Skip to main content Accessibility help


  • Gregory Klass (a1)


The doctrine of promissory fraud holds that a contractual promise implicitly represents an intent to perform. A promisor's conditional intent to perform poses a problem for that doctrine. It is clear that some undisclosed conditions on the promisor's intent should result in liability for promissory fraud. Yet no promisor intends to perform come what may, so there is a sense in which all promisors conditionally intend to perform.

Building on Michael Bratman's planning theory of intentions, this article provides a theoretical account of the distinction between “foreground” and “background” conditions on intentions in general and then explains why foreground conditions on a promisor's intent to perform are likely to result in material promissory misrepresentation, while background conditions are not. The difference between foreground and background conditions lies in whether the agent accepts the satisfaction of the condition for the purposes of her practical reasoning. A promisor's nonacceptance of a condition on her intent to perform is material because it is likely to affect her preperformance deliberations and investment in the transaction, as well as her willingness to seek agreement with the promisee on how to fill contractual gaps.



Hide All

1. Hillcrest Center, Inc. v. Rone, 711 So. 2d 901, 903 (Ala. 1997).

2. Id. at 906.

3. Ian Ayres & Gregory Klass, Insincere Promises: The Law of Misrepresented Intent 26–29 (2005).

4. This is laid out in Michael E. Bratman, Intention, Plans, and Practical Reason (1987).

5. Ayres & Klass, supra note 3, at 35.

6. I use “cause” here in the broadest sense possible. I mean to suggest nothing about the sort of causal mechanisms at work.

7. Aditi Bagchi, The Accidental Promise: Remaking the Law of Misrepresented Intent, 2008 U. Ill. L. Rev. 985 (reviewing Ayres & Klass, supra note 3; Craswell, Richard, Taking Information Seriously: Misrepresentation and Nondisclosure in Contract Law and Elsewhere, 92 Va. L. Rev.565 (2006).

8. Yaffe, Gideon, Conditional Intent and Mens Rea, 10 Legal Theory273 (2004).

9. Yaffe does not need to, for he argues that the line for criminal liability cuts across that distinction. Id. at 290–298.

10. Luca Ferrero, Conditional Intentions, Noûs (forthcoming) (references are to paragraph numbers in version 2.5, Nov. 30, 2008, on file with the author). I am especially grateful for Ferrero's permission to cite and discuss a paper that he is still in the process of revising. It might turn out that some of my comments or criticisms do not apply to the published version of Ferrero's article. If so, they should be read as snippets from the middle of a conversation—one from which I have greatly benefited.

11. See, e.g., Scott, Robert E., The Case for Formalism in Relational Contracts, 94 Nw. U. L. Rev.847 (2000). For an overview, see Kimel, Dori, The Choice of Paradigm for Theory of Contract: Reflections on the Relational Model, 27 Oxford J. Legal Stud.233, 238, 250–253 (2007).

12. Campbell, Kenneth, Conditional Intention, 2 Legal Stud.77, 89–90 (1982). Jack Meiland similarly argues that if an agent does not realize that the nonoccurrence of condition C would affect his decision to X, “C does not enter into his intention at all, although C might become part of his intention if he realized that C was relevant to the doing of X or to his other purposes.” Jack W. Meiland, The Nature of Intention 20 (1970).

13. Michael E. Bratman, Intention, Belief, Practical, Theoretical, in Spheres of Reason (J. Timmerman et al. eds., forthcoming) (Sept. 22, 2008 draft) §4, at 12–13, available at\_IBPT\_2008\_final.pdf.

14. I ask the reader for the sake of the example to grant me that Jones would not be able to buy her way out of the sales contract.

15. Bratman discusses the difference between intending and having a belief about one's intentions in Bratman, Michael E., Cognitivism about Practical Reason, 102 Ethics117 (1991), reprinted in Faces of Intention: Selected Articles on Intention and Agency 250 (1999), and in Bratman, Intention, Belief, supra note 13. The difference between intending and ascribing an intention to oneself is also central to Wilfrid Sellars's work on intentions. See, e.g., Sellars, Wilfrid, On Reasoning about Values, 17 Am. Phil. Q.81, 84 (1980).

