Weberian legitimacy and its discontents

While it is true that modern empirical approaches to legitimacy can largely be traced back to Weber, the above quotation about social-scientific legitimacy being just belief in legitimacy is deceptively simplistic. Critiques of Weber’s approach abound, especially when it comes to his typology of forms of legitimation and his silence on the processes through which social support is formed and sustained.Footnote ¹⁰ But we need not delve into those controversies, insofar as the scholars involved remain committed to the broadly Weberian project of construing legitimacy as an explanatory concept for the phenomenon of authority, that is, the willingness to comply with or not oppose a power structure for reasons other than mere fear or calculations of advantage. According to Alba Ruibal, “the widely shared view across contemporary social sciences that the problem of legitimacy is defined in a social relationship, and that it is not possible to evaluate the legitimacy of a system of power without considering the views of the ruled, continues to be the enduring Weberian legacy.”Footnote ¹¹ Indeed, the focus of contemporary social-scientific approaches to legitimacy is often an extension of the Weberian approach, in the sense that it goes beyond a description of the different ways in which structures of authority may enjoy support and focuses on the dynamics through which “the structures and processes of social entities become aligned with collectively supported norms, values, and beliefs.”Footnote ¹²

Partly building on that type of work—and more relevantly for our present concerns—some contend that Weber’s account of legitimacy is not merely descriptive or explanatory, but also contains normative commitments that can take it into the philosophical camp. Two related sets of arguments that may be deployed to construct a Weberian theory of normative legitimacy stand out in the literature. The first one is due to Tamsin Shaw, who puts forward an intriguing critique of Weber’s skepticism about democratic rule as well as an alternative account broadly in keeping with Weberian commitments.Footnote ¹³ The second one is due to Amanda Greene, who resourcefully argues that Weber’s account of legitimation provides a viable moral standard for the evaluation of political regimes.Footnote ¹⁴ My starting point is thus that I will argue that broadly Weberian normative commitments—which have variously been called “civic alignment” or “responsiveness”—are not sufficient (on their own) to get an empirically grounded normative theory of legitimacy off the ground.

In so doing I will, whenever possible, try to range over various methodological controversies about how one may or may not empirically measure the properties of a system of authority that give rise to legitimacy. However, for reasons that should soon become clearer, it is probably best not to ignore another social-scientific debate that is closely connected to the issue at stake here, namely, the one on responsiveness and other indicators of the quality of political—especially democratic—institutions. The claim that developing a normative version of Weberian legitimacy is unusual and controversial is likely to surprise those who follow the empirical literature on democracy. In this literature, it is standard to think of “responsiveness”—essentially, alignment between public opinion and policy outcomes—as a key indicator of a well-functioning democratic polity.Footnote ¹⁵ The step from democratic responsiveness to Weberian legitimacy seems fairly short: one can take responsiveness as a proxy or at least an indication of broad support for authority.Footnote ¹⁶ As Andrew Sabl notes, something like this view often appears to be the implicit normative theory adopted by empirical scholars of democracy, but it is far from anything political theorists and philosophers think about the matter.

Empiricists typically claim or assume that “democratic theory” tells us how important responsiveness is. However, to the extent that democratic theory means the reflections of political theorists who study democracy, this claim is essentially false. Political theorists do not and never have regarded responsiveness as the central measure of democratic quality. When they have imagined a perfectly responsive regime, they have judged that this would be a bad thing. Democratic theorists, put simply, are not playing the role that empiricists have written for them and empiricists are not interested in the play that theorists are actually starring in.Footnote ¹⁷

What is going on here? Weber’s own famously skeptical if not despondent views on democracy and popular rule provide an answer that anticipates this theoretical-empirical divide. In short, Weber thought that modern states, even when they take the form of electoral democracies with mass participation, cannot afford any meaningful degree of popular rule. Modern states are too large and too complex for the people collectively to be at the helm of the political process. For that would require direct democracy, which is only a live option in much smaller polities that do not require as much division of labor as modern societies.Footnote ¹⁸ So much political power evaporates as it makes its way through a capillary bureaucracy that escapes the control of anyone’s political will.Footnote ¹⁹ To the extent that any power is wielded, it can only be wielded by charismatic, Caesarist leaders in ways that may find support with the people—in Weber-legitimate ways, that is—though that may involve a significant amount of demagogic manipulation. This is why, for Weber, modern democracy is far from being rule by the people.

