History in Political Philosophy: Refutation and Imagination

Introduction

The importance of the history of political thought is often underestimated in contemporary political philosophy. Many political philosophers operate under the assumption, largely implicit, that historical fluency is an optional extra. Insofar as historical research appears in their inquiry at all, it is used to validate isolated claims for accuracy. In such cases, history sits alongside other modes of empirical enquiry and has no special status. Accordingly, the political philosopher feels under no special pressure to get to grips with the history of the concepts and ideals she works with. I want to resist this way of thinking about history’s relevance to political philosophy because it unduly limits the scope of history’s relevance in this area of inquiry. Historical inquiry doesn’t only supply evidence that undermines the content of explicit political claims. Rather, it supplies imaginative resources that expose limitations in the broader frameworks of understanding that political philosophies express, including the questions and concepts that political philosophers assume as starting points.

To motivate my position, §1 starts by considering examples—drawn from the Political Realist literature—of philosophers invoking historical research when seeking to undermine positions in contemporary political philosophy. Subsequent sections develop the view that these Realist criticisms are targeting a different—and ultimately, deeper—kind of epistemic failure than a theory’s fact-insensitivity. Instead, they target what I will call imaginative failures. Section 2 gives an account of this imaginative function to historically informed criticism, using David Graeber’s criticism of the Barter Theory of the conceptual origins of money as an illustrative example. Section 3 argues that this account captures how a number of thinkers from across the humanities have understood history’s relevance to political philosophy, including R. J. Collingwood, Quentin Skinner, and Michel Foucault along with Raymond Geuss, David Graeber, and Bernard Williams. Section 4 considers the most obvious challenge to what I am proposing here, which is that the imaginative function is ultimately no different from the refutational function. I conclude this section by arguing that the imaginary function of historical research is doing something different from, and more than, just supplying refutational counter-evidence. I finish, in §5, by arguing that this imaginary function cannot easily be duplicated through ahistorical methods of normative inquiry, such as state of nature accounts, or normative intuition pumps.

1. Historical Criticism in Raymond Geuss’ and Bernard Williams’ Political Realism

John Rawls and Robert Nozick begin their best-known works with bold assertions about key political concepts. Rawls begins A Theory of Justice with the claim that “Justice is the first virtue of social institutions.” Nozick begins Anarchy, State, Utopia by claiming that “Individuals have rights, and there are things no person or group may do to them (without violating their rights).” Neither claim is defended explicitly. Readers are expected to grant them based on the combination of their supposed intuitive appeal in our modern historical context and the cogency of the theory that emerges if the claims are taken as axiomatic. These claims are supposed to have enough prima facie appeal that readers won’t insist on scrutinising them before reading on. They are meant to strike us as credible candidates for universal assent.

Raymond Geuss criticises this feature of Rawls’s and Nozick’s work. With Rawls, Geuss says there are no real historical grounds for thinking that the concept of justice is properly central to politics.

Historicising Justice

No particular saliency had been attributed to ‘justice’ in the political philosophy of the modern period … Prima facie it seems highly unlikely that the analysis of a concept like ‘justice,’ which is so highly dependent on shifting forms of economic activity and on historically extremely variable conceptions of the good life, could give one any real grasp on the central phenomena of politics. (2006a: 3–4)

And as for Nozick’s declaration about rights, Geuss says he would like to ask Nozick,

Historicising Rights

why one should assume that the proper starting point for political philosophy should be a set of subjective rights at all … ? One possibility might be that we simply could not imagine any way of going about social life … but … we know that [another form of social life] did actually exist … the very concept of a ‘subjective right’ in anything like its modern form is an invention of the late Middle Ages. In particular, historians have argued that this conception arose during the discussion about the so-called vita apostolica. (2008: 64)

Understanding the shape of Geuss’ criticisms is not trivial. Taken literally, Geuss is proposing that an explanation for the structure of Anarchy, State, Utopia’s argument is Nozick’s inability to imagine a different social world. This seems at best hyperbolic. After all, Nozick acknowledges the relevance of history in shaping normative political theories—“at this late stage in history, to dream up a description of the perfect society, is not of course the same as starting from scratch” (1974: 313)—and speaks favourably of taking lessons from the successes and failures of past social experiments and experiences (313).

I think Geuss’s remark is best understood as placing some rhetorical pressure on Nozick. Geuss is trying to make vivid the absence of historical fluency in Nozick’s decision to simply take for granted that individual rights are a natural starting point for political philosophy. Whatever other lessons he has extracted from history, Nozick’s argument starts from an assumption that has intuitive purchase within a modern context, and concludes—without further justification—that this is an assumption that anyone should accept. Geuss finds this troubling. His concern is not that, as a matter of fact, many people would not have accepted this as a starting point for political philosophy. It is that Nozick seems to be doing more than just disagreeing with these people. He is writing them out of the discussion entirely.

These kinds of complaints are echoed in Bernard Williams’s review of Thomas Nagel’s The Last Word. Nagel asks whether the justification of the modern Liberal worldview can be grounded in universal principles, or whether it is always relative to a contingent perspective. Nagel favours a universalist view. Williams is critical of the trans-historical ideal of reason that underlies this view.

“To reason is to think systematically in ways anyone looking over my shoulder ought to be able to recognize as correct,” Nagel says near the beginning of his book. Anyone? So I am reasoning, along with Nagel, in a liberal way, and Louis XIV is looking over our shoulder. He will not recognize our thoughts as correct. Ought he to?—or, more precisely, ought he to have done so when he was in his own world and not yet faced with the task of trying to make sense of ours? (2014: 385)

If reason per se entails Liberalism, Williams asks, why have so many societies throughout history, and today, been confident in their illiberal views? Ultimately, in Williams’s view, Nagel

lacks an explanation of something that cries out for one… If we come to understand historically and psychologically how our own and others’ ethical thoughts came about, this can change the way we think about the status of our thoughts, and about their relation to other people’s. To neglect this possibility does seem to me to constitute a form of dogmatism in Kant’s sense, a refusal of the kind of critique that has made modern philosophy … what it is. (2014: 385)

These three examples are united in methodologically critiquing ahistoricist philosophers, by invoking history. But how does this link between history and criticism operate?

One possibility is that history’s job is to refute or debunk factual deficiencies in a political philosophy. If history has no distinct interest for the political philosopher amongst other empirical disciplines, it seems natural to think of its relevance to political philosophy in such broadly refutational terms. Political philosophies may be criticised for being ignorant of, or unable to explain, certain political-historical facts. These philosophies would have been different (and better) if their authors had done their historical due diligence. Call this the refutational function of historical research in political philosophy. History’s relevance to political philosophy is to unearth evidence which undermines some features of a normative view or programme in political philosophy.

There are plenty of examples of political philosophers that explicitly use historical analysis within their philosophical arguments in this broadly refutational way. For instance, in The Last Utopia Samuel Moyn (2010) criticises contemporary philosophical accounts of human rights for interpreting human rights as a timeless and universal normative reference point. On Moyn’s view, the modern ascent of human rights theory is precisely due to its historically contingent utility in addressing various problems of post-war diplomacy and political debate. In a related case, Noam Chomsky’s Responsibility to Protect (2009) criticises moral arguments for humanitarian intervention because they simply ignore the long history of humanitarian justifications for imperialistic actions by powerful states. Ignorance of the history of our normative concepts and frameworks, for both of these authors, can lead us to implausible judgements and conclusions, with respect to both practical and theoretical moral problems. One could also mention in this connection Elisabeth Anderson’s Private Government (2017) and Katrina Forrester’s In the Shadow of Justice (2019). In each of these cases, the authors are arguing that a particular normative political claim or theory is unjustified, because it in some sense misrepresents the world that it purports to theorise about, and the appeal to history is used to demonstrate this factual inadequacy.¹ Moreover, the refutationalist function is consistent with the views of methodologically ahistoricist philosophers. Indeed, my preceding discussion of Nozick provides a reason for seeing him as embracing history’s refutationalist function; I will presently offer some comments that support the same conclusion about Rawls’ later work.²

This kind of refutational function obviously has its limits. Although certain aspects of a political theory allow for empirical evaluation, historical evidence cannot refute normative claims directly. What historical evidence refutes is something about the applicability or ‘fit,’ that connects a work of normative theory to the real-world political phenomena that it purports to be a theory of. Moyn isn’t saying that philosophical theories of human rights are incoherent, he’s saying that there’s some kind of mismatch between their claims about the nature and content of human rights, and what human rights as we find them in real political life actually are. Charles Mills (2019) famously presses this kind of objection against Rawls. Rawls’s theory of justice is meant to apply to a society that’s understood as a cooperative venture for mutual advantage. Mills’ claim is not that it is incoherent to conceive of society in this way, but that the kinds of societies—like post-war America—to which Rawls (and his followers) thought the theory was supposed to apply cannot be plausibly seen as such mutually advantageous ventures. Yet regardless of whether refutational criticism proceeds through direct or indirect challenge, the important point is that under the refutational function historical insights are being cited as a form of counter-evidence, with a view to refuting or debunking political claims.

