Art of the Possible?
Feasibility and Compliance in Ideal and Nonideal Theory
Naima Chahboun
Naima Chahboun Art of the Possible?
Stockholm Studies in Politics 191
Doctoral Thesis in Political Science at Stockholm University, Sweden 2020
Department of Political Science
ISBN 978-91-7911-282-0 ISSN 0346-6620 Naima Chahboun
This dissertation engages with the methodological debate concerning ideal theoretical approaches in normative political theory. In four free- standing articles, I examine when and why noncompliance due to agents' motivational limitations constrains what justice can demand, and how this affects the status and usefulness of ideal theory. My findings shed new light over the debate's underlying conflict lines, and urge debaters to render explicit and argue for the assumptions upon which they rest their judgments about ideal theory.
Art of the Possible?
Feasibility and Compliance in Ideal and Nonideal Theory
Naima Chahboun
Academic dissertation for the Degree of Doctor of Philosophy in Political Science at Stockholm University to be publicly defended on Friday 27 November 2020 at 10.00 in Aula Magna, höger hörsal, Frescativägen 6.
Abstract
In the past decade, the value of so-called ideal theory has become a major point of dispute among political theorists. While critics of ideal theory accuse this approach of “idle utopianism”, its advocates fault the critics for conceding to “cynical realism”.
This dissertation examines two charges against ideal theory. The demandingness charge states that ideal theory fails to acknowledge the constraints on justice set by the empirical conditions that prevail in our world, and that it therefore produces invalid principles. The uselessness charge states that ideal theory, even if it tells us what justice would require under exceptionally favorable circumstances, offers no information valuable for guiding action in the nonideal circumstances characteristic of today’s societies. The two charges target the idealized assumptions made in ideal theory, in particular the assumption of full compliance. By assuming full compliance, the critics argue, ideal theory ignores the way real-world agents’ motivational limitations render the pursuit of its proposed principles infeasible or undesirable.
In four free-standing articles, I examine when and why noncompliance due to motivational limitations puts constraints on justice, and how this affects the status and usefulness of ideal theory. I argue that motivational limitations constrain justice in ideal theory if we hold that justice is action-guiding in the sense that it confers actual duties on individual agents, and that the distribution of collective duties to individuals requires reasonable expectations of others’ compliance. In nonideal theory, adopting an actualist standpoint will lead us to conclude that not only the noncompliance of others, but also our own foreseeable noncompliance constrains what justice can demand. I further argue that how this affects the usefulness of ideal theory depends, on the one hand, on how we interpret crucial concepts such as “action-guidance”, and, on the other, on which task we expect political theory to perform. My findings shed new light over the complex conflict lines that underlie the current dispute, and urge debaters to render explicit and argue for the assumptions upon which they rest their judgments about ideal theory.
Keywords: Ideal theory, nonideal theory, feasibility, compliance, justice.
Stockholm 2020
http://urn.kb.se/resolve?urn=urn:nbn:se:su:diva-185748
ISBN 978-91-7911-282-0 ISBN 978-91-7911-283-7 ISSN 0346-6620
Department of Political Science
Stockholm University, 106 91 Stockholm
ART OF THE POSSIBLE?
Naima Chahboun
Art of the Possible?
Feasibility and Compliance in Ideal and Nonideal Theory
Naima Chahboun
©Naima Chahboun, Stockholm University 2020 ISBN print 978-91-7911-282-0
ISBN PDF 978-91-7911-283-7 ISSN 0346-6620
Cover image: 'The Uncomfortable Watering Can' ©Katerina Kamprani 2020 Printed in Sweden by Universitetsservice US-AB, Stockholm 2020
For things will never be perfect, until human beings are perfect - which I don't expect them to be for quite a number of years!
Thomas More, Utopia
Acknowledgements
Short before entering the PhD program, I met with a professor who told me that his years as a PhD student had been the best ones of his life. And though I hope that the best years of my life are yet to come, I can honestly say that being a PhD student has (mostly) been a truly enjoyable experience. To a large degree, this is thanks to the great number of people who in different ways have contributed to the completion of this dissertation, and to whom I here wish to express my gratitude.
My first thanks go to my supervisors, Ludvig Beckman, Alexandra Segerberg, and Eva Erman. Ludvig, you were
(together with Johan Tralau) the one who once upon a time first introduced me to the magical world of political theory – a world to which I have kept coming back ever since. You have supervised my Candidate and Master theses, and now also my dissertation. And throughout the years, you have kept that combination of rigorous thought and curious mind that so deeply impressed me during those undergraduate seminars in Badhuset in Uppsala. Thank you for keeping up with me all this time, and for the many laughs we shared along the way!
Alexandra, I am so glad to have had you by my side during these years. You have in many ways been the ideal supervisor;
dedicated, careful and attentive. The thoroughness with which you approach every task – from outlining a whole dissertation to proof reading an application – has been priceless. Though I will never live up to your high working standards, you will always serve as a role model for me. Thank you for all the efforts you have put into this dissertation, and for your continuous support!
Eva, you came on board when this project was already running, and gave the dissertation a well-needed boost at the crucial, final stages. Your endless energy and enthusiasm kept my moods up, while your expert knowledge of the field helped me bring the dissertation to completion. Thank you for joining me on this adventure!
I also want to thank the directors of studies, Kristina Boréus and Hans Agné, for patiently guiding me through my first confusing time as a PhD student, and for showing confidence in my capacity to complete this project at a point in time when most things spoke against it. Hans, special thanks for your detailed comments on my thesis manuscript.
My roommate Livia Johannesson shared for many years not only spaces but also everyday experience with me. To see you smile from the other side of the desk made it worthwhile coming to the office even on dark November days, and getting a close view of your academic practice served as a great
inspiration to me. Thank you for your support and for your friendship!
Martin Westergren has not only been my most committed pep-talker and sparring partner, he also travelled around Europe attending numerous graduate courses and conferences with me. To hear you knock at my door, knowing that I was about to be introduced to some new, intriguing problem, has been a welcome disruption to my daily work. Thank you for sharing your thoughts and your time with me!
I am deeply grateful to my fellow PhD students in the Political Theory working seminar, Max Fonseca, Jasmina Nedevska, Markus Furendal, Carolina Janson and Viktor Elm- Schulin. Without the comments and criticism you have offered on various occasions – informal as well as formal – this
dissertation would not have been what it is today. And though your knowledge and intelligence often made me display the vice of envy, your comradeship has rendered this time less lonely, and a lot more fun. Max, special thanks to you for your generous support and advice, on teaching and beyond. You are – despite your choice of sport – a true team player!
Some people made specific contributions at various stages of my dissertation work. In particular, I want to thank Niklas Möller and Jonas Hultin Rosenberg, who on different occasions commented on my manuscript. Thank you for lending me your sharp-sighted eyes! I also owe special thanks to Professor David Estlund, who sponsored my stay as a visiting scholar at Brown University in spring 2015. The semester I spent at Brown was one of the highlights of these years, and the supervision that Dave on several occasions generously offered made a great impact on my thesis.
