How I Learned to Stop Worrying and Love Political Normativity

Do salient normative claims about politics require moral premises? Political moralists think they do, political realists think they do not. We defend the viability of realism in a two-pronged way. First, we show that a number of recent attacks on realism as well as realist responses to those attacks unduly conflate distinctly political normativity and non-moral political normativity. Second, we argue that Alex Worsnip and Jonathan Leader-Maynard’s recent attack on realist arguments for a distinctly political normativity depends on assuming moralism as the default view, which places an excessive burden on the viability of realism, and so begs the question. Our discussion, though, does not address the relative merits of realism and moralism, so its upshot is relatively ecumenical: moralism need not be the view that all apt normative political judgements are moral judgements, and realism need not be the view that no apt normative political judgements are moral judgements.


Introduction
Where do normative claims about politics come from?From what sorts of premises do they follow?Most post-Rawlsian Anglo-American political philosophers maintain that most normative political judgements are derived from or reducible to moral normativity.Over the last decade, a realist challenge to that orthodoxy has emerged.In different ways, realists reject the centrality or even the pertinence of moral normativity to political philosophy.In a recent paper, Jonathan Leader Maynard and Alex Worsnip try to put the realist challenge to rest.They discuss five realist arguments for a 'distinctively political normativity', and contend that all five 'fail to establish a sense in which political normativity is genuinely separate from morality' (Leader Maynard and Worsnip, 2018: 764).This and other attacks have led some realists to distance themselves from the idea of a non-moral political normativity.'The origins of the realist/moralist distinction turned not on the demarcation of different normativities', they argue, 'but on how political theory understands the relationship between morality and political practice' (Sleat, 2021: 2).Relatedly, others argue that realists should be 'either indifferent or openly sceptical of the metanormative thesis [about political normativity]', and focus on offering 'practical guidance' instead (Bagg, 2022: 6).Others suggest relocating the discourse away from normativity and towards empirical political psychology -which, it turns out, lends plausibility to realism (Kreutz, 2022a).More perplexingly, some even say that the distinctiveness of political thought cannot be defended philosophically, by engaging 'in the way that Leader Maynard and Worsnip seek', but is rather to be found in 'a more discursive, allusive method, often leaning on stories of one sort and another' (Jubb, 2019: 8).We argue that those lines of defence are overly concessive: realists need not abandon political normativity, let alone exacting philosophical argumentation.We make that case in two independent and mutually supportive ways.
First, we show that the stakes in this debate are lower than Leader Maynard and Worsnip suggest: even if their argument were to succeed, it would not doom realism, but only a very specific version of it, because non-moral political normativity is not the same as distinctively political normativity -a simple distinction that, nonetheless, even some of the realists quoted above miss.Distinctively political normativity has been characterised in several ways.Arguably, some loose remarks in a paper by Enzo Rossi and Matt Sleat (2014) are at least partly responsible for the metanormative turn in the debate we address here.Political realism, they say, is defined 'on the basis of its attempt to give varying degrees of autonomy to politics as a sphere of human activity, in large part through its exploration of the sources of normativity appropriate for the political' (2014: 1). 1 This was taken by several commentators to suggest 'distinctness' in an ontological sense.But political normativity doesn't have to be its own ontological category.There are politically salient normativities, such as epistemic or prudential normativity, which are distinct from moral normativity without being constitutively or ontologically political.Even the strong versions of realism that seek to eliminate morality from political theory altogether do not need to posit a sui generis political normativity: epistemic and/ or prudential normativity are enough for their purposes. 2Such normativity may not be political in some ontological sense, but it is political in the salient sense that it can generate normative political judgements.
Second, we show that Leader Maynard and Worsnip's argument does not succeed on its own terms.They aim to proceed from a 'position of neutrality between realism and moralism' (Leader Maynard and Worsnip, 2018: 764), but we argue that they don't, because their argument for the redundancy of distinctively political normativity depends on unwarrantedly assuming moralism as the default position in the debate.We advance this argument while suspending judgement on the existence of distinctively political normativity, and so on whether versions of realism that rely on such a normativity constitute a plausible method alongside or in opposition to moralism.We merely argue that Leader Maynard and Worsnip do not succeed in ruling out distinctly political normativity.They do not succeed because moralists are not entitled to demand of realists that they produce the same kind of judgements about politics that an 'ethics first' approach produces, namely, judgements about duties, all-things-considered oughts and the like.Moral normativity should not be the gold standard for politically salient normativity.For instance, as we will see below, realist normative political judgements will often take an evaluative rather than prescriptive form.That is a feature, not a deficiency.It is a feature of a different way of doing political theory.And this difference is not a legitimate ground for complaint, at least if we want to place realism and moralism on a level playing field.
The upshot is that realism and moralism can be understood as alternative methodological stances, and not necessarily as diametrically opposed claims about the only or primary kind of normativity in political theory, pace what both some moralists and some realists claim.And so we do not need to adjudicate the controversy about the relative merits of realism and moralism.What we want to show is that realism has characteristic and viable ways to make normative judgements about politics without relying on moral normativity, that is, collapsing into moralism.In which case, the controversy between realism and moralism can be adjudicated by looking at each approach's achievements, and not through a debate on metanormativity. 3