16. Cartwright, J.P.W., Conditional Intention, 60 Phil. Stud.233, 240 (1990). Donald Davidson makes a similar suggestion:

[B]ona fide conditions are ones that are reasons for acting that are contemporary with the intention. . . . Changing his mind is a tricky case, but in general someone is not apt to view a possible future change of intention as a reason to modify his present intention unless he thinks the future change will itself be brought about by something he would now consider a reason.

Donald Davidson, Intending, in Essays on Actions and Events 83, 94–95 (1980).

17. Cartwright, supra note 16, at 240.

18. Id. at 241–242.

19. Id. at 242.

20. This is a first, rough take on the distinction. For a more nuanced description of it, see Section II.B.

21. Ferrero notes the distinction as well. Ferrero, supra note 10, ¶¶2.2–2.4. While he also uses the term “enabling conditions,” Ferrero calls “restrictive conditions” what I designate “elective conditions.”

22. Ferrero suggests that “[t]here is a perfectly acceptable sense in which the satisfaction of an enabling condition counts as a reason in support of the intended action.” Id. ¶2.4. If so, this is a rather attenuated sense as compared to elective conditions.

23. Compare Ferrero's (DS-E). Id.

24. Thanks to Gideon Yaffe for emphasizing this point to me.

25. Mark Bittman, Food Matters: A Guide to Conscious Eating with More than 75 Recipes 72 (2008).

26. Bratman, Intention, Plans, supra note 4, at 34; see also id. at 24–26.

27. This theory is laid out in id., passim.

28. See id. at 32–33.

29. See id. at 134–138; Yaffe, Gideon, Trying, Intending, and Attempted Crimes, 32 Phil. Topics505, 512–514 (2004); Ferrero, supra note 10, ¶5.1.

30. Bratman, Intention, Belief, supra note 4, at 1.

31. Bratman, Intention, Plans, supra note 4, at 33.

32. Id. at 33.

33. Bratman, Intention, Belief, supra note 13, at 1.

34. Bratman, Intention, Plans, and Practical Reason, supra note 4, at 34.

35. See Michael E. Bratman, Practical Reasoning and Acceptance in a Context, in Faces of Intention: Selection Essays on Intention and Agency, at 15 (1999). Bratman also discusses the basic idea in Bratman, Intention, Plans, supra note 4, at 36–37.

36. Bratman also discusses acceptances tied to special relationships to others, such as loyalty, and the acceptance of propositions that are preconditions of practical reasoning, such as that I have a free will. See Bratman, Practical Reasoning, supra note 35, at 21–26.

37. Id. at 29.

38. Id. at 20.

39. Ferrero generally assumes that elective conditions are intended as sufficient conditions, while enabling conditions are necessary ones. See Ferrero, supra note 10, ¶¶2.2–2.3.

40. A more rigorous formalization would put a proposition on each side of the operator; that is, (2) should be understood as shorthand for “A intends (C → A xes).”

41. Policy-based intentions can also be strictly conditional or biconditional.

42. Luca Ferrero suggests in correspondence that this might mean that “conditioned intentions” are not really intentions as the agent has not yet decided what to do in scenarios in which the condition is satisfied but only in scenarios in which the condition is not satisfied. This would explain why the contrapositive is the more natural expression. According to this objection, only strictly conditioned intentions are true intentions, while what I call “conditioned intentions” are mere constraints on an agent's future deliberations. This objection is of a piece with Ferrero's rejection of the truth-functional reading of conditional intentions. Ferrero would allow that A might intend (not-C → not-x) but rationally not intend the contrapositive, (x → C).

I am not convinced that a conditioned intention expressed as such is not an intention, although of course I grant that it does not specify what the agent shall do if the condition is satisfied. The question is in part a semantic one: How should we use the word “intention”? Conditioned intentions are plans that constrain one's future deliberations and actions. I argue in Section III that they do so in ways that parallel strictly conditional and biconditional intentions. And because I advocate a truth-functional reading of conditional intentions, I am willing to grant that a conditioned intention is equivalent for the purposes of the agent's deliberations to the strictly conditional contrapositive—which would make it odd to say that it is not an intention. All of this recommends, to my mind, a use of “intention” that is more inclusive.

