1. 3 William Blackstone, Commentaries on the Laws of England, 4th ed. (1771), 208.

2. See Richard Posner, The Economic Analysis of Law, 4th ed. (Boston: Little, Brown, 1992), 163-167. The standard historical explanations for the rise of negligence are that it favored enterprise generally or that it amounted to a judicial subsidy to infant industry; see, e.g., O. W. Holmes, The Common Law, ed. M. Howe (Cambridge: Belknap Press, 1963), 77; Charles Gregory, "Trespass to Negligence to Absolute Liability," 37 Virginia L. Rev. 359, 365 (1951);John Fleming, The Law of Torts, 8th ed. (Sydney: Law Book Co., 1992), 101-102; Morton Horwitz, The Transformation of American Law, 1780-1860 (Cambridge: Harvard University Press, 1977), 85-108.

3. U.S. v. Carroll Towing, 159 F.2d 169 (1947); for Justice Hand, a defendant was negligent if the costs of the accident discounted by the probability of its occurring exceeded the costs of avoiding the accident.

4. For a historical account, see George Priest, "The Invention of Enterprise Liability: A Critical History of the Intellectual Foundations of Modern Tort Law," 14 J. Leg. Stud. 461 (1985).

5. See, e.g., Zeppa v. Coca-Cola, [1955] 5 D.L.R. 187 (Ont. C.A.); Roe v. Minister of Health and Another, [1954] 2 Q.B. 66.

6. Gold v. Essex C.C., [1942] 2 K.B. 293; Yepremian v. Scarborough Hospital, 110 D.L.R. 3d 513 (Ont. C.A. 1980); Ormrod v. Crossville Motor, [1953] 1 W.L.R. 1120 (C.A.).

7. Nettleship v. Weston, [1971] 3 All E.R. 581 (C.A.). Some claim that liability insurance has also affected the magnitude of damages awarded in personal injury cases; see Patrick Atiyah, Accidents, Compensation, and the Law (London: Weidenfeld & Nicolson, 1970), 262.

8. Fleming James, Jr., "Accident Liability Reconsidered: The Impact of Liability Insurance," 57 Yale L. J. 549-550 (1948).

9. Greenman v. Yuba Power Products, Inc., 377 P.2d 897 (Cal. S.C. 1963); a forerunner of this development is Justice Traynor's concurring opinion in Escola v. Coca-Cola Bottling Co., 150 P.2d 436, 440 (Cal. S.C. 1944).

10. See Restatement (Second) of Torts (St. Paul: American Law Institute Publishers, 1965), sec. 402A.

11. See generally Guido Calabresi, The Costs of Accidents: A Legal and Economic Analysis (New Haven: Yale University Press, 1970).

12. James, "Accident Liability Reconsidered," 569.

13. See the discussion and citation of various studies by Atiyah, Accidents, Compensation, and the Law, 299-304.

14. See Woodhouse Report, Royal Commission of Inquiry, Compensation for Personal Injury in New Zealand (1967).

15. Richard Epstein, "A Theory of Strict Liability," 2 J. Leg. Stud. 151 (1973); "Defenses and Subsequent Pleas in a System of Strict Liability," 3 J. Leg. Stud. 165 (1974).

16. George Fletcher, "Fairness and Utility in Tort Theory," 85 Harv. L. Rev. 537 (1972).

17. Ernest Weinrib, The Idea of Private Law (Cambridge: Harvard University Press, 1995); "Toward a Moral Theory of Negligence Law," 2 Law and Phil. 37 (1983); "Causation and Wrongdoing," 63 Chicago-Kent L. Rev. 407 (1987); "Understanding Tort Law," 23 Valparaiso Univ. L. Rev. 485 (1989).

18. Above n. 16.

19. For criticisms of this part of Fletcher's theory, see Ernest Weinrib, "The Special Morality of Tort Law," 34 McGill L. J. 403, 404-406 (1989); Izhak Englard, "The System Builders: A Critical Appraisal of Modern American Tort Theory," 9 J. Leg. Stud. 27, 65-66 (1980).

20. The alternative was seen this way by James Barr Ames, "Law and Morals,"

22. Harv. L. Rev. 97, 99 (1908). See also Holmes's utilitarian argument for fault in Common Law, 77.

21. Fletcher, "Fairness and Utility," 564-569.

22. The loss spreading achieved by enterprise liability can be justified on utilitarian grounds if one accepts the postulate of the diminishing marginal utility of money (see Calabresi, Costs of Accidents, 39 ff.). The argument is that a $1 loss suffered by 100 people is less burdensome overall than a $100 loss suffered by one person because the disutility of each dollar lost increases the less one has. However, this reasoning could justify shifting accident costs to a small group of the very wealthy as easily as it justifies loss spreading over the entire community and so pro-rides no specific rationale for the latter. Loss spreading is better understood as serving the aim of the welfare state generally: that of guaranteeing to all agents the basic material conditions of an autonomous life. This goal requires loss spreading rather than a search for the deep pocket, since the latter (imposing as it does a burden disproportionate to one's obligations under distributive justice) is forced philanthropy and so inconsistent with autonomy.

