Pragmatism, critical realism and the study of value

ABSTRACT

This paper examines the relationship between pragmatism and critical realism, first as alternative philosophies for the social sciences in general, and second, as an illustration, in the social study of monetary valuation. The paper argues that both traditions are internally diverse. Hence, the relations between the two are complex, with both substantial overlaps and real differences revealed in encounters between them. Perhaps the most significant difference is pragmatism’s distrust of invocations of structural power in social explanations, whereas realism encourages them, in interaction with other explanatory elements. The paper problematizes claims that recent work in the study of value is predominantly pragmatist. Nevertheless, it argues that pragmatist influence has encouraged valuation studies to focus on the micro level at the expense of the macro. From a realist perspective, however, there is much to be gained from an approach that embraces both micro and macro levels and the relations between them.

KEYWORDS:

• Critical realism
• foundationalism
• pragmatism
• valuation

Introduction

This paper examines the relationship between pragmatism and critical realism as alternative philosophies for the social sciences. These two traditions have often been positioned as competing, and usually as conflicting, with each other (e.g. Baert 2005; Kemp 2017; Kivinen and Piiroinen 2004; Martin 2013b; Porpora 2015). Yet neither tradition is monolithic: pragmatism makes a virtue of its internal pluralism but critical realism is also somewhat diverse. Indeed, one might argue that intellectual traditions in general tend to be composed of many different elements, with different advocates combining different elements in different ways. Hence, there may be diversity and disagreement within traditions and also elements that different traditions (or versions of them) share, though they may not understand them in exactly the same way as a consequence of embedding them in different frameworks. Both pragmatists and critical realists, for example, advocate a fallibilist understanding of knowledge. Furthermore, critical realists have acknowledged debts to pragmatist thinkers, and there are realist elements in the pragmatist mix. The relations between traditions, in other words, are often more complex than simple portrayals of them as complementary or contradictory might suggest (Collins 1998).

Although there are substantial overlaps between pragmatist and realist approaches to the social sciences there are also some real differences. Perhaps the most significant is that many pragmatists distrust and discourage invocations of structural power in social explanations, whereas realism encourages them, in interaction with other explanatory elements. This arguably produces a tendency for pragmatist work to focus on the micro level at the expense of the macro, although in practice some pragmatist authors stretch these boundaries.

The paper approaches these issues at two levels: first, by outlining as briefly as possible some key elements of the two traditions and reviewing some of the previous encounters between them, and second, by discussing how these differences manifest themselves in the study of one particular set of phenomena: the social study of monetary value, in particular the monetary value of financial assets. Although the study of value overlaps with social studies of finance, which are arguably more influenced by actor-network theory, it has become common in the recent literature in valuation studies for authors to assert a pragmatist perspective; indeed Barman argues that ‘It is widely acknowledged that the study of value in the social sciences has taken a pragmatist turn’ (Barman 2015, 12). However, there is relatively little debate in these fields about what pragmatism entails or which version of it is being deployed. This paper asks what it means to adopt a pragmatist perspective not only in general but also more specifically in the study of financial value. By contrast, explicitly realist perspectives are relatively absent in this area of study. Realism is either ignored or occasionally, and usually not in print, disparaged as an ‘other’ of pragmatism. Yet contemporary critical realism offers a promising alternative approach. The paper therefore examines the distinctions between pragmatist and critical realist approaches, their relevance to the field (a concept that will be problematized) and the consequences of the influence of pragmatism. It will also question the claim that the emerging synthesis in valuation studies is distinctively pragmatist, given that it is increasingly intertwined with elements of the actor-network theory tradition. Nevertheless, these two traditions converge on one point in particular: the neglect, or even the outright denial, of structural influences. The paper will argue that this has undermined the study of value, particularly value in finance, and contributed to a bifurcation of the field between these micro-focussed traditions and the macro orientations of the political economy of finance.

The paper has seven main sections, which move from more philosophical content to material that is more specific to the study of value. The first sets the scene by summarizing some key themes from pragmatist philosophy and the second provides an overview of relevant aspects of critical realism. The next three consider encounters between the two traditions, primarily in the philosophy of the social sciences. The final two sections bring us to the study of financial value, with section six discussing how pragmatism has been employed in the field and section seven discussing the consequences.

Elements of pragmatism

Pragmatism is famously plural; as long ago as 1908 Arthur Lovejoy claimed there were thirteen different versions (1908a, 1908b), though Lovejoy’s thirteen pragmatisms might be better understood as thirteen elements of the tradition, fitted together in various combinations by different pragmatists. The diversity in the views of even the founding figures – Charles Sanders Peirce, William James, John Dewey and George Herbert Mead – is widely acknowledged by pragmatists themselves (Bernstein 1992, 824). Nevertheless, there are some core themes and common (though not universal) features of pragmatist philosophies. The prominent contemporary philosophical pragmatist Richard Bernstein suggests five (1989). I will subdivide one and then add five more to give eleven in total. This is far from a comprehensive overview of pragmatism, but it should be enough to contextualize the paper.

(1)	Bernstein’s first theme, and perhaps the most fundamental to pragmatism’s sense of identity, is anti-foundationalism, beginning with Peirce’s critique of ‘the idea that knowledge rests upon fixed foundations, and that we possess a special faculty of insight or intuition by which we can know these foundations’ (Bernstein 1989, 7). Foundationalists argue that although many of our beliefs are justified by reference to other beliefs, ultimately there must be some ‘privileged set of statements that serve as the ultimate foundation of knowledge’ because they do not need to be justified in terms of prior belief (Poston n.d.). The classic example is René Descartes’ cogito: ‘I think therefore I am’. Having found every other basis for knowledge unable to withstand sceptical doubt, Descartes argued that this claim established an unchallengeable truth, and then sought to build others on top (Descartes 1975). Peirce’s critique positions pragmatism against Enlightenment philosophy’s search for a basis for absolute certainty, from Descartes onwards. Pragmatists question not only the possibility that philosophy could rest on certain foundations but also the need for it to do so (Bernstein 1992, 813–4). This remains a core element, in particular, of the more recent pragmatist work of Richard Rorty (e.g. 1980).
(2)	Pragmatism, however, usually seeks to avoid the sceptical or relativist response to the impossibility of absolute certainty, by arguing that we can still have worthwhile knowledge although we must always recognize that our knowledge claims might turn out to be false (see Holmwood 2011, 20, 25). This fallibilism is applied to all sorts of knowledge, including philosophical claims themselves (Bernstein 1989, 8–9).
(3)	Bernstein’s next theme may be divided into two. First, the most sociological of his themes, we have ‘the social character of the self’ (Bernstein 1989, 9 emphasis added) which we may oppose to atomistic notions of homo economicus or the closely related asocial individuals of rational choice theory. This theme is perhaps most closely associated with Mead, whose work on the socially shaped ‘Me’ is echoed in more recent work by thinkers like Pierre Bourdieu, but also balanced by his concept of the responsive ‘I’ (Mead 2015).
(4)	The second part of Bernstein’s third theme is the role attributed to ‘a critical community of inquirers’ as the arbiters of knowledge (1989, 9). We might connect the two parts through the idea of the social character of the knower: rather than individual reason or individual experience as the ultimate ground of knowledge, Peirce offers the shared agreement of the community (perhaps some ultimate shared agreement in the long term) as the measure of knowledge, and indeed of truth. Different pragmatist accounts of truth, however, will appear below.
(5)	Bernstein’s next theme is the conviction that our universe and our lives are marked by radical contingency (1989, 9) and thus unpredictability, which as we shall see feeds into pragmatist accounts of human action.
(6)	Finally, Bernstein argues that for pragmatists ‘there can be no escape from plurality – a plurality of traditions, perspectives, philosophical orientations’ (1989, 10). This paper’s account of pragmatism itself as plural clearly coheres with that theme. Combined with an awareness of fallibilism, this contributes to a commitment to open constructive dialogue with other points of view that is rooted in the effort to understand the opponent rather than merely dismissing their argument (1989, 15–18). As Bernstein himself observes, however, pluralism is not always so open; we must be wary, for example, of polemical pluralism, which invokes the concepts of pluralism and dialogue but only as weapons employed without ‘a genuine willingness to listen and learn from others’ (1989, 15).

Bernstein’s summary is strongly oriented to philosophical issues, and arguably leans towards the Peircean version of pragmatism, so it is useful to supplement it with some other themes that relate more to the social sciences and other leading figures in the tradition.