During my PhD program, I have been lucky to participate in
a number of international exchanges, courses and conferences.
This was made possible through generous financial support from Helge Ax:son Johnsons stiftelse, Elisabeth och Herman Rhodins minnesfond, Widar Bagges stiftelse, and Stiftelsen Siamon. I cannot tell how important the input from these exchanges have been for my dissertation work, and for my intellectual development at large. Thank you!
Over the years, many people have passed through the corridors on level 5 in the F-building. The current pandemic has made it painfully clear how much our chats in the lunchroom and by the copy machine have meant to me.
Thanks, everyone! I also want to thank the department’s administrators, in particular Lena Helldner, Anneli Lindén, Pernilla Nordahl, Emma Bergström, Christian Möllerop, and Schauki Karim for taking care of all the practical stuff, allowing me to devote my time to theory.
Architect Katerina Kamprani kindly allowed me to use her Uncomfortable Watering Can as the cover image for this book.
Its impossible design brilliantly captures the conflict between theoretical and practical purposes, and offers a striking
illustration of the critics’ view of ideal theory as a self-enclosed system. Thank you, Katerina!
My final thanks go to my family and friends. Without the love, jokes, babysitting, dinners, and wine you have provided during these years, I doubt that I would ever have come this far.
I want to thank my mother Elise for always encouraging me to follow my curiosity, even when this has led me in directions other than those she might have preferred. My life partner Karl-Johan has offered continuous support and, in particular at the final stages of writing, taken on a lot more than his fair share of household tasks. Thank you for always being there, and for not once asking me when I was actually going to finish!
Many thanks also to my children, Ezra and Abel, for being a constant reminder that writing a dissertation is neither the most important nor the most challenging task one may face.
Finally, I want to thank baby X, whose upcoming birth set a non-negotiable deadline for my work.
Contents
Introduction
I Three Feasibility Constraints on the Concept of Justice II Motivation as Constraint
III Nonideal Theory and Compliance – A Clarification
IV Ideal Theory and Action-Guidance: Why We Still
Disagree
2
3
Introduction
Politics has a dual nature. On the one hand, its aim is to change the world we live in, to turn the dreams of today into tomorrow’s reality. By setting up vi- sions and working to realize them, political agents push the limits of what life and society could be like. On the other hand, politics is, in Bismarck’s famous words, the art of the possible. Seen in the light of the 20th century’s totalitarian experiments, this is not merely a statement of facts, but a warning: If allowed to guide political practice, visions of the ideal society that fail to cater to ex- isting realities may be both unhelpful and dangerous. Since the costs of futilely trying to implement utopian ideals are often high, political agents must base their agendas on careful estimations of what it is possible to achieve in the conditions where they find themselves. While inherently visionary, politics is also constrained by what is attainable in the real world.
What does the dual nature of politics imply for normative political theory?
Is this, too, an “art of the possible”? We generally turn to political theory for visions, or ideals, for society. Its aim is not to tell us what society is like, but what it ought to be like. To complete this task, it is necessary to look beyond the unjust conditions that prevail in today’s societies. While real-world poli- tics often take place against a background of ignorance, greed, or unequal power relations, a political theory that catered to those conditions would con- cede to injustice and incorporate an unjustifiable bias toward status quo. Ra- ther than taking them as givens, we want the ideals produced by political the- ory to condemn the flaws of individuals and societies as moral failures. Nev- ertheless, a theory that disregards every imperfection of the world cannot claim to offer ideals for society, only utopian fiction. If our theories presup- pose that resources are abundant and infinite, or that people are fully informed and perfectly rational, they will arguably fail to offer the kind of guidance we seek from political theory. When asking what the ideal society would be like, theorists must consider only proposals that are practicable in our world, given the empirical conditions that prevail here.
This idea is well captured by the concept of “realistic utopia”, originally coined by John Rawls. In his ground-breaking work A Theory of Justice (1971), Rawls sets out to identify the principles that govern the fully just so- ciety. His account strives to probe the limits of practicable political possibility by disregarding contingent obstacles that may hinder the realization of justice temporarily, but that could, and should, be overcome (Rawls 2001: 4). In par- ticular, Rawls’ theory brackets problems that arise due to noncompliance with
4
justice, operating on an assumption of full compliance. In his view, ideal or strict compliance theory of this kind provides the only basis for a systematic grasp of the more pressing problems that arise in the real, nonideal world (Rawls 1999: 8).
Rawls’ theory has dominated the field of political theory since the 1970s, and the methodological framework he developed arguably constitutes his most important contribution (Floyd 2017: 367). Yet, in the past fifteen years or so, a growing number of theorists have expressed discontent with what they take to be inherent flaws of the ideal theoretical approach. Since ideal theory oper- ates on idealized assumptions, such as the assumption of full compliance, crit- ics argue that it ignores the ways in which real-world conditions may render an ideal unrealizable or undesirable. In particular, they worry that ideal theo- rists fail to recognize the motivational limitations displayed by real-world peo- ple, and the effects of noncompliance on their principles’ applicability (see e.g. Mills 2005, Sen 2006, Farrelly 2007, Schmidtz 2008, Philp 2010, Hall 2013, Wiens 2015, Galston 2016).
This dissertation examines the claim that ideal theory fails to offer norma- tive standards that are applicable in real-world conditions. I will distinguish between two overarching charges against ideal theory. The first, which I call the demandingness charge, states that the principles produced by ideal theory do not qualify as principles of justice for our world as we know it. Since ideal theory fails to acknowledge the way normative ideals are constrained by the factual conditions – including that of people’s limited moral motivations – that prevail in this world, it leads us to affirm conceptions of justice that are unre- alistically demanding and hence invalid. While ideal theory strives to probe the limits of the realistically practicable through assuming favorable back- ground conditions, critics claim that it adopts an unrealistic account of what the best realistically foreseeable conditions are (Farrelly 2007: 853). Ideal the- ory ignores the motivational limitations real-world agents display and, hence, produces ideals that are not valid for human societies, only for one populated by angels (Hall 2013: 178, Galston 2016: 248, Laurence 2018: 440, see also Wiens 2016b: 333). Sometimes, this charge goes together with the conceptual claim that the standards of ideal theory are not standards of justice, since jus- tice is, in Hume’s words, “the cautious, jealous virtue”. From this viewpoint, the demand for justice arises only due to the fallibility of human nature, and hence a theory that assumes impeccable abidance by highly demanding rules circumvents the very problem a theory of justice should respond to (Miller 2013: 25-6, Hall 2016: 93, Levy 2016: 317-8).
