1. David Hume, A Treatise Of Human Nature, bk. III, pt. 2, sec. 5 (Oxford: Clarendon Press, 1978). For some recent philosophical discussions of the issues raised by Hume, see, e.g., P. S. Atiyah, Promises, Morals and Law (Oxford: Clarendon Press, 1981); Michael H. Robins, Promising, Intending, and Moral Autonomy (Cambridge: Cambridge University Press, 1984); Judith Jarvis Thomson, The Realm of Rights (Cambridge, MA: Harvard University Press, 1990) ch. 12; J.E.J. Altham, “Wicked Promises” in Ian Hacking, ed., Exercises in Analysis (Cambridge: Cambridge University Press, 1985) 1; G.E.M. Anscombe, “Rules, Rights, and Promises” (1978) 3 Midwest Stud, in Phil. 318; R.S. Downie, “Three Accounts of Promising” (1985) 35 Phil. Q. 257; Christopher McMahon, “Promising and Coordination” (1989) 26 Am. Phil. Q. 239; F.S. McNeiUy, “Promises De-Moralized” (1972) Phil. Rev. 63; Joseph Raz, “Promises and Obligations” in P.M.S. Hacker & Joseph Raz, eds., Law, Morality, and Society: Essays in Honour ofH.LA. Hart (Oxford: Oxford University Press, 1977) 210; Thomas Scanlon, “Promises and Practices” (1990) 19 Phil. & Publ. Affairs 199; Georg H. von Wright, “On Promises” in Practical Reason: Philosophical Papers Vol. I (Ithaca, NY: Cornell University Press, 1983) 83.

2. See, e.g., Charles Fried, Contract as Promise (Cambridge, MA: Harvard University Press, 1981); James Gordley, The Philosophical Origins of Modem Contract Doctrine (Oxford: Clarendon Press, 1991); Pall S. Ardal, “Ought We to Keep Contracts Because They Are Promises?” (1983) 17 Val. U. L. Rev. 655; P.S. Atiyah, “Contracts, Promises, and the Law of Obligation” in Essays on Contract (Oxford: Clarendon Press, 1986) 10; Kent Bach, ‘Terms of Agreement” (1995) 105 Ethics 604; Randy Barnett, “A Consent Theory of Contract” (1986) 86 Colum. L. Rev. 269; Richard Craswell, “Contract Law, Default Rules, and the Philosophy of Promising” (1989) 88 Mich. L. Rev. 489; Anne De Moor, “Are Contracts Promises?” in John Eekelar & John Bell, eds., Oxford Essays in Jurisprudence (Third Series) (Oxford: Clarendon Press, 1987) 103; E. Allan Famsworth, “The Past of Promise: An Historical Introduction to Contract” (1969) 69 Colum. L. Rev. 576; Margaret Gilbert, “Is an Agreement an Exchange of Promises?” (1993) 90 J. of Phil. 627; Charles J. Goetz & Robert E. Scott, “Enforcing Promises: An Examination of the Basis of Contract” (1980) 89 Yale L.J. 1261; J.E. Penner, “Voluntary Obligations and the Scope of the Law of Contract” (1996) 2 Legal Theory 325; A.D. Woozley, “Promises, Promises” (1981) 90 Mind 289.

3. See, e.g., John Searle, Speech Acts (Cambridge: Cambridge University Press, 1969); Andrzej Boguslawski, “An Analysis of Promise” (1983) 7 J. of Pragmatics 607; Allesandro Duranti, “Intentions, Self, and Responsibility: An Essay in Samoan Ethnopragmatics” in Jane H. Hill & Judith T. Irvine, eds., Responsibility and Evidence in Oral Discourse (Cambridge: Cambridge University Press, 1993) 24 at 42–43; Jef Verschueren, “On Boguslawski on Promise” (1983) 7 J. of Pragmatics 629.

4. See, e.g., Fried, supra note 2; Michael O. Hardimon, “Role Obligations” (1994) 91 J. of Phil. 333 at 354.

5. Speech act theory, first developed by John Austin, is based on the recognition that utterances can be operative—as when one says, “It’s a deal,” to conclude a contract. See J.L. Austin, How to do Things with Words, 2d. ed. by J.O Urmson & Marina Sbisa (Cambridge, MA: Harvard University Press, 1975) at 4–11. Despite the importance of legal examples in the development of speech act theory, its application to law and jurisprudence has been limited. For a few examples of such application, see Robert Samek, “Performative Utterances and the Concept of Contract” (1965) 43 Australasian J. of Phil. 196; Dick W.P. Ruiter, Institutional Legal Facts: Legal Powers and Their Effects (Dordrecht: Kluwer Academic Publishers, 1993).

6. An example is Downie, supra note 1. The author comes close to recognizing that there are two aspects to promise, but, because he fails to clarify the notion of “obligation,” Downie does not recognize that only one aspect gives rise to obligations.

7. At this point a disclaimer is necessary. Promise is both a complex subject and a topic of widespread interest. Each person brings to the examination of promise his or her own scholarly or practical perspective. And each person who writes on promise, whether a jurisprudent, philosopher of language, anthropologist, or man in the street, tends to seek answers to the questions that are of greatest interest from that perspective.

Although this article makes use of tools originally developed to handle technical problems in linguistics and the philosophy of language, it is written mainly from a jurisprudential perspective. In order that the argument be comprehensible to the interested, but philosophically nonexpert reader, the philosophical tools and distinctions we make use of are not as detailed or refined as they potentially could be. Nonetheless, they are sufficient for the present task of understanding and analyzing the complexities of promise.

Linguists and philosophers of language may be inclined to criticize parts of the argument on the ground that its discussion of meaning is too simplified. These readers might criticize the article for assuming debatable positions, such as whether the distinction between semantics and pragmatics can ultimately be maintained; or for failing to address subtle issues of concern to those whose professional task is to shed light on language phenomena. Such controversial or technical issues are indeed interesting. But to address them would make this a very different article, written for a very different audience.

8. The following discussion substantially follows the explanation of senses of “meaning” advanced by Gilbert Harman in his “Three Levels of Meaning” (1968) 65 J. of Phil. 590.