43. If A accepts the satisfaction of C in the sense described in the next section, then she might simply intend to x. If she does not accept the satisfaction of C, then she might rationally be able only to try to x. For complications in the distinction between intending and trying, see Bratman, Intention, Plans, supra note 4, at 113–22; Carl Ginet, Trying to Act, in Freedom and Determinism 89 (J.K. Campbell et al. ed., 2004); Yaffe, Trying, supra note 29.

44. The theory I describe in this and the next paragraphs builds on David Pears's analysis of implicature for expressions of intent in Pears, Intention and Belief, in Essays on Davidson: Actions and Events 75, 84–88 (Bruce Vermazen & Merrill B. Hintikka eds., 1985).

45. See Ayres & Klass, supra note 3, at 26–29.

46. See Ferrero, supra note 10, ¶¶4.2–4.3, 6.3, 7.2, 8.2.

47. As an example, suppose A intends (x → C) and that A accepts that C is not satisfied. On the proposed theory, her conditioned intention (x → C) is therefore background-conditional. Because A accepts that C is not the case, in that practical context she should also accept that not-C is the case; that is, she should accept satisfaction of the condition of her functionally equivalent strictly conditional intention (not-C → not-x), which is therefore also background-conditional. It is not difficult to show that the rational constraints imposed by A's intention (x → C) where A accepts the nonsatisfaction of C are the same as the rational constraints imposed by an intention (not-C → not-x) where A accepts the satisfaction of not-C. Similar arguments can be constructed for strictly conditional and biconditional intentions.

This is not a merely technical point; it corresponds to how we describe intentions. Suppose Jones has a policy-based intention of donating found money to charity. If she reflects on the matter, she might decide that if she ever wins the lottery, she will give the money away. Because she never buys lottery tickets, she accepts that the condition of winning the lottery will not be satisfied. But her attitude is not inert. Though it creates no rational pressure to intend the means of donating any winnings (such as choosing a charity), it imposes consistency constraints on Jones's practical reasoning. The attitude entails, for example, that it would be irrational for Jones to plan on funding renovations to her coffee shop by playing the lottery (and not only because of the low probability of winning). Jones's intention to give any lottery winnings to charity is background-conditional because she accepts the nonsatisfaction of the condition.

Thanks to Luca Ferrero for pointing out the complications raised by the contrapositive on my approach.

48. See Yaffe, Conditional Intent, supra note 8, at 292–293.

49. See id. at 294–295; Ferrero, supra note 10, ¶5.1 n.37.

50. See Hector Neri Castañeda, Some Reflections on Wilfrid Sellars' Theory of Intention, in Action, Knowledge and Reality: Critical Studies in Honor of Wilfrid Sellars 27, 34–35 (H.N. Castañeda ed., 1975).

Ferrero advances a different argument against a truth-functional reading: that it does not capture what it means to carry out an intention.

The only way in which a conditional intention [to ϕ if C] can be genuinely carried out is by ϕ-ing in those circumstances when C obtains by the time of the intended action. There is nothing that counts as carrying out a conditional intention when, by the time of action, C can no longer obtain. . . . The alleged discharging of a conditional intention when C can no longer obtain would not amount to a practical achievement.

Ferrero, supra note 10, ¶2.8; see generally id. ¶¶2.5–3.2.

My claim here concerns only the deliberative constraints that conditional intentions impose. The argument is not designed to show that a truth-functional reading captures every aspect of conditional intentions—though it would be an important piece of such a showing. Still, I am not persuaded by Ferrero's argument, which I think proves too much. Suppose that I unconditionally intend to stop smoking and that shortly after I adopt the intention, the government bans the sale of tobacco products. I can no longer smoke. In this case, as in Ferrero's example, the world has conspired to realize the object of my intention without any effort by me. This does not mean that I did not intend to stop smoking. It rather reflects the fact that carrying out an intention, whether conditional or not, requires more than moving into a state of the world in which the propositional object is true.

51. Thus it might seem strange to say that Jones might realize an intention to expand if the theater opens (a strictly conditional intention) by taking actions to prevent it from opening, or odd to say that Jones could realize her intention to expand only if her architect approves of the space (a conditioned intention) by going out of her way to convince the architect that it does not suit her needs.