Lowering the Stakes
Despite the title of their paper ('Is There a Distinctively Political Normativity?'),Leader Maynard and Worsnip oscillate between two characterizations of the realist position they attack.At times they indeed question whether political normativity is 'its own distinctive kind of normativity' (Leader Maynard and Worsnip, 2018: 759).But sometimes they conceive of the realists' arguments as just 'attempting to isolate a sense in which political normativity is distinct from moral normativity' (Leader Maynard and Worsnip, 2018: 764).We submit that, by conflating distinctively political normativity and non-moral political normativity, Leader Maynard and Worsnip misconstrue realism, even in its stronger versions, that is, versions that, roughly speaking, reject any role for morality in normative political theory.We should note, though, that Leader Maynard and Worsnip are not alone in this mistake: even some realists commit it. 4 This is how Leader Maynard and Worsnip (2018: 762) delineate the issue at stake: . . .if a collapse into merely verbal debate is to be avoided, this way of distinguishing different kinds of normativity must be able to distinguish the moralist view that political principles are a part of morality, albeit perhaps a distinctive part, from the realist view that political principles are of a different, nonmoral normative kind altogether.
It is clear from that quotation that they do not wish to consider the middle position of a non-moral politically salient normativity that is not its own normative kind -an understanding of political normativity that is compatible with it being reducible to any of the other normativities Leader Maynard and Worsnip (2018: 756) do acknowledge: 'epistemic normativity, prudential normativity, "aim-given" normativity, and aesthetic normativity'. 5Indeed, even if realism's aim is to eliminate moral judgement from normative political theory (an aim not all realists share), it's far from clear that it needs a sui generis political normativity, though realists may sometimes have been guilty of suggesting as much through poor terminological choices.Realism is less demanding: 'if realists can show that they can make normative political judgments that [. ..] don't share the sources of normativity of moral principles, then they will have made room for their view' (Rossi, 2019: 640).
The challenge for realists, then, is to show that it is possible to make normative judgments about politics by using non-moral (and not necessarily sui generis) normativity.We submit that epistemic, prudential and 'aim given' normativity are strong candidates for this role.This is not a claim that we have the space to defend in full here.But all we need to establish is that this claim would have made a better target for Leader Maynard and Worsnip's argument.And that is easy to see: removing moral normativity from normative political theory would be a sufficiently disruptive methodological development to warrant attention to realism.In fact, the critics of realism seem to acknowledge as much when they take their positive claim to be just that 'political justification is irreducibly moral' (Erman and Möller, 2015: 1) or that it has 'its ineliminable roots in morality' (Leader Maynard and Worsnip, 2018: 787). 6 What we need to establish, then, is just that there is a prima facie case for grounding at least some significant normative political judgements in epistemic or instrumental rationality. 7We can do so by pointing at the literature.First, it is worth noting that a recent article by Carlo Burelli and Chiara Destri (2021: 7) defends the role of those normativities within realism extensively and exhaustively: they clarify, inter alia, that Leader Maynard and Worsnip fail to consider the option of epistemic normativity altogether, and take instrumental normativity to be immaterial to the question at hand. 8Second, we can point to various contributions to the 'radical realist' current, some of whose exponents explicitly rely on epistemic normativity alone: the rough idea there is that empirical evidence can be mustered to show that some normative beliefs about political power structures are caused by those very structures, and so lack epistemic warrant due to vicious circularity (e.g.Aytac, 2022;Aytac and Rossi, forthcoming;Cross, 2022;Prinz and Rossi, 2017;Rossi, 2019;Rossi and Argenton, 2021).
Note, though, that epistemic normativity does not yield strict prescriptions.It is an evaluative normative theory, not a prescriptive normative theory (Aytac and Rossi, forthcoming;Rossi, 2019). 9A moralist may take this to be a disadvantage of realism vis-à-vis moralism.The realist need not concede this point.Focusing on evaluation rather than prescription is just a different modus operandi.Indeed, it is worth pointing out that the focus on prescriptions is peculiar feature of contemporary Anglophone political philosophy, whereas, for example, large swathes of the 'critical' or 'continental' traditions have no interest in or even reject the identification of obligations and such like as a task for political theory.So, there is a sense in which at least some strands of political realism break with some commitments of much Anglophone or 'analytic' political theory.That should not be seen as a weakness.It is just a way in which the mode of realism and moralism differ.

