¹ For recent examples of the ‘obstacles to democracy’ approach, see Rueschemeyer, Dietrich, Stephens, Evelyne Huber and Stephens, John D., Capitalist Development and Democracy (Chicago, 1992), Chapter 6Google Scholar, ‘Central America and the Caribbean’, and Torres-Rivas, Edelberto, El tamaño de nuestra democracia (San Salvador, 1992), pp. 19–25Google Scholar. The civil-military relations literature is cited below.

² Giddens, Anthony, A Contemporary Critique of Historical Materialism, vol. II (Berkeley, 1987)Google Scholar, The Nation-State and Violence, pp. 1–5, 22–23, 30; emphasis added. Smith, Peter H., ‘A View from Latin America’, Journal of Interdisciplinary History, vol. 12 (1981), pp. 3–27CrossRefGoogle Scholar.

³ Historically grounded studies of proto-state violence as an autonomous category in this sense have been scarce for the rest of the continent as well. Silvio R. Duncan Baretta and John Markoff's observation nearly two decades ago that, ‘As a background condition violence is readily forgotten’ by scholars who tend to see violence as ‘straightforward and uncomplicated’ is still an accurate one: their study of what they called ‘the tradition of violence’ on the cattle frontiers of Latin America is practically unique. See ‘Civilization and Barbarism: Cattle Frontiers in Latin America’, Comparative Studies in Society and History, vol. 20, no. 4 (October 1978), pp. 587–620. Their claim was echoed much more recently by Fernando Coronil and Julie Skurski's comment that ‘Although political violence has played a central part in the formation of nations, its historical constitution and its role in representing nations have received scant attention.’ Violence is too often ‘reduced to a practical tool used by opposing social actors in pursuit of conflicting ends. Whether treated as a cause, function, or instrument, violence is generally assumed rather than examined in its concreteness’. ‘Dismembering and Remembering the Nation: The Semantics of Political Violence in Venezuela’, Comparative Studies in Socciety and History, vol. 33, no. 2 (April 1991), pp. 288–9.

⁴ While ‘authoritarianism, militarism and violence’ have been ‘transitory’ features elsewhere in Latin America (Brazil, Argentina, Chile, Uruguay) they have been ‘quotidian, traditional and permanent’ in Central America, rooted in the isthmus' ‘economic reality, social organisation and political culture’. Cruz, Rodolfo Cerdas, El desencanto democrático: Crisis de partidos y transitión democrática en Centra América y Panamá (San José, C. R., 1993), pp. 21, 23Google Scholar.

⁵ A species of caudillo, identified as ‘oligarchic dictators’ in Lynch's quasi-evolutionary schema of caudillismo, endured in Central America beyond their natural lifespan elsewhere; while retaining caudillist features such as personalism, patronage and the use of violence, the post-1870s dictators worked in a more centralised environment and confronted more complex social forces. ‘The sanction behind the modern dictatorships, it is true, was still violence and state terror, but the political process was no longer as crude as that of their predecessors.’ Lynch, John, Caudillos in Spanish America, 1800–1850 (New York, 1992), pp. 429–30CrossRefGoogle Scholar. A descriptive treatment of this period of ‘caudillismo frente a constitucionalismo’ that highlights its centralizing tendencies, especially in the deployment of violence, is Arriola, Arturo Taracenca, ‘Liberalismo y poder político en Centroamerica (1870–1929)’, in Las repúblicas agroexportadoras (1870–1945), ed. Ortega, Víctor Hugo Acuña, Tomo IV of Historia general de centroamérica (Madrid, 1993), pp. 167–254Google Scholar. For an analysis of the earlier period of caudillo rule that emphasises the ‘devastating’ rivalry among regional strongmen, see Woodward, R. L., ‘Central America from Independence to c. 1870’, Cambridge History of Latin America, ed. Bethell, Leslie, Vol. III, From Independence to c. I8JO (Cambridge, 1985), pp. 471–506CrossRefGoogle Scholar.

