1. Paul Schenck and Dwight Saunders, Petitioners v. Pro-Choice Network of Western New York, et al. No. 95–1065. United States Supreme Court Official Transcript Wednesday, October 16, 1996. 1996 LW 608239 (U.S.Oral.Arg.), at 23. Although it is the policy of the Supreme Court not to identify which justice poses which questions, since they are ostensibly posed by the Court as a whole, the context often makes clear who the questioner is. And, where it does not, reporters, such as National Public Radio's Nina Totenberg, sometimes give the questioner's name, finding this information newsworthy because it can indicate the way in which a justice may be inclined to vote on a case. In the above exchange, Totenberg quoted Scalia as making the first statement, and the transcript clearly indicates that Breyer made the last statement, and it is a matter of speculation which justices made the statements in between.

2. Cohen, Joshua, Freedom of Expression, 22 Phil. Pub. Aff. 207–63 (1993).Google Scholar

3. I do not address here the question of indirect injury in which a speaker says something to a listener who, as a result, harms a third party. This kind of harm is discussed insightfully in Schauer, Frederick, The Phenomenology of Speech and Harm, 103 Ethics, 642–46 (1993).CrossRefGoogle Scholar

4. For overviews of recent work in philosophy of mind, see Rosenthal, David M., ed., The Nature of Mind (1991)Google Scholar, and Warner, Richard & Szubka, Tadeusz, eds., The Mind-Body Problem: A Guide to the Current Debate (1994).Google Scholar

5. I acknowledge that this characterization of Scalia's interpretation of the First Amendment is not exactly fair, as it is based on an impromptu comment made during oral arguments, possibly even in the guise of devil's advocate. His position in R.A.V. v. St. Paul, 120 L.Ed.2d, 305 (1992)Google Scholar, is certainly more sophisticated, although, even there, he underestimates the differential harm inflicted on victims of, for example, racist “fighting words” as opposed to other kinds.

6. Mill, John Stuart, On Liberty 9 (1978).Google Scholar

7. Schauer, Frederick, Free Speech: A Philosophical Enquiry 7–8 (1982).Google Scholar To state what a free-speech principle requires is not to state that such a principle is justified. In recent writings, Schauer has evinced a certain amount of skepticism about whether a distinct principle of free speech can be defended. In The Phenomenology of Speech and Harm, supra note 3, for example, he notes that his conclusion, viz. that we should reject the hypothesis that speech, as a class, causes less harm than nonspeech conduct, “puts more pressure on the positive arguments for a free speech principle, and perhaps no such argument will turn out to be sound,” at 653. For insightful arguments against the existence of any justifiable general principle of free speech, see Alexander, Lawrence & Horton, Paul, The Impossibility of a Free Speech Principle, 78 Nw. U. L. Rev. 1319–57 (1983)Google Scholar and Fish, Stanley, There's No Such Thing as Free Speech—and Ir's A Good Thing, Too (1994).Google Scholar

8. This view is defended by, among others, Scanlon, Thomas, A Theory of Freedom of Expression, 1 Phil. Pub. Aff. 204, 204–26, (1972).Google Scholar

9. To see this quickly, let P=causes physical hurt; R=can be legitimately restricted. (1) can be symbolized as x (Px→Rx) or, for any given instance of x, Pa→Ra; (2) can be symbolized as x (-Px→-Rx) or, for any given instance of x, -Pa→-Ra. It is a simple logical error to infer (2) from (1), or vice versa.

10. Mill, , supra note 6, at 9.Google Scholar

11. An immunization I choose to have may (momentarily) hurt me physically, but it does not constitute a harm if I have chosen to be immunized. And defamatory statements that damage my reputation may harm me even if they cause me no physical or even psychological injury (as in the case in which I don't find out about them). See Feinberg, Joel, The Morai Limits of the Criminal Law: Volume I. Harm to Others (1984).Google Scholar

12. This statement of the principle is still overly simplified. Among other things, in order for the restriction to be justified, it must not itself cause a greater or unacceptable variety of harm.

13. Feinberg, , supra note 11.Google Scholar

14. Furthermore, not all physical assaults cause physical hurts. Contrary to Justice Scalia's claim that “sticks and stones will break my bones” (my emphasis), they may do so, but, then again, they may merely graze me or hit with such slight impact that I am not, in Feinberg's sense, harmed, since my interests are not invaded.

15. Feinberg, , supra note 11, at 37.Google Scholar

16. See id. at 37.

17. This is the term used by Cohen, Joshua, supra note 2.Google Scholar

18. 578 F.2d 1197 (1978), at 1210.

19. The full proverb, rarely quoted in its entirety, is “Beneath the rule of men entirely great/The pen is mightier than the sword.” Bulwer-Lytion, Edward George, Richdieu, act 2, scene 2, line 307 (1839).Google Scholar Under such a wise and benevolent regime, one can imagine there being no use for swords.