The second charge, which I call the uselessness charge, challenges the view that ideal theory is a prerequisite for nonideal theory. Ideal theory, the critics argue, is neither necessary nor sufficient for guiding action in the nonideal conditions that prevail in today’s societies. Even if the principles produced by ideal theory are practicable under favorable yet realistic conditions, then, knowing what justice would demand under such conditions tells us nothing about what it demands here and now. Insofar as we seek guidance for solving
5
actual, nonideal problems, ideal theory is useless or worse than useless (Sen 2006: 220, Stemplowska and Swift 2012: 377-8). This objection challenges Rawls’ view that ideal theory is a necessary prerequisite for nonideal theory.
Rather than deriving nonideal standards from ideal ones, critics argue that the former can, or even must, be developed independently of ideal theory, starting from precisely those empirical contingencies that ideal theory tends to bracket.
Some theorists advance a redundancy charge against ideal theory, stating that whatever information can be gained through this approach is readily available through other measures (Sen 2006: 221, Wiens 2015: 441). Others go even further, arguing that ideal theory is not only redundant but also misleading, since it obscures and thereby contributes to reproducing existing injustices (Mills 2005: 172).
In response to these charges, ideal theory’s advocates have argued that its critics fail to grasp that its purpose is not to offer recommendations for action here and now, but to identify aims that are desirable even if not immediately accessible and that we should strive to realize in the long run (Stemplowska 2008: 332). Although the critique highlights potential problems associated with the ideal theoretical approach, this gives us no reason to abandon ideal theory as such, but rather to scrutinize the assumptions made in specific ideal theories (Valentini 2009: 348-9).
The disagreement between advocates and critics of ideal theory reveals a tension at the heart of the “realistically utopian” approach. On the one hand, when proposing principles of justice, political theory should not bend to peo- ple’s weakness of will or immoral inclinations. A theory that takes every fail- ure to comply to be requirement-blocking is hopelessly biased towards status quo and will be rightfully accused of “cynical realism” (Valentini 2009: 339, Gilabert 2012: 50, Estlund 2016: 355). On the other hand, a theory that com- pletely ignores the fallible nature of human beings will not produce ideals that are practicable in the real world. If anything falls into the category of “idle utopianism”, it must be a theory with which we have reason to believe that no human will ever comply (Simmons 2010: 8, Galston 2016: 239, Southwood 2016: 8, but see Estlund 2011: 220). Although we generally hold instances of noncompliance to constitute moral failures, then, it would seem as if we some- times have reason to treat foreseeable noncompliance as part of the “back- ground conditions” to which a realistically utopian theory must cater. This gives rise to the following puzzle: What makes it the case that noncompliance sometimes invalidates a proposed ideal, while sometimes constituting a justice violation that warrants blame and calls for rectification? Why, that is, do cer- tain kinds of motivational limitations put constraints on justice, while others are dismissed as mere weakness of will?
Aim and research questions
The aim of this dissertation is to shed light over the disagreement between advocates and critics of ideal theory. In particular, I examine how diverging
6
judgments about ideal theory’s status and action-guiding capacity reflect dif- ferent perceptions of how noncompliance due to motivational limitations con- strains what justice can demand. While one part of my work consists in the identification and mapping of various positions that we may take vis-à-vis this issue, an equally important part consists of the critical assessment of their jus- tificatory grounds. The following research questions will inform my study:
1. When, and why, do motivational limitations put constraints on what justice can demand, and what does this imply for ideal theory?
2. Which kinds of noncompliance give rise to “nonideal
circumstances”, and what does it mean for ideal theory to be action-guiding in such circumstances?
The first question relates back to the demandingness charge outlined above.
Given that realistically utopian political theory must cater to motivational lim- itations of some sort, we need to know more about when such limitations are requirement-blocking, and when they are simply expressions of “bad will”.
With regard to ideal theory, we must know to what extent it fails to recognize the relevant motivational limitations. The second question relates to the use- lessness charge. In order to determine whether ideal theory fails to be action- guiding in nonideal circumstances, we need to know both which circum- stances these are and what it means for a theory to be action-guiding. In par- ticular, we need an explanation for why noncompliance with ideal principles would deprive those principles of their action-guiding capacity.
Answering these questions constitutes a first step toward settling the issue of whether ideal theory is a worthwhile intellectual enterprise. Only when we see which assumptions and commitments lead us to affirm or reject ideal the- ory can we fully realize what is at stake in the current debate. Rather than arguing for a position, my intention here is to offer careful reconstructions and assessments of the various options at hand. The debate about ideal theory has largely been conducted in polemical terms, with authors categorically dismiss- ing each other’s views (see e.g. Mills 2005, Geuss 2008, Rosenberg 2016).
The backside of this is that the underlying assumptions often remain unartic- ulated: What one side takes for granted, the other dismisses as not worth con- sidering. By conducting an open-ended inquiry, aimed at identifying the cru- cial conflict lines rather than defending any particular view, my thesis unpacks the arguments of both sides and highlights the key points of dispute between them. As will be shown, advocates and critics of ideal theory tend to rest their arguments on different understandings of what crucial concepts, such as fea- sibility and action-guidance, entail. Further, they tend to attribute different weight to the idealizations made in ideal theory, and to the risks these give rise to. I suggest that their divergent outlooks reflect different views of what the tension between individual and collective action implies for political theory, as well as different perceptions of the role of normative principles in generat- ing and explaining our moral judgments. Uncovering the many conflict lines
7
that intersect in the disagreement between advocates and critics of ideal theory brings clarity to the ongoing debate and constitutes a necessary step towards solving the intriguing and highly relevant problems at its core.
This introductory chapter outlines the starting points for my work and pre- pares the reader for the four articles where I present my findings. In the re- mainder of this section, I situate my study in relation to other parts of the broader discussion on “realism” versus “idealism” in political theory. The fol- lowing section develops a working definition of ideal theory and discusses this concept in relation to Rawls’ theory and beyond. After that, I introduce the two charges that constitute my study’s main focus – the demandingness charge and the uselessness charge – and explain how each of my four articles relates to them. The subsequent section offers brief summaries of the articles. In the final section, I present my dissertation’s main conclusions and contributions.
Delimitation of focus
The current disagreement between advocates of more “realist” and more “ide- alist” approaches to political theory has several historical predecessors.
Among the most famous we find Machiavelli’s critique of moralized accounts of authority in The Prince (1532), as well as modern classics, such as Karl Popper’s critique of political utopianism in The Open Society and Its Enemies (1945), and Hans Morgenthau’s realist account of international relations in Politics Among Nations (1948). The publication of Amartya Sen’s influential article “What Do We Want from a Theory of Justice?” (2006) is often seen as the starting point for the contemporary debate. While discontent with the Rawlsian approach predates Sen’s critique (see e.g. Honig 1993, Waldron 1999, Mouffe 2005), his article redirected and narrowed down the debate, highlighting the methodological issues at its core (Jubb 2016: 393). In partic- ular, Sen’s claim that identifying the characteristics of the fully just society is neither necessary nor sufficient for finding out what justice demands here and now sparked a debate that focused specifically on the practicability of the prin- ciples generated by ideal theory. The resulting critique has come in many dif- ferent shapes, only some of which will be under scrutiny here. To delimit the scope of my study, I will distinguish between the criticism of concern and other objections that have been raised within the various debates conducted under the label ideal versus nonideal theory. The positions outlined here are not mutually exclusive, and authors may consistently advance claims belong- ing to several categories. Rather than a categorization of authors’ positions, the distinctions I offer aim to identify different charges against ideal theory.
