Berkeley, by opening up the implications of subjectivism, was able to develop a more thorough scepticism towards sense-perception and materialism. (1710,1713) The Common Sense philosophy of Reid and Hamilton called into question epistemic scepticism as the key tool in the construction of knowledge; conclusions could be valid other than those drawn from systematic doubt. These positions suggested that the theory of knowledge was neither the indubitable route forward nor the necessary prior one; the question, how do we know, did not have to be answered before knowledge was possible. (Maund 1937) Within a relatively short period the approach to knowledge through establishing an indubitable process had given way to much more affirmative views. (Clive 1960 57-73)
The emphases in German-speaking areas, primarily under the influence of Kant, was different again. Kant placed a great emphasis on the problem of metaphysics, doubting whether it is capable of giving us knowledge of reality and even calling it “rotten dogmatism”. Whereas previous philosophy had sought an ontological basis for metaphysical certainty, Kant’s focus is on how reason can a priori establish the grounds of thought. He was convinced that synthetic a priori thought provided the necessary foundation for phenomenal knowledge. It was conceived as the structured categories of experience. The organisation of these categories thus became the central epistemic issue together with a formal elucidation of the rules of thought which governed this process. Along with these went a concern for the consistency of these categories as a basis by which empirical knowledge could be organized by the thinking ego. The external world was seen less in terms of sense impressions than phenomenally. (Kant 1787) Greek and Thomist conceptions of knowledge also had more weight further East. However, what is quite astonishing is how quickly sceptical epistemology gave way also in this culture to Idealist philosophies which were working from a different point of departure. (Copleston 6II 220-9)
In Britain Humean scepticism was rapidly short-circuited by larger trends like the Romantic movement and the Evangelical Revivals. They addressed more the question, “How does understanding change us?”, asking how knowledge can be fully appropriated to human growth, rather than How do we know? The Romantics emphasised personal identification of the human spirit with that which is to be known. Evangelicals uncovered areas of personal distortion and blindness which called in question the ability of people to know God, themselves or people around them truly. They also emphasised belief and faith as the mode of knowing and the centrality of God in all true acts of knowledge. Both groups therefore challenged the assumption that knowledge was to be conceived as independent cerebral ideas and affirmed a more total personal involvement in that which is known. The events of the Revolutionary Wars and industrialisation seemed also to require this involvement. The concerns of Wordsworth, Shelley, Wollstonecraft, the Wesleys, Whitfield, Scott, Paine and Burke were in different dimensions from the sceptical epistemological emphasis of the Enlightenment; the moves were from Reason to Feelings, Deduction to Intimation, Natural Goodness to Sin, Ratiocination to Living, Natural Rights to Innate Freedom, Natural Understanding to Revelation, Thinking to Faith and Sense Experience to Vision. The changes wrought by these perspectives were pervasive and eclipsed the writings of Locke and Hume for much of a century.
Post-Kantian developments in German thinking exhibited a similar pattern. Kant believed that he was eliminating metaphysics through a process of critically reflecting on the laws of thought but within a short time a number of new developments had bypassed this direction. Hamann was critical of Kant’s seeming dissociation of thought from language and the distinctions which this generated. By contrast he sought unifying principles which emphasised the coherence of thought. (Beiser 1987 39-43) With Herder and the Sturm und Drang movement Reason was subsumed within a richer set of responses focussing on the Spirit. At the same time although metaphysics per se had to be abandoned, it was possible to see natural science as culminating in purpose, life and direction. Herder used Spinoza to develop such a pantheistic response. Even Reinhold, friendly and sympathetic to Kant, was soon trying to rework his foundations by developing an understanding of representation which could undergird the Kantian framework. (Beiser 1987 226-265) Philosophers moved on to what was largely a reawakened metaphysics in the thinking of Hegel, Hölderlin, Schelling, Steffans, Novalis and Schleiermacher. This group, it could be said, had been compelled by Kant to focus on the difficulty of grasping the ding-an-sich in rational thought; their response was to find ways in which the human subject could fully lay hold of the objects of her/his thought. This was possible by suspending the gegenstand of the knowing subject in relation to the world which was so central to the 18th century epistemological stance and, instead of addressing it as a problem, taking it as an inner relationship. Thus Hegel in the Phenomenology of Mind showed the gulf fixed between the self as absolute subject or consciousness and the object of knowledge in and of itself, and set out to bridge it. (Hegel 1807, Dancy 1985 227-30) This aroused hope that intuitively or rationally the nature of reality could be grasped, and the gulf to the ding-an-sich could be crossed. (Hegel 1812-16) Again, the epistemological mode of approaching knowledge was disavowed.
At about this time the growth of disciplinary knowledge in history, economics, sociology and other areas began to give yet another focus to the formation of knowledge. Scholars in these areas brought a more direct faith to their study. Despite repeated setbacks the underlying commitment throughout the late 18th to the mid 19th century was of progress in knowledge by exploring the relationship between Man and Nature: if the human mind could confront the nature of these realities, and unlock their structures and principles, then Man would understand the world in which he lived. The curtain would be thrown back and the true nature of the given reality revealed; each new version had this driving optimism. There was assumed to be an inevitable progress in knowledge; it could never be lost. Fact-gathering in the 19th century was gradgrindian, not 18th century epistemological empiricism. It was cumulative, subject to the law of addition, a conquest which would occur through an uneqivocal direct understanding of the world. We have examined the way this optimistic faith foundered in Classical economics, but here and in other areas it had a period of hegemony when natural laws were expected to reveal themselves unequivocally.
The Foundational epistemological response in the late 19th century grew from the failure of this optimistic naturalism, as it produced a range of conflicting representations. Reality was variously seen in materialist terms, as a metaphysical system, as a spiritually intuited whole, as Nature, as an organism, as a system in conflict, as mechanical laws, as an historical entity or as an expression of the human spirit. (Mosse 1963 13-289) The jangle of sounds was confusing, the multiple images flickered. The promise of directly grasping reality to the soul which was offered by Romanticism, Realism, Materialism, Marxism and some forms of Christianity had degenerated into the problem of the dogmatism of knowledge. Heterodoxy was occurring everywhere; in art Romanticism, Neo-Classicism, the Pre-Raphaelite Movement, Naturalism and Realism vied to show different visions of the world which were not compatible. In Politics Liberalism, Conservatism, Statism, Marxism and Socialism all claimed a correct understanding of the nature of political life. Mechanical, Materialist, Creationist, Evolutionary, Vitalist and Organic views of Natural Science also produced conflicting views. Theology faced the competing claims to orthodox belief of Evangelicals, Tractarians, Catholics, Christian Socialists, and Theological Liberals and the growing numbers of sects. Sociology included Comtean Positivists, Marxists, Conservatives, Evolutionists and Utopians. These multivocal manifestations created a number of problems. The authority of each position was challenged. They were seen to be biassed or ideological in their approach to the world and each other. Their central intuition after a while was seen as fallible, until finally the underlying claim which each made of grasping reality became suspect. Other visions could and would arise claiming the same status of understanding at least part of reality authoritatively. All, and none, of them were credible. Among some groups at least the heart went out of this mode of response; conviction receded and scepticism became part of the zeitgeist. In reaction and as a solution to this breakdown Foundationalism took shape. It sought to establish the proper foundations of disciplinary knowledge rather than push particular ideologies. For now ideology was suspect. Groups appeared who were ideologically committed to one view of the world which was radically and irreconcilably at odds with other views. They existed within the disciplines of economic, politics, psychology and sociology, and the epistemological challenge therefore became to establish some neutral ground on which knowledge could be established which was uncontaminated by ideology. All these groups, claiming or assuming to own the truth, could not simultaneously be correct. Each dogma was only right in its own eyes, sowing often bizarre ideas. Later with Fascism, Nazism and Stalinism the pus finally broke out in patterns of unreason. The fight of the foundational epistemologists of the late 19th and 20th centuries was therefore largely against ideology rather than, for example, religious belief.
For those in philosophy the assumed perspicacity of the enterprise of grasping the true nature of reality was called into doubt. Articulations of an answer in terms of appearance and reality (Bradley), intuition (Bergson), art (Santayana), absolute idealism (Royce), monism (Haeckel) and consciousness (Green) failed to command any long-standing support. Rather than carrying the conviction that they were addressing the new disciplinary issues, they conveyed the idea that philosophy was a separate mode of discourse from disciplinary knowledge. In reaction and as a solution to this problem with substantive views of the world younger philosophers were likely to look to the epistemological scepticisms of the 18th century. Within this reaction epistemological foundationalism took shape. Yet it was not just a reaction, for the knowledge around which the change in views took place was primarily that of the now strongly growing human and natural sciences.
This summary shows a number of different orientations to the “problem of knowledge”. These include egoistic deductivism, the naturalist building of knowledge of Locke, the sceptical empiricism of Hume and the a priori rationalism of Kant. But it also shows that for other philosophers at the same time and later “the problem of knowledge” was not approached as the key question requiring a systematic methodological response. Especially it shows the eclipse of epistemological priority in much of the 19th century; this way of thinking can obsess a generation of thinkers and then recede from consideration with the next, showing that the question, How do we know? is only one way of relating to knowledge. The summary also identifies a range of responses which can arise within the rubric of the “theory of knowledge”, a term which has its own limited location. It brings out the possibility of dissociating knowledge from understanding. It identifies an apologetic concern, which focusses especially on the questions of whether and how we can know God. It shows that subjectivity, objectivity, laws, process, rules, categories of thought, consistency and certainty can be key or recessed issues in considering the issue. Even this cursory study discountenances the idea of a continuous and equally weighted concern with the “problem of knowledge”.
Foundationalism, as defined here, is therefore a study of a specific historical response to the issues of the late 19th century. Although it drew on earlier epistemological traditions, its concerns were with disciplines, not the ego, with knowledge, not understanding, with public validation rather than the process of human understanding and with the problem of ideological thinking, not with religion and apologetics. It is to this specific response that we now turn.

Epistemological Foundationalism.
The sitzproblem which generated Foundationalism is now clearer. Although it is articulated in this book in terms of economics, it was actually a wider phenomenon, reflecting a 19th century crisis in the belief that the nature of the world could be directly grasped and understood. Classical economics was one example, among other possible ones, where ideological grasp was in disarray. The particular epistemological issues which were at the fore at this time were the kinds of knowledge which were valid in the human sciences, and it was in these disciplines that foundational patterns of understanding most easily took root. The need for academic respectability and the collapse of philosophy and theology as possible sources of integration for knowledge created additional pressure for this kind of development to occur. Yet, of course, there was nothing inevitable about this move. How these problems were interpreted largely depended on what scholars were prepared to reconsider and what they retained as assumptions. The feebleness of christian contributions at the time ruled out many possible re-evaluations, and the greatest tenacity was perhaps shown to the idea that human knowledge should in one form or another be unassailable. This led to a particular outcome which occurred on a wide scale in many different contexts in this overall post-Enlightenment cultural milieu.
The movement involved a switch to a driving concern with how knowledge should be gathered and evaluated, which allowed, it was believed, open agnosticism for the time being about what the world was like. It sought independence from subjective bias, from national, class and personal slanting of scholarship to a neutrality which was proof against prejudiced ideologies. It hoped to rescue late 19th century culture from a warring, self-validating relativism by establishing educational certainties which rested in the validity of the process by which knowledge was gathered. The transition was not automatic or immediate. The internal problems of the 19th century European world-view were so deep that it took time to meet such a dilemma in the culture of the era and fight for a new solution. Some people came to the issues early; others decades later.
The transition can be illustrated by one of its earlier forms. The advent of Impressionism was a major revolution in painting and is usually linked with the advent of Modernism. The Romantic and Realist schools of the earlier period had painted heroic scenes, socialist themes, everyday life, classical scenes, eastern subjects and political caricatures. They had different styles, but the main emphasis, as the label, “realist”, indicates, was on what they painted. After the exhibition at the Salon des Refuses in 1863 a new emphasis was present. It can be illustrated by the great series of canvases which Monet did of haystacks, the front of Rouen Cathedral and his lily pond. He did not paint 40-100 canvases of these subjects just because he was obsessed by them as objects, the nature of which he was depicting, but because he was concerned with how one perceives and paints. The emphasis on direct perception, the new theory of colour and the method of painting which later developed into Pointillism are well known. The Impressionist Movement involved a change from concern with the substantive subject matter of art to an involvement with the process of how one sees and paints. As such it was also part of this great transition.
In one sense the move was a retreat. It involved a new scepticism about the claims to understanding made by these multivocal positions, and its deliberate intention was to be agnostic about the nature of the world, society, politics, the mind or the economy. The self-confidence of statements of belief made by ideologues, metaphysical thinkers, churchmen, in art, theatre, literature and music prompted the reaction, On what grounds do they base their belief? This was the proper location of Yeats’ “The best lack all conviction, but the worst are full of passionate intensity”. Similarly, social theorists were retreating from ideological claims, committed statements, judgements and ethical principles, towards patterns of supposedly neutral scholarship, as the methodological studies of Durkheim, Weber and Pareto show. (Durkheim [1895], Weber [1904-18], Pareto [1906]) . If the status of knowledge depended not on what it asserted to be the case, about which there was, and could be, substantial disagreement, but on the mode of its formulation, then the ideological and metaphysical precommitment which generated so much controversy and failure to communicate could be obviated. In an era when ideological conflict was becoming more and more intense, this seemed no small gain.
In another sense it was a new optimistic movement making authoritative claims. If one could establish how one knows, sees, thinks, forms theories or makes propositions, then an unequivocal method was tooled up for well-formed knowledge. (Storkey 1979 93-106, 314-6) The productive metaphor is apposite, because this was the beginning of the era when knowledge was to become an industry. It was a question of reculer pour mieux sauter. The foundational method of knowing would in the end produce a range of well-formed and reliable knowledge which far outran the hotch-potch of opinions which was now available. The new emphasis was on a foundation which established what reliable knowledge was. This was the Holy Grail; there had to be some utterly trustworthy basis for knowing. If the kernel of the process of obtaining infallible knowledge could be prototypically built and properly used, then incontrovertible, scientific understanding would follow in spite of subjectivity, ideologies and opinions. Such a move would provide coherence in the face of the fragmentation of conclusions and would achieve the indubitability which would separate knowledge from mere belief. It was also needed as a basis for objectivity and value-freedom in the face of positions which seemed precommitted. And of course, the foundation would also be the way of establishing the integrity of the disciplines as areas of study on a similar basis to those of the natural sciences. Finally, it laid down a procedure of scientific distancing which human studies seemed to require for themselves as they moved into the academies and universities. Thus, independently, from all kinds of backgrounds, scholars began to search for this required foundation.
What, more rigorously stated, was the movement to Foundationalism? First, it was concerned with knowledge, usually that located in the sciences (rather than the Enlightenment concern with understanding, the relationship of the knowing ego to the world). This knowledge was generally conceived as theory, specialised propositions about an area of study which were seen as part of the public domain. The characteristic stance was therefore to seek to establish the nature of well-formed knowledge abstracted from the knowing subject and the nature of the world. This was done by establishing some basis or foundation through which knowledge could be validated and to which appeal could be made as a criterion of distinguishing it from unfounded opinion, hence the central use in this study of the term Foundationalism.
It should have some infallible or authoritative status as knowledge irrespective of the knowing subject and the given world. Crucial was the distancing which dissociated from the knowing subject to treat understanding and belief at one remove. The stance was of the onlooker who understood the process of constructing knowledge but did not have to adopt the position of direct understanding and belief. (Storkey 1979 94)

This reflexivity seemed to allow belief and understanding to be handled dispassionately. The thinking ego was subordinated to the foundational requirements in a way which eliminated subjectivity as something which might bias the construction of knowledge. The stance was concerned with the public accountability of knowledge irrespective of the persons involved. The question is less whether one person has inner certainty of propositions as true, and more whether they can be publicly validated; belief is dissociated from knowledge. We shall describe this as the objectivity requirement or the onlooker stance.
Another part is the understanding of knowledge as encyclopaedic and cumulative; it can be gathered into bodies which can be handed on and learned in appropriate ways. Thus, knowledge is seen as making progress, but of a certain kind. The assumption is of a process which is almost additive whereby the building of knowledge will grow ever higher. This process must be more or less unassailable and guarantee the development of the science for its recipients. For knowledge to have these kinds of cumulative properties, it must have secure foundations and never incorporate error which might need demolition. This was the requirement for universality or generality.
When, as often happened, knowledge of the special sciences was challenged by an alternative view, there was a need to show how and why this knowledge was well-formed and authoritative. This, of course, also implied criteria for establishing theory which was reliable and repudiating that which was not. There were normally two levels of criterion; the first was concerned with establishing what propositions and theory had followed the proper procedure of construction as knowledge, while the second established whether, given that form, it was valid or defective theory. At least at the first level it should therefore be possible to be agnostic about the content of theory and propositions and merely concentrate on procedural criteria. The method insisted on epistemological priority; how knowledge is obtained was the question prior to what it was.