It would seem, then, that Weberian legitimacy comes apart from the ideal of popular rule that does much of the normative work in underpinning democratic regimes. Legitimacy ends up looking like a description of a passive attitude tinged with irrationality; as such, it is not clear how it can compete with the typically more demanding ideals put forward in philosophical theories of legitimacy. This is arguably why philosophers typically consider Weberian legitimacy normatively sterile or at least inadequate.

Upending that standard judgment is the challenge Shaw takes on when she tries to construe a broadly Weberian theory of democratic legitimacy. The problem, as she puts it, is one of whether there is any sense in which the people can be said to have power in modern democracies. Her ingenious solution is to reject Weber’s understanding of political power as voluntary control, while retaining Weber’s insight that legitimacy consists in a form of alignment between the rulers and the ruled. Simplifying somewhat, we may say that Shaw’s claim is that while Weber is right to say that democracy cannot be rule by the people, he is wrong to dismiss the power that people may wield if democracy turns out to be rule for the people. To demonstrate why that may be the case, Shaw argues that Weber had at his disposal an alternative account of political power, one not centered on an account of freedom “derived from the conception of the methodical-rational personality produced by inner-worldly asceticism.”Footnote ²⁰ That is to say, Shaw ascribes Weber’s rejection of the democratic ideal of popular self-rule to his commitment to a view of rule as a form of freedom, in turn understood as individual rational control. As an alternative that is faithful to other general Weberian themes, Shaw puts forward an understanding of democracy that divorces self-rule from an ideal of freedom: “[W]hilst it may be the case that democracy is instrumentally useful in preserving freedom, we do not have to see it in itself as the realisation of freedom.”Footnote ²¹ Rather, “more significantly from the point of view of Weber’s concerns, democratic participation might help to make political life more responsive to human values … making political rule maximally responsive to the avowed values of a population.”Footnote ²² This, for Shaw, would be an account of democratic legitimacy that is closer to Weber’s own reading of ancient democracy and one that could be adapted to modern democracy by shedding “inherited post-Christian encumbrances”Footnote ²³ such as Weber’s “nostalgia for the systematic self-control, the autonomy, that ascetic Protestantism fostered.”Footnote ²⁴

To bring Shaw’s argument into a closer conversation with our present concerns, we may restate its conclusion as follows. Rulers’ responsiveness to the people’s values gives normative weight to Weberian legitimacy, thus making it more than a mere description of a mechanism through which compliance is produced. Put differently, people power or popular rule—though probably not popular sovereigntyFootnote ²⁵—can be realized through responsiveness, that is, when political power is exercised in accordance with values widely held by those over whom it is exercised. Lack of direct popular control over political decision-making does not preclude other Weberian avenues for normative evaluation of such decision-making, so measuring the axiological distance between ruler and the ruled yields a substantive normative criterion of legitimacy.

However, as anticipated, I do not find that strategy fully satisfactory, albeit for reasons that, at least in part, differ from those of most other political theorists. Shaw herself points in the direction of my main worry, which has to do with how genuinely the ruled support the ruler.Footnote ²⁶ When extolling responsiveness of political rule of “the avowed values of a population,” she goes on to say that “we might independently hold that these ethical commitments should be protected as far as possible from ideological manipulation, in which case, again, we arrive at a more fully articulated account of liberal democracy.”Footnote ²⁷ Set aside the issue that liberals have historically been uninterested in, if not contemptuous of, ideology critique.Footnote ²⁸ After all, I am interested in a broadly Weberian account of legitimacy tout court, not liberal-democratic legitimacy. Rather, the problem can be provisionally stated as follows. While protection from ideological distortion seems indispensable if we want ruler-ruled value alignment to be more than just a description of one way in which the rulers rule (that is, by inculcating their values in the population), it is not clear that problematizing the quality of beliefs in legitimacy is compatible with Weber’s idea that legitimacy just is or supervenes on ruler-ruled value alignment. It is not clear, that is, how we may introduce ideology critique as a normative filter without depriving the theory of its distinctiveness and turning it into a standard philosophical account of legitimacy, that is, one that gets its normative force from commitments beyond what can be empirically observed. At the very least, Shaw does not tell us how introducing ideology critique would be compatible with what we may call Weber’s resignation or even complacency about the top-down nature of rule and its attendant expectation that reinforcing or even producing support for authority in the ruled just is part and parcel of ruling.Footnote ²⁹

