Reorienting Economics: Some epistemological issues

In setting out some reactions to Tony Lawson’s new book, Reorienting Economics, the first point to emphasize is how welcome the volume is. It is an evolution from Economics and Reality in a variety of ways. It sets out the development of Lawson’s thinking, often explicitly in response to reactions to the earlier book. In particular, Lawson has taken the opportunity to address (mistaken) dualistic readings of the earlier book (a common feature of reactions to methodological work of this sort) and so to emphasise the character of the middle ground between positivism and relativism. He reiterates the critical realist critique of orthodox methodology, but much more succinctly, allowing more latitude for looking beyond critique and addressing issues arising within the critical realist approach. Not only does this succinctness further clarify the critical realist project, but it also means that the tone is more constructive, as is proper for a maturing set of ideas. The key message of Reorienting Economics, as of Economics and Reality, is the importance of ontological argument. In the Preface, Lawson specifies the aims of engaging in ontological argument as being:


INTRODUCTION
In setting out some reactions to Tony Lawson's new book, Reorienting Economics, the first point to emphasize is how welcome the volume is. It is an evolution from Economics and Reality in a variety of ways. It sets out the development of Lawson's thinking, often explicitly in response to reactions to the earlier book. In particular, Lawson has taken the opportunity to address (mistaken) dualistic readings of the earlier book (a common feature of reactions to methodological work of this sort) and so to emphasise the character of the middle ground between positivism and relativism. He reiterates the critical realist critique of orthodox methodology, but much more succinctly, allowing more latitude for looking beyond critique and addressing issues arising within the critical realist approach. Not only does this succinctness further clarify the critical realist project, but it also means that the tone is more constructive, as is proper for a maturing set of ideas.
The key message of Reorienting Economics, as of Economics and Reality, is the importance of ontological argument. In the Preface, Lawson specifies the aims of engaging in ontological argument as being: 1. to identify the conception of reality implicit in economic practice 2. thereby to draw attention to inconsistencies between that conception and what practitioners actually conceive as the nature of realitythis is aimed particularly at orthodox economics 3. to elaborate a general conception of the broad nature and construction of society 4. to build knowledge pluralistically on the basis of this ontology, with a division of labour employed within social science, and within economics, according to differences in ontological commitment.
In this essay, I will take it as given that it is important to focus on the ontological level in order to discuss epistemological and methodological issues. From this base, I want to raise two (related) sets of issues with respect to the specific aims noted above. One is the status of the 'general conception of the broad nature and construction of society', and whether difference in 'ontological commitment' is a satisfactory way of understanding schools of thought. How far is it reasonable to expect agreement at the ontological level? The underlying issue is the extent to which it is possible to focus on ontology without straying into epistemology. This is not to differ with Lawson over the importance attached to ontology, but rather to raise questions about its scope in pure form. The second issue concerns the epistemological impact of the book itself, and raises questions about persuasiveness among different audiences. If the scope for ontological agreement is less than Lawson implies, then the argument may come across as being more positivist than it actually is. This provides further reason for addressing the scope for ontological differences within the grey area between ontology and epistemology.