9. Concepts and semantic meaning are closely allied. It seems a truism that the concept of fruit is much the same as the meaning of the term “fruit” (or some equivalent word in another language, such as “Frucht”) and that to ask for the meaning of a term is to ask about the corresponding concept. It also seems a truism that a word has no meaning if there is no associated concept, and that it has more than one meaning if it can express multiple concepts. In light of this close correspondence, we will refer interchangeably to semantic meaning and concepts.

10. The fundamentality here is conceptual, rather than epistemic or psychological. To say that semantic meaning is fundamental is not to say that one must first grasp it before one can engage in communication—that would be absurd, since one learns meaning in this sense only through its being communicated. The fundamentality of semantic meaning lies in the fact that an adequate analysis of meaning in other senses presupposes an adequate account of meaning in this sense. See Harman, supra note 8 at 593.

11. The discussion in the text glosses over the difference between an individual’s concept of X and the public concept of X (or the meaning of “X” in the relevant language), and the relationship between them. This topic is extremely technical and well beyond the scope of the article. Our concern is with the public concept.

12. A classic exposition of meaning in this sense is Wilfrid Sellars, “Some Reflections on Language Games” in Science, Perception and Reality (New York: Humanities Press, 1963) 321. See also Gilbert Harman, “Language, Thought, and Communication” in (1975) 7 Minn. Stud, in the Phil. of Science 270 at 290.

13. See, e.g., Wilfrid Sellars, “Empiricism and the Philosophy of Mind” (1956) 1 Minn. Stud, in the Phil, of Science 253, reprinted in Science, Perception and Reality 127 (1963); Hilary Putnam, “The Meaning of Meaning” (1975) 7 Minn. Stud, in the Phil, of Science 131.

14. On pragmatics generally, see Steven C. Levinson, Pragmatics (Cambridge: Cambridge University Press, 1983); Steven Davis, ed., Pragmatics: A Reader (Oxford: Oxford University Press, 1991).

15. Pragmatic meaning is important to law and legal theory. Very often the question of what the law is comes down to the question of what a statement or document means; i.e., what it communicates. When one interprets a statute, contract, or other writing, one clarifies its meaning in this pragmatic sense. Sometimes one does so by eliciting the “legislative intent” (or “intent of the drafter”); and this is pretty much what, in pragmatics, is called “speaker’s meaning.” More commonly, however, one interprets a legally significant document by considering what the language ordinarily means when it is used in the given context. In doing this one elicits “plain meaning,” which is pretty much what in pragmatics is referred to as “plain” (or “literal”) meaning. For one effort to apply pragmatics to statutory interpretation, see Geoffrey P. Miller, “Pragmatics and the Maxims of Interpretation” (1990) Wise. L. Rev. 1179. See also Jay Conison, “ERISA and the Language of Preemption” (1994) 72 Wash. U.L.Q. 619.

16. Austin, supra note 5.

17. J.O. Urmson, “Performative Utterances” (1977) 2 Midwest Stud, in Phil. 120 at 126.

18. Searle, supra note 3 at 64–67. Certain other technical conditions should be added, but they need not be elaborated here.

19. Elocutionary force can be conveyed in a variety of ways: for example, by inflection, word order, a prefatory verb, or simply context. Searle, ibid, at 30.

20. Note, however, that the social or other conventions may themselves call for the use of language. Urmson, supra note 17 at 125–26. For a further discussion of the role of social and other conventions in the analysis of performatives, see Dan Sperber & Deirdre Wilson, Relevance: Communication and Cognition (Oxford: Blackwell, 1995) at 244–46.

21. For an account of performatives in the law, see Neil MacCormick, “Law As Institutional Fact” (1974) 90 L.Q. Rev. 102.

22. See, e.g., Joseph Raz, Practical Reason and Norms (Princeton, NJ: Princeton University Press, 1990) at 97–106.

23. MacCormick, supra note 21 at 114–21.

24. Joseph Raz, “Voluntary Obligations and Normative Powers” (1972) 46 Proc. Arist. Soc’y (Supp.) 79 at 92–101.

25. See Urmson, supra note 17 at 124. Urmson does not use the notion of a power norm, but it clearly is the type of rule his analysis tries to capture.

26. See, e.g., Linda Coleman & Paul Kay, “Prototype Semantics: The English Word Lie” (1981) 57 Language 26; Gilbert Harman, “(Nonsolipsistic) Conceptual Role Semantics” in Ernest Lepore, ed., New Directions in Semantics 55 (Toronto: Atlantic Press, 1987); Putnam, supra note 13. See also Julius Kovesi, Moral Notions (New York: Humanities Press, 1967) at ch. 1–2.

27. Frederick Waismann, “Verifiability” (1945) 19 Proc. Arist. Soc’y (Supp.) 123. The term “open texture” was subsequently adopted by H.L.A. Hart to refer to a different linguistic phenomenon, see H.L.A. Hart, The Concept of Law (Oxford: Clarendon Press, 1961) at 121–32.

28. By contrast, a spurious concept is one for which no rule or principle gives guidance in cases beyond the paradigm. Although one can say that Jones’ cat Fluffy, the tallest sycamore tree in Central Park, and the final movement of Schumann’s Spring Symphony are all kenzires, unless we have a rule to help determine what other items are too, there is no concept of kenzire. See Kovesi, supra note 26 at ch. 2.

29. A semantic theory in some respects similar to that sketched in the text may be found in Coleman & Kay, supra note 26. The authors of that article (who are linguists) develop a semantic theory based on prototypes: for a given word, X, a finite set of characteristics such that (a) an item’s possession of all the characteristics ensures that people will consider the item an X, and (b) an item’s possession of some, but not all, of the characteristics will lead a significant number of people to consider the item an X. The notion of prototype is similar to the notion of non-ostensive paradigm developed in the present article.

The authors recognize, as they must, that the prototype alone cannot guide use of a term. For example, they observe from studies that the usage of “lie” (in the sense of prevaricate) is guided also by a speaker’s consideration of whether conduct that partially satisfies the prototype is deserving of certain types of criticism. However, rather than treating the principle that lying is blameworthy as a further part of the semantics of “lie,” the authors consider it a matter of social convention, a subject for ethnography, Ibid. at 35–38.