The normative-descriptive gap still seems as wide a chasm as ever. To see exactly why that is the case and what my proposed remedy is, it will be useful to consider another attempt to extract a normative theory of legitimacy from Weber.

We can read Shaw’s suggestion to fortify Weberian legitimacy with ideology critique as signaling a normative dissatisfaction with an aspect of Weber’s view, namely, the self-justification of power. As will become clearer below, even though I do not think Shaw does enough to show how her solution can work within her framework, I share this dissatisfaction. Others do not. Greene and Ilaria Cozzaglio bite the bullet, arguing that even when political power generates its own support among the ruled, this support has significant normative weight.Footnote ³⁰ They partly establish that point via a critique of Bernard Williams’s realist theory of legitimacy, in which he proposes a critical-theoretic amendment to Weber—much like Shaw or vice versa. While I share some of Cozzaglio and Greene’s reservations about Williams’s proposal, in the next section I will try to show that there is a viable combination of Weberian legitimacy and ideology critique. To get there, though, we first need to consider some arguments that say we do not need to or cannot go there.

According to Greene, Weberian legitimacy is “morally valuable”Footnote ³¹ insofar as it brings about “‘stable civic alignment’, which is the realisation of political stability with minimal brutality and intimidation.”Footnote ³² The idea is that most mainstream philosophical theories of legitimacy tend to conflate the value of legitimacy with other commitments, “such as justice, liberalism, and democracy.”Footnote ³³ In contrast, the Weberian approach manages to isolate a more modest but more distinctive account of the normative significance of legitimacy, one that can illuminate precisely how some widespread moral commitments “can come apart from political legitimacy.” What is more, this approach also has the added benefit of “maintain[ing] alignment with the empirical study of legitimacy by social scientists.”Footnote ³⁴

This normative reading of Weber, then, has one key idea in common with Shaw’s, namely, that the focus on civic alignment (essentially, what I have been calling responsiveness) enables us “to approximate more closely the moral ideal of a voluntary association,”Footnote ³⁵ while avoiding the Scylla of unattainable literal voluntarism and the Charybdis of hypothetical agreements unmoored from reality.Footnote ³⁶ In a sense, on this view civic alignment or responsiveness are the best we can do in terms of legitimacy—as opposed to justice and other values—if we care about the normative significance of what citizens think and are serious about what a modern polity may look like—or at least a modern polity that is a state.Footnote ³⁷

Not coincidentally, a number of political realists have recently been developing normative theories of legitimacy broadly along those lines. The general angle of what I have termed the “ordorealist” current of realism is that legitimacy is best thought of as a relatively permissive standard, aimed primarily at securing relatively weak forms of assent or acquiescence to political power, which in turn constrains the worst excesses of political power.Footnote ³⁸ In so doing, these theorists emphasize Bernard Williams’s “Basic Legitimation Demand,” namely, his idea that the political order must merely be stable and “make sense” to those subjected to it, which is a far weaker demand than what is typically found in standard liberal theories of legitimacy and one that can be grounded in an account of what politics (as opposed to raw domination) is, rather than in moral commitments such as liberty or autonomy.Footnote ³⁹ These same theorists, however, deemphasize or ignore another key element of Williams’s theory of legitimacy: his “Critical Theory Principle.”Footnote ⁴⁰ This is the idea that political power does not really make sense to those subjected to it if this making sense is the product of the power itself. As we have seen, Shaw’s proposal to supplement Weberian legitimacy with ideology critique proceeds from this sort of insight. Radical realists agree; they center various versions of this idea even as they criticize its coherence and other shortcomings.Footnote ⁴¹ However, it is not always clear on what grounds ordorealists reject or downplay this aspect of Williams’s realism, unless one puts political concerns ahead of philosophical ones, a move that may well suit the ordorealist temperament.