There was also strong pressure for knowledge to be neutral in relation to what were seen as a variety of potential biasses. It was not to depend on or be beholden to anything else like metaphysics, religion, ideology or dogma which would shape it in a certain way, but was to stand sui generis. The charge of Marxist analysis that ideas reflected the relations of production seemed to require a defence against automatic bias. Similarly the pressure to change the world in socialist analysis was resisted by those who wanted to dissociate theory from understanding which required any kind of commitment or normative response; there should be a detachment, a take it or leave it quality about knowledge. This was not only a negative disposition, but also a positive commitment to the value of disinterestedness. The knowledge could be publicly traded with commodity status across all value commitments and ideologies. As a transferable resource it could bind together the fragmented commitments of late 19th century culture with a pattern of public understanding which was capable of being universal. Such an approach allowed, for example, the creation of textbooks.
Thus, Foundationalism understood in this sense denoted a genus of responses which have a common pattern. On the one hand they seek to remain agnostic about substantive claims to knowledge and to avoid dogmatism in asserting what is the case. But on the other they aim to establish sufficient grounds by which knowledge can be judged to be valid and stand as neutral, scientific and objective. It was part of a quest for certainty, expressed in a new way. (Dewey 1929, Aaron 1971 3-88) The older quests for what could be known about the nature of the world and for inner certainty in the knowing subject were substantially abandoned for a certainty that knowledge per se was certain. This is actually a momentous change. The hinge of the position was the optimistic conviction that a basis for obtaining well-formed knowledge could be established which would consequently engender such knowledge. There must be a Foundation. It might be logical necessity, unassailable evidence, the definitive process of understanding, the method of science or a demarcation of knowledge from unfounded belief, but it has to be absolutely secure. It should offer freedom from ideological contamination through neutrality, objectivity, validation as scientific, a prescribed methodology and a basis for coherence in a corpus of knowledge. It should offer a way out of the diverging convictions of the late 19th century humanist worldview.
The epistemological concerns of the late 19th and 20th centuries are therefore different from the 18th century For the earlier English empiricists the question of how we know was largely pre-theoretical and addressed to the individual psyche. In the later period it was posited of already existing bodies of theoretical knowledge, looking for criteria for distinguishing well and badly formed knowledge. The earlier context was concerned with establishing valid reasoning; the latter with the philosophy of science and specific disciplines. While the early philosophers were speculative and leisured, the later ones were busy setting the agenda for mushrooming disciplines and subdisciplines. The earlier reaction was against revelation as a source for knowledge, but the later against ideology. The early epistemological formulations were born of optimism about processes of generating “natural” understanding, whereas the need for a foundation for knowledge was the pressing result of the pessimism of late 19th century ideas.
There was no shortage of people committed to this direction. Indeed, the term itself was unavoidable and comes from the literature of the period. In psychology Wundt (1873), Mach (1875), Muller (1878), Lipps (1883), Kulpe (1893), Wundt (1896), Lehmann (1912) and Murchison (1929) all used the idea, “Foundations”, in the title of their texts. (Flugel 1933 363-72) Similar patterns occurred in economics, politics, history, mathematics, philosophy and sociology. It was a concept that a lot of people began to use with meanings close to that defined above. Its use here follows from these dozens of earlier examples from Marx’s Grundrisse onwards.
But more important than the term is the commitment which it expresses. Indeed, Marx’s use of the term is not fully foundationalist in the sense meant here, because it is in part concerned with the substantive foundations necessary to understand the nature of the economy. He precedes the methodological commitment of later foundational moves of which their are many other expressions. “The unity of all science consists alone in its method, not in its material.” (Pearson 1892 12) “This pure theory of economics is a science which resembles the physio-mathematical sciences in every respect.” (Walras 1874 71) “In order to follow a methodical course, we must establish the foundations of science on solid ground and not sinking sand.” (Durkheim 1895 46) “We shall therefore try to see whether we can arrive at a theory of all truth by starting from truths of fact” (Schlick 1909) “Moreover, transcendental phenomenology is not a theory, devised merely as a reply to the historic problem of Idealism, it is a science founded in itself, and standing absolutely on its own basis; it is indeed the one science that stands absolutely on its own ground.” (Husserl 1913/31 13) “I am not prepared to admit that there is any problem which cannot be solved by quantitative methods.” (Cobb 1927 921) These were but some of the assertions, but behind them were many others who assumed this was the direction forward. Gradually these streams of commitment gathered into rivers of orthodoxy, picking up adherents automatically with the currents of academic fashion. In time it was to become the mainstream which dominated much of the public production of knowledge for the next hundred years. If this way of viewing knowledge was flawed, it was indeed a serious problem.
Although the concept, “foundation”, has been widely used throughout the period by theorists in the human sciences, those using it more recently in philosophy have interpreted it in a rather different way, focussing on the justification of belief rather than the validation of knowledge. They have used it in a variety of ways, which can broadly be classified as wider and narrower. There are clearly family resemblances between the definition in this study and a wider view, as expressed, for example, by Bernstein. He links it to “Cartesian anxiety” and a long tradition of epistemological concern. (Bernstein 1983 8-11, 16-19) In this definition the focus is somewhat more general, involving the search for “an ahistorical permanent matrix for grounding knowledge”. (10) The vagueness of this definition, however, makes it less useful for the tasks undertaken here; this study suggests there is much distinctive about the last hundred years, like the disciplinary base and the retreat from ideology, which needs its own analysis. Wolterstorff similarly identifies a tradition including the Enlightenment era, Aquinas and even Aristotle. (Wolterstorff 1976 30-4) He defines foundationalism partly in terms of “theory” and “science” (which do not easily carry back to the Middle Ages) and proposes that the acceptance of a theory as part of genuine science is dependent on it being justified by a foundational proposition which is true in its own terms. Here the idea of scientific knowledge being dependent on a validation process is brought out, but the concept of “foundational proposition” is seen very widely; it includes Aquinas’ “natural light of reason”. (30-1) The explanation is that Wolterstorff’s concern is with theistic belief and whether justification of it is required in terms which could be described as foundational. (Flew and Macintyre 1955 96-9, 109-30, Mackie 1982) This agenda within the philosophy of religion requires an apologetic frame of reference which is different from the focus on the philosophy of the human sciences in this study. (Plantinga 1975, Plantinga and Wolterstorff eds 1983) Its concern is much more with the justification of belief than the validation of knowledge. While foundational views have been used to attack belief in God in the philosophy of religion, in the human sciences they have normally not at all been concerned with that issue and have directed their efforts to establishing an agnostic position in relation to all ideological and normative views.
Alongside these wider views are a range of narrower philosophical uses of the term. For example, Lakatos discussed the problem of the foundations of knowledge, but his definition hinged on a distinction between foundations, that is the epistemic infallible base of knowledge, and the method of discovery of knowledge, which proceeded on his view according to a different rationale. (Lakatos 1968 315-417) This distinction, owing much to Popper, (Popper 1979 378-82 [1930-3]) was part of a particular foundational view within the positivist tradition which we examine later. This focus fails to come to grips with the full range of foundational positions which occur outside the Falsificationist/ Hypothetical Deductivist development. Alston similarly does not look to a wider frame of reference for the meaning of foundationalism. He, Chisholm and others tend to have a deontological view of the term (Alston 1989 115-52) which identifies two kinds of belief, those which are non-basic and need the support of other beliefs through inference, and those which need no support and are foundational because given directly by the senses. (Albert 1985 12-38, Dancy 1985 53-84) This meaning, given currency by Chisholm and C.I.Lewis was criticised by Feyerabend in the paradigm debate. (Feyerabend 1961, Couvalis 1989) Its limitation in relation to this study is the way it does not address knowledge which claims to be founded on logic, testing, experiment, behavioural analysis or instrumental rationality and does not relate to wider methods of knowledge validation in the human sciences. The foundational model of directly justified beliefs is constructed around personal belief systems and does not usually relate to patterns of gathering knowledge in disciplines. Since this is our main concern, there seems no value in this context in retaining the narrower meaning.
It is also worth relating the concept used here to recent work in the philosophy of economic science. The concerns in this study are close to those of Caldwell, who looks both at the positivist tradition in the philosophy of science and its economic embodiments. There are differences in this study: in the identification of three major traditions, in taking foundationalism as the core problem and in seeing its expressions as more formative, dogmatic and destructive of economic knowledge. These points will be taken up in the last chapter. (Caldwell 1982) Boland uses the term, “Foundations”, more directly but uses it to mount methodological critiques of the foundations of neoclassical and other forms of theory. The result is a different mode of analysis from that used by Caldwell or here. (Boland 1982) Wong’s study of Samuelson is also explicitly foundational, but puts less emphasis on the formative role of epistemological stances in shaping theory. (Wong 1978) Carabelli’s study of Keynes emphasizes fully the epistemological formation and this study comes to similar conclusions in relation to this theorist. There is therefore a good body of work leading in this kind of direction, identifying the word and the idea of foundation as of importance to the discipline.
Foundationalism therefore occurs at the confluence of a set of influences. Its origins lay in the confusion of 19th century beliefs about the nature of the world, evident in the breakdown of classical economics. Its context was the professionalisation of the human sciences and their need for academic status. It took place when some disciplines had territorial ambitions on others and there were disputes about the boundaries. It occurred when disciplines first became self-conscious about their methods of acquiring knowledge and were developing theory quickly. At that time the philosophies of Science and Social Science were emerging from general philosophy as an area of study, and it also happened when the massive 20th century expansion of universities and colleges was being shaped and extruded. Consequently, no one domain of analysis is adequate for its full study, and what follows will merely examine a limited part of the phenomenon in greater depth.
Finally, this study suggests foundationalism is fundamentally not what it claims. It involves a set of assertions which are not necessary epistemic conclusions and therefore constitute misguided beliefs which must be discarded in their own terms. Yet these beliefs are also ones in which the adherents live as a matter of pre-theoretical commitment. Despite the fact that various formulations of foundationalism were refuted or discredited, there were repeated attempts to re-establish the underlying premisses. They therefore lack many levels of evidential support and have been often held as dogma in the sense of being beyond self-critical reflection and awareness. This kind of scientific knowledge, it was believed, did not have the dependent and responsive character of a faith, since it required no external referent. It believed it was beyond belief. Part of the credo of foundationalism was its commitment to the separation of belief and knowledge; the former could be categorized as subjective and probably ideological, while the latter was in some sense objective and ideologically neutral. In asserting the internal autonomy of such knowledge, foundationalism was therefore potentially repudiating characterisation as a belief or faith, but the failure of the foundational positions opens up the question of whether faith and knowledge can be so categorised, or separated at all. It may perhaps be that the formulation of the status of different kinds of knowledge and the criteria by which they are assessed can only be established in terms which involve shaking the foundations. Part of the task in looking at this supposed credo-free approach to knowledge is therefore to uncover its failure in self-description.

The Failure of Foundationalism.
Later in the study the failure of foundational claims will be examined in a number of specific areas. Here it will be examined in more general terms. Because of the widespread espousal of foundational scientific theory, there were few who stood outside it and were able to offer a critique of the direction which was being adopted. Only recently as the collapse of these positions has become more obvious have they begun to be critically addressed in more general terms. There are, however, a few scholars who have seen more strategically and systematically the weaknesses of this approach. Here we shall examine two who analysed some of the weaknesses of foundationalism during its earlier stages of development.
One of the first to identify some of the problems was Leonard Nelson. He did not fit easily into the mainstream of philosophy; his stance was that of a Socratic post-Kantian commentator on the developments he observed. In “Über das sogenannte Erkenntnisproblem” (1908) Nelson analysed the reversions of Schultz, Jacobi, Maimon, Schiller, Reinhold, Schulze, Fichte, Schelling, Hegel and others to various pre-Kantian positions, and found Fries different in avoiding their weaknesses and really carrying through Kant’s critical method. In particular he liked he way Fries moved out of the epistemological dilemma of justifying each item of knowledge by an external ground of validity. (Nelson 1971 II 164-73). On the contrary Fries showed that knowledge always required specific justification and participated in metaphysical or world-view judgements. (Nelson 1973 59-393,esp 92-152) In grappling with Fries’ argument Nelson stood outside many of the contemporary foundational developments, not merely as a reactionary, but as someone who had thought through in principle the epistemological stances to which others were now becoming committed. Thus he asked if Mach could be successful in formulating a view of natural science which was free of metaphysics. (Nelson 1974)
By 1911 he had reflected on this preoccupation with establishing the definitive science of knowing and asked if the problem of a foundational epistemology admitted a solution in principle. Could the theory of knowledge be securely established? He came to the conclusion that a foundation for knowledge was not logically possible. For in order to solve this problem there would need to be a criterion by the application of which one could decide whether or not a cognition was true, a validity criterion. This criterion would itself either be a cognition or not be a cognition. If it were, it would fall within the area of what is problematic, the validity of which is first to be established by the criterion, therefore it could not be a cognition. But if the criterion were not a cognition, it would still have to be known, that is, one would have to know that it was a criterion of the truth. But in order to gain this knowledge one would have already had to apply it, which would be impossible. The conclusion was that the problem is insoluble. (Nelson 1946 185-205)
Nelson’s understanding of why the problem was insoluble is interesting and addresses a recurring theme in this study. It was concerned with the mediation of knowledge. If knowledge is always validated on the ground of another cognition, then no knowledge per se would be possible. Nelson therefore argued that all knowledge required a judgement by the subject. If, however, knowledge is always validated by a criterion which is a matter of noncognitive faith, no knowledge is without some, however small, substantive content of belief and breaks down the possibility of distinguishing in any fundamental sense knowledge from belief. It also means that the foundational position of principial agnosticism with regard to the substantive content of theory is ill-founded. The conclusion is interesting in the light it throws on the relationship between cognition and non-cognitive belief, for of course the whole argument hinges on the premiss that cognition and belief obey the law of the excluded middle. For those who hold this position the argument is a refutation of foundationalism, but a more sober conclusion might involve questioning more fundamentally the dissociation of knowledge and belief.
Another part of the weight of this argument lies in the way it fits problems which occurred as the search for certainty in epistemological terms developed within foundationalism. Thus, for example, Russell and Whitehead’s attempt to reduce mathematical systems to a logical foundation faced Gödel’s demonstration that even within quite simple mathematical systems, the internal consistency of the system is not demonstrable without bringing in other equally questionable principles of inference. (Gödel 1931 in Heijenoort 1967 592-617, Newman 1668-95) Russell’s problems with classes which are members of themselves exhibits another such issue; there always seems to be an irreducible element or metalanguage which defines the classes under consideration in some other terms than they define themselves. Similarly, when members of the Vienna Circle and others set forward the Verification Principle stating that only propositions which could be verified by the senses were meaningful, it turned out to be self refuting and defied successful reformulation. (Ayer 1936, 1959, Feigl and Brodbeck 1953, Hanfling 1981) When Ayer and others tried to make the analytic/synthetic distinction fundamental to knowledge, they defined that which is analytically true by virtue of identity or synonymity, as true in all possible worlds and therefore beyond disproof, but because the terms in which the synonymity was expressed were part of a specific world and language, they could never be divorced from the framework out of which the analytical terms purportedly broke free. (Ayer 1940, Quine 1953) Wittgenstein at the end of the Tractatus, having stated an epistemological position which centres on the validation of propositions by the senses, has to admit that the propositions which he has formulated are themselves nonsense in the terms he has just proposed, and must be abandoned. One must, so to speak, throw away the ladder after one has climbed up it. (Wittgenstein 1921 151) In all of these cases the formulation of a foundation for knowledge seems to undermine itself; the criteria lose their status as knowledge and become arbitrary. This suggests there is substance in Nelson’s critique; this move is one which does not deliver what it promises. If the basis for knowing is specified within the corpus of knowledge itself, something slips out as an extraneous judgement. If it is externally given, it has no justificatory basis completely outside that for which it purports to provide a basis.
Nelson’s approach was within the Kantian tradition of critical philosophy. He was less directly concerned with theoretical knowledge in the special sciences. With the burgeoning of scientific conceptions of the disciplines and the philosophy of science a range of additional issues occur, in particular the status of theoretical knowledge within particular disciplines. Consumption theory involves this dimension. The Dutch philosopher Dooyeweerd, working in the 1930s to the 1960s when the disciplines were more developed, addressed this issue more specifically.
His first argument approached meaning or knowledge in a radically different way. As a christian he took the scope of meaning present in creation as given by God. Scholarship and indeed immediate day to day knowledge was therefore essentially responsive and dependent in nature. It was dependent on God in many different dimensions, although a lot of scholars did not acknowledge this, and it also had an inherent relational and dependent quality which was provided by the given meaning of creation. Because the meaning of knowledge and understanding was given by God and through the creation, knowledge which was truly articulated had to acknowledge this dependent status. The fear of God is the required beginning of wisdom. Yet those who did not acknowledge this dependence, Dooyeweerd argued, had to find an alternative source of meaning for their knowledge, which would undergird or support it, an origin or Archimedean point. Within most philosophies or views of knowledge this meaning base had to be located within the creation, or more specifically within a specific mode of thought about the creation. Because its locus was immanent to the creation, it had both to misstate its own meaning as self-sufficient and distort the meaning relationships of other parts of the creation by stressing their relation to the supposed Archimedean point. This point addressed much of what occurred in the period leading up to the first crisis of ideological proliferation; a range of different perspectives each seeking their origin of meaning in different parts of the creation were at war with one another. By not facing this fundamental religious issue, foundationalist thinking was obscuring the issue it had to face.
