Collective Violence as Social Control Roberta Senechal de la Roche Sociological Forum, Vol. 11, No. 1. (Mar., 1996), pp. 97-128. Stable URL: fip:flinks jstor-org/sici sici=0884-897 | 281996039291 1%3A 1 3CITHIACVASCHIE2.0,CORSB2-L Sociological Farum is curently published by Springer Your use of the ISTOR archive indicates your acceptance of ISTOR's Terms and Conditions of Use, available at flip: feworwjtor org/aboutterms.htmal. ISTOR's Terms and Conditions of Use provides, in par, that unless you fave obtained pcior permission, you may not dowaload an cnt isus of @ journal or multiple copies of articles, and you may use content inthe ISTOR archive only for your personal, non-commercial uss. Please contact the publisher cegarding any further use of this work. Publisher contact information may be obtained at bhupsferw.jstoc.org/joumals/springer. html Each copy of any part of a JSTOR transmission must contain the same copyright notice that appears on the screen or printed page of such transtnission. ISTOR is an independent not-for-profit organization dedicated to creating and preserving a digital archive of scholarly journals. For more information regarding ISTOR, please contact support @jstor.org- hup:thrww.jstor.orgy Fri Aug 25 11:46:31 2006Socotagical Farum, Vol U1, No, 1 1996 Collective Violence as Social Control’ Roberta Senechal de la Roche? Collective violence is often social control: self-help by a group. It Yypically defines and responds to conduct as deviant. When unilateral and nongovernmental, it appears in four major forms—Iyncking, rioting, vigilantism, and terrorism—each distinguished by its system of liability (individual or collective) and degree of organization (higher or lower} Following Donald Black's paradigm of pure sociology, the central assumption. is that collective violence varies with its location and direction in social space—the conflict siracture. I offer ten propositions that predict and explain the likelihood and seventy of collective violence in general and the four forms of collective violence in panicular Conflict structures with « high degree of relational distance, cultural disiance, fuxctional independence, and inequality Getween the adversaries are associated with collective violence in general. Each ‘of the four forms depends on the degree of social polarization beeen the parties as well as the continuity of the deviant behavior to whick the violence responds KEY WORDS! colcctive violence; socal control canfler sruatuie, Foal polation, cantinity of deviant bchavio INTRODUCTION Collective violence is personal injury by a group? Most is sacial con- trol: @ process by which peaple define or respond to behavior as deviant ‘Badier vesons of tis paper were pretend at the Annual Meeting of the American Sociological Association, Pitsbuegh, Penrsyaria, Aogust 1992, and the Centennial Congest ofthe frcerational Insti of Sociology, Universi af Pare (Sorbonne), Par, June 1993 SDeparicient of History, Washington and Lee Unvecaty, Cexiogtn, Virginia 24450 ‘For present purposes, & group's ive or more people acting in concert. (compare, € 2, Olzat, 43G92:55-S7, TO, Til, 197869-85, 174-196) Included ace bth overt and covert violence a5 ell as thar which i threatened, attempted, oF merely paints! sithout physical damage of lasting natuce (Oak, 19925625), ‘Althovgh the fous on physical hare othe person, collective agaresion against property, 7 ‘4 aL 9K6H OFA © 5 Pha Ft, Caprn% Senechat de ta Rocke (Black, 1976:2, 1984:1).* Deviant behavior is not merely conduct an outside observer might regard as odd, abnormal, or illegal, but any action—however seemingly trivial, inoffensive, or innacent—that is subject to social control Collective violence, then, is commonly a moralistic response to deviant be~ havior. And, aptly enough, it is sometimes described as “popular justice.” ‘THE GHOST OF GUSTAVE LE BON Since the 1960s, a conception of collective violence as deviant behavior rather than social control has increasingly been rejected as ideological. A well- known example of the earlier view is Gustave LeBon's (1895/1960) charac- terization of violent crowds 4s irrational, fickle, and destructive riffraft and scoundrels (see Rudé, 1964:3-10, 252-257; Rule, 1988:91-95). Collective vio- lence was also treated as a form of mass hysteria and the dictatorial domi- nation of a group mind (Blumer, 1946; Chaplin, 1959; Foster, 1991; LeBon, 1895/1960) In contrast, collective violence is now commonly regarded as a form of protest, a guest for justice, and the purposive expression of “real griev- ‘ances over underying social, economic, and political issues” (Gurr, 198913; see also Rule, 1988:170-172, 287-290; compare Marx, 1972). One analysis of Euro- pean crowd violence, for example, asserts that “justice .. lies at the heart of violent conflict” (Tilly er af, 1975:85). Another claims that riots by American blacks in the 1960s “can only be understood 2s 2 manifestation of the griev- ances of the black ghetto” (Fogelson, 1970:142-143; see also Feagin and Hahn, 1973). Similar generalizations appear in a voluminous literature. on @ wide. variety of provestotiented actions, from 17¢h- and {8th-century food and tax riots to labor disorders and enlleetive resistance to colonialism (€.2, Rercé, 1990; Edgerton, 1989; Mars, 1970; Rudé, 1964; Tilly et al, 1975) ‘Yet the ghost of Gustave LeBon still stalks the field. While most social scientists have abandoned conceptions of violent workers, peasants, and mi- norities 8 pathological riffraff in need of disciptine and punishment, some continue to characterize collective. violence of some kinds by some people not only 2s deviant behavior but also as undesirable and blameworthy—irra- tional, pathological, or criminal. Some scemingy deserve sympathy while oth ers do not. The violence af some is more rational and comprehensible. than ‘whieh offen occurs in rioting and cerrrism, will occasionally be mentioned as part of collective incident ‘Por other works that treat vilenee as sia eontol, ee Baumgartner (1584); Blac (1983); ‘weiburd (1580). Sta the 1960s, the interpretation of violence shifted decisively from he political right 60 the left Tn this “moral redefinition of the subject matte,” one bias replaces another (Rule, 1988:182-184, 283-290)Caective Violence * that of others. In particular, while collective violence “from the bottom up" by socially subordinate groups tends to be portrayed as reasonable and pos- sibly even worthy of sympathetic understanding, that “from the top dowa” bby dominant or majority groups against minorities, the poor, or members of labor movements, is not, Indeed, violence by social superiors is portrayed in ‘a manner reminiscent of LeBon’s irrational mobs. In one analysis, for exam- ple, anti-Semitic attacks in Germany are called “‘purposeles mob violence” (Tilly ef ab, 1975:235), In another, attacks by whites against blacks in America are said t0 be less “understandable” than assaults by blacks on whites during the 19605 ghetto riots (Fogelson, 1970:154-157). And southern lyachings of blacks by whites are attributed to “a mental condition, a disorder of the mind in which internal problems are projected upon external persons" (Williamson, 1984:151-152). Whereas a number of social scientists and historians acknow- ledge if not applaud the goals, grievances, or motives they attribute to those ‘who use violence. against social superiors, those who use violence against s0- cial inferiors are frequently regarded as illegitimate and deserving of punish- ment. The subject thus appears to be distorted by ideological elements not explicitly acknowledged by the investigators (Berk, 1972; Foster, 1991:464- 467; Gratam, 1989; Marx, 1972:49, 464-467; Rule 1988:183, 280). Collective violence directed at social inferiors and marginals is com- monly not even regarded as sociologically meaningful in itself, but rather is characterized as a symptom of something else, 1¢ is not the punishment of deviant behavior, as its participants claim. Instead, it i, for example, an in- direct byproduct of macrostructural disruptions or strains such as urbaniza- tion, unemployment, or competitian—conditions that are said to frustrate or otherwise predispose individuals to aggressive behavior (Beck et al, 1989 Beck and Tolnay, 1990; Blalock, 1957; Bonacict, 1972; Hency and Short, 1954:51-60; Hovland and Sears, 1940; Bonacich, 1972; Olzak, 1989, 1990, 1992; Tolnay and Beck, 1995) Given an “excuse” (such as a crime or dispute, the importance of which is usually trivialized as a “precipitating” incident), a “mob” vents its anger not at the real source of its misery (such as dectining economic conditions or increased competition), but at the most vulnerable targets in its immediate environment.” One study, for example, describes AL igs here is the conception of collective violence found in these studies and not the ‘ality or importance of Casings showing that rates of viqlence vary with macrosructtal ‘Concitions ofan economic, social, or paliteal natuce (ate eg, eck and Tole, 1980, Henry ‘and Short, 1954; Ozak. 1992: Talngy and Beck, 1995) These saistiea relationships are not Inconsistent with the present anaiss. Thor validity and inmpertinee ate Independent othe imagery of collective violence invoked, interpretations of the meanitg af the Violence, OF alleged It there is fo be social contol, he offense Ise, bet minaeinfacion oF a9 act200 Senechal de ta Roche lynching as “a form of murder” unrelated to social control: “The events pre- cipitating lynchings are only excuses for the conflict which ultimately would have occurred on virtually any pretext” (Henry and Short, 1954:51; see also ‘Tumer, 1964393; compare Sencchal de la Roche, 1997). Another describes it as a process of “black victimization” (Beck and Tolnay, 1990:530, 533, 537). In short, colictive violence is not always what collectively violent people say it is, Sometimes it is the deviant behavior of miserable and misguided people who take out their frustrations either on innocent victims or on alleged of- fenders who would otherwise have been left alone or at least handled more Jeniently. It is @ form of scapegoating (e.g, Downey and Hiyser, 199%; Hodes, 1993; Inverarity, 1976; Raper, 1933; Williamson, 1984:140-222) A PURE SOCIOLOGY OF SOCIAL CONTROL Although the ideological origins of an explanation do nat necessarily undermine its validity (see Dahrendorf, 1968), a generat theory of collective violence requires that we be as consistent as possible in applying concepts and as encompassing as possible in our formulations. Alt behaviorally similar actions should ideally be treated alike, regardtess of how the actors, their targets, or their goals might ideologically attract or repel an observer. ‘The present discussion therefore self-consciously departs from conceptual and ‘explanatory practices that do not treat collective violence as a single subject, matter. Guided by Donald. Black's paradigm of pure sociology, it seeks to ‘contribute to a general theory of eallective violence as social control (see, eg, Black, 1976, 1979, 1995) Pure sociology predicts and explains variation in human behavior with its location and direction in social space (Black, 1976, 1995). A multidimen- sional social space. is defined by the social characteristics of everyone en- gaged in an instance of human conduct. In its vertical dimension (measured conuidered serous crime, does na by isl entiely dotetmine what If snything, wil happen fn is wake For entepies of studies that apply oc extend Black's paradigm to other tyes of scial contrat hejond violence, see Baumgariner (1984), Cooney (1994), Horvitz (1982. 