⁶ By ‘violence’ I mean physical harm inflicted on people or their property; for a discussion of violence in these terms, see Telia, Torcuato S. Di, Latin American Politics: A Theoretical Framework (Austin, 1990), pp. 76–9Google Scholar. By ‘limits’ I mean the freedom of state agents to (a) both define and control (through sanctions or inducements) enemies of the state, (b) avoid punishment for committing ‘illegal’ acts (i. e., impunity), and (c) informally deputise non-state agents in the exercise of state-sponsored violence (death squads). Variations in the elasticity of these freedoms would constitute shifts in the limits of state-sponsored violence. While such variations would be most conspicuous during periods of intense change, it is above all the secular character of these limits, their routinised, taken-for-granted quality over the course of decades and centuries, that would be of greatest relevance to the study of violence in the region. I borrowed the phrase ‘upward displacement’ from Anderson, Perry, who applied it to the process by which power was transferred from feudal lords to the absolutist state in Lineages of the Absolutist State (London, 1979), p. 19Google Scholar.

⁷ Tilly, Charles, Coercion, Capital and European States, AD 990–1990 (Cambridge MA, 1990) pp. 68–9Google Scholar.

⁸ Tilly, Charles, Big Structures, Large Processes, Huge Comparisons (New York, 1984), pp. 9–10Google Scholar.

⁹ For standard interpretations of the transition see two articles by Mariscal, Nicolás, ‘Régimenes políticos en El Salvador’, Estudios Universitarios, no. 365 (marzo 1979), pp. 139–52Google Scholar, and ‘Militares y reformismo en El Salvador’, Estudios Centroamericanos, no. 351/352 (enero-febrero 1978), pp. 9–27; McDonald, Ronald H., ‘Civil-Military Relations in Central America: The Dilemmas of Political Institutionalisation’, in Wiarda, Howard (ed.), Rift and Revolution: The Central American Imbroglio (Washington, 1984), pp. 129–66Google Scholar; Rouquié, Alain, The Military and the State in Latin America, trans. Sigmund, Paul E. (Berkeley, 1987), Chap. 5Google Scholar; Lynch, Caudillos, pp. 425–30; Pastor, Rodolfo, Historia de Centroamérica (Mexico, 1988Google Scholar; reprint, Guatemala, 1990), pp. 191–196, 210–12 (page references are to reprint edition); Véjar, Rafael Guidos, El ascenso del militarismo en El Salvador (San Salvador, 1988)Google Scholar; and a critical analysis of some of the theories by Steve Ropp, C., ‘Teorias sobre el comportamiento de los militares centroamericanos’, Estudios Centroamericanos, no. 451/452 (mayo-junio 1986), pp. 411–30Google Scholar.

¹⁰ Caudillos in their relation with the popular sector were largely the agents of the elite, whose interests they protected from popular insurgency and social unrest; caudillos were ‘the necessary gendarme’. Lynch, Caudillos, pp. 234–6.

¹¹ Tilly, Coercion, Capital, pp. 103–4, 122–4.

¹² Tilly himself proposes this possibility for Third World societies after World War II; Coercion, Capital, pp. 218–23.

¹³ Lynch, Caudillos, p. 406.

¹⁴ Ibid., pp. 433–7, 4–5.

¹⁵ The classic expositions of this model are Wolf, Eric, ‘Aspects of Group Relations in a Complex Society: Mexico’, American Anthropologist, vol. 58 (1956), pp. 1065–78CrossRefGoogle Scholar and Powell, John Duncan, ‘Peasant Society and Clientelist Polities’, American Political Science Review, vol. 64, no. 2 (06 1970), pp. 411–25CrossRefGoogle Scholar. For a subsequent refinement, see Eisenstadt, S. N. and Roniger, L., Patrons, Clients and Friends: Interpersonal Relations and the Structure of Trust in Society (Cambridge, 1984), pp. 231–44CrossRefGoogle Scholar.

