Terrorism in the Heartland: Vigilantism, Lynching, and the Klan in Indiana, 1858-1930

Download 117.99 Kb.
Size117.99 Kb.
  1   2   3
Terrorism in the Heartland:

Vigilantism, Lynching, and the Klan in Indiana, 1858-1930

Richard Hogan, Associate Professor

Purdue University, Sociology Department, College of Liberal Arts

West Lafayette, IN 47907-2059 hoganr@purdue.edu

A version of this paper was presented at the Indiana Academy of Social Science meeting,

Evansville, IN, 9 October 2015. Comments and suggestions from E. M. Beck and Sarah Mustillo were particularly helpful.

keywords: terrorism, repertoires, collective violence, Indiana history, KKK

9,7477 words (41 double spaced pages, including abstract, bib and appendix, 8 figures and 4 tables—a second appendix has been provided for the editor and reviewers.

Terrorism in the Heartland:

Vigilantism, Lynching and the Klan in Indiana, 1858-1930

The general public has learned nothing from the social movements and political protest researchers who cut their teeth on the failure of mass society and collective behavior theories to explain the late (nineteen) Sixties anti-war movement. Now a band of neoliberal (conservative) scholars are returning to the cultural and psychological roots of these old theories of political protest. They tell us that resource mobilization and political process theories are outdated and cannot explain the new social movements—particularly, Occupy. In this regard, they are more distracting than dangerous, but the popular image of the terrorist is particularly susceptible to the ahistorical, cultural and psychological reductionism of these theories of the Sixties. In the interest of gaining some historical and macro-institutional perspective on the War on Terrorism, this paper looks at terrorism in Indiana between 1830 and 1930, focusing on the National Horse Thief Detective Association and the Ku Klux Klan, during two waves of terrorism: Civil War: 1858-1876 and Post-Reconstruction: 1877-1930.

We can blame it on 9-11, I suppose, but the confusion surrounding terrorism is reversing progress made since 1970 in the study of social movements and social change. Scholars and the mass media have returned to old theories of the Cold War era—mass society and collective behavior, to explain the “radical Islamic terrorist campaign,” which is corrupting women and children in Western Europe and even here in America’s Heartland (Callimachi 2015). How can we hope to turn the tide of fear and ignorance that seems to be perpetrating the twin threats of conservative reaction and anti-intellectual hysteria?

Let’s start by dealing with the confusion on concepts, starting with the concept of “radical” as opposed to “reactionary” and the difference between “terrorism,” “genocide,” and “rebellion.” Then we can appreciate the difference between “terrorism,” which is a strategy, and violence, which is a tactic. Lest we become thoroughly confused in meta-theoretical discourse, let us ground this exercise in the material world, following a structural but historical path, focusing primarily on the U.S.A., 1775-1930, and, particularly, on Indiana, 1858-1930. Thus we shall come to appreciate that terrorism is at least as old as the Declaration of Independence and as American as apple pie. We shall also come to see that terrorism is not necessarily violent or primitive, and that violence is not used exclusively by terrorists or primitives.

Eventually, we will consider the difference between the old repertoire of terrorism in the USA—particularly vigilantism and lynching, which we will examine in Indiana, 1858-1930, and the new (social movement) repertoire of terrorism, which includes Klan marches, public displays, and private meetings and ceremonies, characteristic of the second wave of the KKK, specifically in Indiana in the early 1920s. We shall consider the extent to which the same petit-bourgeois morality (interest in public order—see Author 1990) and even the same Horse Thief Detective association and the same Democratic or Republican parties1 (organizations) provided the base for both rounds of collective action. As I will suggest, it was not interests or organizations but repertoire that explains the widely divergent political geography of lynching victimization and KKK events (both estimated from newspaper reports, coded by colleagues who provided the populations). Terrorists might use violence or not. They might use old school or thoroughly modern tactics, but they are still terrorists if they are attempting to strike fear in the heart of a civilian population.2 We will consider how this perspective might inform our view of ISIS and foreign terrorists after we gain a better appreciation of terrorism in American as history and consider the possibility that terrorists may be returning to old school tactics, as was the case during the Civil Rights Movement (Morris 1984), with GOP candidates inspiring the reactionary wings of the party to defend the borders against the barbarians.

