) What accounts for the observed distribution of international law instruments? This question was raised by Saul Levmore in his comments, at the April 2001 onference, on James Morrows paper. Treaties may be unilateral, bilateral multilateral with reservations allowed or multilateral with reservations prohibited. And international law need not take treaty form, but can develop as customary law and command wide acceptance. Furthermore, it appears that the distribution of treaties may vary by issue area- for example, multilateral treaties providing for uniform treatment have been extensively developed in trade over the last half-century, but not in the area of taxation Julie roin argues in her paper. Could a general theory, based on rational- choice premises, to developed to explain the observed distribution? 2)Under what conditions is international law tightly connected or, on the contrary, decomposed? 3 James Morrow discusses the concept of"firewalls' between sets of constraints on military strategies to prevent violation of one agreement from leading to a complete collapse of constraints. Similarly Downs and Jones discuss the segmentation of reputation. But sometimes firewalls exist and reputations are segmented, and sometimes the"dominoes fall. What are the conditions under which we should expect one or the other result? 3)In a screening model, what level of screening is desirable under what conditions? As Andrew Guzman asked at the april conference, what are the tradeoffs between having high standards and low standards and what criteria should be used to decide on the optimal screening level? Presumably the higher the risk of a "false negative, the higher the standards should be -as increases in airline security after September 11, 2001, suggest ) When is international law better- interpreted as designed to provide information about obligations and behavior, as conjectured by informational heories; and when it is better interpreted as designed to obfuscate issues, thus creating pooling equilibria that are favorable to opportunists? Could a general theory be constructed to account for variations in the precision of international law, or in the degree that it could be used to separate sincere players from insincere opportunists? 5)How fundamental are the differences between types of international obligations enforced largely by reciprocity and those not so enforced? In the former category are trade and arms control agreements; in the latter category, human rights treaties and environmental treaties involving pure public goods For the latter category of agreements, withdrawal of benefits and targeted 32 Julie Roin, "Taxation without Coordination. "This volume, pp Herbert A. Simon, The Sciences of the Artificial. Third edition(Cambridge, MIT Press, 1996), pp. 197 204 Compare the line of analysis represented by Keohane, After Hegemony, and Abbott and Nidal, "Hard and Soft Law in International Governance, "cited above, fi. 2 and fn. 5, with that represented by Goldsmith and Posner in this volume11 1) What accounts for the observed distribution of international law instruments? This question was raised by Saul Levmore in his comments, at the April 2001 conference, on James Morrow’s paper. Treaties may be unilateral, bilateral, multilateral with reservations allowed, or multilateral with reservations prohibited. And international law need not take treaty form, but can develop as customary law and command wide acceptance. Furthermore, it appears that the distribution of treaties may vary by issue area – for example, multilateral treaties providing for uniform treatment have been extensively developed in trade over the last half-century, but not in the area of taxation, as Julie Roin argues in her paper.32 Could a general theory, based on rational￾choice premises, to developed to explain the observed distribution? 2) Under what conditions is international law tightly connected or, on the contrary, “decomposed?33 James Morrow discusses the concept of “firewalls” between sets of constraints on military strategies to prevent violation of one agreement from leading to a complete collapse of constraints. Similarly, Downs and Jones discuss the segmentation of reputation. But sometimes firewalls exist and reputations are segmented, and sometimes the “dominoes fall.” What are the conditions under which we should expect one or the other result? 3) In a screening model, what level of screening is desirable under what conditions? As Andrew Guzman asked at the April conference, what are the tradeoffs between having high standards and low standards, and what criteria should be used to decide on the optimal screening level? Presumably the higher the risk of a “false negative,” the higher the standards should be – as increases in airline security after September 11, 2001, suggest. 4) When is international law better-interpreted as designed to provide information about obligations and behavior, as conjectured by informational theories; and when it is better interpreted as designed to obfuscate issues, thus creating pooling equilibria that are favorable to opportunists?34 Could a general theory be constructed to account for variations in the precision of international law, or in the degree that it could be used to separate sincere players from insincere opportunists? 5) How fundamental are the differences between types of international obligations enforced largely by reciprocity and those not so enforced? In the former category are trade and arms control agreements; in the latter category, human rights treaties and environmental treaties involving pure public goods. For the latter category of agreements, withdrawal of benefits and targeted 32 Julie Roin, “Taxation without Coordination.” This volume, pp. --- . 33 Herbert A. Simon, The Sciences of the Artificial. Third edition (Cambridge, MIT Press, 1996), pp. 197- 204. 34 Compare the line of analysis represented by Keohane, After Hegemony, and Abbott and Snidal, “Hard and Soft Law in International Governance,” cited above, fn. 2 and fn. 5, with that represented by Goldsmith and Posner in this volume