20. Lawrence, Charles R. III, If He Hollers, Let Him GoGoogle Scholar, in Words That Wound: Critical Rvi Theory, Assaultive Speech, and the First Amendment (Matsuda, Mari et al. eds., 1993).Google Scholar

21. 340 U.S. 290.

22. Concurring opinion in Whitney v. California, 274 U.S. 357 (1927), at 377.Google Scholar

23. In Frederick Schauer's terminology, the view that speech either has no costs or else lower costs, as a class, than nonspeech conduct is the “lesser harm hypothesis.” See Schauer, , supra note 3.Google Scholar

24. Here again I am referring only to the direct injury caused by verbal assaults, not to the indirect injury that is alleged to result when someone's speech causes a listener to physically assault a third party. Throughout this article, I use the nouns “hurt” and “injury” interchangeably, continuing to distinguish them both from harm, as defined by Feinberg.

25. Schauer, , supra note 3, at 648.Google Scholar

26. For a clinical description of PTSD, see Diagnostic and Statistical Manual of Mental Disorders (3rd ed. rev., 1987).Google Scholar

27. For discussions of such harms, see Brison, Susan, Surviving Sexual Violence: A Philosophical Perspective, 24 J. Soc. Phil. 5–22 (1993)CrossRefGoogle Scholar; Herman, Judith, Trauma and Recovery (1992)Google Scholar; Janoff-Bulman, Ronnie, Shattered Assumptions: Towards a New Psychology of Trauma (1992).Google Scholar

28. For examples, see Matsuda, et al. , Words That Wound, supra note 20.Google Scholar

29. One could give similar analyses of the other characteristics of pain described by Bentham (The Principles of Morals and Legislation): certainty, propinquity, fecundity, and purity. It could be argued, along the lines suggested above, that psychic and physical pains cannot be distinguished along these dimensions either.

30. Schauer, , supra note 3, at 635–53.Google Scholar

31. Rutzick, Mark C., Offensive Language and the Evolution of First Amendment Protection, 9 Harv. C.R.-C.L.L. Rev. 7 (1974).Google Scholar

32. Id.

33. The Supreme Court used the term “sensibilities” in Street v. New York, asserting that the state had “an interest in protecting the sensibilities of passers-by who might be shocked by appellant's words about the American flag.” 394 U.S. 576, at 591.

34. This is done by connecting a pair of electrodes to the subject's skin and to a polygraph. (See Damasio, Antonio, Descartes' Error 207 (1994).Google Scholar

35. Rutzick, , supra note 31.Google Scholar

36. In Freedom of Expression, Joshua Cohen rejects no-cost minimalism because of its denial of the direct and indirect costs of speech. But he does not present arguments intended to persuade those who deny such costs, (supra note 2).

37. Post, Robert, Managing Deliberation: The Quandary of Democratic Dialogue, 103 Ethics 654–78 (1993).CrossRefGoogle Scholar Post lists an additional kind of harm, which he calls the “intrinsic harm,” or the “elemental wrongness” of hate speech. Such expression, on Post's view, violates the respect for equality that is at the core of the Fourteenth Amendment. I believe this reduces to harm to individuals or harm to groups, since the “respect for equality” is not the sort of thing that can be harmed, given the definition of “harm” I am using.

38. From Matsuda, et al. , supra note 20, at 67.Google Scholar The Stanford code was struck down by the California Superior Court of Santa Clara County (in 1995) on the grounds that the code restricted constitutionally protected speech. A California state law known as “the Leonard Law” requires even private educational institutions in California to abide by the U.S. Constitution.

39. This definition of “fighting words” comes from Chaplinsky v. New Hampshire, the 1942Google Scholar U.S. Supreme Court case upholding the conviction of a Jehovah's Witness for violating a New Hampshire law forbidding any person to address “any offensive, derisive or annoying word to any other person who is lawfully in any street or other public place.”

40. Several arguments against the use of the “fighting words” doctrine in hate-speech codes are (a) the doctrine can be used to provide counter-speakers with a heckler's veto; (b) it presupposes that the only reason for regulation of hate speech is the public interest in preserving the peace, whereas a more compelling reason is to secure the intended victim's right to equal protection; (c) it fails to provide protection to those most vulnerable and least likely to fight back; and (d) it assumes that people can be harmed only by face-to-face insults.

41. Most people can be socialized, however, not to respond violently to even the most extreme provocation. It is consistent to hold people responsible for their actions while recognizing that their emotional reactions may be out of their control.

