La Théorie de la justice de John Rawls se présente explicitement comme une conception "paroissiale" dont l'application est limitée aux démocraties libérales occidentales. Lorsqu'il s'agit de définir les principes d'une justice globale, il s'en tient au respect des droits fondamentaux des individus et, pour le reste, à la tolérance à l'égard des différentes pratiques culturelles. Cette approche a fait l'objet de nombreuses controverses au sein de la philosophie politique libérale contemporaine. Une des plus fécondes contributions à une conception cosmopolite de la justice est apportée par l'approche des "capabilités", qu'explorent Amartya Sen et Martha Nussbaum.
Dans cet article, fort bien documenté, Elisa Hörhager, titulaire d'un Master en Sciences Politiques, diplômée de l'Institut d'Etudes Politiques d'Aix-en-Provence, fait le point sur ces débats hautement sophistiqués. Je la remercie infiniment de m'avoir autorisé à le publier.
As of the 1990s, the persistence of global socio-economic inequality shifted political theory towards a new debate about a potential remedy for this situation on a global scale. This debate is called the cosmopolitan debate on social justice. For several cosmopolitan theorists in the contractarian current, this global shift has been a continuation of the logic implicit in the 20th century’s fundamental theory on “justice as fairness” by John Rawls. Uniting social justice with liberalism, Rawls initially limited his theory to the confinements of a nation state. At a later stage, he acknowledged the challenge of global justice and tried to take it into account by formulating a law of peoples to be developed out of his own idea of liberal justice (Rawls 1999, 3).
The cosmopolitan debate on global justice is based on the Rawlsian premise that the way natural endowments are distributed in morally arbitrary. Cosmopolitan theorists agree with Rawls that “we do not deserve (in the sense of moral desert) our place in the distribution of natural endowments” (Rawls 2001, 74). However, unlike Rawlsian theory of justice, cosmopolitanism is also based on the assumption that the birthplace of a person or a person’s nationality is morally irrelevant for thinking about justice. The “chance event of being born in one nation rather than another pervasively determines the life chances of every child who is born” (Nussbaum 2006, 224), and is something which cosmopolitanism wants to overcome through a globally applicable concept of justice.
As cosmopolitan approaches confer upon each individual, equally, the status of “ultimate unit of concern” on a global scale and in complete generality (Pogge 1992, 48), they have taken up the methodological individualist standpoint inherent in political liberalism. However, extended to a global scale the meaning and requirements of political liberalism are far from clear. Legitimately arguing that a concept of justice is globally applicable on the basis of liberal methodological individualism poses numerous problems, from context sensitivity to responsibility. The goal of this article is to examine the compatibility between global justice and political liberalism.
Through a closer look at the way Rawls conceived of justice on a global scale, the challenges of incorporating political liberalism into cosmopolitan theory become clearer. Justice as fairness is politically liberal in that it can be accepted by different individuals each with their own comprehensive views on the good life. Rawls however refused to extend the reasoning and principles of justice as fairness to the global scale, arguing instead for an alternative law of peoples based only on human rights duties. It is therefore necessary to re-examine the reasons why Rawls refused to extend his contractual concept to a global scale. Political liberalism, as conceived by Rawls and as employed by cosmopolitanists, consists of two main principles. When applied to an individual inside a politically liberal society, they can be termed negative and positive freedom. On a global scale of justice, both toleration and pluralism are values inherent in liberalism. Hence, depending on the scale of justice, these principles vary in terminology yet only slightly in content. Globally, the dichotomy is of a larger order and concerns not the content of liberalism, but the validity of political liberalism as a doctrine. It questions the capacity of such a “higher-order” liberalism to deal with the global empirical context, which contains liberal as well as non-liberal societies.
These two principles inherent in political liberalism do not only complement each other, but also are alternatively prioritized in different versions of social justice. A first step in this article is to compare on the scale of a liberal society Rawls’ version to an alternative theory of liberal justice which has recently drawn much scholarly attention. Martha Nussbaum’s capability approach is deliberately conceived as an alternative to Rawls’ resourcist, rationalist contractarianism. However, the capability approach faces the possibility of a trade-off on negative freedom while prioritizing empowerment.