While I accept that history can serve this function, I want to resist a purely refutational understanding of Geuss’ and Williams’ history-based criticisms of philosophers such as Rawls, Nozick, and Nagel. Whatever its legitimate uses are, focusing solely on the refutational function overlooks what is most distinctive, and often most interesting, about what the history of political thought has to offer normative inquiry. The developments in Rawls’s own theory of justice serves as a helpful illustration of the refutational function’s limitations.

Starting from the “Dewey Lectures”, Rawls starts framing his conception of justice in a contextualist framework, as an elaboration of the “basic intuitive ideas that are embedded in the political institutions of a democratic society” … [that] have their origins in the Wars of Religion following the reformation and the development of the principles of toleration. In contrast with A Theory of Justice, Rawls does not take justice as fairness as valid for all societies anymore, but just “within a democratic society under modern circumstances.” (Testini 2020: 2)

This description of Rawls’ shift to a contextualist theory does not entail that the developments in Rawls’ theorising were driven by reflections on history. Neither does it render the validity of Rawls’ later theory of justice hostage to the truth of an historical account about the Wars of Religion. Nevertheless, in committing him to the existence of some historical account about the origins of the modern framework upon which the assumptions of his later theory of justice relies, it reveals something important about history’s relevance to political theorising. Rawls accepts that the context of reference to which his theory refers is underpinned by a set of empirically demonstrable facts. As Rawls puts it in the Dewey Lectures,

[C]onditions for justifying a conception of justice hold only when a basis is established for political reasoning and understanding within a public culture. (1980: 517)

What justifies a conception of justice is not its being true to an order antecedent to and given to us, but its congruence with our deeper understanding of ourselves and our aspirations, and our realization that, given our history and the traditions embedded in our public life, it is the most reasonable doctrine for us. (519)

In the case of modern democratic society, to which Political Liberalism refers, these conditions consist in a set of facts that includes both ideational elements—the idea of society as a fair system of cooperation, of citizens as free and equal persons, and of a well-ordered society—and social elements, like the fact of reasonable pluralism. Any historical account concerning the origins of this set of facts is thus relevant for the assessment of the theory and its rivals, including early-Rawls’ universalist theory of justice.

Recognising the responsiveness of political judgments to contextual historical factors seems like progress in a project of inquiry, as far as it goes. But if we focus solely on the refutational function, the implications of historical research may end up being unduly limited. The refutational function of history is ultimately a way of checking a theory for errors and “bugs.” It doesn’t necessarily call for deep engagement with an historical perspective.³ When treated in a superficial way, it just invites some late-stage reviewing to see how broadly or narrowly applicable a particular theory is, depending on its historical generality, or lack thereof. The universalism of Rawls’s (1999) theorising in A Theory of Justice is criticisable because it overestimates how broadly applicable the theory was; nonetheless, historical fluency about political thought and experiences that existed before the advent of the modern democratic society play no part in justifying the theory of justice found in Political Liberalism. Rawls’s updates to his theory of justice thus reflects a strategy for accepting the limited importance of history for political theorising: accept a contextualist framework about one’s domain of inquiry, while treating further inquiry into a full historical account of that domain as a merely antiquarian curiosity.

In contrast to this Rawlsian contextualism, I want to focus on thinkers from across the humanities whose account of history’s relevance to political philosophy is not satisfied by this relatively confined engagement with history. For reasons that we will shortly encounter, I think the impetus to reject a more confined engagement with history is best characterised as the result of historical inquiry’s imaginative function to political philosophy. The case of David Graeber’s critique of Barter Theorists in economics provides an illustrative example of the kinds of imaginative claims thinkers in the humanities have pressed, and which I think offers a more plausible model for the kind of epistemic failure that is being targeted by Williams’s and Geuss’s criticisms of Rawls, Nozick, and Nagel.

2. David Graeber’s Critique of the “Myth of Barter”

David Graeber has criticised a certain brand of philosophical accounts about the conceptual origins of money. Barter Theory is a conceptually-pragmatist answer to the question of why money became so central to our thinking about how to solve co-ordination problems concerning the distribution of material goods in societies across the globe. Under these accounts—such as that of Adam Smith’s The Wealth of Nations—the practice of exchanging goods for money is theorised to arise as a natural response to local practical pressures faced by all but the most materially uncomplex societies. It is a two-stage account.

The Theory of Barter

First-stage: The practice of bartering—a new concept—arises in virtue of the practical necessity that members of a society have to acquire a set of basic materials while lack the capacity to source all those goods themselves.

Second-stage: As time goes by, and the complexity of material needs in the society continue to grow, a new practical necessity becomes salient. Economists call it the problem of the double-coincidence of wants: in order for a successful barter exchange to occur both parties must at the same time be in possession of a good which the other desires. Bartering reliance on such temporal coincidence becomes a problem because the practice no longer assures a sufficiently efficient model for co-ordinating the distribution of material goods in line with individual needs.

As with the first-stage of the account, the need for a new concept emerges — the concept of ‘money.’ Physical instantiations of this concept display three functions—it serves as a medium of exchange, a unit of account, and a store of value—which conjointly resolve the efficiency problem.

Graeber believes that the Barter Theorists’ account is a complete work of fiction, and a particularly politically infelicitous one at that — hence his name for it, the “Myth of Barter.” One claim in Graeber’s critique is that there simply is no anthropological evidence that supports the existence of an early society whose primary solution to the problems of growing complexity in collective material needs came via the practice of bartering.⁴ Importantly however, this aspect of Graeber’s critique does not exhaust his complaints against the Barter Theorists. In a key excerpt, Graeber centres his criticism around an imaginative deficiency claim:

Recall here the language of the economics textbooks: “Imagine a society without money.” “Imagine a barter economy.” One thing these examples make abundantly clear is just how limited the imaginative powers of most economists turn out to be. (2011: 33)

Here’s how I understand Graeber’s argument in the larger excerpt from which the above fragment is extracted.

Historicising Money (a.k.a. The Myth of Barter)

• (i)

  The Myth of Barter aims to give an explanation, internal to the discipline of economics, for the ubiquitous use of the concept of ‘money’ in the discourse of so many societies across the contemporary world. To count as a valid explanation within economics certain axiomatic doctrinal assumptions must be instantiated in the account’s contents, such as the neat division of life into the marketplace and the sphere of consumption.

• (ii)

  Graeber views the situation described in (i) as problematic — the doctrinal axioms of economists do not describe assumptions shared by the very societies whose social practices Barter Theory aims to explicate. This fact is concerning because it means that the range of practical solutions available to people in those societies does not properly align with those the Barter Theorists are willing to contemplate.

• (iii)

  Graeber believes that this misalignment explains why the economists’ argumentation for Barter Theory are disposed to display a failure of imagination which, once observed, undermines its cogency.