¹⁶ In their classic theoretical treatment of caudillismo, Eric R. Wolf and Edward C. Hansen affirmed that caudillo leadership rested heavily on the demonstration of masculinity, which they defined as two closely related attributes: the domination of females and the readiness to deploy violence; see Wolf, Eric R. and Hansen, Edward C., ‘Caudillo Politics: A Structural Analysis,’ Comparative Studies in Society and History, vol. 9, no. 2 (01 1967), pp. 174, 177CrossRefGoogle Scholar. For thick (and entertaining) description, see Chasteen's, John Charles recent reconstruction of the careers of the Saravia brothers, Heroes on Horseback: A Life and Times of the Last Gaucho Caudillos (Albuquerque, 1995), ch 9Google Scholar.

¹⁷ Eisenstadt and Roniger, Patrons, Clients and Friends, p. 49.

¹⁸ Alvarenga, Ana Patricia, ‘Reshaping the Ethics of Power: A History of Violence in Western Rural El Salvador, 1880–1932’, unpubl. PhD diss., University of Wisconsin, 1994, pp. 231–65Google Scholar, 267, 289, 365–78, 62, 71–2.

¹⁹ Walter, Knut and Williams, Philip J., ‘The Military and Democratization in El Salvador’, Journal of Inter-American Studies and World Affairs, vol. 35, no. 1 (1993), P. 47CrossRefGoogle Scholar.

²⁰ Carmack, Robert M., ‘State and Community in Nineteenth-Century Guatemala: The Momostenango Case’, in Smith, Carol A. (ed.), Guatemalan Indians and the State: 1540 to 1988 (Austin, 1990), pp. 116–36Google Scholar; quote is at p. 121. For the deployment of Indian community militias in his own personal defence by Manuel Estrada Cabrera (1898–1920), see Adams, Richard N., Etnicidad en el ejército de la Guatemala liberal (1871–1915) (Guatemala, 1995), pp. 28–30Google Scholar.

²¹ Schroeder, Michael J., ‘To Defend Our Nation's Honor: Toward a Social and Cultural History of the Sandino Rebellion in Nicaragua, 1927–1934’, unpubl. PhD diss., University of Michigan, 1993, pp. 114Google Scholar, 125–6, 196, 202, 209, 353–4, 77. See also his article in this issue of the Journal.

²² Gould, Jeffrey L., ‘Vana Ilusión! The Highlands Indians and the Myth of Nicaragua Mestiza, 1880–1925’, Hispanic American Historical Review, vol. 73 (08 1993), pp. 420–1CrossRefGoogle Scholar.

²³ Walter, Knut, The Regime of Anastasio Somoza, 1936–1956 (Chapel Hill, 1993), pp. 57–58, 97Google Scholar.

²⁴ Barahona, Marvin, ‘Caudillismo y política en Honduras (1894–1913)’, paper presented at the conference, ‘Estado, participation polftica e identidad national en Centroamérica, siglos XIX y XX’, San José, Costa Rica, 23–25 02 1995Google Scholar.

²⁵ Bailey, Norman A. elaborated inventively on international expressions of caudillaje in ‘The United States as Caudilto’, Journal ofInter-American Studies, vol. 5 07 1963), pp. 315–24Google Scholar. The essential role of arms transfers in international caudillaje was developed by Christopher Shoemaker, C. and Spanier, John, Patron-Client State Relationships: Multilateral Crises in the Nuclear Age (New York, 1984), pp. 14–15Google Scholar.

²⁶ A quantitative analysis of the pace, magnitude and character of US collaboration can be found in Holden, Robert H., ‘The Real Diplomacy of Violence: United States Military Power in Central America, 1950–1990’, The International History Review, vol. XV, no. 2 (05 1993), pp. 283–322CrossRefGoogle Scholar.

²⁷ For an excellent review of the problem see Gerner, Deborah J., ‘Weapons for Repression? United States Arms Transfers to the Third World’, in Stohl, Michael and Lopez, George A. (eds.), Terrible Beyond Endurance? The Foreign Policy of State Terrorism (New York, 1988), pp. 247–80Google Scholar.

