Abstract

The impressive growth of antidiscrimination law in liberal democracies in the past few decades belies the inadequacy of the normative bases on which it has been sought to be justified. Popular ideals such as rationality, equality and dignity have been unsuccessful in providing a coherent liberal framework for the fundamental aspects of the practice of antidiscrimination law. In this thesis, I have argued that a unified normative framework comprising autonomy and dignity-as-autonomy does a markedly better job of justifying the most fundamental aspects of these laws. The ideal of personal autonomy is understood here as a principle that seeks to guarantee an adequate range of valuable options to individuals. Dignity-as-autonomy is understood to be an expressive norm, which forbids certain persons from expressing contempt for the autonomy of another. These ideals have different forms: autonomy is a non-action-regarding principle, while dignity-as-autonomy is action-regarding. They are also distinct substantively: it is often possible to violate one of them without affecting the other. When these ideals make incompatible demands, I argue that those made by autonomy should prevail. Mandating positive action and reasonable accommodation on the one hand, and prohibiting indirect discrimination and harassment on the other, are essential features of a model of antidiscrimination law based on this framework. Further, under this framework, antidiscrimination law is not vulnerable to objections such as ‘levelling down’ and responds well to claims of discrimination on ‘intersectional grounds’. Furthermore, it is not essential to find an ‘appropriate comparator’ in order to prove discrimination. This model also explains when, and under what conditions, can some forms of discrimination be ‘justified’. Finally, on an autonomy-based model, antidiscrimination law is only one of several complementary tools that should be employed to protect and promote personal autonomy.