Veblen,Marshall and neoclassical economics

Abstract

Alfred Marshall is often depicted as a pioneer of neoclassical economics almost as if this is a label he embraces and promotes. Yet neoclassical economics is not a category Marshall deploys but a term Thorstein Veblen introduces in characterising Marshall. Veblen coins the term neoclassical to identify an ontological discrepancy in the work of a specific group of his contemporaries, a prominent figure among whom is Marshall. Veblen’s view is that Marshall and other neoclassicals discern features of social reality that suggest a tentative recognition of a causal processual social ontology of the type Veblen associates with modern evolutionary approaches and yet also remain staunchly committed to a taxonomic conception of science underpinned by a quite different set of ontological presuppositions. Veblen’s assessment of Marshall is brief and assertive. In this paper it is argued that the ontological discrepancy interpretation of Marshall, that Veblen first sketched, can convincingly be filled out, has substantial merit and is of importance in developing an adequate appreciation of Marshall.

Introduction

Alfred Marshall is often depicted as a pioneer of neoclassical economics, almost as if this is a label he adopts, embraces and promotes. ¹ Yet neoclassical economics is not a category Marshall deploys but a term Thorstein Veblen introduces when characterising Marshall’s contributions. Veblen does so in the final paper in his The Preconceptions of Economic Science series. Lawson (2013, forthcoming) demonstrates that Veblen coins the term neoclassical to identify an ontological discrepancy in the work of a specific group of his contemporaries, a prominent figure among whom is Marshall. Veblen singles Marshall out as one of the more worthy, interesting and insightful contributors belonging to this group. Lawson shows that Veblen uses the term neoclassical to capture an inconsistency at the level of ontological commitments between those that can be discerned in an endorsed evolutionary vision and those implicit in an assumed taxonomic conception of appropriate scientific method. Marshall and the others of his group tentatively emphasise a causal-processual social ontology of the type that Veblen himself associates with modern evolutionary approaches. But they simultaneously remain staunchly committed to a taxonomic conception of science underpinned by a quite different set of ontological presuppositions. The term neoclassical is coined by Veblen to express this tension, the term classical being used for those that wield taxonomic methods, with the term neo being appended to those in this group that also adopt an evolutionary vision, albeit failing to recognise that their continuing reliance on taxonomic methods is inconsistent with this vision. ²

The objective of the current paper is not to challenge Lawson’s interpretation of Veblen’s intended meaning in coining the term neoclassical. Lawson’s analysis is well grounded in the body of Veblen’s writings, and convincingly argued. There have been a number of papers that have critically reflected on Lawson’s interpretation of Veblen on these matters, including by Veblen scholars (see Mayhew, 2016) and none have challenged Lawson’s reading of Veblen’s use of the term neoclassical. The focus here is different. Lawson, in the relevant papers, is not interested in Marshall’s writings in themselves but only in how Veblen interprets Marshall. At the same time Veblen’s own treatment of Marshall is brief. The aim of this paper is to examine Marshall from the point of view of Veblen’s specific characterisation of him as neoclassical. That there is inconsistency and related failings in Marshall’s later writings in particular is noted by many commentators. ³ Few though have located the problem as an inconsistency of ontology. ⁴ For the purposes of this paper, I shall accept the meaning of neoclassical as the ontological mismatch Veblen identifies, ⁵ and question whether it is a fair characterisation of Marshall’s contributions, exploring whether there are nuances and qualifications to be noted. Mestrovic (2019) observes in the context of current debates about classical sociology that it is Veblen’s work that especially needs to be rediscovered and applied to modern culture. The evaluation of Veblen’s interpretation of Marshall as a neoclassical economist together with consideration of the contemporary relevance of the category of neoclassical economics in its original meaning is set out here as a contribution to such a broader project of rediscovery.

The paper is structured as follows. In the second section a summary is provided of Veblen’s characterisation of Marshall as neoclassical. The sections that then follow advance an assessment of the appropriateness of this characterisation of Marshall. In the third section the extent to which Marshall can be understood as having an evolutionary vision is explored. In the fourth section Marshall’s commentaries on science and method are examined and an assessment of Veblen’s claim that he retains a taxonomic conception of science provided. In section five the focus is on whether there is evidence in Marshall of the types of complications and inconsistencies to be anticipated of an author who recognises the open, processual nature of social reality but retains a taxonomic conception of science with quite different and incompatible ontological presuppositions. Concluding remarks follow where the implications of Veblen’s assessment of Marshall for modern social and economic theorising are considered.

Neoclassical economics and the Marshallian ontological-discrepancy thesis

Lawson notes that in the ‘preconceptions’ essays Veblen’s primary focus is not substantive theory but the metaphysical or ontological preconceptions underpinning economic theorising. Veblen (1899a, 1899b, 1900) aims to identify the form the presentation of results must take for scientific contributions to be accepted as having been completed successfully. Veblen draws out the norms of valid scientific presentation that are, if often implicitly, accepted by participants in particular research communities and guide their activities.

In the three ‘preconceptions’ papers Veblen (1900) argues that the ‘changes which have supervened in the preconceptions of the earlier economists constitute a somewhat orderly succession’ (p. 240). Veblen (1900) emphasises that the most interesting feature of this trajectory has been the gradual change over time in the received ‘grounds of finality’ presupposed in economics:

The feature of chief interest in this development has been the gradual change in the received grounds of finality to which the successive generations of economists have brought their theoretical output, on which they have been content to rest their conclusions, and beyond which they have not been moved to push their analysis of events or their scrutiny of phenomena. There has been a fairly unbroken sequence of development in what may be called the canons of economic reality; or, to put it in other words, there has been a precession of the point of view from which facts have been handled and valued for the purpose of economic science (p. 240).

Structuring Veblen’s analysis in these papers, as Lawson points out, is a contrast between two separate grounds of finality for science both of which he saw as being influential among his contemporaries. Veblen distinguishes taxonomic from modern evolutionary science.

For Veblen, as Lawson reports, the basic tenet of a taxonomic science is that the fundamental basis for the systematisation of knowledge is something like a natural law, an association of phenomena, an empirical generalisation or correlation with exceptions then typically interpreted as disturbing factors. This contrasts with an evolutionary conception of science that presupposes nothing beyond cumulative causal sequence. In adopting an evolutionary orientation the assumption that outcomes conform to some pre-ordained pattern or regularity or serve some laudable purpose is abandoned and instead it is merely presupposed they are caused by something that went before them.