1990), Moral (1982, 1995), Malus (1995), Radelt (1989), Regoli rat (1984), and Tucker (1988). sboule be noted thatthe paradigm is apalcabie not nly to law arc other social cartrol ut © variation in any an al frm of seca fe, rom scence, meccine ana raigion to at, alts, aud sport (Black 1979, $995). ‘The shape of a muleditversional social space as the source of variation teanseends mare familia orts of anaes, W may apply co phenomena lavelengoaly ane or 4 few eicual, fr a thove involving an orgatdzanoe, acighhorhoad, community, region or enti society AAs Black (1998883) States, “Socal space #nesther small nor lange nities & mzeoeant BO? 4 macrocosm, neither o person nor a socket. Ts size is varlable, Ws boundaries ae variate Ihe duration te variable"Collective Violence 10 by social status, including wealth), for example, an action has a higher ele- vation witen the parties involved are all of higher status, and 2 lower cle- vation when they are of lower status. When the parties arc unequal in status, they are vertically distant. An action also may have an upward direction {toward a higher status), downward direction (toward a lower status), or a lateral direction (Coward an equal). The horizontal dimension of social space, includes, for example, relational distance (the degree of intimacy between, people), and the symbolic dimension includes cultural distance (Such @s a difference: in language or religion; Black, 1976:37-59). An action between strangers of different ethnicities thus involves a greater degrce of both re~ ational and cultural distance than an action between intimates of the same. ethnicity, In this sense, every human action has a social structure, ‘The form and quantity of social control that occurs, whether law. self- help, avoidance, mediation, or nothing at all, depends upon the social Lo cation and direction of the conflict: the relative positions of all parties (the offender, aggrieved, and any third parties), such as whether they are equal or unequal, intimates or strangers, and culturally similar or different (Black, 1976, 1984, 1990, 1993, 1995).!9 When social control through collective. vio lence is treated as a dependent variable, the central question becomes: What conflict structures predict and explain this phenomenon?!" Collective violence is often an extreme form of self-help, a species of social control that entails the handling of a grievance by unilaterat aggees- sion (Black, 1990:43-44), Self-help is not always violent, but also includes lesser forms of aggression such as an audience booing a speaker or a group of children ridiculing a playmate (Clayman, 1993). All manner of insults, taunts, criticisms, and protests, as well as vandalism, arson, and other dam- age to property may be instances of collective selt-help.!* Though itis likely By contrast, macrosruciral studies of collective vilence use ene comers, egions, ‘or rations a8 thet nits of anains ard eck to-explain variation (3 rales of violecee with ‘haractersies of those units such a tel ev! of ecorante prosperity ar degree of political ‘or economic competion (eg, Beck and Talnay, 199% Henry an Shae, 1996; Olak, 1989, 4990, 1992; Tily aca, 1978, Toleay and Beck, 1985) ‘A smalét unit of analysc—the eroud—is used in many pre-I96s studies of collective violence. These seek to explain bow and when a crome becomes violent with teferetce 19 the presence or communication of varus seul or interactions benween crave members (eg, Bluner, 1946, Granoverer, 1978 LeBor, 1895/1960, Turner, 1964 ace so Rule, 3988 43-49, 91-108, 116-117, 241-245, 386-200), (9{ eter Reve 10 eotflies aver right and wrong rather than enacts of ineeest, such as competion for power or economic esoutees (compare Olzak, 1992; Till, 978), ‘The concept of the social suenite of a enact for conf trette) fs appears in Black (1988-20-26, "Escuded i collective aggression that isnot social conta, such as collztive predation (eg, raiging loci, ar rosbery by 2 group) ar Feceation (eg, destruction of property ater 4 spore victory or during festivals anc holiday). On aoamoraintsagavesio, see Cutvn and Porter (1995), Cie CONF I8=2, 246-240), and Mare (1972)ry ‘Senechal dela Roche that the formulations introduced in the following pages have broader ap- plications, 1 limit my focus to four forms of unilateral violence: lynching, rioting, vigilantism, and terrorism. Excluded are. not only nonviolent forms of selthelp but also bilateral and reciprocal forms such as brawling, blood feuds, and warfare, along with goveramental violence of all kinds." Re- sponses (0 collective violence, by the state or others, are excluded as well. The formulations presented below order collective violence without regard to the motivational and other psychological factors often emphasized by students of the subject (e.g, Grimshaw, 1972; Gurr, 1970, 1989). In this sense, the approach is purely sociological: How people perceive, interpret, intend, or experience collective violence is beyond its scope. It docs not assume or deny that social conditions frustrate or otherwise predispose peo- ple to aggression, for example, or that people learn the violence they inflict, that violence is goal directed, socially constructive or destructive, or that the participants rationally weigh or calculate its costs and rewards (com- pare, ¢g, Brown, 1969; Gurr, 1979; Smelser, 1963; Tilly, 1978; see also Graham, 1989; Rule, 1988). Neither does it address ritualistic aspects of collective violence of assume or deny that it serves a larger function, suc 4 the clarification of moral boundaties in a community or society (com- pare, ¢, Downey and Hyser, 1990; Inverarty, 1976). Rather, the analysis proceeds ftom an entirely different theoretical perspective that yields highly general and yet readily testable propositions about the likelitood and. se- verity of collective violence in specific cases of conttict. FOUR VARIETIES OF COLLECTIVE VIOLENCE, ‘Semantic confusion surrounds the terms lynching, rioting, vigilantism, and terrorism, but these. forms of collective violence ean readily be defined ‘and distinguished long two dimensions: their breadth of liability and de- agree. of organization, Liability is « condition of accountability for a griev- ance, such as the accountability of 4 burglar for his crime or a motorist, for an accident (see, ¢.g., Black, 1987:566, 573-576; Koch, 1984), Liability Recipracal collective violence has a different socal strztace fom unilateral forms of the phestorenon. Set the section at inequality helo. Tie possible that propositions sila wo those jtoduced here could be used to predict, and expin the doeurence and seventy of police ard military silence agsnst evans (Black, 1290, 1998:127-157}. ‘fhe severiy ox quantiy of voles isles, for example, when 2 threat ennsive of mece words than wien accomparied by the beandithirg af weapons, a beating 1 les volenee ‘an aking, and klirgs theraselves vary in tele severity from a quick dispatch, say with funlie, c prolonged and lethal tortore, and, arnong Filing, hose followed by tastlagan Of the conpse as wellCatlective Violence us to collective violence may be collective (when a group or members of @ group or social category are held accountable for the conduct of an indi- vidual) or individual (when only a wrongdoer is held accountable) (Koch, 1984; Moore, 1972:86-95; see also Black, 1983:38). Collective violence may also be relatively organized, even formally so, of relatively unorganized, In this section I classify lynching as an unorganized form of collective violence ‘with individual fiabitity, rioting as an unorganized form with calfective li- ability, viglantism 5 an organized form. with individual liability, and ter rorism as an organized form with collective liability (see Fig. 1) Lynching and vigilantism may be defined partly by their logic of dividual ability: Only the alleged wrongdoer is accountable, while unin- volved members of the wrongdoer’s group or social category are not subject to punishment or other social control. Like a modern court, lyachers punish only the alleged offender, then disband. And though they may kill or injure many different individuals over a period of time, vigilantes judge each of- fender individually, however brief or superficial the process may be. In contrast to lynching and vigilantism, rioting and terrorism may be defined partly by the presence of a logic of collective liability by which a group or members of an offender's group or social category are held ac ‘countable for the offender's conduct. Those held collectively liable might include, for example, a race, religion, ethnic group, nationality, political party, labor organization, family, clan, or tribe, Literally any member of a social category, including women, childrea, and the elderly, may be vulner- able to attack by rioters or terrorists. The form of collective violence may further be defined by its degree of organization—the capacity for sustained collective action. Lynching and rioting aré distinguished by their relatively low level of organization and Vigilantism and cecrorisen by their high level of organization, Lynching aad rioting are temporary as well a5 informal. Even though collective violence that qualifies as a riot occasionally lasts more than a day and what qualifies as lynching may last as much as 2 week or more, greater degrees of or- ganization would ultimately change the classification of a riot to terrorism or a lynching to vigilantism. Informal organization also implies relatively ‘open and fluid membership: Participation may be possible for nearly eve: ryone in a given community. Even law enforcement officials may join in te violence. In the frequent ethaic riots of South Asia in recent times, for example, the “participation during tiots of police, army, or other security forces... as vigorous participants favoring the cause of ane side or another is 4 fact of life recorded in country after country" (Banerjee, 1990:55-57, See. also Greenberg, 1976:19-26, 50-54; Grimshaw, 1963:271-289; Marx, 1970; Tambiah, 1990:743, 747-748). And in the American South of the fate 19th ‘and early 20th centuries, whole communities of whites—men, women,14 Seneehal de la Roche children, and police—might attend and celebrate the Jynching of an alleged black offender (see, eg, Brundage, 1993:36-45, 64-65; Raper, 1933, Wyalt- Brown, 1982:458). Finally, though riotiag and lynching may include some degree of planning and oxganization, the behavior defined in this analysis as rjoting ang lynching is considerably more situational, spontaneous, and, decentralized than vigilantism or terrorism.’ Unlike Synehers and rioters, vigilantes are sufficiently organized to handle multiple cases over a period of time, and terrorists to launch a series of attacks. Some groups are formally organized, with centralized decision raking, officers, and limited access to membership, while others may or- ganize only for a matter of weeks or even days before they dishand. The degree of organization in vigilante and terrorist groups is itself variable, hhowevec. In the carly American West, for example, the San Francisco Vigi- Jance Committee of 1856 had an executive committer, titled officers, & ‘membership list, clear chain of command, a written constitution, press te- leases, and elaborate tribunals. They remained active for over three months, ‘whereas ranchers or farmers who organized to eliminate livestock thieves typically operated for a week or two (Senkewicz, 1985:170; see also Ban- croft, 1887; Brown, 1975:127-128, 134-143; Johnson, 1981 578-579). Ter- yorist groups such as the pro-Palestinian terrorists of the Middle East or the Trish Republican Army of Northern Ireland during the 20kh century may endure for yours, while others may conduct only a few attacks before disbanding. Figure 1 classifies the four forms of colletive violence. accordingly “The tendency of previous scholars ta focus on oxpaaization alone in thet attemprs °9 distinguish viilansn fom Iyeching has resulted insignificant controversy and contusion. ‘As David A, Johrsan roted in discussing Richard Manvell Beowe's atempls (9 cefine ‘igilantism, “The distincuon berween ‘ergatized vigilance committees ard ephemera’ lynch patles - difcltaf eat ipossbie to make in nase crpiical cases. Many seli-sled Sigiance commitees were unerganized and ephesiral 20d alleged (yeh paris were often aie highly organized” (Johan, 1981'$60). Fohrson thus chooses ¢0 treat vigilance Committees, popular inbunals, people's courts, and Iytchings as ane category, and uses the terms eynonymously. Further, Brown's category of “instant vgiantism” (hese enes whee otent raups lacked office's, Wil, ele, and whose members orzanized and acted Swtly invwone-shat viglante actions") certainly includes seany instances of yechirgssecorcmg {2 the definion used here (see Brown, 1975103; McGrath, 198), Here viglantes are defied as those who deal with more tan one iestance of deviant bpehavio overtime. Lynchers, even if they do tem themselves ¢“viglanee committee” or equator,” are. not viglantes IC thee abjex is to punish a. siagle offense. Thus, “rpontanenus vgilation,” whieh has been roughly defined as the ave igtervention af 2 srOup of sanders who, upon witnessing a cime, purse, capture, and mete out punishment {o the wrongcoer on the spot, i hy our defiton, Insting T nwo individual Ua ity dad indormal organization The bystanders do not argaize to eeahce cesta! bebation #6 the neighborhood at large (Sholand, 1976:30-32),Collective Violence 10s uaeiLiry Inaviguat Calective Low lynching fiatng ‘ORGANIZATION igh Viglantisn terri Four ors of callectve violence ‘THE SOCIAL STRUCTURE OF COLLECTIVE VIOLENCE Collective violence sometimes arises where Jaw is lacking, weak, or openly partisan, a pattern consistent with existing theory suggesting that Where Jaw is Weak or absent, other means of social control tend to arise (Black, 1976:6-7, 105-111). Yet collective violence also occurs where legal agencies are well established and may even compete with law, a phenome- non seen when suspects are forcibly abducted trom police or jails and pun- ished by private citizens. We therefore need to explain why collective violence rather than some other form of social control (including law) ac- curs in response to 4 grievance. Black identifies five general responses to deviant behavior: self-help, avoidance, negotiation, scitlement through a third party, and toleration (1990:43-69). Each has a distinctive social struc- ture comprised of the social characteristics of all of the pasties associated with a conflict—offenders, complainants, partisans, and anyone else with knowledge of it! What conflict structure, then, is conducive to collective violence in general? And what conflict structures are conducive to each of the four forms—lynching, rioting, vigilantism, and terrorism? First consider collective violence in general. Sar the five major forms of social control svailae in conflicts and the srscsural ‘characterises conducive t0 each, se Black (1990) See also Horwitz (1990),106 Senechal de ba Roce As a generic phenomenon, unilateral collective. violence arises with specifiable combinations of the following variables: (1) relational distance, Q) cultural distance, (3) fnctional independence, and (4) inequalig:” Bach variable differs in degree from one conflict to the nest, and the probability and severity of collective violence varies accordingly. In particular, where relational and cultural distance, functional independence, and inequality are greatest, [ predict the greatest likelihood and severity of collective vio- lence: Relational Distance Relational distance, or intimacy, is “the degree to which [people] par~ ticipate in one another's lives,” measurable by such variables as the number of ties between people, the frequency and duration of contact betwen them, and the age and nature of their relationship (Black, 1976:40-48), The greater the relational distance between the parties to a conflict, the greater is the likelihood and severity of collective violence: Collective valence varies directly with relational distance'® Groups handle unknown offenders more harshly than those with whom they are acquainted. Regnlarized contact in a workplace, organiza- tion, or neighborhood increases intimacy and makes collective violence tess likely (Gluckman, 1956:19-24; Lewis, 1990:98; Senechal de la Roche, 1990:141=15]; Weishurd, 1989:J24-126). Moreover, such intimacy moder ales the severity of collective violence witen it does occur, Food riots swept Japan in 1918, for example, and their severity varied directly with the de~ agree of relational distance between the parties involved. Whereas cities with “an absence of face-to-face relationships between rioters and the targets of their attacks” experienced extensive looting, property damage, and vio- ent clashes with tie police, in rural areas and villages where the parties hhad closer ties, rioters were said to be “almost mannerly” (Lewis, 1990:98). Similarly, in the Miami riot of 1980 and the Los Angeles riot of 1992, busi- nesses awned by outsiders were more likely to suffer total destruction than those owned by people with more ties to the neighborhoods that experi: enced the violence (Porter and Dunn, 1984-131; Tierney, 1994:151-152). Likewise, in the postbellum American South, lynching af a white or a black was more likely t0 occur and more likely to be lethal when the alleged me variables described here are drawn fom Black's theovetical werk, pariulatly is anasis of selchelpdiscpline, rebellion, and vengeance (1990448), "efor each propastin i is assumed tht all else te other vacanles specified here andthe, type of offense i held constant,Caltective Violence w offender was a stranger to the community (Brundage, 1993:81-82, 90-91; Senechal de 1a Roche, 1997), Changes in communities that increase relational distance between ‘groups and individual also increase the probability of collective violence. In the American South in the late 19th centusy, for instance, contact between ‘blacks and whites gradually lessened over time. Intimate ties associated with slavery diminished as former masters and slaves died off and as blacks and ‘whites increasingly lived and worked in different settings: “A new generation ‘of blacks and whites faced each other across an ever-widening chasm. The ‘best’ whites and blacks seldom had contact with one another, as both races increasingly withdrew into their own neighborhoods and churches” (Ayers, 1984:182, 236-238, 241; sce also McMillen, 1989:6, 11, 23-24). As the degree of relational distance between blacks and whites grew, the ratc of lynching increased.” By contrast, in Brazil, where blacks and whites have long inter- mingled and intermarried, interracial collective violence has been virtually nonexistent (sce, e.g., Degler, 1971:95-96, 254-255). Population movements such a8 in-migration and desegregation mul- tiply the number of strangers in a community. Conffict between people separated by considerable relational distance therefore becomes more likely, and this in tuen increases the likelihood of collective violence. It follows, for example, that the rate of lynching in the American South should have been higher in areas with higher rates of in-migration by blacks. And it was (Ayers, 1992:156-157, 496-407, Brundage, 1993:143-149; McMillen, 1989:259-261; Senectal de la Roche, 1997).# One historian thus notes that Iynchings “tended to flourish where whites were surrounded with what they called ‘strange niggers,’ blacks with no white to vouch for them, blacks with no reputation in the neighborhood, blacks without even other blacks to aid them” (Ayers, 1992:157; see also Brundage, 1993:81-84, 90-91). A heavy in-migration of blacks also led to a brief but dramatic increase in the rate ‘of interracial rioting in northern American cities during and immediately after World War 1 (eg, Rudwick, 1964; Tuttle, 1970), Because the deseg- regation of a city or neighborhood similarly increases the number of strang- ‘es, American cities with greater desegregation of racial or ethnic groups should have had higher rates of collective violence as well. And they did (Olzak ee at, 1993; see also Glzak and Shanaban, 1994). "Foe vaiaon in th ate of ting i the Sou, e, oe ample, Tay and Beck 8S: ap. 2) "Bymile the average county in the South berween 1880 sed 1910 saw is bisck popelaion increase by 48% counties in the Gulf Coat subceyionsteiching tom Florida to Texas sav back poptlatiog mcreases ranging from approximately %-130%. Jt was this Oxlf Coast subregion that had the highest (ching tate over this period. The ane exception ws LLouisana, which received 2 far smaller share of black migrans and had oocrespondimety ‘ot Tyechieg (Ayers, 1992156).108 Senechal dela Roche Riots virtually always begin with. a conflict between relationally distant parties, and those who are strangers aod outsiders to the aggrieved group are usually the most vulnerable to attack. In Sri Lanka in 1983, for instance, Sinhalese rioters tended to be stcangers to thei Tamil victims, while Sinhalese neighbors of Tamils—those relationally loser —cither remained neutral or of fered protection to potential targets of thei fellow Sinhalese (Kanapathipillat, 1990:321-344), Likewise, when I8th-century American cities experienced oc- «assional collective attacks on houses of prostitution (usually after a customer ‘was robbed or killed}, mast of the participants came from distant neighbor- hoods: “When New Yorkers rioted against bawdyhouses,” one historian notes, “it was anything but a local affaic” (Gill, 1987:90). The frequent involvement ‘of sailors in 170h- and 18th-century English and American urban riots is simi- larly understandable in part with their relational distance: from their targets “Seamen were 1 usally natives of the city where the tiot developed. They were outsiders” (Rediker, 1987:249). In the Springfield, Illinois, race riot of 1908 (precipitated by the alleged rape of a white townswoman by a black stranger), the vast majority of white rioters worked in all-white occupations, lived in all-white neighborhoods, and were otherwise relationally distant