Should we care about ideological distortions in the acceptance of power? Can we do so while remaining faithful to Weber’s broadly realist orientation?Footnote ⁴² Cozzaglio and Greene, as anticipated, offer an argument for why a realist need look no further than Weberian responsiveness to establish legitimacy. They do so by building on Greene’s account of the normative significance of Weberian legitimacy. The starting point is the observation that Weber’s accomplishment was to establish a distinction between legitimate and illegitimate political domination without the need for moral commitments. This is an improvement over Thomas Hobbes’s “might makes right” account of legitimacy:

The key normative but not moralized step is that domination, or rule, can be conceptualized as something other than raw coercion or self-interested calculus: “[P]olitical domination is legitimate when its exercise of power is viewed by subjects as corresponding to their values.”Footnote ⁴⁴ This proceeds from an explanation of the social phenomenon of compliance with power, not from a moral postulate. Cozzaglio and Greene refer to this requirement “to maintain congruence between the mode of legitimation and the underlying values in a specific political community” as “the ethic of rule.”Footnote ⁴⁵ They also identify a second dimension of evaluation that yields “a basis for developing a realist form of political normativity” in what Weber himself calls “the ethic of responsibility,” namely, an ethic of fitting means to ends, however those ends may be determined.Footnote ⁴⁶ For Cozzaglio and Greene, the ethic of responsibility is particularly interesting when combined with the ethic of rule, as it yields another ground on which power may be criticized:

We thus end up with a Weberian theory of normative legitimacy consisting of two necessary and jointly sufficient conditions:

Note the words I emphasized above, for this is where Cozzaglio and Greene part company with Williams: “[F]rom Williams’s perspective, Weber’s view is defective because it appears to allow for power to produce its own acceptance…. But according to Weber, power producing its own acceptance is consistent with legitimate political domination.”Footnote ⁴⁹ Cozzaglio and Greene are adamant that given Williams’s realist commitments, he is not entitled to supplement the Basic Legitimation Demand with the Critical Theory Principle. On their view, realism can support the increase in normative demandingness from Hobbes’s to Weber’s theory of legitimacy, but not the further requirement introduced by Williams. They argue for that conclusion by showing that Williams has three options to make good on his claim, but two are unpalatable—inconsistent and circular—and the third one is obscure to the point of near hopelessness. The first option would be for Williams to argue that “it is wrong to use power in order to influence values and beliefs.” However, Cozzaglio and Greene note, “it is moralistic to say that some uses of political power are moralistic as such.”Footnote ⁵⁰ The second option would be to argue that “when power produces its own acceptance, it is inconsistent with a relationship of political subjection,” but that seems circular, because “[i]t claims that there is something inherent in the idea of political power that entails Williams’s interpretation of legitimation, as opposed to that of Weber.”Footnote ⁵¹ After all, Cozzaglio and Greene argue, it is not as if Weber does not have an account of power that affords a critical standard for its evaluation. Finally, the third option would be to show that “a violation of the critical theory principle shows that the first political question has not been solved [that is, that the Basic Legitimation Demand has not been met].” This “seems very hard to do, if all the requirements of legitimacy must arise from the idea of political order,” for even Hobbes’s account of legitimacy requires order. Cozzaglio and Greene conclude that “[m]aybe the normativity in Weber … is the most that political realism can have.”Footnote ⁵²

I agree with Cozzaglio and Greene that the first two options are unviable. Indeed, I and Janosch Prinz argue elsewhere that Williams seems to take the first one, since his Critical Theory Principle ultimately rests on “an aspiration to the most basic sense of freedom, that of not being in the power of another.”Footnote ⁵³ I am even tempted to agree with Weber that politics may fundamentally be about some people being in the power of others. Still, I do not think that the third strategy is as arduous as Cozzaglio and Greene suggest, so in the next section I will show how an epistemic form of ideology critique can succeed where Williams’s crypto-moralized Critical Theory Principle fails. What is more, I will suggest that without the addition of such a normative requirement, Weberian legitimacy cannot fulfill the role of a normative theory.Footnote ⁵⁴

Critical responsiveness

How, then, is it possible to say, without introducing moral commitments, that a stable political order that produces its own acceptance is illegitimate? My suggestion is that to answer that question we need to understand Williams’s “first political question” as being to a significant extent an epistemic question, that is, a question about the preconditions of inquiry into how social orders function and what social orders are possible.

