1. Perspective And Summary
15A. Phasing Propositions and Their Evidence on International Conflict
Democratic Peace page
...the policy-maker must be concerned with the best that can be achieved, not just the best that can be imagined. He has to act in a fog of incomplete knowledge without the information that will be available later to the analyst. He knows--or should know--that he is responsible for the consequences of disaster as well as for the benefits of success.
---- Kissinger, 1973:527
Conflict may remain latent between international actors. Leaders may change, domestic interests may shift, new ideologies may become ascendent, weapons purchases or developments may alter relative capabilities. But patterns of interaction may persist. Emerging interests may become absorbed into or submerged by overriding goals.
For example, the need for Western technology and aid has persuaded Soviet leaders to ignore some short-run interests that may create new conflicts with the West. The uncertainty of American commitments and support following the fall of South Vietnam in 1975 caused President Marcos of the Philippines to ignore latent conflicts with the Soviet Union and China, and seek new patterns of cooperative relations with them. Directors of multinational corporations like IBM have often submerged their many interests in conflict with communist leaders in order to make large, profitable deals with them. And so on.
Still, situations of conflict do involve active, empowered forces towards conflict. In spite of overriding interests or because the situation is independent of such interests, pressures in the situation may burst forth in actual conflict behavior. Two conditions bring this about: some event (occasion) and the will's consequent decision to manifest the opposing interests.
The second phase is that of uncertainty. It is the initiation of conflict phase, with all the objective uncertainty about an object's real goals, capabilities, responses, and stakes in the conflict. Although an actor may be subjectively certain of success, may calculate or intuit that the costs of the conflict will be acceptable, such expectation is always conditional. The precise moves of the other are unsure; the subsequent course of actual events speculative.
My intent here is to make more precise the theoretical elements and their relationships in this situation of uncertainty.
Now, this phase is initiated via a decision of the will triggered by some event. This trigger transforms the situation of conflict into a situation of uncertainty involving the will and preparations to assert one's interests. How can this trigger be made more precise?
First, consider that a trigger is a random event. It may be some final straw, some objectively minor or major happening that finally tilts an actor towards action, provokes such action, or makes it unavoidable (like the capturing of one's ship, as in the Pueblo or Mayaquez incidents involving the United States). It may be unintended communication from the other by word or deed that shows a lack of will or ability to defend some opposing interest. Any event is potentially a trigger, depending on the situation of conflict, on the subjective fields of those involved, on the timing. And these events may be unique, only important to or perceived by the actor. Therefore, we must deal with a space of potential events; we must define a trigger probabilistically and within the actor's perceived situation.
Let Tg denote the occurrence of a trigger stimulating the will to action in a situation of conflict g. Let Tgf be a specific trigger f which would provoke i's will within this situation. Then,
- Tg = 0, if Tgf = 0, for all f;
- Tg = 1, if Tgf = 1, for at least one f,
- 0 = nonoccurrence;
- 1 = occurrence;
- g = situation of conflict;
- f = a specific, potential trigger.
The question now concerns the probability of Tg = 1. To assess this consider that situations of conflict can be more or less empowered. They can be tense, hostile situations existing barely beneath the surface, overlaid by a patina of routine, cooperative interaction, an explosion seeking a spark. Little in the way of a trigger--an excuse, really--is needed to surface this conflict. Then there are situations that involve weak, secondary and tertiary interests: a situation one is willing to live with unless these interests are provoked by some rare, and really disruptive event.
Therefore, the likelihood of a trigger occurring is a function of the situation's intensity. The stronger the forces towards action, the more explosive the associated situation, the more powerful the situationally perceived interests and capabilities (d),
- pgt(T = 1) = g + g(tegt)(gid,i-j),
- Pgt(T = 1) = the probability at time t that T = 1, in situation g;
- , = situational parameters;
- egt = the tth time interval e (e.g., days, months, years) since the situation g of conflict came into being.
The notation egt defines the passage of time. According to the equation, the probability of T being unity is a linear function of the product of the passage of time since the inception of the situation and the intensity of the disposition (d) towards action. No matter how much time passes, if there is no situation of conflict, there can be no trigger; if the situation of conflict is intense, only a small passage of time should suffice, theoretically, for a trigger to occur.
The probability pgt(T = 1) is therefore a transition probability--the likelihood of a latent situation of conflict being transformed into conflict.
Equation 8.3 can now be elaborated to include the trigger:
- h,ij = ghig(Tggid,i-j + Qg,ij).
Thus, if no trigger occurs for situation g, then Tg = 0 and,
- h,ij = higQg,ij,
and from Equation 8.2
- wg,ij = Qg,ij.