from blacks. Those who worked with or lived mear blacks generally stayed at home during the violence (Senechal de la Roche, 1990:141-151). A similar pattern, prevailed in the Detroit race rit of 1943 (Capeci and Wilkerson, 1991:105-107, 108; Lee and Humphrey, 1943:17, 28, 130-133, 140) ‘Those who live in intensely intimate settings such as hunter-gatherer bands or tribal villages rarely engaged in colletive violence against theit fellows (see, eg., Gluckiman, 1965, Roberts, 1979). Strangers in large num- bers occur relatively late in human history, and 0 does collective violence. It arises historically with the growth of tawas and cities where people en- counter others across greater distances in relational space. It arises with increasing geographic mobility, as migration and improving transportation ireulate stangers within and across nations. It arises with segregation, which reduces or severs ties between co-tesident populations. I¢ arises with in-migration and desegregation, which suddenly infuse strangers into com- munities and neighborhoods, Strangers may come in peace as migrants or refugees or with colonial regimes or as slaves purchased in distant lands Regardless of their mission, however, they bring more collective violence. Cultural Distance Cultural distance is measurable by differences berween individuals and ‘aroups in the expressive aspects of their social life, such as language, dress, religion, and art. As these differences increase, 50 do the probability andCatlective Violence 109 severity of all forms of collective violence, whether lynching, rioting, vigilan Aism, or terrorism: Collective violence varies directly with cultural distance. In tribal societies where people share a single pattern of culls, col- fective violence is rare or absent. Cities with ethnic and religious enclaves have more, In South Asia, with its great religious and ethnic diversity, for instance, conflicts over stch matters as straying livestock, the playing of re- ligious musie in public, elapements, and traffic mishaps may explode into ri ting when the parties are culturally distant (See, ¢.8,, Banerjee, 1990; ‘Hussain, 1990; Kannangara, 1984; Roberts, 1990; Roy, 1994). An extreme ‘case was the interethnic rioting in India that followed the 1984 assassination of Prime Minister Indira Ghandi, a Hindu, by two of her Sikh guards. Hindus in New Delhi responded with several days of mass violence against Sikhs, looting, burning, and killing over 3000, including women and children (Tam- biiah, 1990;744-748). Yet Sikhs, Hindus, or Muslims among themselves virtually never use such violence against their fellows. In the 19th-century American and Canadian West, whites sometimes used collective violence against the Chinese Gee, eg, Chen, 1980:47, 89-94, 137-141; Quinn, 1967:82-80; Torrance, 1986:30- 31), Typically the targets were temporary immigrants, soon to return to China, ‘but permanent immigrants were culturally removed by traditional customs, ‘cuisine, religion, and dress as well, and they too were vulnerable. Grievances against pariah groups (culturally distant, nonindigenous minority groups) are often handled with special severity. What might seem to an outside observer trivial offenses or unfounded rumor might prompt ‘extreme amounts of bloodletting. A public display of alicn culture by itself, such as a minority religious or funeral procession, may trigger rioting. Re= ligious holidays and processions, for cxample, have long been associated ‘with rioting in South Asia between Hindus, Muslims, and those of other faiths (c.g, Roberts, 1990). Similarly, in the late 1960s, a Chinese political celebration in Malaysia sparked anti-Chinese riots that left hundreds dead and many homes and businesscs looted and burned (Von der Mehden, 1976:223-226). In late 19th- and early 20th-century Russia, a rumor of ‘wrongdoing by a Jew was sometimes sufficient to precipitate a casc of anti- Jewish rioting by gentiles known as a “pogrom” (Greenberg, 1976:50-54;, Weinberg, 1993-15-20, 164-187). Pogrom-style violence, however, was nat employed against fellow Christians. In 16th-century France, a period of re- ligious ferment, public religious events by Catholics and Protestants often prompted collective attacks (Davis, 1975:170-173). And in New York City in the early 18005, Irish Catholic religious holidays witnessed collective at- tacks by Protestants who sometimes paraded with effigies mocking Catholic saints. To reduce rioting, the city ultimately passed an ordinance to ban all effigy processions (Gilje, 1987:128-135), In contrast, grievances among those of the same faith were virtually never handled with collective violence.M0 Sonechal de la Roche When people are both relationally and culturally distant, the potential for extreme violence is all the greater. In Nigeria, for example, a seemingly trivial incident led (0 4 lynching when & member of an ethnic minority ac cidencally brushed against another bus passenger and was angrily accused of inflicting a curse to render him sexually impotent, The rest of the pas- sengers then beat the culturally distant stranger to death (Mazrui, 1976213). In Uganda, culturally and relationally distant automobile drivers involved in auto accidents may be beaten or killed on the spot by angry local citizens (Maztui, 1976:208-211). Among fellow tribesmen, however, Jethal lyaching is rate (see, ¢.2, Edgerton, 1972). In the late 19th-ceatury American South, a high degree of both cultural and relational distance was present in some of the relatively rare cases of white-on-white lynching. In Georgia, Alabama, Kentucky, and elsewhere, for instance, Mormon mis- sionaries—alien not just because of their beliefs but because of their largely northern background and manners—sometines met with beatings, whip- pings, and killings. Itinerant ministers of the South’s two dominant faits, Baptist and Methodist, by contrast, virtually never risked lynching (Ayers, Sessions, 1976:212-225). In the late 20th century, American levels of ethnic and religious diversity experienced more collective violence than mare culturally homogeneous settings. Korean busi- nesses were targeted for destruction by rioting blacks and Hispanies in Los ‘Angeles in 1992, for example (Freer, 1994; Tierney, 1994), And the acci- dental killing of a black child by an Orthodox Jewistt motorist in Brookya it 1991 prompted several days of attacks against Jewish citizens by blacks, while arrests and killings by culturally distant police in the 1980s and 1990s sparked rioting by both blacks and Hispanics in cities such as Miami and Washington, DC (Levin and MeDevitt, 1993:137-148, 235-257; Porter and Dunn, 1984). Grievances against shopkecpers, crrant motorists, and police of the same culture, however, rarely result in collective violence, When collective violence docs occur between parties who are cultur- ally close, it tends to be less severe. Thus, in 18th-century France and Eng- land, foad riots against culturally similar millers, bakers, and other vendors who allegedly boarded scarce food and raised prices during shortages were relatively mild, were normally limited to the scizure and distribution of food or its sale at a “fair” price, and only rarcly resulted in beatings or killings (Bereé, 1990:170-179; Rude, 1964:19-45). In 1918, a wave of food riots in culturally homogeneous Japan was similarly restrained (Lewis, 1990). Ri ting in American cities also rarely resulted in severe injuries or killings ‘until theit cultural homogeneity declined with the appearance of new ethnic and retigious groups in the early nineteenth century (see, e.g., Gilie, 1987), Vigilantism tends to be less severe between culturally clase. parties as well, When worker vigilantism during labor conflicts in 18th- and 19th-cen-Catlecive Violence m tury England and America was aimed at culturally similar targets, for ex- ample, it was relatively mild. Weavers, shoemakers, hosiers, and ather crafts- ‘men verbally harassed and threatened strikebreaking workers and ‘occasionally destroyed their working materials, but when the strikebreakers were of an ethnic group different from the vigilantes, beatings were more common (Gilje, 1987:175-176, 198-199; Laurie, 1980:125-126, 157-158; ‘Thompson, 1975:270-273), In the American West, alleged offenses by Chi- nese or Hispanics against whites were more likely (0 result in collective vio fence and entail more severe violence than cases that crossed less cultural distance, between white setters. Thus, from the 1840s to the 1900s, “non- whites, primarily Hispanics, composed a dramatically disproportionate per- centage of those who experienced the wrath of the ‘people,’ [and] the victims ‘of crimes for which Iynch courts sought retribution were overwhelmingly white" (Johnson, 1981:570-575). Vigilantes also dealt out harsher punish ments—typically severe beatings, mutilation, and execution—to culturally alien offenders. In early California, for instance, “mining camp punishments fell with special severity on Mexicans, Chileans, and Chinese” (Caughey, 1957-228). In the 1980s, Arab youths who threw rocks at Israeli settlers’ ve~ hicles might attract collective punishments against their entire community ‘Though Jewish settlers sometimes tried to locate and beat only the offending individuals, more often they held Arab communities collectively liable, as ‘when bands of settlers closed down or damaged Arab schools and businesses and destroyed Arab property or assaulted Arab residents indiscriminately (Shipler, 1986:124-129; see also Rigby, 1991:67-70, 122; Shalev, 1991:123~ 126; Weisburd, 1989:54-85), Among the Jewish settlers themselves, however, violence involving logic of collective liability & tare, if not entirely absent. Independence A thitd variable associated with the likelihood and severity of collec- tive violence is functional interdependence—the extent to which individuals and groups cooperate with one another economically, politically, militarily, ‘or otherwise: Collective violence varies direcly with functional independence. People are comparatively unlikely to attack those who are indispensable to their well-being, but prefer to kill or maim those they can do without. For example, extensive trade ties or alliances reduce the. likelinood of collective violence between groups involved in conflict (Black, 1990:45-47). Wien col- lective violence does accut, functional interdependence tends to reduce the severity of violence involved. In traditional India, for instance, “untouchables” ‘or others in fowly occupations of lineages sometimes offended high-caste vil- lagers by adopting visible emblems of higher-caste status. When the trans-ry Senechat de la Roche atessors provided services indispensable to the comunity, they were merely shunned or ignored, but when few or none depended on their services, they zmight be beaten or killed (Mandelbaum, 1970:378-379, 458-459, 474, 477). During European food riots of the 18th century, crowds might seize and re distribute bakers’ and millers’ wares, but seldam eliminated the offenders whose services benefitted the community (Bereé, 1990:170-179; Rudé, 1964:19-45). More extreme forms of violence were reserved for those from. whom they were more independent, such as tax collectors and soldiers (Bercé, 1990), The likelihood and severity of lynching in the American South also depended in part on the degree of interdependence between an alleged black offender and aggrieved whites. Black vagrants and transicnt workers were usually completely independent of any whites they might offend, and they were more vulnerable to lethal Iyachings, But a black who offended his white employer—a situation of greater interdependence —would generally receive nothing more than a beating. An employer also might intervene. to protect, 2 black employee from punishment by other aggrieved whites (Brundage, 199345, 63-64, 84; Senechal de la Roche, 1997). Variation in the degree of functional independence between workers and employers also helps explain patterns of collective violence in indus- trial societies. In the early stages of industrialization, independent skilled craftsmen sometimes attacked the new factories and machines that they regarded as @ threat to their income and autonomy. In London in the 1760s, for example, silk weavers destroyed newly introduced engine looms. Later, in the English Midlands Luddite outbreaks of 1811-1816, hand- weavers wrecked steam looms and burned factories (Rudé, 1964:66-91).. By the 19th century, however, workers rarely destroyed the facilities upon which they largely depended for their livelihood. Instead, most violence during labor conflicts shifted to attacks against those from whom workers were entirely independent: strikcbreakers and company guards (sce, ¢., Dubofsky, 1985:33-55; Taft and Ross, 1969:281-395), For their part, em- ployers in the 19th century lost their dependence on particular workers as the need for skilled employees declined and as an oversupply of labor made most individuals ¢asily replaceable. And as they became increasingly independent of their workers, employers proved more willing to encourage violence against strikers from their own factories (sce, cg, Taft and Ross, 1969:281-395). Inequality Inequality of status, measured by wealth and other variables, is also associated with the likelihood of lynching, vigilantism, rioting, and ter-Calletive Violence 1 rorism. The more inequality between parties in conflict, the more these forms of sacial control occur: Collective violence varies directly with in- equality Recall that the present analysis pertains only to unilateral forms of collective violence. Inequality, moreover, is the major social characteristic that differentiates conflict structures associated with unilateral collective violence from those associated with bilateral (or reciprocal) forms such as feuding and warfare. Bilateral collective violence may—like unilateral violence—have 4 conflict structure with a high degree of relational and cultural distance as well as functional independence (Black, 1990:44-45; 1995:855, n. 130), but unilateral violence is not likely to arise where the partis are equal in size and resources. As the vertical distance between antagonists increases, unilateral collective violence becomes more likely. Extralegal beatings and executions, whether inflicted by lynchers or vigilantes, are usually reserved for poor and subordinate offenders. Riots, however, may be aimed either dowawardly at inferiors or upwardly against those of higher status, American anti-black and anti-Chinese riots illustrate. the former (see, eg., Chen, 1980;89-94; Rudwick, 1964; Senechal de la Roche, 1990), whereas some politcal riots, tax riots, worker violence, and American ghetto riots are examples of the latter (see, eg, Kerner ef al, 1968; Tilly, 1969:5-7, 16-21, 26-27; Tily et al, 1975). Terrorism is usually upwardly directed, although sometimes poorer members of dominant groups ‘may be included as targets Lynchings are more likely to be physically violent and lethal when the offender is poor or 2 member of an inferior ethnic group, race, oF caste. The alleged rape or attempted rape of a respectable white woman by a black in the late 19th-century American South, for example, was more likely to provoke. a capital lynching than the same offense committed by a white (ee, eg. Brundage, 1993), But a black’s alleged rape of a low-status white woman regarded as promiscuous ot otherwise unrespectable—a case between relative equals—was more likely to go unavenged (Brundage, 1993:63-64; Hodes, 1993:411; see also Senechal de la Roche, 1997; Wright, 1990:55), On the other hand, the most violent lynchings followed the mur der of a white police officer by a black, a dramatically upward offense against the polity itself. In these cases, torture as well as execution of the offender often resulted (Brundage, 1993:73, 75-77, 79; see also Dowaey and Hyser, 1990; Horwitz, 1990:108-113). Black-on-black and white-on- white assaults or killings, which typically involved parties of equal status, rarely resulted in death or torture for the alleged offender (Ayers, 1984:231;, Senechal de la Roche, 1997). In the relatively few cases where whites lynched other whites for assaults or killings, the offender commonly was14 Sencha de ta Roche of very low status (Brundage, 1993:86-102; Senechal de la Roche, 1997). In traditional India, attempts to rise in caste were more likely to provoke severe responses when aspirants were “untouchables” or others nearest the bottom of the social hierarchy (Mandelbaum, 1970:378, 474, 477) Vigilantism, rioting, and terrorism vary with inequality as well. Up- ward offenses by subordinate individuals or groups may spark rioting, as, ‘was often the case with American race riots before World War IT (see, 4g, Quinn, 1967; Rudwick, 1964; Senechal de la Roche, 1990), Actions, by the police in handling individuals or crowds have occasioned extensive rioting in American black neighborhoods in the 1960s and during work ers’ protests and street demonstrations in Europe (Feagin and Haha, 1973:144-146; Kerner er al, 1968; Marx, 1970; Tilly ef at, 1975; Tilly, 172-177). Moreover, when vigilantes are drawn fom wealthier seg- ments of communities in response to offenses by social inferiors, the se~ verity of penalties increases. An unusually lethal ease in Montana in the 1880s, for example, involved large ranchers against diverse inferiors: “a ‘motley assortment of unemployed whiskey traders, wolf hunters, woodchop- pers, and trappers” (Malone and Roeder, 1976:122). While conflicts among large ranchers over ownership of land and cattle in rarc cases led to feuding, they did not result én vigilantism, Homeless children who beg or steal in commercial districts of Brazilian cities run the risk of being Killed by “death squads," some of which may be run by off-duty police officers (Rosenbaum and Sederberg, 1976:10), but otherwise respectable shoplifters run no such risk, Modem terrorists often have high-status, powerful adversaries, includ- ing governments, and the severity of their violence reflects the degree of inequality involved. Terforists in Norther Ireland or the Middle East, for example, often count the killing of high-status members of targeted social categories (British or Israeli troops and goverament officials, for example) as a special achievement, Indiscriminate attacks are also made against politically and economically dominant populations, as when Arab terrorists bomb offices, transportation facitties, and crowded public places in Israel |AS the degree of inequality between the partics in 2 conflict less- ens, so does the amount of violence. When the Ku Klux Klan of the Reconstruction South operated as vigilantes, they tended to mete out different punishments to blacks and whites who supported the rising Re- Leas severe Iynhings without sever Inury or execution oceur in confi where offenders 2nd the apgieved ace tlalvely close a sates An example 6 the chavary 2 tlalized public punishment of devians by members of small communities widespread in Europe fram Breient times tha aso occured in rural America, especialy in the South, through the 20ch century (Brown, 197525, 190-181; Davis, 199597-933, Liat, 196889, Wyar-Biown, 1sa2435-451),Cotleetive Violence us publican Party. Poor blacks received the harshest treatment —often death Gee, eg., Trelease, 1971; Wade, 1987; Wright, 1990).’The second Ku Klux Klan of the 1920s often served as vigilantes who punished moral offenses such a8 fornication and drunkenness by low-status whites with tarring and feathesing and whipping, while wealthier whites—those closer in status to Klansmen—were rarely handled in this fashion for similar offenses (Wyatt- Brown, 1982:449-450). In episodes of vigilantism in the 19th-century ‘American West, when offenders were poor and avengers wealthy, punish- rent was most likely to be lethal (see, e.g, Brown, 1991:87-127, Malone, and Roeder, 1976:122-123)? But what kind of collective violence is likely to occur? ‘THE DISTRIBUTION OF COLLECTIVE VIOLENCE ‘Whereas all four forms of collective violence—Iynching, rioting, visi- Jantism, and terrorism—are more likely to appear in cases of conflict with the social characteristics discussed ahove, the specific form occurring de- pends on at least two additional variables: (1) the degree of social polari- zation and (2) the continuity of the deviant behavior at which the violence is directed. Social polarization refers to the combined degree of relational distance, cultural distance, inequality, and independence. present in a case of conflict, whereas the continuity of deviant behavior is measured by its frequency and duration! Social polarization and the continuity of deviant behavior predict the breadth of liability (collective or individual) and the dearce of organization (high or tow) of social control and therefore predict—in eombination—the form of collective violence likely to occur in a particular confit. Collective liability (expressed in rioting and terrorism) oecurs where social polarization is greater, whereas individual liability (expressed in lynching and vigilan- tism) appears where it is lower. A high degree of organization (seen in vigilantism and terrorism) appears when alleged offenses are chronic or ‘enduring, whereas more informal organization (seen in lynching and riot ing) typically arises when they ate infrequent or transitory. rye mode presented Reve sssimes thatthe confi suture is alo chaccteried by iemobly, That bs shes hare a eing and ace aczeble wo ove athe. The dqiee Of immobity sfect the iteinood of the we of seltelp 40 soil contol (ack, Sana). iq her sty of the banding of American housng-code violations, Mleski (1971) disinguhes between continooss deviate and those that ae dctete vents tka ed ina retsveyy short tre uch sb murder, tape, and robber}a6 Senechal de la Roche Polarization and Collective Liability Social polarization, which refers to the. degree of relational and cul- tural distance, inequality, and functional independence, predicts whether collective liability is likely to appear. As social distance of all kinds in- creases, so does the breadth of liability. More polarized conflicts have a greater likelihood of collective lability: Collective liability to collective vio ence varies directly with social polarization. Thus, where those in conflict are relationally and culturally extremely distant, have little or no interde- pendence, and differ sharply in status, forms of collective violence with a