23. Fletcher's only response to this development is a note saying that it confuses corrective with distributive justice; see "Fairness and Utility," 547, n. 40. But there is no confusion. Enterprise liability ignores corrective justice in the same way that the reciprocity paradigm ignores distributive justice. The argument thus assumes what must be demonstrated—that some part of accident law must be concerned with correcting wrongs.

24. Fletcher's particular formulation of the nonreciprocal risk principle has exposed him to criticisms that do not touch the principle itself. Fletcher seems to determine whether risk is nonreciprocal by focusing on the specific roles that plaintiff and defendant perform in their transaction. Thus, a dynamite blaster imposes a nonreciprocal risk on an adjacent cattle farmer and is therefore strictly liable for injuries caused. But then, it may be objected, a car driver always imposes a nonreciprocal risk on a pedestrian and so should, on Fletcher's theory, also be strictly liable for injury even if the pedestrian was negligent; see Jules Coleman, "Moral Theories of Torts: Their Scope and Limits, Part I," in Michael Bayles and Bruce Chapman, eds., Justice, Rights, and Tort Law (Dordrecht: Reidel, 1983), 62. Such problems are avoided, however, if we view plaintiff and defendant not as particular role players but as moral agents for whom any role is theoretically possible (so the car driver of today may be the pedestrian of tomorrow and vice versa). Now we must judge whether risk is nonreciprocal by measuring it against a social background of ordinary risk—that is, by asking whether the defendant imposed on the plaintiff a risk greater than that ordinarily adjunct (and to which he is thus reciprocally exposed) to social interaction at a given stage of urban and technological development. This is the version of the nonreciprocity principle that I will later endorse.

25. In a subsequent article, Fletcher criticized his risk analysis of tort liability for failing to capture the notion of trespass, which involves a focus on the defendant's actual intrusion on the plaintiff; see "Synthesis in Tort Theory," in Bayles and Chapman, eds., Justice, Rights, and Tort Law, 120: "Risk analysis has powerful synthetic power, but it cannot serve to reduce the currently conflicted ex ante and ex post modes of tort analysis to a single framework for analyzing liability." I mean to show that the two sorts of analysis—corresponding roughly to negligence and trespass—are not conflicted.

26. The reaction against monistic theories produces "mixed" theories that sacrifice unity to multiplicity; see, e.g., Peter Cane, "Justice and Justifications for Tort Liability," 2 Oxford J. Leg. Stud. 30 (1982).

27. Richard Epstein, "Strict Liability," 151, 153-154.

28. Morriss v. Marsden, [1952] 1 All E.R. 925; Restatement (Second) of Torts (1965), sec. 283B.

29. Epstein, "Strict Liability," 158-160.

30. For criticisms of Epstein's theory, see Weinrib, "Causation and Wrongdoing," 420-424; Coleman, "Moral Theories of Torts, Part I," 52-57; Stephen Perry, "The Impossibility of General Strict Liability," 1 Can. J. Law and Jurisprudence 147 (1988).

31. Epstein, "Strict Liability," 159.

32. Id., at 164.

33. Id., at 179.

34. Id., at 172.

35. Epstein acknowledges this in "Nuisance Law: Corrective Justice and Its Utilitarian Constraints," 8 J. Leg. Stud. 49, 50-51 (1979).

36. Richard Epstein, "Causation and Corrective Justice: A Reply to Two Critics," 8 J. Leg. Stud. 477, 499-501 (1979).

37. See Richard Epstein, Takings: Private Property and the Power of Eminent Domain (Cambridge: Harvard University Press, 1985), 74.

38. Cf. Weinrib, "Causation and Wrongdoing," 424. An intentional destruction of another's property would, however, amount to an usurpation of proprietary sovereignty, and so such an act is, as Epstein has argued, morally indistinguishable from a taking; see Takings, 38.

39. Epstein thinks that strict liability achieves a reconciliation between the right to act and the right to security, inasmuch as everyone is free to act without taking account of others as long as he does not cause harm; see "Strict Liability," 203-204. However, this amounts to saying that the agent has a right to create risk for others as long as the risk does not materialize. It is difficult to see what a right to create risk would mean if it did not confer on the agent an immunity from responsibility for at least some materializations of risk.

40. An argument that loss and gain are not correlative in this situation has been put forward by Jules Coleman. The gain to the actor, he argues, equals the savings realized by foregoing precautions, and this gain would accrue whether or not anyone suffered a loss; see Jules Coleman, "Corrective Justice and Wrongful Gain," 11 J. Leg. Stud. 421, 424-427 (1982); cf. Stephen Perry, "The Moral Foundations of Tort Law," 77 Iowa L. Rev. 449, 457-461. Coleman has lately modified his views somewhat; see "The Mixed Conception of Corrective Justice," 77 Iowa L. Rev. 427 (1992).

41. "Toward a Moral Theory of Negligence Law," above n. 17; "Causation and Wrongdoing," above n. 17; "Right and Advantage in Private Law," 10 Cardozo L. Rev. 1283 (1989); "Understanding Tort Law," above n. 17; "The Special Morality of Tort Law," above n. 19; The Idea of Private Law, above n. 17.

