Understanding Political Violence in Italy's Protest Cycle, 1966-1973 - Prof. Pilati, Schemi e mappe concettuali di Sociologia

Scarica Understanding Political Violence in Italy's Protest Cycle, 1966-1973 - Prof. Pilati e più Schemi e mappe concettuali in PDF di Sociologia solo su Docsity! European Journal of Political Research 14: zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA607-632 (1986) zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 0 Martinus Nijhoff Publishers. Dordrccht vutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA- Printed in the Netherlands zyx tsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Unwanted children: Political violence and the cycle of protest in Italy, 1966-1973 Donatella DELLA PORTA’ & Sidney TARROW’ zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA European University Institute, San Domenico di Fiesole. Italy; Cornell University, lthaca, zyxwvutsrqponmlkji fedcbaZYXWVUTSRQPONMLKJIHGFEDCBAUSA Abstract. The article focuses on the cycle of protest that developed in Italy during the late 1960s and early 1970s. Some hypotheses on the evolution of the repertoires of action are tested with the aim of explaining the emergence of political violence during a cycle of protest. Newspaper-based data are presented on the proportional presence of violent forms of action, on the social and ideological groups involved in political violence. and on the grievances expressed during violent protests. The widespread political violence that developed in Italy in the early 1970s is explained as an internally differentiated strategic adaptation within the social movement sector, during a cycle of protests that was disorderly but far from violent. In the wake of the tragic kidnapping and murder of Aldo Moro in 1978 by the Red Brigades, Rossana Rossanda, a journalist for the newspaper Manifesto, wrote that the terrorists were like the unwanted children in the ‘Family Album’ of the Left. The reactions to Rossanda’s comment were predictable: on the Left, the PCI- ever concerned for its respectability -denied that it bore any responsibility for the terrorism of the mid-l970s, while the Right thought it had found an admission of a direct connection between the mass protests of the late 1960s and the organized terrorism of the next decade. But both missed the core of Rossanda’s insight: that terrorism could not be understood except as part of the cycle of protest that gripped Italian society in the decade after 1965, for which even terrorism’s enemies - and the PCI was surely among these - bore some responsibility. It is towards understanding the relationship between that cycle and its violent outcomes that this article will be devoted. In it, we shall argue that violence cannot be understood except in the context of a country’s cycle of mass protest; that it is a function of the competition between components of the social movement sector; and that it occurs - not as the essence of a period of mass protest - but as a sign of its decline. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA A previous version of this article was presented to the Workshop on the Dynamics. Strategy and Efficacy of Political Protest of the ECPR Joint Sessions, held in Gothenburg, Sweden, in 1986. 608 I. Introduction zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA We shall not focus in this article on Italian organized terrorism itself zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA- it already has a long and tormented bibliography' - but on the less organized violence that preceded it and that we think provided a training ground for its development [della Porta, zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA19851. Nor will we argue that political violence was the only - or even the most important -outcome of the Italian protest cycle.' What we will try to show is that the widespread political violence that de- veloped in Italy in the early 1970s can best be understood as an internally differentiated strategic adaptation within the social movement sector during a cycle of protest that was disorderly but for the most part was far from violent. The theoretical literature on political violence is rich in 'structural' contribu- tions that stress economic social. political and cultural variables' and in 'con- junctural' ones that underline the peculiarities of historical periods of individ- ual countries.' 1. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAStructural explanations, which are frequently based on cross-national comparisons of aggregate data, point to a society's level of development [Russett et al., 19641, the presence of ethnic or class cleavages [Barrows, 19761. the country's political culture [Bandura, 19731 and wider economic ine- qualities [Sigelman and Simpson. 19771. 2. Conjunctural explanations, which are most often based on historical narrative or configurative case studies, find that modernization phases are related to violence [Huntington, 19681. as are intermediates stages of eco- nomic development [Feierabend and Feierabend. 19661, periods of ineffective state coercion [Tilly, 1969; Feierabend and Feierabend. 19721 and rapid shifts in value systems. The first important feature of the literature is that much of it is either excessively macro-structural - and therefore has difficulty tracing the transfor- mation of structural configurations into patterns of mobilization - or is exces- sively conjunctural - attributing responsibility for political violence to idio- syncratic features of a particular society in which it happens to be common. Between the macrostructural models found in cross-national studies and the conjunctural ones found in the national narratives, there has been little attention to the specific conflicts, constraints and opportunities for advantage that encourage the appearance of collective action. This can be understood in two senses: temporally, as a political opportunity structure that develops and changes in the course of a cycle [Tarrow. 