logic of collective liability (riots and terrorism) are more likely than those with a logic of individual liability (lynchings and vigiiantism), ‘Consider, for example, pariais, These nonindigenous minorities typi- cally live in highly polarized social settings: Differences in dress, dict, and religion, along with limited macital and business tics to the indigenous com- ‘munity imply a high degree of relational and cultural distance, and the level of functional independence frequently is high as well. In conflicts with the majority population, therefore, parias are especially vulnerable to col- lective liability. In Southeast Asia and Africa, for example, Chinese and Indians were important economic fixtures in their host societies, but like the Jews of Europe. in earlier times, they were also often held collectively able during conflicts and have experienced riots for decades (Von det Mehden, 1976:218-233). Native groups are far tess likely to handle. conflicts, among themselves in this fashion. More generally, in the growing cities of South Asia, with their distinct mosaics of ethnic, racial, tribal, and religious gtoups, polarized conflicts with a logic of collective liability ate endemic. ‘The sins of the few may damn the many with lightning speed. In Karachi, Pakistan, for exampte, extensive interethnic rioting in 1985 was precipitated by the death of a schoolgirl struck by an errant bus driven by a man of differing ethnicity (lambiah, 1990:748-749). Similar mishaps between mem- bets of the same ctimnic group who know one another rarcly, if ever, lead to such actions. Terrorism and rioting between Arabs and Jews in Isracl and the occupied Arab territories express enormous polarization 3s well (see, €.g, Shipler, 1986:79-137, 181-221; Weisburd, 1989.64-85). But ter- rorism and rioting aimed at fellow members of Arab or Jewish communities is virtually unknown, When collective violence docs occur within these com- munities, it follows a logic of individual liability. During the Palestinian In- tifada (the uprising that began in 1987 against Israeli occupation. of the West Bank and Gaza areas), for example, Arab vigilantes beat or executed those suspected of collaboration with the Israeli authorities, but typically uci otters a ‘ropesiton:“Lisbilty varies siectly with social distares™ (198775),Collective Violence uw spared kin or others close to the alleged offender (Hunter, 1991:113-115,, 202-205; Rigby, 1991:43-45, 156; Shalev, 1991:70-72), They sometimes evacuated relatives from the alleged collaborator’s home before attacking (Hunter, 1991:114). Proximity and Individual Liability Individual lability is more likely in less polarized conflicts. Hence, a degree of social proximity—intimacy, shared culture, interdependence, and equality between the parties—renders terrorism and rioting less likely and lynching and vigilantism more so. Consider, for example, collective violence in frontier settings. From the 15th century onward, Eutopean colonization in the Americas, Africa, and elsewhere created highly polarized settings of newcomers and indige- nous populations. Conflict between the nto sometimes ended with mass expulsion or virtual extermination of native groups, Functional interdepend- ence between settlers and natives, however, significantly ceduced the prob ability of cotlective liability during disputes. Where trade ties entailed ‘mutual dependence, as in the French Canadian fur trade, collective liability was far less likely than in the British colonics to the south, where trade and the use of Indians 4s laborers were tess developed (Blalock, 1967:76— 79; Nash, 1974; see also Fredrickson, 1981:37-38, 47). Fronch settlers ia 17th- and 1Stb-century Canada, who were mostly young, single males, also depended on Indians for a supply of wives—another form of interdepend- ence that reduced the collective liability of the Indians (Nash, 1974). During the late 19th-century gold mining of the Canadian Yukon Territory, “a ‘man’s relations with the Indians could be as consequential for his well-being and survival as his relations with his fellow miners” (Stone, 1988:86). In- dians who stole from or murdered whites were typically treated with a logic ‘of individual liability (Stone, 1988:95-100). Retaliatory violence did not spread to an Indian offender's family, village, or tribe. But in aocthwestera, California, where some tribes kept aloof from settlers and were. unimpar- tant to whites, miners migit hold an entire Indian settlement collectively liable, The miners might even destroy the offender's entire village and its inhabitants. If his home village was unknown, they might level the Indian village nearest the site of the offense (Hurtado, 1988:117-122). Tin the roral American South in the late 19th and early 20th centuries, lynching was, s0 to speak, a sign of closeness between blacks and whites. ‘Though the races drifted apart during this period of increasing segregation, they remained sufficiently intimate, culturally close, and interdependent tat Iynching, and not rioting or terrorism, was the favored form of collective vio-a8 Senechal de la Roche lence against blacks who allegedly offended whites (see, e.g, Ayers, 1992:157- 158; Brundage, 1993). Rarely did whites punish or injure blacks who were. merely associated with the allegedly guilty party, In the urban settings of the American South and North, however, blacks lacked intimacy and interde- pendence with whites and therefore were more likely to experience indis- criminate mass attacks—riots—by whites (see, eg, Gilje, 1987; Rudwick, 1964; Senechal de la Roche, 1990, 1997). Social proximity similarly under rmined collective liability n the 1918 Japanese food riots mentioned eattier. Some so-called food rioters in rural areas and villages acted as vigilantes: Highly or ‘aanized protesters carefully singled out for abuse only those local merchants be- lieved responsible for inflated food prices. Aggression against these well-known food vendors invalved minor property damage at most. In large cities, how- ‘ever, where the degree of intimacy and interdependence between the parties was far less, collective liability—rioting—prevailed (Lewis, 1990). Chronic Deviance and Organized Violence ‘The continuity of deviant behavior is low where offenses are infre- quent and momentary and high where offenses are frequent or chronic. Examples of the former arc isolated acts of theft, vandalism, or murder. The latter would include so-called epidemics of crime or rowdyism as well as single deviant acts that endure over time, such as the maintenance of a house of prostitution, a landlord's violation of housing standards, or an in- dustry’s pollution (Mileski, 1971:10). colonial regime similarly creates an enduring grievance. among, the conquered by its mere presence. ‘The organization of collective violence varies direetly with the continuity of deviant behavior. Vigiamtism, for example, generally arises with multiple instances of deviant behavior, such as repeated livestock thefts, robberies, and violations of moral standards. Vigilantes are “repeat players” (rather than “one-shotiers”) who deat with a number of offenders over time. Even so, vigilante groups tend to be short-lived compared to terrorist groups, some of which may last for many years, a in the cases of the Irish Republican Army or the Palestine Liberation Organization. The United States produced an estimated 300-500 vigilante groups from the 18th to the carly 20th centuries, but few lasted as much as a year (Little and Shef- field, 1983:803-804). One short-lived but effective vigilante group, nick- ‘tamed Stuarts Stranglers after its leader, organized in Montana in 1884 after ranchers suffered prolonged and severe livestock losses, In less than 2 month, the group hunted down and executed as many as 60 alleged out- Maalanter (1974) uses the terms “repeat players” and “one-shoters” to classify wsees ofCotective Violence a9 laws, some of whom were seized from the authorities, When cattle theft disappeared, Stuart's Stranglers disbanded (Malone and Roeder, 1976:122- 123). In Uganda, shortly after independence in 1962, ardent nationalists organized as vigilantes to punish whites and Asians who did not support the new nation. They kidnapped a European newspaper editor, for exam- ple, because he hosted “a racialistic cocktail party” at which he and other white guests allegedly lamented “the demise of the empire” (Mazrui, 1976:198). When whites and Asians became more circumspect, the vigilan- (es disappeared. In East Africa, young vigilantes supporting “decency in dress” roughed up and humiliated young women who wore miniskirts, wigs, ‘or tight trousers in public during the late 19605 (Mazrui, 1976:198). Single incidents provoked no such systematic forms of social control. Whereas vigilantes withdraw when predators or other deviants disap- pear or modify their behavior, terrorism normally arises with intractable ‘affenses of longer standing, Like vigilantes, terrorists are repeat players, bbut they commonly have only a single grievance, be it social, political, or economic. The condition they define as unjust or immoral cannot be reme- died with the punishment of one. or several deviant individuals or with a single attack (as in a riot or lynching), for itis nearly always a large and powerful group that is defined as blameworthy in a continuing fashion, In- vasions, military dictatorships, and other patterns of domination provide fertile conditions for terrorism. In the West Bank of Israel, for instance, years of anti-Israeli terrorism followed wiat Palestinians defined as an ua- just invasion and dispossession by European Jews. Another example was the terrorism by the Ku Klux Klan and kindred groups in the American South immediately follawing the Civil War: Confronted with a massive change—impased by the North—in social relations with their former slaves, some whites turned to collective violence, targeting both blacks and whites who supported the new order. The Klan repeatedly engaged in killing, beat- ing, property destruction, and occasional wholesale expulsion of blacks from their homes (see, ¢.,, Wade, 1987:19-77; Wright, 1990). Such measures are unlikely when deviant behavior is more transitory and isolated ‘Transitory Deviance and Unorganized Violence Unlike vigilantes and terrorists, rioters and Iynchers are “one-shotters” rather than “repeat players” (Galantet, 1974): A single incident typically precipitates a single instance of collective retaliation or punishment. Here again the organization of collective violence reflects the continuity of the behavior: Rioters and lynchers are less organized than vigilantes and terror- ists. In the late 19th-century American South, for example, where an alleged120 Seoechal de Ia Rocke lalling of a white by a black preceded many Iynchings (Brundage, 1993:72; Tolnay and Reck, 1995), some southern communities could rapidly raise a white crowd to avenge one of its residents, Yet once the alleged offender was dead, the Iynchers neatly always dispersed. On rare occasions, such as when a poor black allegedly killed a higher-status white or a police officer, the crowd might not only Iynch the alleged killer but also go on to attack the black community at large—a case