30. On the need to take rules into account in explaining the meaning of empirical concepts, see Wilfrid Sellars, “Concepts as Involving Laws and Inconceivable Without Them” (1948) 15 Phil. Sci. 287. On the role of rules in the meaning of social or artifactual concepts, see Kovesi, supra note 26. For another discussion of the role of rules in meaning, see Margaret Gilbert, On Social Facts (Princeton, NJ: Princeton University Press, 1989) at 69–70.

31. The analysis admits of the possibility that, in certain cases, either paradigm or point is disproportionately important to meaning. At one end of the spectrum would be substantially indexical terms—those whose meaning is given almost entirely by an ostensive paradigm. At the other extreme are terms that are substantially functional, ones whose meaning is given mainly by point.

32. For present purposes, it suffices to take the difference between descriptive and normative concepts to be a difference in the rules or regularities that enter into the point. See Kovesi, supra note 26 at ch. 1. The simplest way to capture the distinction between descriptive and normative rules is as follows. Suppose R is a statement of a rule and F is a fact inconsistent with R. Then, if/? is a descriptive statement, R is incorrect, and if R is a normative statement, F is incorrect. For a colorful elaboration of this distinction, see G.E.M. Anscombe, Intention, 2d ed. (Ithaca, NY: Cornell University Press, 1976) § 32. Much more, of course, can be said about the difference between descriptive and normative concepts and rules. However, there is no need to pursue the topic here.

33. See A.I. Melden, Rights and Right Conduct (Oxford: B. Blackwell, 1959).

34. To forestall confusion, we must emphasize that a “semantic context” for a term is not the context of its utterance, or any other social context. By “semantic context” we mean an organized body of practical or theoretical knowledge, concerning or involving the relevant term. For example, one semantic context for “tissue” would be histology, a body of medical knowledge; another might be what we call the grocery business, a practical body of knowledge and action. Of course, a semantic context is usually implicit in any occasion of communication—grocers discussing tissues will draw on the meaning of “tissue” in the semantic context of the grocery field. But our present interest is in the context for semantic meaning.

Although the label “semantic context” is invented here, the notion is largely commonsensical; dictionaries make use of it freely. For example, examine the dictionary entry for “point” and one will find references to the word’s meaning “in” music, astronomy, time measurement, sports and games, bridge, securities markets, typography, (Australian) meteorology, jewelry business, geometry, hunting, heraldry, sculpture, mountaineering, negotiation, punctuation, tatting, topography, military arts, animal breeding, dancing, nautical science, archaeology, electrical mechanics, theater, cricket, lacrosse, trial practice, and falconry. The New Shorter Oxford English Dictionary (Oxford: Clarendon Press, 1993) at 2266–67.

35. Anscombe, supra note 1 at 319–20.

36. Compare Raz’ notion of normative effect, which is based on the distinction between the result and the consequence of an act:

That John wakes up is the consequence of my turning on the light, but [is] the result of my waking John[,] which I may have done by turning on the light…. Similarly, bullying somebody to promise to do A with the consequence that he does make the promise is causally effective in creating the norm that he ought to do A, but only his act of promising affects this norm normatively.

Raz, supra note 24 at 94. See also Raz, supra note 22 at 103.

37. One may call this pattern the institutional pattern. It has been invoked to explain an array of social phenomena, ranging from tables to legal institutions. Following the lead of G.E.M. Anscombe, “On Brute Facts” (1958) 18 Analysis 69, we may call ϕ a brute fact; and following the lead of MacCormick, supra note 21, call \|f an institutional fact. For explanatory uses of the institutional pattern, see Kovesi, supra note 26; John Searle, The Construction of Social Reality (New York: Free Press, 1995).

38. For present purposes, it suffices to rely on commonsense notions of “describe” and “explain.” No doubt the distinction between descriptive or practical laws, on the one hand, and explanatory laws on the other, can be much refined. But it is unnecessary to do so here.

39. See, e.g., Bernard Williams, Ethics and the Limits of Philosophy (Cambridge, MA: Harvard University Press, 1985) at 140–45.

40. Ibid, at 129–30.

41. Some have noticed this fact, although they have expressed it in other ways. See, e.g., von Wright, supra note 1 at 85:

The urge to establish (decide, find out) whether a promise has been given or not, usually comes from the claim on the part of an alleged promise-receiver that an alleged promise-giver should keep his word. The urge to establish the existence of the promise, one could also say, usually comes from a claim that a certain agent is under an obligation to another agent. The existence of the promise is the justification of this claim. Anscombe takes a slightly different approach, contending that it is misleading to treat “Jones promised to ϕ” as an independently meaningful proposition. Rather, she argues, such a statement must be understood as part of a statement schema of the form: “Jones should (or should not) …, because Jones promised ….” Anscombe, supra note 1 at 322.

42. E.g., Restatement (Second) of Contracts § 2(1); The New Shorter Oxford English Dictionary, supra note 34 at 2375; E. Allan Farnsworth, “Decisions, Decisions: Some Binding, Some Not” (1994) 28 Suffolk L. Rev. 17 at 19.

43. See, e.g., John Searle, “How to Derive ‘Ought’ from ‘Is’” (1964) 73 Phil. Rev. 43. On the rationale for treating “ought to ϕ” as equivalent to “has a reason to ϕ.” See Raz, supra note 22 at 29–32.

44. See Urmson, supra note 17 at 124–25; Neil MacCormick, “Voluntary Obligations and Normative Powers” (1972) 46 Proc. Arist. Soc’y (Supp.) 59 at 60.

45. Cf. Duranti, supra note 3 at 43 (claiming that promising in Samoan society requires a public commitment in specific social settings).

46. Kurt Baier, The Rational and the Moral Order (Chicago: Open Court, 1995) at 316, seems to recognize the need to take both aspects into account, but makes little use of the speech act character of promising. Others who seem to recognize the dual character of promising, but without fully appreciating the implications, are Altham, supra note 1, and McMahon, supra note 1. Arguably, Hume recognized it when he wrote that:

When a man says he promises any thing, he in effect expresses a resolution of performing it; and along with that, by making use of this form of words, subjects himself to the penalty of never being trusted again in case of failure. A resolution is the natural act of the mind, which promises express: But were there no more than a resolution in the case, promises would not create any new motive or obligation. They are the conventions of men, which create a new motive, when experience has taught us, that human affairs would be conducted much more for mutual advantage, were there certain symbols or signs instituted, by which we might give each other security of our conduct in any particular incident. After these signs are instituted, whoever uses them is immediately bound by his interest to execute his engagements, and must never expect to be trusted any more, if he refuse to perform what he promised.