19831 and sectorally, in a social movement sector that reacts differentially to these changes in opportunity [Garner and Zald, 19831. By political opportunity structure, we intend the potential for gain and loss through disruptive direct action represented in the political and social environ- ment of insurgent groups. By social movement sector we intend, following trigger political responses in a logic of action and reaction. We differ from most students of political violence in regarding the dynamic of the cycle zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA- and not its structural causes - as the most important element in the development of violence within a protest cycle. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA“ I Typically, the early phases of the cycle are composed of mass protests which arise out of interpersonal networks and categorical groups using existing repertoires of action to publi- cize precise grievances and demands for representation. These attract the external support of ‘conscience constituents’. disrupt routines, disorient elites and encourage the formation of organizations to represent the emerging social movements. These forms of contention are diffused to new actors and regions and are adapted to different kinds of claims. At this stage, elites are divided, political groups are confused. and instru- ments of effective repression have not been developed. When they are used, it is mainly as a result of zealous and inexperienced police action and not of general policy. In a general climate of open access. police violence in response to protest is neither general nor effective, leading only to outrage and further protest. often in solidarity with the original protestors. Existing organizations- trade unions and parties- are also taken by surprise at the vehemence and generality of protest. though they often unwillingly provide the cadres and symbols around which protest is organized. Though ‘new’ actors and new modes of action begin to appear, they often begin their careers within and in conflict with existing organizations [Zald and Berger, 19781. The success - or at least the survival - of the early phase of mass protest leads new social actors to enter the movement sector on behalf of their own aims and activates those whose interests have been adversely affected by the original actions. Sequences of action and reaction become common in a climate in which protest is legitimized and ordinary means of redress appear to be ineffectual. This brings disruptive behaviour into secondary sectors of society in which protest is not traditional and in which elites are not accustomed to dealing with contestation. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA As protest becomes more general and the number of social actors and political organizations grows. a process of competitive tactical differentiation begins. As the impact of the original. routine forms of action declines and elites learn how to adapt to them. innovation occurs in the forms of action employed and in the types of claims made [McAdam. 19831. Spreading social conflict and an air of innovation encourages ‘new’ issues - the environment. religion and culture, family and gender conflicts- to be placed on the agenda. Competition for a following in the presence of expanding mobilization leads movements to develop confrontational, mass forms of action. These confrontational forms are usually disruptive but non-violent; never- theless. some are bound to lead to violence, particularly as small political 612 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA organizations operating on the margins of large public actions try to establish themselves against their competitors. They lead to negative reactions. both on the part of antagonists and by the forces of order. The former develop more disciplined forms of action and specialized cadres whose task is to control demonstrations, while the latter - under pressure from the public to stop disorder - respond to specific provocations with general violence. Finally, there is a tendency for disruptive behaviour to reach into the institutionalized sector, as insurgents in the trade unions, professional associ- ations and interest groups are attracted by the new demands and more direct tactics. Established organizations compete with the movement sector for support. at times by copying its tactics and adopting its demands, but more often by routinizing and tempering them. This gives the new movements incentives to create more conflictuul forms of action and to state their demands in more general and more extreme form. The appearance of social protest in new movement sectors, its adoption by existing organizations and a mush- rooming of political groups gives the impression of total uproar just as rnobi- lization begins to decline. The contradictory aspect of this situation requires some underscoring. Protest has been diffused to different parts of the society; it has led to the creation of a number of competing social actors and political organizations, which employ a broad repertoire of competitive tactics. It thus gives the impression to the public