The ontological, epistemological and social aspects of the intellectual divide

In document Disciplined reasoning Styles of reasoning and the mainstream-heterodoxy divide in Swedish economics Hylmö, Anders (Page 55-63)

Chapter 2. Mainstream and

3. The ontological, epistemological and social aspects of the intellectual divide

The previous section surveyed some of the major positions that discuss intellectual divides in economics in substantive or ontological terms. However, other authors place emphasis on other aspects of the divide. I suggest that we can think of three relevant aspects of the divided structure of an intellectual field like economics.

These are the ontological, epistemological, and social aspects. I will first briefly summarise the first, ontological aspect already touched on in the previous section.

I will then turn to the role of methodological conceptions and practices emphasised by Lawson, or what I suggest we think of as an epistemological aspect of the intellectual divide. This is the notion that mainstream economics is fundamentally about a particular orientation towards certain methods and views

on how proper knowledge should be produced. Third, I turn to more sociologically familiar ground with the social aspect, emphasising the importance of understanding the relational nature of intellectual conflict and the practice of boundary work involved in establishing professional and scientific legitimacy and authority.

The first aspect: Ontological assumptions and presuppositions

The first aspect relates to differences in core beliefs or axioms of social ontology, that is, fundamental ideas about how the aspect of the world one studies is made up. The core axioms of any science constitute its objects of study and direct attention to certain aspects of the world while excluding other aspects. It is useful to think of different authors drawing lines at slightly different levels of abstraction.

Where to draw the line depends on what we wish to highlight. Are we interested in theoretical differences between schools of thought, for example in how they conceive of rationality? Or are we interested in similarities at the more fundamental level of social ontology, that is, about the basic presumptions about the nature of actors, markets, and so on? Here, I think it is useful to take the Arnsperger-Varoufakis approach, of thinking in terms of meta-axioms. This illuminates what many heterodox critics find to be a common ground, what they take to be their common problem with the neoclassical mainstream.

We may think about these distinctions as layered in a hierarchical way, in the sense that a few basic meta-axioms may be compatible with broader substantial theories. With such a model, we may then think about the formation of consensus and disagreement at different levels of abstraction. That is, there may be disagreement on the level of substantial theories, although there is still consensus about the underlying fundamentals. It is then a matter of shared beliefs about ontology rather than theory. These axiomatic ontological beliefs, that is, beliefs about fundamental aspects of reality that are taken for granted and never questioned as such, may then also be conceived at different levels of abstraction.

We can think about them in terms of preceptions (as Veblen did), that is, the lower level meta-axioms that are not necessarily even explicit or explicable. This just means that holding some belief presupposes another at a higher level of abstraction. This is exactly the point that Arnsperger and Varoufakis and Lawson make, although from slightly different angles. Whereas the former authors are interested in meta-axioms as theoretical presuppositions, Lawson shifts his focus onto the ontological presuppositions of certain methods.

The second aspect: Epistemology and methodological ideals

The second aspect of the intellectual divide is epistemological. This is exemplified by the approach taken by Lawson, and as we have seen, also mentioned by other authors: the understanding of mainstream economics as a research programme unified on methodological rather than axiomatic ontological grounds. Lawson (2013) turns against not only the use of neoclassical to describe the core of mainstream economics, but also any attempt at defining mainstream economics in terms of some set of substantial theoretical assumptions or axioms (what I call ontology here).

He argues that this common approach is not useful from the point of view of a critical heterodoxy, and that it only leads to a superficial critique that does not get to the heart of the problem. Instead, he points to the centrality of the shared methodology of mainstream economics, and as a consequence of that, what he argues are the implicit ontological assumptions that comes as part and parcel of certain methodologies. In order to understand the problems of the mainstream, it is necessary to inquire into the more basic problem of the relation between methodology and ontological assumptions.