But this problem becomes further confounded when confronted with further supposedly neutral theoretical development. Dooyeweerd argued that religious forms of idolatry often present themselves as theoretical dilemmas, which are insoluble. The religious dialectic, in other words, entangles life and theory in a dialectic that is utterly incomprehensible when measured with the yardstick of the theoretical dialectic. Unlike the theoretical dialectic, the religious dialectic lacks the basis for any real synthesis.
Let no one, therefore, try to correct the religious dialectic by way of the theoretical dialectic – the method of the Hegelian school. That approach is an utterly uncritical form of dialectical thought, because at the root of its overestimation of the theoretical dialectic lies a religious dialectic that is hidden to the thinker himself. (Dooyeweerd 1979 13 [1945-8])
Thus, attempts by the Hegelian dialectic or any other theoretical method, to resolve the problem of the false origin for the meaning of life in theoretical terms will not work. In the terms of this study, you cannot solve the problems of the first crisis, by some kind of epistemological revolution. This is because the supposed self sufficiency of meaning of theoretical thought cannot hold. In claiming a “foundational” role (Dooyeweerd did not use this term), theoretical knowledge was mistaking the locus of many of the problems it tried to solve, which were more deeply religious. (Dooyeweerd 1935-6 I 1-21) Thus, for example, the idea of a logic of utility which is the basis of much consumer theory is not purely analytic; it is based upon a commitment to choice and satisfaction which rules out of consideration many of the factors which actually play into consumption decisions like shortage of time, greed and social conformity; choice and utility are (falsely) sanctified. The theoretical framework is therefore already locked into preconceptions which invalidate conclusions and misrepresent what is happening. He believed that the patterns of distortion could actually be mapped with reference to a coherent Christian perspective to avoid many of the theoretical errors we shall later be examining.
The second argument concerns the limited status of theoretical knowledge. This mode of thinking occurs as a result of a deliberate process whereby theory is put over against what is given, the Gegenstand relationship. Dooyeweerd distinguishes theoretical understanding from naive experience, which involves any direct response of human subjects to objects in the world in an undifferentiated way. The theoretical stance abstracts from other aspects of reality and addresses its subject matter inquisitorially, but it therefore remains relative to that over against which it stands. This distinction may be exaggerated, (van Riessen 1958, McIntire 1985 143-66, see different meaning of foundationalism 145n) but it points to a vast area which may make theoretical thought problematic. For example, it involves abstraction, the decision to treat as of no account vast areas of subject matter which may be relevant to the theoretical issue under consideration. Because the making of theory is an intentional act, not an ontological reality, all kinds of issues arise. (Dooyeweerd 1935-6 I 39) Theory ignores. It also addresses subject matter. Decisions are made about the domain of cases which it addresses. If the theory gets these decisions wrong, if the domain of cases is seen as too limited or too extended, or is misinterpreted, theory formation is distorted. Simply to assert the pretended autonomy of theoretical thought is to cover up all these issues and rule out the possibility of addressing them. The disjunction must be resolved with full respect to the subject matter which presents itself. Both the processes of disjunction, or gegenstand, and synthesis must be deliberative acts. Baldly, theory must be formulated with awareness of its limitations and its essentially responsive nature. By contrast when it pretends to self-sufficiency, it cuts itself off from real engagement with its subject matter. Theoretical syntheses, including those generated by foundationalism, thus remain unselfcritical. (Dooyeweerd 1935 I 34-51) Later, this study examines some of the problems created by the assumed self sufficiency of theory.
The third argument begins with a recognition of various modes of human experience which are generically different and studied by the disciplines of the modern university. These modes or aspects are partial; a person may in one act of buying a good have a certain pulse rate, emotions, spatial and geographical location, a history of that act, energy transfers, ethical calculations, financial and work considerations, aesthetic evaluations, social relationships, legal constraints, patterns of communication, thought and calculation and levels of wisdom. Each of these aspects can be investigated by one discipline, but no one mode of study can grasp even this simple act. The geography, psychology, economics or aesthetics of buying are relative and partial studies. This requires, as we have already established, that the act of abstracting one mode of study, say the economic, be deferentially and cautiously made, so that respect for all the other aspects of the act of buying are maintained in the partial theoretical analysis. Dooyeweerd thus underlined the way in which all analysis is interdisciplinary, if even only by default. Defining theoretical analysis in foundational terms often compromised this inescapable truth.
This perspective of the modes of human experience also affected the way disciplines themselves were defined. The meaning of disciplines is not theoretical, but arises from the area of life which they address, and the central norms of this area infuse and shape the theory of the discipline. Thus Dooyeweerd refused to grant the autonomy of the theoretical domain and required the continual interaction of theory and life; this, as we shall see, addresses the underlying problem of otherworldliness which has beset consumption theory and other theoretical areas during the last century or so. Economics is a viable study because people plan the use of resources according to principles and personal priorities and not because general equilibrium models can be solved. He saw that this mode of life related to other areas; there is economy of thought and energy, (II 66-68) and therefore retained a basic integration between the way people live in all its varied richness and the theoretical constructs which respond to this pattern of living.
By contrast, foundational approaches, premissed on the autonomy of theoretical thought, require the meaning of the theory to be located within the disciplinary analytical frameworks. Often it will use disciplinary matrices, like psychologism, historicism, materialism, logicism, economism and biologism. These look to some Archimedean point which is partly located in an area of life, but also partly depends on the theoretical structures which have articulated this as a basic commitment. Thus, for example, we shall later study Becker’s claims for the power of utility maximization as a fundamental social analytical tool; I believe there is a link between the economistic belief which says money is the measure of all things and the belief that utility calculus is the key analytical framework for childbearing. Thus, Dooyeweerd insisted on the pretended autonomy of theoretical thought; its commitments remain embedded in the analytical systems, but unacknowledged in a way which means the theory is other than it usually claims to be. (Clouser 1991)
But another problem is more straightforward. Because disciplinary theory is presented in foundational, or autonomous, terms, it loses contact with other subject matter which it important to the domain of study. For example, maximization is an important theoretical tool in consumption theory, but maximization only poorly comes to terms with all that is involved in eating; dieticians operates with the very different principle of “what is good for you”. Thus, autonomous thought continually fails to relate to the full subject matter it faces, which functions by other principles. By contrast Dooyeweerd argued that the norms of living should infuse and shape the theoretical concerns of the various disciplines. They are life driven. But at the same time the theoretical domain must be comprehensively defined to allow all the relevant modal issues to be brought to bear on a subject. A comprehensive theoretical framework is necessary, one which only the biblical doctrine of creation is big enough to provide, but it is to be integrated with the underlying view of life which gives it meaning, not claim some kind of autonomous status. (Dooyeweerd 1935-6 II) This is part of the task of the fifth chapter.
A fourth focus was the failure of foundational thought in its pretended autonomy to reflect on the role of the subject in the activity of theoretical formulation. In no way could the relationship between the subject and the objects of theoretical reflection and analysis be reduced to a so-called objective, logical, neutral status. The subject always had ontological and epistemological commitments which shape all her/his thinking. Any assumption of neutrality prevented a deeper penetration to the character of the subject’s contribution, especially in terms of the faith and worldview which s/he articulates. In bald terms, if the subject can be ideologically committed, s/he can no less be epistemologically committed in ways which reflect her/his beliefs. Nor are these commitments just personal, but reflect vast cultural trends, or ground-motives out of which people respond, whether they know it or not. (Dooyeweerd 1935-6 I 114-527, 1945-8) Polanyi and others have opened up the personal dimensions of the theory of knowledge, (Polanyi 1958) but Dooyeweerd formulated this point, like all the others, in a bigger context. It was the created necessity of a human response to God which required a faith response and denied its reduction to a neutral status. Similarly, the dependence of the creation on the creative law of God required acknowledgement of God in relation to all human knowledge; otherwise it misstated its origin and meaning. Clearly, this commitment to a specific view, both of the world and theory, was counter to the tenor of foundationalism, and needs further discussion.
The context of these points needs some elaboration. They were not constructed as a critique of foundationalism, but of the presumptions of autonomy in theoretical thought. This involved analysis of both the substantive weltanschauungen of theoretical perspectives and some of the epistemological positions which are examined in this study in foundational terms. It is worth pointing out a difference of approach between Dooyeweerd’s concerns and those of the present study. His philosophical aim was to identify the ways in which epistemological positions led to a distorted or incomplete understanding of coherent reality seen in Christian terms. This study remains more strictly concerned with following through the implications of foundational positions. (Until the last chapter, when some of the gaps in consumption theory are more directly explored from a Christian viewpoint.) Especially valuable in view of the focus of this study is Dooyeweerd’s continual questioning of theoretical autonomy, the neutrality of scholarship and the problematic nature of theory at a time when few others were doing it systematically. Neither Nelson nor Dooyeweerd was in a position to study the later developments arising from this basic move. With hindsight we are in a better position to spell out in more detail the structural problems of foundationalism.

The Structural Problems of Foundationalism.
To conceive of knowledge in these foundational terms, whether to establish the scientific status of economics or to have an incontrovertible basis for theory was thus a big step. Here we attempt to examine more systematically what that step might involve. Each foundational position, of course, varies in its consequences, but the suggestion here is that there is a structural pattern to those consequences which can be traced through the position. Before the detailed examination begins, and without prejudging what it might find, these implications are explored.
1. Foundational Validity.
The foundational claim is that certain kinds of knowledge constitute valid (economic) theory; they are well-formed. By contrast other forms are invalid or not properly scientific, or not even knowledge at all. It is the ground of this validity on which the position rests. Normally, it is seen as self-evident, incontrovertible, or basic to a definition of economic science. “Analysis cannot but proceed from what is logically the case.” “Economics must study the way people behave economically.” “You can’t argue with the facts; every economic theory must be judged against them.” “All economic events must be the result of prior causes.” “Of course economic theory has to be consistent.” The grounds on which these claims are made vary, but they all assert some kind of validation which establishes what normal economic science is taken to be.
As Nelson has shown the claim to establish the ground or criterion on which knowledge is to be based is more ambiguous than it might seem. Often it involves a process of self validation which can be examined in a number of ways. First, its own inner argument for irrefutability can be challenged as invalid; it is not the cognition it claims to be. Second, it can be shown not to have the consequences which it is deemed to have. Third, an alternative basis for knowledge can be brought to bear on it which shows up weaknesses and gaps which have not been considered before. Thus, positivism can be criticized by showing that the Verification Principle is self-refuting, by exploring the limited conclusions which can validly be drawn from positively defined “facts”, or by showing that positive knowledge excludes most of normal science, as Popper argued. (Popper 1965 281)
The question arises as to what critique is deemed sufficient to invalidate a foundational position? Those holding a position tend to feel that only a refutation of the inner logic establishing the foundation is adequate. Although the failure of supposed inner certainty occurs in all the positions examined here and therefore invalidates them in their own terms, there are wider considerations to take into account. First, the phenomena to which these positions lay claim extend far beyond the foundation itself, and whether their characteristics are addressed with full integrity is a relevant argument. Second, the terms of the inner (self-defined) logic are open to question, since other logics, or consequential arguments, can at least claim to be interpretations of the positions. Third, there may be consequences of the foundation which are not divined in the logic of validation which also need consideration as evidence of what that position implies. The way these positions have fallen apart therefore varies. Often with hindsight it is amazing how seriously the claim to authoritative forms of knowledge was taken by those who deferred to them. Some positions claim two or more kinds of knowledge based on different foundations. These positions face the question of what makes these different grounds for knowing, and only these, valid, and what is the relationship between these autonomously defined categories of knowledge.
Examining the foundation is not merely a matter of critique, but also of investigating why the position was adopted and what makes it false. This can be carried out in terms other than those in which the position is stated. In a sense we are always travelling back to the start of the journey, but with insight about where to go from there, and with a different map.
2. Foundational Dogmatism.
Because different foundational positions each have their own grounds of validation which are self-supporting and claim to be without external dependence, they stand in a dogmatic relationship to one another. The relationship between the competing bodies of theory which they generate or the authority which each claims for its kind of knowledge seemingly cannot be discussed except in terms of one or the other position. Often, the other position cannot be articulated and discussed in terms of the first, because the incontrovertible foundation of the first makes the second position unknowable. Thus a lack of communication is built into the theoretical enterprise at every level; as Blaug suggested, it is like playing tennis with the net down. (Blaug 1975)
For consumption theory this implies that the formative theorists, Jevons, Pareto, Marshall, Keynes, Hicks, Samuelson, Friedman, Becker, Scitovsky and Katona have had alien visions of their enterprise, and have largely been unable to articulate why they have these fundamental differences of theoretical definition. Because they are all supposed to be engaged in economic science, the presumption of a common enterprise is there, but the substance is not. The extent to which this is the case will be examined in the body of this study.
The dogmatism arises because there is are no theoretical grounds on which the incompatibility of the foundational positions can be discussed. The pretheoretical or prescientific commitments which shape the positions are ruled out of the domain of academic discourse. The positive task of this study is therefore to open up the modes of discourse through which these differences can be debated. Wong tried to do this through what he called the “method of rational reconstruction” which aimed to present theory as a solution to an epistemological/methodological situation. (Wong 1978 9-24) Caldwell, after some important foundational analysis, evokes methodological pluralism as a way of crossing the divide, on the assumption that no universal, “logically compelling method of theory appraisal exists.” (Caldwell 1982 244-52) This methodological pluralism involves rational reconstruction of the positions, critical assessment, an examination of the strengths and weaknesses of each position and a pluralist comparative response. Although Caldwell partly resolves the problem of dogmatism, to offer short cuts across the fields to other roads, this study concludes that a more radical examination of the problem of foundational dogmatism and its effects on theory is necessary if a rediscovery of open debate in consumption theory is to be possible. The reason is because the dogmas are embedded in theoretical constructions which prevent an easy disengagement to a pluralist position. Further, a pluralist epistemology must receive other sources of guidance than those hitherto provided by defective epistemologies.
3. Theoretical and Methodological Prescription.
The foundational emphasis on the process of obtaining well-formed knowledge also leads to prescribing the kind of theory which should be produced, and the methods and criteria which apply to knowledge formation. The differences are immense; theory either should, or should not, have structure, data, assumptions, testing, generalisations, historical and logical components. Its aim should be to predict, model, isolate variables, organise information, expose alternatives, be tested, construct a logical framework, determine behaviour or be capable of solution. Because the shape of the theory is potentially so different, it is difficult to relate one to the other. Nor can economists test one another’s theory, because they differ in the methods by which theory is meant to be assessed and evaluated. This problem is reflected in the experience which economists often have when reading articles and papers; it is not a question of detailed criticism of the argument or conclusions, but a more general awareness that they would not have approached the issue in these terms at all. This perception is too widespread to be ignored any longer. (see Dow 1981, Weintraub/Dow 1982-3 295-308, Blaug 1975, 1980b, Weisskopf 1979) It has resulted in journals which address only their own methodological subculture and accept the fragmentation as normal. Because the methodologies are not neutral, they also encourage biasses towards certain economic philosophies and ideologies; later, for example, we shall look at the congruence between positivism and laissez faire individualism. (Finn 1979, Friedman 1953 cf Friedmans 1979)
This prescriptivism constricts the potential for theory generation and theoretical communication, and the positive task is therefore to open up new kinds of theory, and to formulate ways in which the theories may be openly debated and evaluated which are not foundation dependent.
4. The Problem of Excluded Knowledge.
The exclusiveness of the foundational positions also involves rejecting much of what is on offer as knowledge which is not compatible with the foundation. What is discarded, which of course differs for each foundational position, may include a rich variety of material which is important for understanding the area under consideration. The theorist goes fishing with a net having a certain mesh, and everything which does not fit the mesh swims off into the unknown. It may even be that what is caught is not fish, but only objects with a certain size square head… This narrowing of the legitimate corpus of knowledge is actually one of the most serious consequences of adopting foundational positions. It suggests the task of incorporating the material which has been ignored may be a big one. At the same time competing methodological nets cannot be taken as plural contributions to a broader methodology because they are at root incompatible; they claim to be fishing in different seas! The problem of what dimensions of knowledge each approach excludes is therefore an important analytical task.
However, there is a process whereby the prescriptivism of the foundational positions can be used to identify the kinds of knowledge which are excluded. The richness which is available is not the sum of theories available from the different foundational positions, but it can be uncovered by reversing the constrictions which have been imposed by the exclusive positions and then constructing good theory. At the same time, however, ways of judging falsity, accuracy, interpretation, scope and location of theory are needed and must be provided on some other terms than those which foundationalism presents.