Levelling the Playing Field
Let us now move on to consider Leader Maynard and Worsnip's case against distinctively political normativity, rather than non-moral normativity.The general pattern of their argument is this.First, they introduce a set of premises accepted by the realist (P1) from which the realist infers that (C1) there is a distinctively political source of normativity.Second (P2), Leader Maynard and Worsnip profess that there is nothing keeping the moralist from also accepting the realist's premises (or something sufficiently similar), and then drawing moralist conclusions.Therefore, they conclude that (C3) the realist is not warranted to infer, given (P2), that (C1) is true.
For example, this is how Leader Maynard and Worsnip reconstruct a realist argument concerning the realist's distrust in the legitimacy of the unrestricted enforcement of universal moral claims: 10 (P1) Just because a moral principle is either true or reasonable it doesn't follow that it is legitimate to enforce it politically.In other (realist) words, even if A morally-ought to ϕ it doesn't follow that B politically-ought to force A to ϕ.
Leader Maynard and Worsnip note that this is an argument realists may make in light of Bernard William's (2005: 7) thesis that the 'Basic Legitimation Demand' requires a political power-holder to provide a legitimation story acceptable to those over whom the power-holder wishes to have authority, and that such legitimation stories float free of morality.Granting the prima facie plausibility of Williams' claim that his machinery involves a non-moral concept of acceptability, 11 the realist account of legitimacy escapes the confines of morality.Thus, (C1) political realists can claim to have found a non-moral yet normative political value, legitimacy, and thus to have unveiled a genuine political source of normativity. 12Leader Maynard and Worsnip then agree, qua moralists, with the realist's observation that (P2) Just because a moral principle is either true or reasonable it doesn't follow that it is legitimate to enforce it politically.In other (moralist) words, even if A morally-ought to ϕ it doesn't follow that B morally-ought to force A to ϕ.
But then, since the (in)acceptability of forcing others to abide by some true or reasonable moral principle 'can itself be moral in nature' (p.767), Leader Maynard and Worsnip contend that (C2) The normativity of (P2), the observation that sometimes a true and reasonable moral principle shouldn't be enforced on others, springs from a moral source of normativity.
And on that basis, they conclude that (C3) Distinctively political normativity is redundant.
C3 supposedly follows because moralists can assert P2.As Leader Maynard and Worsnip put it, it can be true that actor A morally ought to do action X, while also being true that actor B morally ought not to force actor A to do action X.Since this is evidently coherent, even reading both 'oughts' as moral, there is no need to introduce a distinctive political 'ought' to make sense of such a structure (Leader Maynard and Worsnip, 2018, emphasis added). 13  Yet we don't know what would have to be the case for political philosophers to have a need to introduce distinctive sources of normativity.And this debate is not about what political philosophers need, but about what they may plausibly do.It is about trying to establish the presence of a viable methodological approach.There is no inconsistency nor redundancy in developing a realist account from a set of premises that the moralist accepts.The fact that despite agreeing on the premises (P1 and P2) realists and moralists develop two different accounts just indicates that realism is another prima facie viable position.
If Leader Maynard and Worsnip didn't assume moralism as the default view, their claim (P2) that a moralist can also account for the realist's observations on legitimacy (P1) would lead not to the alleged refutation of the realist conclusion (C3), but to a stalemate between the realist and the moralist.Realism's critics aren't entitled to expect from non-moral normativity the same sorts of things one expects from moral normativity.Put differently, without assuming moralism as the default view and forbidding the proliferation of normativities, (P1) and (P2) allow for several different conclusions -which hardly undermines the case for distinctively political normativity.It just puts it on a par with moral normativity.We shouldn't ask whether political normativity can do what moral normativity does, but whether it enables us to do normative political theory in a different way.
Our point, then, is that the deep disciplinary embeddedness and conventionality of moral normativity can't be used to adjudicate this question.This is because assuming that political philosophy's core aim is to generate all-things-considered prescriptions is tantamount to stacking the deck in favour of moralism.That is to say, it is inappropriate to demand from non-moral normativity the same 'authority' or 'force' as prescriptive moral normativity (a point similar to P2, where it is argued that non-moral normativity has nothing to add over and above moral normative verdicts): 14 other kinds of normativity, and even some versions of moral normativity, simply do not aspire to make this kind of judgement. 15But it must not be.Likewise, it's not appropriate to require realists to show that non-moral normativity overrides moral normativity.'Overridingness' is a desideratum of certain accounts of moral normativity (Stroud, 1998), and so of their application to political philosophy, but there is no reason to take this as a yardstick to measure all ways of rendering normative judgements about politics.
