In procedural-law scholarship as well as in the theoretical analysis of the notion of proof as a result of the joint assessment of all items of evidence introduced in a trial, reference is frequently made to notions such as the conviction, belief, or certainty of a judge or a jury member about what happened. All these notions underscore the mental states involved in the process of determining the facts on the part of a judge or a jury. In this analysis, I look at the links between beliefs and the justification in the findings of fact provided by the judge or jury in her or its verdict.

In a recent paper, “How Facts Make Law,” I launch an attack on a fundamental doctrine of legal positivism. I argue that nonnormative facts cannot themselves constitutively determine the content of the law. In a response published in this journal, Ram Neta defends the view that nonnormative social facts are sufficient to determine normative facts, including both moral and legal facts. Neta's paper provides a useful opportunity to address a spelled-out version of this view, which in various forms is widely held in philosophy of law and other areas of philosophy. I begin by addressing Neta's attempts to show that descriptive facts can alone determine moral and legal facts. First, Neta's account of why it is wrong to break promises fails. In addition to other problems, it begs the question by taking for granted that a person's desires or other motivational states necessarily justify the actions that they motivate. Next, I turn to Neta's attempt to provide a counterexample to my view about law. In my original paper, I claim that the nature of the constitutive determination relation in the legal domain is what I call rational determination. Roughly speaking, a full constitutive account of the legal facts must include reasons that explain the relation between the determining facts and the legal facts. The facts on which Neta's putative counterexample depends cannot be reasons of the required sort because they take for granted what they are supposed to explain—the way in which nonnormative social facts contribute to the content of the law. Finally, I address the larger issue of how far my argument applies to other domains. I consider and reject Neta's argument that purports to show that all normative domains have the relevant features of the legal domain. I then sketch a competing picture of some normative domains.

This paper questions the fairness of our current tort-law regime and the philosophical underpinnings advanced in its defense, a theory known as corrective justice. Fairness requires that the moral equality and responsibility of persons be respected in social interactions and institutions. The concept of luck has been used by many egalitarians as a way of giving content to fairness by differentiating between those benefits and burdens that result from informed choice and those that result from fate or fortune. We argue that the theory of corrective justice, along with its institutional embodiment of tort law, is at odds with an egalitarian commitment to fairness because it allows luck an unjustifiable role in determining dissimilar liability for similar wrongs and dissimilar compensation for similar losses to bodily integrity. Many egalitarian political theorists have also recognized, if not defended, the notion of distinct forms of justice, namely corrective, retributive, and distributive. Although theorists of these different forms of justice have been concerned with negating unfair luck inside the operations of each form of justice, there has been little attention to the way in which luck operates to sort cases into each form of justice. We claim that there is a significant way in which luck operates to subject different people to principles of corrective, retributive, and distributive justice—thereby assessing dissimilar liability for similar wrongs and disparate compensation for similar losses—which flies in the face of the egalitarian value of fairness. After surveying the arguments put forward by theorists defending a categorical distinction between corrective justice and retributive and distributive principles, we argue that although analytical distinctions can be made between different forms of justice (although, we also suggest that the distinctions are not as sharp as some commentators suggest), there is no good reason to defend an acoustic separation between these forms of justice when doing so creates unfair outcomes.

What does “speech” mean in constitutional First Amendment law and in ordinary language and the philosophy of language? Under what circumstances does intentional action count as speech? Can communication be unintentional? And what follows (in law) from the fact that almost any action can be made expressive?

This essay addresses these and related questions.

Many contemporary legal positivists have argued that legal theory is evaluative because it requires the theorist to make judgments of importance. At the same time they argue that it is possible to know “what the law is” without resort to evaluative considerations. I distinguish between two senses of “what the law is”: in one sense it refers to legal validity, in another to the content of legal norms, and I argue that legal positivism is best understood (as indeed some legal positivists have explicitly said) as a claim about legal content. Understood this way, however, it is open to the objection that knowing the content of legal norms requires evaluative considerations for reasons similar to those offered by positivists for thinking that legal theory is requires evaluative considerations. I then distinguish between evaluative considerations in general and moral considerations and argue that because of the subject-matter of legal norms, there are good reasons for thinking that it is moral considerations, and not just any other evaluative considerations, that are required for knowing the content of legal norms.

