1. The book is as much a “primer” for the prisoners and their counterparts in the real world as it is an analysis of their problems. Axelrod himself described an earlier article in which he reported on the first round of his computer tournament as “a ‘primer’ on how to play the Prisoner's Dilemma game effectively.” See his “Effective Choice in the Prisoner's Dilemma,” Journal of Conflict Resolution 24 (March 1980), p. 3Google Scholar. For other related articles by Axelrod, , see “More Effective Choice in the Prisoner's Dilemma,” Journal of Conflict Resolution 24 (09 1980), pp. 379–403CrossRefGoogle Scholar; (with Hamilton, William D.), “The Evolution of Cooperation,” Science 211 (27 03 1981), pp. 1390–96Google Scholar; and “The Emergence of Cooperation among Egoists,” American Political Science Review 75 (June 1981), pp. 306–18.Google Scholar

2. Hardin, Russell, Collective Action (Baltimore: Johns Hopkins University Press, 1982), p. 125.Google Scholar

3. Axelrod does observe that “it is possible that the results from pairwise interactions will help suggest how to undertake a deeper analysis of the n-person case as well but that must wait” (fn. 3, p. 216). See also his discussion of the differences between two- and n-person PD games in fn. 3, p. 221.

4. See, for example, the conclusion of Snyder, Glenn H. and Diesing, Paul, Conflict among Nations: Bargaining, Decision Making, and System Structures in International Crises (Princeton: Princeton University Press, 1977)Google Scholar, that two-by-two games need to “be considerably reinterpreted to provide adequate explanations of real decision processes” (p. 83).

5. Russett, Bruce, The Prisoners of Insecurity: Nuclear Deterrence, the Arms Race, and Arms Control (San Francisco: Freeman, 1983), p. 115.Google Scholar

6. See, for example, Lipson, Charles, “International Cooperation in Economic and Security Affairs,” World Politics 37 (10 1984), pp. 1–23.CrossRefGoogle Scholar

7. Jervis, Robert, “Security Regimes,” International Organization 36 (Spring 1982), p. 359CrossRefGoogle Scholar. Conybeare, John A. C., “Public Goods, Prisoner's Dilemmas and the International Political Economy,” International Studies Quarterly 28 (03 1984)CrossRefGoogle Scholar, adds: “Since World War II, rivalries have not usually led to trade wars. Indeed, a casual reading of the financial press suggests that the prisoner's dilemma outcome of mutually debilitating retaliation is the exception rather than the rule” (p. 13).

8. Stein, Arthur A., “Coordination and Collaboration: Regimes in an Anarchic World,” International Organization 36 (Spring 1982), p. 320CrossRefGoogle Scholar. (Stein attributes this to Jervis in fn. 44, p. 320.)

9. See Jervis's, careful discussion of these distinctions in his “Cooperation under the Security Dilemma,” World Politics 30 (01 1978), pp. 167–214.CrossRefGoogle Scholar

10. See, for example, Modelski, George, “Long Cycles and the Strategy of U.S. International Economic Policy,” in Avery, William P. and Rapkin, David P., eds., America in a Changing World Political Economy (New York: Longman, 1982), pp. 97–116Google Scholar; Christopher K. Chase-Dunn, “International Economic Policy in a Declining Core State,” in Ibid., pp. 77–96; Gilpin, Robert, U.S. Power and the Multinational Corporation: The Political Economy of Foreign Direct Investment (New York: Basic, 1975).CrossRefGoogle Scholar

11. Hegemons do not always eliminate security dilemmas, of course: witness Greek and Turkish rivalries over Cyprus.

12. It may also be that Japan did not retaliate because it profited from the restrictions. See Yoffie, David B., Power and Protectionism: Strategies of the Newly Industrializing Countries (New York: Columbia University Press, 1983).Google Scholar

13. The power of this criticism in relation to Axelrod's work obviously depends on the extent to which international politics is characterized by the presence of dominant states. Available evidence suggests that the structure of the international system varies across time and space. See, for example, Modelski, George, “The Long Cycle of Global Politics and the Nation-State,” Comparative Studies in Society and History 20 (04 1978), pp. 214–35.CrossRefGoogle Scholar