5. Nonscientific and Scientific Knowledge.
Another problem arises from the division which foundationalism creates between experience and scientific knowledge. There is a division between people’s day to day experience, plans, values, reactions and assessments and the understanding which foundational views of science claim to offer. The latter approach deliberately cuts itself off from values or ideology as part of the response to the first crisis examined earlier. This engenders a basic division. Scientific theory, validated in its own terms, is a form of knowledge which has a different reference point from experience and cannot be expected to guide, inform or reinterpret that experience. Specifically the purpose and terms of reference of Consumption Theory are different from those of Marketing, Sales, Advertising and Consumer Education. It aims to present pure theory, data summaries, proofs of arguments or causal analyses which are valid in terms which are independent of the lives of the subjects of its study. This theoretical or scientific autonomy lives in tension with the polytechnic or pragmatic areas of study, which teach people how to respond to their daily experience. Because the division into the two kinds of knowledge occurs at the root definition of the academic enterprise, integrating the two kinds of knowledge often becomes difficult. (Storkey 1986 110-6)
This prevents two valuable kinds of relationship. First, the experience of those involved in the area of theoretical study is a rich potential source of knowledge. Yet consumers actually contribute little of their experience and understanding to many of the theoretical perspectives which are examined in this study. Conversely consumer theory is not available to guide and inform consumers about this area of their lives, nor does it consider this to be one of its tasks. Why this is the case and the remedies for these two lacunae will be considered later.
6. Otherworldliness.
The autonomy of foundational knowledge in its forms of validation means that it is cut off at the root from daily living and experience. It has another reference point. Thus in economics the focus is often theoretical equilibrium, consistency, closure or elegance. This is not merely a product of theoretical abstraction, but arises from positing an independent scientific gnosis and operating within it. When knowledge is formed in these terms, it points to its gnosis and away from the world it purports to examine; it is otherworldly. Its point of reference is not that which it seeks to know, but that which constitutes valid knowledge. Hicks, by no means outside the problem himself, has described the effect on growth theory in the following terms.
For with every step that we have taken to define this Equilibrium model more strictly, the closer has become its resemblance to the old static (or even stationary) Equilibrium model; its bearing upon reality must have come to seem even more remote. It has been fertile in the generation of class-room exercises; but so far as we can yet see, they are exercises, not real problems. They are not even hypothetical real problems of the type “what would happen if?” where the “if” is something that could conceivably happen. They are shadows of real problems, dressed up in such a way that by pure logic we can find solutions to them. (Hicks 1965 183)
This problem of otherworldliness is replicated in different forms in consumption theory. The existence of the demand function becomes just such a problem area. There are a number of demand functions which satisfy the requirements of different epistemological positions, and the literature is full of discussions of these which do not touch what people do when they are shopping. Many issues in consumption theory, like the convexity of indifference curves, continuity of preferences and corner solutions are conceived within theory rather than with prime reference to what is being studied. The locus of reference is the gnosis rather than consumption.
This is no small matter. If consumption theory has its own reified terms of reference, then it will sooner or later be discarded by those directly involved with the subject matter. Already marketing, advertising, consumer protection and advice agencies, credit institutions, those concerned with consumption ecology, retailers and others treat consumption theory as of little relevance to their concerns.
The problem is rooted in the autonomy of theoretical thought, and the failure to deliberate over the gegenstand choices which are being made. It is important not to confuse this failure with the theory/praxis dualism growing within Hegelian and Marxist thought, which tends to have as its concomitant the devaluation of theory. Here it is not Reason which is autonomous, but theoretical analysis; some of these frames of reference claim to be very practical, like Becker’s utility maximization. Uncovering these points of autonomy and showing the failure to discriminate in the theoretical moves undertaken is another of the positive tasks which requires addressing at the end of this study.
7. Economic Boundary Disputes.
Another foundational problem arises when the positions define the boundaries of the discipline in contradictory ways and engender conflicting understandings of the scope and identity of the discipline. Thus formal and logical rationalists tend to see economic science in terms of pure theory surrounded by a penumbra of applied economics. They operate with fairly traditional and autonomous disciplinary boundaries. Means-ends rationalists obliterate the pure/applied distinction in favour of a boundary definition including all efficiency orientated behaviour, which can extend into social, political or psychological activity. Positivists tend to focus on observables like exchange or monetary transactions to define the discipline, and causal theorists have difficulty with any tight boundary. Because consumption theory is often seen as standing on the edge of the domain of economic theory, itself a peculiar and interesting idea, it is touched by these differences. We shall see how ambiguous its position actually is within the competing foundational positions.
The problem arises from the way in which epistemological and methodological positions are used to define the boundary of discipline and subdiscipline. If, for example, consumption theory is identified with the logic of choice of goods, what happens after choice, the actual utilisation of goods, is excluded from the subdiscipline, and also substantially from economics. Or if it is seen in terms of observable transactions, the economic strategies and motives which generate those transactions are largely seen as intruding on the analysis of relationships between the key variables. One way out of the problem is to focus boundary definition more fully on what the subjects give priority in their orientation to events and actions. The mode of address responds to the subject matter rather than to the methodological prerequisites of the foundation adopted.
8. Disciplinary Autonomy and Imbalance.
The problem is wider, because each epistemology posits different relationships among the disciplines. Logicists tend to see psychology as extrarational and therefore not capable of formulation as knowledge, while behaviourists see psychology, viewed in behaviourist terms, as firmly located close to the centre of economic analysis. Clearly, there is the potential for all kinds of distortions among the disciplines.
One is the process whereby the mode of analysis of one discipline is given a definitive or controlling role within the analysis of another. It is possible to reduce economic analysis to mathematical, logical, psychological, social, political or kinetic terms. But buying a pair of shoes is not reducible to a calculus, the interaction of forces, or the resolution of feelings. The integrity of the economic act needs recognition together with the subsidiarity of the modes of analysis coming from other disciplines. Yet often the foundational epistemology dwells in other disciplines, or is predatory from economics on other disciplines.
Another form of distortion occurs when the prescribed theoretical mode is formulated autonomously within the discipline in a way which allows no contact with other aspects of human activity. Then the relationship between the disciplines is seen in terms of tectonic plates which rub up against one another only in interdisciplinary studies, but otherwise are isolated continents of meaning. There is no theoretical point of reference between economics, psychology, sociology and other disciplines. This contrasts with the intimacy which occurs between economic, psychological, logical, and other aspects of our actual behaviour. Buying the bunch of flowers involves an aesthetic judgement, a social relationship and an economic calculation more or less at the same time. Rather than just relating to the economic mode of analysis which the foundation requires, it may be that theory should translate to other modes. Ecological consciousness has made us aware of the intimacy between consumption, geographical and biological analysis and show the dangers of autonomous disciplinary theory.
9. The Subjectivity of the Economist.
Finally, the position and faith perspective of the economist tends to be obscured when a foundational position is adopted. It cuts economists off from a recognition of the way in which their values, culture, professional location and worldview mould their theoretical work. It requires them to ignore the reasons working within their minds and hearts which generate categories, research agendas, and arguments. This study will be full of the Lausanne School, Cambridge England and Cambridge Massachusetts, Austrian economics, the Chicago School and other geo-cultural positions which have often issued in disputes and misunderstanding. If this pseudo-neutrality was recognized, economists could relate their perspectives in a contributory way. Potentially this is a source of strength and creates a more healthy basis for theoretical communication within the discipline.
The subsequent analysis will follow through the implications of foundationalism at these nine problem levels, not in a neat, predetermined way, but as they are seen to be of significance. Repeatedly at least some of these levels of failure will be evidenced in foundational positions.

The Early Development of Consumption Theory.
The subsequent analysis will explore ways in which foundational views of theory have influenced the development of consumption theory. The influence seems particularly clear in this subdiscipline of economics, partly perhaps because its formative development occurs wholly within the era of foundationalist thinking. Here, in order to prepare for the later analysis, we look at the initial, slow growth of this area of theory.
The relationship between popular and academic awareness is very interesting. Popular awareness of consumption issues during the 19th century was strong. Britain was already known as a nation of shopkeepers to Samuel Adams and Napoleon. Especially with servants, domestic accounting was important and much talked about from Mrs Beeton onwards. More surprising was the Anti-Corn Law League, formed in 1838, as an urban consumer movement concerned about the price of bread, and concluding that trade restrictions were responsible for its high level. It was an articulate, nation-wide movement with detailed theoretical backing. On December 21, 1844 the Rochdale Co-operative opened. By 1851 there were over 150 substantial Co-ops which soon moved into manufacturing (1854) and wholesale buying (1863). This vast movement thus extended from a consumption base to penetrate the productive organisation of the British economy. (Kallen 1936, Webb 1920, Carbery 1969) In 1851 the Great Exhibition at Crystal Palace was an international display of the consumer goods which Britain was generating. At all kinds of levels therefore, especially in the classes of people who were generating economic theory, consumption was a substantive reality which generated much immediate interest.
By contrast classical theory seemed slow to respond. Much has been written about the bias of the classical model towards production and its relative deprecation of consumption. If urbanization meant a substantial movement away from subsistence forms of production towards purchasing, and if the new needs of the urban populations decisively shaped the development of industry and commerce, why did demand and consumption only have a desultory role in the classical model? It appears in Mill’s Principles in a very subordinate role. (Mill 1848 37-9, 51-2, 436-50) He qualifies consumption as productive or unproductive; it is the former only if it is going to productive workers to help them do more work. He further insists, almost perversely, that consumption expenditure cannot increase the demand for labour, but only redistribute it. (79-90) Markets and purchasers were everywhere evident, but were obviously not considered significant within the economic framework. Why was this the case?
One explanation is epistemological and offers an introduction to the kind of themes we shall be examining later. The underlying epistemology of the classical tradition was causal; it is examined in more depth in Chapter Four. The Inquiry into the Nature and Causes of the Wealth of Nations gave primacy to production from which followed consumption. The antecedent is used to explain the consequent through laws. This pattern left consumption as the effect. Sometimes, the consumer was seen as cause, as with Malthus’ consideration of “unproductive consumers”, but largely consumption was a dependent variable, the result of the system. Viewing consumption as consequence meant that the independent, purposive, directional activity which could have been associated with it had to be put aside. Jevons, in changing this pattern, felt he had to fling aside the English School and espouse the French. (Jevons 1871 xliv-v)
The change which allowed a consumption perspective to develop had, however, its own particular bias which was important to the developments described in the next three chapters. It was the new subjectivism which was evident towards the end of the century. After the aggressive macrotheory of Hegel, Comte, Marx, Mill, Darwin and Spencer there was a new intension of thought evidenced in many areas. (Hughes 1979 33-89) Neo-Kantian dualism emphasised subjective choice and ethics; Lange, Kroner, Windelband, Rickert, Dilthey and others all wrestled with this, and it later emerged in the emphasis of Troeltsch, Meinecke and Weber on subjective meaning. (Willey 1978 83-181) In Vienna Brentano developed an important awareness of intentionality, influencing both the Austrian School of economists and Husserl. Wundt and Freud developed within the same new awareness. (McAlister 1976 80-175) Sidgwick, Green and Bradley reflected a similar trend in England. Although Jevons broke through to a robust post-Benthamite view of choice, it is this subjectivism which filters through into consumption theory towards the end of the century. The individual’s construction of meaning is significant in its own right in purchasing decisions, rather than just being classicist consequence. Later, this theme will be examined in more detail, but again the epistemological changes seep into the formative period of consumption theory.
It seems therefore that, like the marginal revolution, the emergence of consumption theory was less uniform than the Schumpeterian interpretation has suggested. Jaffé’s conclusion on Walras is clear: “It cannot be emphasised enough that what Walras was after was the completion of his competitive market model, and not the elaboration of a theory of subjective valuation in consumption.” (Jaffe 1983 315) However, Menger’s concern with individual valuation left him more open to develop consumption theory, and Jevons had to make a major change of frame in order to move to his new view of economics. Although marginalism was part of the framework, the key change was the location of his basic theory of value with the individual. As his Journal shows, he suddenly flipped from a labour to a consumption theory of value within a period of 16 days in 1860; a move like that only occurs when there is a major conversion in perspective, but it also does not usually happen at the underlying point of a person’s thinking. (Jevons VII 120, LaNauze 1953) As Blaug points out, “there is more marginalism in Ricardo than Jevons or Walras.. it took twenty or thirty years to complete…it was not a marginal utility revolution and it did not happen in the 1870s.” (Blaug 1985 306-7) None of this suggests a decisive breakthrough.
In Britain the change to consumption theory was subtle and long-term. It required the development of theories of consumer discrimination, which in turn required an action or choice frame of reference. Bentham’s formulations had been largely exorcised from economic analysis by Mill, for whom choice was subjectively real, but not causally significant. The ethical limitations of the Benthamite position stood in the way of a more choice orientated frame of reference being developed. Only with Sidgwick did this possibility occur. In The Methods of Ethics (1874) Sidgwick looked at egoistic hedonism, a priori intuitionism and utilitarianism as three different methods of ethical argument. He thus drew back from the dogmatic claims of Bentham towards a more behavioural and relativist ethical position which allowed that individual choice could be premissed on utility, duty, benevolence and impartiality and studied with greater detachment. All of this was vital preparatory work for Marshall, who more radically moved over to consider human action and decision making as the raw material of his analysis. Laws were to be defined as regular patterns of activity which could be premissed on moral, collective, altruistic and selfish activity. Sidgwick’s position allowed Marshall to be open to the ethical aspect of activity and to treat the decision-making processes involved in consumption with rigour. By separating psychological and ethical hedonism, and establishing the independence of ethical propositions, he also gave Marshall the scope of an ethical position and room for direct economic analysis. Jevons relationship with utilitarianism was a good deal tighter and more problematic, because he was more closely tied to the Benthamite inheritance.
Yet the main point which arises from these considerations is much more straightforward. When a particular branch of theory develops affects how it develops. The early development of international trade theory and production theory, and the late development of income theory, have effected their concerns and articulation. Consumption theory seems to have developed slowly in relation to the phenomena it studied and had no classical orthodoxy to which to appeal. It was also formulated when a concern for the professional status of economics was dominant. The scientific status of the discipline, independence from ideological contamination, the autonomy of the domain, incontrovertibility and the need for an epistemological foundation therefore featured strongly. Because these, and especially the last, shaped the development of consumption theory so decisively, it is to examine them in detail that we must now turn.

Foundationalism and Consumption Theory.
We are now have the tools in place for the subsequent analysis. In the next three chapters evidence is set out that consumption theory during its formative period has been decisively shaped by foundational epistemologies. In their professional concern with well-founded knowledge economists constructed consumption theory according to theories of knowledge which have subsequently turned out to be invalid; each version has tended to create dogmatic theory and produce results of limited usefulness. The task now is to describe the different kinds of foundational consumption theory and examine the consequences of this way of constructing theory.
Yet the analysis cannot be too neat; in the transmission of the idea of foundationalism there was no one pattern. Some, like Pareto, Weber and Robbins reflected directly on epistemological issues and developed their own responses to them. However, in Anglo-Saxon circles this was not very common. Others developed their approach to consumption theory in the light of the broader popular conception of science about at the time, using the ideas of philosophers of which they became aware. This often involved a time-lag in the transmission of positions in the philosophy of science of a decade or more. Others related to a tradition of economic thought which already embodied a foundational view and articulated their own position in relation to it. Others adopted two incompatible positions without realising it and tried in a variety of ways to effect a compromise. The description will not therefore always work neatly through from the epistemological base to the consumption edifice. Yet whatever the route of transmission, the concern for the development of neutral, value-free, economic science has been so great that the ground plan can usually be read from the finished building.
In the next three chapters we shall be looking at nine positions which have had a great importance in twentieth century consumption theory. They can be grouped into three broader categories, rationalist, positivist and causal, which are the organising categories of the next three chapters, but within each of these broader classes it will be evident that there are substantial disagreements.
Modern foundational rationalism is examined in Chapter Two, where it is shown to have three powerful developments in consumption theory. The first is associated with the rebirth of pure logic, hence the term Logicist, in the mid-nineteenth century. The figures who led this development into British economics are Boole, De Morgan and, obviously, Jevons. Although it was a powerful influence in economics with Edgeworth and others, and in philosophy in Britain, it was to some extent eclipsed in the early decades of this century. However, Pareto pioneered a similar development at Lausanne, and it was through his influence on Hicks and Allen especially that logicism most directly impinged on modern British consumption theory. Within this tradition consumption theory is seen to be founded on the logic of choice. The second tradition had Kantian and Cartesian roots and aimed to establish the synthetic a priori framework for economic knowledge, including consumption theory. This tradition is described as Formal Rationalism and its pioneer was Walras, who despite his philosophical naiveté drew on his father and Cournot to create a new understanding of economic knowledge. Its main aim was to create a consistent analytical framework which provided the basis for all pure theory. It was transmitted via Schumpeter to Samuelson who used it in the Foundations of Economic Analysis. Within this perspective the focus of consumption theory is the stability conditions which are the sine qua non for the whole rational system.
The third position grows out of the neo-Kantian epistemological debates in Germany at the end of the century and is best described as Means-Ends Rationalism. Its rational focus is not systemic, but takes as its foundation the instrumental calculations made by the purposive subject as a universal process within economics which can be studied objectively and neutrally. A key transitional figure was Max Weber, but it was fully brought within the corpus of economics by Robbins and into consumption theory by Gary Becker.