Hume, supra note 1 at 522.

47. For a criticism of speech act analyses of promising on the ground that they fail to account for the motivational feature of promises, see McNeilly, supra note 1 at 76.

48. John Searle, “A Taxonomy of elocutionary Acts” in John Searle, ed., Expression and Meaning (Cambridge: Cambridge University Press, 1979) 1 at 12–13.

49. Ibid, at 13–14.

50. Ibid, at 14. We are here concerned with vowing and swearing in their everyday senses, and not in their more specialized religious or legal senses. Thus, we are not concerned with, e.g., marriage vows.

51. Deciding and resolving, like vowing and many other illocutionary acts, can be performed either as mental acts or as illocutionary acts. So far as concerns their function of giving the speaker (or thinker) reasons for action, nothing relevant to the present inquiry turns on the difference.

52. Searle, supra note 48 at 9. Of course, expressing the intention to § can be a speech act.

53. The analysis of decisions given here largely follows that of Raz, supra note 22 at 65–69.

54. Ibid, at 36.

55. E.g., Donald Davidson, “Actions, Reasons, and Causes” in Donald Davidson, ed., Essays on Actions and Events (Oxford: Clarendon Press, 1980) 3; Anscombe, supra note 32; Robert Audi, Action Intention and Reason (Ithaca, NY: Cornell University Press, 1993); Raz, supra note 22 atch. 1.

56. However, the presence of underlying reasons is not a necessary feature of decisions. When pressed to explain why he decided to ()>, an individual may deny that there is any further reason for his ϕ’ing.

57. For an account of weighing reasons in the context of judicial decision making, see Steven J. Burton, Judging in Good Faith (Cambridge: Cambridge University Press, 1992).

58. Raz, supra note 22 at 70–71.

59. Of course, the fact that the promisor is under an obligation may provide additional reasons for her to do what she promised. But the discussion in the text shows that she ordinally will have reasons to do what she promised, even without appealing to the presence of any obligation.

60. Speech act analyses of promise tend to understate the role of the hearer. Speech act theory, of course, recognizes that illocutionary acts call for a hearer. But it fails to appreciate that a full explanation of promising requires consideration of the hearer’s perspective on the act. On speech act theory’s neglect of the role of the hearer, see John R. Searle, “Conversation” in Herman Parret & Jef Verschueren, eds., (On) Searle on Conversation (Philadelphia: J. Benjamins Publishing Company, 1992) 7 at 7. See also Erving Goffman, “Footing” (1979) 25 Semiotica 1 (speech act theory tends to understate the potential complexity of the status of hearer).

61. Note that, just as with the speaker-based illocutionary point, there is nothing in this formulation of hearer-based illocutionary point that necessarily involves the notion of obligation. Although one can say that the hearer has a “right to rely” on the promise, this is to say no more than that the hearer may rely on the promise if she so wishes. The term “right” is not used here in a sense that involves a corresponding duty on someone else’s part.

62. We are assuming, contrary to conventional classification schemes, e.g., Searle, supra note 48, that a speech act can belong to more than one usefully identifiable class.

63. On the relation between assurance and reliance, see Jay Conison, “Assurance, Reliance and Expectation”, Southern Cal. Interdisc. L.J. (forthcoming).

64. For a more comprehensive analysis, see ibid.

65. Note also that when a speaker says “I assure you that I will ϕ,” he ordinarily promises to ϕ. A systematic account of promising as a species of assuring is given in Thomson, supra note 1. Thomson largely agrees with our analysis of assuring, but contends that it must also involve the hearer’s uptake. Ibid, at 296–98. It is unclear just what Thomson means by “uptake.” She has in mind something more than merely hearing the speech act, but less than relying on it; “uptake” seems to be an analog to the contract-law concept of consideration. Thomson’s account of promising differs from ours in two other ways. First, she neglects the structure of speaker reasons resulting from the illocutionary act. Second, she believes that the obligational aspect of promising ipso facto results from the hearer’s uptake. Ibid, at 302.

66. Raz, supra note 22 at 90–91.

67. The similarities and differences can be elaborated further through a simple example. Suppose that Jones, a private in the army, is hiking in the woods as part of a military exercise. He comes to a stream. He can cross at that point only by walking on a log of uncertain strength. If, while pondering what to do, he sees Sergeant Smith on the other side, and Smith orders Jones to cross on the log, Jones has an exclusionary reason by virtue of which he ought to disregard first-order reasons for not walking on the log. If, instead, he had seen Private Wilson on the other side, and Wilson had said, “I assure you Jones, the log will support you,” Jones would have an exclusionary permission by virtue of which he may disregard first-order reasons for not walking on the log. Whether Jones does disregard those first-order reasons will depend on Jones’ assessment of whether, all things considered, he ought to do so.

68. On reliance as the exercise of the exclusionary permission, see Conison, supra note 63.

69. The reason to believe that p, which results from an assurance, appears to be a generalization of the notion of expectation, which, along with reliance, is of great interest to contract theorists. Much of the difference between expectation and reliance can be accounted for in terms of the difference between first-order reasons and second-order reasons. See ibid.

70. Every illocutionary act must have some function for the speaker.

71. The speaker’s intent to get the hearer to understand that she has an exclusionary permission is not the same as an intent that the hearer should actually rely. The former is part of the illocutionary point of promise; the latter is an adventitious intent which may or may not accompany any act of promising. See Raz, supra note 22 at 99. Theories of promising that purport to explain the binding character of a promises in terms of reliance on it often err by focusing on the non-essential intent to induce reliance. See, e.g., MacCormick, supra note 44.

72. This analysis bears some similarities to the analysis of “low promising” in McNeilly, supra note 1 at 71–75. The most important difference is that McNeilly locates the motivation for doing what one has promised to do in some external sanction, rather than in rationality. For this reason, as will be seen more clearly below, McNeilly has not “de-moralized” promising, as he sought to do.

