1. See the debate between the “managerial” perspective that emphasizes centralization but not enforcement, Chayes and Chayes 1995, and the “compliance” perspective that emphasizes enforcement but sees it as decentralized, Downs, Rocke, and Barsoom 1996.

2. Hart 1961.

3. Hart 1961, 79.

4. Hart, of course, observed that in form, though not in substance, international law resembled a primitive legal system consisting only of primary rules. We sidestep that debate, noting only that the characteristics we observe in international legalization leave us comfortable in applying Hart's terms by analogy. We also observe that the international legal framework has evolved considerably in the decades since Hart wrote. Franck reviews these changes and argues that international law has developed a general rule of recognition tied to membership in the international community. Franck 1990, 183–207.

5. On the “obligation” dimension, jus cogens refers to an international legal rule—generally one of customary law, though perhaps one codified in treaty form—that creates an especially strong legal obligation, such that it cannot be overridden even by explicit agreement among states.

6. Gamble 1985.

7. The declaration has also contributed to the evolution of customary international law, which can be applied by national courts as well as international organs, and has been incorporated into a number of national constitutions.

8. Interestingly, however, while the formal mandate of the OSCE High Commissioner on National Minorities related solely to conflict prevention and did not entail authority to implement legal (or nonlegal) norms, in practice the High Commissioner has actively promoted respect for both hard and soft legal norms. Ratner 2000.

9. Kaplan 1957.

10. Bull 1977.

11. Compare Goldstein, Kahler, Keohane, and Slaughter, this issue.

12. Victor, Raustalia, and Skolnikoff 1998, especially chap. 4.

13. Setear 1999.

14. In linking obligation to the broader legal system, we are positing the existence of international law as itself imposing a body of accepted and thereby legitimized obligations on states. If the ultimate foundation of a legal system is its acceptance as such by its subjects, through a Kelsenian Grundnorm or an ultimate rule of recognition, then we are positing the existence of that acceptance by states with regard to the existing international legal system. The degree of obligation that we seek to measure refers instead to acceptance by subject states of a particular rule as a legal rule or not, that is, as binding or not binding as a matter of international law.

15. Under accepted legal principles, many of which are codified in the Vienna Convention on the Law of Treaties, the intent of the parties to an agreement determines whether that instrument creates obligations that are legally binding, not merely personal or political in effect, and that are governed by international law, rather than the law of some nation. Intent is sometimes explicitly stated; otherwise it must be discerned from the overall context of an agreement, its negotiating history, the nature of its commitments, and its form. As a practical matter, however, legalization is the default position: significant agreements between states are assumed to be legally binding and governed by international law unless the parties indicate otherwise. U.S. practice on this score is summarized in the State Department's Foreign Relations Manual, pt. 181.

16. Zaring 1998.

17. Although precise obligations are generally an attribute of hard legalization, these instruments use precise language to avoid legally binding character.

18. See Baxter 1980; and Schachter 1977.

19. In addition to the explicit escape clauses considered here, states are often able to escape from the strictures of particular provisions by filing reservations, declarations, and other unilateral conditions after an agreement has been negotiated.

20. These avenues of escape are quite precisely drafted and are supervised by the European Commission and Court of Human Rights, limiting the ability of states to evade their substantive obligations.

21. In contrast to the European Convention on Human Rights, this withdrawal clause is self-judging, increasing its softening effect. Nonetheless, the clause was originally inserted to impose some constraints on what might otherwise have been seen as an unconditional right to withdraw.

22. Some agreements authorize particular conduct rather than requiring or prohibiting it. Such provisions are usually couched as rights, using the word may. Gamble 1985.

23. See Chinkin 1989; and Gruchalla-Wesierski 1984.

24. This discussion also applies to instruments adopted by organizations with law-making competency but outside prescribed procedures. A significant example is the European Social Charter, adopted by all members of the EC Council except the United Kingdom. These states bypassed a unanimity requirement to avoid a U.K. veto, adopting a softer instrument to guide subsequent legislative action.

25. Palmer 1992.

26. A precise rule is not necessarily more constraining than a more general one. Its actual impact on behavior depends on many factors, including subjective interpretation by the subjects of the rule. Thus, a rule saying “drive slowly” might yield slower driving than a rule prescribing a speed limit of 55 miles per hour if the drivers in question would normally drive 50 miles per hour and understand “slowly” to mean 10 miles per hour slower than normal. (We are indebted to Fred Schauer for both the general point and the example.) In addition, precision can be used to define limits, exceptions, and loopholes that reduce the impact of a rule. Nevertheless, for most rules requiring or prohibiting particular conduct—and in the absence of precise delegation—generality is likely to provide an opportunity for deliberate self-interested interpretation, reducing the impact, or at least the potential for enforceable impact, on behavior.

27. Franck 1990.

28. Franck labels this collective property “coherence.” We use the singular notion of precision to capture both the precision of a rule in isolation and its precision within a rule system.

29. Morrow 1997 and 1998.

30. Simma and Paulus 1999.

31. Fuller 1964.

32. The standard regime definition encompasses three levels of precision: “principles,” “norms,” and “rules.” Krasner 1983. This formulation reflects the fact that societies typically translate broad normative values into increasingly concrete formulations that decision-makers can apply in specific situations.

33. Kennedy 1976.

34. Similarly, agreements administered by the WTO can, with similar legitimacy and effectiveness, specify detailed rules on the valuation of imports for customs purposes and rely on broad standards like “national treatment.”

35. Operationalizing the relative precision of different formulations is difficult, except in a gross sense. Gamble, for example, purports to apply a four-point scale of “concreteness” but does not characterize these points. Gamble 1985.

36. The State Department's Foreign Relations Manual states that undertakings couched in vague or very general terms with no criteria for performance frequently reflect an intent not to be legally bound.

37. Abbott and Snidal, this issue.

38. U.S. Releases Proposal on Elements of Crimes at the Rome Conference on the Establishment of an International Criminal Court, statement by James P. Rubin, U.S. State Department spokesperson, 22 June 1998, <secretary.state.gov/www/briefings/statements/1998/ps980622b.html>, accessed 16 February 1999.

39. Law remains relevant even here. The UN Charter makes peaceful resolution of disputes a legal obligation, and general international law requires good faith in the conduct of negotiations. In addition, resolution of disputes by agreement can contribute to the growth of customary international law.

40. Abbott and Snidal 1998.

41. Boisson de Chazournes 1998.

42. Shihata 1994.

43. Schauer and Wise 1997.

44. Sunstein 1986.