¹ For example, Niccolò Machiavelli, The Prince and The Discourses; Henry Taylor, The Statesman.

² Lehrer, Tom, “New Math,” ASCAP 1965Google Scholar.

³ Explanation and prediction have the same structure but different starting points. When we explain something we start from an observed outcome and work backwards to explain it. (X has occurred; it occurred because of Y.) When we predict something, we start with a set of principles and work out the consequences they imply. (Because of X, Y will occur.) We are working inductively when we start with some observed facts and construct a generalization(s) compatible with them. We are working deductively when we logically derive consequences from some system of generalizations.

⁴ For example, see: Rapoport, Anatol and Chammah, A. M., Prisoner's Dilemma (Ann Arbor 1965)CrossRefGoogle ScholarPubMed, and Shure, Gerald H. and others, “The Effectiveness of Pacifist Strategies in Bargaining Games,” Journal of Conflict Resolution, IX (March 1965), 106–17.CrossRefGoogle Scholar

⁵ The relevant chapters are those by Frohlich and Oppenheimer, Morse, Choucri and North, Tanter, Allison and Halperin.

⁶ See above, p. 95.

⁷ See above, pp. 53–54.

⁸ The formulation which provides a source from much of the thinking in the Frohlich-Oppenheimer essay, Olson's, MancurThe Logic of Collective Action (New York 1968)Google Scholar, has been subjected to at least one carefully designed test. The findings support Olson's theory. See Burgess, Philip M. and Robinson, James A., “Alliances and the Theory of Collective Action,” in Rosenau, James N., ed., International Politics and Foreign Policy (rev. edn., New York 1969), 642–53.Google Scholar

⁹ See above, p. 55.

¹⁰ See above, p. 103.

¹¹ See above, p. 21.

¹² It is not clear if Tanter views these alternative formulations as theories or even as more historically limited closed logical structures. If they are neither, they are not open to test. If they are one or the other, the comments which follow apply.

¹³ The most accessible formulation is that by Allison, Graham T., “Conceptual Models and the Cuban Missile Crisis,” American Political Science Review, LXII (September 1969), 689–718CrossRefGoogle Scholar. One can of course try to extend the organizational process formulation to handle organizational processes within alliances. That seems inappropriate in this case for two reasons. First, the formulation remains to be tested in the initial area of concern, i.e., within a nation. Second, the extension would require modifying the event interaction formulation to exclude interactions between allies.

¹⁴ On this and other problems and possibilities see Boudon, Raymond, “A New Look at Correlation Analysis,” in Blalock, Hubert M. Jr. and Blalock, Ann B., eds., Methodology in Social Research (New York 1968), 199–235Google Scholar.

¹⁵ See above, pp. 22–24.

¹⁶ See above, p. 28.

¹⁷ When we apply a theory to situations outside of its scope conditions we are engaged in one of two enterprises. We may be trying to establish that the scope conditions are unnecessarily restrictive. In that case we can make no claims for the applicability of the theory until after our attempt to relax the scope conditions has proven successful. Alternatively, we may wish to analogize between situations to which the theory applies and those which it does not. That is, we choose to use the theory as a metatheory. Accordingly, it should be judged as such.

¹⁸ The problems posed by confirmation based on highly controlled tests, which are the only ones likely to meet rules of correspondence, are presented with unusual clarity and insight by Chapanis, A., “The Relevance of Laboratory Studies to Practical Situations,” Ergonomics, X, No. 5 (September 1967), 557–77CrossRefGoogle Scholar.

¹⁹ The subsets of interest to policy users are frequently isomorphic with their role in the bureaucratic division of labor. If the bureaucracy is organized along geographical lines, interest and relevance go to a geographically bounded part of the set analyzed. If the organization is along functional lines (as used by the State Department), the part of the analysis pertinent to the function receives attention.

²⁰ Projections rest of course on a variety of continuity, in a sense stability, assumptions. If we have no basis for accepting or rejecting such assumptions, projections, as Hume pointed out some time ago, are useful if worrying.

²¹ I have deliberately stressed information-handling methods as distinct from data and findings. For helpful examples of such methods and their applications, see: Gillespie, John V. and Nesvold, Betty A., eds., Macro-Quantitative Analysis (Beverly Hills 1971)Google Scholar; Mueller, John E., ed., Approaches to Measurement in International Relations (New York 1969)Google Scholar; and Singer, J. David, ed., Quantitative International Politics (New York 1968)Google Scholar. do not mean to imply that all the articles in these references are instances of quantitative history or indeed that any are pure cases.

²² For example, one can see potential policy value in the CASCON and CACIS systems discussed by Tanter without sharing any of his theoretical views. For discussion of the considerations involved in academic and official development and use of computer based information systems, see: Bobrow, Davis B., “Data Banks, Foreign Affairs, and Feasible Change,” prepared for delivery at the Conference on Data Banks for International Studies sponsored by the U.S. Department of State and the International Studies Association, Washington, May 14–16, 1971Google Scholar; Tanter, Raymond, “The Policy Relevance of Model Building in World Politics,” prepared for delivery at the American Political Science Association Annual Meetings, Chicago, September 7–11, 1971CrossRefGoogle Scholar.

²³ Simon, Herbert A., The Sciences of the Artificial (Cambridge, Mass. 1969)Google Scholar.

²⁴ Ibid., xi.

²⁵ I am indebted for many fruitful ideas to Minnesota colleagues Frederick Bailey, Robert Holt, and James Lyday.

²⁶ On the need for process models, see: Brunner, Ronald D. and Liepelt, Klaus, “Data Analysis, Process Analysis, and System Change,” Institute of Public Policy Studies, University of Michigan, Discussion Paper #20 (1970)Google Scholar; and Bobrow, Davis B., “International Indicators,” prepared for delivery at the American Political Science Association Annual Meetings, New York, September 2–6, 1969Google Scholar.

²⁷ The authors stress the need for further development. I do not mean to imply that they are unaware of the limitations I suggest.

²⁸ See above, pp. 103, 121.

²⁹ See above, p. 121.

³⁰ See above, pp. 77–79.

³¹ Problems of and strategies for influence are discussed in two recent papers: Charles F. Hermann, “The Knowledge Gap: the Exchange of Information Between the Academic and the Foreign Policy Communities,” and Bobrow, Davis B., “Analysis and Foreign Policy Choice: Some Lessons and Initiatives.” Both were prepared for delivery at the American Political Science Association Annual Meetings, Chicago, September 7–11, 1971Google Scholar.