73. See Searle, supra note 3 at 60–61, 177–82; Searle, supra note 43. Elsewhere, Searle describes a promise as a “linguistic fact,” which needs no extralinguistic institution to account for it. John Searle, “How Performatives Work” (1989) 12 Linguistics and Philos. 535, reprinted in (1991) 58Tenn. L. Rev. 371 at 384.

74. For criticisms of this premise, see Toni Vogel Carey, “How to Confuse Commitment With Obligation” (1975) 72 J. of Phil. 276; Altham, supra note 1 at 3–12; Jaako Hintikka, “Some Main Problems of Deontic Logic” in Risto Hilpinen, ed., Deontic Logic: Introductory and Systematic Readings (Dordrecht: Reidel, 1971) 58 at 93–98.

75. It is enough that the question of Jones’ obligation be open, since that alone defeats the claim that the premise is a “tautology.” Searle, supra note 43 at 46.

76. Bronaugh has similarly argued that “the obligation of promises is not fully content-independent.” Richard Bronaugh, “Promises” in L. Becker, ed., Encyclopedia of Ethics (New York: Garland Press, 1992) 1020. In Bronaugh’s view, whether an utterance of “I promise to ϕ” (or the equivalent) gives rise to an obligation to ϕ depends on the speaker’s belief about the hearer’s belief in the lightness of ϕ’ing. This view appears broadly consistent with the statement in the text, since both the hearer’s views about ϕ’ing and the speaker’s belief about the hearer’s views will ordinarily depend on norms of the society in which they live. Furthermore, Bronaugh’s proposal helps show why cases of cross-cultural promising can be morally ambiguous. If speaker and hearer are members of substantially different cultures, the speaker may have no belief (or, at least, no reasonable belief) about the hearer’s belief about the rightness or wrongness of the promised action.

77. Raz, supra note 1 at 219–21.

78. Raz agrees that reasons for action must motivate. Raz, supra note 22 at 32–34.

79. Raz contends that if, in a society in which there is no social practice of promising, “a man communicates to another his intention to undertake by the very act of communication, an obligation to perform an action and confer a corresponding right on his interlocutor, I cannot see how we can avoid regarding his act as a promise.” Raz, supra note 1 at 214. This highlights a tension in Raz’ treatment of the rule behind promissory (and other types of) obligation. On the one hand, the mandatory rule must be accepted by the obligated individual in order for it to give her the relevant reasons. Hence, Raz notes that “[wjhich acts are held obligatory depends on the substantive practical principles one adopts.” Ibid, at 225. On the other hand, the content and validity of the rule must ultimately be a matter of social fact, in order that it not just be an expression of personal preference. Thus Raz urges that a promising rule must be justified on the ground that “the creation of [certain types of] special relationships between people is held to be valuable.” Ibid, at 228. Raz’ discussion of promissory obligation slips back and forth between a view of the promising principle as a social rule, and as a rule that someone has accepted.

80. See William K. Frankena, “Obligation and Motivation in Recent Moral Philosophy” in A.I. Melden, ed., Essays in Moral Philosophy (Seattle: University of Washington Press, 1958) 40; Williams, supra note 39 at 193.

81. For a somewhat different challenge to views that assimilate obligational and non-obligational commitment, see McNeilly, supra note 1 at 76–78.

82. Although the circumstance of an obligation’s having no motivational potential is more common in cases of obligations thrust upon a person—such as obligations to family members—it can occur even in cases of promising and other obligations that are willingly assumed. For example, it might be that Jones promised Smith that he would ϕ, but then subsequently forgot the promise. Not only might Jones be unable to recall having made the promise; he might be unwilling to believe he made it. When asked by Smith to ϕ, he might be completely unmoved. In such a case, Jones no longer can be said to have a formal or exclusionary reason to ϕ. Yet one would not contend that memory lapse should affect his obligation.

83. See generally Philippa Foot, “Reasons for Action and Desires” (1972) 46 Proc. Arist. Soc’y (Supp.) 203 at 206–08.

84. See, e.g., Bernard Williams, “Ought and Moral Obligation” in Bernard Williams, ed., Moral Luck (Cambridge: Cambridge University Press, 1981) 114 at 121.

[W]e do not first have a determinate notion of moral obligation … to which the notions of blame and related reactions are then added. The class of moral obligations … just is the class of oughts to which the notions of blame and related reactions are added.

For a similar view, see H.L.A. Hart, “Legal And Moral Obligation” in Melden, supra note 80 at 82. Cf. Oliver W. Holmes, Jr., “The Path of the Law” (1897) 10 Harv. L. Rev. 457 at 458: “[A] legal duty so called is nothing but a prediction that if a man does or omits certain things he will be made to suffer in this or that way by judgment of the court.”

85. For a discussion of this point, see Williams, supra note 39 at 192.

86. We are here concerned with “obligation” in its everyday sense. Although the concept of legal obligation, too, essentially involves criticism or blame (e.g., damages), a full explanation involves complexities that cannot be pursued here.

87. Hume sought to explain the obligational character of promising in terms of the desire to avoid the reputation for untrustworthiness which supposedly would result from failing to keep a promise. Said Hume: “by making use of this form of words, [an individual] subjects himself to the penalty of never being trusted again in case of failure.” Hume, supra note 1 at 522. But this explanation is implausible. Although a chronic promise breaker likely would earn the sanction of a bad reputation, the failure to fulfil a promise in a single instance rarely has this effect. To the contrary, we accept the fact that promises are not always kept, and we do not consider occasional promise-breaking to be a character defect. One who always kept his promise, no matter what, would likely be thought to have a severe character defect, or even be mentally ill. For a modern attempt to explain the obligational features of promising along Humean lines, see McNeilly, supra note 1 at 72–73.

88. See, e.g., Craswell, supra note 2 at 499–500.

89. Williams, supra note 39 at 178; Melden, supra note 33 at 42.

90. See Anscombe, supra note 1, for a similar explanation.

91. While obligations provide a basis for criticism, they do not necessarily require it. In any given case, there may be countervailing considerations that would make criticism or blame inappropriate. Jones may promise Smith that he will have lunch with her on Monday, but if Jones cancels because he is too busy, Smith and everyone else who knows of the failure may decline to criticize Jones for this breach of obligation because, e.g., they acknowledge the importance of Jones’ getting his work done.