This contrast had featured in Veblen’s (1898) earlier essay on ‘Why is economics not an evolutionary science?’ where he characterises the evolutionary scientist as being ‘unwilling to depart from the test of causal relation or quantitative sequence’ (p. 377) and one who interrogates what has taken place and what is taking place and formulates answers in terms of cause and effect. The thought and action of the taxonomic economist is guided quite differently, in particular ‘this ground of cause and effect is not definitive’ (Veblen, 1898: 378). From this point of view the underlying premise is that there is a trend in the association of phenomena, an empirical generalisation, that is regarded as ‘natural’ or ‘normal’.

Lawson notes that Veblen in the preconceptions series reads all earlier strains of economic thought as conforming to a taxonomic point of view. For earlier taxonomic economists the content of normality was vital, but that conception weakened over time. For later taxonomic economists, for whose contributions Veblen reserved the label classical economics, scientific preconceptions of normality took the form essentially of correlations or event regularities, albeit regularities about the normal or natural case, understood as that which common sense determines as desirable. Lawson shows that Veblen is careful to trace through how preconceptions of normality and regularity have changed and been rationalised in different periods culminating with those of the classical economists of the recent past and contemporary period. But Lawson notes that Veblen’s central interest is in whether conceptions of normality are being retained or abandoned in a context where a thoroughly modern evolutionary orientation is being adopted in other fields of research. To the extent that presuppositions of regularity and normality are retained Veblen considers how the development of evolutionary economics is held back. Veblen’s (1898) interest in exploring the extent to which the seen to be appropriate methods of economic science are informed by a taxonomic orientation is stated in the ‘Why is economics not an evolutionary science?’ essay and then taken up in the later ‘preconceptions’ papers:

The question of interest is how this preconception of normality has fared at the hands of modern science, and how it has come to be superseded in the intellectual primacy by the latter-day preconception of a non-spiritual sequence. This question is of interest because its answer may throw light on the question as to what chance there is for the indefinite persistence of this archaic habit of thought in the methods of economic science (p. 379).

Lawson notes that the term neoclassical economics is introduced by Veblen near the end of the preconceptions essay series after an extended historical analysis of the, largely tacitly held, metaphysical postulates of earlier economists. The term features as Veblen considers the prospects of an evolutionary approach to economics displacing the taxonomic perspective which he shows had itself, in the form of classical economics, come to take a less animistic orientation. ⁶ Veblen, in introducing the term neoclassical economics, is focussed on current and future developments in economics and especially with whether it could be anticipated that an evolutionary orientation with its causalist rather than associationist ontology would soon achieve a position of dominance.

In exploring whether and to what extent an evolutionary approach was gaining ground among economists Veblen concentrates on those contributions that seemed particularly advanced in promoting this view. Crucially though, as Lawson emphasises, Veblen argues that even the very best of these contributions ultimately made only limited progress towards a coherent evolutionary perspective. Lawson notes that at this point in his analysis Veblen maintains that despite these cutting edge contributions promoting an evolutionary vision they adopt and see as valid methods conforming to taxonomic conceptions. These methods in turn imply ontological commitments that are at odds with the evolutionary stance the relevant authors advance. Lawson sees Veblen as pinpointing a rather fundamental tension, one which he takes Veblen as being concerned about since he recognises it as a constraint on the progress of evolutionary thinking in economics.

Lawson notes that, when elaborating upon this tension and exploring how it serves to check the development of an evolutionary perspective, Veblen (1900) considers it ‘necessary to limit discussion, for the present, to a single strain, selected as standing peculiarly close to the classical source, at the same time that it shows unmistakable adaptation to the later [evolutionary] habits of thought and methods of knowledge’ (p. 261). Veblen (1900) examines the nature of this ‘strain’ by considering the contributions of two of its most able proponents: John Neville Keynes and Alfred Marshall:

For this later development in the classical line of political economy, Mr. Keynes’s book may fairly be taken as the maturest exposition of the aims and ideals of the science; while Professor Marshall excellently exemplifies the best work that is being done under the guidance of the classical antecedents (pp. 261–262).

Marshall’s contributions represent, for Veblen, a specific strand of economics, the one that constitutes the best work done within this line of thinking, in effect moving it away from its classical heritage.

Lawson’s argument is that Veblen’s main point in discussing this Cambridge school is that no matter how ready members of this group are to acknowledge causal processes, and in particular causal histories of structures like institutions, in line with the kind of causal processual ontology that Veblen takes as underpinning evolutionary science, even Neville Keynes and Marshall are unable in practice to break with the taxonomic ideal of science, particularly at the level of method. Veblen (1900) suggests: ‘There is a curious reminiscence of the perfect taxonomic day in Mr. Keynes’s characterisation of political economy as a “positive science,” the sole province of which is to establish economic uniformities’ (p. 264) and observes ‘in this resort to the associationist expedient of defining a natural law as a “uniformity,” Mr. Keynes is also borne out by Professor Marshall’ (Veblen, 1900: 265). For Veblen, on Lawson’s reading, certain antiquated beliefs about what the work of the scientist involves constitute a constraint on Neville Keynes and Marshall shaping the methods they see as legitimate and ultimately preventing them from advancing a meaningful account of the genesis and developmental continuity of institutional phenomena.

The taxonomic approach that had guided the classical school is retained by Neville Keynes and Marshall with their notions of normality taken to relate to those situations, now considered to exist at the level of the actual course of events, that conform to empirical regularities. Lawson highlights that this is all quite inconsistent with Veblen’s conception of evolutionary thinking. Indeed, although Marshall is seemingly more adapted to modern science than most economists, he is interpreted by Veblen as being especially inconsistent on these matters. For, Veblen (1900) observes that although Marshall devotes time to investigating the nature of institutions and is enthusiastic about incorporating insights of evolutionary thinking, throughout Marshall’s work the ‘taxonomic bearing is after all the dominant feature’ (p. 264).

