Fact-Centric Political Theory,Three Ways: Normative Behaviourism,Grounded Normative Theory,and Radical Realism

Abstract

In the last two decades, Anglophone political theory witnessed a renewed interest in social-scientific empirical findings – partly as a reaction against normative theorising centred on the formulation of abstract, intuition-driven moral principles. This brief article begins by showing how this turn has taken two distinct forms: (1) a nonideal theoretical orientation, which seeks to balance the emphasis on moral principles with feasibility and urgency considerations, and (2) a fact-centric orientation, which seeks to ground normative conclusions in empirical results. The core of the article then compares and contrasts three variants of fact-centric political theory: normative behaviourism, grounded normative theory and radical realism. The upshot: normative behaviourism achieves focus on observable behaviour at the cost of status quo bias, grounded normative theory achieves radicalism at the cost of endorsing an activist orientation to theorising and radical realism combines a non-activist orientation with the potential for far-reaching critique of the status quo.

Introduction

If there is one overarching trend in Anglophone normative political theory in the twenty-first century, it is the turn away from armchair philosophising, and towards approaches that give greater role to empirical descriptions of the social and political world. In this article, I compare and contrast different ways in which the facts can come to play an important role in political philosophy.

Some readers may notice a conspicuous absence in the article’s subtitle, namely nonideal theory. By way of introduction, let me clarify my focus by spelling out the selection criteria for the three approaches I will discuss: normative behaviourism, grounded normative theory and radical realism. I submit that those three approaches share the feature of being fact-centric, or empirically grounded, whereas nonideal theory is fact-sensitive, or empirically informed. ‘Fact-centric’ political theories draw normative – prescriptive or evaluative – conclusions by pointing out the normatively salient features of an empirical description of a state of affairs. Jonathan Floyd’s normative behaviourism does this by establishing a link between certain observable behavioural pattern and the justifiability of political arrangements. Crudely, low levels of crime and insurrection indicate legitimacy (Floyd, 2017: 168ff). Grounded normative theory, on the other hand, uses empirical observations to uncover some ‘insights, claims, interests, and actors’ that will feature in normative argumentation (Ackerly et al., 2021: 5). Finally, radical realism seeks to empirically uncover self-justifying power structures, and then criticises them for their deleterious epistemic effects, that is, for how they distort our perception of social reality and, therefore, our capacity to make good political decisions. ¹ What the three approaches have in common, then, is the centring of empirics, in the sense that those three approaches couldn’t make the distinctive sorts of claims they make without empirical input: roughly, claims about observable preferences (normative behaviourism), previously invisible perspectives (grounded normative theory) or epistemic distortions (radical realism).

Whereas empirics play a different role in nonideal theory: again roughly, they modulate pre-existing abstract moral commitments – for example, to justice, democracy, equality – and their application by adding feasibility constraints, considerations about motivation, claims about the relative priority of various moral ills and the like. That is the sense in which I maintain that nonideal theory is fact-sensitive rather than fact-centric, and so outside of the scope of this article. To be sure, this sort of characterisation of an approach is by necessity somewhat stipulative. Some approaches that self-identify as nonideal theory may resemble grounded normative theory more than nonideal theory as characterised here, say. But I think the general distinction holds: even for the theorists who maintain that nonideal theory does not require ideal theory, it is still the case that the main task of nonideal theory is to evaluate states of affairs or courses of action on the basis of pre-existing moral commitments, much as in ideal theory, including ideal theory of the ‘fact-free’ variety (Cohen, 2003): ‘. . . our reasoning to a set of directive political principles for our nonideal world has two equally basic inputs: a set of moral evaluative criteria and a specification of the feasible set’ (Wiens, 2015: 445). The only important difference is the addition of certain empirical constraints – what I’ve called fact-sensitivity rather than fact-centricity.

Having set nonideal theory aside, in the next section I turn to the first of the three fact-centric approaches, namely normative behaviourism. The next two sections are devoted to grounded normative theory and radical realism, respectively. I will argue that normative behaviourism achieves focus on observable behaviour at the cost of status quo bias, grounded normative theory achieves radicalism at the cost of endorsing an activist orientation to theorising and radical realism combines a non-activist orientation with the potential for far-reaching critique of the status quo.

Normative Behaviourism

Before formulating my two objections, it is worth clarifying which parts of normative behaviourism I do not wish to attack, at least for the sake of the present argument. I will not question Floyd’s empirical evidence for his claim that crime and insurrection are lowest under social-liberal-democracy, nor the related and more controversial claims that ‘illiberal, undemocratic systems tend towards collapse’ and ‘once prosperity reaches a certain point, democracy is soon demanded’ (ibid: 184).