92. See John Rawls, “Two Concepts of Rules” (1955) 64 Phil. Rev. 3.

93. In the society of the Norse sagas, the rules defining social roles and the obligations attaching to them included the rules of vendetta. Obligation is tied to kinship so closely that a near kinsman is a “skyldr fraendi” [“skyldr,” a word etymologically related to “should,” means “owe”]; and “o-skyldr,” which might be naturally, if clumsily, translated as “not connected to one by way of obligation,” means “unrelated.”

Alasdair Maclntyre, “Ought” in Against the Self-images of the Age (Notre Dame, IN: University of Notre Dame Press, 1978) 136 at 143. Maclntyre points out “that promising is an institution in our society comparable to vendetta in Icelandic society, in which the established rules are such that if you make a promise to someone, you ought to do whatever it is, in this sense, that you owe it to him.” Ibid, at 153.

94. Melden, supra note 33; Hardimon, supra note 4.

95. See, e.g., Alasdair Maclntyre, “What Morality Is Not” in Against the Self-images of the Age, supra note 93 at 96 and 106.

96. Williams, supra note 39 at 185. This highlights another difference between reasons and obligations. Negative obligations, such as those relating to murder, should enter into practical deliberation only in exceptional circumstances.

97. On this type of obligation, see generally Melden, supra note 33.

98. Hardimon, supra note 4 at 334–35. For an examination of obligations of this type, and a defense of their centrality in ethical reasoning and everyday life, see ibid, at 342–47.

99. Although a person is ordinarily born into citizenship, one can voluntarily become a citizen of a country through immigration and naturalization. Similarly, although the obligations of fatherhood are ordinarily imposed on a person, one can voluntarily become a parent through adoption.

100. Thus arises the supposed “puzzle about how we can commit ourselves to a course of conduct that absent our commitment is morally neutral.” Fried, supra note 2 at 11.

101. Of course, two individuals can make promises to each other. But in such a case, there are two one-sided promissory obligations.

102. For a similar distinction, see Penner, supra note 2 at 327–28. Cf. Atiyah, supra note 2 at 39:

The purpose of a promise, far from being, as is to often assumed, to create some wholly independent source of obligation, is frequently to bolster up an already existing duty. Promises help to clarify, to quantify, to give precision to moral obligations, many of which already exist…. The promise which is given without any independent reason for it is a peculiarity ….

103. See Joseph M. Perillo & Helen H. Bender, eds., 2 Corbin on Contracts (St. Paul, MN: West Publishing Co., 1995) §5.5.

104. A similar observation is made in Bronaugh, supra note 76, who also distinguishes releasing the speaker from relieving him of responsibility.

105. For another argument against the view that die promisee must benefit from, or want, the promised act, see Raz, supra note 1 at 213–14.

106. Similarly, if H had responded: “OK, I’ll rely on you to ϕ,” she would not have strengthened or confirmed the obligation. She simply would have created another ground for criticism in the event S should not $.

107. Consider the following scenario:

Smith: “You don’t have to pick me up at the airport. Just forget your promise.”

Jones: “No, no. I’ll do it.”

Smith: “You don’t have to. I won’t fault you if you don’t.”

Jones: “I understand, but even though I don’t have to do it, I will. You can count on me.”

Smith: “Okay, okay, I’ll expect you to pick me up. But remember: you don’t have to.” Smith can rationally continue to rely on die promise, even though no one can now criticize Jones for breach of obligation if he fails to pick up Smith.

108. See, e.g., Margaret Gilbert, “Agreements, Coercion, and Obligation” (1993) 103 Ethics 679.

109. Arguably, some coerced promises do give rise to obligations in our society. See Altham, supra note 1 at 10–12. If so, the analysis is unaffected, since it still remains a matter of social fact which duressed promises do, and which do not, give rise to obligations. The analysis would simply have to be limited to those promises that do not give rise to obligations.

110. It is natural and understandable for the coerced individual to want to resist performing the speech act of promising, precisely to avoid the reason-generating features of the act. Why else would the coercer not be content with an immediate, casual statement by the victim that he promises to do what the coercer wants. It would be obvious in a such a case that the speaker had performed a sham speech act, and had not given himself a reason to do what the coercer wants.

111. Cf. Woozley, supra note 2 at 289–90:

Conceptual questions about promises … become the questions of how far we can deviate from the paradigmatic case of a simple two-person promise, in the direction of dropping one element or the other, before it becomes reasonable to deny that the performance was a promise.

112. The same conclusions appear to hold about promises to perform acts prohibited by a canonical obligation. In our society, a sincere promise to murder does not create an obligation, although it does give rise to the usual set of reasons.

113. Cf. Bronaugh, supra note 76 (in the case of a threat, the hearer has no power of release).

114. See von Wright, supra note 2 at 88–89; Vera Peetz, “Promises and Threats” (1977) 86 Mind 578. Of course, one who fails to carry out a threat may be open to criticism on other grounds.

115. Alternatively, the speaker might create the exclusionary permission but make it unreasonable to exercise it, i.e., make it unreasonable to rely.

116. Employee Retirement Income Security Act of 1974, Pub L. No. 93–406,88 Stat. 829 (substantially codified as amended at 29 U.S.C. §§ 1001 et seq. and in various portions of the Internal Revenue Code).

117. See generally Jay Conison, “Foundations of the Common Law of Plans” (1992) 41 DePaul L. Rev. 575 at 588–89, 599–600.

118. Ibid. at 601–10. A similar, contemporary example involves employer representations regarding the permanence or security of employment, possibly at variance with its at-will character. Courts are developing law to govern the issues of when an employer has made a “promise” (in a specialized sense) that limits its power to terminate employees, and when the promise is enforceable. See, e.g., Rood v. General Dynamics Corp., 507 N.W.2d 591 at 606–07 (Mich. 1993).