Marshall’s contributions are not dismissed by Veblen. As Lawson notes, Veblen even suggests that despite ‘survivals of the taxonomic terminology, or even of the taxonomic cannons of procedure the latter do not hinder the economists of the modern school from doing effective work of a character that must be rated as genetic rather than taxonomic’. Veblen (1900) compares Marshall to the botanist Asa Gray ‘who, while working in great part within the lines of “systematic botany” and adhering to its terminology, and on the whole its point of view, very materially furthered the advance of the science outside the scope of taxonomy’ (p. 265). Yet the steps towards an evolutionary approach remain, according to Veblen, hesitant and uncertain. The evolutionary thinking to be found in Marshall is ultimately rather superficial and he makes little progress in fashioning relevant methods of analysis. The stream of neoclassical thinking developed by Marshall is distinguished by Veblen in order to show that even this advanced form fails to break decisively from taxonomic science, especially at the level of method. ⁷

Veblen’s original deployment of the term neoclassical is thus shown by Lawson to capture contributions that are characterised by a tension and specifically an inconsistency at the level of ontological commitments. It is a tension between the ontological presuppositions of methods sponsored by a taxonomic conception of science and a causal processual social ontology seen as underpinning an evolutionary view of science. Veblen acknowledges that in these contributions there is a greater awareness (than in the work of earlier classical economists) of issues that are central to the evolutionary approach of the sort he favoured. ⁸ However, even here taxonomy in terms of method remains dominant. The taxonomic aspect is the classical part expressed in the deployment of methods that presuppose event regularities and an associationist ontology and the evolutionary awareness is the ‘neo’ component. The ontological discrepancy thesis as it relates to Marshall is, at best, only sketched by Veblen and is not Lawson’s primary concern. In the following sections I explore whether Veblen’s reading of Marshall can be substantiated. ⁹

Marshall’s evolutionary vision

In characterising Marshall as neoclassical Veblen suggests he, albeit only partially and rather loosely, adopts an evolutionary vision. In this section I explore what Veblen means by an evolutionary perspective and consider whether Marshall adopts such a vision. Veblen primarily uses the term evolutionary to refer to a method or conception of science but in doing so ties it to a causal processual ontology. In considering whether Marshall had an evolutionary vision it is necessary to explore both some of his explicit methodological remarks and his underlying ontological orientation.

In his essay on ‘Why is economics not an evolutionary science?’ Veblen (1898) describes any course of events appropriate for evolutionary analysis as conforming to something like ‘impersonal or mechanical sequence’ (p. 379) and any process in which such an impersonal sequence is produced as one of ‘dispassionate cumulative causation’ (p. 381). It is the ontology of processes of cumulative causation that Veblen sees as grounding an evolutionary science of economics.

Veblen elaborates in a later contribution on the understanding of causality that he associates with modern evolutionary approaches by contrasting it with earlier conceptions. He writes: ‘The Law of causation as it is found at work in the maturer science of the eighteenth and early nineteenth centuries, comprises two distinguishable postulates: (1) equality (quantitative equivalence) of cause and effect; and (2) similarity (qualitative equivalence) of cause and effect’ (Veblen, 1904: 364). He then provides further detail:

In the conception of the causal relation as it may be seen at work a hundred years ago, cause and effect are felt to stand over against one another, so that the cause controls, determines the effect by transmitting its own character to it. The cause is the producer the effect the product. Relatively little emphasis or interest falls upon the process out of which the product emerges, the interest being centred upon the latter and its relation to the efficient cause out of which it has come. The theories constructed under the guidance of this conception are generalisations as to an equivalence between the producing cause and the effect product (Veblen, 1904: 365–366).

In advanced modern science Veblen argues the conception of causality is quite different. Those progressing modern evolutionary science have come to think in terms of technological process:

The process is always complex; always a delicately balanced interplay of forces that work blindly, insensibly, heedlessly; in which any appreciable deviation may forthwith count in a cumulative manner, the further consequences of which stand in no organic relation to the purpose for which the process has been set going (Veblen, 1904: 368).

Hence, he notes: ‘The characteristically modern science does not inquire about prime causes, design in nature, desirability of effects, ultimate results, or eschatological consequences’ (1904: 370). He concludes that ‘that work of research which . . . effectively extends the borders of scientific knowledge is nearly all done under the guidance of highly impersonal, mechanical, morally and aesthetically colourless conceptions of causal sequence’ (Veblen, 1904: 372). For Veblen an evolutionary perspective involves a shift towards such an understanding of causality.

Marshall can be understood as at times adopting something like the concept of causation that Veblen associates with an evolutionary approach. Marshall recognises that regularities or strict event associations are rare occurrences and that the isolation of a single stable causal factor within a closed system is effectively infeasible in social contexts. Marshall (1923 [1929]) argues explicitly that the progress of knowledge is ‘throwing increasing doubt on the possibility of attributing any event to a single cause. The more an event is studied the larger is generally the number of causes, by which it is seen to have been influenced: and there is seldom an easy and decisive means of isolating the influence of any one cause’ (p. 274).

Marshall emphasises that social phenomena are typically not atomistic in the sense of having the same independent, invariable and quantitatively equivalent effect whatever the context. The material different sciences address varies in this regard. For Marshall (1923 [1929]) ‘many classes of movement in the physical world are exact copies of movements that have gone before’ (p. 351). This is not the case when it comes to the material that biological and social sciences consider. In the Principles he notes that ‘the matter with which the chemist deals is the same always; but economics, like biology, deals with a matter, of which, the inner nature and constitution, as well as the outer form, are constantly changing’ (Marshall, 1890 [1961]: 772).

Failure to recognise and respond methodologically to the reality that social phenomena are inherently and continually in process is a limitation that Marshall associates with certain of his prominent predecessors. According to Marshall (1885 [1925]) economists need to focus on both the way in which human societies change and how agents themselves are in process:

The chief fault, then, in the English economists at the beginning of the century was not that they ignored history and statistics; but that Ricardo and his followers neglected a large group of facts, and a method of studying facts which we now see to be of primary importance. They regarded man as, so to speak, a constant quantity, and gave themselves little trouble to study his variations (pp. 154–155).

For Marshall the social realm is highly relational and constituted by phenomena subject to change and therefore its very mode of being is intrinsically dynamic. Marshall (1898) elaborates as follows:

The catastrophes of mechanics are caused by changes in the quantity and not in the character of the forces at work: whereas in life their character changes also. “Progress” or “evolution,” industrial and social, is not mere increase or decrease. It is organic growth chastened and confined and occasionally reversed by the decay of innumerable factors, each of which influences and is influenced by those around it; and every such mutual influence varies with the stages which the respective factors have already reached in their growth (pp. 42–43).

The qualitative equivalence between cause and effect that Veblen associates with pre-evolutionary approaches is an assumption Marshall (1890 [1961]) appears to avoid:

It happens that in economics, neither those effects of known causes, nor those causes of known effects which are most patent, are generally the most important. “That which is not seen” is often better worth studying than that “which is seen”. Especially is this the case if we are not dealing with some question of merely local or temporary interest, but are seeking guidance in the construction of far reaching policy for the public good; or if, for any other reason, we are concerned less with immediate causes, than with causes of causes (pp. 778–779, from fifth edition 1907).

There is a recognition here of the importance of underlying causal factors that can be qualitatively distinct from events. It seems likely that Veblen would see in such commentary an attempt by Marshall to break away from the conception of causality associated with earlier taxonomic perspectives. Marshall is interested in issues of growth, variation and process – the class of questions that Veblen associates with modern evolutionary approaches.