(7)	In the social sciences, perhaps the most characteristic feature of pragmatist interventions is a focus on human action. Action is seen as oriented to solving practical problems as they arise, through drawing creatively (when required) on established habits (Gross 2009, 366). Again there is a parallel to Bourdieu, indeed Bourdieu himself recognized this (Gross 2009, 367). There is a sense of people constantly adjusting to the radical contingency of events. Hans Joas, influenced by Dewey, rejects the focus on human goals and purposes in other traditions, suggesting that these are constantly updated in reaction to changing circumstances (Joas and Beckert 2001). Other pragmatists emphasize the social character of the influences on individual action (Gross 2010, 342). Taking this focus on action to the extreme, Andrew Abbott even argues ‘that actions, not actors, are the primitives of the social process’ (2007, 7), also suggesting a sub-theme – a tendency towards process ontology.
(8)	Sometimes, though not always, this focus on action is accompanied by a rejection of social structures as causally significant. Randall Collins suggests that Peirce, not being a sociologist and typically of American thinkers at the time ‘had little conception of social structure’ (1994, 252), implying that it was an omission rather than a deliberate exclusion. Mead, as we have seen, saw social forces as shaping individual action through shaping habits, though without talking of them as structures. Contemporary pragmatists rarely invoke social structures. At one extreme we find outright rejection of social structure, for example by Osmo Kivinen and Tero Piiroinen, who read James as a methodological individualist and reject ‘reifying modes of interaction into social entities in the way Bhaskar [the founder of critical realism] does’ (2004, 234). At the other we find ambivalence: Neil Gross welcomes ‘the very thinness of the [pragmatist] model at the meso- and macro-levels’ (2009, 368), but also insists that Pragmatism is not a form of methodological individualism; it does not require that mechanisms operating at the meso or macro-levels be explained exclusively in terms of the actions of the individuals involved … It does insist, however, that the potential contribution of individual action to the operation of mechanisms be taken into account (2009, 369).
(9)	Pragmatism is also somewhat ambivalent about realism (Holmwood 2011, 20). The early Peirce, in advocating the scientific method, argues that ‘There are real things, whose characters are entirely independent of our opinions about them’ (1877, 21). The later Peirce seems equally realist. Applying his pragmatist maxim that we can best understand concepts by tracing out the practical consequences of affirming them (1904, 56), he argues that ‘the real becomes that which is such as it is regardless of what you or I or any of our folks may think it to be’ (1904, 58). Still, he may not have understood the real in the same way as contemporary realists: Howard Mounce argues that Peirce was an objective idealist, who saw the world as ‘constituted by an order which is mental in character but which is quite independent of the human mind’ (Mounce 1997, 9). Perhaps contemporary sociological pragmatists are more influenced by Dewey, who argues that ‘the chief characteristic trait of the pragmatic notion of reality is precisely that no theory of Reality in general … is possible or needed’ (1917, 222) – arguably more a rejection of theorizing about reality (see the next theme) than a denial of reality itself.
(10)	Both Peirce’s pragmatist maxim and Dewey’s attitude to realism bring us to perhaps the most common understanding of pragmatism: that it is an anti-theoretical tradition, in a variety of ways. Meaning is just a matter of the practical implications of a belief or claim, practical knowledge is to be preferred over theoretical inquiry (Mounce 1997, 13), and thought should be judged purely by whether it provides an adequate cognitive basis for solving the problem at hand (Holmwood 2011; Little 2009). James himself calls pragmatism ‘anti-intellectualist’ (1907, 54) and Dewey calls for a philosophy that ‘ceases to be a device for dealing with the problems of philosophers and becomes a method, cultivated by philosophers, for dealing with the problems of men [sic]’ (1917, 230). Richard Rorty (despite making a broad range of claims about the nature of the world and the pursuit of knowledge) argues that pragmatists ‘do not require either a metaphysics or an epistemology’ (1985, 3). In the philosophy of the social sciences, Patrick Baert opposes pragmatism to attempts to develop theoretical frames of reference that apply across multiple settings (2005, 194). And yet even here there is room for ambiguity: despite rejecting a great deal of existing metaphysics, Peirce retained a life-long commitment to metaphysics as a project, which he sought to place on a scientific basis (Haack 2007).
(11)	Another contender for the best-known element of pragmatism is James’s take on knowledge or truth. Peirce argued that ‘The opinion which is fated to be ultimately agreed to by all who investigate, is what we mean by the truth’ (1878, 45). James, by contrast, took the maxim Peirce had applied to meaning, and applied it to truth instead: knowledge claims are to be judged true or false by tracing out the practical consequences of affirming them (Mounce 1997, 37). Famously, this led to James’s claim (in an argument justifying belief in religious truths) that ‘The true is the name of whatever proves itself to be good in the way of belief’ (1907, 76). Or, similarly, that ‘‘The true’ … is only the expedient in the way of our thinking … Expedient in almost any fashion; and expedient in the long run and on the whole of course’ (1907, 222). Formulations like these can seem to suggest that truth for pragmatists is whatever it is in our interests to believe (cf Mounce 1997, 37), although it is possible to read James to mean that beliefs are useful (and therefore true) when they work, and that they work when we can validate them against our experience (Porpora 2015, 205). Almost immediately, however, James adds that The “absolutely” true, meaning what no farther experience will ever alter, is that ideal vanishing-point towards which we imagine that all our temporary truths will some day converge … Meanwhile we have to live to-day by what truth we can get to-day, and be ready to-morrow to call it falsehood’ (1907, 222–3).

James’ absolutely true seems to correspond to Peirce’s truth, while James’ day to day truths seem to be more like our everyday concept of belief. Similar disagreements have continued in more recent pragmatism. Hilary Putnam, echoing Peirce, says that ‘to claim of any statement that it is true … is, roughly, to claim that it could be justified were epistemic conditions good enough’ (1990, vii). Rorty, on the other hand, advocates James’s sense of truth as ‘what it is good for us to believe’ (Rorty 1985, 3), and prefers Alfred Tarski and Donald Davidson’s ‘homely and shopworn sense of “true”’ to Putnam’s philosophical sense (Rorty 1980, 308–9). Overall, pragmatists have taken a variety of conflicting views on truth (Porpora 2015, 204–5).

Critical realism¹

Critical realism is a much more recent tradition. While it draws on earlier work by Rom Harré (notably Harré and Madden 1975), the first authoritative statement of critical realism appeared in Roy Bhaskar’s book A Realist Theory of Science (1975). Although there is also diversity within critical realism, Bhaskar remains the central figure in the tradition, and most if not all of its central themes originate from his work. Indeed, Bhaskar himself took up different positions in different phases, without discarding any of the earlier work, and much of the diversity within critical realism is a product of different critical realists working with themes from his different phases: original critical realism, dialectical critical realism, and his work on metareality. Critical realism has been adopted primarily by social scientists, who tend to work with ideas from original critical realism, which will be the focus here. Again, let me begin with his general philosophy and then move towards the implications for the social sciences.

Bhaskar’s first book on the one hand criticizes empiricist positivism for collapsing causality into our experience of its effects, and on the other attacks philosophical tendencies to collapse the world into our knowledge of it. As an alternative to both he develops a realist ontology of causal powers.

The critique of positivism takes the form of what Bhaskar calls a transcendental argument from the practices of laboratory science.² He takes Hume as the archetype of positivism, and takes issue with his argument that causality is nothing more than the constant conjunction of one type of event with another. Bhaskar argues that science (in particular, laboratory science) would not make sense if this was all that causality was. When scientists design laboratory experiments, they are trying to produce empirical regularities by creating conditions in which factors that would normally prevent those regularities are suppressed. The purpose of doing so, however, is to make claims about the causal implications of their findings for events outside the lab in which those other factors may also be operating (Bhaskar 1975, 13). This behaviour is only intelligible if scientists have good reason to believe that the causal forces observed in the laboratory continue to operate outside it, even when they do not produce the exceptionless regularities that occur there. In other words, those causal forces operate as tendencies generated by objects of the type they are observing, rather than necessarily producing the same outcomes in all cases (Bhaskar 1975, 14). To put it in Bhaskar’s terms,

The real basis of causal laws are provided by the generative mechanisms of nature. Such generative mechanisms are, it is argued, nothing other than the ways of acting of things. And causal laws must be analysed as their tendencies. Tendencies may be regarded as powers or liabilities of a thing which may be exercised without being manifest in any particular outcome (1975, 14).