²⁸ See Holden, ‘The Real Diplomacy of Violence’, pp. 307–11.

²⁹ This is not to argue that Latin America is ‘more violent’ than the rest of the world. The United States and the European states (including Russia) have devoted more resources to perfecting strategies of violence than to any other conceivable state activity. Countless imperialist forays, genocidal campaigns of extermination directed against internal ‘enemies’, two world wars and a half century of nuclear arms production are enough to overshadow Latin America in any accounting of state-sponsored violence. The subject here, of course, is not violence between nation-states nor technical capacity but the intensity of overt violence accompanying statemaking within national frontiers, a realm in which the Latin American countries have clearly excelled since independence.

³⁰ These two premises permeate the social science literature on the Latin American military, and multiple examples could be cited. Among many others, that of Alfred Stepan must be mentioned, for its influential character; see especially Rethinking Military Politics: Brazil and the Southern Cone (Princeton, 1988). Other recent examples: Varas, Augusto, ‘Las relaciones civil-militares en la democracia’, in Kruijt, Dirk and Torres-Rivas, Edelberto, coord., América Latina: militares y sociedad, vol. II (San José C. R., 1991), pp. 153–80Google Scholar; Fitch, J. Samuel, ‘Armies and Politics in Latin America: 1975–1985’, in Lowenthal, Abraham F. and Fitch, J. Samuel (eds.), Armies and Politics in Latin America (rev. ed., New York, 1986), pp. 26–58Google Scholar; and Zaverucha, Jorge, ‘The Degree of Military Political Autonomy During the Spanish, Argentine and Brazilian Transitions’, journal of Latin American Studies, vol. 25, no. 2 (05 1993), pp. 283–300CrossRefGoogle Scholar. Similar assumptions underlie the work of historians who write about the military; prominent examples include Potash's, R. A. wo volumes on the Argentine army –The Army &Politics in Argentina 1928–1945: Yrigoyen to Perón (Stanford, 1969)Google Scholar and The Army & Politics in Argentina 1941–11)62: Perón to Frondiz (Stanford, 1980) – and the work of Nunn, Frederick M. – Chilean Politics 1920–1931: The Honorable Mission of the Armed Forces (Albuquerque, 1970)Google Scholar and The Time of the Generals: Latin American Professional Militarism in World Perspective (Lincoln, 1992).

³¹ Among the insights that made Huntington's, Samuel J.Political Order in Changing Societies (New Haven, 1968)Google Scholar a pathbreaking achievement was its claim that it was not the military itself but the ‘political and institutional structure of society’ that explained military intervention in politics. However, Huntington and his disciples largely confined themselves to studying the contemporary role of the military as an institution, focusing not on the broader question of state-sponsored violence but on ‘military rule’, and all but excluding any interest in the historical dimension. Of course, they also launched a very different argument from that platform than the one that I am proposing to make, claiming that increasing political disorder called into being an interventionist military institution with middle-class roots that would somehow midwife ‘modern’ civilian political institutions, thus subduing disorder and conflict (p. 194). For the originality of Huntington's observation at the time, see Berghahn, Volker R., Militarism: The History of an International Debate 1861–1979 (New York, 1982), p. 76Google Scholar.

³² Giddens, Anthony, The Nation-State and Violence, vol. II of A Contemporary Critique of Historical Materialism (Cambridge, 1985), p. 251Google Scholar.

³³ Kalmanowiecki, Laura insightfully links police corruption and violence in ‘Police, People, and Preemption in Argentina’, in Huggins, Martha K. (ed.), Vigilantism and the State in Modern Latin America (New York, 1991), pp. 47–60Google Scholar.

³⁴ This line of argument is bolstered by Michael Mann's dictum to think about societies not as ensembles of ‘subsystems’ but as ‘multiple overlapping and intersecting sociospatial networks of power,’ in which organisation and function are not coterminous but highly elastic, capricious or ‘promiscuous’ over time. The Sources of Social Power, vol. I, A History of Power from the Beginning to A.D. 176O (Cambridge, 1986), Ch. 1.