First, I will distinguish between the philosophical critique of ideal theory and the claim that political theorists should prioritize the more urgent prob- lems of nonideal theory to this project. Given the grave injustices that pertain in the world, some critics argue that researchers ought to spend their limited time and resources solving nonideal problems rather than searching for per-
8
fection or engaging in abstract discussions about the relative strength of vari- ous ideal standards of justice, none of which will come anywhere close to re- alization in the foreseeable future. Sometimes, the proposed research agenda is accompanied by the claim that scholars are required, as a matter of profes- sional ethics, to abstain from ideal theorizing (Stears 2005, Hall 2015, Frazer 2016). This critique may be both forceful and pertinent. However, rather than mounting a principled objection against ideal theory, it constitutes an appeal to political theorists to go out and get their hands dirty. While problems of professional prioritizations are relevant in their own right, my present concern is with the philosophical issue of what justice can demand. In my forthcoming discussions, I will therefore set this kind of objections aside.
Second, I will distinguish between the specific methodological critique that arises from problems of practicability and the more thoroughgoing objection that ideal theory asks the wrong kind of questions and therefore ends up with the wrong kind of answers. The latter objection is advanced by what we may call “hard-core” realists, who challenge not only the methods employed in ideal theory, but also its aim to identify principles of justice that are to govern society. Realists of this kind hold ideal theory to confuse political theory proper with “applied ethics”. They urge political theorists to abandon the quest for justice and instead look for a normativity inherent to the political domain of conflict, force and power struggle (see Williams 2005, Geuss 2008, Galston 2010, Horton 2010, Philp 2010, Rossi 2012, Sleat 2013, Sleat and Rossi 2014, Jubb 2015, Finlayson 2017). The hard-core realist view has both normative and methodological implications. However, while often presented as a critique of ideal theory, realism’s rejection of “moralized” understandings of politics challenges not only this particular approach, but also any theory that seeks to identify principles of justice. Rather than an objection on the grounds of prac- ticability, hard-core realists propose a radically different way of thinking nor- mativity within the political domain (Baderin 2014, Sleat 2016). Their critique offers a refreshing challenge to the project of contemporary political theory.
Yet, this challenge takes us far beyond the criticism from practicability of con- cern here. The problem this critique ascribes to ideal theory is not that it at- tempts to identify principles of justice, but that it fails to deliver what it prom- ises. To allow my study to concentrate on the methodological questions that arise within the domain of justice theory, I will leave the more wide-ranging critique advanced by hard-core realists aside.
Instead, I will focus on the critique that states that ideal theory fails to tell us what justice demands. Critics advancing this kind of claim – let us call them
“moderate realists” – share the ideal theorists’ aim of identifying principles of justice, but object to the methods they employ to reach them. If we want to know what justice demands in this world, they argue, we must descend from the ivory tower of abstraction and engage with the messy facts of everyday life. In particular, we must attend to people’s shortcomings and limited moral motivations, which ideal theory tends to dismiss as empirical contingencies.
According to moderate realists, ideal theory disregards too many facts, some
9
of which constrain what justice can demand rather than hinder its realization.
What they request is not a radical shift away from justice theory, but closer attention to the imperfect nature of real-world agents and societies (see e.g.
Mills 2005, Farrelly 2007, Robeyns 2008, Herzog 2012, Jubb 2012, Wiens 2015b). As previously explained, the critique may target either the validity of principles produced in ideal theory or the usefulness of information about full justice in nonideal circumstances. The first charge states that the normative principles produced in ideal theory are not valid principles for our world. The second charge states that even if they are, knowing what ideal justice requires is unhelpful for guiding action in the nonideal conditions that prevail in to- day’s societies. While some critics propose alternative approaches for finding out what nonideal justice requires (Sen 2006, Wiens 2012), my focus will be on their negative claim that ideal theory is neither necessary nor sufficient for guiding action in nonideal circumstances.
The methodological critique highlights specific aspects of the contested re- lationship between facts and norms in political theory. In particular, it calls attention to the so-called feasibility constraint on the concept of justice (Bren- nan and Pettit 2007). While critics and advocates agree that the demands of justice must be feasible in order to be action-guiding for political agents, they draw contrary conclusions regarding ideal theory’s capacity to cater to this constraint. This calls for further investigations of what the feasibility con- straint on justice entails, and how different understandings of this constraint affect our judgments about ideal theory. Feasibility is, however, not the sole object that requires further examination and elaboration. Other crucial con- cepts, such as action-guidance and nonideal circumstances, are similarly un- der-specified. While it is commonly agreed that normative principles cannot offer exact and immediate guidance for specific situations – in order to arrive at action-guiding recommendations, principles require application and inter- pretation – there is no agreement on what kind of information an action-guid- ing account must minimally provide. Does it suffice that a principle points to a value that should be granted consideration in deliberations about what to do, or must it also tell us something about how this value can be realized, or how much weight it should be attributed compared to other concerns? In the ab- sence of clear answers to these and similar questions, critics and advocates of ideal theory tend to talk past each other, or retract to positions that are easily defendable yet fail to address the other side’s objections (see e.g. Swift 2008:
382). By bringing attention to such unclarities and offering suggestions for how they may be resolved, my study fills a gap in the existing research and paves the way for fruitful future debates on ideal theory’s status and useful- ness.
10
Ideal theory
What, then, is ideal theory? This question turns out to be trickier than one would expect, and advocates and critics of ideal theory tend, to some extent, to give different answers to it. For my present purposes, it is important that my understanding of ideal theory does not settle the dispute between advo- cates and critics prematurely by casting ideal theory as clearly unproblematic or inherently flawed. By contrast, the understanding I opt for should help us realize both what it is about ideal theory that gives rise to the critics’ objec- tions, and why ideal theorists find this approach attractive in the first place.
Rather than stipulating my own definition of ideal theory, this purpose is best served through adopting a tentative working definition, whose primary aim is to provide a starting point for coming discussions. This section outlines such a definition, and discusses the relationship between ideal theory in general and Rawls’ ideal theory in particular.
Previous proposals
Let us start by looking at some prominent attempts to define ideal theory.