Let me explain. Williams says: “I identify the ‘first’ political question in Hobbesian terms as the securing of order, protection, safety, trust, and the conditions of cooperation. It is ‘first’ because solving it is the condition of solving, indeed posing, any others.”Footnote ⁵⁵ My idea is that the formulation I have emphasized also refers to the question of whether the order is a genuine political order and not a case of raw domination or suspended warfare. The first political question remains unsolved unless we are in a position to judge properly whether it has been solved. We cannot know whether the first political question has been solved, unless our understanding of social reality is good enough, but power that generates its own support is bound to distort this understanding. For example, Williams argues that there can be cases of ostensibly stable political order that are, in reality, “pure cases of internal warfare,” such as the situation of the Helots in ancient Sparta.Footnote ⁵⁶ However, there can be cases in which an ideological distortion succeeds in disguising successful internal warfare as mere rule, at least with a sizeable part of the population. Apartheid—an example to which I return below—may well have been such a case. At any rate, this is why “the acceptance of a justification does not count if the acceptance itself is produced by the coercive power which is supposedly being justified.”Footnote ⁵⁷ We cannot properly pose, let alone answer, any questions about which social arrangements are justified if our grasp of how the world works is distorted by a ruling ideology. We need a theory of ideology to tell us when those epistemic distortions are bad enough to compromise the way in which an order answers the first political question. The first political question, then, is not a moral question but it is, in part but irreducibly, an epistemic question. If that is the case, my epistemic interpretation of Williams’s Critical Theory Principle shows how this normative requirement is “derived entirely from the nature of politics,” and so satisfies Cozzaglio and Greene’s desideratum for a properly realist standard of evaluation.

What is more, my approach turns the table on Cozzaglio and Greene’s suggestion that realists committed to something like Williams’s Critical Theory Principle risk collapsing into a Hobbesian position where political subjection just is political legitimacy.Footnote ⁵⁸ Rather, it is those who only look at Weberian assent or support without considering the real conditions in which it is given that risk being unable to distinguish between a genuine political relationship and one of raw domination or suspended warfare. That is the sense in which, as anticipated, my argument shows that pure Weberian legitimacy is not an adequate normative standard. However, Weberian responsiveness plus an epistemic form of ideology critique is still much closer to the facts than are most alternative normative theories. Importantly, epistemic ideology critique is itself a form of empirical inquiry into the genealogy of beliefs in legitimacy, so the approach proposed here—critical responsiveness—should be able to bridge the normative-descriptive gap discussed at the outset of this essay.

To see better why that is the case, we should unpack further the idea of critical responsiveness. I take the “responsiveness” part to be covered by the discussion above of social-scientific theories of legitimacy. It is the “critical” part that needs some elucidation, so let us begin by setting out some desiderata for the form of epistemic ideology critique that underpins this view. We may call the approach “radical realist social analysis.” It is a nonmoralized form of genealogical inquiry. Its normativity is epistemic in origin. It is centered on an account of a distinctly political type of epistemic distortion, namely, the self-justification of power, so the approach can generate evaluations of political orders without drawing on moral commitments.

The general idea behind this epistemic account of ideological distortion is that self-justifying power causes “legitimation stories” to become epistemically suspect, regardless of their truth and of whose interests they effectively advance. A legitimation story is a set of beliefs and other “cultural technes” deployed in support of political practices or institutions.Footnote ⁵⁹ To provide a normative edge for the sort of realist theory of legitimacy outlined here, then, radical realist social analysis must satisfy three desiderata:

1. (i) Debunking desideratum: The analysis must afford debunking judgments on legitimation stories, avoiding what has come to be known as the genetic fallacy, namely, mistaking the faulty causal history of a belief with the lack of alternative arguments in its support.