Because every event is multiply determined by many interacting causal powers, on any given occasion any given causal power may be frustrated by the effects of others. The consequence is that causality operates without producing exceptionless regularities.

Bhaskar identifies causality with the powers of things, powers that are produced by (or are synonymous with) generative mechanisms that depend on the structure of the kind of thing concerned.³ He calls these causal powers emergent, in the sense that they are powers of the thing that are not possessed by its components – a sense of emergence that has also been invoked in the pragmatist tradition, notably by Mead (2015, 198, 329) and the later Dewey (Pratten 2019). This is not emergence in the sense of a mysterious unexplainable appearance from nowhere, but rather emergence as the consequence of organization: these powers come about as a consequence of the relations and interactions between the components that only occur when they are organized into the type of higher level structure that has the power concerned (Elder-Vass 2010, 16–18).

On the other hand, Bhaskar is also critical of many earlier anti-positivist philosophers, whom he accuses of committing the epistemic fallacy, ‘the view that statements about being can be reduced to or analysed in terms of statements about knowledge; i.e. that ontological questions can always be transposed into epistemological terms’ (1975, 36). This critique is directed against the denial of reference and representation that is common in postmodern and radical constructionist thinking but which is also implicit in some strands of foundationalist philosophy, under the influence of Immanuel Kant, which ‘exchange the structure of the world for the structure of the mind’ (Goodman 1978, x). For Kant, we can only know the empirical or phenomenal world, the world as we experience it, and this knowledge is shaped and limited by our preconceptions. By contrast, we cannot have knowledge of the external or noumenal world that we think we are experiencing. For Bhaskar, on the other hand, we can have knowledge of parts of the world that exist independently of our beliefs about them, although that knowledge is necessarily fallible. As I have put it elsewhere:

Unlike Kantianism, this is a non-dualistic understanding of our relation to the world: we are part of the same world that we have knowledge about, and the distinction between our knowledge and the world beyond us is internal to the world; both me (including my beliefs) and the things I can see, for example, are different pieces of the same world (Elder-Vass 2021).

Bhaskar distinguishes between the intransitive dimension and the transitive dimension of knowledge (1975, 21–4). He uses the concept of the intransitive dimension to refer to those objects of a person’s knowledge that exist independently of her beliefs about them. The concept of the transitive dimension, by contrast, refers to the beliefs, including concepts, theories, and the like, that the person has about the objects in the intransitive dimension. These beliefs develop through an iterative process of interaction between the transitive and the intransitive dimensions, in which a person’s experience of interacting with the intransitive can lead her to revise her transitive beliefs. The transitive/intransitive distinction, however, does not relate to two different types of objects; rather, it is a relational distinction that refers to the role an object plays in any given knowing relationship: your beliefs, for example, are part of your transitive domain but part of my intransitive domain. This becomes important when we turn to the social realm, as Bhaskar does in his second book The Possibility of Naturalism (1979).

Here, Bhaskar extends the model of emergent causal powers to social structure and to human agency. The powers of social structures are seen as depending on the activities of people and on their understandings of what they are doing (1979, 38). Nevertheless, they are also emergent in the sense that they are not powers of the individuals, but depend instead on how those people come to be related to each other in the structures concerned as a result of their understandings. One consequence is that as the beliefs and activities of individuals change, so do the structures, in what Bhaskar calls the Transformational Model of Social Activity (1979, 33–6). Margaret Archer’s theory of the morphogenetic cycle, developed independently, is a close analogue of the TMSA (Archer 1995, 154–9). This approach to social structure should help to make clear why Bhaskar’s transitive/intransitive distinction is more plausible than the naïve version of the realist claim: that the world is entirely independent of human beliefs about it. The naïve version fails when we recognize that the social world is not independent of our beliefs about it – that social structures like organizations, for example, only work because people recognize and accept their existence and alter their behaviour in response to these and related beliefs. The intransitive/transitive distinction, however, allows the organization theorist to distinguish between those beliefs of other people that sustain the organization concerned (which are independent of the theorist’s beliefs and thus fall into her intransitive dimension) and the theorist’s own theories (which fall into her transitive dimension). This makes it possible to recognize both that social structures depend on the conceptions that people have, but also that their existence is independent of any given scholar’s beliefs about them.

The central tenets of critical realism can be summed up in the three principles of ontological realism, epistemic (or epistemological) relativism, and judgemental rationality (Bhaskar 1986, 24, 2016, 25–6). These are closely interconnected. Given a form of ontological realism that recognizes (a) a real world that is largely independent of any one observer’s beliefs about it, and (b) that this world can influence our beliefs about it, then it follows that further experience can always undermine any given belief. Bhaskar’s ontological realism therefore entails fallibilism about belief. He also sees belief as open to influence by the social context of the believer (1979, 57), and thus advocates a philosophy that ‘would unashamedly acknowledge as a corollary of its realism, the historicity, relativity and essential transformability of all our knowledge’ (1989, 155). His realism thus also implies epistemic relativism: knowledge claims are necessarily uncertain and influenced by social factors. On the other hand, the same realism, in particular the same recognition that our beliefs can be influenced by our interaction with its objects and their effect on us, leads to a rejection of extreme relativist claims that see beliefs as completely determined by social forces. Such relativisms eliminate any objective basis for making judgements, but Bhaskar’s form of epistemic relativism leaves space for what he calls judgemental rationality. This is the claim that our capacity to interact with the intransitive dimension entails that we can have the capacity to make reasonable, justified, though necessarily fallible, judgements about the objects that populate it (Proctor 1998, 361). Our interactions with the intransitive world beyond us provide evidence that may be influenced by our transitive preconceptions but is not fully determined by those preconceptions and therefore provides us with the means to make reasonable judgements that can improve our understanding. One might argue that classical pragmatists also practice a form of judgemental rationality, endorsing those beliefs that it is judgementally rational to hold because they work as a basis for action. Realism adds something here: that when our knowledge works for us, it is often because it does represent external reality successfully. At another level, though, the traditions are similar: both see knowledge as both fallible yet also potentially reliable.

In the social sciences critical realism has made an important contribution to debates on the question of structure and agency. As well as social structures, Bhaskar also sees human individuals as having causal capacities to influence social events. Similarly, Archer sees people as having ‘personal emergent properties’ that can interact with structural and cultural emergent properties to shape social events (1995). In other words, the larger ontological framework in which events are seen as multiply determined also applies to the social world and allows us to see social events as produced by interactions between both structural and agential powers (and also the powers of material objects) (Elder-Vass 2010). Critical realism also provides extended accounts of human agency, notably several books from Archer on the nature of human reflexivity, drawing on, amongst others, Mead’s notion of the internal conversation (Archer 2000, 2003). Other critical realists have developed accounts of agency that combine Archer’s orientation to conscious deliberation with more disposition-oriented accounts of human action such as Bourdieu’s, with its echoes of James’ focus on habit and Mead’s notion of the socially-shaped ‘Me’ (Elder-Vass 2010, 87–114; Sayer 2005, 22–51). Both critical realism and pragmatism can therefore be seen as ‘humanism[s] strongly committed to consciously centred agents’ (Porpora 2015, 196).

Encounter one: Rorty vs. realism

The analysis so far suggests some non-trivial similarities between pragmatism and critical realism, including a shared commitment to anti-foundationalism and fallibilism, a belief in the contingency of events, and the importance of socially-influenced individual agency in their approaches to social theory. There are also partial but important overlaps, such as the commitment of Peirce to realism. Nevertheless, pragmatist theorists have tended to insist on a radical opposition between the two. Although there are certainly important differences, this section and the next two will argue that pragmatist critics of realism have often mis-identified those differences. This section begins with Rorty’s critique of realism (in general, rather than critical realism specifically) in his paper ‘Objectivity or Solidarity’ (1985). The paper proposes a binary division of philosophers into pragmatists and realists, a bold rhetorical move that inevitably misrepresents much of the field (Proctor 1998, 365). On the realist side, he argues that

Those who wish to ground solidarity in objectivity – call them “realists” – have to construe truth as correspondence to reality … They also must argue that there are procedures of justification of belief which are natural and not merely local … On their view, the various procedures which are thought of as providing rational justification by one or another culture may or may not really be rational. For to be truly rational, procedures of justification must lead to the truth, to correspondence to reality, to the intrinsic nature of things (Rorty 1985, 3)

Many realists, including most critical realists but not Bhaskar, do advocate a correspondence theory of truth (e.g. Elder-Vass 2012; Groff 2000; Porpora 2015).⁴ There are many different varieties of correspondence theory (David 2016), but the clearest realist version states simply that a claim is true when it accurately represents the state of affairs it describes (Alston 1996; Groff 2000, 426). Crucially, this is a theory about the meaning of the concept of truth; it is not a criterion against which we might judge any particular truth claim. As Ruth Groff puts it, this means that ‘questions about truth must be distinguished from questions about what Alston terms “epistemology.” It is the fundamental error of epistemic theories of truth generally … that they collapse this crucial distinction’ (Groff 2000, 426). This is just what Rorty does in the quote above: he moves from ascribing the correspondence theory to realists to suggesting that this entails a commitment to procedures of justification that lead to the truth. But critical realists would argue that procedures of justification produce knowledge, not truth, and that knowledge is not the same thing as truth: knowledge is a claim held for a good reason that may or may not be true. Even our most rational procedures of justification therefore do not necessarily lead to the truth, as realists understand it.