THE SCOPE FOR AGREEMENT ON ONTOLOGY
Critical realist analysis builds crucially on transcendental analysis. Here there is only scant reference (confined to footnotes) to Roy Bhaskar, whose transcendental realism had a more central role in earlier expositions. Indeed the transcendental argument is put here in a way which connects more directly with economists' own experience, thus avoiding some of the philosophical issues which have arisen over Bhaskar. Lawson explains (in Chapter 2) that transcendental analysis involves a special form of retroductive argument, applying pure reason to generalized conceptions of human experience in order to work out what must be the case about social reality for knowledge of causal mechanisms to be possible. Lawson is careful to specify (page 34) that transcendental arguments are contingent on the experience to which the arguments are applied, and our conception of it. They are fallible. But, far from drawing the implication that different transcendental arguments and thus different ontologies are possible, Lawson focuses on aspiring to a common, open-system, ontology.
The emphasis in Economics and Reality was on the inconsistency between the methodology of orthodox economics, suited to a closed-system ontology, and the open nature of reality which transcendental analysis demonstrates to be the case. Here the argument moves on to address the ontological foundations of heterodox economics. Lawson makes the interesting argument (in Chapters 8 and 9) that institutional economics and feminist economics, by emphasizing particularity of experience, go too far in the direction of relativism. But what I want to focus on here is the argument (pp. 180-3) that heterodox schools of thought (actually or potentially) share a common ontology and differ only in 'ontological commitment', that is, having different concerns and asking different questions about that shared ontology. Schools of thought therefore, according to Lawson, differ 'at a level of abstraction below that of social ontology but above that of relatively concrete social-scientific explanations of highly specific phenomena' (p. 181).
One way of expressing the generalized conception of social reality which could be said to be held in common within most (at least) of heterodox economics is that there is a shared view that social reality is an open system. It is on these grounds, for example, that the different heterodox schools of thought adopt methodologies within which mathematical formalism plays only a partial (if any) role. The subject matter requires a more pluralistic methodology. Schools of thought are better identified at the level of methodology than at the level of theory. To that extent I agree, and an attempt at a more elaborate spelling out of what can provide shared foundations for heterodox economics is welcome.
But we need to consider further what we mean by open-systems social ontology. We may agree that reality, independent of our knowledge of it, as far as we can tell (given the nature of the human condition), is open. But this agreement, fundamentally important though it is, applies to a relatively 'thin' layer of understanding, referring to what we may call 'pure' ontology. Any further discussion requires that we depart from this transcendental notion of openness. As soon as we start conceptualizing the economic system, we inevitably invoke closures (normally of a provisional, incomplete, sort). Epistemology cannot be conceived of as an open system in the same pure sense as social ontology.
The very words we use involve closures. Our conception of reality requires categorization (identifiable individuals, products, firms, institutions and so on). There is intrinsic and extrinsic openness, but we cannot even discuss that without some form of specification of the system and its parts (Chick and Dow 2001;Chick 2004). Further, epistemological categorizations which are embedded in institutional design and which frame knowledge and behaviour, in turn have consequences at the ontological level. Lawson's ontological account of social reality draws heavily on real structures which derive from the epistemological level. Rules of the road are designed, for example, on the basis of knowledge about the real workings of road travel. Real hardship may arise from unavailability of credit due to monetary policy, where the institutional structure within which monetary policy is framed, as well as the policy itself, derive from monetary theory. Epistemology, which inevitably involves (provisional, partial) closures, feeds back into ontology.
Even if our conception of generalized experience leads us to conclude that ultimately reality is open in a pure sense, there is no reason to expect that there will be a shared conception of that open reality, even among heterodox economists (as Lawson himself points out, transcendental argument is contingent on particular perceptions of experience). Indeed the evidence supports the contrary view, that, while the methodology of heterodox economics is consistent with an open-system ontology, there are significant differences within that general methodology and the theories which are generated which are consistent with different conceptions of the nature of that open system (Dow 1985(Dow , 1990). It is not just a matter of asking different questions (different ontological commitments, in Lawson's words), but of different conceptions of reality (whether reality is understood in terms of the individual or class, exchange or production, stability or instability, and so on). A recognition of a shared open-system 'pure' ontology is important for cross-fertilization of ideas among heterodox schools of thought, and the evolution of theorizing in relation to a changing reality more generally. But both reason (about the impossibility of pure opensystem epistemology) and evidence suggest that schools of thought are associated with different versions of open-system ontology.