The second section of this paper focuses on the global scale. Far from simply rejecting liberalism on a global scale, John Rawls refused to elaborate a cosmopolitan version of his politically liberal justice as fairness concept out of respect for the liberal principle of toleration. This position can be criticized on several accounts. Most fundamentally, it betrays the primary assumption of methodological individualism at the center of liberalism and of Rawls’s own theory. In light of this trade-off on pluralism in favor of toleration, the capability approach is put forward as an attractive alternative. It attempts to be at the same time a “thick” and a “vague” concept of justice, both context sensitive and globally applicable. In conclusion, the capability approach overcomes Rawls’s stunted version of global justice by consistently and effectively applying political liberalism. From the perspective of the individual, where Nussbaum’s approach might not seem liberal enough is less in its approach to global justice than in problems of individual applicability.
What political liberalism means for social justice
There are two perspectives for conceiving of the abstract principle of freedom which forms political liberalism. Liberalism means respecting both the “positive” freedom and the “negative” freedom of individuals, meaning the freedom to act and the freedom to be free of constraints (except for the constraints imposed by the positive freedom of others). In a concept of social justice, positive freedom means leaving room for individual concepts of the good life – “free agents must be, in a fundamental sense, self-governing” (Christman 1991, 344). Negative freedom in liberalism is traditionally formulated both as state neutrality towards individual ideas of the good life and neutrality between the individuals as a function of their individual freedom. It is important to recall that these two perspectives on freedom, conceptualized separatly in this abstract manner, are two sides of the coin of political liberalism and mutually enable one another. As will become clear, however, theories of justice tend to epistemologically favor one of the two perspectives.
What it means to enable true individual autonomy through social justice differs largely with the anthropological concept of a theory. Rawls recognizes that in order to provide effective autonomy freedom from constraints has to be complemented by basic rights and resources. Thus surpassing a purely formal definition of liberalism, he has gone to great lengths in a procedural contractarian construct so that the result would be a module of justice which could be accepted by and fit in to various reasonable comprehensive doctrines. In the original position, his is a theory of rational choice (on this point see DeLue 1980) based on a situation of complete equality under the veil of ignorance. “Aristotelian” elements of human beings in their social reality are therefore left out of the picture while the principles of justice are formed. This is exemplified in his rejection of civic humanism, where Rawls writes “in the strong sense, civic humanism is (by definition) a form of Aristotelianism: it holds that we are social, even political [...] beings“(Rawls 2001, 142-3). Aristotelianism is rejected as a “comprehensive philosophical doctrine and as such incompatible with justice as fairness as a political conception.” (id.)
Rawls’ liberal concept has been widely criticised for “de-individualizing“ the contracting parties and focusing one-sidedly on economic resource distribution (for an overview see Kersting 2002, 63-69). This is not completely accurate. In a later stage of justice as fairness, the reasonable, disinterested individuals formerly under the “veil of ignorance” (Quote) acquire social and moral characteristics. This is why Susan Moller Okin can claim that at the center of justice as fairness “(though frequently obscured by Rawls himself) is a voice of responsibility, care and concern for others.“ (Okin 1989, 230). For Nussbaum, this is the point in Rawls’ theory where it becomes obvious that no concept of justice can afford to ignore the social situation of individuals and the injustices and limitations they face in reality. With regards to liberalism, a reductionist anthropological foundation to justice will lead to privileging negative freedom. By trying too avidly to avoid imposing any particular idea of the good life, Rawls’ reductionist justice theory fails to provide for any effective autonomy.