52. Jones is ignorant of the contractual duty of good faith.

53. This claim is similar to Wilfrid Sellars's comments on the difference between the “logic, strictu sensu, of intentions” and the conceptual truth that some states of affairs “belong in their content by virtue of being believed to be the case or about to be the case,” while others “are considered by the agent to be ‘up to me.’” Wilfrid Sellars, Conditional Promises and Conditional Intentions (Including a Reply to Castañeda), in Agent, Language, and the Structure of the World 195, 213 (J. Tomberlin ed., 1983). Sellars argues that the latter differences result from the particular reasoning that produces the intention. Id. at 217–219. It is also worth keeping in mind Sellars's suggestion that our intuitions in these cases sometimes concern “not the logic of intentions as such, but the logic (in a suitably broad sense) of conditional commitments,” which one can undertake with an expression of intent. Id. at 206.

54. This thesis marks another point where my approach differs from Ferrero's. Ferrero would analyze the content of a contingency plan, for example, as including the fact that it is subordinated to a different, primary pursuit. Ferrero, supra note 10, ¶6.6 n. 60.

55. See, e.g., Bratman, Intention, Plans, supra note 4, at 28–35, 64–70, 87–91.

56. Compare Bratman's comments on the inconsistency of intending x and believing that one will not x. Id. at 37–41.

57. Ferrero, supra note 10, ¶6.5. Ferrero puts this point in terms of the agglomerativity constraint, but it holds for consistency as well.

58. This last point argues that Jones's background conditioned intention to expand only if the coffee shop is still in business imposes consistency constraints of the same sort as her belief that if she expands she will need to hire more employees. While one refers to a precondition of expanding and the other a necessary side effect of doing so, both are entailments of performing the intended act and so figure into what is consistent with so acting.

This explains how known side effects enter into the content of future-directed intention without, in an important sense, being intended. A's intention to x with side effect S screens out an intention to y, where a condition of the possibility of ying is not-S. While this is beyond the scope of this article, I believe this point requires refining Bratman's claim that while choice is holistic—encompassing the entire scenario intended—future-directed intentions are not. See Bratman, Intention, Plans, supra note 4, at 153–155. Bratman is right to observe that known side effects do not impose a pressure to intend the means of their realization. But they do figure into the consistency constraints that the intention imposes. Future-directed intentions are also holistic, though not every element in the intentional object plays the same role.

59. Ferrero, supra note 10, ¶5.2.

60. The same holds for Jones's belief that no condemnation of the property is a necessary condition of purchasing if she then discovers that it will not be condemned. It makes no difference whether the necessary condition is elective or enabling.

61. But not in all cases. If, for example, A has a reason to intend or attempt to x other than the desirability of xing, such as in toxin-puzzle-type cases, a foreground enabling condition might not affect how much it is rational for A to invest in the means of xing.

62. Yaffe makes a similar point. Yaffe, Conditional Intent, supra note 8, at 294.

63. Ferrero, supra note 10, ¶5.2.

64. Bratman, Intention, Plans, supra note 4, at 33. It is an interesting question how we should describe cases in which satisfaction of a necessary condition on an intention to x is so unlikely that it reduces the rational pressure to intend necessary means of xing to the point where, given the costs, it is no longer rational to intend those means. Bratman suggests that in such cases, the agent no longer intends to x:

Granted, my uncertainty about the efficaciousness of my plan for stopping at the bookstore might reasonably lead me not to bother with that plan, especially if I find the alternative path home nicer in other respects. But if I am persuaded by this then what I should do is give up on the intention to stop at the bookstore, not intend to stop there but still not intend known necessary means.

Bratman, Intention, Belief, supra note 13, at 32. But should we reach the same conclusion if there is no “alternative path”? Yaffe argues that “in this case, my conditional intention places me under rational pressure to adopt the means. . .; it is just that I am under stronger rational pressure not to decide on any action from which I can expect a loss.” Yaffe, Conditional Intent, supra note 8, at 294. If the conditional intention still relaxes the consistency and agglomerativity constraints on the agent's deliberations, this seems right. If, however, the conditional intention does not relax those constraints, the agent has accepted the nonsatisfaction of the condition and the attitude is better characterized at most as a background-conditional intention.

65. Where the intention is foreground biconditional, the agent should also intend (not-C →F not-x). If not-xing takes some effort on the agent's part, means-end coherence will also require that the agent eventually intend the means of not-xing.