Four positivist developments are examined in Chapter Three. They occurred around the beginning of the century, despite earlier roots, with a strong foundational programme developed initially in relation to the philosophy of the natural sciences by the Vienna Circle and other groups. The first position, which is designated Crude Positivism, argues that the key conceptual constructs of consumption theory can be obtained directly from the data. From the early philosophical positions of Schlick, Wittgenstein and others this is transmitted via Bridgman’s operationalism to Samuelson in his Revealed Preference mode, although by this time the position had already been destroyed philosophically.
The second position is Logical Positivism, developed by Carnap and others which identified protocol statements and logic as two independent epistemological bases. (Suppe 1977 6-27) From the Vienna Circle it spread to be quite influential after the Second World War, and Samuelson in a later epistemological reincarnation was also affected by it.
The third retreats somewhat from the total claims of crude positivism by dissociating theory and data; It establishes its foundation in the process of testing the former against the latter. One version is described as Hypothetical Positivism. Hempel and Oppenheim develop it philosophically. Another is Popper’s Falsificationism. Hutchinson was a key interpreter of it to economics, and Friedman was its strongest early proponent within consumption theory. The “F-twist” debate between Friedman, Samuelson and others is seen as largely a clash between the second and third positions in this positivist tradition, and the failure in mutual understanding is explored in these terms.
The fourth tradition can be called Frequency Inductivism, or Probabilistic Positivism and grew out of the work of Carnap, Reichenbach and Richard von Mises. It issued in the kind of inductivist theory developed especially in econometrics, where data is used within an inductive predictive framework. Here the belief was that probability theory would provide a secure basis for coming to conclusions which, although not certain, were scientifically reliable within probabilistic terms.
In Chapter Four a tradition which has been less fully identified than the positivist one is examined. It goes back to the Enlightenment and nineteenth century reformulations of naturalism, and is also evident in the classical economic concern with causes. But it receives a new foundational emphasis, first in the epistemological work of John Stuart Mill, who espoused a form of Causal Determinacy. Although his economic theory remained largely classical in conception, Mill’s System of Logic moved over to a foundational method in the social sciences, which was taken up by Marshall, W E Johnson, John Neville and John Maynard Keynes into some of the most significant developments in consumption theory. This epistemological position is the defining characteristic of the Cambridge School in the 20th century.
Another expression of causal analysis was Behaviourism, which focussed on experimental procedures in the mechanical sciences and saw them as definitive for scientific method in psychology and the other social sciences. Behaviourism is examined in consumption theory through the work of Scitovsky, who moved from the logicism of Welfare and Competition (1952) to the psychological behaviourism of The Joyless Economy. (1976) Finally, the work of Katona and the Michigan School is examined, to show a tradition which espoused causal methodological research but moved beyond foundationalism in many of its approaches. It was able, for example, to be open to the sociology and psychology of consumption.
Although these categories are not exhaustive, they do include most of the foundational emphases which have shaped 20th century consumption theory. Another example of a causal foundational position is methodological Marxism, but this has not become so direct a contributor to consumption theory in the West. By examining these positions and the way they have been articulated into consumption theory we shall be able to see the scale of the problem created by this approach. There is an historical dynamic to these developments. One of the recurring themes is the attempt to eliminate problems created by earlier foundationalist positions. Consequently, there is a retreat from the dogmatism and assertiveness of the early positions which is more marked especially during the last decade or so. Nevertheless, the underlying problem is not thoroughly understood and stays gripping consumption theory throughout its modern history. Only when we have travelled through these vicissitudes will a more promising way for consumption theory become evident.

Chapter two: Rationalist Foundations in Consumption Theory
Introduction.
The main argument of the following three chapters is that the content of various kinds of consumption theory is deeply structured by the epistemological precommitments of the theorists, whether they acknowledge it or not. This chapter traces the impact of the positions which can be described as rationalist; they may be broadly defined as ones which see the construction of mental categories as basic to understanding. Three traditions can be identified, not because they are the only philosophical positions which exist, but because their impact on economics is the most evident. They are Logicism, Formal or A priori Rationalism and Means-Ends Rationalism. In each of the sections sketching these positions a number of tasks will be undertaken. First, the epistemological basis of each position is outlined together with some of the consequences which follow from it. These include some of the structural problems outlined in the previous chapter – foundational dogmatism, theoretical prescription, otherworldliness and so on. Second, the longer term historical development of the positions is opened up so that the modifications of each view can be explored. Third, the implications of the positions for the consumption theory of different economists are identified, and fourth the problems which each of these views of theory generate are analysed. Since the story of each tradition is different, there is, however, no one neat way through each development.
In this chapter, as in the later two, we find the epistemological differences have many of their roots in national philosophical traditions. The Logicist was mainly British and Italian, the Formal, French and the Means-Ends tradition Austrian and German. Yet part of the story is the way these traditions travel through migration, personal contact and scholarly conviction. Sometimes, indeed, there is no obvious pattern of transmission, but a situation where different theorists arrive at similar conclusions because it is one of the most attractive options in the face of an underlying problem. This is effectively the conclusion which these three chapters adduce. Foundationalism was not a response which was always directly propagated, but one which sprang up at a lot of different points as a seeming solution to problems in substantive economic theory.
Although the focus of this chapter is Rationalism, this too is more problematic than is often assumed. It is easy to forget how recently the conviction has grown that human rational understanding is the key to knowledge. In the not too distant past revelation, experience and various kinds of wisdom have been seen as central. The recent dogmatic weight of rationalism has obscured how tendentious the idea of rational understanding might be to other ways of understanding. It is important to recognize that the character of the basic positions described in the next three chapters is not of theoretical options which can be used as alternatives, but of epistemological empires which claim prior allegiance. Just as the Marxist-Leninist philosophy of Russia and Eastern Europe deliberately outlawed most forms of rationalism, like for example the Polish tradition of logic, so, albeit more politely, rationalism has also engaged in patterns of intellectual conquest. It has often been assumed by rationalists that the territorial gains are justified, but whether intellectual conquest is fair is a more difficult question, and we now have, so to speak, the advantage of post-colonial hindsight. This chapter will be probing whether the rationalist empire which has undoubtedly grown has always understood how it has fought or even what territory it should own.

Rationalism as a Foundational Epistemology.
The Greek tradition of using reason to discover the truth was reintroduced during the Renaissance and spread again throughout Europe, reinforcing earlier scholastic patterns which had the same origin. Over a period of time it hardened into a doctrine which can loosely be called rationalism, suggesting human thought was the central and authoritative source of understanding. The scope of this movement was vast, influencing the culture of Europe deeply. However, our concern is more limited. It is with the rationalism which took a more epistemological turn, which implied that mental categories were the necessary basis for constructing valid knowledge. This approach had a patchy development throughout Europe. The impact it had in France, especially through Descartes, was considerable. In Britain the tenor even during the Enlightenment was largely antirationalist, favouring Common Sense, Empiricist and Naturalistic forms of understanding. In Germany there was a strong tradition including Leibnitz and Wolff before Kant, but he became the great impetus to this kind of thinking throwing up waves of development through the next century and a half. In this chapter, therefore, we are looking at three quite dissimilar patterns of origin for the forms of rationalism under examination. In Britain the philosophy of logic and logicism grew in quite hostile soil largely in the 19th century. In France strong Cartesian roots fed a formal rationalism which was also influenced by Kant in the mid 19th century, while in Austria and Germany post Kantian epistemological developments fired another foundationalist form. Let us look at each of these developments in turn.
In Britain, although much thinking was rationalist in a broader sense, there was little epistemological rationalism either in philosophy or economics. British classical economics espoused a naturalist style of reasoning about the economy; it made sense of it as a natural system, the mechanism of which could be uncovered by reasonable men. Rational principles or logic were not important as the source of understanding, but only a kind of systematic common sense. Adam Smith believed there was a natural order holding together the motives and transactions of the economic system which understanding could uncover. The Utilitarian tradition opened up an optimistic expression of rational individualism; Bentham’s calculus presumed to show how individuals acted and thence to provide the deductive key for wider economic policy, but it was not strongly developed. Despite James Mill’s advocacy of Bentham, Ricardo did not really move away from a Smithian framework. (Hollander 1985a 1-46) Other economists like Malthus were critical of this tradition as imposing too much of a rational order on the working of the economy. (Pribam 1983 169) John Stuart Mill continued Ricardo’s emphasis on a given natural order of the economy. More seriously for our concerns in The System of Logic Mill hijacked the meaning of logic to make it into a causal crunching machine which had no rationalist point of contact at all. (see ch 4) Although this emphasis was shaken by Jevons, it was not really challenged while Marshall was the dominant influence in British economics. Rationalism therefore remained much on the fringe of British economic theory and philosophical thought.
The tradition which did develop in British philosophy was marginal. The place of Hamilton in the Scottish school of philosophy exemplified this weakness of rationalism in Britain. He was quite an isolated voice, out of tune with Hutcheson, Hume, Smith, Reid and Stewart and one of the few in Britain to come to grips with the Kantian school. Yet he did establish more than any other at the beginning of the 19th century a concern with the philosophy of logic. (McCosh 1975 415-60) This discipline, which he, De Morgan, Boole and others developed largely in terms of analysing the limits of what could be determined by logic, was taken up by Jevons in an entirely different sense. Only in the 1860s does this foundational meaning arrive, and then dramatically in Jevons work. For him it became the basis on which scientific knowledge was constructed. A similar understanding of logic then influenced Pareto and Pantaleoni through whom it came back to England in the work of Hicks and Allen in the 1930s who very self-consciously imported it. Throughout this period logicism had been largely peripheral to the development of economics, and only through Hicks in post World War Two textbooks did it become orthodox consumption theory.
The other two forms have many of their origins in Kant. It is worth even at this stage dwelling on how deep a foundational emphasis he had. This is clearly evident on the opening of his Berlin Academy Prize Essay of 1763, “The Only Possible Basis of Proof”, where he states the following.
The question posed is of the sort that, if it is properly solved, higher philosophy will have to take on a definite form. If the method by which the utmost certainty in this sort of knowledge can be attained is established, and the nature of this conviction is well understood, an unchangeable methodological rule will necessarily aid thoughtful minds in similar endeavours, in the place of the everlasting instability of opinions and schools, just as Newton’s method in the natural sciences transformed the confusion of physical hypotheses into a sure procedure guided by experience and geometry. (quoted in Cassirer 1981 66-7)
This change away from Wolffian metaphysics to focussing on the underlying rule by which understanding is established in human thinking was fundamental to most subsequent rationalist foundational positions. Kant’s influence spread in geographical circles rippling out over a period of a century or so, reaching its greatest influence in France and Britain at the end of the 19th century. At the edge of the expanding circle around the middle of the 19th century an epistemological form of Kantianism mixed with French deductivist and scientistic thought was absorbed by Cournot, Auguste Walras and Léon Walras. Just as Kant claimed to construct the necessary a priori categories of thought, so this tradition believed it could construct the necessary fundamental categories of economic thought. Léon Walras developed it into a powerful foundational approach to economics which shaped consumption theory strongly. Strangely its main apostle in the second half of the 20th century has been an Austrian, Schumpeter, who has suffused this approach through the English-speaking world and made it one of the dominant orthodoxies. We shall examine its influence on the consumption analysis of Samuelson, Wald and others working in the General Equilibrium tradition.
Meanwhile at the centre of the circle in Germany Kant’s epistemological rationalism gave way to the more metaphysical forms of Hegel, Lotze and Fichte which were dominant for many decades. In turn this way of approaching philosophy was attacked by a number of positions which questioned the centrality of the mind: the economic materialism of Marx, Hartmann’s emphasis on the unconscious will, Darwinist evolutionary thought and historicist views. They all undermined the idea that rational categories gave definitive shape to life. Although many of the idealist philosophers continued to exert influence, it was evident to the leading thinkers that metaphysical constructs had no compelling rational justification. Those involved in the Geisteswissenschaften focussed more directly on the kind of human understanding which is sought in the human sciences. They wanted a method of obtaining knowledge which took account of subjective motives and values and yet focussed on the logic of purposive action in relation to these values. Gradually statements of these concerns emerged from the work of Dilthey, Weber and Husserl in Germany, and Menger, Brentano, Meinong and others in Austria which were more subjective than the formal rationalist tradition. Rather than being concerned with a macrocosmic rational order, they delineated the way human beings organize their thoughts as a basis for action. Rational categories denoted what true efficient behaviour would be and therefore created the necessary foundation for analysis. There were hints of this foundational position in Brentano and Menger. We shall focus more on the transmission of these Austrian and German traditions to the English-speaking world through Robbins conception of economics as a science of rational means-ends calculations. From Robbins it received fuller expression in Becker and some members of the Chicago School who have made many contributions in consumption theory.
Thus, even in these traditions which could be loosely described as rationalist there were divergent visions of what constituted well formed knowledge and what kind of theory economics should generate. Let us therefore examine each in turn to see where their foundational faith lay.

Section 1 The Logicist Tradition.
Jevons and the Changing View of Logic.
The British understanding of logic grew out of the dominance of empiricism, which gave it a different flavour from the views coming from a more Aristotelian view. Bacon’s New Organon emphasized learning from nature, and both Locke and Hume, by emphasizing that impressions and ideas came from experience, left logic as essentially an empty, formal science. (Locke 1690 344-55, Hume 1739 180-7, 1748 40-53, Ayer 1980 60-1) Hume was clear that “relations of ideas”, or logic, was without dependence on “what is anywhere existent in the universe” (1748 40), or as Ayer puts it later but in the same tradition, “So far as logic is concerned, anything may produce anything.” (1980 60) This was the origin of the British idea of analycity, the tautology which was true in all possible worlds. Seen in these terms logic was a thin science with little impact on anything, and it is therefore not surprising that it had a difficult time within the British traditions of philosophy. This meaning of logic shaped much of its 19th century development, but there were some thinkers who pushed against this trend.
Whateley’s Elements of Logic (1825) marked a renewed interest in the subject, but Hamilton’s work in Scotland really opened up a new direction. He emphasized moving to an analytical logic which made propositions into equations and thus made subject and predicate necessarily identical. A few years later he criticized Whateley on the grounds that the other’s approach to logic was not scientific enough, because it still included important elements of judgement. Hamilton wanted logic to be the science of the laws of thought as thought. “Of the truth or falsehood of propositions in themselves, it knows nothing, and takes no account.” (Hamilton 1833 249) However, the emphasis of De Morgan, Thompson (1842), Boole and others was to move yet further in the same direction. They wanted logic to be a pure science, uncontaminated by any covert assumptions and therefore able to stand incontrovertibly on its own terms. This is the first introduction to the idea of purity which we shall see is important in the construction of foundational thinking.
Yet the development of pure logic received a massive jolt with the publication of Mill’s System of Logic, for it was operating on entirely different principles. Mill, whose position will be examined in more detail in Chapter Four, was largely the foil against which Boole and De Morgan took their stance. He had developed his position in opposition to Hamilton’s opening up of formal logic, because he wanted to retain it as a process of ascertaining inferred truth (Mill 1828, 1865, 1872 book 2 ch 3). Although he meant logic to be the Queen of the Sciences, the science of science itself, his approach to logic undermined its rational foundational intent, for he allowed inductive and a priori reasoning and even presuppositions like the law of universal causation to be part of its corpus. (Mill CW12 78-9) These left logic interwoven with beliefs about the world and dependent on external givens. De Morgan especially set out to purge the discipline of these elements and saw himself as opposed in principle to the approach adopted by Mill because it undermined the absolutely incontrovertible basis of logic.
De Morgan was a mathematician and logician, and probably the predominant influence on Jevons who adored him as a teacher. In his calculus course Jevons would have covered its application to geometry of two dimensions and partial differentiation. (De Morgan 1842 341-87, 709-36, and 1852) He would have all the technical equipment which he would need later for marginal utility theory. But more important was De Morgan’s overall conception which moved to another basic foundationalist tenet. He believed in logic as a method and was agnostic about its truth claims.
The study of logic, then, considered relatively to human knowledge, stands in as low a place as that of the humble rules of arithmetic, with reference to the vast extent of mathematics and their physical applications. Neither is the less important for its lowliness: but it is not everyone who can see that. Writers on the subject frequently take a scope which entitles them to claim for logic one of the highest places: they do not confine themselves to the connection of premises and conclusion, but enter upon the periculum et commodum of the formulation of the premises themselves. In the hands of Mr Mill, for example (and to some extent of Dr Whately) logic is the science of distinguishing truth from falsehood, so as both to judge the premises and draw the conclusion… (De Morgan 1947 46)
He saw the method as neutral, objective and pure.
Logic is the science, the act, the theory and the practice, of the form of thought, the law of its action, the working of its machinery; independently of the matter thought on… Logic is formal, not material: it considers the law of action, apart from the matter acted on. It is not psychological, not metaphysical: it considers neither the mind in itself, nor the nature of things in itself, but the mind in relation to things, and things in relation to the mind. (De Morgan Syll 117, 153)
Thus logic could be separated from all psychological and metaphysical issues and was not modified by the content of the terms and operations. This opened the way for logic to be seen as the indubitable formal framework of the sciences, the substance of which was given by the specific disciplinary content. When Jevons had absorbed this conception from De Morgan, he was then in a position to give it a stronger foundationalist expression.