The capability approach builds on this criticism of contractarianism to propose a holistic view of tolerance for the individual which takes into account personal concepts of the good life. This attempt is termed a “thick vague theory of the good” (Nussbaum 1992, 214-5). Instead of conceiving of justice as a module fitting in to the different reasonable comprehensive doctrines, it represents more of a foundation on which these different doctrines and views of the good life can build. Nussbaum has taken up “Aristotelian” elements to be found also in justice as fairness and incorporated them in her outcome-oriented normative approach. The anthropological picture her theory paints is one of the individual not only as a rational being, but also as an individual characterized by different degrees of vulnerability and different dimensions of belonging. Together with Amartya Sen, Martha Nussbaum is not contented with limiting the reach of social justice to the availability basic human goods. It is not what people have but rather “what people are actually able to do and to be” (Nussbaum 2000, 5) that determines their capacity to take advantage of the freedom at the heart of political liberalism. According to Martha Nussbaum, by conceiving of the individual as an empty unit or a procedural agent of justice and leaving out essential elements inherent in human interaction, social justice is blocked from doing what liberalism requires of it: enabling effective autonomy. To this effect, Nussbaum presents an elaborate list of basic human capabilities, a minimum threshold of which should be at every individual’s grasp. These capabilities are necessary for a flourishing life of “truly human functioning” (Nussbaum 2000, 13) identified in a historically sensitive internal essentialism (Nussbaum 1992, 207). An individual’s autonomy rests on the capabilities to function in a way she chooses. For Nussbaum, enabling agency means paying attention to the “animal and material underpinnings of human freedom” (Nussbaum 2006, 88). Nussbaum has broadened the Aristotelian principle which Rawls introduced to become a normative justification of the “Marxian/Aristotelian idea of truly human functioning” (Nussbaum 2000, 13). Autonomy of a person means that she develops capabilities with the help of justice in order to achieve her own version of functioning.
There are three aspects of the capability approach which indicate, however, a trade-off in negative freedom. In other words, the capability list, through its “thickness”, might just impose on the individual concepts of the good life. The first difficulty here is that of autonomy and preference. There is tension between the social responsibility of justice to enable a person’s flourishing life and the individual responsibility for one’s well-being inherent in autonomy. Traditionally, liberalism has tried to expand the latter as much as possible (Alexander 2007, 107). Nussbaum tries to uphold this dimension through choice of relevant functioning. This means that once the capability provided, an individual can choose if and how to mobilize the relevant functioning attached to it (such as “being able[...] to play“ and not “to play“). In her reasoning, termed internal essentialism, identifying essential, common functionings in human life is the basis for her normative approach. Only in a second step does she then reduce the demands placed on social justice and requires it to provide only the capabilities to function in order to conserve the liberal nature of her theory (see Den Uyl and Rasmussen 2009b). In so far as capabilities are abstracted from functionings, they contain a certain concept of what the good life entails, already providing a frame limiting individual choice. Nussbaum brings choice into line with her essentialist account of a flourishing human life by qualifying preferences. At the same time as the capability list is “sensitive to people’s actual beliefs and values” (Nussbaum 2000, 158), certain preferences which are destructive in the sense that they go against the type of flourishing life envisaged are put off as “adaptive preferences”. Adaptive preferences are qualified as such because they are thought to have not been formed on a basis of autonomous choice and actually limit a subject’s autonomy. Put more strongly, an individual might have formed self-destructive or harming principles in reaction to a warped and repressive social environment in which she has long been deprived of relevant capabilities. Basing her argument on the previously utilitarian theorist Richard Posner and in line with a feminist current on preferences (see Walker 1995), Nussbaum points out that it wouldn't be valid to take such warped preferences into account when leaving room for choice in the capabilities list, because they validate an “unjust” status quo (Nussbaum 2006, 73). This step of qualifying preferences is highly controversial from a liberal perspective. Nussbaum offers no theoretical standard for qualifying when preferences become adaptive and stop being expressions of agency. John Christman has forwarded such a possible principle, according to which preferences are not adaptive “when an agent is in a position to be aware of the changes and development of her character and of why these changes come about” (Christman 1991, 348). However, even this definition does not necessarily dismiss what in the capability approach are put off as adaptive preferences. Nussbaum is aware of the „problematic tension between a normative sorting of preferences and liberal-democratic values“ (Nussbaum 2000, 115). She has chosen the explicitly normative point of view of insisting on the capabilities list even when it goes against individual preferences. The way she justifies this is by putting autonomy and the major liberties by themselves on the capability list. In adition, as mentioned above, the functioning relative to a capabiltiy is up to an individual's choice, thus limiting the normative view of the good life inherent in the approach. John Christman formulates a standard for content neutrality on which to measure a justice theory’s respect of negative freedom: „While it may be an additional value to me that my free actions conform to what turn out to be the correct standard of value, it is hardly essential to the value of free action itself that my actions so conform.“ (Christman 1991, 358). On several occasions, Nussbaum admits to a teleological form of reasoning, in which she departs from a view of the good life through certain functionings and from there derives the capabilities (Nussbaum 1992, 222; 1997, 288; 2000, 87). Thus a life without functioning is not a flourishing life. There is value inherent in functioning itself. Den Uyl and Ramussen, adamantly question the approaches’ liberal nature; they point out that it is conceptually very difficult to actually define the limit between capabilties and functionings (Den Uyl, Rasmusen 2009a). In an empirical application of the capability list, this differentiation would become even more problematic. Nussbaum allows for numerous exceptions to impose functionings, for example for children and the mentally impaired. Paired with the incomensurability of the capabilities on the list (no trade-off is allowed between them), for economic reasons and aspects of justice including taxation and resource redistribution it is easy to imagine situations in which it would be more socially legitimate to provide directly for functioning. In its application, there is a strong case to be made for questioning the capability approache's neutrality and respect of individual choice and responsibility. Amartya Sen's refusal to define a list of capabilities points to this ambiguity when attempting to provide for effective autonomy (Sen 2003, 45).