For Veblen a distinguishing feature of an evolutionary approach is its rejection of the assumption that certain situations that come about are predetermined to do so and hence laudatory or legitimate in some significant sense. For Veblen (1898) the ‘notion of legitimate trend in a course of events is an extra evolutionary preconception and lies outside the scope of an inquiry into the causal sequence in any process’ (p. 392). Marshall recognises such implications of an evolutionary view noting for example that the question of worth is irrelevant to evolutionary analysis. Marshall (1919 [1923]) writes:

Darwin’s ‘law of the survival of the fittest’ is often misunderstood; nature being supposed to secure through competition, that those shall survive who are fittest to benefit the world. But the law really is that those races are most likely to survive who are best fitted to thrive in their environment: that is, to turn to their own account those opportunities which the world offers to them. A race of wolves that has well organised plans for hunting in packs is likely to survive and spread: because those plans enable it to catch its prey, not because they confer benefit on the world (p. 175).

There is then not only a general emphasis in Marshall on the processual nature of the material that the biological and social sciences address but also, at least at times, a specific recognition – one Veblen would associate with an evolutionary perspective– of the non-predetermined, non-teleological nature of biological and social change.

According to Veblen, for any school of economics to reach an evolutionary state of development dated metaphysical presuppositions depicting human agents in overly passive terms as responders to narrowly economic prompts need to be overcome. The individual Marshall (1890 [1961]) argues is not to be reduced to an isolated atom independently calculating and pursuing their own narrow ends:

In all this they [economists] deal with man as he is: not with an abstract or “economic” man; but a man of flesh and blood. They deal with a man who is largely influenced by egoistic motives in his business life . . . but who is also neither above vanity and recklessness, nor below delight in doing his work well for its own sake, or in sacrificing himself for the good of his family, his neighbours or his country; a man who is not below the love of a virtuous life for its own sake. They deal with man as he is (pp. 26–27, from third edition 1895).

Marshall, when he responds to those who worry about the influence on economists of Hedonism or Utilitarianism, emphasises the need to recognise a broad range of motives for action and the influence that notions of duty may have. ¹⁰

Marshall (1890 [1961]) argues, when addressing economic problems, the individual needs to be recognised as an active being participating in broader groups, a component part of a whole. He acknowledges the existence of emergent powers of more complex social entities that are irreducible to those of their parts considered separately. He emphasises that it is necessary to understand the action of individuals in relation to the positions that they occupy in the social totalities they are constituent members of:

As a cathedral is something more than the stones of which it is made, as a person is something more than a series of thoughts and feelings, so the life of society is something more than the sum of the lives of its individual members. It is true that the action of the whole is made up of that of its constituent parts; and that in most economic problems the best starting-point is to be found in the motives that affect the individual, regarded not indeed as an isolated atom but, as a member of some particular trade or industrial group. . .. economists study the actions of individuals, but study them in relation to social rather than individual life, and therefore concern themselves but little with personal peculiarities of temper or character. They watch carefully the conduct of a whole class of people, sometimes the whole of a nation, sometimes only those living in a certain district, more often those engaged in some particular trade at some time and place (pp. 25–26, from third edition 1895).

According to Veblen the modern evolutionist when addressing social phenomena is concerned to explore the canons of conduct imposed by the group’s scheme of life ¹¹ and Marshall at times expresses an interest in pursuing such investigations. Marshall emphasises that the focus of interest in economics is not only agents who act but also the various conditions that provide the basis for the potentially creative acts of individuals. Thus, Marshall (1890 [1961]) writes: ‘Although the proximate causes of the chief events in history are to be found in the actions of individuals, yet most of the conditions which have made these events possible are traceable to the influence of inherited institutions and race qualities and of physical nature’ (p. 723, from second edition 1891).

Marshall provides some partial insights into the nature of these constraining and facilitating structures. In his discussion of the development of free enterprise, for example, he not only considers certain physical conditions as necessary but also emphasises the role of custom and routines in early societies. According to Marshall customs there make themselves felt on agents through divergences from ancestral routine being met ‘with the opposition of people who had a right to be consulted on every detail’. Energetic and inventive individuals are he suggests ‘unwilling to face the anger with which their neighbours would regard any innovation, and the ridicule which would be poured on any one who should set himself up to be wiser than his ancestors’ (Marshall, 1890 [1961]: 727, from first edition 1890). Marshall (1890 [1961]) often depicts custom as a significant constraint on industrial progress. He argues that ‘the greater part of custom is . . . a crystallised form of oppression and suppression’ (pp. 725–726, from first edition). He develops this theme by claiming that ‘there often grew up a complex network of rules, by which every cultivator was so rigidly bound, that he could not use his own judgement and discretion even in the most trivial details. It is probable that this has been the most important of all the causes which have delayed the growth of the spirit of free enterprise among mankind’ (Marshall, 1890 [1961]: 727, from first edition).

Despite the heavy constraining force that is custom, Marshall emphasises that agents can respond in novel, creative and unpredictable ways. Hence, for Marshall (1919 [1923]) custom is not completely fixed or fully determining, he writes:

If custom had been absolutely rigid, it would have been an almost unmixed evil. But the resistance which it offered to the bold reformer resembled that presented by a glacier to anyone who might try to change its shape: custom and the glacier are plastic, but both refuse to be hurried in the adjustments. Custom has discouraged any attempt at improvement which involved a sudden breach with tradition: but, except in some ceremonial matters, it has been tolerant of modification in substance, form and method which did not obtrude themselves (p. 197).

Marshall maintains that custom should develop slowly, this is consistent with his broader insistence not only on the processual nature of social reality but also the idea that change is typically gradual. ¹² For Marshall if custom remained fixed this would impede the development of industrial techniques and the achievement of higher standards of moral conduct. Yet Marshall (1919 [1923]) recognises moral and technical progress cannot be expected without a body of tradition that can in a piecemeal fashion be transformed:

Had each [individual attempting a change] put his individual fancies into practice without restraint, few would have followed his erratic movements: there would have been no corpus or body of general thought, in which they could have been merged; and, in the absence of written record, they might probably have perished without leaving direct successors. But custom supplied a permanent body of general design, on which each fresh mind might try to make some variation for the sake of economy of effort, of increased utility, or more pleasing effect. And custom under favourable conditions was able to make utility, economy and artistic delicacy work harmoniously together in improving the standard commodity of common life (pp. 197–198). ¹³

On occasion Marshall appears eager to address the social conditions facilitating and constraining individual action and consider how social structures impact upon individuals by placing them in material circumstances – positions with associated rules, tasks, duties, rights – that motivate them to act in certain ways (via the perceptions of their interests).