By contrast with correspondence theories, epistemic theories of truth do provide criteria for judging truth claims. For epistemic theories, there is ‘an epistemic protocol or methodology that produces the epistemic state of certainty’ although different epistemic theories invoke different protocols or methodologies (Porpora 2015, 78). For these theories, in other words, a knowledge claim is true if it meets a specific epistemic criterion, or if it is produced by some specific procedure of justification. James’ view that truth is what works, for example, wraps together a definition of truth with a criterion of which claims are true. For such theories, knowledge cannot be separated from truth, so what is true today could turn out to be false tomorrow, whereas for a realist if we find that a claim is false tomorrow that implies that it was also false before that, whatever we thought at the time. Rorty appears to hold an epistemic theory of truth, which perhaps explains why Rorty assumes that realists also conflate knowledge and truth. The net result is that Rorty implies that realists cannot be fallibilists, which as we have seen is clearly not true of critical realists.

In lay discourse the concept of truth has a rhetorical force which arises from the conflation of two different understandings of what it means. On the one hand, to assert that a statement is true derives its force from invoking a correspondence-type concept of truth as an ideal relation between a claim and its object, in which the truth of a belief is independent of what we may think about it. On the other hand, no-one who makes such a claim can know with absolute certainty whether the claim they are making meets that standard, though they may have a great deal of confidence in it. Hence, they use the assertion that a claim is true as a kind of rhetorical device, what Rorty calls ‘merely an expression of commendation’ (1985, 4). Realist philosophers committed to the correspondence theory split these two elements apart and when they are able to resist the temptation to slip back into everyday usage they reserve the word true for truth as an ideal, while using knowledge instead for those particular beliefs we think we have good reason to endorse.

This difference lies at the root of some misunderstandings between pragmatists and realists. Realists are often bemused, for example, by James’s assertion that beliefs that work are thereby true because they assume he means that these beliefs are true-as-ideal, but he is employing an epistemic theory that does not recognize truth as independent of knowledge. James then confuses things further by employing the notion of ‘absolute truth’ for what correspondence theorists would generally call ‘truth’. James’s ambivalence reflects a tension between epistemic theories of truth and fallibilism. Strict epistemic theories of truth cannot distinguish between knowledge and truth, whereas fallibilism does: what the knower believes she knows may turn out to be false. When a fallibilist calls today’s knowledge true, she is using the term as a commendation, and understands that this knowledge may fail to live up to the standard of truth-as-ideal. Fallibilism entails, to put it more directly, that some knowledge is false, whereas there is no room for this possibility in strict epistemic theories. One must therefore doubt whether all versions of pragmatism fit entirely on the pragmatist side of Rorty’s binary division of philosophy.⁵

Encounter two: Bhaskar vs. Rorty

The most substantial critical realist response to pragmatism is Bhaskar’s book Philosophy and the Idea of Freedom (1991) and the overlapping paper ‘Rorty, Realism and the Idea of Freedom’ (available in Bhaskar 1989; and Malachowski 1990). It is perhaps an indicator of the significance of pragmatism for Bhaskar that this is by far his most extensive engagement with any contemporary philosopher. He agrees with much of Rorty’s ‘eloquent critique of the epistemological problematic, from which contemporary philosophy is gradually emerging’ (Bhaskar 1991, vii) – in other words, with his attack on epistemological arguments designed to provide certain foundations for knowledge (Bhaskar 1991, 31). Rorty sees the epistemological problematic as embedded in the vocabulary of the discipline of philosophy (Rorty 1980, xi) and notably suggests that we have ‘Kantian philosophy built into the fabric of our intellectual life’ (Rorty 1978, 314). Rorty’s work can be seen as an attempt to escape this fabric but Bhaskar claims that Rorty ‘has provided us with only a partial critique of a problem-field, to which he remains in crucial respects captive’ (Bhaskar 1991, vii).

Bhaskar argues that the epistemological problematic is marked by a pair of fallacies. On the one hand, we have what he calls the ontic fallacy, which corresponds roughly to Rorty’s critique of ‘the effective ontologization or naturalization of knowledge, the reduction of knowledge to, or its determination by, being in what may best be regarded as a species of compulsive belief-formation’ (Bhaskar 1991, 33). It is untenable, in other words, to regard knowledge as being directly and completely imposed on us by the objects of our experience. Bhaskar argues, though, that Rorty fails to see that the epistemological problematic also commits the epistemic fallacy: the substitution of (claims about) knowledge for (claims about) the objects of knowledge. Being is replaced, in this problematic, by our beliefs about being. Kant, for example (in the usual interpretation of his work) substitutes the phenomenal world, and thus our experiences of the world, for the noumenal world, or the world as it exists independently of us, in his discussions of human knowledge. The epistemic fallacy, according to Bhaskar, paves the way for

the ontologization (eternalization and divinization) of knowledge in a subject-object identity or correspondence theory. Such a theory effectively welds together the transitive or social-epistemic and intransitive or ontic dimensions of science … On it knowledge is naturalized and being epistemologized (Bhaskar 1991, 32) [see note 4 on Bhaskar’s understanding of correspondence theory]

Bhaskar’s point is that the tradition Rorty is criticizing can only see belief-formation as compulsive (knowledge as determined directly by being) because it has already re-cast being (or, in accounts like Kant’s, the only aspects of being that we have access to) as an aspect of experience. The gap between knowledge and its object that is central to realism and fallibilism had already been eliminated by the epistemic fallacy, and so when Rorty criticizes the ontic fallacy he addresses only part of the larger problem: that the epistemological problematic falsely conflates the ontic and the epistemic by refusing to accept that there is anything beyond our knowledge and experience that it is knowledge and experience of.

The consequence, Bhaskar argues, is that Rorty over-reacts. In rejecting compulsive belief-formation, he argues instead for ‘effectively unconstrained belief-formation’ (Bhaskar 1991, 32). Perhaps it is unfair to suggest that Rorty sees belief-formation as entirely unconstrained, since Rorty recognizes the beliefs of the community as a constraint, but Bhaskar’s core point is surely correct: Rorty fails to recognize that although the objects of our knowledge do not fully determine our beliefs about them, they do influence our beliefs about them. Rorty therefore ends up advocating a kind of relativism about knowledge that eliminates the realist possibility that we can make judgementally rational claims about the world based on our experience of it (although influenced by our cognitive presuppositions). Instead, our knowledge loses all moorings to what it is about and instead is rooted only in the beliefs of our community. While both Bhaskar and Rorty advocate epistemological relativism, they are very different kinds of epistemological relativism.

Encounters in social science

Turning to encounters between the two traditions in the social sciences, this section will discuss three themes in pragmatist responses to critical realism – the claims that it is foundationalist, builds unnecessary complex ontological frameworks, and is dualist – and then close by considering more conciliatory approaches.

Baert, for example, has accused Bhaskar of being foundationalist, on the grounds that he seeks to uncover ‘the logic of inquiry that all successful scientific activities purportedly have in common’ (Baert 2005, 192). No doubt if Bhaskar’s project was to uncover a universal logic of inquiry in order to assert that it provides a basis for infallible truth claims, that would be foundationalist. But that is not his project at all. He considers the logic of inquiry of laboratory science in particular, in order to construct an argument about what the world must be like for that logic to work, and at no point does he suggest that this logic is universal, nor that it leads to infallible knowledge. Indeed he and other critical realists are pluralists about scientific methods, and Bhaskar is ‘unblushingly fallibilist and historicist about science’ (Bhaskar 1991, 30). Bhaskar’s project is an ontological one rather than an epistemological one.