• (iv)

  The cause of this failure of imagination is not personal, but disciplinary. It is economists as a group, and as economists, who are prone to field arguments displaying imaginative deficiencies, for two reasons:

  • a.

    Economists, by virtue of their characteristic research methods, remain ignorant of the plurality of strategies which actual historical societies have used for dealing with problems endemic to a growing complexity in its members’ material needs. Consequently, economists have difficulties conceiving of ways a society might approach the problem of satisfying its material needs that are different to their (the economists’) own: viz. solutions that hypostasize a sphere of the marketplace that can be isolated from other spheres of human life.

  • b.

    Economists have a vested interest in finding accounts that show the world to compare favourably with the disciplinary assumptions that underpin contemporary economic theory. Moreover, they exist in a society so-organised as to make this feature of their thinking obscure to them, by making it hard to imagine any other possible arrangement.

None of Graeber’s claims in this argument are about the fact deficiency of the economists’ account; the validity of his criticism is logically independent to the establishment of any factual errors in the Barter Theorists’ historical claims. Rather, as I suggested above, Graeber thinks that the Barter Theorists’ account involves an imaginative deficiency in virtue of which they are epistemically criticisable. According to Graeber, they fail partly through a lack of interdisciplinary curiosity and partly because of their objectionable use of prudential reasoning.

A fair question at this point is whether Graeber’s argument, directed as it is towards the first-stage of the Barter Theory, really is effective at undermining the Barter Theorists’ main contention. Even if they are wrong about the history of bartering, their conceptual analysis of money may still be valid. After all, the practical problem of co-ordinating the exchange of goods is real, and every existing society of a certain size and material complexity has adopted the concept of ‘money’ to solve it. Why then couldn’t the Barter Theorists start the argument from practical necessity at the second-stage, dismissing questions of historical-accuracy about the first-stage as irrelevant?

In the subsequent chapter Graeber addresses the challenge directly; the shape of his response provides further reasons not to treat Graeber’s argument as refutational.

Why not simply write off the myth of barter as a quaint Enlightenment parable … ? The answer seems to be that the Myth of Barter cannot go away, because it is central to the entire discourse of economics. (2011: 43)

I have argued that the point of identifying issues of historical accuracy with the first-stage of the Barter Theory was not to refute the account, but to challenge the reliability of the economists’ judgments about what is ‘practically necessary’ by showing that their evaluation of ‘necessity’ is predicated upon idiosyncratic framing assumptions. Graeber’s Myth of Barter argument is focused on the first-stage, but the intention is to undermine our evaluation of the economist’s judgments about practical necessity in general. By itself, however, this response risks begging the question: in the retort we are considering, the economists’ are precisely claiming that Graeber’s argument is ineffectual because the arguments in the two stages are independent. As I read him, Graeber accepts that the impact of his arguments about the Myth of Barter is dependent on his book’s wider success in establishing that the reliability of the economists’ judgments concerning the arguments in these stages are precisely not independent from one another.

In brief, here is his strategy to do so. Barter Theorists, like Adam Smith, like to think of judgments about economics as independent from historically parochial conceptions of human relations (see point (iv) above). Graeber argues that in order for this view to be tenable, the direction of explanation of social practices needs to run from practices around things like property, money, and markets to centralised political institutions like the state, noting that “[i]n this Smith was the intellectual property heir of the Liberal tradition of philosophers like John Locke, who had argued that government begins in the need to protect private property” (2011: 24). However, Graeber contends, the history of the conceptual origins of money, shows us the exact opposite. His discussion of the Myth of Barter provides some support for this claim, but it does not exhaust the supporting evidence Graeber provides. In the next chapter Graeber focuses his discussion on the history of the origins of money-based exchange proper (as opposed to barter). For example, he claims that the advent of ‘money’ is not best explained as a response to a general problem of the co-ordination of goods, but a much more specific practical problem the State faced—i.e. a problem that only came to exist after the advent of states—in attempting to maintain a (large and costly) standing army.⁵ Moreover, he claims that subsequent transitions to money-based exchange practices came through the imposition of exogenous forces that were already familiar with, and held a prudential interest in expanding, those practices.

Graeber’s subsequent chapter thus provides further historical examples which undermine a key assumption made by Barter Theorists: that ‘the market’ was a practically-necessary, spontaneous development. Admittedly, these further examples discussed in the next chapter do not provide evidence that directly contradicts the Barter Theorists contention that utilising the concept of ‘money’ is the only practicable solution to the problems of co-ordinating the exchange of goods in societies of a given size and material complexity. This simply reinforces the claim that we should read Graeber’s historical argumentation in Debt: The First 5,000 Years as primarily imaginative—aimed at revealing our susceptibility to imaginative failures—rather than refutational. Indeed, this emphasis fits better with his description of the project in the introduction.

This book is a history of debt, then, but it also uses that history as a way to ask fundamental questions about what human beings and human society are or could be like … . [T]he book begins by attempting to puncture a series of myths, [like] the Myth of Barter, … that in one way or another form the basis of our common-sense assumptions about the nature of economy and society … . The one thing that all these misconceptions have in common … is that they tend to reduce all human relations to exchange.

History’s role in the book is thus to make us take seriously the possibility that our inability to think of an alternative solution to the co-ordination problem of the distribution of goods in complex material societies is not due to unavoidable practical necessity, but to a failure of imagination in our attempts to conceive of human relations to exchange under a different framework than the one we have inherited from the Liberal tradition.

To be clear: the reader is free to disagree with Graeber that the imaginative function can be effective at establishing this conclusion. It has not been my intention to convince the reader that Graeber has offered us a compelling philosophical explanation for why we should put as much credence as him on the imaginative function. Ultimately, I will offer (in §5) my own view of why the imaginative function has a central role to play in political philosophy. It is a line of reasoning that I think is compatible with, but ultimately not articulated in, Graeber’s text. For now, all I want to convince the reader of is that we can correctly identify in Graeber’s text a critical function of history that is distinct from the refutational function.

A different question the reader might have is whether Graeber’s discussion of the Barter Theorist’s argument is relevant for philosophers, even accepting that it might be relevant for economists. Since Graeber’s argument is rejecting a descriptive account—how did the concept of money gain its current prominence in our thinking—one might worry that the structure of criticism it displays will be impotent against normative arguments. Given Barter Theory’s status as a foundational myth in economics, I think it would be false to claim that Graeber’s argument, even if valid, would not impact normative arguments in political philosophy. But rather than try to convince you of this claim, let me instead provide what I believe to be a more trenchant counterexample to the general worry. One can import this same Graeberian structure of argumentation against Nozick’s reasoning in his thought-experiment about the structural features of a ‘stable association.’ That thought-experiment has clear normative ambitions given its central role in justifying what Nozick thinks is the most fruitful conception of ‘Utopia’ available to political philosophy. Thus, by extension, one of his book’s overarching arguments for a Libertarian minimal state uses the same structure as a Barter Theory argument.

Nozick’s thought-experiment centres around imaginary worlds—i.e., worlds imagined by you, the experimenter—populated by you and other, similarly imagined, agents where certain rules are stipulated to apply. One of these rules is that any of your co-inhabitants must also be imagined having the power to leave the current world and inhabit a world of their own imagining (2013: 299). Within the terms stipulated by this thought-experiment, a stable association is defined as the associations represented by a stable imaginary world—that is, a world in which “all of the imagined population will choose to remain” (299). Nozick then asks,

What are such stable associations like? Here I can offer only some intuitive and overly simple arguments. You will not be able to set up an association in which you are the absolute monarch, exploiting all the other rational inhabitants. For then they would be better off in an association without you. (299)

Notice that since Nozick’s thought-experiment forms part of his overarching universalist argument in favour of a Libertarian minimal state, it needs to apply to all agents – both contemporary and historical. Nozick must have been aware that many historical agents did not share Modern intuitions about monarchy. And yet, given its transhistorical scope, the validity of his argument requires him to be justified in assuming that these very agents would not choose to remain in a Monarchically organised association. Some might feel that this discrepancy calls for an explanation. Nozick is content to take it as evident that our (Modern) intuitions give us solid intellectual ground to dismiss the possibility that pre-Modern agents might not judge themselves ‘better off’ in an association without a monarch (when under the conditions stipulated by the thought-experiment).