and to elites of broad and increasing mobilization. But it has also brought part of the movement under the control of organiza- tions which continue to rely on traditional repertoires and make limited claims. And because it has led to the adoption of more aggressive and con- frontational forms of action by some, it leads to at least partial demobilization by others and to a decline in support by ‘conscience constituents’. It is thus the magnitude of participation across the system and the number of competing social actors and political organizations that have increased, and not the strength of any particular movement or of mobilization in general. In fact, the ability of the movements to mobilize a mass following weakens as more groups crowd into the movement sector and the intensity of the cycle reaches its peak. Thus the intense forms of action adopted by individual movements or sectors is a substitute for high levels of participation, and not its evidence. And their adoption of general goals and ideological programmes is an attempt to broaden their appeal from a situation of increasing weakness. One important outcome is factionalization, which is not only the result of ideological differentiation, but of political organizations manoeuvring for turf in a crowded arena in which participation is declining. As these differences develop, they polarize the movements. with more aggressive participants gravitating towards organizations that have opted for confrontational forms of action and less aggressive ones either shifting towards more peaceful forms or ’ 613 leaving the movement sector altogether for private life. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA A second result is a backlash on the part of public opinion and a more deliberate repressive policy on the part of the police. The public does not know that mobilization is in decline; what it sees is the increased number of social groups and political organizations that have exploded on to the public scene and the increasingly confrontational tactics they use to gain public attention. Political elites, once divided about the usefulness of the new social groups and the legitimacy of repressing them, begin to respond to public pressures to maintain public order. One result is to further demobilize some and to push others towards the only forms of action that can succeed in such a climate zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA- covert violence. Third, semi-clandestine organizations dedicated to organized violence ap- pear both within and against the early movements, bringing a new kind of conflict - violent encounters between opposing groups - into prominence. This in turn has a further debilitating effect on mass protest and provides a training ground for the most aggressive of these organizations, which even- tually adopt organized terror as their major tool. A related effect is that protests on behalf of group claims decline as the social movement sector turns in upon itself. The logic of a protest cycle is that institutionalization and increasing vio- lence accompany and feed upon one another as the cycle winds down. That both these trends augur the completion of the cycle should by now be obvious. The culture of protest, which so recently appeared to be both general and transformative, retreats into art forms, fashion and religion. Political elites reweave the fabric of political society, perhaps adding a client group here and deserting another there. Mass demobilization, institutionalization and the isolation of extremists into a clandestine ghetto make it possible for economic elites to regain their hegemony. In the next section, we shall, first, lay out the main phases of the Italian protest cycle, attempting to trace the number of protest events, the intensity of conflict reflected in the newspaper articles, and the proportion of violent forms of action, presenting a more detailed analysis of forms of action, to show when violent repertoires appeared and how they changed during different phases of the cycle. In Section 111, we will deal with the social and ideological groups that were involved in political violence and with how they differed in the forms of action that they employed and with the sectors in which violence occurred. In Section IV, we will turn to the 'why?' of violent protest, attempt- ing to sort out the main grievances of violent protests both in terms of target of the action and of the nature of the demands. We shall close with some observations that illustrate our interpretation of the quantitative data and some speculations about the implications of these findings for understanding the organized terrorism that began towards the end of the period we shall study. 616 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 350 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA- 300 - 250 200 ca * 100 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 0 violent zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPON LKJIHGFEDCBAe v e n t s # t o t a l zyx vutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAeVe"t.8 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA n 0 1 9 6 6 1 9 6 7 1 9 6 8 1 9 64,clr&3..p\L0 1 9 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA7 1 1 9 7 2 1 9 7 3 Fig. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAI . Protcst and violent protcst per year A third phase of the cycle starts at the beginning of 1972 and lasts at least up to the end of 1973. The number of total occurrances of violence decreased after a last high peak in the winter of 1972. We cannot of course affirm that the cycle was definitively over after 1973; it is nevertheless indicative that in the winter of that year the total number of protest events reached only slightly more than half of that of the same season the year before. But the decline is much less sharp for the violent component of the protest curve. Here, in fact, the percentage of violent cases was higher than the average for the entire period. Analyzing this trend in the light of the evolution of the protest cycle, the incidence of political violence is strictly connected with collective action in at least two ways. First, it grew in total numbers during the whole cycle. Second, its percentage weight was average in the beginning, low during the upswing and high in the declining period of the protest wave. Violent forms were, therefore, part of the protest repertoire from the beginning and their presence tended to grow in total numbers during the whole cycle. But it was when the wave of collective action declined that their percentage distribution increased. This is a very rough description of the temporal trend of the violent sub- repertoire. It does not explain why violence grew in importance during the cycle. Moreover, the description will not be detailed enough to be useful when we turn to a causal analysis. In order to be exhaustive, we need more informa- tion on the different forms that violence assumed in Italy. 617 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA A . The degree zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAof violence The degree of violence is not an easy matter to define. The solution suggested by Wilkinson (1977) has never found a satisfactory empirical translation. Our operational definition of the degree of violence is based on the aggregation of a combination of forms and outcomes under four different categories: incidental violence, attacks on property, violent encounters, and attacks on people. In the category of incidental violence we include those events in which the protesters made use of peaceful forms, which only in the process of interaction produced violent outcomes. Attacks on property include rampages and ac- tions oriented to property damage. Among violent encounters. we include violent conflicts, either between contending groups or between protesters and police. The proportions of the four types in the eight year period we studied was as follows: incidental violence accounted for only 4% of the violent events (and zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 1% of the total events); attacks on property follow, with 21% of the violent events (and 7% of total events); attacks on people were present in 33% of violent actions (and 12% of total events); while violent encounters were the most widespread form of violence, with 42% (accounting for 15% of the total events). If we combine the first and the last types and the second with the third, we can see that the proportion of planned attacks on people or things was far from high, reaching less than 50% of the total violent events. As can be seen in Figure 2, the four types of violence follow quite different patterns of evolution over time. Let us look first at the two purposive kinds of violence zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA- against property and people. The two curves can be described as being roughly parallel from 1966 to 1969. The increase in number was then faster for attacks on people than for attacks on property, which declined in 1971. But the more notable differen- ces appear when we look at what happened in the last two years, between 1972 and 1973. While the number of attacks against property declined dramatically, the number of attacks against persons kept growing as the overall magnitude of protest declined. If we look also at the other two curves we notice that attacks on people were the only type of violence whose presence rose in absolute terms after 1971. Its influence is even more evident in percentage terms, with attacks on people jumping from the 24% of 1968 to the 45% of 1973. Let us turn our attention to non-purposive violence. Not much can be said with regard to incidental violence, whose values are in general quite low and with a weak fluctuation during the whole period. Violent encounters were the more widespread form of violence up to 1972. Their number grew regularly from 1966 to 1972, with a very high peak in 1968 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA(55% of the violent events). It reached its peak in absolute terms in 1972 and declined in the following year, 1 60 0 violent encounters attacks against people zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA + attacks against property - incidental violence zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 0 - - - - - - - 1966 1967 1968 1969 1970 1971 1972 1973 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA years zyxwvutsrqponml ihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Fig zyxwvutsrqponmlkjihgfedcbaZYX VUTSRQPONMLKJIHGFEDCBA2. vutsrqponmlkjihgfedcbaZYXWV LKJIHGFEDCBAViolent types per year. zyxwvutsrqpon jihgfedcba ONMLKJIHGFEDCBA when it accounted for only 30% of total violence. At this point. our description of the evolution of violence during the cycle can be enriched by some specifications. First. a high percentage of violent events took the form of violent encounters. It is very difficult to say if they were planned or not. But what is certain is that they took place between competing groups or with police - and not against targets in civil society. Their