1. Simon, , The Sciences of the Artificial (1969) 3.Google Scholar

2. (16·17) Aleyn. 26, 82 E.R. 897 (Common Pleas).

3. (1863) 3 B. & S. 826. 122 E.R. 309 (Q.B.).

4. Id. at 312 per Blackburn J.

5. On a second reading, implied terms are terms that do not come to mind but, literally, “go without saying” in the sense that they do not come to one's mind at all—either as a way station to writing or speaking or as a stopping point. Instead, they form the obvious background against which people make explicit claims. On the second reading, there are endlessly many such implied terms, facts in the absence of which it would make no sense to make agreements at all. We return to this reading below.

6. Davis Contractors [1956] A.C. 696. (II.L.) Per Lord Radcliffe at 728.Google Scholar

7. Capital Quality Homes Ltd. v. Colywyn Construction Ltd. (1975) 9 O.R. (2d) 617 (C.A).Google Scholar

8. Id. at 623.

9. Id.

10. There are plainly reasons of legal process and institutional competence that stand in the way of having courts regularly step in to rewrite contracts. Our present point is simply that the power to make private arrangements that contract law confers should be taken seriously.

11. In his casebook on contract law, Lon Fuller identifies the same problem in the frustration cases, and says that the problem has not been satisfactorily resolved by psychology. See Fuller, and Eisenberg, , Basic Contract Law 4th ed. (1981), 848–52Google Scholar. We are offering a philosophical account that Fuller's other writings lead us to suppose he would welcome.

12. Metropolitan Toronto Police Association (unreported decision of arbitrator Swan, K., 01 12, 1987).Google Scholar

13. Bork, R., Neutral Principles and Some First Amendment Problems, 47 Indiana L. J. (1971).Google Scholar

14. Bork's view of the relationship between democracy and the Constitution commits him to the view that, wherever the framers had not considered a case, the Constitution sets no limits on legislative power. This approach is of a piece with die approach to frustration that charges courts with implementing public policy goals in circumstances not anticipated by the parties to a contract. Bork's view differs only in the particular alternative to interpretation that he suggests. Views about democracy lead to majority rule as a default in constitutional cases; views about public policy lead to efficiency as a default in private arrangements. Both views share the core idea that the meaning of a contract reaches only as far as the parties explicitly contemplated.

15. Wittgenstein, , Philosophical InvestigationsGoogle Scholar (trans. Anscombe, G.E.M., 1972), 33.Google Scholar

16. Arithmetic provides a philosophically useful example for Wittgenstein's purposes, because it seems like the easiest case for those who think that knowing a rule is a matter of having something in mind. Even the rules that seem to he clearest and most explicit depend on what Wittgenstein calls “forms of life” if they are to guide behavior. But its very usefulness for that purpose makes its extension to more difficult examples, such as contractual interpretation, less than straightforward. Arithmetic is a useful case for Wittgenstein because all agree that there is a right answer to every question; the philosophically difficult questions all revolve around what makes those answers right. When the teacher has successfully taught the student addition, her intention has been communicated for all possible cases. And whatever exactly is involved in grasping a mathematical rule, once it is grasped, it applies to all possible pairs of numbers, and gives a definite answer for each. Those who know how to add know how to go on in the same way in unanticipated cases. And if one knows only how to add some of the time, one doesn't know how to add at all.

Arithmetic and contracting differ in many important ways, including the ways in which disagreements are resolved. For our purposes, the crucial similarity is that the correctness of an answer in each doesn't depend on what is going through anyone's head.

17. Wittgenstein, , supra note 15, at 95.Google Scholar

18. See, e.g., Marmor, A., Interpretation and Legal Theory (1992).Google Scholar

19. Since Descartes, philosophers of mind have reflected on dreaming to understand how thought can be meaningful, for dreams appear to be meaningful but false. Davidson provides an easy way beyond this picture, for he points out that, in order for my dream to be about the things I think it is about—say, sitting in front of the fire with my computer—most of the things I believe in it had better turn out to be true. For example, as I draw nearer to the fire, I get warmer; fires burn wood; normally, the things I do to the keyboard systematically affect what happens on the screen, and so on. Without these banal truths, and many others like them, my belief that it is a fire I am in front of turns out not to be about fires at all. Thus I can only be mistaken about particulars if I get a great many facts, both general and specific, more or less right. Too many mistakes and I end up changing the subject. If I dream that I'm sitting by the fire, but that fire doesn't consume wood, or that it is cold, and so on, at some point the thing in front of which I take myself to be sitting is not a fire.

20. Davidson, , A Nice Arrangement of Epitaphs, in Truth and InterpretationGoogle Scholar (LePore, E., ed. 1986), p. 433.Google Scholar

21. Carlile v. Carbolic Smoke Ball Co. [1893] 1 Q.B. 256 (C.A.).Google Scholar

22. Esso Petroleum v. Mardon [1976] Q.B. 801 (C.A.).Google Scholar

23. Cundy v. Lindsay (1978) 3. App. Cas. 459 (H.L.).Google Scholar

24. (1871) L.R. 6 Q.B. 597.

25. We are not told how the defendant could identify the age of the oats on deliver), but not in the sample.

26. Of course, you might have simply made an honest mistake that I could not detect. You thought you were at the sheep-shearer's new oats auction. In those circumstances, we have to ask who should bear the risk of mistake. The answer will depend on our understanding of the sources of your error, and the information available to you, as well as on whether I had cues that might have told me what you thought.

27. (1899) 29 S.C.R. 450.

28. I.e., 40 acres.

29. Id. per King J. at 468.

30. Id. per King J. at 467.

31. (1864) 2 H. & C. 906, 159 E.R. 375 (Exch.).

32. Davidson, , Communication and Convention, in Inquiries into Truth and Interpretation (1985), 265.Google Scholar

33. See Treitel, , Frustration and Force Majeure (1994), 13–22Google Scholar, for an extended discussion.

34. The same point can be made in terms of estoppel. Parties are estopped from claiming that they meant something different because what they meant is given by what they said and did, not by what they may or may not have privately thought. By saying something publicly, they are making a representation, on which others are entitled to rely. Thus, they cannot go back on it The only question for contractual interpretation is, “Which representation was made”? And that is a public, if richly complex, question.

35. Some lawyers may balk at this suggestion. They are too familiar with examples of “strict” interpretation in an old-fashioned sense. They are likely to think of this form of reasoning as opposed to “purposive” interpretation and they will remind us that it is a common common-law view that, for example, criminal statutes are to be construed strictly. This is, in one sense, all fair enough. There clearly are examples of bad statutory interpretation. This is what strict interpretation usually amounts to. The interpreter stops short in his or her effort to connect systematically the drafter's words to the world, forgetting that statutes are passed never for the sake of their details, but in order to do something. What of the idea that criminal statutes are to be construed strictly? This idea makes perfect sense (and may be a good idea), but it in no way challenges (he idea that all interpretation is, in our important sense, “strict.” It might well be sensible to adopt an interpretive strategy in criminal cases such that if an accused can offer a plausible (although perhaps not the most sensible) interpretation of the law that leads to acquittal, then that is, for various good reasons, a reason to acquit. But all of that depends on the same mode of interpretative exercise we have outlined.