It is worth at this stage considering the relationship between logic and mathematics in the thinking of De Morgan, Boole and Jevons. De Morgan saw logic as considering the laws of thought, while mathematics was concerned with the necessary matter of thought. (De Morgan 1860 184) Boole, however, despite his desire to formalise logic, moved slightly in a different direction, seeing arithmetic as the way of constructing a logical language of symbols, signs of operation and expressions of identity. (Boole 1854 58, 1855/6 232-3, Frege 1979 12) This assumed a more foundational role for mathematics. Although Jevons was unsure exactly what the definition and relation of mathematics and logic was in Boole’s thought, he was clear where he stood.
But I do not care to pursue the subject because I think that in either case Boole was wrong. In my opinion logic is the superior science, the general basis of mathematics as well as all the other sciences. (Jevons 1874 156)
Jevons was thus stronger than De Morgan in insisting that logic was the foundation of all science, including mathematics. Within the logicist tradition this remains a central principle; theory is basically logical, while algebra and geometry enter as developments or illustrations of that logic, and numerical, statistical or econometric analysis belongs to the area of applied theory. Although this point is somewhat obscured by Jevons championing of mathematical economics in the Theory of Political Economy, (1871 13) his view of the relation between logic and mathematics was quite clear.
I cannot assent, indeed, to the common notion that certainty begins and ends with numerical determination. Nothing is more certain than logical truth. The laws of identity and difference are the tests of all that is certain throughout the range of thought, and mathematical reasoning is cogent only when it conforms to these conditions, of which logic is the first development…. Number is but logical discrimination, and algebra a highly developed logic. Logic resembles algebra as the mould resembles that which is cast in it. Boole mistook the cast for the mould. Considering that logic imposes its own laws upon every branch of mathematical science, it is no wonder that we constantly meet with the traces of logical laws in mathematical processes. (Jevons 1874 I 153-6)
and
But logic treats of those principles and forms of thought which must be employed in every branch of knowledge. It treats of the very origin and foundation of knowledge itself…. There is in short something in which all sciences must be similar; to which they must conform so long as they maintain what is true and self-consistent; and the work of logic is to explain this common basis of all science. One name which has been given to logic, namely the Science of Sciences, very aptly describes the all extensive power of logical principles. (Jevons 1870 5-6)
Thus Jevons ascribed a triumphalist foundational role to logic in its relationship to mathematics and the other special sciences. His Pure Logic (Jevons 1864), with its distinction between quality and quantity had given him the first vision of this view of the sciences, and he carried it with him throughout his later work as the basic view of knowledge which he espoused.
It is worth dwelling on the change of outlook which Jevons experienced during this period, because it was something akin to conversion. He describes it thus in a letter to Henrietta in early 1859.
To attempt to define the foundations of our knowledge of man, is surely a work worthy of a lifetime, and one not excelled in usefulness or interest by any other. Why then should anything beyond my necessary moral obligations debar me from it? While I should never consent to sacrifice them, why should I care to sacrifice my own present ease and amusements? Why should I care for money, for fine possessions, for present name and position, or even for the real pleasures of Scientific study, while there is such an important and interesting work evident to me? Others will not for years know or appreciate my real purposes, but it is not to be expected that they should. (Jevons II 362 my italics)
Thus we see, explicitly, the foundational drive, and the place which logic had in it. Inevitably this vision shaped his economic theory. At the same time as he was reporting the new “true theory of economy, mathematical in principle”, he was also stating the crucial conclusion which spelt the end, for him, of substantive metaphysical philosophy. “The ultimate question of philosophy, that between idealism and materialism, is necessarily an insoluble one.” (Jevons 1973 I 410-11) Logic, based solely on the likeness of things, was to provide an alternative way forward. “I cannot disbelieve, yet I can hardly believe that in the principle of sameness I have found that which will reduce the whole theory of reasoning to one consistent lucid process.” (Jevons 1972 II 179, 185-6) The Principles of Science are but an extension of that basic vision matured into a philosophy of science.
Our immediate concern in the next section will be with the way Jevon’s logicist vision was taken up into his economic theory, but the later influence of this philosophic tradition is also important and is best sketched here. The position taken up by De Morgan, Boole and Jevons developed much more strongly at the end of the 19th century in the work of Venn, Peano, Frege, Russell and Whitehead. One of its aims was the incorporation of mathematics into logic. If it could be shown that logic is the foundation of mathematics, one of the greatest puzzles of human knowledge would be solved. Mathematics is merely a construction of logical consequences based on self-evident definitions; there was no need to recur to intuition, a priori assumptions, metaphysical ideas of number or to statements about the relationship between mathematics and the universe. Frege pushed this line of development forward in a series of studies which allowed definitional concepts to uniquely define specific numbers. (Frege 1894, 1969 203-250) In his view of logic Frege was especially keen to emphasize its independence from the subject and its purely assertoric nature. (Frege 1969 1-8, 185-202). He developed this emphasis by insisting that the referent of every sentence was its truth-value, and by making the latter the focus, its coherence was guaranteed. (Frege 1892 214-8esp.) The development of Frege’s approach in Russell and Whitehead’s Principia Mathematica incorporated the claim that pure mathematics could be developed from logic. Its aim was to construct a formal system of uninterpreted signs which was noncontradictory and which would generate the formulas needed for arithmetic. They were expected to be true in all possible worlds, not depending on independent axioms. It was the crucial test for the logicist foundational position.
Gödel’s exposure of the limitations of this position are well-known. (Gödel 1929-36) He showed, by mapping meta-mathematical statements about the formal system into the system itself, that if the axioms were consistent, formulas would be undecidable and the system would be incomplete. Yet we could also ask why this foundational hope fails? At this stage it is not possible to discover a full answer, but there are some clues. If a move to state arithmetic or other mathematical systems were possible which had recourse only to logical necessity, then those systems would be inherently analytical. Whatever the systems claimed to “represent” in specific applied situations as theorems would therefore have analytical form, and the claim to represent number and operations would be open to question at every point. Second, to introduce metamathematical statements which are making claims (at another level of statement and syntax) about the original statements is to ignore the significance of this new level of interpretation and the relationship it has with the original. The construction of metastatements is the construction of this relationship, and its intepretive significance cannot be reduced to logic. Third, the emphasis on formal and uninterpreted systems which have only analyticity means in the end that the signs and other components of the system have only the character of being posited, and they become in that sense arbitrary. As Quine puts it: “In the end it is perhaps the same to say, as one often does, that the laws of mathematics and logic are true simply by virtue of our conceptual scheme.” (Quine 1952 xiv) What is necessarily the case is, so to speak, neither here nor there. Thus the promise of an indubitable foundation is seen to founder into mere convention, and some of the internal contradictions which this approach generates begin to become evident. At the very point where foundational logic seeks to become unassailable, it fails, either because there is a chasm between itself and what it seeks to interpret or because other assumptions and intepretive commitments are necessary.
The later development of modern symbolic logic has actually undermined the foundational claims which Jevons’ and Russell’s programme involved. Initially it developed strongly within the analytic/synthetic divide which is characteristic of this position and which has had such a profound effect on economics. Ayer, although not a symbolic logician, popularised this position in England. (Ayer 1936 96-115) However, Lukasiewicz and others came to show the variety of systems which share the tautological form which is seen to be basic to logic. (see Haack 1978 for a presentation of this development.) Tarski tried to find a way out of the impasse by moving to a semantic framework where the metalanguage could both refer to itself and the object language, but the inability of this tautological framework to discriminate sense and nonsense showed the problems of this consistent metalanguage. ‘”Snow is red” is true if and only if snow is red.’ remains metalinguistic unless it is used as a criterion, when its validity becomes open to question like any other correspondence theory. (Tarski 1944 cf Popper’s claims 1965 223-8) Quine’s demolition of the analytic/synthetic divide similarly marked another departure from the view that pure analytical statements exist. (Quine 1953) Thus the more recent development of symbolic logic has recognized the compromised independence of the foundational view. No longer can logic provide a method which is the basic framework for generating knowledge, because the specification of the meaning of signs is far more complex and depends on so many other assumptions. Its greatest contemporary strength lies in the far more humble role which it has traditionally adopted of specifying the limits of what we can conclude from any propositions or signs.
Although modern symbolic logic is much more circumspect in its claims, the foundational impetus which developed philosophically was very powerful and has continued to influence economic theory, including consumption theory. We shall examine these developments later, but first we must examine how Jevons logicist vision played into his economic theory.

Jevons and Consumption Theory.
Jevons was obviously one of the founders of consumption theory; his The Theory of Political Economy (1871) jerked classical English economics out of a snoring indifference to consumption by developing a marginal utility theory of consumer choice and making it basic to his system. How, then, do we see his significance?
At this stage it is worth querying whether the interpretation of Jevons’ part in the “marginal revolution” as seen by Schumpeter, Stigler and others is correct. (Schumpeter 1954, Stigler 1950 307-27, 373-96, Black 1973 305-20) On this view the marginal revolution is seen as an important technical advance in the discipline which was discovered independently but roughly at the same time by Menger, Jevons and Walras with Gossen as antecedent. The discovery was then incorporated into the discipline and allowed the incontrovertible development of modern utility theory as neo-classical orthodoxy. However, Howey has pointed out that “marginalism” was a neologism of 1914 (Black 1973 15-36) and Blaug has suggested that the coherence of the revolution is less marked than is supposed and is related to the professionalization of economics. (Blaug 1985 294-327, Black 1973 3-14) Others see the change as more epistemological and related to the philosophy of science; Fisher has interpreted it in Lacatosian terms, which maybe uses a late 20th century agenda anachronistically. (Fisher 1986, Black 1973 59-77) We have already examined the way in which the crisis of substantive classical theory and the professionalisation of the discipline put pressure on economists to develop a scientific foundational approach. This study suggests that although the views which shaped the work of Jevons, Menger and Walras remained distinct in many respects, the coherence in their approaches grew out of the common need to develop a rationalist foundational method. (It might even be that the problems Jevons and Walras had with one another’s economics lay in the different kinds of rationalist epistemology which they espoused…) Let us begin to explore these possibilities in more detail.
How did Jevons’ consumption theory develop? By about 1860 he had been deeply affected by De Morgan’s views on logic, “the laws of action of thought”, and mathematics, “the necessary matter of thought”. He had been imbued with the notion that “inference has nothing to do with the truth or falsehood of antecedents, but only with the necessity of the consequences.” (De Morgan 1966 184 [1860]) But he, unlike De Morgan, made this idea central to his academic perspective and faith. During this time he had two major themes running concurrently. The first was a reworking of Boole’s Algebraic Logic (1847, 1854), which allowed inferences to be expressed by manipulating a “logical alphabet”. The study was published in 1864 as Pure Logic, or the Logic of Quality Apart from Quantity, and reflected Jevons conviction that logic was foundational to all thought. As he stated it later, “[the laws of identity and difference] are the prior conditions of all thought and all knowledge, and even to question their truth is to allow them true”. (Jevons 1870 7, Wood I 11-25, 167-187, 212-32, 251-67) With this key he was ready to unlock the other big challenge with which he was involved.
For at the same time, but at a more basic level, he was also working at Political Economy, with a nagging feeling that there was something wrong with the discipline and he was the one to sort it out. If we accept the diary date of early 1860 for the breakthrough and the paper to the British Association meeting at Cambridge in September 1862 as the early statement, this was the crucial transition. Why did Jevons base the whole theory of value on utility? The diary tells us that he moved in his work on Political Economy on February 3-5, 1860 from basing value on labour to a “true” comprehension of value, presumably basing it on utility, by February 19th. (Jevons VII 120, LaNauze 1953 357) This complete about-turn in his central substantive concept suggests consumption was less central to his concerns than the method he was employing. Indeed, he never really showed a detailed interest in consumption as a subject – what people bought, when, why, with what and for whom. Utility was a solution to a foundational problem, not the first step in opening up a subdiscipline.
By 1871 the theory with the mathematical functions which show the logic of consumption had been fleshed out. Using Bentham as his substantive philosophical base, he was able to describe economics as a calculus of pleasure and pain. (1871 vii) He may also have been influenced by the work of Macleod, who had a strong commitment towards a scientific view of the discipline involving a single central conception focussing on exchange rather than wealth. (Macleod 1862 215-30) Using a mathematical logic of utility he had turned the classical system on its head. Beginning with the theory of Pleasure, Pain and Utility and the logical relations which followed in Exchange, he was able to banish the classical concerns of Labour, Rent and Capital to the end of The Theory of Political Economy. It was an amazing revolution, not because it had explicitly included a theory of consumer choice in the main corpus of economic theory, but because it had created a different kind of economic theory which had an essentially logical form.
Jevons was quite precise about the relation of logic and mathematics in his theory. Economics as a science must be mathematical simply because it deals with quantities, and calculus provides the tool whereby the measurements can be scientifically worked through. But the logic of more and less, of discrimination, provides him with the basis for analysing utility; when the ratio of the intensities of the last wants is proportional to their prices, then equilibrium is established. Mathematics is necessary, but logic is basic. He is often charged with retaining cardinal, and therefore Benthamite utility, but this may be a failure to understand his logic, mathematics and economics. His logic had excluded all such metaphysical assumptions and merely differentiated more and less. His use of arithmetic was largely because goods and money were numerical, and if the utility calculations of more and less had in the end to be weighed against the relative prices of goods in the thinking of the consumers concerned, there was no special virtue in eliminating the cardinal utility merely to establish a foundation which was presumed free from metaphysical judgement but no better conveyed what consumers actually do. For consumption must do with number, because it always does involve numbers of goods and amounts of income. In this particular Jevons awareness of the place of mathematics in consumption theory, which later post-cardinal logicism lost, had some value; whereas consumer preference has universally fled into algebra, every consumer remains with the arithmetic of counting and accounting, and it is a loss to theory that this is not recognised. Cardinal utility may have no special virtue, but no consumer has yet escaped from cardinal prices and accounting.
However, the overwhelming drift of the position was foundational. Jevons was strongly convinced of the correctness of his formulations and did what he could to establish it as the new orthodoxy. The drive was to a scientism of logico-mathematical certainty which offered a perfectly general mode of reasoning. In The Theory of Political Economy other areas of theory are developed in only a sketchy way, but it is worth assessing what this way of doing theory implied.
One of its consequences was to cement a number of assumptions into this mode of theorising about consumption. For example, there was the individualism of the analysis. The logic of utility maximization was always predicated of an individual and summed into aggregate market responses, although Jevons failed to show how utility functions could be aggregated. In part it reflected the Benthamite origins of Jevons’ calculus, but, as MacLennon suggests, (Jevons/Wood I 260) he was not inside Utilitarianism, but just using it methodologically. The dominant reason for adopting it grew from his logical method; a person-thing logic could not cope with other-person referential frameworks and retain its status as an infallible foundation, because taking other people into account introduces values which are less than self-evident and spoil the conclusive outcome of the more-less logic. (Sen 1982 84-106) It was the method which required the assumption of non-relational individualism. It is only with the work of Sen (esp 1970), Collard (1978) and others that the arbitrary nature of this assumption has been fully exposed. Previously, it remained a tenet of the logicist tradition through Jevons, Edgeworth, Pareto, Slutsky, Hicks, Allen, Arrow and Lancaster. That this approach, allowing a neat determinate consumption map, ignored the social relationships involved in consumption decisions seemed not to worry theorists within this tradition. Logical closure was their concern.
Another assumption was the centrality of the margin, the logical margin, as the locus where decisions were supposed to be made. Yet many decisions are not made at the margin, especially in income-glutted economies where many people do not feel they have to engage in a marginal calculus, but just buy goods anyway. Of course, later logicist theory recognises closed preference sets and “bliss points”, but even this modification does not address the deeper issue. Purchases change people; they demand subsequent commitments and even set the terms for later patterns of substitution. Thus, for example, the new owner of an expensive car with high fuel consumption will probably relax his terms of substitution in relation to petrol, and the overeater will continue that way. This again is an important lacuna in the theory. Many of the most important consumption decisions do not take place at the margin, but much earlier, and a methodological system which is tied to the margin introduces its own bias. Again the centrality of the prescribed logic rules out of consideration consumption activity which may be very important.
It was similar with normativity. Foundational logic knew no norms, except those of utility discrimination; it was anormative and ethically neutral except with respect to the logic of more and less. It was this logic which circumscribed consumption theory and (other) normative prescriptions (for more and less are prescriptive) were banished to the end of the book, ready to be hived off into what later became welfare economics. Here then was the beginning of the chasm between pure theory and welfare judgements, with the latter always seen as ex post and external to the main body of theory. Since we shall observe a number of different is/ought divides, it is worth noting that this one is based on the logical impossibility of drawing normative conclusions from nonnormative premisses. The problem with Jevons and later theorists is the unjustifiable privileged status given to the “logic” of more and less; their chosen logic was impure, but claimed to be whiter than driven snow.