Taking a second bite at the cherry, Baert adopts an extended definition of foundationalism from Steven Seidman and Jeffrey Alexander, who argue that in the social sciences foundationalism seeks ‘to uncover unchanging foundations of an all-embracing framework or science of the social’ (Baert 2005, 194). Baert identifies critical realism as the purest case of this supposed crime: ‘the belief that theory provides an objective base for a powerful frame of reference, one that is applicable to different, if not all, settings, cultures and times’ (Baert 2005, 194). In some respects, Baert is on firmer ground here. Critical realists do indeed seek to develop a frame of reference that is applicable to many different settings, but it is an open and flexible frame that recognizes difference and the difference it makes. It is hard to see, however, how developing theoretical frameworks is in itself foundational in anything like the original sense criticized by the classical pragmatists. Critical realism’s frameworks provide tools for thinking through how causality might be working in particular cases, but they do not mandate particular answers nor do they provide reasons for taking the answers that any particular research finds as certain. Frameworks like this are potentially useful, and it is not only thinkers in realist traditions that develop them. Thinkers in traditions that many pragmatists value also develop general frameworks and find them useful – arguments about the role of language in constituting the social, for example, or indeed Gross’s explicitly pragmatist A – P – H – R framework for thinking about social action (2009).

It may be that some of these critiques are (mis)informed by the problem that Bhaskar ascribes to Rorty: a failure to shake off the epistemological problematic. Baert’s identification of Bhaskar’s argument from scientific enquiry as foundationalist, for example, perhaps reflects an assumption that this must be an epistemological argument; that claims about the real are necessarily also claims about criteria of the true. We find a similar example in John Levi Martin’s debate with Christian Smith. Discussing his own reasons for abandoning a version of realism, Martin identifies realism with the belief that ‘our qualitative perceptions of other persons are true because we perceive the real world’ (2013a, 16). We may connect this with his more general statement that ‘Realism says not only that there is a real world but that our knowledge is knowledge of this world, and not of a mediating world constructed by our own subjective apparatuses’ (2013b, 6). This second statement is perfectly reasonable, unless one believes that there is no separation between knowledge and truth, in which case it would imply the sort of foundationalist perspective stated more explicitly in the first statement. As I have made clear above, this is not what critical realists believe, and Smith responds equally directly: ‘JLM [John Levi Martin] mind-bogglingly says I conceptually collapse truth and reality, when my repeated argument on the matter to the contrary could not be clearer’ (2013a, 18).

Stephen Kemp, drawing on the work of Popper, offers a more subtle critique of critical realism’s attitude towards knowledge claims (Kemp 2017). Kemp argues for a ‘transformational fallibilism’ in which ‘the investigator should be willing to subject any element to questioning and ultimately revision or reconstruction’ (2017, 193). From this perspective, the critical realist ‘view that practically successful understandings reveal something about the structure of the world involves an overconfidence in these understandings’ (2017, 193). Transformational fallibilism, by contrast,

does not treat the idea that a claim may be untrue as simply a philosophical nicety that an inquirer can acknowledge whilst retaining great confidence in the truth of certain beliefs for the time being. Rather, it argues that in order to develop knowledge, even apparently obvious and hard-to-question beliefs should have their limitations explored (2017, 193).

Kemp picks out two cases where he thinks that critical realists fail to do this. First, he challenges Andrew Sayer’s argument that when basic empirical beliefs work, it is because they reflect the nature of the objects of those beliefs (Kemp 2017, 201). For critical realists, Sayer’s argument seems perfectly reasonable, particularly given that he is thoroughly fallibilist about the beliefs concerned. Second, Kemp picks out my own argument that some epistemological standards may ‘be well-suited to the job of generating veritistically reliable, or approximately true, knowledge’ (Elder-Vass 2012, 232) and that we should favour those standards that are more conducive to doing so. Kemp seizes on what he sees as the implication that we must be able to tell which beliefs are true, but this reading requires some quite uncharitable contortions given that in the preceding pages I make clear that we cannot do so (Kemp 2017, 203–4). Overall, Kemp’s argument reflects a sense among pragmatists that if critical realism is fallibilist, there is some sense in which it is not falliblist enough. But he has to read his critical realist sources against the grain to dig out nuggets that might support such a view. In the end, it is hard to see why a pragmatist would dispute critical realism’s view that knowledge claims are (a) potentially fallible; but ALSO (b) useful to work with unless and until we have good reason to actively doubt them.

A second theme of pragmatist criticism is that critical realism creates unnecessarily complex ontological frameworks, or ‘heavy ontological furniture’ (Kivinen and Piiroinen 2004, 231). Scepticism about ontological claims is common amongst sociological pragmatists, and connects to a belief that they can manage perfectly well without them. A typical response is Gross’s: ‘But what does this mean for explaining stuff in society—you know, the thing sociologists are supposed to do? Beats me’ (2017, 300).

We can ask Gross’s question at two different levels, which are sometimes conflated: as a question about ontological furniture in general, and as a question about specific ontological claims (which is the context of Gross’s quote). Bhaskar replied to the first question by citing Bachelard: ‘all philosophy, explicitly or tacitly, honestly or surreptitiously … deposits, projects or presupposes a reality’ (Bachelard 1953, 411; quoted in Bhaskar 1989, 13). We can extend the point to social science: every claim makes some presuppositions about what the world is like and we all therefore have at least an implicit ontology. Because your ontology affects the kinds of claims you make about the world, there is merit in being explicit about your ontological assumptions and examining whether they are consistent and plausible. Critical realists, for example, would want to question the ontology behind Martin’s claim that ‘reality is more or less the outcome of human agreement’ (2013a, 17). Pragmatists have ontology too; critical realists surface theirs and aim to make it consistent and plausible. When pragmatists do not, a minimalist approach to furnishing is a weakness rather than a strength.

This does not, of course, constitute an answer to the second version of Gross’s question: what the practical implications are of any particular ontological claim and whether it is useful. There is no general answer to such questions, other than ‘it depends’. Any particular ontological claim, whether made under the banner of pragmatism, critical realism, or any other tradition, must be examined on its own merits and even critical realists might ultimately want to judge it on whether or not it paid its way in terms of useful consequences.

Kivinen and Piiroinen organize several criticisms of critical realism around my third theme: the claim that it is dualist. They argue ‘that behind the whole idea of ontology … there is an underlying Cartesian subject-object dualism between the knowing subject and the world of objects to be known’ (2004, 231). They too seem to remain in thrall to the epistemological problematic, advocating a position inspired by Dewey and Rorty which ‘aims at overcoming the subject-object dualism, thereby managing without unnecessary ontological accessories’ (2004, 232). But they overcome the subject-object dualism by collapsing being into knowledge, in an argument that has already been criticized above. This is just as much an ontological position as the one they reject. Equally importantly, this and their further arguments rest on a failure to understand the nature of dualism. Dualisms entail that the world falls into two ontologically separate domains. By contrast, critical realists argue that people are part of the same world as the events and objects they observe and consider: knowledge is a relation between different parts of the same world, not between fundamentally separate subjects and objects.

Within that world, of course, there are different kinds of things, and we cannot make sense of the world without recognizing this. Kivinen and Piiroinen, however, accuse critical realism of ‘Cartesian dualisms’ because it ‘tends to draw binary distinctions between, say, nature and society, body and mind, and agency and structure’ (2004, 234–5). Again, they are off the mark. For critical realists, for example, the social world is part of the natural world, although it is a distinctive part and the nature of the distinction is instructive; the mind like the body is an aspect of the human individual; and agency and structure are both deeply interdependent and non-exhaustive of our ontology. We make distinctions, but distinctions are not necessarily dualisms, even when they are binary distinctions. And that ‘we’, of course, includes critical realists, pragmatists, and everyone else too: none of us can think at all without relying on distinctions.

Perhaps one source of confusion is that Archer describes her account of the relation between structure and agency as analytical dualism, but the word analytical is extremely significant here: it means that for the purpose of making sense of the relationship between structure and agency it is analytically useful to think of the contribution of each separately (Archer 1995, 165–194). Social structure, in Archer’s account, is largely the outcome of past agential interactions with previous iterations of structure, which is reproduced or transformed by those interactions. Nor is her ontology exhausted by structure and agency – we also have culture and material things. So the second respect in which her dualism is analytical is that it represents a moment in which we focus on the interaction of structure and agency while abstracting from the fact, of which Archer is fully aware, that in real social events culture and material things also play a part. To be sure, she thinks of structure and agency as distinct, but also as deeply interdependent. In other words, this is a strategy for picking out one relationship to focus on in order to clarify how it works, not a strategy for dividing the world into two binary orders of being (Archer 1995).