There is then evidence that Marshall (1898) has an evolutionary vision of sorts and draws upon something approximating a causal processual social ontology. Beyond general comments such as his famous remark that the ‘Mecca of the economist is economic biology rather than economic dynamics’ (p. 43) there is also more substantial analysis indicating his adoption of such a vision. His conception of causality is in some respects consistent with Veblen’s understanding of that found within evolutionary approaches, he recognises the non-teleological nature of biological and social change, distances himself from overly passive accounts of human agents and is interested (to use Veblen’s terms) in exploring the canons of conduct imposed by the community’s schemes of life. Aspects of Marshall’s contributions accord with an evolutionary perspective as understood by Veblen and indeed Marshall himself. However, as is shown in the following section these aspects are at odds with the presuppositions of the methods Marshall typically deploys.

Marshall’s retention of a taxonomic conception of science

If Marshall can be understood as having to some extent an evolutionary vision, does he at the same time maintain a taxonomic conception of science and thereby import the ontological preconceptions associated with it? It is this element of Veblen’s assessment of Marshall that will now be considered.

Something of a taxonomic conception of science is evident as Marshall seeks to reconcile his recognition of the processual texture of the social realm with the possibility of pursuing an appropriately scientific approach to economics. If the taxonomic view of science is retained, the recognition of the open processual nature of social phenomena is a threat to the very aims of science, for on this reading the identification of strict event regularities or associations is central to science. The dilemma posed by the processual nature of the social realm, as Marshall (1898) conceives it, is that science is concerned with the elaboration of event regularities and yet social phenomena are mostly generated in systems curtailing the possibility of detailing such associations: ‘If we include in our account nearly all the conditions of real life, the problem is too heavy to be handled: if we select a few then long drawn out and subtle reasonings with regard to them become scientific toys rather than engines for practical work’ (p. 52). Marshall implies that if only the conditions could be sufficiently specified the sought-after event associations would be found to obtain. Acknowledging that such a complete analysis is impracticable he argues that careful interpretation is always required and economists must recognise the potential dangers of such partial analyses as are feasible. Had Marshall abandoned the taxonomic view and the assumption that results must take the form of event associations then his ambition of defending the possibility of social science could have been advanced alongside his acknowledgement of the open, processual nature of the social realm without tension or strain.

For those committed to the taxonomic view (and therefore implicitly tied to its ontological presuppositions), the recognition of openness and process represents a major dilemma. Marshall emphasises that in the face of this constraint on the search for event associations, science characteristically initially adopts what he labels the statical method. Marshall attaches considerable significance to the statical method, and it is in his elaborations of it that the influence of the taxonomic conception is especially pronounced.

Marshall (1898) outlines the procedures associated with the statical method as follows:

Man’s powers are limited: almost everyone of nature’s riddles is complex. He breaks it up, studies one bit at a time, and at last combines his partial solutions with a supreme effort of his whole small strength into some sort of an attempt at a solution of the whole riddle. For breaking it up, he uses some adaptation of a primitive but effective prison, or pound, for segregating those disturbing causes whose wanderings happen to be inconvenient for the time: the pound is called Ceteris Paribus. The study of some group of tendencies is isolated by the assumption other things being equal: the existence of other tendencies is not denied, but their disturbing effect is neglected for a time. The more the issue is thus narrowed, the more exactly can it be handled; but the less closely does it correspond to real life (p. 40).

For Marshall (1898) the statical method initially assumes that certain factors have no systematic influence which leads to a certain loss of realisticness but increased precision. The objective is to move towards greater realisticness which is interpreted as involving the incorporation of more factors previously locked in the ceteris paribus pound. Thus, he writes:

Each exact and firm handling of a narrow issue, however, helps towards treating broader issues, in which that narrow issue is contained, more exactly than would otherwise have been possible. With each step of advance more things can be let out of the pound; exact discussions can be made less abstract, realistic discussions can be made less inexact than was possible at an earlier stage (p. 40).

This statical method carries ontological presuppositions at odds with Marshall’s evolutionary vision. Such a statical method is likely to be relevant only where certain conditions can be shown to be satisfied. Specifically, for the method to be successfully applied the features treated as isolated in thought would need to be real tendencies and the nature of the causal elements considered in isolation as a first step not knowingly portrayed incorrectly. In order for this condition to hold the way a causal element operates as it is found in reality would need to be the same as it operates in isolation from some other factors affecting the total outcome and this runs counter to the social ontological insights Marshall, at a general level, embraces. Also, if analysis of the operation of causal elements in assumed isolation is to reliably advance an understanding of the broader system they in part constitute the effects of the different elements would need to combine in an additive, or merely mechanical, fashion. This amounts to an assumption, one clashing with Marshall’s evolutionary vision, that there are no emergent powers of the more complex entity or whole irreducible to those of its parts considered separately.

Marshall offers an (ultimately unconvincing) defence of the deployment of the statical method which tends to run together methods of abstraction and isolation. Marshall (1898) clearly considers the statical method and its results to be abstract in the sense of being artificial, fictional or imaginary constructs. He writes:

It deals with abstractions: and refers to realities for the purpose of illustration only, not of construction. Its aim is not so much the acquisition of knowledge, as of power; power to order and arrange knowledge, especially with reference to the eternal opposition of forces impelling people to do and forces holding them back (p. 52).

Theory constructed on the basis of his statical method is understood by Marshall as being abstract in the sense of dealing with fictional or imaginary worlds. Such an association of the abstract with the fictional points to a failure to adequately elaborate on the nature of abstraction and differentiate approaches relying on abstraction from isolationist strategies. In theoretical science, the aim is not to explain happenings in the world in all their detail. Rather, the objective is to explain certain aspects of those happenings by abstracting, which involves the stripping away in thought all aspects of those happenings that are not the target of explanation. Abstraction may involve, for example, momentarily concentrating on a feature of the concrete, holding all the rest in place, before refocussing on the whole. The use of abstraction does not imply the kind of distortion associated with those isolationist strategies that treat an open system as a closed one. ¹⁴

While acknowledging the fictitious status of the results emerging from the statical method and conceding the method is only appropriate for preliminary investigations, Marshall insists it is a necessary starting point. He suggests greater realisticness follows as more, previously excluded, factors are brought back into the analysis, with the ‘abstract’ quality of the results achieved gradually diminishing. The success of this method depends crucially upon the conditions of isolation being treated as falsely unreal rather than the causal element itself. Yet it is the latter that the statical method, as deployed by Marshall, seems to involve: assumptions adopted serve to distort rather than merely simplify. Marshall’s analyses based on his statical method do not correspond to the study of the undisturbed operation of a stable mechanism in some painstakingly constructed isolated physical system that contributes to an understanding of what happens outside of that context where numerous forces are in play. Nor do they arise from stripping away in thought inessential details through abstraction properly conceived. The danger for Marshall is that analysis guided by the statical method ends up addressing only invented, rather than real world, systems.