The first aspect of Floyd’s view I do wish to contest is, rather, the inference from the lack of crime and insurrection to the claim that ‘the most appreciated political systems in both past and present political life are those that I am calling social-liberal-democracies’ (ibid: 197). I want to focus on insurrection, partly because crime may be somewhat endogenous: by definition, a liberal regime will criminalise fewer behaviours than, say, a theocracy. ² Floyd says that ‘social-liberal-democracies are, simply put, the people’s choice. They are their choice in the very basic sense that people find them more bearable than any attempted type of regime’ (ibid: 184, emphasis added). I want to put pressure on the inference from choice to bearability. Even setting aside issues to do with straightforward false consciousness, the possibility remains that compliance is due to some aspect of social-liberal-democracy being particularly effective at hindering the cooperation and coordination necessary to rebel (Rosen, 1996: 160ff). I leave open whether this (hypothetical) effectiveness is due to something akin to Chomsky’s ‘manufactured consent’, Marx’s ‘mute compulsion of economic relations’ or social-liberal-democracy being a golden cage – even the latter strikes me as rather worrisome in terms of our ability to radically improve our political predicament.

In more recent work does Floyd (2020: 145) address a similar worry: ‘[normative behaviourism] has to distinguish, as realism does, between genuine acceptance and ideological or coerced acceptance’.. His solution is that behaviour is indicative of real acceptance irrespective of belief. I think this addresses worries about false or unjustified beliefs, but not about coordination problems. Indeed, somewhat ironically, when discussing the relationship between rational choice theory and normative behaviourism, Floyd (2017: 216) says that social-liberal-democracies ‘disincentivise non-cooperative forms of behaviour such as insurrection’. Whereas this only shows that they incentivise one form of cooperation (compliance) rather than another one (rebellion). Indeed, some Marxists argue that liberal (i.e. capitalist) social relations are stable precisely because they create a strategic imbalance between capital and labour (Chibber, 2022; Cicerchia, 2021). But there is a silver lining here: whether the lack of insurrection is motivated by resignation and/or ideology is an empirical question, which should help us advance the project of fact-centric political theory. Indeed, my discussion of radical realism in the last section will start from this very issue.

My second concern also has to do with a potential status quo bias in normative behaviourism. This is an issue Floyd anticipates: ‘Is this order not too static? We would never have achieved social-liberal-democracy without complaint and creativity, so are we now never to improve?’ (ibid: 244). His reply is that ‘social-liberal-democracies are themselves always trying to improve’ and that ‘a crucial part of the explanation of the contentment generated by that system is the fact that it is always open to being significantly altered by means of law-abiding, peaceful political campaigns’ (ibid: 244–245). My objection is basic: it is just not clear why our prospects for improvement should be limited to the framework for social and political experimentation provided by social-liberal-democracy. In his discussion of this sort of objection, there is, in fact, a telling shift from matters of politics to matters of policy (ibid: 246–247). I am unsure whether Floyd thinks we have reached the end of history, or a zenith of sorts, or whether he thinks there are prudential considerations that count against more adventurous experimentation. None of those options strike me as compelling – a concern that is made particularly salient by my previous worry about what may be lurking beneath the relatively calm waters of social-liberal-democracy.

Grounded Normative Theory

Grounded normative theory tends to look more sympathetically upon those who are dissatisfied with the status quo – even the status quo of social-liberal-democracy. Of the three approaches under consideration, grounded normative theory is the oldest and most varied. I will follow Ackerly and colleagues’ recent systematisation of it (2021), and take a recent argument by Ackerly (2018) as an exemplar. Ackerly (2018: 1) is clear about the fact-centricity of her approach: ‘If we want to learn what to do about injustice and how to do it, we should learn from those who are doing something about it’. In a nutshell, the idea is to figure out what justice requires and how to achieve it by empirically studying social movements devoted to its promotion. And so, for example, Ackerly (2018: 147) develops her ‘feminist critical methodology grounded in Third World feminist social criticism’ in a rich qualitative investigation of the practices of feminist social movements: she formulates fact-centric principles of justice by ‘thinking what they are doing’ (Ackerly, 2018: 134).

There is no doubt that this approach can uncover a range of issues and perspectives that would otherwise be invisible to political theory: the texture of qualitative data brings to light normatively salient nuances that are normally lost in the vignettes of standard political philosophy (e.g. ethnography can reveal ‘normalization’: ways in which injustice is compounded by habituation to unjust circumstances; Ackerly, 2018: 9). Furthermore, it can provide new grounds on which to formulate normative principles and so, presumably, lead to the formulation of ‘principles-in-practice’ (Ackerly, 2018: 192–193) that might otherwise never have seen the light of day, or at least would have to be justified. And, given that grounded normative theory enables us (though arguably does not require) to focus on the perspective of activist groups and other political agents, the worry about status quo bias I raised in relation to normative behaviourism evaporates.