Most pragmatists would probably not be so rash as to accuse critical realism of Cartesian dualism on the grounds that it distinguishes between structure and agency, but there is certainly an undercurrent of resistance to realist accounts of social structure. Kivinen and Piiroinen, for example, are ontological individualists, rejecting Bhaskar’s talk of social structure as reification (2004, 234–5). As we saw above, this (less than universal) pragmatist attitude to talk of social structure predates critical realism, but it is certainly also part of pragmatist responses to critical realism.

Critical realists have perhaps tended to be more conciliatory towards pragmatism. Smith, for example, although he responded robustly to Martin’s criticisms of his own work, reviewed Martin’s work in positively glowing terms, suggesting that it was so compatible with critical realism that ‘Martin should simply go ahead and declare himself a card-carrying critical realist’ (Smith 2013b, 15). They have also sometimes acknowledged influence from pragmatist work, as Archer does with Mead. Doug Porpora is more critical, particularly of what he calls ‘the more subjectivist strain of pragmatism’ (2015, 74), although he also criticizes the sparseness of pragmatist social ontology and, above all, its neglect of social structures (2015, 58).

Occasionally pragmatists are also more conciliatory. Omar Lizardo, for example, tells us that he agrees with most of critical realism, although he sees its relationship with Bhaskar as problematic and criticizes critical realists in sociology for their rhetorical approach (2013). From outside both traditions, James Proctor has also argued that despite their differences, critical realism and pragmatism have a great deal in common (1998). He suggests that part of the issue is a difference in ‘temperaments and inclinations’: pragmatists prefer to work with lower-order claims that they find ‘more immediately useful’ while critical realists ‘place much greater value on correct conceptualization of problems as a necessary first step in solving, them, leading them to seek higher-order, structural truths to help explain the empirical situation of interest’ (1998, 368).

While Proctor’s argument is strong, I would summarize these debates a little differently. Pragmatism itself appears to be divided on some of the core points that are also sometimes seen as fault lines between pragmatism and realism. I have not yet found a wholly satisfactory terminology to describe that internal division, but on one side we have pragmatists who are strongly influenced by Rorty and on the other pragmatists who are not. The Rortyans persistently attack critical realism but in terms that suggest they have failed to cross the paradigm divide that separates irrealists from realists and thus to understand the positions they are criticizing. As a consequence, many of their criticisms, and particularly those that ascribe foundationalist views about knowledge to critical realists, are wrong. It is ironic when those pragmatists then lambast critical realists for a lack of dialogical engagement (e.g. Baert 2005, 195). On the other hand, the non-Rortyans fit Proctor’s description much better. Whether or not they recognize that critical realism is also a fallibilist, anti-foundational philosophy, their primary disagreement is based on an accurate understanding of the difference between the two traditions: they entertain serious doubts about the value of the kinds of ontological frameworks developed by critical realists, and often also doubt the value of concepts of social structure in developing sociological explanations.

Pragmatism in the study of value

While abstract discussions can help us make sense of the distinctions and indeed overlaps between theoretical approaches, ultimately the key question for social scientists is what difference the adoption of any given theoretical framework makes to the analysis or explanation of social phenomena. Both pragmatism and critical realism have been applied in many different fields and disciplines, and it would be impractical to survey all of these in a single paper. No doubt there are significant variations between the ways in which frameworks are applied to different fields, but we can at least begin to illustrate the issues arising by considering one of them. This section therefore turns to the role that pragmatism has played in studies of financial value, focusing on work by three of the most prominent advocates of pragmatism in the area, Fabian Muniesa, Emily Barman and to a lesser extent Jens Beckert. A great deal of excellent work has been done on financial value, including theirs, but I will nevertheless raise some concerns here about the overall impact of pragmatism in this area of study, which will be developed further in the following section.

Let me preface the discussion, however, with a comment on the area of study. Influenced by the focus of my own recent work on the topic, I began this project thinking of the study of financial value as a field, but I have been reminded that different scholars experience the segmentation of academic study in different ways. The notion of an academic field has taken on a life of its own: it no longer (if it ever did) refers simply to the work that has been done around any given topic of interest, but rather implies an element of institutionalization, whether simply discursive or buttressed with organizational structures, and the construction of a distinctive identity. The study of financial value sits at the conjunction of two such fields: the fairly recent but relatively well-established field of social studies of finance and the nascent field of valuation studies. I say this not simply to locate the area in the academic landscape but also because it appears that the assertion of a pragmatic perspective by scholars studying value functions as part of a process of fieldisation of valuation studies: an attempt to assert a particular identify for the field.

Muniesa is arguably the most assertive advocate of pragmatism in the study of value, and his paper ‘A flank movement in the understanding of valuation’ is his flagship statement of the case (2011). Muniesa makes it very clear which elements of pragmatism he is drawing on and how they influence his argument. He begins by criticizing the literature as divided into a tradition that sees ‘value as something that something has by virtue of how people consider it (that is, it is socially constructed, a convention, a social representation, a projection …)’ and another that sees it as something ‘that the thing may have as a result of its own condition (what it costs, how it is made, with what kind of labour, money and materials, what it is worth in relation to objective standards and fundamental metrics)’ – what we might call subjective and objective theories of value (2011, 24). Muniesa suggests a way forward inspired by Dewey: ‘Dewey’s “flank movement” consists in a shift in subject matter from value (or values) to valuation, considered explicitly as an action’ (2011, 25). For Dewey, values as such do not exist but rather are references to a quality that things may have by virtue of human action: ‘things which exist independently of being valued … are valuable when they are the objects of certain human activities’ (Dewey 1939, 4). Muniesa therefore argues that ‘Defending a pragmatist attitude in the study of value requires replacing the very notion of value with the notion of valuation’ (2011, 24). Although Dewey was writing a century ago, for Muniesa his message has not been adequately heard, and the study of value remains divided by the ‘“realist-idealist” syndrome’ (2011, 27). He calls for a shift of attention ‘to valuation as an action … in the sense of a process, a form of mediation … value is definitely not something that something just has’ (2011, 32). He and his colleagues also advocate a similar approach to economic capital: ‘we do not principally examine capital, capitalism or the capitalist, but operations of capitalization’ and capitalization itself is ‘considered as a particular form of valuation’ (Muniesa, Doganova, and Ortiz 2017, loc 185)

Muniesa’s argument has been influential. Most strikingly, both the field and its flagship journal have been named valuation studies, and one prominent edited collection invokes Dewey’s notion of valuation as an activity as an organizing theme of the book (Kornberger et al. 2015, 1). One might argue that the argument is taken too far: if valuation is an activity, what does it produce, if not value? And doesn’t value still remain in Dewey’s formulations as the outcome of valuation – not value as a thing, but value as a quality of things? Aren’t valuation and value mutually necessary concepts? Don’t even Marxists with their attachment to objective value see it as the outcome of a process? And doesn’t valuation as a process involve both subjective and objective elements, and indeed social structures? Offering a critical realist synthesis and drawing on the economics of conventions, I have argued, for example, that value claims are socially authorized beliefs about what something is worth that also depend on objective features of the thing being valued (Elder-Vass 2019). From this perspective, valuation activities are part of a process that necessarily involves the subjective, objective and social elements that Muniesa is asking us to move away from.

Muniesa’s argument is inspired by the pragmatist focus on action as well as by Dewey’s more specific arguments about valuation, and I do not question that an adequate account of valuation must take account of human action. Substituting valuation entirely for value, however, misses out important parts of the picture, and a focus on action can be equally problematic if it leads to the neglect of other elements of the larger complex of individual, social and material forces that is involved in valuation.

In an earlier paper, Muniesa seeks to enrol other scholars of value under the pragmatist flag. He claims, for example, that ‘a pragmatist viewpoint can be identified in a large number of contributions’ to the French conventions theory tradition (Muniesa 2007, 391). The sociological side of this tradition is often known as pragmatic sociology, but as a number of commentators have pointed out, pragmatic sociology is not a version of pragmatism (e.g. Hennion 2016, 301). Laurent Thévenot has acknowledged affinities with pragmatism but only read the American pragmatists after the most influential work in the conventions tradition had already been written (Blokker and Brighenti 2011, 397–8). The affinities with pragmatism arise from elements that they learned from other traditions, and Thévenot also affirms explicit influence from the distinctly un-pragmatist Durkheimian tradition.