Historical inquiry cannot directly refute such dismissals, but they can bring other possibilities to our attention. Contrast Nozick’s intuition-based position with the historically informed opinion of historian Benedict Anderson:

These days it is perhaps difficult to put oneself emphatically into a world in which the dynastic realm appeared for most men as the only imaginable political system. For in fundamental ways ‘serious’ monarchy lies transverse to all modern conceptions of political life. (2016: 19)

Views such as Anderson’s give us reason to be suspicious about Nozick’s reliance on intuitions. It cannot be ruled out that contemporary intuitions track normative truths.⁶ But historical fluency suggests another possibility: that Nozick’s reasoning is being limited by a failure of imagination in conceiving of the possibility that pre-Modern agents could coherently fail to view the transition to non-Monarchical society as an improvement. As Williams (2006a, esp. 190–191) has argued, evaluating how historical agents would respond to contemporary forms of reasoning is not a trivial enterprise. Without history, such evaluations are vulnerable to conflating the claim that contemporary forms of reasoning have won, with the claim that they have won an argument (Williams 2006a: 190). The point is not to contest that we Moderns do not have good, normative, reasons to disagree with historical forms of reasoning [(Williams 2006a; 2008a: 9–12), (Queloz 2017)]. It is to emphasize that Nozick’s reasoning displays a specific blindness about methodologically ahistoricist philosophy’s vulnerability to failures of imagination that undermines the effectiveness of his arguments.

Before moving on to argue that Graeber’s imaginative concerns are reflected in thinkers across the humanities, let me offer an important clarification. In diagnosing the economists’ ‘imaginative deficiency,’ I do not believe that Graeber is merely identifying questionable patterns of deliberation like those discussed above, i.e., a lack of interdisciplinary curiosity, and an objectionable use of prudential reasoning. As already mentioned, I take the aim of the imaginative function of historical research to be non-refutational; the strategy cannot therefore be to refute the Barter Theorists account by the identification of explicit deliberative failures. Instead, when Graeber describes the Barter Theorists’ account as imaginatively deficient, he seems to have three thoughts in mind. First, that to the eyes of a historically well-informed evaluator, the assumptions the Barter Theorists make about how the concept of money must have developed will simply seem naïve. Second, although economists believe that their intuitions reliably track available solutions and that therefore the Barter Theory is persuasive, Graeber thinks there is a better explanation. Namely, that the strength of their intuitions about Barter Theory’s explanatory force is more plausibly explained by reference to phenomena like the economists’ limited interdisciplinary curiosity or their inadvertent misuse of prudential reasoning. Finally, the previous two points prompt questions about why a widespread contemporary disposition towards such imaginatively impoverished assumptions about the conceptual origins of money became so ubiquitous in the modern world.

Graeber believes that the ubiquity of this disposition can partly be attributed to the entrenchment of Barter Theory in our collective social-political imaginaries. Moreover, he thinks that the naivety underpinning the Barter Theory’s assumptions gives us a reason to change our approach in inquiring about the conceptual origins of money. In this, Graeber’s thinking seems in line with the view of historical arguments expressed by Geuss:

Historical arguments… are not in the first instance intended to support or refute a thesis; rather, they aim to change the structure of argument by directing attention to a new set of relevant questions that need to be asked… One of the effects that one type of historical account ought to have is that of causing it to seem naive or ‘unphilosophical’ simply to make a certain set of assumptions. (2008: 68)

For Geuss, historical research is not usually focused on a refutational output, but a way of operating an epistemic change on oneself to counteract ‘naivety,’ or ‘unphilosophicality.’ Mental states that comprise conditions of naïve or unphilosophical reflection instantiate what epistemologists call a non-luminous state. A condition is luminous just in case whenever one is in it, one is in a position to know one is in it. When we call someone’s evaluation naïve or unphilosophical, we mean precisely to imply both that one is in an epistemically deficient evaluative position and that one is does not find oneself in a condition to know that one is in an epistemically deficient evaluative position. Thus, the condition of the naïve or unphilosophical thinker is not luminous. To let a little light more light in, we need to get these thinkers to acquire a more lucid and realistic appreciation of the epistemic position they occupy. In the next section, I will argue that a number of thinkers from across the humanities have understood history’s relevance to political philosophy along these lines.

3. The Imaginative Function across the Humanities

Graeber and Geuss are not alone in conceiving of the value of historical arguments in causing our evaluative perspective to suddenly appear naïve. One finds passages that resemble Geuss’ views in the methodological reflections of other prominent Twentieth Century historical thinkers with a concerted interest in the philosophical underpinnings of their practice, such as Quentin Skinner and R.G. Collingwood. Here are two sample passages.

Skinner’s Liberty before Liberalism is a work whose publication led to the establishment—or according to Skinner, the revival—of a conception of freedom as non-domination in the professional philosophical literature. At the end of the book, Skinner offers a series of reflections about the benefits of historical research to political thought that he hopes his book instantiates.

The thought [about the relevance of intellectual history to political philosophy] at which I am gesturing is that, if we examine and reflect on the historical record, we can hope to stand back from, and perhaps even to reappraise, some of our current assumptions and beliefs. The suggestion I want to end by exploring is that one of the present values of the past is as a repository of values we no longer endorse, of questions we no longer ask. One corresponding role for the intellectual historian is that of acting as a kind of archaeologist, bringing buried intellectual treasure back to the surface, dusting it down and enabling us to reconsider what we think of it … .

Here then is one moral implicit in the story I have told:⁷ it is remarkably difficult to avoid falling under the spell of our own intellectual heritage. As we analyse and reflect on our normative concepts, it is easy to become bewitched into believing that the ways of thinking about them bequeathed to us by the mainstream of our intellectual traditions must be the only ways of thinking about them. (2012: 116)

Skinner’s description of the normative structure of his project could just as well be describing the structure of Graeber’s imaginary critique of the Liberal tradition.⁸ Moreover, notice the absence of any allusion to the refutational function of such historical research to political philosophy. Skinner, after all, describes its function as allowing us to access values we no longer endorse and questions we no longer ask, not refute those we do.

Whereas in his intellectual autobiography, the purpose of which was to retrospectively elucidate the most important aspects that unified his thinking, Collingwood argued that

the historian may very well be related to the non-historian [e.g., the normative political philosopher] as the trained woodsman is to the ignorant traveller… The historian’s business is to reveal the less obvious features hidden from a careless eye in the present situation. What history can bring to moral and political life is a trained eye for the situation in which one has to act. (1939: 100)

Collingwood imagines that some opponents might find this description to undermine, rather than reinvigorate, our conception of the utility of history to activities like moral and political philosophising:

This may seem a small gift. Surely, someone will say, we are entitled to ask for more than that. There is not much use in showing us the tiger unless you also give us a rifle with which to shoot him. (1939: 100)

Collingwood offers two observations to palliate this concern. First, that we do well to distinguish between the skill sets of the gunsmith and the woodsman: the former is skilful at crafting weapons, but not necessarily at using them. Second, that only the woodsman’s (i.e., the historian’s) skillset seems to be helpful in determining whether the creatures we can hear in the woods are the kinds of things for which we need guns (because they are the kind of thing that need shooting).