of lynching evolving into rioting, Even then, however, the lynchers-tumed-riotcrs organized only informally and temporarily. Since southern whites seldom faced continuous deviance. by blacks, they seldom organized in the manner of vigilantes or terrorists. In short, cach of the four forms of collective violence varies with both, the degree of social polarization and the continuity of deviant behavior: Lynching varies inversely with social polarization and the continuity of deviant behavior. —Rioting varies directly with social polarization and inversely with the continuity of deviant behavior. —Vigilantism varies inversely with social polarization and directly vwith the continuity of deviant behaviay. Terrorism varies directly with social polarization and the continuity of deviant behavior. ‘Thus, lynching is most likely in less polarized cases of conflict and when the grievance involves a single offense, Rioting is also most likely when a single incident is involved, but where social polarization is greater. Vigilan- ‘ism is most likely in conflicts with less polarization but continuous offenses. Finally, terrorism is most likely in. polarized conflicts where the grievance endures. Figure 2 illustrates these patterns: CONCLUSION Most collective violence is social control: a process that defines and responds (0 conduct as deviant. It is self-help by a group and, when uni- lateral, has four major forms: lynching, rioting, vigilantism, and tertorism. Following Black's paradigm, [seek to develop a pure sociology of collective Violence. My central assumption is that collective violence varies with the social structure of conflict. I offer ten propositions that predict and explain variation in the likelihood, severity, aad form of collective violence. The following propositions address collective vialence in general: 1. Collective violence varies directly with relational distance, 2. Collective violence varies directly wit cultura distanceCallective Violence wa 3. Collective violence varies directly with functional independence. 4. Collective violence varies directly with inequality. sociaL POLARIZATION High Low CONTINUITY OF DEVIANT BEHAVIOR, High Low ‘wore Less: Organization Organization ‘TERRORISM Caleetive RIOTING VIGLANTISM Individual Liability Liabtey More Less Organization ‘Organization LYNCHING “inavidual ‘abit Fig. 2 Socal distibution af fous fours of collective violencem Senccta de ta Roche ‘Two propositions address variation in the breadth of lability and in the degree of organization of collective violence: 5. Collective liability to collective violence varies directly with social polarization. 6, The organization of collective violence varies directly with the continuity of deviant behavior. Four propositions address the combinations that predict and explain the forms of collective violence: 7. Lynching varies inversely with social polarization ated the continity of deviant behavior. 8 Rioting varies directly with social polartzarion and teversely with the continuity of deviant behavior 9, Vigilantisin varies inversely with social potarization. ard directly with the continuity of deviant behavior 10. Terrorism varies directly with social polarization and the continuity of deviant behavior. Although scattered evidence tests and provides preliminary support for my propositions, a fuller assessntent of their predictive and explanatory power will require further investigation. It is likely, morcover, that other structural variables not yet discovered will prove pertinent In any event, further comparative work will clarify the circumstances under which the various degrees of severity and forms of collective violence are found. We might focus on a single variety of collective violence such as lynching or rioting, for example, draw a national, cross-cultural, or longitudinal sample ‘of cases, and determine whether its likelihood and severity follow the pre- dictions implied by my propositions. These and kindred investigations would yield not only a meaningful assessment of my initial formulations, but might also yield new directions for the future, ACKNOWLEDGMENTS 1 thank the following individuals for commenting on previous drafts: Donald Black, W. Fitzhugh Brundage, Stephea Cole, Randall Collins, Mark Cooney, Robert A. Dentler, Allan V. Horwitz, Edwin M, Lemert, Robert W, MeAhren, Calvin Morrill, Jeffery Mullis, David Sciuli, Stewart E. Tol- nay, James Tucker, Elizabeth West, Kendall O. White, Jr, and various anonymous reviewers. Foe example, Black's (1976) analysis of faw includes other varlables, suc as the degree of iuageation and cooventionaly, tha may explain variation in collective vel belp 35 well,Clete Violence ns REFERENCES ‘Ayers, Eaward 1 1984" Vengeanee and Justice: Chime and Punisher in the Nigeteenth Cen. lusy Amesican South. New York: Ox fact University Press. 1952 The Promise of the New South: Life After Reconstruction, New York: OF ford University Press Bancrat, Hubert Howe 1887 Populac Tribute, 2 vols, San Fran esto: The History Company. aver, Ashish 1990. "Comparative curfew. Changing mensions cf cenimanal pelts in In ia Te Veena Das, (ed), Mitrrs of Violence: Coromunities, Riots, and Survivors in South Asi 37-68, New York: Oxford Usiversity Pes. Baumgartner, ML P. 1984" "Social cartrol from below.” Jn Dee. ald Black, (e6.), Towatd a Gereral ‘Theosy of Social Contech, Volume 1: Fundamentals: 27-68, Ovlando, FL: Academie Pes. Beck, EM James L. Massey, and Stewart F-Tatmay 1989" "The gallows, the rob, the vote: Le thal sanctioning of blacks in Nor Carolina and Georgia. 1882-199. Lew and Society Remew 23:37-331, Beck, EM. and Stevact E. Tolnay W990 Te filling elds ofthe Deep South “The markt for cotion ard the fe ing oF Blas, 1852-1930." American Sociological Review 55:56.539 Berk, Richard A. 197 “The controversy surrounding ably ses of ealective violence! Some meth ‘xological notes” In Jers F. Shot, Se, ate Marvin E. Wolfgang (eds), CCttecsve Violence: 112-118 Chieag Aleine: Atherton Bereé, Yves Marie 199)" Biscory of Peasant Revels: The Soc (Ogsins of Rebellion in Foxy Modern France. Ithact, NY: Cornell Unier- sie) Press. Black, Donald 1976" The Betavior of Law, New York: Acs demic Press. 179 “A'stategy af puce sociology.” In Seatt G, MeNall, (ed), Theareveal 988 1934 1980 1980 1998 1995 Perspectives in Sachology: 149-168 New Yorks St, Martin's Press time as soeil control” American Sociological Review 48°34.45, ‘Socal coneal ar 4 dependent vt able" ta Donald Black, (ed), Toward 2 General Theary of Sozial Cantal, Volume t: Fundamencals: 1-36. Or lado, FL: Academic Pres. "Compensation and the socal truce ture of misfortune” Law and Society (view 21 560.554 ‘ae elementary forms of canthet management” In New Ditections is the Study of Justice, Law, and Social Control” 43-69, Prepared by the School af Tuslice Studies, Arizona State Usiversiy. New York: Plenum Press. ‘The Social Scructure of Right and Wrong. San Diego, CA: Academic Pres ‘Tae epistemalogy of pure sociology.” Law and Social Tequiry 20829-8 Blalock, Hubert Moy Je. sr Blam “Towatd 2 Theowy of Minorig-Group Relations. New York: ohn Wiley & Sovs Herbert allective behavior." In Alfred MeClung Lee, (ed), Prieciples of So- cology: 167-122, New Yark: Baenes and Noble, Bonacieh, Edna i A theory of ethaie antegoaitc: The split labor eardel” American Secia logical Review 37547-558. Brown, Richard Maxwell 1969 ws gst “Historical patterns of violence in America” In Hugh Davie Graham and Ted Robert Gr, (eds, The Hie- tory of Violence in America 45-84 New York: Bantam Books, Strain of Violence: Historical Srdies fof American Violence and Viglan: tom. New York: Ovoed Univers No Duty to Retreat: Violence and ‘Values im American History anc Soci+ ery. Normen: University of Otlahorns Presns ‘Bean, Richard Maxwell, ed. 1970 American Violence. Englewood Clif, INE Prentice, Brundage, W. Fitzhugh 1985) Lyrening ip the: New South: Georgia and Vigin’s, 1880-1990, Utara Unk verity f Tinos Pres. Capeci, Dominic J., Jr. and Martha Wilkerson 1901. Layered Violence: The Dettit Riot 613 of 194, Jackson: University ress of Missi Caughey, John W. 1957 “Theie majestiee the mob Pacific Hisotical Review 26:017-234 Chaplin, J.B 1859) Rumor, Feat, and the Madness of Crowds: New York: Ballantine Books Chen, Tack 1580" The Chinese of America. New York: Harper & Row. Clayman, Steven 1998. "Boing: The anatomy of » disatiia: tive response.” American Sociological Review $8:110-8, Cooney, Mark 1995 “evidence as partisanship” Law and Society Review 28:33 388, Curvn, Robert and Bence Porter 1979 "Blackout Looting! New York Ci, Tuly 13, 1977, New York: Garnet Press, Dabrendort Ralt 1968 Values and social scence: The value dispute i perspective” In Ralf Dac heendoxt, Essay in the Theeey cf Sor ety (1960) Stanard, CA: Stanford Universty Press Davis, Natatie Zemon 1975" Scary ard Culture in Ely Modern France. Stanfors, CA: Stnford Unt versity Pres Daler, Carl N. OPT Neither Blacks Nor White: Slavery sed Race Relations ie Brasil an@ the United States New Yorke Macrln Downey, Dennis B. and Raymond M, Hyser 1990" Ne Crooked Desh Coateile, Pere sylvapia and the Lynching of Zachatiah Walker. Urbana: Univesity of Minos Press, “origina! pudieation date Senechal de la Roche Dabo, Mtn 98S Tedustrialism and the American ‘Werer, 1865-1920. 2ad ea. Arlington eights IL: Harlan Davison, agerton, Robert B. 172 Violencein Bast Afecan tba sace- fies." In James F Short, Ie. and Marvin F Wollsang, (eds), Collective Violence: (59-171, Chicago: Aldine Atherton 1989 Mau Mau: An Ariesn Crucible New ‘York: Hallantine Hooks, Fagin, Joe R. and Harlan Ham 1995 Ghewo Revol: The Politics of Vio fence in American Cities, New Yor Macmillan Fogelson, Robert M. 197 "Violence and grievances: Refletions fon the 19st rots” Journal of Soci Tases 2641-188, Foster, Don 199), "Crowes and colletive volenee.” In Don Foster ane John LeawePesgietcr, (is), Social Payhology fo South A ea. Isanco, South Afvic: Lexicon Publishes, Fradeickson, George M. U8i White Supremacy: A Comparative Study in Amenean and South Atiean History, New York: Oxford University Pres. Freer, 1901" “Black-Kocean conflict." In. Mark aldassae, (ed), The Los Angeles Re ots Lessors for the Urban. Farsre: 175-203. Boulder, CO: Westview Pres, Galante, Mare 1974" "Why the “haves' come out ahead: Speculations om the limits of egal change.” Law and Society Review 995180 Gi, Paul A 1587" The Road to Mobocrae: Poplar Dis, srder in New York City, 1763184. Chapel Fil Cavers af North Cato: lina Bess, (Gluckman, Max 1936, Caslam and Contiet in Afees, Ox ford, UK: Basil Blackvell 1965 Politis, Law and ital in Teal So- ciety Chicago: AldineCatestive Violence Graham, Hugh Davis 1889 "Violence, oral theary, and the his tosians: The debate ver concensas and culture in America.” In Ted Robert Gurr, (e.), Violence in Arter fea, Volume 2 Prove, Rebellion, Re fore: 329-381, Newbury Park, CA: Sage. Geanorater, Mark 197R “Theeshold models ef coletive be haviou” American Jouseal of Sociol ogy 85:1000-143, Greenberg, Lonls 1976 The Jews in Russia: The Seugge for Emancipation, Valume 2. (1981") New Yorks Sehockea Rook Grimshaw, Allen D. 1963) “Ations ofthe police andthe eilary in Amecicat tace riots.” Phyion drs. 172. “Interpreting collective wolence: An argument forthe