119. See, e.g., Burten v. Milton Bradley Co., 763 F.2d 461 (1st Cir. 1985); Sylvania Elec. Prods., Inc. v. Brainerd, 166 U.S.P.Q. 387 (D. Mass. 1970); Davis v. General Foods Corp., 21 F. Supp. 445 (S.D.N.Y. 1937); Downey v. Gen. Foods Corp., 286 N.E.2d 257 (N.Y. 1972).

120. However, note the possibility of a visitor making, in effect, an insincere implied promise. Analogously to the everyday insincere promisor who utters “I promise to ϕ” but has no intention of ϕ’ing, a visitor might enter the museum having no intention of doing what he (here implicitly) has said he would do. For present purposes, we are concerned only with the sincere implied promisor.

121. It would be odd to appeal to a supposed promise to explain Jones’ not letting go. If one asks Jones why he does not let go, he would say that there is no reason to do so, or that letting go would hurt Smith, or that it would defeat the whole purpose of his and Smith’s work.

122. The third theme is more significant in the law of waiver of Miranda rights. See, e.g., Moran v. Burbine, 475 U.S. 412 at 421 (1986).

123. Bram v. United States, 168 U.S. 532 at 542–43 (1897). There is, however, no per se rule invalidating confessions induced by promises. See, e.g., Brady v. United States, 397 U.S. 742 at 753–54 (1970).

124. E.g., Holland v. McGinnis, 963 F.2d 1044 at 1051 (7th Cir. 1992), cert, denied, 113 S. Ct. 1053 (1993):

Inflating evidence of Holland’s guilt interfered little, if at all, with his “free and deliberate choice” of whether to confess …, for it did not lead him to consider anything beyond his own beliefs regarding his actual guilt or innocence, his moral sense of right and wrong, and his judgment regarding the likelihood that the police had garnered enough valid evidence linking him to the crime. In other words, the deception did not interject the type of extrinsic considerations that would overcome Holland’s will by distorting an otherwise rational choice of whether to confess or remain silent.

125. E.g., State v. Smith, 512 A.2d 189 at 196 (Conn. 1986).

126. E.g., United States v. Haddon, 927 F.2d 942 at 945 (7th Cir. 1991) [hereinafter Haddon].

127. In the law of confessions, “involuntary” is a catch-all term used to characterize a wide variety of Constitutional infirmities. “A confession is voluntary if it results from the free choice of a rational mind, if it is not a product of coercive police conduct, and if under all the circumstances its admission would be fundamentally fair.” State v. Mikulewicz, 462 A.2d 497 at 501 (Me. 1983). In United States v. Long, 852 F.2d 975 at 980 (7th Cir. 1988), Judge Easterbrook argued in a concurring opinion that confessions induced by promises—which are counted as “involuntary” confessions—are in fact voluntary from the perspective of contract law. The point is true, but irrelevant, since “involuntary” in the law of confessions means something quite different from what it means in contract law.

128. United States v. Rutledge, 900 F.2d 1127 at 1130 (7th Cir.) (Posner, J.), cert, denied, 498 U.S. 875 (1990). The court continued:

The police are allowed to play on a suspect’s ignorance, his anxieties, his fears, and his uncertainties; they just are not allowed to magnify those fears, uncertainties, and so forth to the point where rational decision becomes impossible.

Ibid. at 1130. See also Miller v. Fenton, 796 F.2d 598, 605 (3d Cir.), cert, denied sub nom. Miller v. Neubert, 479 U.S. 989 (1986) (decision to confess must be “a product of the suspect’s own balancing of competing considerations”).

129. See, e.g., Welch v. Butler, 835 F.2d 92 at 95 (5th Cir.), cert, denied, 487 U. S. 1221 (1988) (confession resulting from prayer session with police officer; “[a]t most, the police set up a situation that allowed Welch to focus for some time on those concerns [about salvation and divine forgiveness] with a fellow Christian in the hope that his desire to be saved would lead him to confess”).

130. E.g., Spano v. New York, 360 U.S. 315 (1959) (“official pressure, fatigue and sympathy falsely aroused” invalidated confession). On the other hand, where an apparent distortion of the deliberative process derives from sources other than official conduct, the confession is not necessarily inadmissible. Colorado v. Connelly, 479 U.S. 157 (1986) (mental illness of defendant led him to seek out police and confess).

131. On the other hand, courts allow law enforcement officers a fair amount of leeway in deceiving suspects. “These ploys may play a part in the suspect’s decision to confess, but so long as the decision is a product of the suspect’s own balancing of competing considerations, the confession is voluntary.” Miller v. Fenton, 796 F.2d 598 at 605.

132. E.g., United States v. Walton, 10 F.3d 1024 at 1028 (3d Cir. 1993) [hereinafter Walton].

133. E.g., Streetman v. Lynaugh, 812 F.2d 950 at 957 (5th Cir. 1987).

134. See, e.g., Walton, supra note 132. Where a promise plays no significant role in deliberation, however, a confession which follows it is not thereby invalidated. See, e.g., State v. Walton, 769 P.2d 1017 at 1025–26 (Ariz. 1989), aff’d, 497 U. S. 639 (1990):

It is clear from the context of the whole interrogation that the defendant did not respond to the statement as if he understood it as a promise …. When he at length offered some information, he did so not in response to a promise, but due to the continued confrontation with known details and the effect that had on his guilt or fear of retribution, compounded by the need to provide more facts to extricate himself after he had implicated himself…. The actions of the defendant show that he did not treat the detective’s statement as a promise.

Moreover, there is no black-letter rule that a confession induced by a promise is inadmissible. United States v. Guerrero, 847 F.2d 1363 at 1366 (9th Cir. 1988) [hereinafter Guerrero); State v. Starling, 456 A.2d 125 at 127–29 (N.J. Super. Ct. 1983), aff’d, 504 A. 2d 18 (N.J. Super. Ct. App. Div. 1985), cert, denied, 511 A.2d 658 (N.J. 1986). For example, a law enforcement officer’s promise of a so-called collateral benefit does not render a confession inadmissible. E.g., State v. Holloman, 731 P.2d 294 at 300 (Kan. 1987) (release of suspect’s brother).

135. See, e.g., Guerrero, supra note 134; United States v. Robinson, 698 F.2d 448 at 455 (D.C. Cir. 1983) [hereinafter Robinson].