This passage does not merely work well as an—admittedly metaphorical—explanation of what Geuss might have had in mind when he claimed that historical arguments change the structure of an argument by directing attention elsewhere. It also works well as a rubric through which to interpret Graeber’s critique of the Barter Theorists. Interpreting that critique via Collingwood’s chosen metaphors, we might say that the kind of weapon the economists crafted—a tool for offering a unitary explanation for how all societies escape the problems of growing material complexity—was not appropriate for the circumstances in which they wished to act. For the economists (i.e., the non-historians) crafted a weapon that could only be used against target societies that accept the distinction between the marketplace and the sphere of consumption, but no such targets were to be found in the woods. Moreover, Graeber suggests, what seems needed is not a single kind of weapon—a tool for offering a unitary explanatory account—in the first place. Given the plurality of solutions to the problems of growing material complexity to be found in the historical-anthropological literature, what is needed is a different kind of tool completely—a non-unitary account of the origins of money.

The fact that philosophers, anthropologists, archaeologists, and intellectual historians have all expressed—from different angles, in different professional dialects—similar views about the relevance of history to normative inquiry underscores that taking the imaginative dimension seriously is not an idiosyncratic view. It is however, one whose importance has not been properly appreciated within analytic political philosophy. At least this is what Bernard Williams—admittedly writing about two decades ago—believed:

Collingwood insisted, correctly, that the questions being answered by Plato and Hobbes,⁹ for instance, were not the same, and that you literally could not understand them unless you understood this. The [Oxford] ‘realists’ had very little sense of this… If Collingwood had not been so ignored, it might not be so necessary, as I am afraid it is, to remind people of this now. It might also not have been necessary to rediscover for oneself that the point of reading philosophers of the past is to find in them something different from the present—and that is not just a historical but a philosophical discovery. (2006b: 344)

If Williams assessment remains correct, this is surely not because few philosophers engage in such reading but because they do not engage with those historical texts in ways likely to lead them to find something different from the present. What those readings lack, in other words, is an ability to access the imaginative function of historical research.

4. The Political Imagination in Graeber’s Critique

Let’s take stock. Historical research doesn’t just offer evidence for refuting or debunking views in political philosophy. That is, its relevance to normative theorising isn’t limited to what I call the refutational function. Rather, historical research can also provide imaginative resources for showing that views in philosophy adopt assumptions that are implausible and discrediting or that they focus too narrowly on questions that are wrongheaded or idiosyncratic. In other words, it can fulfil an imaginative function in addition to (or as an alternative to) its refutational function. At any rate, this is what I take authors like Geuss, Williams, Graeber, Skinner, Foucault,¹⁰ and Collingwood to be proposing.

In this section, I consider an objection that you might press when presented with this refutational vs. imaginative distinction. The objection is that talk of such ‘imaginative resources’ is just another way of describing certain types of evidence. And so, you might argue, there isn’t a real distinction here. After all, the supposed imaginative deficiencies of economists to which Historicising Money gestures is precisely to be explained by reference to knowledge of the facts that economists tend to lack, such as the kinds that Graeber mentions in his book about peoples such as the Gunwinngu and the Nambikwara (2011: 9–35). Economists lack facts that show their solution is not the only practicable one. Thus, it seems, their inquiry into why we use the concept of money is deficient because they lack historical facts. Specifically, they lack those which show that their solution for inventing a concept that can serve as a store of value, unit of account, and a medium of exchange, is not the only practical solution that can be given to the set of practical pressures faced by societies with materially complex needs. Put differently, they lack those facts that show the economists’ pre-theoretical assumptions about the world’s pragmatically necessary divisibility into the spheres of the marketplace and the sphere of consumption to be false. After all, it is not as if Graeber is positing that there is some intrinsic property of anthropologists that make them inherently more imaginative than economists, other than their knowledge of facts about peoples such as the Gunwinngu and the Nambikwara. If this interpretation is correct, then perhaps we should, after all, say that whenever historical research is used to criticise views in political philosophy, it is always being used refutationally. It’s just that the refutational evidence is sometimes subtler, or more complex, and thus liable to be described by invoking ideas of framing assumptions. So, when Graeber says that the myth of barter shows us something about the economist’s limited imaginative powers, what Graeber really means is just that economists in general are overlooking some important evidence.

Against such a redescription of Historicising Money, I wish to maintain that the refutational vs. imaginative function distinction is a substantive distinction. My aim in this section is to explain why. I’ll start by saying a bit more about imaginative failure. ‘Failure’ is a success-term. An agent exhibits a ‘failure’ to exercise her imagination when she does not exercise it sufficiently or properly relative to some putative standard. In what follows I assess Graeber’s use of ‘imaginative’ in formulating his imaginative-failure-based criticism.

Here is a natural way of reading the epistemic criticism Historicising Money expresses. The economists’ failure of imagination is a failure to satisfy some putative epistemic standard of justification. Justificatory power is a normative power: when I can—when I have the power to—justify something, I have a (pro tanto) reason to do something. Furthermore, the act of comprehending the justification for some phenomena serves an epistemic function: famously, I must comprehend the justification for my beliefs if I am to claim that I know them to be true. Thus, Graeber’s concern about a lack of justification in the Barter Theorists’ reasoning is concerned with a fault in the epistemically normative claims that they are making. Graeber ascribes to the imagination a normative power of justification which, when it is lacking, results in a criticisable epistemic failure. Failure of imagination as epistemic failure.

Of course, this reading is only coherent if one already thinks exercises of the imagination can be properly invoked in the context of justification. In other words, that it is not circumscribed to the context of discovery. There is disagreement on this point, because there is disagreement about the relationship to truth exhibited by the operation of doxastic processes and the operation of the imagination (Kind & Kung 2016). If one assumes that the output of any kind of operation of the imagination can never be invoked in the context of justification, then Graeber’s criticism of the economists is likely to strike us as not really about imagination after all. (In which case, the labelling proposed will probably be distracting for that reader, but it does not mean that they have lost all purchase on the phenomenon Graeber is trying to describe).

In order to determine whether Graeber’s has identified a plausible case where a failure to exercise one’s imagination has led to an epistemic failure of justification, let’s look more closely at the kind of justification involved. The kind of justificatory failure Graeber describes seems distinctive. Consider, for contrast, Timothy Williamson’s argument against viewing imagining and knowing as possessing non-overlapping kinds of epistemic import. Williamson argues that when the imaginative faculty is used in such a way as to offer selective, reality-oriented attention to practically relevant possibilities it can confer knowledge. Williamson has the following kind of case in mind:

Think of a hunter who finds his way obstructed by a mountain stream rushing between the rocks … . It is vitally important for the hunter to know whether he can jump the stream … . There is a natural human method of gauging one’s capacities in such situations. One imagines oneself trying. (2016: 4f)

This is not the kind of imaginative exercise Graber has in mind at all. While both Williamson and Graeber are content to invoke the imagination in contexts of justification, Williamson’s account defends this by reference to exercises of the imagination that play a direct role in the assessment of counter-factual truth values. This is a perfectly reasonable way to model some exercises of the imagination. Indeed, it seems that the Barter Theorists were familiar with this use of the imagination. For imagining oneself trying to live in a society without the concept of money to justify so many of our material distribution practices on the concept’s practical necessity is precisely what the Myth of Barter thought-experiment was trying to achieve. In relation to Williamson’s model, then, the Barter Theorists’ imaginative faculty was engaged in the right sort of mental operation. In relation to Williamson’s model, the only kind of failure one could ascribe to Barter Theorists is one analogous to committing an arithmetical error when seeking the answer to an exam question: the agent is engaged in the right kind of operation, but an unintentional lapse in concentration meant they came up with the wrong result. This is a failure of some sort, but it is not one born from epistemically irresponsible conduct—the economists, like the arithmeticians, are engaged in exactly the kind of cognitive exercise required and are trying their best—their reasoning just fell short. This is evidently not the kind of failure Graeber had in mind.