Some Definitions and Concepts

Let us begin with a lesson from political science. As indicated in Figure 1, liberals and conservative share general optimism about the institutional order (e.g., republican capitalism), but they differ in their opinions on human nature. A similar pattern distinguishes radicals and reactionaries, except that they are pessimistic about the institutional order. Reactionaries tend to see people as selfish, lazy, and greedy, and they believe that established institutions exacerbate this problem—republican capitalism is a license to steal, and the welfare state encourages sloth. Radicals see people as cooperative and productive but see established institutions as the source of anti-social, inhuman struggles for wealth and power.

Figure 1

Political Attitudes Defined by Opinions on

Existing Institutions and Common People

Opinions About Existing Institutions

Opinions About

Common People









source: adapted from Tom Ruth, California Government class lecture, spring 1970, Grossmont Community College, El Cajon California

When the institutional order is threated by radical revolutionaries or paramilitary bands of reactionaries, liberals and conservatives are inclined to forget their differences and form “progressive” coalitions that might offer some changes (reforms) within the institutional order (e.g. women’s suffrage, collective bargaining, or civil rights) in order to avoid regime change and the attendant dangers of communism, or anarchy, or civil war (Piven and Cloward 1977). When successful, these progressive coalitions include enlightened conservatives and liberals. In the U.S.A. these coalitions tend to tolerate reactionaries and to repress radicals.

Consider, for example, the events leading up to the American Revolution. The Sons of Liberty and the rest of the “Radical Whigs” (Bailyn 1967) were challenging the authority of the Crown, destroying stamps, harassing tax collectors and royal governors, and even throwing tea into Boston Harbor in 1773 (Maier 1974; Morgan and Morgan 1974). At the same time, Judge Lynch was terrorizing Loyalists in Virginia.

“Charles Lynch (1736-1796), a justice of the peace in Bedford County, Virginia, and his followers whipped and occasionally hanged suspected local Loyalists. Apparently, Lynch’s vigilantism was sufficiently notorious that ‘Lynch’s law’ became an American colloquial expression for vigilante violence.” (Brundage 1999: 297)

Although Lynch and his mob did hang people and were, in that sense, killers as well as terrorists, neither they nor their radical Whig counterparts, who also used collective violence to harm persons and to destroy property, were attempting to exterminate the Anglo-British people or the Anglo-Caucasian race. This was not “genocide,” which implies a campaign to exterminate a “genus” or species of persons that might be defined by cultural imaginaries, such as race, religion, or nationality. Genocide includes plans to exterminate Jews or Croats or Palestinians, not to mention Native Americans, who were routinely massacred in what might well be considered genocidal campaigns, justified by the principal that “the only good Indian is a dead Indian.” Generally, the colonial revolt of 1776 was neither terrorist plot nor genocide. It was a bourgeois colonial revolt, which provided cover for a variety of less scrupulous and more noble campaigns, including “workingmen’s and artisan ‘mobs’ of the North End and South End [of Boston]” (Brown 1975: 55), as well as the Southern lynch mobs, not just in Virginia, but particularly in South Carolina, where the Regulators were organized as something like a vigilance committee or citizen militia in the backcountry West of the Appalachian Mountains (Brown 1975: 96).

The differences between the lynching in Western Virginia, the vigilantism in South Carolina and the acts of rebellion in New England are significant. It appears that the Regulators were subsistence-plus “yeoman” farmers (Kulikoff 1989), engaged in a different sort of colonial revolt, a version of “internal colonialism,” in which Regulators and Anti-Regulators represented the conflicting interests of commercial capitalism and labor, albeit capital tied to slave relations of production and labor tied to petit bourgeois landed patriarchal exploitation of family labor (Rubenstein 1970). Here it is not clear that the bourgeois colonial revolt was anything more than an opportunity for east and west to mobilize and fight for new advantages, as Crown and Colonial leaders were fighting amongst themselves.

In contrast, Lynch and the Sons of Liberty defended the interest of bourgeois colonial revolt, in opposition to the Crown. While this interest united them, they differed primarily in target. The target of the Sons of Liberty and their allies were tax officials and other colonial government officials (Morgan and Morgan 1974, chapter XI). Judge Lynch targeted civilians. Like Quantrill’s Raiders, who inflicted the Lawrence Massacre in 1863 (Brown 1975: 9; Hofstadter and Wallace 1971: 89-92), Lynch was a terrorist, using violence to intimidate civilians. John Brown, a white abolitionist, also used violence, both in Kansas and in West Virginia, where his raid on Harper’s Ferry led to his capture and execution, but he was leading an armed insurrection, attacking troops and ultimately a Federal armory (Hofstadter and Wallace 1971: 96-101). Although his interest (abolitionism) was diametrically opposed to the Confederates who attacked Fort Sumter, his tactics and target and general strategy were virtually the same (May 2013: 156-157, 246)