42. Weinrib, "Right and Advantage," above n. 41.

43. The attempt to invoke both Kant and Hegel to support his theory leads Weinrib into difficulties from which he has yet to extricate himself. In particular, it leads him to affirm that the foundation productive of exclusively negative rights is an absolute foundation of right (Kant) but that there are nonetheless positive rights derived from principles external to formal right (Hegel); see "Right and Advantage," 1308-1309.

44. See Immanuel Kant, Critique of Practical Reason, trans. L. W. Beck (Indianapolis: Bobbs-Merrill, 1956), 65-66; for a lucid commentary on this Kantian principle, see Michael Sandel, Liberalism and the Limits of Justice (Cambridge: Cambridge University Press, 1982), 1-14.

45. Weinrib observes this without criticism. "Abstract right," he says, "is the normative structure that corresponds to the most fundamental condition of the will's freedom: the capacity to abstract from any particular object of choice" ("Right and Advantage," 1288).

46. Weinrib, "Right and Advantage," 1288.

47. Id., at 1289.

48. "In abstract right, advantages are valued not for their own sake but only inasmuch as they represent embodiments of an abstractly free will. Instead of being conceived as the particular benefits enjoyed by particular persons, the rights of private law are seen as expressions of the universal nature of the will's freedom" ("Right and Advantage," 1286).

Stephen Perry has criticized Weinrib's formalist theory of tort for its inability to account for tort law's concern for human welfare both at the stage of wrongdoing and at the stage of reparation; see "Moral Foundations," 481-488. Perry's critique, while partly correct, is too one-sided; it misses the dialectical tension in formal right. He is right to say that the principle of formal right excludes welfarist concerns and so renders much of tort law mysterious. However, he does not see that this principle contains an impetus for serf-embodiment, hence makes certain universal interests relevant to tort law, and so contains more than formal right's emphasis on pure agency allows it to accommodate. The problem with formal right is not that it is indifferent to welfare; its problem is that it is avowedly indifferent but actually and inevitably concerned. Formal right has within it the elements of agency and welfare, but because it identifies law's foundation with pure agency, it does not, qua formal right, know how to grasp the elements as a whole. To grasp them as a whole is to surpass formal right. Furthermore, the abstractness of formal right's foundations is not simply a problem for Weinrib's theory of tort law, as Perry believes; rather, it is the problem of classical tort law itself, whose internal contradictions Weinrib has faithfully reproduced.

49. Weinrib, "Understanding Tort Law," above n. 17.

50. This was, in fact, the gist of Fleming James's response to Charles Gregory's critique of his views on contribution among tortfeasors; see Charles Gregory, "Contribution Among Tortfeasors: A Defense," 54 Harv. L. Rev. 1170(1941). For James, contribution was regressive because he thought it would mean shifting losses from superior to inferior loss spreaders; see "Contribution Among Tortfeasors: A Pragmatic Criticism," 54 Harv. L. Rev. 1156, 1165 (1941). Gregory sympathized with the goal of loss distribution, agreed that the fault system was outmoded, but thought that in the absence of administrative compensation it was better to perfect the fault system than to bend it to goals it could achieve only imperfectly. James replied that to prefer a more coherent fault system to a partial realization of social insurance was to "sacrifice good sense to a syllogism—and an outworn syllogism at that." Fleming James, "Replication," 54 Harv. L. Rev. 1178, 1183 (1941).

51. 3 Blackstone, Commentaries, 116.

52. Fowler Harper, Fleming James, Jr., and Oscar Gray, The Law of Torts III (Boston: Little, Brown, 1986), 113-114, 131; Holmes, Common Law, 77-78.

53. Charles Gregory, "Trespass to Negligence to Absolute Liability," 37 Virginia L. Rev. 359, 365 (1951); Horwitz, Transformation, 97-101.

54. See G. W. E Hegel, Vorlesungen über Rechtsphilosophie 1818-1831, ed. K.-H. Ilting (Stuttgart-Bad Cannstatt: Frommann-Holzboog, 1974), IV, 174.

55. 114 So.2d 357 (Fla. S.C. 1959).

56. Id., at 359.

57. The independence of wrongdoing from harming is also reflected in the tort defense of volenti non fit injuria. If disrespect for the freedom of personality is a wrong regardless of the benefit conferred, respect for freedom absolves from wrongdoing regardless of the magnitude of harm inflicted. Suppose A accidentally and nonnegligently kills B in a highly risky surgical operation to which B has consented. A has committed no wrong, for though he has undoubtedly dealt a devastating blow to B's interests, he has shown no disrespect for the freedom of personality. Moreover, since freedom is here defined in utter abstraction from the particular knowledge of the individual, the consent that absolves from wrongdoing is formal consent; in a formal right unmodified by any requirements derived from a more concrete understanding of right, it is irrelevant that the victim was ignorant of the risks associated with the act and that the defendant knew this. Provided the defendant has not deceived the victim, the uncoerced consent of someone with legal capacity is sufficient to negate the wrong even if the consent is uninformed. Thus in Hegarty v. Shine, 4 L.R. Ir. 288, 294 (1878), recovery was denied a woman who sued her lover for assault claiming he withheld information that he was infected with a venereal disease. See also R. v. Clarence, 22 Q.B.D. 23 (1888).