66. The above analysis has consequences for the proper analysis of generic ceteris paribus clauses, such as “provided that it is feasible” or “provided that it is advisable,” which can be attached to any ascription of intention. Ferrero suggests that such generic qualifications should be included in a complete statement of the background conditions on a person's intention. Ferrero, supra note 10, ¶¶8.1–8.2. Translated into my notation, Ferrero would analyze A's foreground unconditional and background-conditional intention to x as:

A intends [(x →B P1, P2. . . Pm, & x is feasible) & (C1, C2. . . Cq, & x is advisable →Bx)],

where “P1, P2. . . Pm” stands for the complete list of “actually articulated” background enabling conditions on A's intent and “C1, C2. . . Cq” stands for the complete list of “actually articulated” background elective conditions on A's intent. (I take this to be a translation of Ferrero's (DS-E*) for a foreground unconditional intention; my parsing assumes that Ferrero's “given” operator is meant to describe necessary conditions when applied to enabling conditions and sufficient ones when applied to elective ones.) Ferrero stipulates that a condition on A's intent is “actually articulated” if A explicitly entertains it or might have entertained it at an earlier time or it is operative in A's psychology. Id. ¶8.1. But once we have included in the analysis of A's intention to x all of the background conditions that are operative in A's psychology, it is difficult to see what the generic feasibility or advisability conditions add to the analysis, not to mention conditions that A has or might have entertained but that are not operative, if such conditions could exist. For any potential condition Ci (or Pi) on A's intention to x, either A is rationally disposed to treat Ci as a condition of her intention to x, or she is not. If A is so disposed, then Ci is “operative” in A's psychology. If A is not so disposed, then the condition does not play a role in her practical reasoning. Ferrero says that the advisability placeholder is meant to capture the agent's acknowledgment of “yet unarticulated conditions that, if taken to obtain, would give her a good enough reason not to ϕ.” Id. ¶8.5. But such an acknowledgment belongs to theoretical, not practical reasoning.

67. Ayres & Klass, supra note 3, at 35.

68. Chung v. Kaonohi Center Co., 618 P.2d 283 (Haw. 1980). Chung affirmed an emotional distress award based on a tort theory of wanton or reckless conduct in the breach. The Hawaiian court has since rejected that rule, abrogating the holding in Chung, and ruled that emotional damages are available only when the defendant's behavior violates an independent tort-based duty. Francis v. Lee Enters., Inc., 971 P.2d 707, 717 (Haw. 1999). But as Ayres and I argue, the Chung defendants' behavior looks like it also satisfies the elements of promissory fraud, which would support the award of punitive damages in the case even under the new rule. Ayres & Klass, supra note 3, at 28.

69. Chung, supra note 68, at 286.

70. Bratman makes a related point: “the dispositions to figure out how to do what one intends, and to settle on needed preliminary steps, provide support for the expectation that an agent will both be in a position to do what she intends and succeed in doing it.” Bratman, Intention, Plans, supra note 4, at 18.

71. Hillcrest, supra note 1.

72. I take this to be implicit in Bratman's description of simplification. Bratman, Practical Reasoning, supra note 35, at 21–22. Compare Bratman's description of the cost-benefit analysis involved in when it is rational to reconsider one's intentions. Bratman, Intention, Plans, supra note 4, at 64–70.

Bratman suggests a number of other possible reasons for acceptance, including special relations with others. Bratman, Practical Reasoning, supra note 35, at 25. Some contractual promises—those that establish a moral relationship between the parties—might create the sort of special relation that makes it rational for the promisor to deliberate as if her conditional intent to perform were unconditional.

73. This is especially the case where the promisor knows more than the promisee about the likely satisfaction of the condition. It is therefore relevant that the Chung defendants were in the business of leasing such properties and therefore probably knew more than the plaintiffs about the chances that a more profitable tenant would come along.

74. See Bratman, Intention, Plans, supra note 4, at 37.

75. Bratman, Michael E., Shared Cooperative Activity, 101 Phil. Rev.327 (1992), reprinted in Faces of Intention: Selected Essays on Intention and Agency 93 (1999); Bratman, Michael E., Shared Intention, 104 Ethics97 (1993), reprinted in Faces of Intention: Selected Essays on Intention and Agency 109 (1999).