This brings us to the second thing to note about Graeber’s invocation of the power of the imagination. He ascribes a creative power to the imagination. One that goes beyond the kind of creative power involved in reality-oriented, counter-factual truth-assessments displayed in Williamson’s discussion. Moreover, for Graeber the imagination’s creative power seems to impact the space of reasons, and consequently the dynamics of the criticism of justification involved in a distinctive way. In the kinds of cases Williamson has in mind certain key features are held fixed. The hunter desires to jump across the stream. No part of the imaginative exercise they undergo seems to put in question the desirability of that objective. By contrast, in the case discussed by Graeber, the economists are described as exhibiting a failure of imagination in being unable to conceive of human relations to exchange under a different framework than the one we have inherited from the Liberal tradition. I have argued that his claim that Barter Theorists lacked imagination is specifically tied to a criticism of their justification for holding this feature of their reasoning fixed. To think within the constraints of such a Liberal tradition—to assume, as Graeber puts it, that “war, passion, adventure, mystery, sex, or death” can have nothing to do with human relations of exchange—is undesirable because, as he puts it in a different argument with a similar thrust (Graeber & Wengrow 2021: 3),¹¹ it makes the past needlessly dull. A needlessly dull conception of the past, in turn, seems to go hand in hand with an impoverished conception of human social experience and thought. This feature of the Barter Theorist’s thought, Graeber suggests, makes it imaginatively deficient and, ultimately, politically infelicitous. The key difference I want to highlight with this comparison, then, is between a process where imagination is used strategically to achieve a previously decided desire, and a process where the imagination is engaged in a manner that is actively producing (or destroying) desires that are relevant to the assessment of how I should act within a given situation.

The specific relationship between imagination and justification present in Graeber’s critique—and in the other authors’ I’ve mentioned—seems well-captured in the following observation by Bernard Williams:

[T]he deliberative process can add new actions for which there are internal reasons, just as it can also add new internal reasons for given actions. The deliberative process can also subtract elements from [the agent’s subjective motivational set]. Reflection may lead the agent to see that some belief is false, and hence to realize that he has in fact no reason to do something he thought he had reason to do. More subtly, he may think he has reason to promote some development because he has not exercised his imagination enough about what it would be like if it came about. In his unaided deliberative reason, or encouraged by the persuasions of others, he may come to have some more concrete sense of what would be involved, and lose his desire for it, just as, positively, the imagination can create new possibilities and new desires. (These are important possibilities for politics as well as for individual action.) (1981: 104–105).

Williams describes a situation where the imaginative faculty’s impact upon active possibilities and desires seems to describe a more robustly creative phenomenon than countenanced by Williamson’s exercises in counter-factual truth assessment.¹² And this is what makes Williams’ suggestion about the importance of imagination to politics so intriguing. It is a view of the importance of the imagination for politics that describes the exercise of the imaginative faculty as related to the exercise of justification. We can, after all, gain new reasons for political actions—but these reasons are not explicable in terms of an interpretation of the evidence that was available to me before the imaginative exercise. For this reason, the imaginative resources that historical inquiry offers to normative political philosophy are not adequately described as evidential at all, and are therefore not reducible to (a more subtle or complex version of) the kind of evidence provided by the refutational function.

5. Why Historians are Uniquely Well-Placed to Excite the Political Imagination

Suppose you grant these interim conclusions but remain unconvinced about their overall significance. Imagination in politics is epistemically oriented, in that it helps us assess justifications for political norms. It is also creative, in that it helps us conceive of new possibilities and desires, that change our perspective on those same norms to the benefit of our political theorising. Still, why think that historically-informed exercises of the imagination—as Graeber construes it—fulfils these tasks particularly reliably, or consistently? And relatedly, why think that other forms of ahistorical inquiry that we find in political philosophy—like state of nature stories or stylised moral intuition pumps—cannot fulfil these imaginative tasks just as satisfactorily? Different still, what should we say about the value of the literary genre of political fiction or of political anthropology about peoples outside our political lineage?¹³

The general shape of what is special about historical inquiry is not difficult to see. History is unique in offering us non-fictional access to political experiences that are distinct, but genetically linked, to our own. And there is prima facie reason to think such access might be valuable to normative theorising. Different experiences—for example, experiences held by people in remote historical circumstances—can elicit different perspectives on normative issues. If we can tap into those perspectives, we may learn something that current-day intuition pumps aren’t well placed to detect. Yet this intuition only gets us so far. Most forms of imaginative exercise worth the name will elicit different perspectives on normative issues in some sense – intuition pumps included. Therefore, the real sticking point, the sceptic about historicism is likely to insist, is the further supposition that there is something special about the imaginative value of what we learn from historical research that we could not also learn by some other (perhaps less costly and difficult) method of inquiry.

To be clear, I certainly do not claim that the exercises of the imagination involved in any of these activities is a normatively toothless way of thinking about political theory. Any exhaustive evaluation of the extent to which one should privilege historical over non-historical activities for imaginative engagement will inevitably require some form of—likely intractable—empirical investigation, which I will not attempt here. Instead, I will argue that reflection on the functions of political theorising suffices to show that (i) normative political thought is always at risk of imaginative failures, (ii) involving issues with political deliberation that normative philosophers should care about, and (iii) that historical inquiry into political thought is the only effective means we have for mitigating this risk. I believe that (i)–(iii) collectively give the political philosopher good reason to engage with historical inquiry, irrespective of whatever other distinct kinds of imaginative failures can be avoided through non-historical forms of imaginative engagement.

Furthermore, my explanation for why (i)–(iii) is importantly different from the kind of argument we find in the debunking literature. As Geuss has put this worry,

a major danger in using highly abstractive methods in political philosophy is that one will succeed merely in generalizing one’s own local prejudices and repackaging them as demands of reason. The study of history can help to counteract this natural human bias. (2006b: 38–39)

It might be tempting to construe the shape of his criticisms in Historicising Justice and Historicising Rights in light of such well-known debunking concerns. In what follows, I want to offer an account of the utility of history that goes further, by grounding our reasons to care for the history of political thought not merely on the psychological assessment that anyone engaged in abstractive methods is at risk of generalising local prejudices, but to explain the development of local prejudices as a necessary consequence of fulfilling a core function of political theorising. To understand why I think this is the case, it is helpful to clarify some background assumptions about the nature of political theorising that I am working with.

First, I take it that part of the remit of political theory is to assess and evaluate the social structures that are active in our society. By a social structure I mean a network of social relations, comprising relations between subjects and relations of subjects to things. I also assume both that at least some central social relations are constituted through social practices, and that social practices paradigmatically comprise collective solutions to problems of distributive co-ordination or collective access to valenced resources. Here are two simple examples of what I mean: promises and money. Definitions of both concepts will refer to social relations. I take it that the relations between promisor and promisee is a social relation constituted through a social practice—i.e., that the act of promising is, in some deep way, dependent upon a social practice of promising—and that it makes sense to explain this social practice in terms of a solution to collective problems of co-ordination. Similarly, the relationship between a subject and her money is a social relationship: agents in the state of nature will never encounter an object whose physical instantiation serves as a medium of exchange, a unit of account, and a store of value. And of those societies that possess the practice of speaking in terms of ‘money,’ that practice is similarly to be explained as a solution to a collective problem of co-ordination.

Sally Haslanger has proposed that social practices can offer solutions to collective problems because they are constituted by

clusters of culturally shared mental states and processes, including concepts, attitudes, dispositions, and such, that enable us to interpret and organise information and coordinate action, thought, and affect (2017: 21).

Adding that

these clusters offer collective solutions insofar as they are internalised by individuals, acting as public ‘schemas’ that enable individuals to act fluently in social contexts.

I think that applying Haslanger’s views to the examples I have been discussing helps elucidate the arguments involved. In Historicising Money, Graeber highlights both (i) that the Barter Theorists all display a shared disposition to think that the world of intersubjective social relations can be divided into the sphere of the market and the sphere of consumption, and (ii) that this disposition is most plausibly thought of as a kind of social practice that is underpinned by a cluster of shared concepts, attitudes, and dispositions which are historically parochial. Similarly, in Historicising Justice, Guess’s criticism centres around the lack of appreciation about the fact that the cogency of Rawls’s early theory of ‘justice as fairness’ presupposes a context of reference to ideational and social elements that was likewise historically particular — I.e., it presupposed the referential availability of shared clusters of mental states and processes that were simply not accessible to all political agents throughout human history. This was something post-Dewey Rawls himself would go on to accept.