In the contemporary period, it seems clear that hegemony remains important to understanding relations within the security blocs of East and West. It also seems clear that U.S. postwar hegemony and current power remain important, though obviously not exclusively so, to understanding the political economy of relations among the major Western industrialized states. Students of international regimes seem to agree, for example, that power remains essential to understanding the creation, though not necessarily the persistence, of international regimes. See Krasner, Stephen D., “Regimes and the Limits of Realism,” International Organization 36 (Spring 1982), pp. 497–510CrossRefGoogle Scholar. Community can also substitute for a central authority in the international system. See, for example, Bull, Hedley, The Anarchical Society (New York: Columbia University Press, 1977)CrossRefGoogle Scholar. Communities that can maintain order effectively must meet stringent requirements that normally do not prevail in most arenas of international politics. For an excellent treatment of community and order, see Taylor, Michael, Community, Anarchy and Liberty (Cambridge: Cambridge University Press, 1982).CrossRefGoogle Scholar

14. Taylor, Michael, Anarchy and Cooperation (London: Wiley, 1976), pp. 93, 94.Google Scholar

15. For an extended discussion of positive and negative altruism, see Ibid., chap. 4. See also Keohane, Robert O., After Hegemony: Cooperation and Discord in the World Political Economy (Princeton: Princeton University Press, 1984), pp. 27, 66, 74Google Scholar; and Margolis, Howard, Selfishness, Altruism, and Rationality: A Theory of Social Choice (Cambridge: Cambridge University Press, 1982), p. 15.Google Scholar

16. Axelrod, states, “The payoffs already include whatever consideration each player has for the interests of the other” (p. 12).Google Scholar

17. Heath, Anthony, Rational Choice and Social Exchange: A Critique of Exchange Theory (Cambridge: Cambridge University Press, 1976), p. 45.Google Scholar

18. Waltz, Kenneth N., Theory of International Politics (Reading, Mass.: Addison-Wesley, 1979), p. 105 (emphasis added).Google Scholar

19. Pure positive altruists–who strive “only to maximize the other player's payoff” (Taylor, , Anarchy and Cooperation, p. 72Google Scholar)–are also irrelevant both to international relations and to Axelrod's analysis, although for different reasons. It is inconsistent with self-preservation for nations to behave as pure positive altruists; within the context of Axelrod's analysis, it is irrelevant because no conflict of interest or PD game exists under these conditions. See Axelrod, p. 6.

20. This is drawn from Taylor's analysis of Games of Difference in his Cooperation and Anarchy, pp. 73–74. To put this another way, the conflict of interest between states may be more severe than Axelrod's matrix represents. See the discussion of the implications for cooperation of increasing the conflict of interest between the players in a PD in Axelrod's, Conflict of Interest: A Theory of Divergent Goals with Applications to Politics (Chicago: Markham, 1970), chap. 3.Google Scholar

21. When the goal is to accumulate more points than the other player, as Behr, Roy L., “Nice Guys Finish Last–Sometimes,” Journal of Conflict Resolution 25 (06 1981), pp. 289, 298CrossRefGoogle Scholar, notes, “one of course must defect more than one's opponent does (though not necessarily before one's opponent does). Thus TIT FOR TAT, which never defects more than the other player, can never win.” “[W]hen either consistent victory or maximum average margin of victory is the predetermined policy objective [in the computer tournament],” Behr adds, “strategies which usually cooperate until provoked, but occasionally take advantage of cooperative opponents are clearly most successful” (emphasis in original). Axelrod footnotes Behr's article (fn. 1, chap. 6, p. 219) with the observation that the objective of maximizing relative gains transforms the PD into a “zero-sum game with all D being the one and only dominant strategy for any value of w.” Not all negative altruists seek only to maximize relative gains, however; consequently, not all games played by negative altruists transform PD into zero-sum games.

22. To put this in Axelrod's terms, the shadow of the future may not loom sufficiently large in relation to the payoffs to render immediate defection irrational. See pp. 59, 127–28.

23. It is also true, says Axelrod, that “both sides would prefer mutual restraint to the random alternation of serious hostilities, making R > (T + S)/2” (p. 75).