From the perspective of the Crown in 1773 or the federal government in 1859 (Harpers Ferry) or 1863 (Lawrence Massacre), the difference between rebels and terrorists (or guerillas) might be blurred. All were challenges that inspired repression, but the meeting that produced the Declaration of Independence in 1776 was equally threatening to the Crown. Similarly, the Ku Klux Klan marches, down the main streets of cities across the country during the 1920s, were terrorism, just as much as the lynching that came before and after (McVeigh 2009).

The point is that terrorism is a strategy for inciting fear within a civilian population. Lynching is a tactic that might prove effective in this endeavor, but it is not the only tactic available. In fact, as we have learned from Tilly (1986; 1995; 2008) and Tarrow (2011), there has been a sea change in the repertoire of contention between the American Revolution of 1776 and the resurgence of the Klan in the early 20th century (1915-1924). There has been a decline in the use of direct action and patronized actions, in which challengers imitate the actions of authorities while taking the law into their own hands—this includes food riots as well as vigilantism and even lynching.

Public meetings and petition campaigns were not invented in the nineteenth century, but they came to predominate in the British repertoire of contention (Tarrow 2011: 47 and 51). Similarly, after the invention of the modern social movement, old forms of direct action and patronized local actions continued, but they were less common, particularly in relatively stable modern democratic states.

For reasons that we might want to discuss in depth, there is growing opposition to the Resource Mobilization and Political Process models of contention that will guide this analysis (with a healthy dose of historical materialism to complement these models). Some of the earliest critics (Useem 1980) took aim at some of the weakest claims, attributing to Resource Mobilization theory the claim that discontent does not matter, but these critics later became more sophisticated in defending social disorganization theory from a state-centered perspective (Useem and Goldstone 2002). Later, as sociologists took the cultural turn (Goodwin and Jasper 1999), critics claimed that Political Process and Resource Mobilization theories were static structural theories that ignored culture. Still others (Einwohner 2003; McVeigh 2009) claimed that these theories could not explain the revolt of the powerless or right wing middle class movements.

More helpful than these largely unfounded criticisms are more micro-level theories that use framing and discourse theory to augment the structural analysis of interests and opportunities (Benford and Snow 2000; Gamson 1992; Snow et al. 1986; Snow et al. 2014; Steinberg 1998; Steinberg 1999). This allows us to consider how the Republican Party, after its flirtation with Progressivism, was able to accommodate if not fully support the reactionary ravings of the Ku Klux Klan and even the most reactionary elements within the genetic engineering movement. At the same time, some of these pioneers of Progressivism have since been claimed by community organizers and pro-choice feminists who are able to frame the questionable aspects of Progressivism within a more liberal, reformist stance that recognizes the racism within early feminism, for example, and the fact that Progressive tools, such as the referendum and the recall, can be used for reactionary as well as radical goals, as evidenced in the California Tax Revolt and the California Coastal Act (Author 2003).

How these framing and discourse theories can help us to steer a path between structural determinism and post-modern whimsy should be evident as we look at three waves of terrorism in Indiana: Civil War and Reconstruction vigilantism, Post-Reconstruction lynching, and the Klan. We will be guided by Wade (2011), who argues that vigilantism, lynching, and Klan were all expressions of a hegemonic whiteness, sustained by the authority of the Indiana Constitution, which sanctioned collective violence wielded initially by the Indiana Horse Thief Detective Association and later by the Klan. The fact that the Indiana Klan used legislative sanction for the Detective Association as a ruse for establishing a vigilante private police force, during the 1920s, makes the case all the more compelling (Chalmers: 1987: 165-166; Jackson 1967: 145-146; McVeigh 2009: 135).