Second, I take it that agents in societies where such public schemas are active will display a strong collective motivation to ensure that those schemas remain internalised by a sufficient proportion of other agents in that society. Another way of putting this collective-motivational point is to accept Haslanger’s further proposition that “although schemas are variable and evolve across time and context, their elements are sticky and resist epistemic updating” (2017: 21). Consider once more the examples of Historicising Money and Historicising Justice. Once again, we might say that the agreement amongst economists concerning the divisibility of spheres of human action, and the agreement between Modern political agents concerning their conceptions of the nature of society and citizenship, can both be explained by these agreements offering collective solutions to problems of distributive co-ordination or collective access to valenced resources. Yet to count as de facto, as opposed to a merely notionally coherent, solutions to such problems, such agreement must be shared by a sufficiently large proportion of the population to allow the management of these problems to remain tractable.

Suppose you concede these two points. That is, you agree that (i) political theorising is concerned with the normative assessment of social structures along the lines I have proposed, and that (ii) the social practices which (either wholly or partly) constitute our social structures are themselves constituted by clusters of culturally shared mental states and processes whose elements—i.e., concepts, dispositions, attitudes, and such—resist epistemic updating. In such circumstance, I would argue that epistemically prudent political philosophers have strong reason to desire a mechanism that can appropriately counterbalance our collective tendency to resist epistemic updates in our political thinking of the kind Haslanger describes. More specifically, political philosophers ought to have an interest in counteracting epistemic overreliance on public schemas when trying to think about the space of conceptual possibilities through which we might seek to address collective problems of co-ordination and access with respect to valenced (political) resources. It is also worth emphasizing that the ideas underpinning this last point generalise beyond the specificities of Haslanger’s analysis of social practices. While I find that analysis to be readily applicable to the examples I have been discussing, this is not to say that it necessary to apply her views specifically to make sense of those examples. Indeed, even if the reader rejects Haslanger’s story, they can still agree that our ways of making sense of the world exhibit radical presentist biases. And if one accepts that, then my central point—that we have good reason to adopt methods of inquiry that counteract the relevant biases—seems to follow.

The previous passages argued in favour of a correction mechanism against epistemic overreliance in historically specific, culturally shared mental states and processes in our collective political deliberations based on certain assumptions about the nature of political theorising. I will now argue for the same conclusion by reasoning from assumptions about the nature of the political imagination.

I have been arguing that a central function of political theorising is to help us to solve collective problems co-ordination and access via the internalisation and normalisation of culturally shared mental states and processes that act as schemas which enable individuals to act fluently in social contexts. One theorist of the political imagination who advocates a theory that is sensitive to precisely this dynamic is Avshalom Schwartz (2021). According to Schwartz, the political imagination is a heterogeneous human faculty comprising a constitutive, a creative, and a critical element. The constitutive imagination is the element of the political imagination responsible for organising the web of social meanings that must exist inside a society for it to remain sustainable and stable. As Schwartz further characterizes it, the constitutive political imagination is “a force that contributes to the maintenance of legitimacy and order. It sets limits: it denotes the boundaries of what is questionable, just as much as it sets the boundaries of what is considered as imaginable” (2021: 3330).

In setting up the boundaries of what is questionable and imaginable, however, the constitutive imagination—if operating in isolation from other human faculties—is disposed to create a situation in which “a political community might be incapable of adapting itself to changing environments and circumstances” (2021: 3340). Hence the need for the creative and critical elements of the imagination which—unlike the constitutive imagination—represent a challenge to the established order (2021: 3328–3329). The creative imagination is responsible for the expansion of the boundaries of the imaginable, whereas

the critical imagination represents a critical reflection of the foundations of social order that aims to expose their imaginary character. Since the constitutive imaginary rarely represents itself as a product of the imagination (but instead aims to appear as natural, ahistorical, and objective), such critical reflection is neither common nor easy. (2021: 3036)

Notice that this description of the function of critical imagination’s relevance to our political thinking is just another way of acknowledging our need for a mechanism that can counteract the form of epistemic overreliance against the presentist biases exhibited in our ways of making sense of the world that I discussed in relation to Haslanger’s thought. They are two sides of the same coin. The risk Schwartz describes with the solitary operation of the constitutive imagination can thus be translated into Haslanger’s parlance. It is in the nature of political theorising (as I have characterised it) to rely on public schemas to resolve problems of co-ordination and access. Yet, to fulfil that role, such public schemas will be resistant to epistemic updating, thereby introducing an epistemic source of error and risk to our theorising. More precisely, I think that the semantic extension of what Schwartz calls the constitutive imagination as picking out some salient, specifiable subset of the semantic extension of Haslanger’s notion of shared cultural clusters of mental states and processes. As such we find, in Schwartz’s reflections about the nature of the critical imagination, the counterpart to the mechanism for avoiding the epistemic risks that, for Haslanger, are indigenous to our dependence on social practices to resolve collective problems of co-ordination and access.

We thus find a common pattern in both Schwartz’s and Haslanger’s theories. Both theories showcase the need for a constitutive and critical element in our theorisations about the political domain. These elements stand in a symbiotic relationship, wherein the critical element takes the products of the constitutive element for its target. In Schwartz’s theory, the political function of the constitutive imagination is to preserve such political goods as order, legitimacy, and identity. But the kind of imaginary co-ordination required to preserve our access to these political resources requires what I will call a self-effacing structure: attaining a sufficient agreement on ‘institutional facts’ (Searle 1995) to allow for the preservation of societal order often requires the (de facto) justification of those beliefs to fade into the background. This creates a risk of “stagnation and ossification” (Schwartz 2021: 3339) in our political thinking—i.e., political thinking that, by virtue of the machinations of the constitutive imagination, exhibits unduly constrained patterns of deliberation—which it is the function of the critical imagination to counteract. Haslanger’s theory is broader in scope, applying to the entirety of the social—not just the political—domain. As such, one cannot as easily pin-point paradigmatic concepts to which it is meant to apply. Nonetheless, the self-effacing structure described above is, for analogous reasons, also present.

It is at this point, finally, that we are in a position to appreciate why historical inquiry is uniquely well-placed to stimulate the political imagination. Such inquiry helps us assess justifications for political norms by providing a methodologically inimitable access-route to new possibilities and desires that change our perspective on those same norms, to the benefit of our political theorising (in the manner described by Geuss, Williams, Graeber, Skinner, and Collingwood). The point is most clearly made in Schwartzian terms. Simply put,

Historical Inquiry’s Unique Selling Point (USP):

• (P1) The constitutive imagination plays an essential role in collective justifications of (our) social structures.

• (P2) The constitutive imagination is justificatorily self-effacing.

• (C1) [From (P1) and (P2)] Any reliable assessment of our social structures requires a mode of inquiry that can uncover that which has been effaced.

• (P3) Historical inquiry is, by definition, the only kind of non-fictional, genetic inquiry.

• (P4) (P3) describes the only form of inquiry that has the capacity to track the shape of that which has been effaced and offer insight into the original reasons for that effacement.

• (C2) [From (P3) and (P4)] Historical inquiry is uniquely well-placed to counteract this justification-based self-effacement effect.

• (C3) [From (C1) and (C2)] Therefore, methodologically historicist inquiry is uniquely well-placed for a reliable assessment of our social structures.

Schwartz’s contraposition of the constitutive vs creative/critical functions thus allows us to get clearer on why the imaginative function of historical research, but not the refutational function, is imaginary in nature. Simply put, it is because the imaginative function contributes towards the proper functioning of the critical imagination: it helps prevent the ossification and stagnation of the public imaginary. The refutational function might also occasionally bolster such operations—some refutations might carry consequences that trigger creative and critical elements of our imagination. Yet, insofar as the refutational function is not designed to distinguish those doxastic attitudes that are a product of our constitutive imagination from those that are not, it will not be near as efficacious at counteracting stagnation and ossification.