24. Ashworth, Tony, Trench Warfare, 1914–1918: The Live and Let Live System (New York: Holmes and Meier, 1980)CrossRefGoogle Scholar. Axelrod states that his analysis “relies upon Ashworth's fine work for its illustrative quotes and for its historical interpretation” (p. 75). Ashworth argues: “obviously, to restrict aggression was to increase chances of survival, since aggression and survival were inversely related: as the volume of bombs, bullets and shells passing reciprocally between the trenches increased, so the chance of avoiding death and injury decreased. The most general motive of exchange was the interest in staying alive common to all combatants …” (p. 130).

25. See Schelling, Thomas C., The Strategy of Conflict (London: Oxford University Press, 1960)Google Scholar, for an excellent discusssion of collaboration games and game theory in general.

26. Ashworth, , Trench Warfare, pp. 20–21Google Scholar. Ashworth later adds: “one reason … why men do or do not fight, concerns the informal social relations which emerge among soldiers on the battlefield, and the present research has explored and confirmed this theme in respect of the First World War. Nevertheless, there is a difference here: whereas other works relate combat performance to informal groups of compatriots, this research relates performance not only to compatriots, but also to informal networks of antagonists. Thus unofficial negotiation among adversaries determined what happened on the battlefields of the First World War” (p. 206).

27. Arms races may sometimes resemble the tacit truce case in their relative invulnerability to resolution via strategies derived from assumptions about the decisive importance of systemic variables. If domestic preferences for arms increases are strong enough, as Downs, George W., Rocke, David M., and Siverson, Randolph M., “Arms Races and Cooperation,” World Politics (forthcoming), p. 5Google Scholar in mimeo, observe, the game being played in some races may not be Prisoner's Dilemma but Deadlock (where mutual defection is preferred to mutual cooperation).

28. See Brams, Steven J., Superpower Games: Applying Game Theory to Superpower Conflict (New Haven: Yale University Press, 1985), p. 88 (particularly fn. 7)CrossRefGoogle Scholar; Russett, Bruce, “International Interactions and Processes: The Internal vs. External Debate Revisited,” in Finifter, Ada W., ed., Political Science: The State of the Discipline (Washington, D.C.: American Political Science Association, 1983), quotation at p. 553.Google Scholar

29. Nicholson, Michael, Oligopoly and Conflict: A Dynamic Approach (Toronto: University of Toronto Press, 1972), chap. 6Google Scholar; Taylor, , Anarchy and Cooperation, p. 96.Google Scholar

30. Schoultz, Lars, Human Rights and United States Policy toward Latin America (Princeton: Princeton University Press, 1981)CrossRefGoogle Scholar, presents a comprehensive review and explanation of the various restrictions on the extension of economic and military aid enacted by Congress in the 1970s.

31. Taylor, , Anarchy and Cooperation, p. 96.Google Scholar

32. Nicholson, , Oligopoly and Conflict, p. 150.Google Scholar

33. Hacking, Ian, “Winner Take Less,” New York Review of Books, 28 06 1984, p. 18.Google Scholar

34. Initial cooperation is a fairly regular pattern in PD supergames, according to Hacking, Ibid.; “typically there is a fairly high chance of cooperation at the start, which declines as one tries to exploit another, and mutual distrust develops. But after a while people realize this is stupid, and on average they slowly begin to cooperate again.” A similar account is in Russett, , The Prisoners of Insecurity, p. 107Google Scholar. Axelrod also observes of his student subjects: “Sooner or later, one student defects to get ahead, or at least to see what will happen. Then the other usually defects so as not to get behind. Then the situation is likely to deteriorate with mutual recriminations” (p. 111).

35. Nozick, Robert, Anarchy, State, and Utopia (New York: Basic, 1974), p. 11.Google Scholar

36. Jervis, , “Security Regimes,” p. 360.Google Scholar

37. Downs, George W., Rocke, David M., and Siverson, Randolph M., “Arms Races and Cooperation” (Paper prepared for the 1984 Meeting of the American Political Science Association,Washington, D.C.,30 August-2 September, pp. 13–14Google Scholar, cited by permission). Jervis notes that sometimes misperception can lead to more cooperation than otherwise. He observes, however, that the positive images that lead to this result arise on “relatively few occasions” (pers. comm., 6 September 1984).