Weintraub, while discussing core theoretical assumptions, also emphasises the role of mathematics in modern economics. Weintraub (2002a) has studied the mathematisation of economics at depth, and points to the increasing role of formal mathematical modelling in modern economics. Colander and his co-authors (2000; Colander et al. 2004), while discussing the changing substantial features of the mainstream, also note the centrality of methodology, if only in passing. Colander finds that the modelling approach, rather than any substantial content, is the central attribute of modern economics. Following Robert Solow and Jürg Niehans, he argues that “the modelling approach to problems is the central element of modern economics” (Colander 2000:137; emphasis in original), and concludes with Niehans that our time should be characterised as “the era of modelling” (2000:141). The mention, in passing, of a core of methodological principles remains the same in the 2004 paper: “Our view is that the current elite are relatively open minded when it comes to new ideas, but quite close minded when it comes to alternative methodologies”(Colander et al. 2004:493). Even if the authors are keen to emphasise the openmindedness of those elite actors in the profession in their argument for a mainstream pluralism, they admit that there is

“unconscious suppression” of heterodox views that does not fit nicely with the elite’s way of thinking.23 In such suppression, methodology is an important tool

23The authors centre their account on the elusive concept of “elite of the profession” as the central actor in the process of intellectual change, a notion they claim is “understood by those in the

(Colander et al. 2004:493). However complexity-oriented and openminded the elite may be when it comes to theory, there seems to still be one right approach when it comes to methodology: “If it isn’t modelled, it isn’t economics, no matter how insightful” (Colander et al. 2004:492).

Lawson comments on this observation, but is not wholly satisfied by the limited role Colander and his co-authors assign to it: “I am not sure they fully appreciate the significance of their observation (they give it little emphasis)”. Instead, he wants to put full pressure on exactly this point. It is this methodological orientation which is the defining feature of contemporary mainstream economics:

The mainstream project of modern economics just is an insistence, as a discipline-wide principle, that economic phenomena be investigated using only certain mathematical-deductive form of reasoning. This is the mainstream conception of proper economics. It is the one feature or presupposition that remains common to (if not always explicitly formulated in) all contributions regarded as mainstream, remaining in place throughout all the project’s theoretical fads and fashions.

(Lawson 2006:492)

This distancing from the Colander-Davis view of mainstream pluralism is repeated with full force in Lawson’s 2013 piece, where he holds that:

whilst the concrete substantive content, focus and policy orientations of [the mainstream tradition] are highly heterogeneous and continually changing, the project itself is adequately characterised in terms of its enduring reliance, indeed, unceasing insistence, upon methods of mathematical modelling. (Lawson 2013:4, emphasis in original)

Lawson’s point is clear. To understand modern mainstream economics, we need to focus on its insistence on mathematical modelling. The problem of mainstream economics, according to Lawson (2013:7), may have many manifestations at the level of substantial theoretical claims and policy advice, but its real source lies at

“the level of methodology and social ontology” Lawson’s central argument is that the mathematical methods used bring with them certain ontological presuppositions that do not match the nature of social reality. This is where Lawson’s critical realist philosophy of science comes into play. He argues that the use of mathematical techniques like functions and calculus presuppose event regularities in closed systems. Lawson calls this methodological approach

profession”. There should be no doubt that the idea and actual importance of a disciplinary elite is very powerful in this profession (cf Fourcade, Ollion, and Algan 2015).

mathematical deductivism, the doctrine that explanations should be framed in terms of such closed systems and event regularities. A social ontology or model world that guarantees such event regularities must, according to Lawson, be a world of isolated “atoms” that, when triggered, have the same effect or action regardless of context. Note that I have talked about an ontological aspect here in much the same way as Lawson talks of “substantial axioms”. However, I think it is fruitful to think of the Arnsperger-Varoufakis formulation of three meta-axioms as a social ontology, that is, a fundamental assumption about the basic stuff of which the social world is made.24

There are also, Lawson (2013:8) notes, exceptions to such unrealistic assumptions, mainly in the form of modern, overly atheoretical econometrics. But this, he claims, just shows that the common ground remains the use of mathematical models. Lawson’s critique hinges on a critical realist argument about the discrepancy between the modellers’ presupposed atomistic deductivist ontology, and the nature of social reality. The problem is not the mathematical modelling per se, but rather that this methodology is not fit for purpose, since it cannot model the fundamental ontological openness of the social world:

The heavy use of these tools in conditions for which they are found to be inappropriate both explains the repeated explanatory failings of the discipline as well as why formulations are of a nature that are typically recognised by almost everyone as rather unrealistic. That, in summary, is the real cause of the discipline’s problems. (Lawson 2013:7)