incidence was greatest at the peak of the cycle and decreased in the last years of the period. As far as the gravity of the effects is concerned, the more bloody and directed forms of violence - attacks on persons - grew continuously up to the end of the cycle, while the less dramatic forms of violence tended relatively to decline. In the Italian protest cycle, incidental and inter-group clashes coin- cided with magnitude with the height of the cycle, while violence directed at persons or things increased as the total magnitude of conflict declined. There is another differentiation in the concept of violence that we need to introduce at this point in order to make our description more precise. Philoso- phers as well as politicians have - sometimes in an over-simplified way - distinguished between an ethically justifiable mass violence z xwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA(e.g. ‘of the people’) and a morally wrong small group violence. Without entering this debate, we can see if the violent events which involved a large number of protesters followed a different pattern from those in which only small groups of people were active. 62 1 The major incidence of violence was in protests involving students or young people, 54% of whose actions engendered violence of some kind. That young people are more likely to become involved in violence than their elders has become a truism of our time. In the Italian case, we find a high level of violence even when they are involved in the same events as industrial workers. Analyz- ing the 4,418 events in which a known social actor participated, there was violence in 51% of which young people were present, as opposed to 15% in which only workers were present, and the same percentage in which neither was present. Popular imagery has students trashing their schools and sequestering teach- ers, but most of the violent youth protests do not take place in educational settings. When we coded the events according to which sector they involved (see Table 2), only 26% of the educational protests involved violence, only slightly more than the events in industry (21%), and not much greater than protests in the service sector (16%). The conclusions are even more striking if the data are disaggregated by the different types of violence. It appears, in fact, that in all three sectors, the great majority of violent occurrences took the form of violent encounters or inciden- tal violence (67% for the industrial sector, 58% for the educational sector, zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA 45% for the service sector). Most of the violence involving young people seems to have been related - not to sectoral demands at all - but to political antagonism, often between competing groups of opposing ideological tend- ency. What do violent groups or groups interacting violently compete about? This moves us closer to what will be dealt with in detail in the next section. At this stage, it is sufficient to point out that protests were far more likely to bring zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Table I. Incidence of violent and non-violent events by actors who participated in the action Social actors Non-violent events Violent events Total events* %, N . zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA"/i N. %, N. Workers 80.6 869 19.4 209 21.6 1,078 Students 47.0 903 53.0 1.019 38.6 1.922 Old middle class 83.7 220 16.3 43 5.3 263 Public cmployces 91.2 548 8.8 53 12.1 60 I New middle cla5s 89.3 343 10.7 41 7.7 3x4 Urban actors 74.9 340 25.1 I15 9.1 454 Others 42.2 419 57.8 573 19.9 992 * The total refers to the numbers of events in which each category of social actors was present. Since more than one actor was often present in the same event. the total of social actors' prcscncc is higher than the numbers of events. 622 about violence when the groups involved had a clear and observable ideologi- cal character. In the zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA1.866 events in which ideologies were clearly manifest, as Table 3 shows, 70% involved violence. If not a cause of violent behaviour, ideology is at least zyxw utsrqpon lkjihgfedcbaZYX VUTSR P N LKJIHGFEDCBAa strong corrolary of it. What kind of ideology is most likely to engender violence? It will surprise no one that extremist groups engage in violence more frequently than moderate ones. Second, violence is almost certain when opposing ideological groups meet. And third, as Table 4 shows, extreme rightwing groups are more likely (%YO) to be involved in violence than leftwing ones (72%). Not only that: there is a sharp difference in the type of violence in which we find extreme left- and rightwing groups engaging. The Left is twice as likely to be present where there is a violent encounter between groups or with the police (41% vs 22%). while the Right is more likely to be involved in property zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Tohk 2. Incidence of violent and non-violent cvents by policy sector of the protest Sector Non-violent cvents Violent events Total events zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA Yo N . Y" N . %" N. Industrial 79.4 714 20.6 I85 18. I 899 Education 74.8 661 2.5.2 223 17.8 884 Service 84.0 1.1167 16.0 203 25.5 1.270 International affairs 51.9 10s 4 . 