Rawls, who coined the term, equated this approach with strict compliance the- ory, bracketing problems that arise due to noncompliance (Rawls 1999: 8). It may, however, be questioned whether Rawls’ theory actually assumes full compliance, at least in the “formal” sense of bracketing every problem of fore- seeable noncompliance with proposed principles (Hamlin and Stemplowska 2012: 55, Levy 2016: 319). Further, Rawls’ account also assumes favorable circumstances, bracketing problems arising from natural limitations and his- torical contingencies (Rawls 1999: 216, Simmons 2010: 13-4). This suggests that his account is more complex than he may have acknowledged, and gives rise to problems of which he was unaware.
Recent proposals for what characterizes ideal theory tend to focus on either (1) theory output (an end-state ideal of justice; an account of the principles governing the fully just society), (2) theory input (false assumptions or ideal- izations), or (3) the theory’s prospects of realization (utopian or unrealistic).
Beginning with the last, the third proposal hardly offers neutral ground on which to proceed. Clearly, ideal theory’s advocates would deny that their ap- proach is per definition utopian or unrealistic (Valentini 2012: 657). A less controversial version of (3) is developed by Zofia Stemplowska, who argues that ideal theory is theory that does not offer AD-recommendations, that is, recommendations that are achievable and desirable in a not too distant future (Stemplowska 2008: 324). This proposal fits the role played by ideal theory today, and it captures many of the complaints raised by its critics. Yet, whether a theory offers AD-recommendations or not depends on the factual conditions that happen to prevail in the world. In conditions that are highly favorable for justice, ideal theory would, on the proposed understanding, cease to be ideal.
If we find this implausible, it is probably because we think that the point and
11
purpose of ideal theory cannot be to produce non-AD-recommendations. What drives ideal theorists is not this aspiration, but the quest for what full justice would require, and (luckily) this does not change as conditions become more favorable.
This leads us to proposal (1), which defines ideal theory in terms of theory output. Ideal theory, on this proposal, is theory that offers an end-state ideal of justice, that is, an account of what full justice would require (Valentini 2012: 660-1). So understood, ideal theory is ideal in the non-technical sense that it proposes an account of the fully just society that we should strive to realize (Valentini 2009: 337). This understanding reflects the ambitions of ideal theorists, and it captures well the uselessness charge, which states that accounts of full justice are uninformative in nonideal circumstances. Consid- ered alone, however, it seems unable to account for the demandingness charge, since it is hard to see that accounts of full justice could be overly demanding as such. We may of course have different ideas of what justice involves, or how distant from the conditions we find in current societies an ideal of justice can be without falling prey to utopianism. But disputes of this kind do not challenge the quest for full justice, only specific conceptions of justice.
In order to see what generates the demandingness charge, we must turn to the input-oriented proposal (2). According to this, ideal theory is theory that makes use of idealized, i.e. false assumptions. Theories designed under ideal- ized conditions are ideal in a technical sense (Valentini 2009: 338). I will re- turn later on to the question of what this entails. For now, let us just note that if we understand ideal theory only in terms of input, this category will include theories that build in false assumptions of extremely unfavorable or unrealistic kinds, such as the assumption that human beings have exclusively sadistic motives, or that their life-span extends to only one day. While it is an open question whether theorizing on such premises can tell us anything of value for our world, our preference for ideal theory arguably does not commit us to an- swering this question in the affirmative. That is, to allow for certain kinds of idealizations does not commit us to the view that any idealization will do (c.f.
Laurence 2018: 443, Southwood 2018: 9). Similarly, ideal theory’s critics need not categorically dismiss every kind of idealization. The idealizations of interest here are those that ideal theorists hold to be necessary for delimiting their object of study and avoiding bias toward status quo. Critics advancing the demandingness charge argue that these idealizations prevent us from ac- knowledging the various conditions that constrain justice in our world.
None of the three proposals thus captures the particularity of ideal theory on its own. Reaching similar conclusions, some authors suggest that ideal the- ory is defined not by one feature, but by a combination of features (Hamlin and Stemplowska 2012: 52). One such account is Laura Valentini’s under- standing of ideal theory as theory that is ideal in both the technical and the non-technical senses (Valentini 2009: 338). I will follow Valentini in assum- ing that ideal theory displays both these features. For clarity, however, I will
12
break up the criterion of non-technical ideality in two. The first criterion em- phasizes ideal theory’s focus on full justice, whereas the second highlights the regulative role this account ascribes to justice. While the former captures the distinction between ideal and nonideal theory, the latter distinguishes between ideal theory and so-called “theory of ideals” (Hamlin and Stemplowska 2012:
55). In the next section, I present my working definition of ideal theory.
A working definition of ideal theory
For the purposes of this study, I will understand ideal theory as theory that a) seeks to identify first principles of justice, b) holds these to be the principles that govern the fully just society, and therefore c) operates on certain kinds of false, or idealized, assumptions. This understanding is under-specific in the sense that it does not describe the idealizations to which the ideal theoretical approach commits us. Since narrowing down on this aspect would risk ruling out some of the critics’ objections on definitional grounds, I will leave this question open at the initial stage. In my individual papers, I can then opt for more specific understandings that highlight the idealization presently under consideration. Granted, however, that my study focuses specifically on how constraints set by noncompliance and motivational limitations affect our judg- ments about ideal theory, the assumption of full compliance will play a prom- inent role in my coming discussions.
According to my suggested working definition, three features are charac- teristic of the ideal theoretical approach. First, ideal theory seeks to identify first principles of justice, that is, principles at the highest level of abstraction.
First principles are not derivative of other principles (of justice), and the stand- ards they set involve no compromises with justice (Cohen 2008: 269, Estlund 2017: 41). If the demands of ideal theory are fulfilled, there is nothing to object to a situation from the point of view of justice. In this respect, ideal theory stands in contrast to nonideal theory of e.g. retributive or reparative kinds, which deal with responses to injustice. It also contrasts to transitional justice and to comparative approaches that ask which of two currently feasible (but less than fully just) alternatives is preferable (but see Hamlin and Stemplow- ska 2012: 52). The aim of identifying first principles further marks the differ- ence between ideal theory and applied theory, which uses principles of various kinds to tease out implications for specific institutions and practices. Ideal the- ory does not prescribe specific institutional arrangements, it only specifies the demands that these arrangements should meet in order to be fully just (Swift 2008: 377). This first feature of ideal theory gives rise to the uselessness charge. While ideal theorists take an account of full justice to be a prerequisite for finding out what nonideal justice requires, their critics hold this to be both insufficient and unnecessary for this purpose.
Second, ideal theory holds first principles of justice to be the principles that govern the fully just society. That is, ideal theory ascribes to justice a regula- tive role. In this respect, ideal theory stands in contrast to what Alan Hamlin
13
and Zofia Stemplowska call “theory of ideals” (Hamlin and Stemplowska 2012: 55). The former takes justice to be an instance of practical reason whose aim is to tell us what to do, whereas the latter perceives of justice in epistemic terms, stating that principles of justice tell us what to think, regardless of whether this has any practical implications (Gilabert 2009: 56, see also Stem- plowska and Swift 2012: 384). According to the Rawlsian account, justice is the “first virtue of social institutions” in that its principles regulate the basic structure of the fully just society. For him, justice is the solution to the practi- cal problem of distribution that arises when people make conflicting claims to scarce resources (Rawls 1999: 3, 110).
