To the editors:
I am pleased that Jason Quinn and Laurel Stone found my central thesis—that high levels of power motivation are associated with conflict escalation—to be obvious. Such a conclusion would likely surprise the numerous ancient and modern theorists and researchers whose investigations of the many and varied putative “reasons, motives, and causes” of war I reviewed in a 42-page chapter.1
At the same time, I would like to offer a few corrections and clarifications. First, power motivation is not measured by counting power-related words, but rather by the frequency of power images or themes. Humans trained to count themes can easily attain 90% agreement with each other, but so far it has not been possible to program computers, which are good at counting words, to reach anything like this level of reliability. The reason is that recognizing things that are similar, but have no identical elements, draws upon high-level cognitive skills. To Quinn and Stone, some examples of these themes that were mentioned “do not immediately stand out as revealing a power motive.” The definition of power imagery is not based on intuitions or impressions, but rather was developed from a series of systematic experiments and cross-validations, in which the power motive was aroused and its effects on verbal expression were observed, defined, and counted.2
Quinn and Stone suggest that most of the eight matched pairs of conflicts analyzed in Chapter 4 “are from the same conflict,” so that the allegedly peaceful resolutions are only a prelude to later wars; thus they were not truly independent. Having cases that are closely matched is important for John Stuart Mill’s method of difference, “comparing instances in which the phenomenon does occur, with instances in other respects similar in which it does not”; Alexander George and Andrew Bennett called this technique structured, focused comparison.3 Nevertheless, at least two pairs involved simultaneous but quite separate conflicts: the Mexican–American War and the Oregon Boundary Treaty in 1845–46, and the Soviet oppositions to national communism in Hungary and Poland in 1956. In one other pair, the war crisis preceded the peacefully resolved one: the Bay of Pigs in 1961 and the Cuban Missile Crisis in 1962.4
The remaining matched pairs were related, but this does not negate their value to explain observed power motivation differences or to predict escalation of future crises. For example, the ill-fated Compromise of 1850 was indeed a failure, as events over the next decade demonstrated. At the time, however, it did counter the secessionist rhetoric of the Nashville Southern Convention and thereby delay war. A decade later, numerous efforts at compromise, especially those of Kentucky’s Senator John J. Crittenden, failed. These differences in outcome are consistent with the observed differences in power motive imagery measured in the Senate debates of 1850 and 1860–1861. And although the outbreak of World War I may have grown out of the same issues as did the 1909 Bosnian Crisis and the 1912 First Balkan War, it does not follow that these earlier peaceful resolutions were failures that merely postponed an inevitable war to 1914.5
Finally, Quinn and Stone suggest that ideological biases led to my dismissing “all modes of institutional constraint, separation of powers, constitutional limits, and decentralization as ineffective” to prevent war. In my book, I do review my doubts about the universal efficacy of these mechanisms for taming power, because they are vulnerable to being hijacked by power itself.6 I also take pains to suggest that for war, “there is enough explanatory blame to go around: for psychology, history, political science, sociology, gender studies, anthropology, economics, the humanities, biology, even theology.”7 Psychology cannot discover or subsume every cause of war. But this is not a reason to dismiss its systematic empirical contributions to answering Albert Einstein and Sigmund Freud’s 1932 question, “Why war?”8