Yet I want to round off this brief discussion of grounded normative theory by pointing out that this particular way of achieving radical potential comes at a significant cost, namely that of giving up the understanding of political theory as a kind of Wissenschaft: an activity whose main criterion of success is the approximation to the truth, or epistemic justification. This is because, on Ackerly’s approach, the normativity of the principles of justice one uncovers and formulates comes from the justifiability of the goals and tactics of the social movement one studies empirically. If one doesn’t share the politics of those movements, the principles-in-practice will lack normative purchase. To her credit, Ackerly makes no secret of this: she espouses what is effectively an activist orientation to political theorising. Again, there is something to be said for the epistemic affordances of such an approach. But one doesn’t need to hold a positivist account of the fact-value distinction to see its significant drawbacks vis-á-vis a more wissenschaftlich orientation to theorising. For one thing, an activist orientation contributes to entrenching the already growing public distrust towards the academic community (Yancey, 2018). This sort of problem should bother even those who are committed to a form of theorising oriented towards political goal, for it undermines the potential public reach of their own scholarship – or at least it trades depth for breadth.

My second worry is more philosophical, and more in keeping with the aspirations of a scholarly rather than activist orientation to political theorising. Quite simply, if normativity is to be read off the praxis of social movements, it is not clear on what grounds we should choose one social movement over another. ³ This is not to doubt the goodness of the cause of Third World feminist activists, say. But the problem remains that there are significant tensions even within those movements (e.g. liberal vs socialist vs Islamic feminists). Not to mention the fact that movements themselves can be prone to ideological distortions that reflect society’s wider power relations. More fundamentally still, it is not clear how taking political sides (beyond the inevitable Weberian issue of how we choose our questions) is supposed to solve what critical theorists usually call ‘the problem of normative foundations’ – the problem of what can make our normative claims true or epistemically justified – rather than simply bypass it.

Radical Realism

The preceding discussion of normative behaviourism and grounded normative theory provided us with two desiderata for fact-centric political theory: radical potential and Wissenschaftlichkeit. This final section provides a brief sketch of radical realism, to show how it can meet those desiderata.

Radical realism, at least in the variant I wish to focus on here (Aytac and Rossi, 2022; Rossi, 2019; Rossi Forthcoming, 2024; Rossi and Argenton, 2021), is a form of ideology critique grounded in epistemic rather than moral normativity. The rough idea is to critique the epistemic consequences of observable patterns of power self-justification. Consider a toy example. In a patriarchal society, people tend to believe that ‘father knows best’, and so comply with the power of senior males. But, as it turns out, this belief is due to paternal inculcation – an instance of power self-justification. One may see this is a moral problem, but radical realists highlight its epistemic dimension, and rely on that exclusively. Very roughly, the epistemic problem here is that beliefs and other cultural elements that result from power self-justification are not good sources of knowledge about society, and so put us in a suboptimal position to make choices about how to organise society. ⁴ Power self-justification allows the powerful to be judges in their own affairs, and judges in their own affairs aren’t likely to reach sufficiently epistemically accurate verdicts. This in turn yields a general epistemic case against social hierarchies: social groups with significantly more power than others are in a position to fog society’s cognitive windscreen in a way that further entrenches their position. ⁵ The aim of radical realist social analysis is to empirically uncover these mechanisms of power self-justification, to criticise the beliefs and attitudes they generate, and so to contest the social practices and institutions they underpin.

How does radical realism fare vis-á-vis the two desiderata we identified? The radical potential should be clear: mechanisms of power self-justification are ubiquitous in society, and even though radical realism refrains from making prescriptions, it seems that it pro tanto counsels rather flat social structures, for the purposes of minimising the ways in which power distort our ability to think clearly about how to organise society. The Wissenschaft issue is more complex, but I submit that there is reason to think radical realism fares better than grounded normative theory, and at least as well as normative behaviourism. This is because radical realism does not need to align itself with political causes, nor even to rely on moral normativity. Its only commitment is epistemic: to improve our grasp of social reality. As such, the enterprise can be considered internal to the practice of social science itself. Not that this practice could or should ever be considered ‘neutral’ – even the staunchest Weberians recognise the normative presuppositions that guide their choice of research focus. The point is just that such an approach is not activist or partisan.

To be sure, that was a hasty sketch of a terrain that is much more variegated than the preceding discussion might suggest. I expect that proponents of both normative behaviourism and grounded normative theory will have a number of promising replies at their disposal. I offer this note as a starting point for further debate, with a view to jointly growing the broad church of fact-centric political theory.