More specifically, the imaginative function of historical research counts as imaginary because, unlike the refutational function of historical research, its raison d’etre is to target the creative phenomenon described by Williams—that “the imagination can create new possibilities and new desires. (These are important possibilities for politics as well as for individual action.)” (1981: 105). This does not amount to a generic claim that historical research gives us more evidence. Rather, it is the claim that historical research helps us appropriately exercise our imaginative faculty so that our assessment is less vulnerable to the machinations of the constitutive imagination. The imaginative function of historical research is a mechanism for facilitating the functioning of the critical element of the political imagination mitigating the risk of imaginative failures.

This completes my exposition of the imaginative function, my argument for its relevance to normative theorising, and my argument for why historical inquiry is uniquely well-placed to stimulate that function. Before turning to my concluding remarks, I would like first to briefly return to Geuss’s criticism in Historicising Justice. Originally, I suggested that it was natural to read that passage in terms of the refutational function of historical research. Now that we have an alternative at hand, I want to contradict that statement; this would be a very poor interpretation of that passage. For notice that nowhere in that passage does Geuss actually impute Rawls for making a fact-deficient universalising assumption over historical intuitions about justice. Rather, it is that historical sensitivity about the different ways that justice has been conceptualised throughout the history of political thought, plus the correlation displayed between historical conceptions of justice and historically parochial ideational and social elements, speaks against Liberal intuitions that simply take for granted that it is natural to view the concept of ‘justice’ as exceedingly central to politics.

To be clear, I am not arguing that Geuss’s argument refutes Rawls. As Geuss himself noted, “[historical arguments] are not in the first instance intended to support or refute a thesis” (2008: 68). What I am arguing is that the intention of Geuss’s argument is to guide our attention to the fact that Liberal patterns of deliberation might be prone to failures of imagination which unjustifiably restrict the range of solutions to collective problems of politics that Liberals are willing to contemplate. The reason why self-described ‘Political Realists’—like Geuss and Williams—call our attention to such possibilities is not merely that they believe Liberals patterns of deliberation about politics might be prone to such failures of imagination. Rather, it is because their own sustained engagement with history of political thought gives them reason to believe that Liberals habitually underestimate just how vulnerable to such epistemic failings they really are. This further concern, I take it, explains Geuss’s and Williams’ motivation to engage in first-hand research into intellectual history, as they do in Public Goods, Private Goods (2001) and Truth and Truthfulness (2010), respectively. Similarly, it is also this concern which I believe centrally motivates their oppositions to the orthodox structure of defences of Liberal theories found in Anglophone philosophy,¹⁴ amongst which is their opposition to Liberal assumptions about what it is epistemically safe to assume that any reasonable person will find justifiable.^15,16

Historical Inquiry’s USP helps us understand how such critiques are supposed to work. Geuss says that his historical and genealogical arguments aren’t in the business of refutation. I’ve tried to put a finer point on what business they are in. It is also what Williams’s preferred form of historical inquiry involves—both authors, after all, have been deeply influenced (albeit to different ends) by Nietzsche’s genealogical method.¹⁷ For Geuss, genealogies challenge the background assumptions that govern and stabilise a shared set of institutions, norms, values, and beliefs. They problematise “some deeply entrenched contemporary item or phenomenon: a practice, institution or identity… that [misleadingly] presents itself as unitary and coherent in that all of its parts fit together smoothly and naturally” (Geuss, 2002: ix). Such inquiry and critique is non-refutational, because its point is precisely to provide the critical distance needed to stop viewing such items or phenomena as ‘natural’ targets for normative thinking.

How, then, does this approach undermine ahistorical normative inquiry? Think again of the historicist concessions Rawls makes to his theory of justice as fairness: “So Modern intuitions about justice are not natural,” he concedes, “even still, they have a normative grip on us that pre-modern intuitions do not. That’s enough to assuage the worries history raises to my theory.”¹⁸ My contention is that once the critical distance afforded by historical inquiry is conjoined with the premises underpinning Historical Inquiry’s USP, Rawls’s response is untenable. It reveals Rawls’s position as vulnerable to the constitutive imagination’s self-effacing structure, therefore revealing his stance as epistemically imprudent, because imaginatively impoverished. It is historical inquiry’s ability to reveal this kind of insight to us, rather than the mere rejection of the universalizability of Modern intuitions, which is the real target of Geuss’s observation.

6. Conclusion

I suggested that the attitude of most methodologically ahistoricist philosophers display a commitment to the limited relevance of research in intellectual history to political philosophy. Such thinkers acknowledge that implementing their political theories would involve ensuring their historical fact-sensitivity, thereby investing historical research with normative relevance to political theorising. Accepting such limited relevance, however, comes methodologically cheap: in only acknowledging the refutational function of historical research it does not actually obligate robust engagement with intellectual history. Moreover, the threat of the refutational function is easily obviated by, on the one hand, accepting an in principle contextualist framework about one’s domain of inquiry while, on the other, waving off the relevance of robust inquiry into any historically distinct contexts about that domain as of merely antiquarian interest.

By contrast, I defended a richer conception of the relevance of historical research to political philosophy by arguing that historical research does not just offer evidence for refuting or debunking views in political philosophy but can also provide imaginative resources. Such imaginative concerns, I contend, are reflected in the methodological reflections of thinkers across the humanities, including R. J. Collingwood, Quentin Skinner, Michel Foucault, Raymond Geuss, David Graeber, and Bernard Williams. Taking David Graeber’s argument against the ‘Myth of Barter’ as an exemplar, I have argued that the issue with the limited relevance claim is that it leaves the theorist open to threats of imaginative failure. Such threats are significant because they suggest methodologically ahistoricist thinkers’ judgments will be epistemically unreliable. The unreliability will be about the degree to which the patterns of deliberation they employ import assumptions that surreptitiously restrict the range of solutions to collective practical problems they are willing to contemplate. This provides normative political philosophers with reasons to reject “the limited relevance of research in intellectual history to political philosophy” thesis.

Moreover, in providing an effective mechanism for the avoidance of imaginative failures, historical research provides a normative function that is unavailable to the methodologically ahistoricist philosopher: this is what I call Historical Research’s USP. Due to the role that social practices play within the context in which political theorising occurs, the normative political theorist ought to be suspicious of self-effacing justificatory mechanisms in our collective political thinking. The reason for this suspicion is due to what Haslanger calls the epistemically sticky nature of the elements comprising these shared cultural clusters. Analogously, it is what Shwartz is referring to when he talks about the looming risk of stagnation and ossification that accompanies the normal functioning of the constitutive political imagination. The only effective response to the threat of such epistemic stickiness, therefore, is through an epistemic mechanism that is able to keep the existence of such self-effacing justificatory mechanisms in our collective thinking within our sights. I argue that historical arguments provide an intellectual mechanism that can cause it to seem ‘naïve’ or ‘unphilosophical’ simply to accept the contents of our constitutive imagination at face value. The execution of such a mechanism is what I call the imaginative function of historical research. It is a function of historical research because only a non-fictional mechanism of genetic inquiry can fulfil such a role.

Acknowledgments

I am grateful to Tom Stern, Angus Gowland, Rob Simpson, Mark Kalderon, Dan Greco, Fiona Leigh, Tim Button, Euan Allison, Chihiro Larissa Tsukamoto, Matt Liston-Jones, Jack Hume, Alice Harberd, Kirstine La Cour, Kenta Sekine, and Kim Leontiev, as well as to the two anonymous reviewers at Ergo in helping me develop the piece into its current form. I would also like to extend my sincere thanks to the Arts and Humanities Research Council (AHRC) for their financial assistance, which provided me with the opportunity to pursue the doctoral research on which this article is based.