Marshall’s defence of the use of the statical method is undermined when the high degree of social inter-relatedness is fully considered, something he partially acknowledges but fails to recognise the full consequences of. To treat some feature of the social realm as if it existed in isolation is unlikely to promote the development of explanatory powerful theory if it is essentially related to excluded characteristics. The statical method with its recourse to ‘breaking up’ the object of study is ill equipped to facilitate the analysis of essentially relational social processes and therefore is poorly adapted to a causal processual ontology and carries presuppositions not at all aligned with it. ¹⁵ Marshall’s attempted defences of the statical method are indicative of the remaining influence of a taxonomic conception. ¹⁶

Thus, to the extent that Marshall has an evolutionary vision it is accompanied by the retention of a taxonomic conception of science. The latter is especially clearly expressed as Marshall seeks to reconcile science, as he understands its essential features to be, with his recognition of the open, processual nature of social reality. Despite the tentative acknowledgement of a causal processual social ontology there is little by way of transformation in Marshall’s underlying taxonomic conception. The continuity at this level is expressed in the desire to maintain consistency between his preliminary discussions and any advanced analysis:

‘The growing prominence of what has been called the biological view of the science has tended to throw the notions of economic law and measurement into the background; as though such notions were too hard and rigid to be applied to the living and ever changing economic organism. But biology itself teaches us that the vertebrate organisms are the most highly developed. The modern economic organism is vertebrate; and the science which deals with it should not be invertebrate. It should have the delicacy and sensitiveness of touch which are required for enabling it to adapt itself closely to the real phenomena of the world; but none the less it must have a firm backbone of careful reasoning and analysis (Marshall, 1890 [1961]: 769, from first edition).

Here Marshall anticipates economists in the future adopting methods appropriate to the nature of the material studied as conceived of in his evolutionary vision yet then places a constraint on this development by insisting that such advanced work and earlier preliminary analysis (which can be shown to be based on quite different ontological presuppositions) must remain mutually consistent.

Marshall’s neoclassicism

Above it has been shown that Marshall partially adopts an evolutionary vision and acknowledges features of a causal processual social ontology. Yet it has also been shown that Marshall retains a residual commitment to a taxonomic view of science. He effectively took for granted that scientific results should take the form of correlations or event associations and thereby imports a quite different set of ontological presuppositions. Is it possible to identify the kind of tensions in Marshall’s work expected of someone whose underlying commitments can be characterised in such a fashion? If no such tensions can be discerned then Veblen’s interpretation of Marshall becomes impossible to sustain.

Marshall argues that economics must deal with human beings as they are and acknowledges that they are motivated by more than a desire to accumulate wealth but also suggests that it is entirely legitimate, indeed productive, for economists to focus narrowly on a limited range of steadily operating motives. The taxonomic view, with its emphasis on the need for results to take the form of correlations, uniformities or event associations, clearly informs Marshall’s discussion of the distinctiveness of economics. A key feature of his characterisation of economics relates to the motives which he sees it as being concerned with, he suggests that these are more powerful and steady in their operation than those considered by other branches of social science. He writes,

[Economists] deal with man as he is: but being concerned chiefly with those aspects of life in which the action of motive is so regular that it can be predicted, and the estimate of the motor forces can be verified by results, they have established their work on a scientific basis. For in the first place, they deal with facts which can be observed, and quantities which can be measured and recorded; so that when differences of opinion arise with regard to them, the differences can be brought to the test of public well established records; and thus science obtains a solid basis on which to work (Marshall, 1890 [1961]: 27, from third edition 1895).

Economics is here understood as being focussed on motives that cause people to act in regular patterns. Measurement facilitates the identification and elaboration of such associations. In setting out the potential of economics as a science, Marshall emphasises that the strength of the motives, though not their quality or character, can often be measured by means of a money yardstick. He notes: ‘an opening is made for the methods and tests of science as soon as the force of a person’s motives themselves– can be approximately measured by the sum of money, which he will just give up in order to secure a desired satisfaction; or again by the sum which is just required to induce him to undergo a certain fatigue’ (Marshall, 1890 [1961]: 15, from fifth edition 1907). Where motives cannot be measured then ‘the general machinery of economic analysis and reasoning cannot get any grip’ (Marshall, 1890 [1961]: 28, from third edition 1895).

Marshall (1890 [1961]) notes that it is the possibility of measuring economic forces that both distinguishes economics from other social sciences and explains what he sees as its relatively more advanced state:

But, for all that, the steadiest motive to ordinary business work is the desire for the pay which is the material reward of work. The pay may be on its way to be spent selfishly or unselfishly, for noble or base ends; and here the variety of human nature comes into play. But the motive is supplied by a definite and exact money measurement of the steadiest motives in business life, which has enabled economics far to outrun every other branch of the study of man (p. 14, from fifth edition 1907).

The claim is that economics addresses forces that allow proper scientific methods to be applied. Marshall sees economics as having made significant progress because it is not directly concerned with inaccessible psychological motives or buried social conditions. Marshall, in seeking to establish and promote economics as a science, emphasises the measurement and tracking of events together with the identification of event associations this makes possible. This preoccupation with regular behaviour patterns and their measurement is, of course, in line with the taxonomic conception but at odds with his evolutionary vision.

Marshall recognises customs as often unacknowledged conditions that provide the basis for (enable and constrain) action and maintains that any atomistic conception of the social is inadequate. However, when considering what economists ought primarily to focus upon social structure, institutional conditions and so on, are not emphasised. Instead of pursuing the study of these various social structures with methods crafted to their natures, he considers them as a constraint on the employment of the appropriate scientific method. He insists that economics is ‘specifically concerned’ with ‘the side of life . . . in which man’s conduct is most deliberate, and in which he most often reckons up the advantages and disadvantages of any particular action’ (Marshall, 1890 [1961]: 20–21, from third edition 1895). Marshall (1890 [1961]) argues that money provides ‘the one convenient medium of measuring human motive on a large scale’ (p. 22, from second edition 1891). Conditioning social structures serve to limit the scope of scientific study because in approaching such phenomena the economist cannot deploy scientific method but must instead rely on instinct. Marshall (1890 [1961]) goes so far as to suggest that such social structures may have to be regarded as unknowable:

There is a large and debatable ground in which economic considerations are of considerable but not dominant importance; and each economist may reasonably decide for himself how far he will extend his labours over that ground. He will be able to speak with less and less confidence the further he gets away from his central stronghold, and the more he concerns himself with conditions of life and with motives of action which cannot be brought to some extent at least within the grasp of scientific method. Whenever he occupies himself largely with conditions and motives, the manifestations of which are not reducible to any definite standard, he must forgo nearly all aid and support from the observations and the thought of others at home and abroad, in this and earlier generations; he must depend mainly on his own instincts and conjectures; he must speak with all the diffidence that belongs to an individual judgement. But if when straying far into less known and less knowable regions of social study he does his work carefully, and with a full consciousness of its limitations, he will have done excellent service (p. 780, from fifth edition 1907).