Amongst further claims for pragmatist influence, Muniesa argues that ‘Debates on the performativity of economics prompted by Callon (1998) have also drawn on pragmatist sources such as John L. Austin’ (Muniesa 2007, 391). Callon (with whom Muniesa has published) is at least as much an actor-network theorist as a pragmatist, and although there are also affinities between ANT and pragmatism, again the influence came from elsewhere and the partial resemblances were noticed only later (Hennion 2016, 37–8). Muniesa’s ascription of pragmatist influence here is directed towards Austin, but as Ian Hacking explains, Austin ‘was definitely not a pragmatist’ in the philosophical sense (Hacking 2009, 37). One must be cautious of ascribing theoretical labels to thinkers based on overlaps and similarities (though pragmatists are not alone in this, as Smith’s response to Martin above suggests). Hacking, notably, found himself labelled a pragmatist and responded that ‘Some of the theses favoured by pragmatists … arise from and seem natural in many other contemporary perspectives … But those selfsame perspectives do not owe much to pragmatism, and do not define one as a pragmatist unless one so chooses’ (2009, 47–8). Muniesa’s claims for earlier pragmatist influence in the field are not entirely convincing when we trace back the associations, but the recent influence is more concrete, and Muniesa has played a significant role in it.

Barman stakes a claim for pragmatist influence in an empirical paper about the development of the market for impact investing, and backs the claim up by listing authors who have made significant contributions to valuation studies from a pragmatist perspective (2015, 12). Invoking Dewey and Muniesa, she identifies the approach as being focused on ‘the negotiation, construction, and objectification of value’ (2015, 12). Contrasting the approach with mainstream economics, she suggests that within it ‘value is understood as a social construction … generated out of the valuation practices, conventions and devices present in the situation of study’ (2015, 10). Overall, she tells us, ‘The pragmatist approach to value provides a particular framework for the empirical investigation of the question’ (2015, 10). Her framework, however, also draws in work from other traditions, notably the discussion of interacting orders of value in French pragmatic sociology (2015, 13) and the performativity literature: as Barman argues, the creation of impact measures helped to grow the market for impact investing (2015, 27).

Unlike Muniesa, Barman does not provide clear lines of descent from the pragmatist tradition to contemporary pragmatist work in valuation studies. Looking back to the central themes of pragmatism, the clearest connection is to studying human action as a process. The empirical focus of valuation studies could also be connected to pragmatism’s doubts about more abstract theoretical work but this is a feature of most work in the contemporary social sciences regardless of meta-theoretical allegiances. One might also invoke the pragmatist attachment to plurality, but in doing so one begins to undermine the claims to this being a specifically pragmatist approach. The focus on socio-material arrangements comes from actor-network theory rather than pragmatism; indeed Porpora suggests that pragmatism has contributed to the resistance of American sociology to taking account of material relations (Porpora 2015, 116–7). Likewise, social constructionism owes no particular debt to pragmatism; and as we have seen conventions theory only recognized an affinity with pragmatism after the fact. Barman, reflecting wider practice in the field, melds many different elements together into a framework (one of those things that pragmatism rejects, according to some of its theorists). It is a productive framework but it is not a distinctively pragmatist framework. It is also noteworthy that, contra Muniesa, value has reappeared as an object of study: ‘value is understood as a social construction’ (Barman 2015, 10).

By contrast with Muniesa and Barman, Beckert wears his pragmatism rather lightly, despite having published with Hans Joas, one of the leading contemporary pragmatists (Joas and Beckert 2001). In his major book Imagined Futures, Beckert acknowledges a debt to pragmatist action theory, for which action is an unfolding process ‘in which ends and strategies are formed and revised based on contingent and changing interpretations of an emerging situation’ and suggests that his own central concept of fictional expectations resembles Dewey’s ends-in-view (Beckert 2016, 14). Beckert returns to Dewey occasionally in the book, but this is only one of many influences he acknowledges. Even in the section on the roots of the concept of fictional expectations, Durkheim receives more attention than Dewey (Beckert 2016, 13–14). Despite his credentials, Beckert does not position himself as a pragmatist, but as a scholar whose influences include pragmatism. The contrast with Barman is instructive. Both cite a similar range of influences that includes pragmatism but also many others, but while Barman claims the whole complex of influences as distinctively pragmatist, Beckert does not.

Discussion

These claims for pragmatist influence on the study of value, then, are somewhat problematic. Clearly each of the authors that makes these claims has taken something from pragmatism, although they have not necessarily taken the same things from it, but these elements of pragmatism have been recombined in larger patchworks with elements of other traditions that owe little to philosophical pragmatism. If the study of value has taken a pragmatist turn, then, it is tempting to suggest that this was only one turn in a spiral of others: conventions theory, actor-network theory, Durkheim and social constructionism could all claim significant theoretical influence alongside pragmatism.

Yet these authors continue to label their frameworks as pragmatist, even though in some cases these frameworks seem to draw at least as much from actor-network theory as from pragmatism. Even some of the apparently pragmatist elements – notably methodological individualism/atomism and rejection of explicit ontological argumentation – were also present in actor-network theory before it started to engage with pragmatism. Having said that, the framework does seem to depart from actor-network theory in the priority it accords to human action and interaction. One could therefore see these frameworks as modifications of actor-network theory, absorbing elements of pragmatism and conventions theory, but then relabelled as a pragmatist framework.

While different actors in this story approach it from different perspectives, there is a sense in which their discourse is contributing to performing valuation studies as a field, and as a field with a particular theoretical orientation. There is a strong parallel with how Science and Technology Studies and Social Studies of Finance were constructed as fields. Fields (which are perhaps proto-disciplines) are not defined purely by topic, but are complex constructions around topics driven by academic-political purposes, and in the cases of both STS and SSF part of the rhetorical agenda was to establish actor-network theory as an obligatory point of passage. Fieldisation not only delineates a space of topics but also institutionalizes a particular perspective on them, at least as a reference point that scholars cannot ignore. If this is the case, then assertions of pragmatist influence on the study of value are at least in part a rhetorical move that, on the one hand, demands recognition of a particular framework in the field, and on the other, attaches an old label to a new synthesis to enrol new allies.

Whatever the balance of influence between different traditions, pragmatism has certainly influenced contemporary work on financial valuation, even if it is only one part of the complex of traditions that have framed it. Its influence is most visible in the attention the field has paid to the role of action in shaping valuation. This attention has been valuable, but this section will argue that it also comes at a cost if the pragmatist perspective leads to neglect of the structural forces that are also at work.

As we have seen, pragmatists have tended towards an individualist ontology, with many denying that social structures can be causally significant, although others are more open to structural explanations as long as they recognize the causal role of individuals. Within the study of finance and valuation Muniesa and his colleagues have advocated ‘trusting the flatness of the world’ and reject talk of ‘hierarchical levels of reality’, very much in the anti-structuralist spirit of Callon and Latour (Muniesa, Doganova, and Ortiz 2017, loc 245). Other pragmatists in these areas have said little explicitly about these questions, but the field is dominated by studies of process at the micro level. This may be a sign of the influence of actor-network theory as much as pragmatism, but whatever the balance of influence, it produces a relative neglect of the larger scale mechanisms these processes might contribute to, of the roles of structure and culture in shaping individual action, and of the macro consequences of processes of valuation. It is hard to discuss social issues at all without implicit recognition of structural actors at least in the form of organizations or institutions, as we find even in Muniesa, where stock exchanges appear as actors, and Barman, who ascribes significant influence to the Rockefeller Foundation, but neither explicitly acknowledges the ascription of causal influence to social structures. This avoidance and/or backgrounding of structural influence becomes part of the habitus of the field – the relatively unexamined orientations implicitly encouraged by the socialization scholars experience from colleagues, reviewers and publications, which contributes to the longer term shaping of study.

The neglect of structure is not universal in valuation studies. Beckert, notably, invokes structural factors including ‘cultural frames, dominant theories, the stratification structures of a society, social networks, and institutions’ as well as ‘powerful actors’ (Beckert 2013, 326). Another important exception is Bruce Carruthers, whose analysis of the history of credit rating agencies and their role in the 2008 financial crisis is framed as an analysis of the role of institutions, reflects on regulatory questions, and invokes the influence of the discursive environment in generating belief in credit ratings (Carruthers 2010). But their explicit acknowledgement of the importance of structural forces makes them outliers in the field.