58. See Yania v. Bigan, 155 A.2d 343 (Penn. S.C. 1959). The "exceptions" to this rule are cases in which plaintiff and defendant stand in a special relationship giving rise to a duty of affirmative action. The special relationships, however, are ones of induced dependence or ones where the defendant has created the danger, so that the failure to help in an emergency is really an imposition of risk indistinguishable from misfeasance. In this sense, the exceptions are really applications of the general rule to special circumstances. See, e.g., Moch Co. v. Rensselaer Water Co., 159 N.E. 896 (N.Y.C.A. 1928); for a discussion of "pseudonofeasance," see Ernest Weinrib, "The Case for a Duty to Rescue," 90 Yale L. J. 247, 254-258 (1980).

59. Epstein argues that the misfeasance/nonfeasance distinction reflects the principle of strict liability that causation is necessary and sufficient for wrongdoing. Thus, misfeasance is a causing of harm, whereas nonfeasance is not; see "Strict Liability," 190-191. This seems questionable, however, since there is nothing in the concept of causation (or at least nothing that Epstein reveals) that prevents one from saying that an egregious omission was the cause of a harm; see H. L. A. Hart and T. Honoré, Causation in the Law, 2d ed. (Oxford: Clarendon Press, 1985), 37-38. It is probably better to say that the distinction between misfeasance and nonfeasance rests on the irrelevance of welfare in formal tort law. If there is no wrong in harming, then afortiori there can be no wrong in failing to benefit.

60. Hegel, Philosophy of Right, par. 155; see also Arthur Jacobosn, "Hegal's Legal Plenum," 10 Cardozo L. Rev. 877, 901-906 (1989).

61. Interestingly, the action for damages seems to have been superimposed on an earlier action for novel disseisin, in which the remedy for dispossession of land was the restoration of seisin but not damages. The early cases in which damages are awarded draw a clear distinction between seisin and damages; see George Woodbine, "The Origins of the Action for Trespass," 33 Yale L. J. 799, 807-8o8 (1924).

62. Here we sidestep the debate over whether the action on the case issued from Westminster II in the thirteenth century or from an independent evolution of the common law in the fourteenth. See E. J. Dix, "The Origins of the Action of Trespass on the Case," 46 Yale L. J. 1142 (1937).

63. See, e.g., Leame v. Bray, 102 Eng. Rep. 724 (1803); Reynolds v. Clarke, 92 Eng. Rep. 410 (1726).

64. Mouse's Case, 12 Coke's Rep. 63 (1608); Ploof v. Putnam, 71 A. 188 (Vt. S.C. 1908).

65. 111 N.W. 1 (Minn. S.G. 1907).

66. See London Borough of Southwark v. Williams, [1971] 2 All E.R. 175 (C.A.). For a discussion of the controversy over the juridical significance of necessity, see Alan Brudner, "A Theory of Necessity," 7 Oxford J. Leg. Stud. 339 (1987).

67. See Richard Dian Winfield, Reason and Justice (Albany: State University of New York Press, 1988), 171-172.

68. This leaves open the possibility that the right to self-possession is subordinate to the state's authority to conscript citizens for military service in order to preserve the legal order. That the state's authority to punish criminals does not contradict the absolute fight to self-possession is shown in the next chapter.

69. Hegel, Philosophy of Right, par. 48.

70. Weaver v. Ward, 80 E.R. 284 (1616).

71. See Gregory, "From Trespass to Negligence," 361-376; Fleming, The Law of Torts, 16-19; Fowler v. Lanning, [1959] 1 Q.B. 426; Letang v. Cooper, [1965] l Q.B. 232, 239-240, per Denning, M.R. For a contrary view substantially in accord with mine, see Cane, "Justice and Justifications," 35-36.

72. That there was ever a shift from a clear strict liability regime to a clear negligence regime is disputed by several writers; see, e.g., Gary Schwartz, "Tort Law and the Economy in Nineteenth-Century America: A Reinterpretation," 90 Yale L. J. 1717 (1981); R. Rabin, "The Historical Development of the Fault Principle: A Reinterpretation," 15 Ga. L. Rev. 925 (1981).

73. See The Case of Thorns, Year Book, 6 Ed. IV, 7, pl. 18 (1466) (as reported in Bessey v. Olliott & Lambert, T. Raym. 467; 83 E.R. 244 [1681]). Here, however, we must remind ourselves that the same extraformalist intuition that produced the negligence-based action on the case for nontrespassory harms may also have insinuated itself into the action for trespassory harms. In fact, there is evidence that by the seventeenth century, a trespassory infliction of harm created only a prima facie case of liability for the harm, one rebuttable by the defendant's showing a lack of fault; see Weaver v. Ward, above n, 70, where it was said that "no man shall be excused of a trespass . . . except it may be judged utterly without his fault." Here "without his fault" was understood to include not only cases of involuntary conduct but also cases where "the defendant had committed no negligence to give occasion to the hurt."