76. As Robert Scott says, “[w]e are all relationalists now.” Scott, supra note 11, at 852.

77. Llewellyn, Karl N., What Price Contract?—An Essay in Perspective, 40 Yale L.J.704, 712–713 (1931). The two classic modern works in the field are Macaulay, Stewart, Non-Contractual Relations in Business: A Preliminary Study, 28 Am. Soc. Rev.55 (1963); and Macneil, Ian, Contracts: Adjustment of Long-Term Economic Relations under Classical, Neoclassical, and Relational Contract Law, 72 Nw. U. L. Rev.854 (1978).

78. The example is adapted from Bratman. See Bratman, Cooperative Activity, supra note 75, at 98–99.

79. For the difference between joint intentional and shared cooperative activities, see Shapiro, Scott J., Law, Plans, and Practical Reason, 8 Legal Theory387, 394–401 (2002).

80. More, for example, about the degree of cooperation, about the extent to which the parties' plans must mesh, and about the nature of the shared project. Any thinking about the degree of cooperation in contractual relationships should take account of Shapiro's work on shared intentions in institutional, authority-based structures. Id. passim. With respect to the nature of the shared project, the above description purposively leaves room for a thicker account of shared projects than Daniel Markovits's description of contractual exchanges as joint intentional activities arising out of relatively thin promissory commitments. Markovits, Daniel, Contract and Collaboration, 113 Yale L.J.1417, 1456–1463 (2003); see Klass, Gregory, Three Pictures of Contract: Power, Duty and Compound Rule, 83 N.Y.U. L. Rev.1726, 1747–1750 (2008) (suggesting an alternative description).

81. Craswell, Richard, Taking Information Seriously: Misrepresentation and Nondisclosure in Contract Law and Elsewhere, 92 Va. L. Rev.565, 628 (2006).

82. Id. at 629.

83. “[I]n Ayres and Klass's strictly economic account, this subjective intent is drained of any agent particularity. As such, Ayres and Klass could do well enough without the complicating notion of intent. . . . It would be cleaner to speak of the probability of performance in purely objective terms without alluding to choice.” Aditi Bagchi, The Accidental Promise: Remaking the Law of Misrepresented Intent, 2008 U. Ill. L. Rev. 985, 992; see also id. at 1002, 1006–1007.

84. Id. at 1007.

85. John R. Searle, Speech Acts: An Essay in the Philosophy of Language 54–63 (1969); see also Ayres & Klass, supra note 3, at 29 (describing moral basis for promissory representations). If this is right, the bare representation of an intent to perform is neither a conversational implicature nor a conventional one. See Stephen C. Levinson, Pragmatics 126–132 (1983) (describing the distinction between conversational and conventional implicature).

86. Bagchi, supra note 83, at 1005 n. 83.

87. See Ayres & Klass, supra note 3, at 61.

88. For a brief account of this Gricean point, see Craswell, supra note 81, at 601–05. See generally Paul Grice, Logic and Conversation, in Studies in the Way of Words 22 (1989); Grice, Further Notes on Logic and Conversation, in Studies in the Way of Words 50 (1989).

89. The above argument is very abbreviated. One would want to say more, for example, about just how Grice's cooperation principle and the rules of conversational implicature interact with the implicit representation of an intent to perform, which, as I observe above, supra note 85, is not a product of conversational implicature.

90. Such a qualification would be required, for example, by Grice's first maxim of quantity: “Make your contribution as informative as is required (for the current purposes of the exchange).” Grice, Logic and Conversation, supra note 88, at 26.

* This article has benefited from the input of Michael Bratman, Luca Ferrero, David Owens, and Gideon Yaffe as well as anonymous referees from Legal Theory. My thanks to all, and to Addison Draper and Gregory Zlotnick for their research assistance and Ian Ayres for our earlier collaboration.


  • Gregory Klass (a1)


Full text views

Total number of HTML views: 0
Total number of PDF views: 0 *
Loading metrics...

Abstract views

Total abstract views: 0 *
Loading metrics...

* Views captured on Cambridge Core between <date>. This data will be updated every 24 hours.

Usage data cannot currently be displayed