2. (ii) Realist desideratum: The analysis must eschew moral commitments or any other commitments that may themselves be comparably ideologically distorted.Footnote ⁶⁰

3. (iii) Justificatory desideratum: The analysis must show why a lack of significant epistemic distortion is sufficient to consider a legitimation story justified.

I contend that social-scientific evidence can be used—in a two-pronged manner—to challenge legitimation stories in politically productive ways. Social science can either uncover instances of self-justifying power directly, by looking at the genealogy of legitimation stories, or indirectly, by revealing that a legitimation story is not what it seems, that is, it does not fit our best social-scientific understanding of the relevant social dynamics, which in turn rings an alarm bell about a possible genealogical problem. In the parlance of contemporary epistemology, the latter kind of epistemic normativity is accuracy-driven, the former is justification-driven.Footnote ⁶¹ Strictly speaking, accuracy-driven considerations do not debunk, but they can be useful when they lead us to uncover justification flaws. Ugur Aytac and I elsewhere defend at length the justification-based account, by combining the social psychology of motivated reasoning and the theory of epistemic circularity to provide a microfoundation for radical realist social analysis.Footnote ⁶² Given the focus and space constraints of this essay, here I will simply lay out the view in an analogical way. Consider this easy case. A scholar may in principle be the best critic of their own work; nonetheless, a journal editor would be epistemically reckless if they were knowingly to use a referee report written by that paper’s author. The underpinning epistemology here is reliabilist: self-refereeing is not a procedure that tends to yield accurate or trustworthy results.Footnote ⁶³ Likewise, legitimation stories with the sort of pedigree characterized above are epistemically flawed, and so ought to be rejected. Importantly, this is not because of whose interests they advance, but because of their genealogy. That, in turn, yields a pro tanto reason to withdraw support from the practices and institutions underpinned by such stories. Yet nothing directly follows from the debunking about the practices and institutions themselves, and so the genetic fallacy is kept at bay and the debunking desideratum is met.

Unlike straightforward cases of self-refereeing, most real-world legitimation stories hide self-justifying power under layers of history and culture, hence the centrality of empirical results to radical realist social analysis. Sometimes, the distortion can be directly adduced to state power or other clear-cut forms of (coercive) agency.Footnote ⁶⁴ Sometimes, the historical and cultural thread will be more garbled, because societies typically reproduce themselves through their existing structures; routinely, and often in epistemically faultless ways, we believe legitimation stories about our society that originated from the society itself, and so from its power structures. Therefore, we need a more fine-grained criterion to identify epistemically flawed cases of cultural reproduction.

The solution to that problem is to distinguish between different types of cultural transmission and social reproduction. The rough idea is that people and social groups who enjoy a distribution of power particularly skewed in their favor (men in a strongly patriarchal society, say) cannot be seen as epistemically reliable producers and reproducers of legitimation stories about their social order. More specifically, we have reason to be particularly suspicious of legitimation stories involving “family values” in patriarchal societies, private property in capitalist societies, and so on.Footnote ⁶⁵ Note that this is not a moral critique of the inegalitarian or unjust nature of such social orders. It is an epistemic critique. If a society is highly stratified, its hegemonic or simply mainstream cultural production cannot be trusted not to be ideologically distorted, typically through power self-justification mechanisms. To buttress their position, those in power are likely (deliberately or not) to distort perceptions of social reality. This is purely about their inclination to retain power, and so whether they are (morally) entitled to their position is irrelevant to the epistemic assessment of the situation. A straightforward example would be the legitimation stories about ethnic origins that officially underpinned Apartheid in South Africa and were easily debunked with historical and ethnographic evidence.Footnote ⁶⁶ Conversely, the more egalitarian the power distribution in society, the less likely such distortions would be, ceteris paribus. To be sure, this view has radical implications; crudely, it amounts to an epistemic argument against hierarchy. Many widely accepted social and political structures—from states to families—are likely to be called into question by it. This is a feature, not a bug, of the view. Yet the aim of this approach is not completely to eliminate the effects of power from the ways in which we make sense of social reality. Rather, the aim is to find out by how much those effects can be reduced. Matt Sleat contends that this is a betrayal of the traditional realist attention to the inescapability of power.Footnote ⁶⁷ I would retort that, on the contrary, it is a doggedly realist attempt to find out how power really works. It is true that the realist tradition contains a complacent or resigned attitude toward the politically familiar. However, realism also contains an aspiration to grasp reality beyond surface appearances; this may not be the realism of Weber or Carl Schmitt, but it is the realism of Friedrich Nietzsche and Karl Marx, among others.Footnote ⁶⁸ In other words, it is far from clear whether Weber was right to maintain that successful political rule must generate its own support by influencing—distorting, as I would rather put it—the beliefs of the ruled. Like Weber, Sleat seems to assume that social order requires hierarchy, whereas radical realism seeks to put pressure on this assumption.