74. Stanley v. Powell, [1891] 1 Q.B. 86; Brown v. Kendall, 60 Mass. 292 (S.C. 1850).

75. To see this, consider again the (admittedly unlikely) situation in which A takes unwelcome sexual liberties with B reasonably believing that B is consenting. Were it feasible for B to seek an injunction against the unwanted contact, can there be any doubt that she would be successful? Moreover, the law is clear that an action for nominal damages will lie for a faultless trespass to land. It would be odd if a landowner could recover nominal damages for a trespass to land but a person could not receive nominal compensation for unwanted intimacies with her body.

76. The Tubantia, [1924] All E.R. Rep. 615.

77. Hegel, Philosophy of Right, par. 40. In this context belongs Blackstone's famous definition of property as "that sole and despotic dominion which one man claims and exercises over the external things of the world, in total exclusion of the right of any other individual in the universe." 2 Blackstone, Commentaries, 2.

78. 3 Blackstone Commentaries, 209: "For the fight of meum and tuum, or property, in lands being once established, it follows as a necessary consequence, that this right must be exclusive; that is, that the owner may retain to himself the sole use and occupation of his soil: every entry therefore thereon without the owner's leave, and especially if contrary to his express order, is a trespass or transgression."

79. Basely v. Clarkson, 3 Lev. 37 (1681).

80. The foregoing account of strict liability is in disagreement with the conventional view in yet another respect. The standard view is that strict liability belongs to a communal or preindividualist conception of fights whereas fault is individualistic because it allows more scope to individual liberty and correspondingly affords less security to common human interests; see Harper, James, and Gray, The Law of Tort, III, 113. However, this view involves an anachronistic reading of ancient strict liability in the light of modern strict liability. Modern strict liability with its emphasis on loss distribution is indeed connected to a social conception of rights, but ancient strict liability is radically individualistic and fault is social in relation to it. Ancient strict liability is individualistic because it is the liability regime correlative to a conception of property rights that sees them as established prior to any social interaction and hence as unmediated by any deference to the liberty of the defendant. Fault is social because the plaintiff's right to security is not absolute but is mediated by the defendant's right to freedom of action.

81. The point is illustrated in Rodgers v. Elliott, 15 N.E. 768 (Mass. S.C. 1888).

82. Colls v. Home and Colonial Stores Limited, [1904] A.C. 179, at 185.

83. See Bamford v. Turnley, 122 E.R. 27 (1862); Losee v. Buchanan, 51 N.Y. 476 (1873). Why does a usufructuary right and not a possessory one require another's deference and hence a reciprocating accommodation to his liberty? The reason seems to be that A's possessory right over Blackacre cannot conflict with B's possessory right over Whiteacre; and so neither needs any self-abnegation from the other. However, given the possibility of incompatible uses, A's property in use requires B's self-denial and hence (since no fight can emerge from a servile relationship) a return recognition of B's liberty to interfere with A's use up to a limit (ordinary use) compatible with A's equal user fights. Of course, A's possessory right might extend over the whole planet and so be incompatible with any possessory right of B. But in formal right, there is only a negative fight to exclude others from things already reduced to the will; there is no positive fight to property.

84. Sturges v. Bridgman, 11 Ch.D. 852 (C.A. 1879); M.J. Horwitz sees the move from the priority rule to the reasonable user rule aa a judicial aid to newer and more efficient uses of land; see Horwitz, Transformation, 33-34. From our point of view, the help to industry is incidental; the legal development is intelligible on its own terms as one to a superior (because more intersubjective) conception of property.

85. Karl Marx, Capital: A Critique of Political Economy, ed. E Engels (New York: Modern Library, 1906), 81-96.

86. L.R. 3 H.L. 330 (1868).

87. Restatement (Second) of Torts (St. Paul: American Law Institute Publishers, 1977), sec. 519-520. The ratio in Rylands has been extended to such activities as blasting, crop dusting, and the storing of flammable liquids; see, e.g., IHB v. American Cyanamid 662 F. Supp. 635 (1987). However, the modern extensions of Rylands (especially the Restatement of Torts, sec. 519) have transcended Rylands's narrow nuisance theory of liability. Liability in Rylands is based on an interference with use caused by an extraordinary use of defendant's land (see for confirmation of this point Rickards v. Lothian, [1913] A.C. 263 [P.C.], per Lord Moulton). The extensions of Rylands have grounded liability on the causing of harm by activities that are ultrahazardous; whether the activity is within the normal use of the locality is only one factor to be considered in determining whether the activity is ultra-hazardous. It has become customary to view such liability as strict, but it is really a form of negligence liability. Negligence consists in the imposition of socially abnormal risk on someone to whom one owes a duty of care. The crucial feature of ultrahazardous activities (including the keeping of dangerous animals) is that, if carried on in populated areas, no practicable precautionary steps can bring the risk down to the socially normal. Thus, neither Rylands itself nor the ultrahazardous activities rubric is exceptional.

88. Hegel, it must be said, chose to emphasize the misfeasance aspect of breach at the expense of its nonfeasance side; see Vorlesungen, IV, 176-177. His argument is that breach of contract infringes a property already established by the "stipulation" or exchange of words. This is true, but recall that the stipulation itself involved one party's self-renunciation for the sake of another's realized freedom. Each party's property came into being through the other's positive act.