In this article, I examine several distinctions that may be relevant to the morality (and conceptual characterization) of terrorism: (1) the state/nonstate agent distinction, (2) the combatant/noncombatant distinction, (3) the intention/foresight distinction, (4) the means/side-effect distinction, (5) the interrelated necessary/nonnecessary means and produce/sustain distinctions, (6) the mechanical/nonmechanical use distinction, (7) the military/political distinction, (8) the harm/terror distinction, and (9) the harm-for-terror/terror-for-goal distinction. I conclude that some of these factors (though not those most commonly cited) account for the prima facie wrongness of terrorism and that the nondistinctive properties of terrorism (which it shares with some nonterrorist acts) are what make it most seriously wrong. I also provide a conceptual examination of terrorism as we commonly think of it and its relation to torture. In the course of discussing the distinctions and also in concluding the article, I consider why terrorism may sometimes be morally permissible.

Private law is generally formulated in terms of right–duty relations, and accordingly, private-law claims are understood to arise from breaches of duty, or wrongs. Some claims are not easy to explain on this basis because the claim arises from an act that the defendant was justified in doing. The violation/infringement distinction seems to offer an explanation of such claims, but it is argued that the explanation is illusory. Claims of this sort are best understood as based not on a primary right–duty relation at all but on a “primary liability” or “right–liability” relation. A primary-liability claim is not a claim arising from the breach of a strict-liability duty. The recognition of primary-liability claims does not involve skepticism about duties or rules or legal relations and it is consistent with the analysis of private law in terms of corrective justice.

In this paper I address the issue of the normativity of instrumental rules (for example, legal rules). On the one hand, I criticize Scott Shapiro's “constraint model” of instrumental rule-following according to which instrumental rules have motivational clout qua rules: the agent conforms to them simply because they are rules. On the other, I argue for a purely epistemic account of instrumental rule-following. According to this account, instrumental rules inform the agent which action she is required to perform but do not get her to act. Rather the agent is motivationally guided by the reasons for which she adopted the instrumental rules.

A central tenet of positivism is that social practices are at the foundations of law. This has been cashed out in a variety of ways. For example, Austin argues that, among other practices, a habit of obedience to a sovereign is at the foundations of law, and Hart argues that at the foundations of law is the converging attitudes and behaviors of a class of relevant officials. Since Hart, some prominent positivists have employed either David Lewis's analysis of conventions or Michael Bratman's theory of shared cooperative activities to develop new accounts of the social practices that are at the foundations of law, whatever those foundations might be. In this paper, I identify five features characteristic of the Lewisean and Bratmanian models of social facts—models of what I call hypercommittal social practices. I then show that models of social facts that have these features ought not to be used to explain the way in which a social practice is at the foundations the law. I conclude that hypercommittal social practices such as Lewisean conventions or Bratmanian shared activities are not at the foundations of law.

This article strives to articulate some of the conventional foundations of law. The article consists of three main arguments. First, between H. L. A. Hart's conception of the rules of recognition and H. Kelsen's theory of the basic norm, Hart's view is more compelling. Second, the paper argues that the rules of recognition are social conventions but not of the familiar coordinative type, as some commentators have claimed. Finally, the article draws a distinction between deep and surface conventions, arguing that there are some deep conventions about what law is, and surface conventions that determine what counts as law in a particular community. This distinction between deep and surface conventions is employed to solve some of the puzzles about the nature of the rules of recognition.

This essay presents a new theory of statutory application that is superior to two competitors. One of the competitors claims that statutory directives apply to act-tokens fitting the legislature's intention. The other holds that statutes apply to act-tokens that are of the genuine kinds named by the classifying words. These theories solve certain problems badly or not at all, respectively: (1) accounting for the capacity of statutes for epistemic guidance; and (2) avoiding literalism. Both do a limited job of accounting for the rationality of statutory promulgation. The theory developed here, social salience theory, does better on all counts. According to social salience theory, the default position is that statutes apply to act-tokens that are socially salient, given the linguistic and social contexts. The departure from that position is also social, produced by conventions to defer to experts on the statute's applications to particular cases.

This essay consolidates some fragments of the contemporary theory of expressive freedoms, bringing together scattered conceptual distinctions (e.g., hurting and harming, tolerating and legitimating) and moves (e.g., the need to rectify hateful speech and to constrain harmful actions legally) into an account that is sensitive to the needs of abused groups but faithful to the libertarian tradition associated with Mill's harm principle. Accepting this principle as the fundamental condition warranting legal control of action, we explore legislative responsibilities for protecting expressive freedoms through three additional presumptions. The chief of them is an expectation of epistemic rationality that limits occasions of harm requiring legal action. A presumption of enablement over punishment shows that criminal sanctions for hateful speech are almost never appropriate. A presumption of multiple responsibilities accepts that social sanctions may be appropriate where political coercion is not.

[Originally published in 11 LEGAL THEORY 363–385 (2005)]