The framework also excluded any need to consider which actual goods and services were bought and consumed. The abstraction from any specific analysis of kinds of consumption was subsumed under the philistine idea that more of anything with positive utility is better. That the economics of consumption does not vary significantly with the items bought is one of the underlying assumptions of this frame, and of much modern analysis. (cf Norris 1941 107-8) Only the variations in the marginal utilities of the subjects buying need to be considered. It is curious to compare this abstract framework with Jevons’ very specific analysis of The Coal Question (1865) Here, of course, on the basis of extrapolation he comes to the conclusion that coal reserves will be totally depleted by the demand for coal by 1965 (261-74) It is understandable at this time that these two modes of analysis did not mesh, but it could be argued that this was the origin of the lack of engagement with specific areas of consumption having widely different characteristics which has marked later logicist theory. The rigid abstraction required by the logic of marginal calculations dictated that this was not relevant.
In conclusion, we note that the “logic” of more and less, which is, of course, a conventional logic, has imposed its terms of analysis deeply on the subdiscipline. The “logic” does not fit subjects who have finite needs and physical and cultural limits to their consumption, for whom enough is often enough, yet it has continued to dominate analysis and possibly define marketing strategies. It involves ignoring what people consume, consumer economizing activity aimed at reducing consumption, and the interdependence of goods and services in lifestyle. The presumed foundation in logic has allowed a particular partial and limited mode of analysis to claim to be the central theoretical construct of the subdiscipline. Of course, Jevons merely started out along the trail, but the foundational claims made this approach into an intellectual empire builder.
These may seen harsh judgements to bring to Jevons work, but they are not addressed to his oeuvre as much as to the direction of this mode of foundational analysis as a basis for consumption theory. Because the logical appeal was accepted as necessary and authoritative, other issues were sidelined. For Jevons the logic of choosing more and less prescribed marginal analysis as the method of the discipline. Because of this foundational vision, the shape of the subsequent theory developed and influenced others, especially Edgeworth, who was also beholden to Boole. (Creedy 1986, Edgeworth 1881, 1891) Without the foundational commitment other possibilities could have been theoretically significant and other paths taken, but they were not.

Pareto’s Consumption Theory.
It seems a big jump to move from an English logician and economist to a different tradition in Italy, yet Pareto was a logicist in a similar sense to that of Jevons. The meaning he attaches to the idea of logic is slightly different from that of the tradition of De Morgan and Jevons, but there were substantial links between the Italian and English schools. Pareto’s close fellow scholar, Pantaleoni recommended at the beginning of his text on Pure Economics, “the best training in logic for students of economics is supplied by such works as those of A. De Morgan, E Schröder, J Venn, W.S. Jevons, A Bain, W Wundt, M.W. Drobnisch, J.N. Keynes.” (Pantaleoni 1889/1898 7, Pareto 1984 I and II) Pareto’s route into logic was a severe alienation from the ideological developments in his own culture, especially socialism. He sought a kind of knowledge which was free from this contamination. Logic gave him a basis for knowing free from metaphysics and without subjectivity, which was scientific and impartial with respect to beliefs and opinions, and which was individualistic in its applications.
At the same time he was a scientist with a fully articulated framework for economics and sociology, and with a background which made him no pale reflection of English logicism. He received a good training in mathematics at Turin Polytechnic, working in his graduation thesis on “The Fundamental Principles of Equilibrium in Solid Bodies” (1869), and in one sense mechanics was his model of good science. He wanted a procedure which would achieve a science, pure economics, which was equivalent as a science to pure mechanics. (1896 I 2) This theme of following the procedures of the natural sciences was much stronger than in Jevons thinking, and it issued in a fully developed understanding of abstraction as a process which is necessary in the study of any concrete phenomena. This abstraction was notional, and simplified, but it was normal to all the sciences in the generation of theory. Pure theory was weaker in the social sciences than, for example, in celestial mechanics, because it corresponded to people’s behaviour only within certain limits. (1909 190-1). Thus, Pareto did not expect pure economic theory to be testable or to fit the facts, because it was only a first approximation to reality, and the process of abstraction required an awareness of the limits and stages of theory formation. (1916 II 27-9)
One way into the methodological issues was through consideration of laws in the human sciences; they were obviously different from mechanical laws. But the models which Pareto had for social scientific laws were not very satisfactory. For a while he was influenced by Comte, but Comte’s law of the “three stages” lacked any kind of rigour. (Tarascio 1966 30-55, cf Cirillo 1979) He was aware that theorists create their own categories. In this sense he was a “nominalist of the nominalists” and displayed a greater awareness of the limitations of theory than those who used his work. (IEP 1953 172-207) Nor did he believe in any metaphysical laws which are part of the nature of things. Apart from his contempt for the “metaphysicalists”, he also had a strongly worked out view of the relationship between economic phenomena and theory; the theory must arise from the phenomena, not be imposed upon it. Thus, it was not easy for Pareto, along with many others at the time, to decide what economic laws were. His answer was to claim they were logical in form, and his argument for doing so was quite idiosyncratic.
Pareto’s view of logic begins with the actual relations among phenomena, what he calls the facts. These “objective” relations, he argues, cannot but be logical in the sense that they must be consistent with one another, an appeal which Wittgenstein also makes in the Tractatus. (1921 6-17) This allows Pareto to believe that logic is not nominal theory, but something of central objective theoretical validity. However, these objective relations are only sometimes correctly perceived by human subjects. (Pareto 1909 31). Thus they may be either logical or non-logical. If they are the former, then they may mirror the objective relations; they deal with verifiable or real phenomena. Yet still they may only be logical in appearance and not in reality, especially if they deal in imaginary facts. In this case they are theoretically unsound and experiment and observation will in time reveal the error of the theory. Non-logical relations are metaphysical, mythological or self-justificatory in what they posit, and can be ruled out of the scientific arena (Pareto 1909 35). The relations posited by the subject which are non-logical Pareto described as sentiments; they are subjective feelings for which there was no obvious external referent. Included in this category were sentiments of morality, religion, law and custom. (Pareto 1909 38-82) What remains after these are excluded is understandings of the relations between persons and things which are objective, or logical, and which therefore work in daily experience, and this is the domain of economic science. It is objectively assessed, value-free, anormative and logical. Questions of sentiment and value are to be considered within Sociology. This is an astonishing move. By using logic in this global sense, as corresponding to reality, Pareto is able to say that logic is what people actually do in economic life including consumption theory. Having made the assertion, like Wittgenstein, he draws up the ladder he has just mounted and becomes fully foundational. The idea of logic is asserted as characteristic of human economic behaviour and the central method of economic science.
The sense that Pareto had of departing from the old metaphysical way of doing things into a new neutrality comes out most clearly in his debate with Croce. (IEP 3) The latter had an overtly metaphysical post-Hegelian position and is criticized by Pareto for not being scientific and introducing his own metaphysics. Croce’s response is a counter critique. To claim economics could be treated on a par with mechanics was, he argued, to ignore the valuations which people made in their activities which were necessarily metaphysical. (IEP3 1953 172, Croce 1966 661-70, 1949 140-52, 1913) It was not possible therefore to make a distinction between hedonism as a substantive philosophy and the scientific logic of economics; egoism and morality are two kinds of orientation to practical economic activity which variously shape what happens. Economic choices must necessarily involve valuation and judgement. Pareto’s position left no room for valuation. And, claimed Croce, this was a metaphysical postulate:
The accusation of being metaphysical will seem to you the last that could ever be brought against you. Your implied metaphysical postulate is, however, this: that the facts of man’s activity are of the same nature as physical facts; that in the one case as in the other we can only observe regularity and deduce consequences therefrom, without ever penetrating into the inner nature of the facts. (IEP3 1953)
Pareto’s response to this was to emphasize the process of abstraction which is involved in theory construction and to underline his own commitment to rigorous scientific analysis. But the point was not really answered; his method had ruled out arbitrarily the consideration of the way subjects orientate themselves to public behaviour. In the Manual Pareto’s position had hardened into one where valuation was included in the subjectivity of sentiments and therefore excluded from theoretical consideration. How was he able to do this?
The key move was, of course, the identification of economic actions with logical ones and the study of sentiments with sociology and politics. The latter are looking at the effect of events on the sentiments of the people involved. (Pareto 1909 82-3, Finer 1966 33-51) Economics is comprised of logical actions because repeated trials ensure that the subjective relations mirror the reality. Mistakes will be rectified and made logical. Thus although Pareto takes illusory behaviour seriously he concludes that it can be ignored in economics, while being fundamental to sociology. Yet whether the logic of marginal substitution has the converting power in consumption activity which Pareto claims of it is an open question; he seems to acknowledge the opposite.
Men follow their sentiment and their self-interest, but it pleases them to imagine that they follow reason. And so they look for, and always find, some theory which, a posteriori, makes their actions appear to be logical. If that theory could be demolished scientifically, the only result would be that another theory would be substituted for the first one, and for the same purpose; a new form would be used, but the actions would remain the same. (Pareto 1909 95)
He is in a bind. Either actually this sentiment influences and shapes all areas of behaviour and therefore needs to be taken into account in economics, which he will not admit, or the distinction is merely heuristic, allowing “logical behaviour” to be dumped in economics and the rest ejected to sociology. Pareto largely accepts the latter position; so his economic theory seems logical, objective and scientific, but on the basis of being in principle cut off from the way people behave. He establishes a logicist boundary to the discipline, which has subsequently been widely accepted, albeit on this faulty arguing. His own commitment to ejecting ideology from economic theory helped generate this arbitrary conclusion; what is surprising it is subsequent widespread adoption.
The next step was to develop a framework of pure economic theory, and Pareto does this around the notion of equilibrium. The movements necessary to reach equilibrium are “real” and logical; those which lead away are “virtual” and can be treated as incidental. (Pareto 1909 109) Thus a logical map is a ground for establishing equilibrium in the pure theory. The science of economics is to be regarded as pure in the sense that it establishes abstract maps of the relations among phenomena. Because the logical equilibrium was timeless, it was designated “static”. Movements of equilibria were dynamic. So the inner core of the logicist model was established. Pure theory was logical and static, because everything which was extralogical was treated parametrically. Finally tastes per se were extralogical, noneconomic and were part of the world of subjectivity and sentiments. Only the logical indices which map those tastes are a necessary part of economics.
Hence the entire theory of economic equilibrium is independent of the notions of (economic) utility, of value in use, or of ophelimity. Only one thing is needed, that is to know the limits of the ratios….(Pareto 1909 394)
He now had a framework purged of metaphysical concepts expressed in mathematical logic which was foundational and pure, despite the weaknesses of the process through which he arrived there.
But this attempt to create a pure foundation for theory had consequences for what could be called the deep structure of Pareto’s theory beyond his immediate concerns and the issues of which he was directly aware. One involved a new conception of the boundary of economic science and its relationship (or lack of it) with other disciplines. Interestingly, consumption is on the boundary, because it is the place where subjectivity of tastes and the objective relations of logic meet. Subjectivity consists of sentiments which sit outside the logic of choice as parameters. Thus consumption occurs in two stages, first, the given set of sentiments and tastes which are rooted in subjective desires, and second, the rational calculations which are part of the actual domain of economic science. The former is part of the general category of religion, metaphysics, ideology and sentiments which is based upon asserted but not verifiable understanding, and the latter is a neutral pattern of calculation which can be considered without the interpolation of values. Because other economic activity tends towards consumption, the latter is often conceived as circumscribing totally the boundary of economics with the social sciences. Actually, of course, the purchasing of goods is no closer to the edge of economics than work, investment or trading, yet this dogma of the boundary has become orthodoxy in many texts. Similarly, the identification of values with the study of society and of “objective logical relations” with economic science has become widespread. Thus the relationship between economics and sociology is defined in terms of logical and non-logical subject-matter, which cannot be considered jointly within the same methodological framework. Questions of welfare and interpersonal valuation are really extra logical and extra economic, as Arrow showed, reflecting this position. (Arrow 1951) Thus, the logical methodology is supposed to show what the edge of economic science is, but because this methodology is (necessarily) contaminated, it goes on to dogmatically define the content of the discipline and what is excluded. Values, ideologies and views of the world, however central to actual consumption decisions, are defined ab initio as outside the domain of economic analysis.
This position has also moved on from the individualism of Jevons to become subjectivist. Persons are excluded from the analysis. “The individual can disappear provided he leaves us a photograph of his tastes.” (Pareto 1909 120) Thus the person appears in consumption theory only through the tastes which represent her/his subjective sentiments; otherwise only the logic of choice remains. The subjectivity of tastes, and the fact that they were placed at the edge of economic theory, meant that many causal relationships were removed from consideration. James Mill, for example, had distinguished between productive and non-productive consumption, bringing out the point that much consumption was needed to allow productive activity to take place. Many people do spend a considerable proportion of their income getting, doing and keeping their work; mining estates usually have very good butchers. Yet because tastes have been reduced to subjective sentiments, it is not possible to consider these economic reasons for consumption or other careful strategies which are part of people’s orientation to consumption as part of economic theory. Tastes were beyond the realm of ordered logical analysis and had to be exogenous variables if the framework was to hold. For Pareto this depersonization was a great virtue in terms of supposedly constructing a neutral science; actually it impoverished the base from which theory could be formed.
The framework also established a special meaning to the idea of equilibrium. Pareto’s background in mechanics made him at home with the idea of equilibrium as the solution to problems, and his treatment of “obstacles” was almost Newtonian in conception. Pareto saw equilibrium as not only possible but necessary, because the logic of choice reflected reality and the indifference curves could exhaustively specify the domain. It consisted of the logical possibility of an infinite number of indifference curves, which were linear and continuous. Thus the whole fabric of equilibrium analysis rests on the logical necessity that the set of indifference curves can be resolved, which further rests on the assumption that the individual prefers a given combination of goods to any combination which is infinitesimally smaller and nothing else is relevant to choice. This gives the idea of logical equilibrium as the central concern of static economics its weight. Thus the only consideration of an unstable equilibrium revolves around the situation where two bargaining individuals do not have a unique logical outcome to their encounter. (Pareto 1909 142-5) Because logically unstable trading and maximizing positions are difficult to conceive within this framework, disequilibrium becomes an afterthought for this mode of analysis.
The artificiality of this justification for equilibrium as a behavioural assumption is evident if we stand back and question it in terms of daily life. In many areas price calculations do not enter into consumption decisions; people do not know the prices of goods they buy, will buy them anyway or have a range of indifference over price variations as high as 50%. If this is the case the “logic” of price and substitution effects is normally indeterminate and the possibility of equilibrium must include other factors like shortage of time, tastes and limited interest; the limitation of the idea of equilibrium to this logical categorisation is therefore a major category mistake. It also excludes from consideration another issue which is an important consumption variable; the extent to which people are indifferent to marginal variations. Inflationary processes may be partly understood in terms of a decline in price sensitivity of consumers. Yet the Paretian model gives a pre-emptive determination of consumer equilibrium which rules this possibility out of consideration. Later in the study we shall examine this indeterminacy more fully.
The logicist approach also meant that the conception of equilibrium was atemporal. Logical equilibrium was instantaneous. Although Pareto sometimes saw logic as the objective realization of the links between means and ends in human behaviour, he normally saw it as a system of preferences. The foundational characteristic of a logic of choice is that it should be supertemporal, it should be necessarily true at any period. Consequently, the theory is developed outside time which is brought in as an unrelated postscript. The definition of static analysis is that which is able to treat temporal matters in ceteris paribus terms. Pareto does not try to establish the process whereby equilibrium might be attained, as Walras did with his tâtonnement theory, but merely assumed that the logic was compelling. Dynamic considerations, like having reliable sources of supply, speculation, consumption strategies, life-stage consumption, the influence of mobility on consumption, personal investment policies, bargaining, situational gratification and the interaction of different logics of consumption are all pushed into the shady area of dynamic analysis, about which, as Pareto conceded, (Pareto 1909 104-5 ) almost nothing was known. (Pantaleoni IEP5 1955 26-57) Pareto’s method of incorporating time in his analysis was to combine successive (logical) equilibria, but only a few “notions” were available in this branch of theory. When he considers changes in tastes, Pareto has to say that the long-term indifference map may be very different from the short-term one. The necessarily static framework thus comes to dominate pure theory. If the basic equilibrium framework is atemporal, then any temporal elements which are introduced are extrinsic to the analysis. A parallel criticism can be made of the way this foundation cannot take spatial influences on consumption into account either. (cf Frege 1979 148 [1897]) The implications of making logic foundational to economic method and thereby excluding consideration of space-time are so considerable that it will frequently recur. We shall see Hicks in his relationship with Keynes struggling for decades with this issue.
This foundational position also raised the question of how theory should be validated. Here we note the importance of the pure-applied distinction, which became basic to Pareto’s and later logicist thinking. (Baumol and Goldfield 1968 52, 63, 83) Pure theory was assessed in terms of its logical consistency. Its validity was to be understood in terms of formal identities and inference, and it had the qualities of pure mechanics. (1896 II ch1) Testing was not an appropriate way of responding to this level of theory. It was only applied theory which was to be assessed in terms of its interaction with and correspondence to reality. Applied economic theory was like engineering However, this latter category had a number of problems in relation to Pareto’s overall framework which need to be made explicit. Pareto acknowledged that his theoretical constructions were only one way of looking at the phenomena.