89. Hegel, Philosophy of Right, pars. 75, 81. 90.

90. Gregory, "Trespass to Negligence," 364.

91. Brown v. Kendall, above n. 74; Stanley v. Powell, above n. 74. The sole relevance of intention in torts is that it imposes a probability of harm approaching one and so subjects the victim to a degree of risk exceeding the socially normal. The intentionality of the harm makes unnecessary the usual tests for determining whether the risk was unreasonable. Because intention is significant only as a place-holder for the unreasonableness of the risk, it follows that the intention of a tort (unlike that of a crime) relates only to the causing of harm, not to the wrongfulness of the harm; see Morriss v. Marsden, [1952] 1 All E.R. 925.

92. Holmes, Common Law, 63.

93. 152 E.R. 402 (Exch. 1842).

94. Id., at 405.

95. Donoghue v. Stevenson, [1932] A.C. 562, 580 (H.L.), per Lord Atkin. Cf. Heaven v. Pender, 11 Q.B.D. 503, 509 (1882), per Lord Brett; Mcpherson v. Buick, 111 N.E. 1050 (N.Y.C.A. 1916).

96. Justice Cardozo put it well in Mcpherson v. Buick, 1053: "We have put aside the notion that the duty to safeguard life and limb, when the consequences of negligence may be foreseen, grows out of contract and nothing else. We have put the source of the obligation where it ought to be. We have put its source in the law."

97. These phrases are, of course, Ronald Dworkin's; see Taking Rights Seriously (Cambridge: Harvard University Press, 1977), 272-273.

98. 132 E.R. 490 (1837).

99. Id., at 492.

100. How are we to explain the fact that the common law adjusts the objective standard to the capacities of children and the physically disabled unless the child is engaged in an adult activity or the disabled person is doing something an ordinarily prudent person with his disability would not do? See McHale v. Watson, 115 C.L.R. 199 (Aust. H.C. 1966); Restatement (Second) of Torts (1965), sec. 283A. We will understand this phenomenon if we recall that negligence consists in the failure to avoid a socially extraordinary or nonreciprocal risk. Part of the everyday risks we run are those of being exposed to the naturally lower capacities for risk appreciation and risk avoidance of children and the disabled when they perform activities that are considerably less risky than those performed by people with average capacities. Thus, adjusting the objective standard to the capacities of the ordinarily prudent child, blind person, etc., within the sphere of comparatively innocuous activities is consistent with an intersubjective right to care: the child still has a duty to avoid a risk (e.g., riding a tricycle in motor traffic) judged unreasonable not by him or even by the average child but by an ideal person who measures the risk against those that children at play ordinarily pose. When, however, the child performs an activity of normal risk (e.g., driving a car), his abnormally low capacity turns this risk into an abnormal one (an ultrahazardous activity, as it were), and so a right to care implies that he must be held legally accountable for any harm he causes even though the average person of his age would not have been able to avert the accident.

101. Above n. 3.

102. [1951] A.C. 850 (H.L.).

103. [1966] 2 All E.R. 709 (P.C.).

104.      162 N.E. 99 (N.Y.C.A. 1928).

105.      105. Id., at 99.

106. Id., at 103.

107. The neighbor principle is sometimes said to give rise to a prima facie duty of care, subject to limitation by "policy" where liability would otherwise be too onerous; see Arms v. London Borough of Merton, [1977] a All E.R. 496 (H.L.). One such supposed limitation occurs with negligently spoken words. One will be liable for negligent misstatement only if one encouraged the plaintiff to rely on one's words and if he did in fact rely; see Hedley Byrne & Co. v. Heller and Part-nets Ltd., [1964] A.C. 465 (H.L.). However, there is no need to regard the rule in Hedley Byrne as a compromise of negligence principles. The risk of being misled by another's mistaken utterances is surely as normal in the conditions of modern life as the risk of being jostled in a public thoroughfare. For the image of crowded public places one has only to substitute that of far-flung connections made possible by modern communications technology. To take the risk imposed by the defendant out of the socially normal and reciprocal, therefore, he must have encouraged the plaintiff to rely on his words (e.g., by giving advice in the knowledge that the plaintiff would rely on it) and the plaintiff must in fact have been at risk, that is, he must have relied. Indeed, the contrast between reciprocal and nonreciprocal risk underlay Lord Reid's reasoning in Hedley Byrne. "The most obvious difference between words and acts," he wrote (at 482-483), "is this. Quite careful people often express definite opinions on social or informal occasions even when they see that others are likely to be influenced by them; and they often do that without taking that care which they would take if asked for their opinion professionally or in a business connection. . . . But it is at least unusual casually to put into circulation negligently made articles which are dangerous."

108. William Prosser, "Palsgraf Revisited," 52 Mich. L. Rev. 1, 20 (1953).

109. [1921] 3 K.B. 560 (C.A.).

110. [1961] A.C. 388 (P.C.).

111. Weinrib, "Causation and Wrongdoing," 414-416.

112. For a judicial suggestion to this effect, see Rivtow Marine Ltd. v. Washington Iron Works, 40 D.L.R. (3d) 530, 552-555 (S.C.C.), per Laskin, J.