³⁵ Arganaras, Fernando Garcia, ‘The Mechanisms of Accommodation: Bolivia, 1952–71,’ Review (Fernand Braudel Center for the Study of Economies, Historical Systems and Civilizations), vol. XV, no. 2 (Spring 1992), pp. 292–94Google Scholar, develops this point to explain the frequency of transactions that ‘blur the boundaries between the state and the private realm’.

³⁶ Munro, Dana, The Five Republics of Central America: Their Political and Economic Development and Their Relations with the United States (New York, 1918), pp. 42–3Google Scholar.

³⁷ Toledo, Mario Monteforte, Centro America: Subdesarrollo y dependencia, vol. II (Mexico, 1972), p. 216Google Scholar.

³⁸ Torres-Rivas, El tamańo de nuestra democracia, p. 45.

³⁹ Rouquié, Alain, ‘Demilitarization and the Institutionalization of Military-dominated Polities in Latin America’, in O'Donnell, Guillermo, Schmitter, Philippe C. and Whitehead, Laurence (eds.), Transitions from Authoritarian Rule: Comparative Perspectives (Baltimore, 1986), part 2, p. 133Google Scholar. For O'Donnell and Schmitter's comment, see ‘Opening (and Undermining) Authoritarian Regimes’, in the same volume part 4, p. 31.

⁴⁰ For a brief critical review of social-science approaches to national-level violence see Warren, Kay B., ‘Introduction: Revealing Conflicts Across Cultures & Disciplines’, in Warren, (ed.), The Violence Within: Cultural and Political Opposition in Divided Nations (Boulder, 1993), pp. 1–24Google Scholar. An excellent review of some recent work on the Andes is Mayer's, Enrique ‘Patterns of Violence in the Andes’, Latin American Research Review, vol. 29, no. 2 (1994), pp. 141–71Google Scholar. Also see Greenberg, James B., Blood Ties: Life and Violence in Rural Mexico (Tucson, 1989)Google Scholar; Graziano, Frank, Divine Violence: Spectacle, Psychosexuality & Radical Christianity in the Argentine ‘Dirty War’ (Boulder, 1992)Google Scholar; Poole, Deborah (ed.), Unruly Order: Violence, Power, and Cultural Identity in the High Provences of Southern Peru (Boulder, 1994)Google Scholar; Corradi, Juan E., Fagen, Patricia Weiss, and Garretón, Manuel Antonio (eds.), Fear at the Edge: State Terror and Resistance in Latin America (Berkeley, 1992)Google Scholar; and some of the essays in Huggins, Martha K. (ed.), Vigilantism and the State in Modern Latin America (New York, 1991)Google Scholar.

⁴¹ Hunger, Interfaith Appeal, ‘Interfaith Hunger Appeal Announces Grants For Overseas Research’, Internet bulletin, 11 13, 1995Google Scholar. The Appeal identified its ‘partner agencies’ as Catholic Relief Services, Church World Service, Lutheran World Relief and the American Jewish Joint Distribution Committee.

⁴² Almond, Gabriel A., ‘Foreword: The Return to Political Culture,’ in Diamond, Larry (ed.), Political Culture and Democracy in Developing Countries (Boulder, 1993), pp. ix–xiGoogle Scholar.

⁴³ Swidler, Ann, ‘Culture in Action: Symbols and Strategies’, American Sociological Review, vol. 51 (04 1986), p. 284CrossRefGoogle Scholar.

⁴⁴ Eckstein, Harry, ‘A Culturalist Theory of Political Change’, American Political Science Review, vol. 82, no. 3 (09 1988), pp. 789–803CrossRefGoogle Scholar. Political culture as a ‘historical creation’, and its reciprocal relationship with political action, are stressed by Baker, Keith, ‘Introduction’, in Inventing the French Revolution: Essays on French Political Culture in the Eighteenth Century (Cambridge, 1990), p. 10CrossRefGoogle Scholar. Ruth Lane analyses the application of the concept since its introduction in 1956 by Almond, Gabriel, favourably assessing its utility for integrating ‘the sociological and the individual’ in ‘Political Culture: Residual Category or General Theory’, Comparative Political Studies, vol. 25, no. 3 (10 1992), pp. 362–87Google Scholar.