ampararce of social Ssruetoee” tn James F Shore rd Marin E. Wolfgeng, (68). Collective Violence” 3-46, Chleago: Alias Atherton, Grimshaw, Allen D. ed. 1965. Racal Viole in the United States CChicage: Alsine. Gurr, Ted Robert 1910 Why Men Rebel, Princeton, NI Princeton University res 1989. “The kistory of protest, r2belion, and feform lo Ametiea: 2, overviei? Tn ‘Ted Robert Gurr, (ed), Viele in America, Volume 2 Proves, Rebel~ Ton, Reform: 11-22. Newbury Park, CA" Sage, apy, Andeew B. and James F. Short 1954” Suicide and) Homicide: Some Bea amie, Sociological ae Psychological ‘Aspecis of Aggression. Glencoe, L: Free Press odes, Martha 1013 "The semulization of Recaratcuction pallies White women and blek men fh the South after the Civil War.” Tournal af the History of Sexaialiy sama Horwity Allan V. 1983" The Social Corerl of Mental tines. New York Acagemic Pres. 1960 The Lagie of Sorial Conical New Yetk: Plenum Press ns Hovland, Cart J. and Robert R. Sears ISHO” "Minor studies of aggression: Corea: tions of econamte indices wich Iyachings.” Zeuenal of Psycnology 930.10 Hunter, F. Robert 1951 The Palestinian Uprising: A War by Other Means, Berkeley: University of California Pres Hurtado, Albert Le 1965" Indian Survival on the California Frontier. New Haven, CT Yale Uni versity Press. Hussain, Akmal 1900) “The Karachi roi of December 1986: Gis of tate and evil society it Paki- stan." In Veena Das, (ed), Mirrors of Violence: Communities, Riots, 2nd Survivors in South Asia: 185-193. New ‘York: Orie Universiy Press. nveraity, ames M. 1975. “Populism and lynching ia Louisiana, 1889-1696." American Sociological Review 41.6280 authority of popular justice inthe Far West.” American Quarterly 35.558 586, Sanapathipila, Vali 1990 "July 1985. ‘The survivor's experi ence” In Veena Das, (64), Mics of Violence: Communities, Riots, and Survivors in South Asia 221-348, New ‘Vouk: Oxtord Univesity Pres, Kaanangara, AP. 088" "The eons of 1915 in Sei Lanka: A study in the roots of communal ie lence." Past and Present 102110-16, Kerner, Oe sl 1968 Repoct of te National Adws0ry Core- ritsion on Civil Dianrders, New Yor Bantam, oct, Klaus-Feiedeich 1984" "Liability and socal steture.” In Doralé Black, (ed), Toward 2 Gen- eral Theory af Social Conta, Vonme 1 Fundamentals: 95-129, Orlando, FL: Academie Press, Laue, Bruce 1880 Working People of Philadelphia, 1809- 1850. Philadelphia, PAL Temple Unie etsy Press136 LeBion, Gustave 1960 "The Crowe: A. Study of the Popular Mind. (19957) New York: Viking Pres Lee, Alfeed MeCtung and Norman D. Humphrey 1945) Rate Rint. New York: Dryden Pres. Lesin, Jack and Jack McDevitt 1993" Hate Ceres: The Rising Tice of Big: fouy and Bloodshed, New York: Ple tum Pres, Lewis, Michael 1980” Rioters and Citlens: Mase Pretest in petal Japac, Betkeley: University of California Press aco, A.W. 1968" Vislence in Republican Rome. Lan an Oxford Usivecsity Pres, Lite, Cealg B. and Christopher. Sheffield 1983" "Frontiers and criminal fuscze: Eag ist private prosecution cities ane ‘American vigilant in he eighteenc dad rineteceth centuries” Americas Sociological Renew 48,796.508 Malone, Michael P. and Richard R. Roeder 1976 Montana: A History of To Centuries Seattle University of Washington Pres Mandelbaum, David G. 197% Society ie Lodi, Volume 2: Change ‘and. Continuity. Berkeley: University ‘of Calforn’s Press, Mary, Gary T 19M “Chi esrder andthe agens of social contral,” Journal of Social Feet 2619-97, 1972 “asusless rite” In James B. Short and Marvio.F- Wolfgang, feds) Col Teatve Violence: 47.59. Chicago Ald ine-Athestan Mazeul, Ala 1976 “Black vigantion ie cultural ans. ‘ion: Violonee snd vishiligy in cepial ‘Airia:” ke Jon Rosenbaum and Peter C, Sedetberg, (eds), Vigilante Politics” 194-217. Baladelpbiat Uni vers of Pennsyivania Pees. McGrath, Roger D. 1584" ‘Gunfighters, Highwaymen and Vigh antes: Violence an the Frontier, Berkeley: University of California Pres. “Oviginat publication date, Senechal de ta Roche MeMtilen, Nit R. 1989 Dark Journey: Black Mississippian ia the Age of lim Crow. Urbana: Univer sigy of lina Press. Mitek, Maureen 1971 “Pale ses landlord: An observa tion stucy oF acministatve conte.” PhD. dissertation, Department of Se- ology, Yale University. Moore, Sally Falk 1972 Legal Uablity and evolutionary inter (prettion: Some aspets of set liabil fy, selCtelp and collective responsibilty.” In Max Gluckman, (e2), The Allocation of Resparihil ity: 53-107. Monehester: Manchester Unwersiy Pees, Mari, Calvin 1982. “Vengeance among executives." To Tames Tucker, (65) Virginia Review of Sociclagy, Volure 1: Taw and Con ‘ict Managertent: 51.5. Gresawieh, CT IAT Press, 1995 The Executive Way: Confit Marage- see iv Corporation. Chicago: Unie verity of Chicago Pres, Moll, Jeffery 1905 “Mecizal malpractice, social seuetue, sand socal comteal” Sociological Fo. am 1035.16, Nash, Gary B. 1974" Red, White, and Black: ‘The Peoples fof Early America. Englewood Cli, Nk: Prentice Hal ‘teak, Susan 1889 Labor unrest, smmigeation, ana eth rc confit in urben America, 1850. 1914” American Journal of Secicingy 94 1805-1385, 1999 "The politcal contex of competition: Lynching and urban racial violence, 122 1914," Social Forces 69°595-401 1992 The Dynamics of Ethnic Competition and Confit. Sanford, CA: Stanford ‘Universi Press. ‘Olzak, Susan and Suzanne Shanatan 1994 "School desegregation, interracial expotsee, ond antibising activity i contemporary urban America.” ‘American Journal of Sociclogy 100-186-241,Collective Violence Olzak, Susan, Suzanne Shanahan, and lizapeam Metneaney 1898 "Sources of cootemporary cacal un rest in the Uniced States. Fram the 1960s chrougle 1990s. sented at the Angus! Meeting of the Soci Science History Association, Balimore, MD. Port, Bruce and Marvin Duna 1884" The Miatek Riot af 1980, Lexington, RY: D.C, Hea Quine, Lary D. 1867 "Chink chink Chinaman: Tae beain of ative in Montana Paci fctnwest Quarterly $8829 adele, Michael L. 1989 “Executions of whites for eximes against blacks- Exceptions te the tule?” Sociological Quasceriy 30529 4, Raper, Arthur 1833 "The Tragedy of Lynching, Chapel il University of Noxth Carolina Reaiker, Mares 1987 fleween the Desl and the Deep Bue See Merchant Seamen, Pirates, and the Anglo-American Maritine Wodd, 1700-1750. New York: Cambrigge Unsersiy Press Regal, Robert M, Andee W. Mirade, Je, ‘aad Eric D. Poole 1954 “Law are social contol in Chins: An application of Black's thesis" Cri ral Tuntice Review 9:15 1ghs, Anarew i881 Living the Intilede. London: Zea Books, Roberts, Michael 1950 “Noise at cukural struggle: Tomtom beating, the British, and communal sisturbances in Si Lanka, 1880-1930" Tn Veena Das, (0, Mirrors of Vio leace: Commurities, Rits ard Surv wore im South Asia. New Yorks Oxiord University Pees Roberts Simon 179 Order and Dispute: An Tetraduetion to Legal Anthropology. New York: St. Maclin’ Press ‘Rosenbaum, H. Jon and Peter C. Selerberg, cis. 1976 Vigilante Potides, Philadelphia: Uni versity of Peonsyvania Pres. m Foy, Bet 198% "Some Trouble with Cows: Making Sense of Social Conflict, Berkeley Univesity of California Pres. Ru, George 1864" The Cloud ia History: A Study of Popular Disturbances in France and England, 1750-1848, New York: Wiley. Rusdvik, flit, 1961" Race Riol at East St Lous, July 2, 1919. Carbondale: Southern iinois Universi Pees. Rote, James 1885" “Theories of Cuil Violence, Berkeley Universicy of Calforeia Press. Senechal [de la Roche], Roberta 1999. The Sociogenesis af a Race Riot Springilc, Minas, i 1908. Uvhara Univrsicy of Minos Pres. Seneca ge la Roche, Roberta 199 "The sociogeness of lynching" la W. Fichugh Brundage, (ed), Under Sen- ence f Death: Essays ot Lynching In the South. Chapel Hl: Uniersiy of North Carolina Press. Focthcoming Senkewicr, Robert M. 1985" Vigilantes in Gold Rush San Fean- sen, Stanford, CA: Stanford Uiver sy Pres, Sessions, Gane A. 1976. "Myth, Mormonism, and eourder in the South.” Sourl Ailatie Quarterly s72i223, Staley, Ane IG91. ‘The Ilifada: Causes and Bltecs Boulder, CO: Westview Pres. ‘Shipler, Davia K 1986 Arab and few: Woutded Spi in the Promised Land, New York: Pengtit Toots, ‘Should, Re Lance 1575 "Spontaneous vigilantes” Society ne Smeser, Nell J 1953 "Theory of Collective Behavior. New ‘York: Fice Press Stone, Thomas 1979" The Mounties as vigilantes: Peceep- tions of carerunity and the tensor mation af (aw in che Yukon, 1895-1897" Law and Society Review twas.ils, 1983. Miners Justice: Migration, Law ard (Order on the Alaska Yakon Eeoatir, 1873-1902, New Yack: Peter Lang.ns att, Philip and Philip Rass 43869 “Americas labor violence: Its causes, charzeter, and outcome" Ta Hugh [Davis Grabam and Ted Rebert Gur, (eds), The History of Violence i ‘America: 281-085, Now York: Bantam Boats, ‘Tamia, Stanley J. 1990 “Presidential adcess: Reflections on communal violence in Sovth Asia.” Tournal af Asian Studies 48°741-760. “Thompson, FP IMTS "The ene of anonymity” Ia Douglas Hay etal (eos), Alsons Baal Tree: CCime and Sociay in Fuhteecth Ceo. tury England. 255-144. New York Pantheon Press. ‘Tierney, Kathleen J 198" "Property damage and violence: A cal- lective behavior analy” In Mark Baldassare, (e4), The Les Angeles i ‘ois: Leste for the Urban Puture: 149-173, Boulder, CO: Westview Pres ly, Charles 19st) “Collective wolenes in European pers spective” Th Hugh Davis Gezhar and ‘Ted Robert Gurt, (eds), Violence in ‘America: Histories! and Comgararive Perspectives: 45. New York Signet Books 1973 From Mebilization to Revolution Reading, MA: Addison-Wesle. Tily, Charles, Louise Till, and Richard Tilly 19 The Rebellious Century, 1830-1930, Cambeige, MA: Harerd University Pres Tolnay, Stewart E. and E. M. Beck 1965" A Festvalof Voletce: An Anais of the Lyaching of Blacks ia che Ame can South, 1882-1990, Urbana: Uni versiy of Minis Press. ‘Toluay, Seorart E, EM Beck, and James Le Massey 1989" "Black (yechings: The power threat hypothests eevisited,” Social Forces ras, Senechal de ta Roche ‘Torrance, Judy M. 1986 "Publi Violence in Carada, 1867-1982 Moctreal McGill-Queens Unnersiy Pres. ‘eelease alan We. 19M White Tere The Ku Klux on Con fhitag and Southern Reconsttuction. New York: Harper & Row. Tucker, James 1989" "Emplovee heft a social conte Deviant Behawor 1319034, ‘Turme, Ralph 1360 "Collective behav.” In Robert EL. Fears (e0), Handbook o¢ Modern 86 Chicago: Rane McNally Mi, Je 1870" Race Rint Chieago in the Red Sum tner of 1919, New York: Atheneum, Yon der Mehden, Frei. 1976) "Pariar communities and violence.” In Ht fon Rosenbaum and Pete C. Severberg, (cds), Vigiante Polis: 2ig-733. Phlaceiphia: University of Peonaiania Pres, Wade, Ceaig Wy 1987 The Bes Cress: The i Ke Kn i “Ameticn New York Simon & Schuse. ‘Weinberg, Robert 1085 "The Revolution of 1905 in Odessa Bloomington: Indiana Universy Pes. Weisbord, Dad 1989" Jewish Seer Violence: Deviance at Social Reaction. University Pat Rennyania Ste Universi Pees. ‘Witiamsoa, Joe 1988" The Crucible of Race: Black-White Relations in the American South Since Emancipation. New York” Ox ford Univer Pes. Weigh, George C 1990" Racial Violence in Kentucky, 1865- 1540" Lyochings, Mob Rute, and "Le- al Lynchings.” Bator, Rouge ui Ste Unversity Pres. WyateBeows, Bertram 1982 Sovthcon Honor: Ethics and Behavioe in the old South. New York: Oxtord Univesy Pees