For Marshall, while the analysis of institutions and customs provides important historical detail it is not at the forefront of his theoretical analysis. He sees such phenomena as not being amenable to truly scientific treatment, requiring the economist to resort to more primitive and unreliable procedures. Marshall acknowledges the existence of social structures and provides some partial insights but minimises their significance when it comes to his characterisation of economics. He even regards their examination as beyond the reach of scientific method. This ambivalence is just what might be expected of a theorist tentatively moving towards a causal processual ontology while simultaneously remaining committed to a taxonomic conception of science.

A further important tension running through Marshall’s contributions relates to his orientation towards equilibrium theorising – he both acknowledges its severe limitations and yet remains reluctant to set it aside. It is worth noting here that Marshall is primarily concerned with equilibrium as an ontic notion. He is interested in understanding the extent to which equilibrium theorising can promote understanding of real world situations or causation. ¹⁷ Marshall (1898) understands equilibrium as a balance of forces within the economic system he is attempting to represent:

The words “balance” and “equilibrium” belong originally to the older science, physics; whence they have been taken over by biology. In the earlier stages of economics, we think of demand and supply as crude forces pressing against one another, and tending towards a mechanical equilibrium; but in the later stages, the balance or equilibrium is conceived not as between crude mechanical forces, but as between the organic forces of life and decay. The healthy boy grows stronger every year; but with early manhood there is some loss of agility; the zenith of his power is reached perhaps at twenty-five for such a game as racquets. For other corporeal activities the zenith comes at thirty or later. For some kinds of mental work it comes rather late; for statesmanship, for instance, it comes very late. In each case the forces of life preponderate at first; then those of crystallisation and decay attain to equal terms, and there is balance - or equilibrium; afterwards decay predominates . . . The balance, or equilibrium, of demand and supply obtains ever more of this biological tone in the more advanced stages of economics (p. 43).

Marshall’s recognition of the open, processual nature of the social realm leads him to view equilibrium with hesitancy and caution at least as a claim about the existing state of affairs. He suggests that the ‘movements of the scale of production about its position of equilibrium’ may be as complex as those of ‘a stone hanging freely from a string [which is] supposed to hang in the troubled waters of a mill-race, whose stream was at one time allowed to flow freely and at another partially cut off, [when] the person holding the string swings his hand with movements partly rhythmical and partly arbitrary’ (Marshall, 1890 [1961]: 346). There is at the same time an emphasis on real world tendencies in the equilibrium direction: ‘the demand and supply schedules do not in practice remain unchanged for a long time together, but are constantly being changed; and every change in them alters the equilibrium amount and the equilibrium price, and thus gives new positions to the centres about which the amount and the price tend to oscillate’ (Marshall, 1890 [1961]: 346–347, from first edition 1890 with revisions from fifth edition 1907).

Marshall (1890 [1961]), even as he defends the deployment of equilibrium concepts on the basis that they aid tractability or the precision of the analysis, emphasises that excessive reliance upon such constructs is likely to mislead:

Pushed to its more remote and intricate logical consequences it slips away from conditions of real life . . . it is especially needful to remember that economic problems are imperfectly presented when they are treated as problems of statical equilibrium and not of organic growth. For though the statical treatment alone can give us definiteness and precision of thought, and is therefore a necessary introduction to a more philosophic treatment of society as an organism; it is yet only an introduction . . . its limitations are so constantly overlooked, especially by those who approach it from an abstract point of view, that there is a danger in throwing it into definite form at all (p. 461 from fourth edition 1898). ¹⁸

For Marshall equilibrium theorising, despite all its dangers, is not to be abandoned, it remains an essential stage that an appropriate, scientifically orientated, economics passes through. The analysis of equilibrium Marshall often regards as an abstract theoretical device that can inform more applied studies of tendencies in the equilibrium direction. He sees the analysis of statical equilibrium as preliminary and claims that eventually he will be able to lead the reader, step by step, from this imaginary world of simple equilibrium configurations towards an understanding of real world problems. ¹⁹ Marshall fails to reveal how hypothetical models of equilibrium can aid understanding of a social world characterised by genesis, growth, variation and process. Revealingly Marshall (1898) claims that even if no real world tendency in the equilibrium direction is present this does not constitute a fatal blow to equilibrium theorising:

Even where . . . there is no true oscillation, the best device that has yet been found, for studying, analysing and comparing the chief economic forces at work anywhere, for ordering and coordinating facts about them, is to isolate them for the time. When so isolated they will almost always show an equilibrium point . . .. The conception of this equilibrium point helps give precision to the ideas (p. 52).

Marshall’s evolutionary vision and concern with the nature of actually existing economic systems encourages him to qualify his reliance on equilibrium theorising. It is his retention of a taxonomic point of view and failure to interrogate sufficiently the (ontological) presuppositions of his own methods that prevent him from abandoning it entirely. For Veblen (1906) the adoption of a modern point of view, one without the taint of taxonomic postulates, involves setting aside equilibrium entirely – ‘In Darwinism there is no such final or perfect term, and no definitive equilibrium’ (p. 582). Had Marshall been able to shrug off the taxonomic point of view his acknowledgement that social reality is fundamentally processual may have spurred him onto a more sustained form of ontological reasoning and led him to anticipate Veblen in perceiving the irrelevancy of equilibrium frameworks.

Due to his recognition of the open, processual nature of the social realm, Marshall, in elaborating his methodological position in the Principles and Industry and Trade, characterises law like statements not as strict event regularities but as tendency statements. In considering Marshall’s remarks on method it is possible to discern two rather disconnected accounts of tendencies.