⁴⁵ Chavez, Alicia Hernandez, La traditión republicana del buen gobierno (Mexico, 1993), p. 9Google Scholar.

⁴⁶ Morse, Richard, New World Soundings: Culture and Ideology in the Americas (Baltimore, 1989), pp. 133–4Google Scholar, 98–106, 122–6. The work of Roberto Da Matta, heavily cited by Morse, should also be mentioned here. Kindred historical interpretations have been offered by Glen Caudill Dealy, who traces the caudillo's ability to ‘to use force with a good conscience’ to a Thomistic ‘dual morality’; Claudio Véliz, who locates Latin America's political culture in a centralist tradition stemming from the Counter-Reformation; and Howard Wiarda, who holds up a distinctly Latin corporatist tradition. See: Dealy, , The Latin Americans: Spirit and Ethos (Boulder, 1992), p. 24Google Scholar; Véliz, , The New World of the Gothic Fox: Culture and Economy in English and Spanish America (Berkeley, 1994), p. 210Google Scholar; Wiarda, , ‘Toward a Model of Social Change and Political Development in Latin America: Summary, Implications, Frontiers’, in Wiarda, H. (ed.), Politics and Social Change in Latin America: The Distinct Tradition, 2nd ed. (Amherst, 1982), pp. 329–59Google Scholar.

⁴⁷ Lechner, Norbert, ‘La democratizatiáon’, in Lechner, (ed.), Culíura politico y democratizatión (Santiago, Chile, 1987), pp. 253–4Google Scholar and Lechner, ‘Presentation’, Ibid., pp. 7–9.

⁴⁸ Loveman, Brian, The Constitution of Tyranny: Regimes of Exception in Spanish America (Pittsburgh, 1995), p. 63Google Scholar. This excellent study of the long history of constitutional protection for tyrannical regimes is basically framed by the familiar civilian vs. military dichotomy so that the problem appears to be one of ‘military rule’. Yet, like Morse, Loveman also recognises the complex origins of a political culture that nourished authoritarianism; he points to the deadly combination of (1) the absence in colonial Spanish America (unlike in Spain itself) of any parliamentary or representative institutions that might have constrained the authority of the monarch and the consequent practice of rule by royal decree, and (2) the late Bourbon policy of turning over internal administrative functions to the military (35–6). The colonial-era practice of yielding broad administrative authority to military powers was adopted and updated by the former colonies (393–5).

⁴⁹ Comisión Especial del Senado sobre las Causas de la Violencia y Alternativas de Pacificatión en el Peru, Violencia y pacificatión (Lima, 1989), pp. 34, 124, 39Google Scholar.

⁵⁰ Comisión Especial, Violencia y pacificatión, p. 43.

⁵¹ Mayer, ‘Patterns of Violence’, pp. 143–4.

⁵² Greenberg, Blood Ties.

⁵³ But Chasteen, in his haste to disavow cultural determinism, argues that ‘cultural patterns are entirely as contingent as economic ones, in process no less constantly and multidimensionally’, thus undermining (it would seem) the explanatory power of the very ‘conditioning factors’ that he is attempting to illustrate. ‘Fighting Words: The Discourse of Insurgency in Latin American History’, Latin American Research Review, vol. 28, no. 3 (1993), pp. 83–112.

⁵⁴ El tamaño, pp. 38, 41, 29. Similarly, Franz Hinkelammert noted that in Central America, ‘the first thing that a government requires of the opposition is to demonstrate that it is not its enemy. The opposition must constantly demonstrate that it is a defender of the social system in order to be accepted.’ Hinkelammert, ‘El concepto de lo politico segiin Carl Schmitt’, in Lechner (ed.), Cultura política, pp. 235–6.

⁵⁵ Torres-Rivas, El tamaño, p. 29.

⁵⁶ See note 4.