136. See United States v. Fraction, 795 F.2d 12 at 15 (3d Cir. 1986) [hereinafter Fraction]. Similar reasoning has been applied to apparent threats. See Neal v. State, 522 N.E.2d 912 at 913 (Ind. 1988) (defendant was arrested along with his girlfriend and told that the girlfriend would go to jail and her four children would be turned over to the Welfare Department; this was “simply a factual statement concerning the situation”, rather than a threat).

137. Cf. Fraction, supra note 136 at 15, where the explanation for the rule is said to be that a promise must be of something “within the control of the promisor.”

Note that, despite what the courts say, a statement about cooperation being brought to the attention of the prosecutor and judge can indeed function as an assurance, and be intended by the officer to do so, even if a fully reflective person would not treat the statement as such. In refusing to deem such a statement a promise, courts may be assessing it under a reasonable person standard. Cf. Haddon, supra note 126 at 946 (coercion to be determined by reference to impact on a reasonable person under the circumstances).

138. For example, one court said of an interrogator’s statement that “cooperating defendants generally ‘fared better time-wise,’” that it merely gave the defendant a ‘“chance to make an informed decision’” about cooperating with the government. United States v. Nash, 910 F.2d 749 at 752–53 (11th Cir. 1990).

139. Fisher v. State, 379 S.W. 2d 900 at 902 (Tex. Ct. Crim. App. 1964); People v. Dandridge, 505 N.E.2d 30 at 32 (111. App. Ct.1987).

140. United States v. Conley, 859 F. Supp. 830 at 839 n. 6 (W.D. Pa. 1994).

141. Robinson, supra note 135 at 455; People v. Stachelek, 495 N.E.2d 984 at 991 (Ill. App. Ct. 1986). If the promise is not kept, however, it may make the confession objectionable on grounds that the police acted unfairly. See Santobello v. New York, 404 U.S. 257 (1971).

142. Drew v. State, 503 N.E.2d 613 at 617 (Ind. 1987) (collecting cases); Ex Parte Siebert, 555 So. 2d 780 at 782 (Ala. 1989), cert, denied, 497 U.S. 1032 (1990); People v. Hendricks, 131 P.2d 1350, 1354 (Cal. 1987).

143. Restatement (Second) of Contracts §1 (1981).

144. Indeed, modem contract law is based on the willingness of courts to find such promises in a contract. In Slade’s Case, the Court of King’s Bench was willing to allow an action of assumpsit to proceed for breach of a simple contract because:

[E]very contract executory imports in itself an assumpsit, for when one agrees to pay money, or to deliver any thing, thereby he assumes or promises to pay, or deliver it….

Slade’s Case, Eng. Rep. 1074,1077 (K.B. 1602).

145. Restatement (Second) of Contracts, supra note 143 at §§17–18.

146. Ibid. at § 1.

147. Ibid. at § 2. The First Restatement is less explicit, but the comments suggest that promise in the everyday sense is meant. Restatement of Contracts § 2 & comment a (1932).

148. Conison, supra note 117 at 593–98.

149. Supra note 143 at § 77, illustration 9. See generally supra note 103 at § 5.27.

150. The Uniform Commercial Code makes this very clear, providing that:

A lawful agreement by either the seller or buyer for exclusive dealing in the kind of goods concerned imposes unless otherwise agreed on obligation by the seller to use best efforts to supply the goods and by the buyer to use best efforts to promote their sale.

U.C.C. § 2–306(2).

151. Restatement (Second) of Contracts, supra note 143 at §§ 17, 18, 71.

152. Ibid. at § 17, comment b. See also supra note 103 at 38.

153. E.g. Fried, supra note 2 at 46.

154. The term “invitation” is chosen as a relatively neutral description for this illocutionary component.

155. Hintikka, supra note 74 at 95. More accurately, the argument shows that, even if (3’) does result from (3), the fact is irrelevant to any obligation which arises from the fulfilment of the condition.

156. The underlying problem with (3’) may simply be that the content is a conditional event, not an action; thus, it cannot be the subject of an obligation.

157. It would seem that the power to release, too, is a point of conditional promises, subject to the emendation that the obligation in question is the one resulting from fulfilment of the condition.

158. Corbin points out that not all bilateral contracts are formed through a sequence like (2’), and that (2”) may in fact accurately represent the formative acts. For example, a mediator may propose that Smith ϕ’s and Jones ψ’s; to which both parties assent. An act like (2”) is often what takes place at a closing. See Joseph M. Perillo, ed., 1 Corbin on Contracts (St. Paul, MN: West Publishing Company, 1993) at § 1.12.

159. Joint acts are very common; everyday examples are walking together, playing a duet, and conversing. Yet, the notion of joint action has proven difficult to analyze. For some efforts, see Gilbert, supra note 30; Searle, supra note 37; John Searle, “Collective Intentions and Actions” in Philip R. Cohen, Jerry Morgan & Martha E. Pollock, eds., Intentions in Communication (Cambridge, MA: MIT Press, 1990) at 404.

160. See Craswell, supra note 2, for what is in effect an argument that theories of the role of promise in contract formation can tell us little about the obligational character of the relationship so formed.

161. See generally Ian R. MacNeil, The New Social Contract: An Inquiry into Modern Contractual Relations (New Haven, CT: Yale University Press, 1980).

162. It seems preferable to call the latter an obligational point, rather than a performative point, since in the case of contract (and other relationships involving voluntary or institutional obligations) the obligation does not result straightforwardly from a performative.

163. Each party also has first-order and exclusionary reasons to take certain actions. However, these illocutionary features play no role in the basic legal notion of contract, and so will not be discussed here.

164. For a theory of a purely contingent relation between rules and their subject matter, see Frederic Schauer, Playing by the Rules (Oxford: Clarendon Press, 1991).

165. This is not a radical insight. For example, the authors of Corbin on Contracts recognize that to ask why a bilateral contract is enforceable is to ask the wrong question:

[W]e must start out with the general proposition that mutual promises are consideration for each other. We do not need a reason for holding that a bargained for promise is consideration. Instead we must find a reason why sometimes it is not and deal with these exceptions as they arise.

2 Corbin on Contracts, supra note 103 at 129.

166. E.g., Goetz & Scott, supra note 2.