In the dominant account of tendencies Marshall in acknowledging the openness of social systems feels forced to conceptualise law like statements as hypothetical and operating only where certain conditions hold. Marshall (1890 [1961]) summarises the content of book 1, chapter 3, section 3 of the Principles with the observation that ‘nearly all laws of science are statements of tendencies’ (p. 31, from fifth edition 1907). In Industry and Trade he writes that science ‘groups together for investigation phenomena which are fundamentally akin so that the study of many particular relations between them may set thought on the track of general rules of “laws” of causation, tendency, or co-existence’ (Marshall, 1919 [1923]: 203). Marshall typically ties the term ‘tendency’ itself to associations of events. Most straightforwardly in this regard Marshall links tendencies to results that would come about in appropriate circumstances. That is, Marshall frequently uses the term ‘tendency’ to refer to a situation that would be actualised under certain ideal conditions. Thus, he writes: ‘A law of social science, or a Social Law, is a statement of social tendencies; that is, a statement that a certain course of action may be expected under certain conditions from the members of a social group’ (Marshall, 1890 [1961]: 33, from second edition 1891). Marshall here seems to restrict the use of law statements to conditions wherein event associations would come about. This is ultimately a very limited and debilitating conception, since the focus of science is then exclusively on situations that are rare in the natural realm and more or less without counterpart in social systems.

The recognition that social phenomena are generated in open systems leads Marshall (1890 [1961]), echoing the view of Cairnes that Veblen had critiqued, ²⁰ to suggest that law like statements must not only be viewed as tendencies but also conceived of as hypothetical:

It is sometimes said that the laws of economics are “hypothetical”: Of course, like every other science, it undertakes to study the effects which will be produced by certain causes, not absolutely, but subject to the condition that other things are equal, and that the causes are able to work out their effects undisturbed. Almost every scientific doctrine, when carefully and formally stated, will be found to contain some proviso to the effect that other things are equal: the action of the causes in question is supposed to be isolated; certain effects are attributed to them, but only on the hypothesis that no cause is permitted to enter except those distinctly allowed for (p. 36, from third edition 1895).

Marshall, at times, seems to be struggling to move beyond a conception of tendencies as a form of empirical generalisation towards a quite different, more sustainable, characterisation, one in which the term is deployed to express a force that is in play and having effects whatever the accompanying conditions may be and irrespective of the resultant actual outcome. For example, in explaining that the law of gravitation is a statement of tendencies he writes: ‘Every cause has a tendency to produce some definite result if nothing occurs to hinder it. Thus gravitation tends to make things fall to the ground: but when a balloon is full of gas lighter than air, the pressure of the air will make it rise in spite of the tendency of gravitation to make it fall’ (Marshall, 1890 [1961]: 31, from fifth edition 1907). Here the law is seemingly interpreted as an unconditional statement about an operative mechanism rather than an exception-prone generalisation – gravity pulls on the balloon even as other, here counteracting, forces are in play. The co-existence of two accounts of tendencies in Marshall is all of a piece with him sustaining simultaneously both an evolutionary vision and a taxonomic conception of science. The influence on Marshall of the taxonomic conception is shown in how underdeveloped this alternative reading of tendencies remains in comparison to his dominant use of the term to refer to a projected result, one likely to come about only under specific conditions. ²¹

Marshall, at times, moves towards a causal processual social ontology, yet he is unable to systematically develop or deploy methods aligned to such ontological commitments. Marshall is not simply constrained by a taxonomic view but is characterised by a fundamental discrepancy at the level of ontology. The methods he sees as most scientifically credible are ones carrying ontological presuppositions that are at odds with his evolutionary vision. This internal ontological clash is expressed in a series of prominent tensions in his work. Marshall insists that economists recognise the motivational complexity of human agents and yet recommends that they focus primarily on those behaviours that are regular and predictable. He recognises the need to incorporate institutional factors into his analysis but says such factors cannot be addressed in an adequate scientific manner. He aims to transcend equilibrium theorising because of its severe limitations but sees equilibrium as an essential building block for scientific advance. The clash is expressed also in ambiguities in the very categories Marshall deploys – he simultaneously views tendencies as second best empirical generalisations and as unconditional statements about operative mechanisms. ²²

Concluding remarks

When considering the origins of the category neoclassical economics Lawson locates it in the last instalment of Veblen’s preconceptions essay series. Veblen coins the term to identify an inconsistency of significant consequence in the work of a group of his contemporaries which included Marshall as one of its most able members. Veblen, having demonstrated the adherence of classical economists to a taxonomic conception of science, proceeds to understand Marshall as struggling and ultimately failing to escape from it. Importantly Marshall is understood as acknowledging features of a causal processual social ontology but unable to develop or deploy methods suited to it due to a residual commitment to a taxonomic view of science that dominates.

In this paper I have shown that the assessment of Marshall which Veblen himself only briefly sketches (and that Lawson reports on but does not critically evaluate) can be defended. There is evidence in Marshall of an evolutionary vision but also of the retention of a taxonomic conception of science and a failure to question adequately the ontological presuppositions of the methods he deploys. There are tensions in Marshall’s contributions that can be understood as generated by a clash of ontological preconceptions. The associationist ontology presupposed by his methods and retained commitment to a taxonomic point of view cannot be rendered consistent with his tentative acknowledgement of features of a causal processual ontology implied in his evolutionary vision. What Veblen means by neoclassical economics can be better understood by seeing how the ontological discrepancy thesis works out in detail in the case of Marshall. It seems likely that recognising this ontological discrepancy is crucial for any adequate understanding of Marshall’s contributions.

Lawson argues that Veblen’s use of the term neoclassical to refer to contributions that are characterised by a discrepancy at the level of ontology is not only of historical interest but remains of contemporary significance. He suggests that three groupings can be discerned among contemporary economists. First there are those who adopt a taxonomic approach to science, promote mathematical deductivism as an orientation to science to be adopted by all and who regard any questioning of this stance as misguided. Then there are those who are aware that social reality is of a causal-processual nature, prioritise the objective of being realistic and craft methods that are consistent with this ontological understanding. Finally, there are, he suggests, those who recognise at some level that social reality is of a causal processual nature and explicitly prioritise the goal of being realistic and yet remain committed to a taxonomic approach (see Lawson, 2013: 978–979). For Lawson this third group, who express – albeit rather loosely – realistic visions of an open system of cumulative causation and yet seek to provide insight using methods that presuppose an associationist ontology, are Marshall’s modern counterparts. Lawson argues that the members of this group are inconsistent at an ontological level and he suggests progress within heterodox economics and the discipline more broadly can be made by continually challenging contemporary neoclassical economists to be more consistent and recognise the methodological implications of the processual social ontology that they, at some level, acknowledge. ²³

It is surely a tribute to Marshall to note that he was always seeking consistency – Nabers (1943: 17) goes so far as to suggest that he had an ‘almost compulsive demand for consistency at all levels of his analysis’. The frustration for Marshall, as Veblen saw, was that he sought to achieve consistency within a framework that privileged a taxonomic conception which served to ensure that the task was one that could not be coherently progressed.