What political liberalism means for cosmopolitan justice
Since the 1990s, Thomas Pogge has prominently advanced a contractual cosmopolitan concept of justice founded on Rawls’ idea of justice as fairness (Pogge 2002). This approach uses the idea of a global contract between individuals to morally justify universal rights or redistributive measures between nations. At the same time, the contractualist approach places far-reaching moral requirements on participation when elaborating transnational principles of justice. Underlying the cosmopolitan global contract is the assumption that the present global structure and social institutions which bind individuals together across the globe exist already. This assumption makes the global contract vulnerable to the arguments which Rawls himself reasoned with when he refused to simply extend the contract within a politically liberal nation to the global scale. This constructivist perspective would bind liberal and non-liberal societies together in a liberal contract. The term “liberal contract” on a global scale can have two meanings. If stringently applying methodological individualism, it would have to impose political liberalism onto non-liberal societies. If applied to that of the communitarian unit, it would impose “higher-scale liberalism”: liberal freedom from constraints would apply to the existing plurality of groups and not within them. It is in light of this ambiguity of what a global liberal contract entails that Rawls's reasoning will be analysed. This will lead to viewing the capability approach of an attractive alternative to the constructivist theories. Its respect for both the freedom of individuals and communities allows us to avoid the contradiction between the scales of liberalism.
Liberalism is the approach most tolerant of plurality, leading Nussbaum to assure that “any universalism that has a chance to be persuasive in the modern world must, it seems to me, be a form of political liberalism.“ (Nussbaum 1999, 9) However, liberalism on a global scale has to defend itself against those authors for whom the liberal principle of toleration forbids its expansion. John Rawls is one of these authors. His realistic utopia of “The Law of Peoples” (Rawls 1999) elaborates on the moral principle of toleration towards reasonable non-liberal peoples. Toleration has liberated itself of the methodological individualism on which political liberalism is based and is given priority over liberalism itself. The asymmetry in comparisson to justice as fairness is that this new concept respects groups of people instead of individuals. These non-liberal societies can be ones with apartheid or other discriminatory practices (Caney 2002, 103). So if it seems obvious that this form of toleration „sanctions intolerance to liberal minorities stranded in decent societies“ (id., 100), why did Rawls take this measure ?