113. Judith Jarvis Thompson, Rights, Restitution, and Risk (Cambridge: Harvard University Press, 1986), 199-202.

114. See also Aristotle's image of the divided line in Nicomachean Ethics, V, 1132a-b.

115.      See J. Coleman, "Moral Theories of Torts: Their Scope and Limits, Part II" in M. Bayles and B. Chapman, eds., Justice, Rights, and Tort Law (Dordrecht: Reidel, 1983), 95-96.

116. It might be thought that liability insurance is inconsistent with the requirement that the plaintiff's remedy move from the defendant and that, since the defendant normally does not pay, there is no reason to insist on fault; see Atiyah, Accidents, Compensation, and the Law, 262-274. However, there is no inconsistency here. After all, the defendant contracts with the insurance company to insure his liability. The fact of liability insurance does not undercut the corrective nexus between plaintiff and defendant; on the contrary, it presupposes that nexus; see E. Weinrib, "The Insurance Justification and Private Law," 14 J. Leg. Stud. 681 (1985). Liability insurance becomes problematic for tort law only if one assumes that a central aim of tort law is to punish the wrongdoer. But tort law does not punish; it merely restores the balance between plaintiff and defendant.

117. See, e.g., Blackstock v. Foster, [1958] S.R. (N.S.W.) 341 (S.C.). Our understanding of the factual cause requirement as an expression of the correlativity of wronging and being wronged is indebted to E. Weinrib, "Causation and Wrongdoing," above n. 17.

118. For recovery for emotional distress, see Hambrook v. Stokes Brothers, [1925] 1 K.B. 141 (C.A.); Dillon v. Legg, 441 P.2d 912 (Cal. 1968), For recovery for "pure economic loss," see Hedley Byrne & Co. v. Heller and Partners Ltd., above n. 107; Caltex Oil (Australia) Pry. Ltd. v. The Dredge "Willemstad," 136 C.L.R. 529 (H.C. 1976); Union Oil Co. v. Oppen, 501 F.2d 558 (9th Cir. 1974); Canadian National Railway Co. v. Norsk Pacific Steamship Co., [1992] 1 S.C.R. 1021. Sometimes awards for economic loss go too far, as when recovery is allowed not only for setbacks to economic well-being but also to make good an expectation of gain despite the absence of privity of contract; see Anns v. London Borough of Merton, above n. 107. Overreaction to Anns has resulted in a recent setback for pure economic loss recovery in England; see Murphy v. Brentwood District Council, [1991] 1 A.C. 398 (H.L.).

119. By social insurance I mean not any particular scheme for dealing with accident losses but any scheme whose aim is the redistribution of the costs of accidents so as to maintain individuals at a humanly decent standard of living. By this definition, American products liability law is as much an example of social insurance as Workers' Compensation statutes; see Harper, James, and Gray, The Law of Torts, III, 132, n. 7. As I am using the term, moreover, social insurance embraces both social security measures that guarantee a basic welfare floor and compensation plans that guarantee a level of welfare approximating the one enjoyed by the victim at the time of the accident.

120. Henningsen v. Bloomfield Motors, Inc., 161 A.2d 69 (N.J.S.C. 1960).

121. See Priest, above n. 4; P. W. Huber, Liability: The Legal Revolution and Its Consequences (New York: Basic Books, 1988), 36-44.

122. See Harper, James, and Gray, The Law of Torts, III, 103-126; Atiyah, Accidents, Compensation, and the Law, 449-477; Stephen Sugarman, Doing Away with Personal Injury Law (New York: Quorum Books, 1989), 3-72.

123. T. Ison, The Forensic Lottery (London: Stapless Press, 1967).

124. Atiyah, Accidents, Compensation, and the Law, 449-450, 453-460.

125. See Weinrib, "Understanding Tort Law," 502-503; Englard, "The System-Builders," 57-59; S. Smith, "The Critics and the 'Crisis': A Reassessment of Current Conceptions of Tort Law," 72 Cornell L. Rev. 765, 778-785 (1987).

126. One aspect of the social insurance critique of tort law can, however, be dismissed. It is often argued that tort law is inadequate when measured against its own objectives, for it wants to condemn moral wrongdoing and yet the objective fault standard takes no account of personal capacities that may negate the actor's blameworthiness (see Harper, James, and Gray, The Law of Torts, III, 115-117); or it is argued that the penalties tort law imposes are often wholly disproportionate to the blame. We have seen, however, that the condemnation of moral failure forms no part of the internal rationality of tort law. The latter vindicates individual rights against external acts inconsistent with them; it does not penalize immorality except as incidentally concomitant with that purpose.