Relatedly, saying that something can be accomplished with moral normativity is not enough to conclude that it cannot or should not also be accomplished with some other kind of normativity: saying that some other politically salient normativity is not needed because we already have moral normativity is not a way to show that arguments for political normativity are not convincing.There are many ways to make normative judgements about politics, possibly as many as there are distinct normativities, and we shouldn't assume that the moralist one is the gold standard.
Indeed, there may just not be a sufficiently neutral standpoint from which to adjudicate whether what we do with moral normativity is equally important or interesting as what we do with other kinds of normativity.So we are not claiming that Leader Maynard and Worsnip advance their argument in an avoidably biased way, but rather that this sort of methodological argument from need and redundancy is bound to be biased, and so should probably not be advanced.
One might reply on Leader Maynard and Worsnip's behalf that they are not trying to establish that there's no conceptual room for distinctively political normativity, but just that there is nothing in political philosophy that would call for such a view.But that depends on the prevailing yet contingent preferences of political philosophers -just what realism set out to change.So, using the current prevailing preference for moral normativity as an argument against distinctively political normativity begs the question of whether the latter is viable.
But doesn't explanatory parsimony caution against multiplying normativities (and methodological approaches) beyond necessity? 16If that is the case, then one may as well turn the tables and replace moral with political normativity, for example, by way of ideology critique: as Raymond Geuss (2010: 42) pithily put it, 'ethics is usually dead politics; the hand of a victor in a past conflict reaching out to extend its grip to the present and the future'.Whether ideology critics are right in their suspicion of morality's ubiquity or whether the moralists are right to take moral language at face value is just what is at stake here. 17But that cannot be taken as a neutral starting point, because it is precisely the position the realist attacks.
Given Leader Maynard and Worsnip's apt insistence on rendering the debate nontrivial, and on escaping a merely verbal dispute, we must assume that, at least in principle, all prima facie plausible sources of normativity should be acknowledged a methodological option, at least until we find any contrary evidence -evidence, that is, other than the contingent pre-eminence of moralism in contemporary political philosophy.We can only heed Leader Maynard and Worsnip's recommendation to not play 'games of burdenshifting tennis' (2018: 764) so long as they agree to play on a level field.A field, that is, where current methodological preferences are not a reason to rule out alternative methodologies that seek to change those preferences.
What we take away from this engagement with realism's critics, then, is that realism and moralism are just different but not necessarily incompatible approaches to political theory: they have different accounts of what should be the main focus of normative judgements about politics.Still, if we are right, moralism need not be the view that all apt normative political judgements are moral judgements, and realism need not be the view that no apt normative political judgements are moral judgements.And so we have not taken up and need not take up the question of whether and to what extent political normativity should replace moral normativity.Pace what many realists as well as many of their critics claim, realism doesn't need to unseat moralism to claim its seat at the table of normative political theory.One may even envisage a generative division of labour between moral and political normativity, or ways to balance their respective desiderata, though doing so would take us far beyond the scope of this article. 18We simply hope to have demonstrated the viability of the realist project of making normative political judgements not grounded in moral commitments.

Aknowledgements
The authors would like to thank Afsoun Afsahi, Shai Agmon, Ugur Aytac, Sam Bagg, Carlo Burelli, Ilaria Cozzaglio, Ben Cross, Cécile Fabre, Vincent Harting, Robert Jubb, Sam Kiss, Jonathan Leader-Maynard, David Leopold, Andrei Poama, Janosch Prinz, Paul Raekstad, Matt Sleat, Amia Srinivasan, Zoltán Gábor Szűcs, and the audiences at the Centre for Social Science of the Hungarian Academy, the Work in Progress Political Theory Seminar at Nuffield College, Oxford, the CEPPA Conference at the University of St Andrews, and the Oberseminar Praktische Philosophie at the University of Tübingen.

Declaration of conflicting interests
The author(s) declared the following potential conflicts of interest with respect to the research, authorship and/ or publication of this article: The authors have no conflicts of interest regarding the contents of this article, though they are scholarly invested in the ongoing debate it addresses.

Funding
Enzo Rossi's research was supported by the Dutch Research Council, as part of his Vidi project "Legitimacy Beyond Consent" (grant n 016.164.351).Adrian Kreutz's research was supported by a DPIR Oxford Doctoral Studentship.