These, then, are the outlines of the epistemological approach to the mainstream-heterodoxy divide in economics, which are an important concern for Colander and Weintraub, and central for Lawson. While Colander focuses loosely on “the modelling approach”, Weintraub has studied the increasing use of mathematics in economics. The mathematisation and modelling approach are also interrelated in important ways, but should not be seen as strictly identical. For example, Lawson (2013:27) points to the development of mathematics in the wake of the emergence of quantum physics in the early twentieth century as an important factor in the development of modern mathematical modelling. Whereas economists had always used some form of mathematics, previously this was

24 Lawson explicitly argues against the “substantial” approach of not only Weintraub’s idea of a

“meta-theory”, but also Arnsperger and Varoufakis’ idea of three meta-axioms, in favour of his own methodological/ ontological approach. However, the resulting view of ontological presuppositions is perhaps not so distant from the meta-axiom view, that is, Lawson arrives at a very similar conclusion if we look at the actual ontological presuppositions or meta-axioms.

strongly influenced by a mechanistic reductive worldview borrowed from physics, where essential properties of nature were expressed in mathematical form. With quantum physics, however, mathematics was increasingly seen as a tool for examining possible realities represented as axiomatic systems in their own right.

The need to interpret and apply the models in terms of real world events was thus effectively minimised: “In particular it was no longer regarded as necessary, or even relevant, to economic model construction to consider the nature of social reality, at least for the time being” (Lawson 2013:28; see also Backhouse 2002:259). If the ontological aspects focus on underlying assumptions about the nature of social reality, the epistemological approach focuses on assumptions, prescriptions and practices of social knowledge production. With that, let us turn to the social and symbolic organisation of actors in economic knowledge production.

The third aspect: The social nature of thought collectives in science The social nature of all knowledge, science included, is one of the oldest insights not only of science studies, but of modern sociology in general. The social aspect of understanding the intellectual field of economics addresses the social nature of knowledge production. A good entry point to this aspect is historian of economic thought Roger Backhouse’s (2004) “A Suggestion for Clarifying the Study of Dissent in Economics”.

Backhouse wants to provide a better understanding of the role of heterodoxy in economics. To do this, he draws upon the sociology of science to emphasise the role that dissent and controversies play in studies of the natural sciences,

“because they reveal things about science that would otherwise remain either concealed or obscured” (Backhouse 2004:262). Furthermore, he reminds us of the fundamental insight from science studies that “the resolution of controversy is a social process that determines the range of acceptable positions within the profession” (Backhouse 2004:264). We cannot just assume that scientific controversies are terminated through the establishment of truth by the objective arbiter of empirical data, necessarily moving us towards the singular truth of the matter. Instead, “interests, preconceptions, and ideology interact with the various theoretical and empirical techniques available to economists to produce an outcome that, in the circumstances, seems correct to those involved” (Backhouse 2004:264). Controversies are social affairs, the resolution of which establish what comes to be taken as truth; in this sociology of scientific knowledge view, truth is the outcome of social controversies, not an external input that determines their outcome.

Backhouse makes the important clarification, also drawing on science studies, that classifications should be naturalistic and workable. That is, they do not tell us which one of warring parties is wrong or right, but should be arrived at sociologically. Furthermore, definitions are not claimed to be absolute or exact, their “purpose is not to serve as the basis for formal logical deductions”

(Backhouse 2004:263); an important point to remember in the heated debate about the exact definition of neoclassical economics, and whether that entity is still hegemonic. Central to Backhouse’s argument about types of dissent is a distinction between, first, everyday disagreement on some topic which, second, may crystallise into more or less prolonged scientific controversy. A controversy may be resolved with one of the warring sides leaving as the victor, establishing the truth of the matter, or, third, a controversy may remain unresolved and turn into dissent. Dissent differs from the two first situations in that it implies asymmetric relations between insiders and outsiders/ dissenters.

Backhouse now helps us to make a distinction between dissent in general, and groups of dissenters that are in some way organised around dissenting views:

heterodoxy. He then suggests that “heterodoxy be defined as involving self-identification, sociology, and core beliefs. [. . .] It is heterodoxy as understood by, for example, the Association for Heterodox Economics” (Backhouse 2000, see also 2004). With such a strict definition, we can see how dissenting core beliefs, the focus of what we have called the ontological approach, are not sufficient conditions for classifying a position as heterodox, according to Backhouse. We must also take social relations (that is, exclusion from mainstream sites of power, but also organisation in alternative networks, journals and conferences), and self-identification into account. The self-self-identification aspect is central to authors like Lee (2009), and also acknowledged by Lawson (2006) as a fundamental aspect of modern heterodoxy. Lee especially emphasises that understanding heterodoxy means understanding heterodox economics in terms of a specific social community. Combining these with Davis’s (2008) position results in a synthesis that allows us to think more clearly about different forms of heterodoxy.