In his concept of a global contract, Rawls abandons methodological individualism, because diversity on a global scale is empirically structured through the national framework of peoples as national collectives – „[...] analogously to the liberal domestic case, Rawls must draw on the global public political culture to find ideas that can be acceptable to all“ (Wenar 2006, 102). In the global political culture, an applicable concept of justice isn’t one which would enforce political liberalism, because it includes nations such as (the fictive) Kazanistan, in which religion is merged with the state (Rawls 1999, 117-120). Rawls points out that applying a concept of justice globally is only feasible if it takes into consideration that political liberalism doesn’t dominate on a global scale. Rawls’ argument gives advantage to applicability, because existing collectives and national identities are taken into consideration. The critical issue then becomes toleration. Rawls views his global contract as tolerant in a way particularly pertinent to the ideal of political liberalism, which as a principle includes both the idea of rights and liberties and the idea of mutual toleration. Having abandoned the individualist standpoint, toleration now becomes toleration of decent, non-liberal peoples (Rawls 1999, 59-60). Nothing less than a humanitarian duty is set in terms of international justice in order to avoid interventionist possibilities (Hahn 2009, 91). By continuing to base his contract on the existing institutional framework, Rawls has stayed true to his approach in the concept of justice as fairness. By doing so, he has at the same time prioritised a statist vision of global justice and brought us towards a debate on the scope and coherence of political liberalism itself (Tan 2006, 76). Is liberalism supposed to be tolerant towards the status quo, even if it is the status quo is detrimental to pluralism because limiting individual liberty? Asking such a question involves seeing the two fundaments of global liberalism – toleration and plurality – as evaluative concepts. Solving this dilemma is also fundamental for justifying the capability approach, as it tries to extend and protect liberal rights and values on a global scale. If the individual is to be the ultimate measure, toleration as a liberal ideal loses any value as soon as the freedom to choose from various (non-liberal) doctrines is absent. Without plurality, there is no difference in the views of a good life which we need to tolerate. Liberalism on a “higher order” is based on plurality which has been made possible in the first place through toleration of plurality on a lower scale. The right, freedom and empowerment to choose in a pluralist environment therefore is prior to and more fundamental to liberalism than is toleration. Rawls’ step away from the individual can therefore be put into question: “[...] the scope of liberal toleration (inside a society for Rawls) does not and cannot extend to alternatives to liberalism itself” (Tan 2004, 79).
Even if from an internal perspective putting tolerance over methodological individualism seems unjustified, there is another aspect of the argument which must be considered. This aspect concerns cultural relativism. The relativist claim is a radical form of pluralism; its challenge is “based upon the claim that moral truth varies with cultural tradition, hence there is no universal standard of justice.” (Moellendorf 2002, 111).This is more or less what Rawls has accepted by prioritizing toleration. Cultural relativism is based on the argument for respecting cultural differences and local traditions. For Nussbaum, “allowing a local tradition to shape people's life chances pervasively“ (Nussbaum 2006, 253) is just as morally contingent as all the other factors Rawls eliminates in the original position. The basic fallacy of cultural relativism is that it conceives of culture as a homogeneous, closed off entity. In reality, cultures are ever-changing and interacting. They harbor internal conflicts and contradictory tendencies such as liberalism as well as patriarchism. Practical reasoning and the idea that each person has rights is not a particularly Western idea (Nussbaum 2006, 254). At the same time, there can be no such thing as a moral concept originating purely from within a single culture.
A reason for insisting on minimum human rights instead of full blown justice for Rawls is to prevent interventionism. He prefers “an absolutist interpretation of the principle of nonintervention“ as „morally superior to the license that the cosmopolitan conception gives to intervention“ (Moellendorf 2002, 111). Apart from violent interventionism, exchange between cultures and societies (something that already takes place) should enable discussion just as it is encouraged between the followers of different comprehensive doctrines in the domestic case (see Rawls 1999, 131-2). The capability approach attempts to be a universal concept of justice while at the same time respecting the principle of state sovereignty. Nussbaum has incorporated toleration of individuals and thus pluralism into her cosmopolitan approach. Instead of a minimum standard of human rights (Rawls 1999, 65), global justice requires a minimum standard of capabilities. The modern nation state is the framework for implementing justice. By distinguishing between “what we can justify morally for all and what we are morally entitled to implement“ (Nussbaum 2006, 260), only positive measures free from constraint are allowed to be employed towards justice in another state.
More specifically, such positive measures are based on the different aspects of pluralism in the capability approach. Plurality is a founding concept in each individual capability, taking into account the plurality of an individual's circumstances and different ways of functioning. The list of capabilities, itself representing a „multiplicity of fundamental principles“ (Alexander 2004, 459) is a minimum threshold, which can be further elaborated and complemented depending on the relevant context. Open ended, the capability list is open to interpetation in an intercultural dialogue. Because of the „multiple realizability“ of her list, it seems „able to account for a variety of cultural, religious and political backgrounds.“ (Jömann, Kurbacher and Suhm 2001, 66). An overlapping consensus would be possible as soon as the capabilities list becomes a shared moral value. The process of elaborating an overlapping consensus also provides a dimension of acceptability to the capability list even when Nussbaum's internal essentionalist normative foundation is rejected. From the abstract starting point of human worth, a reflexive balance is reached in the context of a free and reflective discourse by comparing the „fixed points“ of an individual's moral intuition (Nussbaum 2000a, 101). When talking about a global consensus, stability is an importan issue. For Nussbaum, the capability approach builds on elements implicitly inherent in all relevant political traditions. For her, Rawls' conditions of stability for an overlapping consensus are too far reaching and unrealisable even in a liberal, Western society. On the other hand, „long-standing seeds“ (Nussbaum 2006, 304) of liberal ideas allow hope for an eventual consensus even from individuals in non-liberal societies. To a certain extent, the more or less global acceptance of basic human rights provides a precedent for a dialogue with the potential for a global overlapping consensus. Alexander even goes so far as to claim that the discourse on human rights would come closer to consensus through adding on the capabilities approach. The idea of a global consensus adds practical legitimacy to the toleration of “bottom-up“ plurality through discoursive practice and agency conferred to the individuals involved in elaborating on the capabilities list.