When the results of theory pass into practice, we can be sure that they will always be somewhat modified by other results which depend on phenomena not considered by the theory. (Pareto 1909 11)
This later process was messy, but it was also residual, an afterthought to the main business of theory formation.. Fundamentally, economic behaviour will conform to the logic of theory, and the basic method of evaluating and establishing theory was a consideration of its logical integrity. Pareto believed he had established this, and he has therefore to be met by the central refutation that his logic was not pure because it asserted one conventional way of viewing consumption. Later critiques which bring post-Popperian views to testing to bear on Pareto’s thought are in part working with a different conception of theory testing. (Tarascio 1966, Cirillo 1979) It was in his own terms that Pareto failed.
Thus Pareto was able to construct a position which he believed was objective, logical and free from metaphysics. Although it failed to be these things, it was seen as offering a new way forward. As a result the consequences were adopted into traditions of theory as dogma. It excluded the consideration of interpersonal differences, the consideration of time and space, the influence of personal values and principles, independent personal reasons for consumption, strategies for the organization of consumption, criteria for purchasing and many other potential areas of theoretical development. Yet it gained enough support as an economic method to become a central tradition in Italy with Antonelli, Boninsegni, Barone, Valenti, Supino and Ricca-Salerno. It influenced the Russian economist Slutsky, and then among others had a deep impact on Hicks and Allen. It therefore spread a mode of consumption theorizing which was severely impoverished.

Hicks’ Logical Foundationalism.
Because this study focusses more on Anglo-Saxon traditions of consumption theory, the study of Pareto is in part an introduction to the work of Hicks and Allen. Harcourt records that “Hugh Dalton put Hicks onto Pareto because he ‘read Italian’, so that he was deep into Pareto before he got much out of Marshall.” (Harcourt 1986 17) Pareto had a formative role in the position Hicks developed, probably more so than Edgeworth, Johnson (EJ 1913 483-513) and, of course, Slutsky (Allen RES 1936 120-9). Hicks’ interest in Pareto was methodological (Hicks 1934, 1939 1-19), rather than philosophical or epistemological. Indeed, during this earlier period Hicks had little awareness in these areas, but merely abstracted from Pareto the technical apparatus of indifference analysis and the foundational conception which was lodged there. The power of this framework had an immediate appeal.
That there are a great many such extensions appears at once when we consider how wide is the variety of human choices which can be fitted into the framework of the Paretian scale of preference. What begins as an analysis of the consumer’s choice among consumption goods ends as a theory of choice in general. We are in sight of a unifying principle for the whole of economics (1939 24)
Moreover the foundation of the new method was logic. Hicks contrasts Keynes “superb intuition” with his own “pure logical” method.(1939 4) He saw logical analysis and the survey of economic institutions as comprising most of what economics required. (1939 7) In particular this method involved the purging of ideas of utility from Marshallian analysis, a task which Pareto is decisive in forwarding and which issued in the Marginal Rate of Substitution formulation. The development which Hicks independently pursued was the logical distinction of the substitution and income effects of changes in price, associated with Johnson (1913) and Slutsky (1915). This has, of course, become standard text book analysis. However, it was only a stage in the more logical formulation of demand theory, which was fully completed in A Revision of Demand Theory (1956).
In this study Hicks is uneasily aware of the epistemological gulf between his own tradition and the “econometric” approach to demand theory, which broadly means positivist approaches. In the process of clarifying his orientation he moves behind the substitution formulation to its logical foundation.
The consideration which decides me in favour of the new method, at least as an essential complement to the old, if not as a substitute, is its greater effectiveness in clarifying the nature of the preference hypothesis itself. The demand theory, which is based upon the preference hypothesis, turns out to be nothing else but an economic application of the logical theory of ordering; if we begin from the logic of ordering itself, instead of starting from the geometrical application of it, we begin by getting to our credit a number of distinctions, which will be of much use to us, and will save us from falling into some notorious pitfalls, later on. (Hicks 1956 19)
The foundation seems incontrovertible. Hicks is careful to distinguish between weak and strong ordering; the former allows sets which are not differentiated and is therefore the basis of indifference curve formation. (1956 19-21, 36-46) The assumption of transitivity makes sure that unordered items will not occur.(33) Then, as his logical foundation requires, Hicks focusses on consistency, showing which responses are consistent with changed prices and constant preferences. Using the method of compensating variation (compensating for the income effect) he shows that the logic of choice yields the normal demand curve. The theory is then articulated to consumer surplus, secondary substitution and reciprocity effects to complete a restatement of demand theory in logical terms, devoid of mathematics.
Because the revision is logical it is believed to be of complete generality and benign in its relationships to other formulations. Yet it requires us to accept the logic of ordering. Ordering is a human activity undertaken with the help of certain criteria. A logical ordering takes place when those criteria are consistently applied. But logic has nothing to do with whether one, two or more criteria are chosen. The logic of ordering which Hicks considers is the single one of more and less, but this excludes other criteria of ordering which produce their own logical patterns and will be important in consumption decisions. Criteria of more or less fair, more or less trouble, more or less durable, more or less aesthetically pleasing, higher or lower quality and more or less helpful enter into the way people live and buy. Most families would see the idea that two paintings or two lorries are better than one as ridiculous; aesthetic criteria and nuisance value weigh more heavily than more and less. To put it another way many situations exhibit very weak ordering in Hicksian terms for entirely logical reasons; indifference curves can be very, very thick. Thus, despite its claims, the foundation is not logical, but assertive.
Many of the directions charted by the Paretian approach are also evident in Hicks’ elaboration of his consumer theory, which is primarily conceived in terms of statics. His concern is only with “current prices, prices ruling during the same period as that in which the commodities are purchased.” (Hicks 1956 16) The logic of ordering creates a static framework, which rules out what was happening earlier – “his choice may depend, for instance, upon what he was doing previously. From the standpoint of a static theory, such as ours, such decision is a matter of chance.” ( 21) This in part explains the peculiar relationship with Hicks has with econometrics, which is seen largely in terms of analysis of cross temporal data. Hicks seems to imply that logical (static) theory is needed to define the framework for a dynamic econometric approach, and his normal method of dynamic analysis is time period equilibrium analysis, again following the Paretian mould. (1965 58-113, 1975). Thus the logic of ordering means that the preference map is by definition logical, while what goes on between those periods is extralogical, or as Hicks puts it, a matter of chance. Of course, he was too good an economist to be fully trapped inside his position, but its static equilibrium bias on logicist grounds is clear.
The problems with this position and Hicks’ ingenuity are evident when we consider the introduction of income as another potential mode of ordering. If the logic of ordering allows Hicks only to consider current price effects, how is he able to move from the logic of substitution to consider income, which really involves another frame of reference and cross temporal considerations? Hicks avoids this problem by redefining income in terms of the commodities which are being substituted. One of the casualties of this approach is the demise of the concept of income-elasticity; the income-consumption curve is not really the same thing. Hicks has to find a way of logically sublimating income, and he does this by treating income in commodity terms and then defining a commodity as “general purchasing power”. (1939 33) Such a move avoids introducing an actual measure of income, varying from one person to another, to the analysis, but this is not acknowledged within the abstract framework. This, for example, means that how poor people cope with changes is not normally a matter of concern within this framework; it only addresses income as what can be purchased.
There is a similar weakness in the way the income effect is identified. Hicks constructs a move along the indifference curve to a hypothetical point at which the consumer is supposed to be price indifferent, defining the income effect as residual. There are two fallacies here which need examining. The first concerns Hicks’ belief in the weak logic of ordering. If P is neutral to Q and Q to P, what kind of relationship do they have to one another? Hicks suggests that because neutrality is reversible, they have some kind of indifferent relationship, but he then assumes this is a relation of ordering (1956 31). Yet this can only be established with a transitivity condition. Here, however, he has a basic problem. How can a transitivity condition which is premissed on the possibility of ordering be applied to phenomena which the subject has not ordered. The transitivity of neutrality which he presumes to have established may be no more significant than the statement that neither pigs nor nails nor beauty are Swiss rolls. All logical attempts to make them comparable must fail, despite the comfort that a full domain of indifference curves gives. This presumption is important because many of the situations which consumers experience are ones where no logic of ordering makes sense. They ignore substitutionary comparisons or do not need any more of one good and are therefore indifferent to a change in its price. It is only the spurious linear picture of indifference curves which rules out these other possibilities and artificially creates the required transitivity.
The second fallacy is more straightforward. It is clearly not valid, when non-ordered items are assumed to be indifferent with respect to price, to treat the residual movement as income effect, because the preferred position is strictly (logically) not comparable to the other in income terms. What seems to be income effect may be other things – a change of convenience, search time, more or less complementary purchasing and a loss or increase in security. A logic of ordering can be good fun, but Hicks assumes the income effect is a statement about the real world. Often in the dynamic situation the fall in the price of a single good reflects falling incomes, changing exchange rates, the emergence of new alternative forms of consumption or lower quality goods or services which mean that the consumer needs to take other things into account. To say that what is left when the substitutionary effect of changes in relative prices is removed is the income effect is fallacious and unwarranted in real life. Again the weakness of this logical framework is evident.
The move of this position to high orthodoxy in many textbooks over the next fifty years does not need documenting. Many economists have bought this logic wholesale and not considered its foundational weaknesses or the constrictions which it has placed on theoretical development. Often, too, they have only been dimly aware of the difference in the meaning of theory created by this logicist approach. Far more instructive is the development of Hicks himself. This logicist position dominates the early Value and Capital and A Revision of Demand Theory and forms Hicks’ underlying perspective for much of his work. Yet he also wrestled with the limitations of the position and with the challenges which other perspectives brought to his own. The deepest challenge was from Keynes whose underlying epistemology, examined in chapter 4, was very much different from his own.
Hicks faced Keynes theory, probably unaware of Keynes own earlier epistemological struggles. He initially saw Keynes as intuitive over against his own logical approach (1939 4), but soon came to grips with the fact that Keynes model was dynamic with income effects central rather than residual. It challenged the static equilibrium micro and macro models of Value and Capital. (1939 11-111) Hicks initial response was a temporary equilibrium model, and then a growth equilibrium framework. (Hicks 1939 245-282, 1965 58-75, 131-197) This allowed him to see most markets as conforming to a fixprice model rather than the price flexible form, and on this model the adjustments take place through stocks and balances of work in progress. Consequently, the consumer often may not face price changes, but rationing of supply, a salient part of the consumption map which was ignored in the earlier logicist pattern. Hicks also introduced a non-logicist dynamic element to his understanding of consumption when he investigated the way in which future plans and expectations can have a causal effect on the current situation. (Hicks 1979) Although Hicks has shown a wider perception, especially at the macroeconomic level, of how limited the logicist framework is, it has still shaped much of his theoretical development as it grew from the static logicist equilibrium framework. (Hicks 1985 1-10) His withdrawal from the IS-LM formulation signalled the fuller penetration of dynamic equilibrium thinking into his analysis and indicated the extent of his withdrawal from the primacy of static logical constructs as basic to theory. (Hicks 1980 139-53, Weintraub S 1977) Sadly, it was just undoing some of the consequences of logicism in fixing consumer theory in indifference analysis for decades.
During this process Hicks also orientated himself to positions which were not so far away from his own logicist rationalist basis. He considered the means-ends framework, or the means-objectives framework, as he called it, arguing at first that a reformulation in these terms would not make much difference compared with substitutionary analysis. Yet he acknowledged that shifts in demand, hardly a peripheral issue, create wider problems. He saw objectives as a level of analysis operating behind the actual demand for commodities which act as means to meet these objectives. Obviously changes in means or objectives interact with the price system, but in a way which would be “very complicated” and lead to unmanageable analysis in applied economics or econometrics. (Hicks 1956 167-8) Thus the rejection of means-ends rationalism takes place on the ground that the logical simplicity of consumption theory would be undermined. Hicks failed fully to empathise with the perspective adopted by Robbins because of his prior logicist commitment.
More intriguing still is the way Hicks relates to the revealed preference of Samuelson, which he ambiguously links with “the econometric approach” and empiricism. (1956 4-7) In Chapter Six of A Revision of Demand Theory he looks at preference in terms of the direct consistency test, and reworks the revealed preference idea of recorded behaviour in terms of the responses of an ideal consumer and whether they are consistent. In terms consonant with the Paretian view of testing he feels “obliged to conclude from this that there is in practice no direct test of the preference hypothesis.” (Hicks 1956 58) This is obviously at odds with Samuelson’s attempt to move from the actual revealed preferences of consumers to general theory. Indeed, Hicks is somewhat dismissive of attempts to relate to actual behaviour at all, (1956 55) and throughout the impression is of two theorists who, although they come to similar benign conclusions, are operating on fundamentally different models of theory construction.
The dominance of the Hicksian logicist framework during the 50s, 60s and 70s in consumption theory has been marked. Yet gradually the weaknesses and limitations of this base have been exposed. Later we shall examine Scitovsky’s move from logicism to a completely different foundation, as he found the logicist framework not addressing important aspects of the way consumers behave. (Scitovsky 1952, 1977) But the problems also arose for those who stayed within the logicist framework. Lancaster recognised one of the limitations of the perspective and tried to solve it in logicist terms. As a result of a response by Johnson to a paper “Revising Demand Theory” (Lancaster 1957), he came to see orthodox logicist theory as asserting that “all intrinsic properties of particular goods have been omitted from the theory….goods are what consumers would like more of… goods are what are thought of as goods…” (Lancaster 1966 132) Many goods, he recognized, were actually bought on the basis of characteristics which they possessed, which were more strategically basic than the goods themselves, like taste, beauty, security. Courageously he set out to reformulate the theory in logicist terms. He tried to move the logic of choice back from the actual commodities to the “characteristics” which generated commodity choice. This focus is different from the view treating goods as instrumental to consumer ends, which we shall examine later, but it does see beyond the actual choices to some of the strategic considerations involved in purchasing. Because the logical framework now involved different criteria, corner solutions appeared in the indifference mapping, and as soon as multiple characteristics and criteria developed Lancaster found the whole framework crumbling in amazing complexity. Because the logicist framework had to be unidimensional, it could not cope with the complexity of the analysis. Yet uneducated consumers handle this problem every day with the simple idea of priority. When one accepts that consumers can have competing criteria of value, each with its own logic, the logic of ordering becomes unrealistic as the necessary framework for consumption analysis. (Lancaster 1966, 1971)
Gradually, the sterility of this framework and its lack of contact with the real world of consumption has lost it adherents. At the same time there are still many economists who have grown up within this framework and cannot escape, because they have not yet defined what it is which grips their theoretical development and claims their loyalty. In view of this it is worth examining the problems more systematically.

The Structural Problems of Logicist Consumption Theory.
Logicism thus constitutes a foundational drive in consumption theory which has deeply influenced Jevons, Edgeworth, Pareto, Pantaleoni, Slutsky, Hicks, Allen and Lancaster. It claims to establish an indubitable logical basis for the development of well-formed theory, but actually it has failed to validate that base and the consequent development of the theory has been flawed. The theoretical problems which it has encountered are structurally related to the logical foundation, and in summary we will identify the links.
The authority of pure or formal logic as self-evident rested on conclusions being necessarily the case in all possible worlds. This incontrovertibility made its statements seem uncontaminated by dissent or the possibility of seeing things different ways. It gave such knowledge appeal as a foundation for scientific work. If the incontrovertible base could be constructed, free from controversy, then sound and well-formed theory could be built on it. Yet this certitude was not available in reality. “A is not non-A” is based on particular symbolic conventions. Any propositions which were theoretically relevant had to import assumptions. Either the “logic” was certain, but cut off from the world which it must claim to address, or it was contentious, but without acknowledging its assumptions. What actually happened was that the dogma of constructing consumption theory around the logic of more and less was not properly faced. The foundation was not incontrovertible, nor was it discussed and subject to analysis, but stayed as an unstated dogma standing over against other positions.
When this problem was embodied in consumption theory, it was normally expressed in the form of a disjunction between pure and applied. Whatever logic is constructed is deemed to be pure theory, and incontrovertible as the basis of theory as a result. Indeed, the use of the word, “pure”, has been characteristic of all the economists in this tradition. In another category, because it brings in assumptions about the real world, is applied theory which is really seen as a residual. The importance of the category “pure” in consumption theory is immediately evident. It is conveyed by Pareto.
First we separate the study of ophelimity from that of the various forms of utility; then we direct our attention to man himself, stripping him of a large number of accretions, ignoring his passions, whether good or bad, and reducing him eventually to a sort of molecule which is susceptible solely to the influence of the forces of ophelimity. By following this procedure we shall achieve a science: pure economics, equivalent as a science to pure mechanics.” (Pareto 1896 II 2)
It betokens the process whereby ideology, subjective understanding, metaphysics and philosophy are supposedly cleansed out of the subdiscipline, which can then become uncontaminated science. This is the drive evident throughout the tradition, yet it fails to achieve the uncontaminated science and actually impoverishes the theoretical responses which can be considered.
Indeed, the logic is routinely flawed. One of the key assumptions is transitivity. In any consumption domain, if A is preferred to B and B to C, then A must be to C. This foundational, seemingly self-evident, assertion however only holds provided a single logic of preference is used. But this is actually a very restrictive assumption, certainly one which has no logical status. When consumers have different dimensions of preference, like time-saving, aesthetics, fitness, saving money and buying something which lasts, then A