ISSUES OF PERSUASION
The ontological nature of the differences between schools of thought is an important issue which has implications also for the communication of critical realist ideas. The aim of the book is to persuade economists, by means of reason, to engage in ontological analysis and to explore a shared, open-system ontology. As a result, it is hoped that economists will recognize the inconsistency between an open system reality and orthodox methodology, and that a general open-systems ontology will be developed in such a way as to provide a shared foundation for economic analysis.
It is unlikely, given past evidence, that orthodox economists of long standing will be persuaded by a critique based on exposing a fundamental inconsistency. The approach employed in the Cambridge capital controversies of the 1960s, for example, was to demonstrate an internal inconsistency in orthodox theory. But the exposure of the reswitching inconsistency did not persuade orthodox economists to change approach; it was, in the end, ignored. The recent literature on the history of mathematics in economics (to which Chapter 10 of Reorienting Economics is a contribution), further reveals a changing attitude to consistency within orthodox economics. In the heyday of pure general equilibrium theory, inconsistency between theory and subject matter was regarded as acceptable -it was internal mathematical consistency that mattered, even though it proved impossible, ultimately, to achieve (and in the case of reswitching it did not prove to be decisive). But now that economics has fragmented away from pure general equilibrium theory, the notion of formal mathematical consistency may no longer be seen as paramount (see Weintraub 2002). Has there been a return to consistency between theory and subject matter as the primary criterion for rigour? But then the criteria for consistency between theory and subject matter remain unclear. As Sent (1998) shows, Sargent's struggle to ensure formal consistency did not meet with success.
If the charge of inconsistency is to be effective, then, the concept of consistency requires more detailed attention. Any form of abstraction from reality involves some form of inconsistency between reality and theory. It is therefore important which forms of inconsistency matter and which do not, and this will differ according to the methodology of whichever grouping we are considering. Since consistency is something which derives its meaning from the system at issue, therefore, attention needs to be addressed to different systems.
Let us consider systems which correspond to schools of thought. Probably most economists (including mainstream economists) would agree, at the transcendental level, that the world is open. But it is the closures which are then invoked to structure perception of experience and to allow analysis to proceed, and the way these closures are regarded (provisional or fixed, partial or complete), which characterise different schools of thought. This illustrates the significance of downplaying differences between heterodox schools of thought, distracting attention from the fact that each rests on a set of (provisional, partial) closures. Were Lawson to allow his focus to extend beyond the thin layer of complete openness at the transcendental level, what could be demonstrated is the way in which heterodox schools of thought choose to segment open reality in such a way as to understand some aspects of (open) reality. This can be contrasted with closures in orthodox economics which are invoked for other reasons, and treated in other ways. Orthodox economics does evolve, and does open up to new variables in order to address real issues. But it does so alongside closures which are inconsistent with that evolution and openness. The most obvious closure which is inconsistent with an open system ontology is the depiction of human behaviour in terms of optimization with respect to an exogenous, complete, consistent set of preferences. This is something which Davis (2003) suggests would be the most effective focus of critique.
There is always a generational aspect to persuasion, such that Lawson has a better chance of persuading younger orthodox economists that the inconsistencies in the approach are insupportable. But there is more to persuasion than reason alone (in the narrow, conventional, sense). The approach which Lawson advocates, in terms of understanding human nature within society, itself provides some basis for understanding the behaviour of economists within the profession. There is some discussion in the volume, for example, of fear of openness. Others have taken this further into the realm of psychology (Chick 1995, Earl 2000. Lawson (Chapter 10) also explores the structural, social forces behind the mathematisation of economics. There is scope for further pursuing these lines of enquiry in order to develop an ontological understanding of why different economists structure their thought in such different ways (thought processes -epistemology -being an important part of social ontology). This would inform expectations as to how far economists are amenable to a reorientation of the discipline, as well as indicating the types of argument which are and are not persuasive to different audiences.

CONCLUSION
This set of comments reflects a concern that critical realism still might be (mis)interpreted in a dualistic way. Lawson is careful to point out that he is addressing the level of pure social ontology, not prescribing practice. But there is a danger in pursuing too sharp a distinction between ontology and epistemology. It is suggested here, on philosophical, empirical and strategic grounds that critical realist analysis should extend beyond the level of pure ontology to take explicit account of closures within an open system, in the substantial grey area between ontology and epistemology.
Sheila C. Dow University of Stirling s.c.dow@stir.ac.uk