There is a group of theorists critical of universal justice who would likely agree with Rawls that “higher-order“ liberalisms necessitates tolerance of groups of people in themselves and not just individual people. Unlike cultural relativists, liberal nationalists question the universal and not the liberal side to liberal global justice theories. They do not question methodological individualism but rather argue that in the name of individual rights community rights should be accepted as morally relevant to the individual in liberal justice. By conceding rights to cultural communities, liberal nationalism has sought to justify the moral priority of groups in contrast to global justice (see Brock 2002, 307-327). Even though prioritizing the national level communities, liberal nationalism does recognize a fundamental responsibility for justice between individuals in its approach. As long as a cosmopolitan approach respects the rights of communities along with global justice, the two approaches are not necessarily contradictory. As a moderate form of cosmopolitanism (Tan 2004, 115-9), the capability approach respects the demands of particularist solidarity. Moderate cosmopolitanism does not insist on cosmopolitanism as the only moral value. This position also advocates mild cosmopolitanist distributive justice (Caney 2001, 974-75). The capability approach only insists on the existence of global distribution but does not exclude other distributive measures, as the radical cosmopolitanism of Charles Beitz (quote). At the same time, affiliaition in all its emotional and cognitive forms is a central value of the capability approach. Compassion and care-structures are important, along with different spheres of attachement, (Nussbaum 2001, 300 and 2006, 321). It also recognizes the collective rights of certain groups, always in accordance with individual self-determination and expects the capabilities list to be implemented as constitutional principles elaborated through national politics (Nussbaum 2000a, 105). Recalling the anthropological foundations of the capability approach which acknowledges vulnerability and independence, the state is relevant for liberal reasons, because it expresses choice and autonomy. Where Nussbaum's cosmopolitan approach and liberal nationalism diverge is on the importance of the other spheres of justice and on the minimum level of global redistribution which is an added responsibility for prosperous nations.
Allowing pluralism on multiple levels, but without abandoning methodological individualism, Nussbaum's theory seems to deal with today's global empirical situation just as well as does Rawls. In the name of toleration, she refuses to simply impose global justice but rather relies on an inclusive discourse to produce an eventual global consensus. The openness of the capability approach and the fact that it can be founded on various normative concepts also respects Rawls' criterion of liberalism, which is to be specific in form but less in content. Redistributive measures, always a highly contentious issue, are not removed from the internal political control of citizens in nation states. In conclusion, the capability approach has found a more balanced approach to the two principles of toleration and pluralism of “higher-order“ liberalism. The weakness of “human capabilities liberalism“ (Den Uyl, Rasmussen 2009a, 875) lies more in the individual sphere. Deciding on an individual version of the good life is not completely left to individual autonomy by the neo-aristotelian approach to capability This leads Den Uyl and Rasmussen to the conclusion that it does „not actually advance the cause of liberal properly understood but actually retreats from it“ (ibd. 876). This in turn affects the global applicability of the capability approach. An important aspect of the Rawlsian overlappings consensus is that it includes reasonable, but not necessarily liberal, points of view. Liberalism is confined to the political sphere to which social justice belongs and as such can be accepted by reasonable, non-liberal doctrines because it doesn't challenge their principles in the realm of the good life. Recalling the arguments from the first section, this type of consensus is solely possible insofar as the capabilities list is not seen to be too “thick“. Only if the capabilities approach is accepted as liberal enough in that it is based on essential moral intuitions acceptible by all reasonable comprehensive doctrines can one hope to reach a global consensus.