Not only did the vigilantes and the Klan share the same reactionary, exclusionary, interest—typically petit bourgeois, they availed themselves of the same organization—the Indiana Horse Thief Detective Association, sanctioned by the state legislature and appropriated by the Klan. From a Resource Mobilization perspective (Tilly 1978: 56), these challengers differ only in opportunity/threat and, of course, repertoire (Tilly and Wood 2014)3. For our purposes, Presidential elections represent the most important, recurring opportunity, where divisions between elites and the potential for powerful allies make repression less likely and toleration if not facilitation more likely (Tilly 1978; Tarrow 2011). In the analysis that follows, the critical elections of 1876, the end of the Greenback challenge and of federal support for Reconstruction, 1896 (the Democrat-Populist coalition), and 1920 (Republican-Klan coalition in Indiana) are used to indicate political opportunities, which varied by county according to partisanship. County level variation in interests and organization will be represented by percent black (which could be considered a measure of threat [Beck 2000], as will be explained below) and the number of children enrolled in public schools—a proxy for petit bourgeois campaigns for public order, family, and religion.

The new collective choice theories have adopted racial competition as a predictor of collective action, but they tend to ignore the fact that competition is a relationship. Even the best these studies, which explore the “multi-dimensionality” of ethnic and racial threat—economic, social, and political, rather than simply demographic, still do not model repression/facilitation as a relationship between more or less resourceful and organized (and mobilized) contenders (Cunningham 2013: 8, 234 n8). Even adding “mediation” (Cunningham 2013: 8-10, 234 n9), as I do below when quoting a Democratic newspaper editor, does not change the fact that these are static structural rational choice models, as opposed to dynamic relational models of contentious interaction (Tilly 2008). Particularly in Indiana, where the threat of black or foreign competition was (and still is) more apparent than real, it is misleading to consider threat or even perceived threat independent of the relations between whites and blacks. Thus percent black or even the size of the black or foreign population indicates, at best, the possibility of a threat to the native white constituency of the local branches of the National Horse Thief Detective Association or the KKK. It is only in counties such as McIntosh County, Georgia, during Reconstruction, where black freedmen actually dominated local politics and managed to sustain an economic, cultural and partisan base, that we can reasonably assert that black population and migration threatened white bourgeois hegemony, which was, in fact tenuous in Georgia, even statewide, in 1868 (Author 2011).

In places like Indiana, 1858-1930, blacks and foreigners did not threaten whites in any substantial manner—not even in the same way that the Chinese threatened the Irish Catholic Democrats of Colorado in 1880 (Author 1990). In both cases, it was the carefully constructed partisan image of threat that was effective in mobilizing partisans, who already shared an interest in opposition to both lumpen proletarian and capitalist/employer class fractions. Their artisanal or petit bourgeois interests were rooted in relations of production and relations with other classes. Their willingness to kill or threaten ethnic and racial minorities required more partisan organizational effort than objective evidence of labor market competition. To argue otherwise is to blame the victims or, at best, to accept the claims of the oppressors. In our efforts to take reactionaries seriously, we do not need to agree with their rhetoric or to search for evidence that might support their claims.

Thus the challenge in what follows is to defend two simple assertions that might inform the historical and sociological debates. First, parties represent material interests, but these are frequently contradictory interests that are provisionally allied in a platform or program, whose realization tends to make the contradictions manifest, thereby undermining the coalition (Author 2005). For example, Radical Whigs and Reactionaries in Virginia were accommodated in the bourgeois colonial coalition that defended confederacy in opposition to the crown. The Articles of Confederation, once the colonial revolt was complete, defined the limited terms for coalition in an umbrella government that was designed—like the modern European Community, to facilitate cooperation among independent states. Of course, this government could not govern, which led to extended conflict, compromise, and a certain amount of rebellion and repression in establishing a federal government that, initially, was defended by the Federalists, including conservatives like Alexander Hamilton and liberals like James Madison.

Clearly, the Federalist Party did not survive its victory either, but the two party systems, first Whig versus Democrat and then Democrat versus Republican, seem to have endured and might yet endure deep into the 21st century. Most important, for present purposes, is the fact that none of these parties were ever ideologically homogeneous. Neither all Whigs nor all Republicans were Radical—but some of them were. Not all Democrats were Reactionary—but the Redeemers clearly were. In fact, there is a Reactionary strain that runs deep among Southern Democrats, who have now become Republicans, but that change—effected by Nixon in the election of 1972, need not concern us today.

More important for now is our second assertion. To the extent that the local community is effectively under the control of a hegemonic republican capitalist regime—what Tilly (2007) would call a “democracy,” political challenges tend to be relatively tame and symbolic, compared to the direct action of vigilantes and lynch mobs. This remains true, even when the challengers represent reactionary interests of Redeemers and Klansmen. Lynching will tend to diminish as the civilizing effects of petit bourgeois churches and schools and vibrant family life provide a more suitable base for the repertoire of the modern social movement—including demonstrations and protests by Indiana Pro-Life, which seems in some ways reminiscent of a Klan rally.