The practical role Rawls ascribes to justice is closely related to two of his further commitments. The first is that principles of justice are fact-sensitive, that is, their justification depends on the outcomes they produce in the empir- ical conditions that prevail in our world. The fact that a principle would gen- erate unjust outcomes in a world different from ours does not speak against it or reveal that it is secondary to some other, universally valid, principle of jus- tice (Rawls 1999: 137, 398, Jubb 2009: 345, Miller 2013: 18, Ronzoni and Valentini 2016: 418, Lister 2017: 123). The second commitment is that prin- ciples of justice are subject to a general feasibility constraint (Brennan and Pettit 2007, Gilabert and Lawford-Smith 2012, Hamlin 2017). On the practical account, justice is action-guiding in the sense that it tells us what to do. Justice guides action through conferring claim rights and duties on agents. The special status of claim rights allows their holders to make demands on duty-bearers.
In addition, duties corresponding to claim rights can be rightfully enforced by either rights holders or third parties (Valentini 2017b: 3). The “oughts” prin- ciples of justice place on us are in this sense actual rather than pro tanto, and owned rather than unowned (Mason 2004: 257, Lawford-Smith 2010: 358, Valentini 2017a: 14). To say that a state of affairs is unjust is thus to say that someone’s rights have been violated, and that some agent has failed to fulfill a duty of hers (Anderson 2010: 4, Valentini 2017a: 19-20). Following the Kantian dictum “ought implies can”, no agent can have a duty to do that which s/he cannot do. If demands of justice take the shape of actual duties, then, justice is subject to a general feasibility constraint (Lawford-Smith 2013b:
244-5, but see Gheaus 2013: 463). To be sure, realist critics of ideal theory do not deny that justice is regulative in this sense. Instead, this feature explains why ideal theory is vulnerable to their critique in a way “theory of ideals” is not. While advocates of the latter may simply accept the critics’ charges of infeasibility or practical uselessness, arguing that these are not reasons to re- ject a conception of justice, this reply is not open to ideal theorists (Valentini 2009: 337).
The third constitutive feature of ideal theory is that it is construed under idealized, i.e. false, assumptions, that make social reality appear simpler or better than it actually is (Valentini 2009: 332). In this respect, ideal theory stands in contrast to more “realist” approaches that pay closer attention to em- pirical facts. Considering what idealization entails, Onora O’Neill contrasts
14
this to the related concept of abstraction. While abstraction brackets certain features from consideration, idealization involves the adding of features to the phenomenon under scrutiny. Abstraction, she argues, is an unavoidable and unproblematic feature of all theorizing. Yet, the use of idealizations is, partic- ularly in normative theory, a delicate and potentially dangerous enterprise (O’Neill 1996: 39-42). The distinction between abstraction and idealization has, however, been called into question. As Miriam Ronzoni observes, ideal- izations can often be restated in terms of abstractions. What matters according to her is not whether an assumption constitutes an abstraction or an idealiza- tion, but whether the idealization is good or bad (Ronzoni 2008: 92, see also Goodin 1995: 41-2). An account of what this entails is offered by Valentini, who suggests that a bad idealization idealizes the subject to which the theory applies. While a good idealization allows us to factor the things abstracted away at the level of theory construction back in at the level of theory applica- tion, a bad idealization constitutes an integral part of a theory’s fundamental principles, and hence cannot be removed without also changing those princi- ples (Valentini 2009: 352-3, see also Appiah 2017: 26). The use of idealiza- tions is what gives rise to the demandingness charge against ideal theory. By proceeding on false assumptions, critics argue that ideal theorists fail to acknowledge facts that may render an account of justice infeasible or undesir- able in the real world (see e.g. Hope 2016: 391).
Rawls’ ideal theory
My proposed understanding of ideal theory is tailored to the account Rawls develops in A Theory of Justice. This is no coincidence; Rawls’ theory consti- tutes the most prominent example of an ideal theory, and it is the primary target of recent critique. Indeed, if a proposed definition of ideal theory ex- cludes Rawls’ theory, many would take this to indicate that the definition is wrong (Hamlin and Stemplowska 2012: 51, Levy 2016: 317, footnote 8). Im- portantly, however, my proposal is not identical to Rawls’ own conceptual- ization of ideal theory, which reduces this approach to strict compliance the- ory. Rather, it captures the crucial aspects that distinguish his theory from nonideal theory on the one hand, and theory of ideals on the other – aspects that are central to the critique. Still, since much of the recent debate revolves around Rawls’ account, it is important for coming discussions that the reader is familiar with its basic elements. In this section, I will therefore offer a brief outline of his theory of justice.
The aim of Rawls’ theory is to identify the principles that would govern the fully just society. Rawls adopts a constructivist approach, according to which principles of justice are those that we would choose in a hypothetical ideal choice situation. In his version of this situation, the choosing parties are placed behind a veil of ignorance that deprives them of all specific knowledge about their society and themselves; their social status, talents, preferences, and be-
15
liefs. Thus situated in what Rawls calls the Original Position (OP), the deni- zens are to choose the principles that they would want to govern their society.
Rawls assumes that the choosers in the OP are rational and self-interested, and hence will strive to secure for themselves the maximum of resources that they may need to fulfill their life-plans (whatever these turn out to be once the veil is lifted) (Rawls 1999: 16-7). He argues that two principles will fall out of this procedure: the liberty principle, which grants each member of society the max- imal range of basic liberties that are compatible with equal liberties for all, and the difference principle, which allows inequalities only if they a) benefit the worst-off and b) are attached to positions that are open to all (Rawls 1999:
53).
The principles generated in the OP are then tested in the process aimed at reaching reflective equilibrium. In this process, we compare the outcomes of principles at different levels of abstraction to our considered judgments about justice, adjusting either the one or the other each time they run into conflict.
Here, not only principles of justice, but also the settings of the ideal choice situation are up for evaluation. The aim is to find principles that fully capture our considered convictions about justice, and can explain these by reference to the ideal choice situation. Once our principles and commitments are in full accordance with each other, we have reached what Rawls calls reflective equi- librium (Rawls 1999: 18, 2001: 29, Kurtulmus 2009: 494).