A related question is the field’s relation to social constructionism. As we have seen, Barman suggests that in valuation studies ‘value is understood as a social construction … generated out of the valuation practices, conventions and devices present in the situation of study’ (Barman 2015, 10). The language of social construction, though, is used much less than the language of practices, conventions and devices. This can perhaps be traced to the rejection of social constructionism by actor-network theory. Latour rejected constructionism partly because it neglected the contribution of material objects to our understandings of them, but more explicitly because it ascribed causal significance to larger social forces. In a move that resonates with pragmatism, Latour called instead for social scientists to replace structural explanations with analyses of specific interactions between empirically observable objects – actor-networks (2005). For critical realists, Latour’s call to open these black boxes is welcome up to a point: the old constructionism tended to attribute social construction to abstract forces like language or discourse without examining how those forces could produce the effects ascribed to them. We can understand constructionist claims better if we see construction as a process, or rather as a range of processes that have some similarities, and identify the mechanisms that make it work (Elder-Vass 2012). Once we understand those mechanisms, we can start to see which constructionist claims are viable and what kind of effects it is reasonable to claim for social construction. Latour’s approach is problematic, though, because there is no place in it for structural factors as contributing to these processes, and because their products are stripped of their structural character – the black box is opened only to be discarded entirely once the contents have been examined (Elder-Vass 2008). The focus on interaction between individual agents and material devices in valuation studies reflects Latour’s orientation, but Barman at least takes us one step back towards constructionism by recognizing that practices, conventions and devices do generate social constructions.

Arguments like this lead us towards a realist constructionism that is mechanism-oriented and structure-aware (Elder-Vass 2012).⁶ From this perspective, value and valuation depend on social construction as a process in which discursive forces shape, for example, the principles or conventions of valuation that are found acceptable when we interact with each other, and those discursive forces themselves are influenced by the balance of social power (Elder-Vass 2019). In the markets for fixed-interest securities, for example, there is a convention that the value of a security depends (among other things) on the creditworthiness ratings assigned to it by the major US rating agencies. The value of these securities is thus socially constructed in a process where the rating agencies play a clearly identifiable part, and that part has evolved through an institutional history in which this convention and their role in it were favoured because they served the interests of a range of powerful actors, including banks, governments and international organizations (Carruthers 2010).

A final problem is that this neglect of social structure contributes to an absence that is filled, not within the field, but outside it, through the validation and reproduction of a division of the study of financial value (beyond formal economics) into two distinct fields. Work on the topic has become bifurcated. On the one hand we have the micro-interactionist fields of social studies of finance and valuation studies discussed above. On the other we have the more or less exclusively macro-oriented field of the political economy of finance, dominated by loosely Marxist-influenced thinkers. While I have bemoaned the absence of attention to structure in the former field, the absences in the latter are equally problematic. As Beckert puts it,

While the strength of political economy has been the focus of the institutionalist explanation of macroeconomic outcomes, including social structures such as inequality and skill distribution, its underdeveloped part has been the microfoundations explaining the concrete processes underlying the phenomena observed on the macro level (2013, 324).

Although outstanding work has been done in both fields, one outcome is that with rare exceptions (notably Braun, 2016) neither field addresses the need to bring together macro and micro level influences nor the need to reconcile their explanation.

Conclusion

To conclude, let me return to some of the questions with which the paper started. What does it mean to adopt a pragmatist perspective, in general and in the study of financial value: what does it entail and what does it exclude? As we have seen, pragmatism means different things to different scholars, and there is a substantial set of themes that are associated (in varying combinations) with pragmatist philosophy, both in general and in the social sciences. Yet very few of these themes have played an explicit role in so-called pragmatist work in the area this paper has used to examine how pragmatism influences research in practice: the study of value. There is certainly an orientation to action theory and an aversion to engagement with social structure. We also find occasional reference back to more specific arguments from pragmatist authors, notably Muniesa’s invocation of Dewey’s work on valuation and Peirce’s theory of signs. But themes that have been imported from other traditions, notably material devices from actor-network theory and performativity from Austinian linguistic philosophy, are at least as prominent. What, on the other hand, does the apparent commitment to pragmatism exclude? There is little evidence of explicit exclusion, but the overwhelming focus on interaction appears to lead to a de facto exclusion of many aspects of structural explanation.

It is tempting to draw a parallel with an argument of Barman’s. Discussing professionals in the field of impact investing, she cites the argument that professional conferences are spaces where actors negotiate ‘the field’s identity’ (Barman 2015, 18). We might say the same about academic spaces such as conferences and journals. These spaces have been used to assert a pragmatist identity for the field of valuation studies in particular yet the frameworks that they employ are far from uniquely pragmatist.

What, on the other hand, does critical realism entail? Amongst other things, a commitment to the determination of social events by multiple interacting causal powers including structural forces, but also to reconciling those forces with micro level accounts. What might it exclude? This is a more difficult question. On the one hand, critical realists would challenge the neglect of structural elements, but on the other much of the explanatory work done in the study of value, whether by self-identified pragmatists or others, has made substantial contributions to understanding parts of the valuation process. Our concern would be that it must be complemented with contributions that make sense of the other parts of the process and the structures on which these processes depend. Some steps have already been made in that direction, even by scholars who profess an allegiance to pragmatism, either because their understanding of pragmatism diverges from the core version discussed here and is thus more open to constructionist and structural arguments, or because they have used those kinds of arguments anyway despite the tension between them and the core version of pragmatism. A critical realist perspective would push further in this direction, and might lead us to postulate mechanisms that have been neglected. By examining these structures, we also open the way for a more critical evaluation of the roles that they play.

Financial value, like most social phenomena, does not just depend on micro-interactions between social actors. It is also manipulated, for example, by powerful institutions that can shape prevailing discourses and governmental interventions to produce systems of profit that benefit some financial actors enormously at the expense of the rest of the population. The study of social phenomena is most effective when it traces the connections between these different elements rather than obscuring some of them from view.

Acknowledgements

I am grateful to Dean Curran, Steve Kemp, Line Nyhagen, Doug Porpora, Yufan Sun, and two anonymous reviewers for their extremely helpful comments, and to the virtual audiences who attended presentations at the SASE and IACR annual conferences in the summer and autumn of 2020.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Dave Elder-Vass

Dave Elder-Vass is an Honorary Fellow at Loughborough University, UK. His forthcoming book Inventing Value (Cambridge University Press, 2022) examines how financial entrepreneurs generate the belief that the financial instruments they create have value. Profit and Gift in the Digital Economy (Cambridge University Press, 2016) theorizes economic diversity and its implications for social theory and politics. He has also published extensively on social ontology and social theory from a critical realist perspective, including The Causal Power of Social Structures (Cambridge University Press, 2010) and The Reality of Social Construction (Cambridge University Press, 2012). For more information see his website: https://eldervass.com/.

Notes

1 This section draws heavily on an earlier publication (Elder-Vass 2021).

2 Tuukka Kaidesoja has criticized Bhaskar’s use of the term transcendental argument, which comes from Kant, because Kant regarded transcendental arguments as producing necessary, certain, and indeed foundational conclusions. Bhaskar consistently insists that all knowledge claims are fallible (e.g. Bhaskar 1986, 15). Nevertheless, rhetorically it might have been wiser to call his argument an inference from the best explanation as Kaidesoja suggests (Elder-Vass 2015; Kaidesoja 2013).

3 There is an intriguing parallel with James’s application of the concept of habit to material things, and in particular his point that the habits of a thing ‘are in the last instance due to the structure of the compound’ (1890, 60).

4 Although Bhaskar criticizes correspondence theories (notably in 1975, 249–50), he has an unusual understanding of them, equating them to ‘subject-object identity’ (1991, 32). Rather than realist accounts of truth as a relation between belief and an independent reality, Bhaskar seems to understand correspondence theories as anti-realist accounts in which there is correspondence between subject and object because there is no independent object. He went on to develop an alternative theory of truth, though this has been criticized by other critical realists in favour of correspondence theory (Groff 2000).

5 Haack’s artful counter-positioning of Rorty’s views with Peirce’s provides a nice illustration of the point (Haack 1998).

6 Constructionism, like pragmatism, takes many forms. Although it is widely associated with irrealism most applications of it are consistent with a realist social ontology, and critical realism itself is at least implicitly social constructionist, not least in the argument discussed above that social structures are concept dependent (Bhaskar 1993; Elder-Vass 2012).