Robert R. Dykstra

How the West Got Wild: American Media and Frontier Violence

Five participants of the 1999 Western History Association in Portland, Oregon, reprise their panel session, in which they each assess past and present trends in the violence, perceived and real, of the American West.

Guns, Murder, and Plausibility

Midway through Randolph Roth’s rebuttal (2010, 189), I see that I’m pictured as hostile to methodological progress, that I’ve drawn a “line in the sand against statistical approaches to historical analysis.” Such rhetorical excess is on display in both American Homicide and Roth’s present article: the frothiness of prose descriptions and tendencies to melodrama, to overenthusiastic generalization, and to embellishment of metaphor. As in his depiction of Old England’s ills in the book, this expository style naturally raises doubts about facts and interpretations. The final paragraph of Roth’s rebuttal is equally disappointing. Quoting the late Eric Monkkonen, he avows that “Social science will prevail.” I’m sure it will. I hope it will. But against whom will it prevail? Moi? I’m a product of the so-called “Iowa School” of quantitative history, led by pioneers W. O. Aydelotte, Allen Bogue, and Samuel P. Hays. In 1965, I became the first historian at Iowa to obtain and make use of computer-aided analytical output. That led in 1968 to a Public Opinion Quarterly essay by political scientist Harlan Hahn and myself. A few years later, geographer David Reynolds and I produced and published an elaborate longitudinal voting study innovatively employing factor analysis. In the mid-1970s, I was modestly helpful in forming the Social Science History Association. And my research continued to follow suit. My second book, Bright Radical Star: Black Freedom and White Supremacy on the Hawkeye Frontier, is grounded on, among other statistical techniques, an incredible number of multiple ecological regression estimates, a then-controversial method that J. Morgan Kousser and I simultaneously defended in the pages of Social Science History. All things considered, I think my interdisciplinary social-science bona fides are pretty good. If social science is at all endangered, as Roth implies, it is not by me. The enemy at the gates is instead the postmodern turn of mind that for a couple of decades has been much apparent in the humanities disciplines, especially liter-

Guns, Murder, and Plausibility: Can Historians Trust the LaFree Hypothesis?

The most astonishing item in Randolph Roth’s massive new book, American Homicide, is not encountered until pages 450–51. Here the reader comes across figures 9.4 and 9.5, the author’s usefully tabulated versions of criminologist Gary LaFree’s discovery about murder in the United States. Since 1958, the national homicide rate (killings per 100,000 of population) has risen and fallen in coordination with the percentages of Americans who (1) “trust the government to do what is right” and (2) believe that “quite a few people running the government are crooked.” In sum, murder declines when government is seen as honest, and it rises when government is deemed dishonest (LaFree 1998, 100–104). Explaining this association of variables requires a big leap of the imagination. LaFree hypothesizes that the homicidemistrust correlation is shorthand for a much larger phenomenon: Americans’ rising and falling faith not only in the legitimacy of the government and the political system, but also in the legitimacy of the prevailing economic order and the traditionally structured family. These three societal elements—economic system, political system, and family—are key “institutions,” which LaFree defines as “the patterned, mutually shared ways that people develop for living together” (71). When institutional trust is healthy, people are more or less at peace with one another. When the opposite is true, people (mostly men, of course) indulge a lethal impulse to prey on others. There is something very Freudian about this, but LaFree’s book contains no reference to the figure who taught us that “the tendency to aggression is an innate, independent, instinctual disposition in man” and that a civilized society requires this tendency’s diversion toward socially sanctioned outlets like competitive capitalist accumulation or cutthroat politics or law enforcement or military service (Freud 1958, 9). If LaFree snubs Freud, Roth in turn is not very generous in acknowledging his debt to LaFree. He grants the criminolo-