42. Whether of not we consider people able to control their reactions depends, in part, upon normative views about what they ought to be able to do. (See Schauer, , supra note 3.)Google Scholar

43. The view that the ability to reason is dependent on having the right degree and kind of emotion is defended in Damasio, , supra note 34, at 52–53Google Scholar, and in de Sousa, Ronald, The Rationality of Emotion (1987).Google Scholar

44. See Matsuda, et al. , supra note 20.Google Scholar

45. Robinson, Jenefer, 92 Startle, J. Phil. 64 (1995).CrossRefGoogle Scholar

46. See Damasio, , supra note 34, at 226.Google Scholar

47. Id., at 224.

48. For more examples of harmful hate speech, see Ehrlich, Howard J., Campus Ethnoviolence: A Research Review (1992)Google Scholar and Lederer, Laura J. & Delgado, Richard, The Price We Pay: The Case Against Racist Speech, Hate Propaganda, and Pornography (1995).Google Scholar

49. I am obviously struggling with terminology that presupposes the very mind-body dualism I am attempting to undermine. The best I can do for now is to put misleading terms in scare quotes, as this is the only language I've got (apart from one or two others, which are no better in this respect). One thing that makes the mind-body problem so intractable is the forced dichotomy between dualism and monism. The question is posed: Is the mind a physical thing—part of the body—or is it a mental thing, distinct from the body? Both of these alternatives are unsatisfactory, but no other options are given. The question—and any sensible answer to it—assumes that either dualism or materialism is true. However, as Wittgenstein observes, “Thinking is not an incorporeal process which lends life and sense to speaking, and which it would be possible to detach from speaking, rather as the Devil took the shadow of Schlemiehl from the ground.” Ludwig Wittgenstein, Philosophical Investigations 109 (Anscombe, G.E.M. trans., Macmillan 1953Google Scholar). Wittgenstein adds: “But how ‘not an incorporeal process’? Am I acquainted with incorporeal processes, then, only thinking is not one of them? No; I called the expression ‘an incorporeal process’ to my aid in my embarrassment when I was trying to explain the meaning of the word ‘thinking’ in a primitive way.” I thank Robert Fogelin for bringing this passage to my attention.

50. Cohen (supra note 2) rejects maximalism because of what he perceives to be its sectarian commitment to the value of autonomy. My own view, defended in The Autonomy Defense of Free Speech, 108 Ethics (01 1998) 312–339Google Scholar, is that no plausible account of autonomy yields a defense of free speech. Here, I approach the issue from a different angle, asking, Why is it assumed (by some theorists, e.g., R. Dworkin, Nagel, and Scanlon) that assaultive speech does not cause harms of the sort that would count as violating someone's right to autonomy?

51. Scanlon, , supra note 8, at 204–26.Google Scholar

52. Brison, , supra note 50.Google Scholar

53. 721 F. Supp. 852 (Mich, E.D.. 1989).Google Scholar

54. 771 F.2d 323 (1985).

55. Id.

56. Likewise, any indirect injury resulting from the speech is considered to be under the control of, and hence the sole responsibility of, the audience members who choose to act on the basis of the speech. This presupposes, however, an implausible “bucket theory” of responsibility. The term “bucket theory” comes from Robert Nozick, who criticizes such accounts of responsibility in Anarchy, State, and Utopia 130 (1974). It refers to the view that there is only so much responsibility to go around, so if we hold the person who acts on the basis of the speech 100% responsible, we cannot also hold responsible the speaker who incites or in some other way causally contributes to the audience's action.

57. To see that not everything “mental” is under our control, consider the degree of control you possess over the perceptual images you have of your environment right now, or the degree of control you have over whether or not to believe something for which you have considerable inductive evidence, such as that the sun will rise tomorrow.

58. Thomson, Judith Jarvis, The Realm of Rights 253 (1990).Google Scholar

59. Unfortunately, we do often hold victims of some crimes responsible if they did not take what we consider to be reasonable precautions. Rape victims are still told, by some, that they were “asking for it” by their behavior, dress, location, etc.

60. Ellis, Anthony, Thomson on Distress, 106 Ethics (1995) 112–19.CrossRefGoogle Scholar

61. See Damasio, , supra note 34, at 40.Google Scholar I thank Ann Bumpus for drawing my attenuon to this book and to this passage.

62. Dorsen gave this response in the discussion of a paper he presented at New York University in 1988 or 1989 to the New York Group of the Society for Philosophy and Public Affairs. On the view he expressed, just as beauty is in the eye of the beholder, injury is in the psyche of the reader, viewer, or listener. A more defensible view of the “appropriate” response of hate-speech victims is offered by Kent Greenawalt “As with direct personal threats, it is doubtful whether ‘courageous citizens’ should be expected to swallow such abuse without deep hurt, and it is also doubtful whether being the victim of such abuse contributes to one's hardiness in ways that count positively for a democratic society.” Kent Greenawalt, Speech, Crime. and the Uses of Language 301 (1989).

63. Might it be that the practice of self-mutilation, surprisingly common in victims of abuse, comes from a desire to make one's inner pain visible and palpable? An interesting question to explore is. How does the physical pain, of, say, cutting oneself, make the psychic pain more bearable?