That said, one might still ask whether this epistemic critique of power asymmetries is a violation of the realist antimoralist desideratum. It is not—at least insofar as it is possible to use a nonmoralized account of power, especially when power is observably coercive. Crudely, to say that an agent has power over another need not imply any moral judgment about this situation. Even though the concept of power itself may be essentially contested and so shaped by one’s prior beliefs, it is not obvious that those should be characterized in moral terms as opposed to terms of interests or identity.Footnote ⁶⁹

Before turning to the third and final desideratum, we should pause to consider a possible objection, namely, that the epistemic normativity at play here is probably not entirely politically innocent. We have seen how some realists regard ethics as dead politics. However, politics has a way of insinuating itself nearly everywhere, so it would be naïve to presume that epistemology is not also, to some extent, dead politics. Yet there is ample reason to think that it is incomparably less so than ethics. After all, given the pervasiveness of moralism in political philosophy, it is not even a desideratum for much ethics to be politically innocent (or as politically innocent as possible), but it certainly is for much epistemology, pragmatic encroachment notwithstanding. As Quill Kukla puts it in a recent overview of literature critical of old-fashioned epistemic purity: “We cannot do epistemology without fundamental, central attention to social identities, power relations, and the social institutions and structures within which epistemic practices happen. But … this result is of no threat to our usable notions of objectivity, justification, and the like.”Footnote ⁷⁰ What is more, it is worth reminding ourselves of the different functions of ethics and epistemology qua social practices. Ethics is directly and primarily concerned with regulating behavior, so its links with social hierarchy are direct and strong.Footnote ⁷¹ Epistemology plays a role in underpinning the social order, but only rather extreme forms of postmodernism would posit that its primary aim is not truth or knowledge and that it is as politically contaminated as ethics.Footnote ⁷²

Finally, let us consider the justificatory desideratum. Suppose we are satisfied that a population’s belief in the legitimacy of the regime they live under is not epistemically distorted—or rather, more precisely and as we have seen at the outset of this section—not epistemically distorted to a degree that casts doubt on whether the basic legitimation demand has been met. Does that imply that the regime is legitimate? In other words, even if the citizens have formed the belief in legitimacy without the sorts of distortions caused by steep power asymmetries, how can we conclude that this belief is justified, and so that they have reason to comply with the authority? To be sure, moral answers are readily available. Citizens should be free, autonomous, sovereign, and so on. But a realist must eschew such answers and, instead, probe the limits of epistemic normativity. Here is a thumbnail sketch of how that might work. For reasons of space and focus, I submit it as a question for further research more than as a firm thesis. The core idea may be assimilated to an empiricist form of standpoint epistemology. Quite simply, those subjected to political power get to decide whether the power is justified because they are epistemically best placed to understand the import and consequences of that power. Radical realist social analysis is merely a tool to help them maintain cognitive clarity.Footnote ⁷³ That seems the most epistemically cautious default position, if only because often the forms of expert knowledge that may claim to override the standpoint of those directly affected turn out to be invested in some of the very hierarchies that would otherwise be called into question.Footnote ⁷⁴ Ultimately, then, the normative grounds of radical realism are to be found in the aspiration to improve the epistemic position from which we make decisions about society.