Rawls’ two-step justificatory process includes several idealizing assump- tions. The denizens in the OP are cast as free and equal, rational bearers of ends, unaware of their individual characteristics and the specifics of their so- ciety, but in possession of full knowledge about the general facts and laws that govern the social world. They base their choice of principles on the assump- tions of closed societies populated by capable adults, favorable natural and historical circumstances, and full compliance (Rawls 1999: 7-8, 17, 2001:
101). Some of these idealizations aim to ensure that the theory does not incor- porate a bias toward status quo. If, for instance, the choice of principles were to reflect real-world injustices such as race or gender inequalities, the theory would arguably not be just at all (Valentini 2009: 346). Others serve to delimit the topic of investigation and thereby simplify the task of the theorist. Rawls did not premise his theory on the assumption of closed societies because he held interchanges over borders to be inherently unjust, but because he thought (rightfully or not) that the problem of domestic justice could be solved without attending to the international context.
What about the assumption of full compliance? This assumption constitutes a simplification in that it brackets issues such as retributive and reparative jus- tice, and thereby delimits the topic of investigation. Yet, per definition, non- compliance also constitutes a violation of justice. Unlike pre-theoretical com- mitments to individuals’ status as free and equal, or societies as closed systems of cooperation, the content of full compliance is defined by the principles that are generated under this assumption. Rather than a background condition for justice, the assumption of full compliance is a heuristic device that allows us
16
to assess principles in the light of their outcomes in the world. According to the Rawlsian account, it is when these outcomes fully capture our considered convictions about justice that we reach reflective equilibrium. Since only the outcomes of complying with a principle are rightfully described as outcomes of the principle itself, the justificatory process must be conducted on the as- sumption of full compliance (Simmons 2010: 8, Stemplowska 2008: 334).
Importantly, what matters for the assessment of principles are the outcomes they produce in real-world conditions. A possible objection against the differ- ence principle is that it would justify great inequalities if such inequalities im- proved the access of the worst off to primary goods. Rawls’ reply to this ob- jection is that in a competitive economy with open class relations, excessive inequalities will not be the rule (1999: 137, see also Lister 2017: 122). Princi- ples of justice need not capture our considered convictions about justice under any conceivable conditions, only under those that actually prevail in our world. “Conceptions of justice”, writes Rawls, “must be justified by the con- ditions of our life as we know it or not at all” (1999: 398).
The outcome of applying a principle in conditions of full compliance will of course be radically different from that produced under partial compliance.
This raises a concern among critics that ideal theorists misjudge the actual outcome of principles, and hence come to affirm principles that do not capture our convictions about justice if applied to real-world conditions. This concern gives rise to objections based on either demandingness or uselessness. The first states that the assumption of full compliance blinds us to the constraints justice is subject to in the real world, leading us to affirm conceptions of jus- tice that are unrealistically demanding and hence invalid (Galston 2010, Schmidtz 2011, Wiens 2016c, Levy 2016). The second states that once we lift the assumption of full compliance, we have no use for the principles that were generated using this assumption. In other words, ideal principles cannot tell us what to do in circumstances where noncompliance renders full justice unachievable (Sen 2006, Farrelly 2007, Robeyns 2008, Wiens 2015a).
Ideal theory beyond Rawls
Given the prominent role played by Rawls’ theory in the current debate, it is legitimate to ask to what extent the critique under consideration challenges ideal theories other than his. Indeed, one may even wonder which other ac- counts fall into the category of “ideal theory”. Two approaches often men- tioned in the debate are those of Ronald Dworkin and Robert Nozick. In this section, I give a tentative account of how each of these fits into my proposed understanding of ideal theory, and how the critique presently under consider- ation targets them.
Like Rawls, Dworkin offers a constructivist account of justice. In his ver- sion of the ideal choice situation, shipwreck survivors on a deserted island are given equal numbers of clamshells that they can use to place bids in an auction dividing the island’s resources between them. The resulting distribution is just
17
if it passes an envy test, that is, if no auctioneer prefers the bundle of goods of someone else to that he ends up with (Dworkin 2000: 66-7). The initial distri- bution resulting from the auction is then supplemented by an insurance scheme, which offers sufferers of bad brute luck (e.g. handicaps) compensa- tion to the extent that the average person – unaware of her turnout in the nat- ural and social lotteries – would have chosen to spend some of her clamshells on insuring against this particular misfortune (Dworkin 2000: 77-8).
Dworkin’s theory appears to be ideal in both the technical and non-technical senses. It offers an account of full justice, and it uses idealizations to arrive at this account (Valentini 2009: 338-9). However, it makes no use of the assump- tion of full compliance, since it does not propose principles that are justified in terms of their distributive outcomes. Unlike Rawls’ OP, which is just inso- far as its principles produce a distribution that captures our considered con- victions about justice (i.e. it does not allow for “excessive inequalities”), Dworkin’s auction is a pure procedure whose outcome would be just whatever it turned out to be.
While the procedural take arguably renders Dworkin’s account immune to the objection that his principles are not justified in the light of real-world con- ditions, a different objection arises from the fact that his theory remains silent on the outcome of his procedure. Ingrid Robeyns notes that Dworkin holds a just distribution to be choice sensitive and endowment insensitive, granted that people’s choices reflect their authentic and independent preferences. In the real world, however, our preferences are often distorted by ignorance, preju- dice and preference adaptation. Under these conditions, knowing that a distri- bution is choice sensitive and endowment insensitive tells us little, if anything, about its justice (Robeyns 2008: 356). Since Dworkin does not tell us what the outcome of his ideal choice situation would be – indeed, he argues that it would be absurd to think that we could predict the outcome of the auction he envisions if conducted in the real world – we have no way of knowing what a just distribution demands here (Dworkin 2000: 163). This raises a practicabil- ity concern of a specific, epistemic, kind (Howard-Snyder 1997: 242, but see Hughes 2018: 448). In contrast to Rawls’ principles, the problem is not that Dworkin’s theory places unrealistic demands on us, but that we cannot know which those demands are (see also Farrelly 2007: 857-9).
In Anarchy, State and Utopia (1974), Nozick sets up two principles for just entitlement: the principle of justice in acquisition, which regulates the appro- priation of previously unowned things, and the principle of justice in transfer, which regulates the exchange of things already owned. According to Nozick, any distribution that follows from the application of these principles is just, and no one is entitled to any holding beyond such application (Nozick 1974:
151). Nozick’s theory also strives to identify first principles of justice, and is thus ideal in the non-technical sense. Yet, it is not clear that it involves ideal- izations of the kind that would render it ideal in the technical sense. Nozick takes principles of justice to be governing in that they set the limits for state intervention. He thus attributes to justice a regulative role. Like Dworkin,
18
however, Nozick denies that principles are justified in the light of the distrib- utive pattern they produce when applied to the world. For this reason, his ap- proach, too, seems invulnerable to the objection that it leads us astray if ap- plied to nonideal conditions. To see this, consider an objection to Nozick’s theory developed by David Wiens. According to him, Nozick’s understanding of rights as side constraints (i.e., rights must not be violated even for the sake of a greater protection of rights) is plausible only against the idealized starting point of a Lockean state of nature where people generally tend to satisfy moral demands. Once we substitute the Lockean state of nature for a Hobbesian war of all against all – where rights violations are the rule rather than the exception – we may accept the use of coercion even for purposes that go beyond the protection of rights, such as promoting social cooperation (Wiens 2015a, ap- pendix). Wiens’ claim appears to be the following: Given that we share Nozick’s concern for rights-protection and individual autonomy, the outcome of his principles would strike us as unjust if applied to nonideal (Hobbesian) conditions. Recall, however, that Nozick does not hold that principles of jus- tice are justified in the light of their outcome. To understand rights as side- constraints is precisely to deny that rights-protection is a goal that the state should strive to maximize (Nozick 1974: 29). Since nothing commits Nozick to the view that justice must secure effective rights’ protection or opportunities for autonomous action in the real world, Wiens’ challenge is rather a refutation of his theory on substantive normative grounds than a methodological objec- tion.
