This paper discusses the relationship between justice and identity. While it is widely agreed that justice requires us to go beyond loyalty to our simplest identity – being just oneself – there is less common ground on how far we must go beyond self-centredness. How relevant are group identities to the requirements of justice, or must we transcend those too? The author draws attention to the trap of confinement to nationality and citizenship in determining the requirements of justice, particularly under the social-contract approach, and also to the danger of exclusive concentration on some other identity such as religion and race. He concludes that it is critically important to pay attention to every human being's multiple identities related to the different groups to which a person belongs; the priorities have to be chosen by reason, rather than any single identity being imposed on a person on grounds of some extrinsic precedence. Justice is closely linked with the pursuit of impartiality, but that pursuit has to be open rather than closed, resisting closure through nationality or ethnicity or any other allegedly all-conquering single identity. Christian List

In criticizing communitarian views of justice, Amartya Sen argues that identity is not merely a matter of discovery but an object of reasoned choice subject to constraints. Distinguishing three notions of identity – self-perception, perceived identity and social affiliation – I claim that the relevant constraints implied by this argument are minimal. Some of Sen's arguments about perceived identity and social context do not establish any further constraints. Sen also argues that a model of multiculturalism and some forms of education can restrict, or fail to promote, reasoned and informed identity choice. This argument is better understood in the light of Sen's work on capability and justice, notably his concern with ways in which underdogs can adapt and his emphasis on public reasoning. It highlights limitations on information and opportunities for reasoning. I suggest that these lead to genuine constraints on (reasoned and informed) identity choice. The paper focuses on Sen's work, though its claims are also relevant to George Akerlof and Rachel Kranton's analysis of identity.

This paper presents Sen's theory of agency, focusing on the role of commitment in this theory as both problematic and potentially illuminating. His account of some commitments as goal-displacing gives rise to a dilemma given the standard philosophical theory of agency. Either commitment-motivated actions are externally motivated, in which case they are not expressions of agency, or such actions are internally motivated, in which case the commitment is not goal-displacing. I resolve this dilemma and accommodate his view of commitment as motivation by developing a broader descriptive theory of agency, which recognizes both agent goal-directed and goal-displacing commitments. I propose a type of goal-displacing commitment, which I call ‘tacit commitment’, that can be seen to fit between the horns. Tacit commitments regulate behaviour without being made conscious and explicit. This resolution suggests a means of bridging the normative/descriptive gap in social-scientific explanation.

A capability approach prescribes paternalist government actions to the extent that it requires the promotion of specific functionings, instead of the corresponding capabilities. Capability theorists have argued that their theories do not have much of these paternalist implications, since promoting capabilities will be the rule, promoting functionings the exception. This paper critically surveys that claim. From a close investigation of Nussbaum's statements about these exceptions, it derives a framework of five categories of functionings promotion that are more or less unavoidable in a capability theory. It argues that some of these categories may have an expansionary dynamic; they may give rise to widespread functionings promotion, which would defeat the capabilitarian promise that paternalist interventions will be exceptions to the rule of a focus on capabilities. Finally, the paper discusses three further theoretical issues that will be decisive in holding this paternalist tendency in check: how high one sets threshold levels of capability protection, how lengthy one's list of basic capabilities is, and how one deals with individual responsibility for choices resulting in a loss of one's capabilities.

Capability theorists have suggested different, sometimes incompatible, ways in which their approach takes account of the value of freedom, each of which implies a different kind of normative relation between functionings and capabilities. This paper examines three possible accounts of the normative relation between functionings and capabilities, and the implications of each of these accounts in terms of degrees of paternalism. The way in which capability theorists apparently oscillate between these different accounts is shown to rest on an apparent tension between anti-paternalism (which favours an emphasis on capabilities) and anti-fetishism (which favours an emphasis on functionings). The paper then advances a fourth account, which incorporates a concern with the content-independent or ‘non-specific’ value of freedom. Only the fourth account would remove all traces of paternalism from the capability approach. Whatever reasons advocates of the capability approach might have had for rejecting this fourth account, those reasons are not internal to the capability approach itself.