How much must we sacrifice in order to stop strangers from suffering serious losses, and does distance from them alter our obligations? When may we harm some people to help others? How can we best reason about these issues? These are three general questions—the first two are substantive ones, the third a methodological one—that Peter Unger discusses in his book Living High and Letting Die (hereinafter LHLD) and that I discuss in this article.1

Every theory of justice requires a first-order theory specifying principles of justice, and a second-order view explaining why those principles constitute the correct principles of justice. According to John Rawls, political liberalism is committed to the two principles of justice specified in its first-order theory, “justice as fairness.” Justice as fairness, according to Rawls, in turn presupposes the second-order view that justice is a political conception. A political conception of justice treats the principles derived from the fundamental ideas in the public political culture as the correct principles of justice. Political liberalism, however, nowhere offers a defense of the view that justice is a political conception. Indeed, it even strives to avoid the admission that it presupposes that justice is a political conception by stating only that it uses a political conception of justice, while allowing that justice might not actually be a political conception. As to the truth of its second-order presupposition, political liberalism chooses to remain agnostic. Rawls claims that political liberalism has no choice at all. To do otherwise, he argues, would lead to an internal contradiction.

Recently, legal and social thinkers have turned to the idea that actions possess a nonlinguistic meaning, called “expressive meaning.” In this article I examine the idea of expressive meaning and its role in legal reasoning. My focus is on a series of U.S. Supreme Court cases involving constitutional challenges to election districts drawn on the basis of race. The Supreme Court used the idea of expressive meaning in striking down the districts. After explicating the idea of expressive meaning, I explain and criticize the Court’s reasoning. I distinguish the approach of Justices Thomas and Scalia, who hold that all uses of race in districting do constitutional harm, from that of Justice O’Connor, who distinguishes uses of race that do constitutional harm from those that do not. I contend that Justice O’Connor is right to make the distinction but she draws the line using a questionable standard. A more defensible standard would be more accommodating to the districts that the Court invalidated.

In “The Model of Rules I,”1 Ronald Dworkin argues that judges are often bound by principles that do not derive their authority from having been formally promulgated by a judge or legislature in accordance with a rule of recognition. But, on Dworkin’s view, the existence of such principles is inconsistent with positivism’s pedigree thesis, according to which propositions are legally valid solely in virtue of having been promulgated in accordance with a rule of recognition. Thus, Dworkin concludes that positivism is false.