It is, however, easy to think of other objections based on uselessness that could be advanced against Nozick’s approach. Nozick proposes that situations where the distribution of holdings are the result of unjust acquisition will be regulated by a principle of rectification. Leaving the details of rectification unexamined, he suggests that this would involve the tracking of justice viola- tions and an estimation of what would – counterfactually – have occurred in their absence. If the current distribution does not match any of the expected outcomes, a redistribution must take place. Importantly, if the estimation yields more than one description of holdings, other principles, such as a prin- ciple of equality, may come into play for choosing between them (Nozick 1974: 152-3). Now, given that most of the holdings in today’s world are the result of unjust acquisition at some point in history, the nonideal principle of rectification would basically do all the normative work in real-world condi- tions. And since the injustices are so pervasive that almost any distribution might have followed had they not occurred, we will choose between an innu- merable set of distributions. Since Nozick’s theory does not rule out any dis- tributive principle for this purpose, it is arguably useless for guiding action in the nonideal conditions of today’s societies.
While Dworkin’s and Nozick’s approaches thus share many features with Rawls’, they lack some of the characteristics included in my working defini- tion of ideal theory. In this section, I have pointed to how these differences may bear on the way certain objections advanced against Rawls’ ideal theory
19
affect their approaches. In my concluding section 5, I return to the question of what this implies for the scope of my findings.
The objections
Critics of ideal theory argue that this approach fails to identify principles of justice that are valid for our world, or to offer action-guiding recommenda- tions in the nonideal conditions that prevail in today’s societies. In this section, I present the different charges and explain how each of the four papers in my dissertation relate to them.
The demandingness charge
The demandingness charge states that ideal theory fails to acknowledge cer- tain conceptual constraints on justice, and hence leads us to affirm invalid principles. The idealizations employed within the ideal theoretical approach blind theorists to obstacles that render conceptions of justice infeasible or un- reasonably costly in real-world conditions. In particular, failures to recognize the motivational limitations of human beings result in ideal theory ending up with unrealistically demanding, and hence utopian, ideals (Schmidtz 2008: 8, Galston 2010: 395, Schmidtz 2011: 778, Wiens 2015a: 442, Wiens 2016b:
349, Wiens 2016c: 167).
The view that justice must cater to motivational limitations is not alien to ideal theorists. Indeed, Rawls agrees that the conception of justice we adopt must be “for humans”, and hence not rely on the moralities of “the saint and hero” (Rawls 1999: 419). If compliance with the demands of justice requires courage, altruism or endurance of a kind that exceeds what most people are capable of displaying, the theory is no longer realistically utopian, but utopian tout court. Rather than being for or against a motivational constraint on justice, the dispute between advocates and critics turns around what this constraint entails. Its content is, however, a matter closely related to its justification. If we want to identify the core of this disagreement, we should therefore take a step back and ask why accounts of justice should cater to motivational limita- tions in the first place.
The most common justification of the motivational constraint holds this to form part of a general feasibility constraint on the concept of justice. This constraint follows from the view that justice is an instance of practical reason and hence strives to be action-guiding, combined with a commitment to
“ought implies can” (Brennan and Southwood 2007: 1, Gilabert 2017: 99).
The idea is that if humans cannot motivate themselves to comply with certain demands, they also cannot comply with them (Wiens 2016b: 341). This raises questions about what it means that we “can do” something, and hence what the feasibility constraint on justice entails. It also highlights the importance of how we perceive of the bearer of duties. Is it feasible for us to stop global
20
warming? The answer to this question depends heavily on who constitutes the relevant “us”. To stop global warming may be feasible for a collective made up of world leaders, yet infeasible for a group of teenage activists.
Often, that which is infeasible for a single agent is feasible for a group, working jointly to achieve a desired outcome. A group of agents may thus be assigned duties that would violate “ought implies can” if ascribed to a single person. We should, however, resist the temptation of assuming that feasibility automatically increases with the number of agents. To stop global warming may seem clearly feasible to us if “we” refer to the world population as a whole. Yet, upon closer look this may not be so, since the feasibility of col- lective action comes with its own challenges. Importantly, it depends not only on the individual abilities of the agents forming the collective, but also on how these agents coordinate and communicate their actions (Stemplowska 2016:
284). Further, if not all agents are equally motivated to participate in the col- lective enterprise, its success will depend on the ability of other members to get them on board. If some members fail to contribute their share, the outcome may become infeasible for the remaining members, calling for a change or annulment of the duties that fall on them (Lawford-Smith 2012: 457).
Assuming that any outcome that would result if all individuals did their share is feasible thus ignores two problems often held to be constitutive of politics; the problem of collective action and the problem of governance.
These problems may seem to concern nonideal theory exclusively. Still, if jus- tice demands that with which most people will never comply, or where com- pliance could never be ensured through institutional arrangements, we may worry that the collective duties assigned by ideal principles will not distribute to individual agents even in the best foreseeable, yet realistic, conditions (but see Goodin 2012). If mutual expectations of others’ noncompliance prevent obligations from arising even in the most favorable conditions we can realis- tically expect, a “normativity gap” between the idea of injustice and the idea that someone treats others unjustly (i.e. fails to fulfill their obligations) will arise (Stemplowska and Swift 2012: 385, see also James 2003: 139). Such a gap deprives justice of its political force, since the recognition that something constitutes an injustice does not necessarily prompt any action or call for a change (Mason 2004: 253, Wiens 2014: 304). If we want our ideals to be ac- tion-guiding for political agents, we must therefore attend to the problems of collective action and governance even in ideal theory.
This insight reveals that the idea of a feasibility constraint on justice is more complex than has often been acknowledged. Not only may motivational abil- ity restrict which principles we can comply with; unless we can realistically make others display the proper motivations, collective duties of justice will also fail to distribute to individuals. Rather than a single constraint, this gives rise to several proposals for a feasibility constraint on justice. My first paper,
“Three Feasibility Constraints on the Concept of Justice”, distinguishes be- tween three understandings of this constraint, and discusses how each of these may be justified. My second paper, “Motivation as Constraint”, takes a closer