64. Langer, Lawrence, Holocaust Testimonies: The Ruins of Memory, 53 (1991).Google Scholar

65. Id. at 102.

66. Cohen v. California, 403 U.S. 15, at 25.Google Scholar

67. Alexander, Larry & Horton, Paul, The Impossibility of a Free Speech Principle, 78 Nw. U. L. Rev. (1984)Google Scholar, note that speech “does not denote any particular set of phenomena. Everything, including all human activities, can ‘express’ or ‘communicate’ and an audience can derive meaning from all sorts of human and natural events,” at 1322.

68. Goodman, Nelson, Ways of Worldmaking (1978).Google Scholar

69. This example comes from Sunstein, Cass, Social Norms and Social Roles, 96 Colum. L. Rev. 904–68 (1996).CrossRefGoogle Scholar

70. The first-person accounts presented by critical race theorists of the effects of racist hate speech have the potential to change dramatically our understanding of the harm of such speech, just as the testimonies of victims of rape, battering, and sexual harassment have, only recently, altered society's assessment of the harms of such forms of victimization. At the same time, as Martha Minow has pointed out (in Surviving Victim Talk, 40 UCLA L. Rev. 1411–45 (1993)Google Scholar), victim talk tends to spawn more victim talk, sometimes on the part of the victimizers, and not all claims to victimization are on a par.

71. See Brison, , Herman, Janoff-Bulman, supra note 27Google Scholar, as well as Elias, Robert, The Politics of Victimization: Victims, Victimolocy, and Human Rights (1986).Google Scholar

72. See supra note 68.

73. See Duff, R. A., Intention, Agency and Criminal Liability: Philosophy of Action and the Criminal Law (1990).Google Scholar

74. See Schauer, Frederick, Uncoupling Free Speech, 92 Colum. L. Rev. 1321–57 (1992)CrossRefGoogle Scholar, and Meyers, Diana T., Rights in Collision: A Non-Punitive, Compensatory Remedy for Abusive Speech, 14 Law & Phil. 203–43 (1995)CrossRefGoogle Scholar, for discussions of the unequal distribution of such costs and proposals for redistribution of the costs, including compensation for victims.

75. Matsuda, Mari, Public Response to Racist Speech: Considering the Victim's Story, in Words That Wound, supra note 20, at 48.Google Scholar

76. My position is that such laws should stand or fall together. I find it odd that, in striking down the Stanford hate-speech code, the court suggested that, as an alternative, “A penalty enhancement scheme along the lines of Mitchell might also be a means of eradicating racism and hate on campus without ‘adding the First Amendment to the Fire’” (quoting Scalia in R.A.V. v St. Paul), Corry v. Leland Stanford Junior University, Cal. Sup. Ct. at Santa Clara, case no. 740309, 02 27, 1995.Google Scholar

77. UWM Post v. Board of Regents of U. of Wisconsin, 774 F. Supp. 1163 (E.D. Wis. 1991)Google Scholar, note 9, at 1174.

78. Freedom of Thought as Freedom of Expression: Hate Crime Sentencing Enhancement and First Amendment Theory, Crim. Just. Ethics 39 (Summer/Fall 1992).Google Scholar

79. The policy ruled unconstitutional in Doe v. University of Michigan was called the Policy on Discrimination and Discriminatory Harassment of Students in the University Environment. 721 F. Supp. 852 (Mich, E.D.. 1989), at 853.Google Scholar

80. Excerpted from the UW Rule, UWM Post v. Board of Regents of U. of Wisconsin, 774 F. Supp. 1163 (E.D.Wis. 1991), at 1165 (my emphasis).Google Scholar

81. Lawrence, Charles R. III, If He Hollers Let Him Go: Regulating Racist Speech on CampusGoogle Scholar, in Matsuda, et al. , supra note 20, at 61.Google Scholar

82. Id. at 62.

83. Likewise, efforts to eliminate discrimination in the workplace are focusing more on practices of hiring and promotion and less on “diversity training.” As Edward N. Gadsden, a black man who is Texaco's director of diversity puts it “I am not in the business of attitudinal change. We need to establish a culture that specifies the behaviors you will exhibit” The strategy being adopted by managers is increasingly becoming that of making “even prejudiced people promote talented women and members of minorities whether they like them or not” Deutsch, Claudia H.. Just Shut Up and Hire, N.Y Times, 12 1, 1996, “The Week in Review,” at 4.Google Scholar

84. The pineal gland was considered by Descartes to be the place in the brain where body and mind interacted. As the mind, according to Descartes, does not exist in space, however, this is a mysterious meeting place indeed.

85. However, the fact that speech involves social norms and communication with others does undercut the individualism of a Cartesian account of language as expressing a person's private, inner thoughts.