On that note, with the promise to return to thinking about terrorism in the 21st century, let us turn to the Indiana frontier and look for evidence that might support these assertions. We will proceed slowly, beginning with some pictures, moving to a chart, then some simple descriptive statistics. Then, we’ll look at maps of Indiana counties, 1870-1920, before turning to some zero inflated negative binomial equation (zinbe) models, to test for the effects of social and political correlates of terrorism.
A Methodological Aside

Readers who wish to hear more about the data gathered by Pfeifer (2014) and McVeigh (2009) should turn to these sources. For our purposes, there are only a few points worth noting. First, both datasets came to me as XL files that I read into Access and manipulated in various ways to get counts by counties over time. Eventually, these data were written back into XL files and then read into STATA, where I was able to generate Zero Inflated Negative Binomial Estimate (zinbe) models predicting lynching victims or KKK events. The cases are the 92 Indiana Counties, and these boundaries did not substantially change between 1870 and 1920, which makes the mapping and the statistical analysis relatively simple.

Both datasets purport to be populations, although they are based on the coding of newspaper reports (by Pfeifer 2014 or his colleagues, or by McVeigh 2009). Pfeifer’s (2014) data report victimizations rather than events (as will become clear below). McVeigh (2009) used the official national KKK newspaper to code events by city (when possible). I later attributed county to cities and treated events not identified with code-able places as missing data. McVeigh’s (2009) events probably over-estimate KKK events, since that is the point of the national paper. Pfeifer’s (2014) victimizations are probably low estimates and probably do not indicate subtle differences in method of execution (as will become clear).

One last qualification is in order. Working with small datasets that are putatively populations (as opposed to samples) gives me more confidence in the validity of my measures—at least, on the dependent variables. The independent variables were downloaded from ICPSR, and the county map boundaries were downloaded from NHGIS. Descriptive tables and maps, combined with predictive models, also should contribute to some comfort on matters of validity. We can see which counties were lynching or Klan centers—at least in quantity of victims or events, and we can imagine how the independent variables distinguish these counties (maps or tables on independent variables are available on request).

The more serious problem here, with only 92 cases, is reliability, particularly since these are essentially ecological correlations. The estimates have large standard errors, which, combined with problems of collinearity, suggest that we should interpret these results with skepticism.4 A model that seems to capture inter-county differences in lynching before the end of Reconstruction (1876) does not do so well after Reconstruction, and it might well be that the differences are rooted in the relationship between total population (or other measures of urban places) and black population—the denominator and numerator (respectively) of percent black. Thus we should take solace in those effects which seem to be particularly robust, although we can certainly argue about model specification. For those (like myself), who enjoy fiddling with models and attempting to find the best fit, I have included an appendix with a table that offers the best model I could find to predict KKK events in Indiana in 1924. This is not the model used in the text, since the goal was to offer different estimates using essentially the same model in order to understand how the phenomenon under examination or the mechanisms that make it more or less likely changed over time within place.

We could, of course, spend more time explaining how negative inflation and other binomial effects should be interpreted. Personally, I have never found the algebra to be particularly enlightening, but I might just say that inflation effects are estimated as bias in populations where most cases have values of zero. Variables that predict zero are included in order to estimate and control for this bias. Thus a negative binomial inflation factor—number of blacks, for example, indicates, in this example, that where there are few blacks there are likely to be few lynching victims. Then, the main effects of the model, for example, percent black, are unbiased by the inflation factor, so the positive effect of percent black would indicate that counties with a proportionally greater black population would have more lynching victims.5

We will return to these considerations when we look at the zinbe models, but first, let’s look at some pictures.
Terrorism Hoosier Style

If you have ever seen pictures of an Indiana lynching, you probably saw this picture.

Figure 2

Marion Indiana Lynching, 1830

source: Madison 2011 (courtesy of the Indiana State Historical Society).
This famous Indiana lynching took place in Marion, Indiana, in Grant County. Accounts of this event focus on the extent to which this was not a typical lynching. Consider, for example, the account of James H. Madison (2011), the Thomas and Kathryn Miller Professor of History at Indiana University (whose account included this photo).
A Lynching in the Heartland:

Download 117.99 Kb.

Share with your friends:
  1   2   3

The database is protected by copyright ©essaydocs.org 2022
send message

    Main page