That is, without a trigger the disposition to behave in a situation g is a result of interests and capabilities independent of the situation of conflict, and common behavior in this situation is then due to this independent disposition weighted by situational expectations
The decision to confront or to initiate conflict may involve an immediate threat or opportunity, or long-range one. For the long range, armaments may be increased and new programs under-taken; alliances, and defensive pacts may be negotiated.
The situation of uncertainty is detailed in Table 11.1. It is the transition stage between latent and manifest conflict, between accepting existing relationships and the balancing of powers, between Phase I and III of the conflict helix shown in Figure 29.1 of Vol. 2: The Conflict Helix. Before turning to the latter phase, however, the manifest reality of preparations in the situation of uncertainty should be assessed. If, indeed, such is a distinct element, then it should be evident as a separate component among conflict behavior.
In doing this two concepts must be kept clear. Compulsory components refer to those defining conflict behavior in the overall behavior space-time, that is, in the context of diverse nonconflict behavior. "Compulsory" is a theoretical term. A conflict component, however, as I will use it here, is an empirical component of conflict behavior. Only the conflict behavior region of space-time is analyzed to delineate these components. Conflict components describe the covariation between conflict behaviors; compulsory components underlie the covariation between all international behaviors.
Now, relevant analyses have been done for all the reported conflict behavior of state actors and dyads for 1955 and 1963.
Table 11.2 summarizes the results of these and other similar analyses.
A situation of conflict develops; a trigger provokes the will into action; preparations for conflict are made. This is summarized in Table 11.1.
* Scanned from Chapter 11 in R.J. Rummel, War, Power, Peace, 1979. For full reference to the book and the list of its contents in hypertext, click book. Typographical errors have been corrected, clarifications added, and style updated.
1. I am omitting the subscripts on the terms if context allows. It should be understood, therefore, that d means the summed product of an actor's perceived situation weighting his distance vectors from others, as in Equation 8.2.
2. I have tested Equation 11.3 in several unpublished projects (Appendix I: Projects 4.2-4.8). The results overall were mixed, and especially ambiguous because the underlying model was usually inappropriate (Model I rather than Model II). Nonetheless, when the concept of trigger is properly formulated in propositional form and related to the variety of analyses available, the ideas developed here have considerable empirical support. See Appendix 16B, Propositions 16.1, 16.10, 16.28, and 16.29.
3. Of course, this logic applied to the scalar equations also applies to vector equations, such as 8.4, as well.
4. Analyses of conflict behavior for dyads generally find such warning and defensive acts intercorrelated highly and consistently defined as an independent component of conflict. See Chapter 15, and particularly the Uncertainty Proposition in Appendix 15A. Among 45 relevant empirical analyses, 84% favor the proposition. Table 11.2 incorporates these findings.
5. Alliances and arms increases have also been found as empirically distinct patterns of conflict behavior, as should be the case if these assertions are correct. That alliances are a separate pattern can be seen from Table 4.1. The sources of evidence for this are given in Table 15A.2 for Proposition 15.1.
6. The compulsory components shown in Table 4.1 were delineated for dyadic conflict events and cooperative dyadic flows and structures. On these data, see Appendix II. The concern here, however, is in conflict events alone, and in the conflict components which may have been lost in the macroview of overall behavior space-time when several dozen conflict and nonconflict behaviors are analyzed together. To be clear on my use of the concepts "conflict behavior" and "compulsory," see Section 4.6.
7. A focus on events excludes alliance structures and arms increases, except in terms of their initiation. Therefore, alliances and armaments are not included in Table 11.2.
8. The sources of the 1955 data were The New York Times Index, New International Yearbook, Keesing's Contemporary Archives, Facts on File, and Britannica Book of the Year. The data collection indicated that these sources generally were redundant against The New York Times, and only this source was used for the 1963 data.
9. Oliva and I (1969) analyzed 24 conflict variables for 107 nations on 1963 data; I analyzed (1967a) 26 conflict variables for 82 nations on 1955 data.
10. Hall and I (1970) analyzed 24 conflict variables for 275 dyads for 1963 data.
11. Phillips (1969) analyzed (super-p analysis) 23 conflict variables across the 12 months of 1963 for 267 dyads.
12. Orthogonal and oblique rotations were done; component and image analyses also were applied.
13. See Hall and Rummel (1969, 1970), Phillips (1969), Oliva and Rummel (1969), Keim and Rummel (1969).
14. See those listed in Table 15A.3.
15. For an explanation of these different components (why do they occur? what do they mean regarding conflict?), see Chapter 15.