Backhouse also mentions a further aspect of dissent, where the dissenting party leaves the field altogether for another neighbouring discipline. Instead of staying within the discipline and struggling to change it, this means pursuing one’s goals elsewhere, for example in business schools or economic history departments. To this, one could add the creation of new academic disciplines, such as the creation of economic history as a discipline in its own right, strongly influenced by what had earlier been the German historical school in economics. In Backhouse’s terminology, Davis’ idea of the new mainstream heterodoxy (clearly distinct from traditional heterodoxy) could perhaps be understood as a dissenting mainstream

view, but not heterodox in the proper sense. It should also be noted that both authors point to the differences in how one relates to orthodox core beliefs.

Whereas critics of the traditional heterodox variety tend to be ruthless in their attacks on these beliefs, with the ultimate aim of having them replaced (at least partially, under the banner of pluralism), the mainstream heterodox seeks rather to reform some of orthodoxy’s core beliefs. Davis’s (2008) explanation of the existence of a marked orthodoxy-heterodoxy split, and the active use of this terminology in economics, deserves attention. He claims that periods of orthodoxy produce heterodoxy through a splitting process. Drawing on the sociological literature on symbolic boundaries between scientific fields and boundary work (Abbot, Bourdieu, Collins, Gieryn), Davis claims that we must understand the process of how sciences try to achieve greater autonomy as well as scientific and professional legitimacy. In any such process, the production of legitimacy is connected to the conventional approaches in each science. That is, the use of a common, generally acknowledged fund of knowledge works as a professional resource to increase the legitimacy of that common project. A science structured around such dominant conventional approaches enables the formation of the profession as a coherent social group (Davis 2008:351).

But why is this tendency so marked in economics? Davis claims that the explanation should be sought in the high level of policy exposure, where the high stakes in the form of the “wide scope and profound impact of the market in modern society” in combination with the high uncertainty of prognosis in economics (Davis was writing before 2008) leads to a situation where the orthodoxy/ heterodoxy split functions as a sort of defence mechanism that “allows economics to claim economics is scientific by dismissing heterodoxy as unscientific” (Davis 2008:352). He also points to the fact that the value-laden nature of economics (like any science) is obvious in a pluralist situation where different schools may debate substantial issues, while a situation with a dominant orthodoxy may suppress such impressions.

The focus of studies on boundary work in science (Gieryn 1983; Lamont and Molnár 2002) is on how social groups are created in interaction with the creation and reproduction of symbolic boundaries. The central argument in the literature is that the boundaries of science have to be maintained in practice through the continuous symbolic work of demarcating science from non-science. Symbolic boundaries are related to social boundaries, when the boundaries of a social group like a profession are maintained and determined through symbolic boundaries that defines belonging and status. As Davis (2008) argues, boundary work in economics functions through the equation of the mainstream approach with science, and heterodoxy with non-science. Another way of saying this is that

mainstream economists “command of the language of science and objectivity”

(Morgan 2015), and the claims of heterodox economists can be refuted as unscientific and thereby not worth taking seriously. In this way, symbolic boundaries are turned into social boundaries. These symbolic boundaries are also an integral part of the construction of collective identities, both the identity as just “economist” without any adjective, and the identity as “heterodox economist”, constructed in explicit opposition to mainstream identity.

The social aspect of understanding the mainstream-heterodoxy divide thus involves the notion of distinct thought collectives or social groups that hold different beliefs that stand in specific relations to each other, are involved in controversies where they employ various resources to promote their views, and whose members identify as members of the group. However, there is also a deeper sense in which these social relations and their dynamics are understood as social, namely as actively-maintained boundaries that link social groups and symbolic categories.

In document Disciplined reasoning Styles of reasoning and the mainstream-heterodoxy divide in Swedish economics Hylmö, Anders (Page 55-63)