127. The milestones on the road to strict liability are Escola v. Coca-Cola Bottling Co., above n. 9 (where the idea appeared in a concurring opinion by Justice Traynor), Henningsen v. Bloomfield Motors, above n. 120 (which imposed liability on the theory of breach of warranty but which waived the privity requirement and invalidated a clause excluding liability), Greenman v. Yuba Power Prods., Inc., above n. 9 (which explicitly put liability on the foundation of tort rather than contract), and sec. 402A of Restatement (Second) of Torts (1965). The logical culmination of this process must, as George Priest has argued, be absolute liability for injuries resulting from the use of products; see Priest, "Enterprise Liability," 527. Products liability law retains a vestige of fault in the requirement that the injury be caused by a defect in manufacture or design. In particular, negligence ideas have been reintroduced in the idea of a design defect, which must be defined in relation to ordinary consumer expectations of safety or to alternative designs whose benefits outweigh their costs; see Barker v. Lull Engineering Co., 573 P.2d 443 (Cal. S.C. 1978). The logical momentum of products liability law is toward elimination of the requirement of a product defect Moreover, the loss-spreading rationale of no-fault liability also implies the elimination of defenses of contributory negligence and voluntary assumption of risk.

128. For a discussion of the problems inherent in allocating costs to their sources, see M. Trebilcock, "The Social Insurance-Deterrence Dilemma of Modern North American Tort Law: A Canadian Perspective on the Liability Insurance Crisis," 24 San Diego L. Rev. 929, 987-980 (1987); see also the discussion by Perry, "Moral Foundations," 461-467.

129. See Guido Calabresi, "Concerning Cause and the Law of Torts: An Essay for Harry Kalven, Jr.," 43 Univ. Chic. L. Rev. 69, 84-87 (1975); John Fleming, "Probabilistic Causation in Tort Law," 68 Can. Bar Rev. 661, 667-669 (1989).

130. Cook v. Lewis, [1951] S.C.R. 830; Summers v. Tice, 199 P.2d 1 (Cal. S.C. 1948).

131. Sindell v. Abbott Laboratories, 607 P.2d 924 (Cal. S.C. 1980).

132. Allen v. U.S., 588 F. Supp. 247 (1984); In re "Agent Orange" Product Liability Litigation, 597 F. Supp. 740 (1984).

133. The irrelevance of causation is finally admitted in Hymowitz v. Eli Lilly, 539 N.E.2d 1069 (N.Y.C.A. 1989), where liability was apportioned according to the percentage of the total risk created to the public, and the defendant could not avoid liability by showing it did not cause the injury.

134. S. Sugarman, "Doing Away with Tort Law," 73 Cal. L. Rev. 555 (1985).

135. At least three broad types of mixed scheme are possible. One uses the tort system as a backup to the insurance scheme (modified no-fault); another uses insurance as a backup to the tort system (add-on scheme); and a third allows everyone a choice between alternative regimes (elective scheme). Under the first model, insurance benefits are generous and the tort system is retained only to allow victims to sue their faulty injurers for the difference between social insurance benefits and full compensation; victims suffering losses below a threshold are barred from seeking vindication of their private rights in court. Under the second, accident victims immediately receive benefits (typically less generous than under modified no-fault) with no prejudice to their right to sue in tort. Unsuccessful plaintiffs continue receiving insurance benefits, while successful plaintiffs reimburse the fund or have collateral benefits deducted from the damage award. Under the third type, each individual elects either to receive guaranteed compensation for pecuniary loss or to take his chances for full recovery in court; for a specific proposal exemplifying the third model, see J. O'Connell, "Statute—A Draft Bill to Allow Choice Between No-Fault and Fault-Based Auto Insurance," 27 Harv. J. Legis. 143 (1990).

136. It has been argued that a dual system best achieves loss-spreading and deterrence goals by obviating the need to compromise either in a single system. Deterrence is achieved by a system that allocates liability according to fault without regard to loss spreading, while the best spreading of losses is achieved through social insurance funded from general revenue; see S. Stoljar, "Accidents, Costs and Legal Responsibility," 36 Modern L. Rev. 233 (1973); cf. W. Blum and H. Kalven, Jr., "The Empty Cabinet of Dr. Calabresi," 34 Univ. Chic. L. Rev. 239, 270 (1967).

137. Thus, of the three mixed models outlined in n. 135, only the add-on model and the elective model are consistent with the justifying theory presented here. The modified no-fault model is an unjustifiable infringement of rights. The choice between the two permissible models can be determined by considerations of cost, deterrence, and the like.

Most North American auto plans (e.g., those of New York, Massachusetts, Michigan, Florida, Ontario) are modified no-fault schemes. Inasmuch as they abolish the right to sue in tort for less serious injuries, they are inadequate to the ideal dual system my argument points to. It is easy to think of a tort action as an expensive frill if the victim is already nearly compensated by the fund, particularly since very few individuals in this situation are likely to exercise their right to sue. However, this sort of thinking betrays the one-sidedly communitarian mind-set that sees tort law as another compensation mechanism rather than as an institution for identifying and correcting wrongs against the person. Given a contest between the right to vindicate personality's worth and guaranteed levels of welfare above the minimum compatible with human autonomy, it would not be eccentric to suggest that the latter rather than the former is the luxury.

138. See Huber, Liability, 3-5.

139. Id., 202.

140. Add-on auto schemes (see n. 135) are in place in Delaware, Maryland, Oregon, Arkansas, Texas, South Carolina, Washington, and Saskatchewan. New Jersey and Pennsylvania have adopted elective schemes.