1 I00 4.2 208 Political antagonism 5.3 43 94.7 775 16.4 818 Others 64.4 5SO 35.6 320 1X.1 900 Total 63.7 3.173 36.3 1.806 100.00 4.979 Table 3. Incidence of violence in ideological events by ideological tendency of participants ~~ Ideological participants Violcnt events Total events* Yo N . %I N . Extreme left Left Center-left and center Center-right and right Extreme right Others Opposite tendcncies Total 32.2 13.1 0.5 0.9 78.2 13.2 12.0 70.4 419 171 6 11 367 I72 I57 1.314 31.7 18.8 1.3 I .0 22.4 16.4 8.5 100.0 591 35 I 24 18 418 306 158 1,866 * Excludes all events in which no ideological participant was found 623 damage (26% vs 18%), and when violence is directed at people (51% vs zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA40%). Finally, Table zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA5 shows that the Left is more likely to become involved in violence during large group events than is the Right (in 33% of the cases against 12%), while the Right is present in 60% of the small group violent events against 31% for the Left. We therefore obtain a picture of extreme leftwing groups which organize larger events, at which violence can occur and clashes with the police are more likely, while the Right -with smaller numbers and more aggressive tactics zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA- is less likely to encounter police repression. Whether this is a result of police preference for the Right - a common thesis in Italian interpretations [Pas- quino. 19831 - or of more evasive tactics on the part of the Right cannot be deduced from the data at our disposition. Table zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA4. Types of violent forms in ideological events in which extreme left and extreme right participate. Type of event Extreme left Extremc right Total events* "/u N. Yo zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAN. Y" N. Only violent outcomes 1.4 6 1.1 4 1.3 10 Attacks against people 39.6 I66 51.2 188 45.0 354 Violent encounters 41.1 I72 21.8 80 32.1 252 Attacks against property 17.9 75 25.9 95 21.6 170 Total 53.3 419 46.7 367 100.0 786 * Excludes all the events in which no ideological participants was found Table zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA5. Size of participant groups in violent ideological events in which extreme left and extreme right participate. Group size Extreme left Extreme right Total events* Yo N . "0 N. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA'Yo N. Mass violence 32.5 107 11.8 34 22.9 141 Small size group violence 31.3 103 59.7 172 44.5 275 Medium size group violence 36.2 1 I9 28.5 82 32.6 20 1 Total 53.3 329 46.7 288 46.7 617 * Table excludes all events in which: a) an estimate of size could not be given and b) no ideological tendency was mentioned. 626 position. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAImplemerztative events are those which aim at the application of already approved laws or agreements. Proactive evenrs are those seeking new policies or advantages. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAA fourth category includes comperitive zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAevents in which no change is requested or opposed, but a competition with adversaries is expressed.’’ The distribution of violent and non-violent events for these different types of grievances can hc found in Table 7. Not surprisingly, the highest percentage of violent events is found in the actions in which competition between groups is the reason for the event. In a category which covers 14% of protects. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA95% were violent. A different distribu- tion characterized the events oriented toward the activation, implementation or repeal of precise policies. In this case, in fact, the percentage of violent events is always below the average. But also among this group of grievances interesting differenences emerge, especially comparing the two larger catego- ries: reactive and proactive grievances (respectively, 44% and 37% of the total events). While the proactive grievances are generally peaceful zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA- with 82% of non-violent events - for the reactive actions the percentage of peaceful actions decreases to 65%. These data also deserve some brief comments. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAA hypothesis common in the social movement literature explains violence as a result of the reaction of traditional strata to an aggravation in their market condition and of the contemporary push of emergent interests to gain political access. Violence would, therefore, result from the interaction among the social actors who are being excluded and those who push to be included in the political arena [Tilly, 19781. Our analysis gives some support to those statements. I t stresses, in fact, that violence emerges especially during direct competition between adversaries and also that the actors are more likely to choose non-peaceful forms of action when they are defending menaced privileges or acquired rights. A deeper analysis would be needed to define the nature of the groups involved, especi- Tuhle 7. Incidcnce of violent and non-violent evcnts by type of protest grievance. Grievance Non-violcnt cvents Violent events Total cvents 1% N . ‘?o N. % N. Reactive 11.2 l..W 43.3 762 43.9 2.144 Competitive 1.2 37 36.0 633 13.7 670 Proactive 47. I 1.471 18.1 319 36.7 1.790 Others 0.X ‘6 1 .o 17 0.9 43 Implementation 6.7 109 I .6 28 4.9 237 Total 64.0 3.175 36.0 1.759 100.00 4.884 627 ally as far as the relationship between social interests and political actors is concerned. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA V. Some conclusions We have shown in this article that the curve of violence in the Italian protest cycle was by no means isomorphic with the curve of protest as a whole: it rose later, was more likely to take small group forms than to involve masses of people, and was not an important aspect of the intense height of the cycle zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA- between late 1967 and early 1970, when students, workers and others were simultaneously mobilized. Violence was also not an important aspect of industrial conflict, nor of protests in the service or educational sectors (the three most important nodes of struggle at the time), but was mainly limited to young people’s conflicts, mainly of a non-policy variety. zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBAAs conflicts over policy among competing groups and between groups and authorities subsided, violent forms of protest and violent conflicts between groups became more important. The connection between ideology and violence and between abstract issues and violent conflict deserves some underscoring. The over 1,800 events in which we were able securely to identify the presence of an ideological group were far more likely than the sample as a whole to involve violence. Does this mean that ideological politics is inherently violent because its targets are abstract? We found mixed evidence for this proposition. We think it more likely that violence became a mode of expression for small competing groups in a climate of general demobilization. where mass protests were no longer possible or profitable. As for the particular ideologies of violent groups, these were more likely to inspire violence when the group came from the Extreme Right than from the Extreme Left. At least in our period, extreme leftwing groups were more likely to get into trouble with the police in the context of large gatherings, while extreme rightwing ones were more likely to attack property or people. We can try, at this point, to draw some generalizations on the correlation between violence and the type of targets of violent action. With the exception of political institutions, the potential for violence does not seem to increase with a high level of abstraction of the actors whose intervention is requested by protesters. It is instead more correlated to the lack of room for negotiation. The choice of targets with low propensity to bargain-or the absence of a target - can work for social movement organizations as a device to provide an outlet for action with low substantive but high symbolic power. The adoption of violent forms can, in this case, be at least partially explica- ble by their high expressive capacity. Starting a chain of reaction, violent 628 zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA action helps to strengthen collective identity by providing a context in which the group can zyxwvutsrqponmlkjihgfedcbaZYXWVUTSRQPONMLKJIHGFEDCBA- and in fact must - count, which explains why protest leaders often seek out violent encounters. In the spiral of radical interactions, the elevated costs paid in violent tactics discourage participation among the weak- kneed, the pragmatic and the moderate, while at the same time increasing the sense of solidarity among people who have faced state repression and violent enemies together. As far as the aims of collective action are concerned, some provisional conclusions can be offered on the most common grievances found in violent events. There is a tendency toward peaceful forms in those actions aiming at activation or implementation of concrete policies by precise targets. When violence occurs during this type of protest, it usually springs up incidentally during mass actions. Violent occurrences tend, instead. to appear especially during action with no policy goals or where little room for bargaining exists. They are used, in particular, by groups reacting to threats against their acquired positions and are particularly oriented against political adversaries belonging - like them- selves - to the social movement sector. Violent actions can, therefore, be explained in the following ways: as a reaction of social and political interests facing a weakening of their power: as a result of expressive actions oriented to the collective cohesion of political organizations; as an expression of the competition among different groups without the involvement of a precise policy. Furthermore, it is highly probable that these three explanations interact with each other. Finally, what do our data suggest about the later phases of organized terrorism in the mid-to-late 1970s? By then, of course, the political oppor- tunity structure had radically changed, with a Communist party that was entering the governing circle (and itself becoming an object of violence) and an economy that was in so deep a crisis that most organized groups were more concerned with preserving their positions than making new gains [Graziano and Tarrow, 19791. If violence was a central part of mass protest, we would expect it to decrease in this period, but it does not. The reasons can partly be found in the period we have studied. First, since most of the protest groups we studied were not oriented to purposive aims, there is no reason to suppose that economic decline would cool their ardour for protest. Second, a decline in mass protest in itself was a stimulus to violence - which is the only form of serious disruption that is possible in a climate of demobilization. Finally, the competitive protest of our period provided a socialization into violent forms of action for a generation of young people: it was inevitable that some of them would later ‘graduate’ into organized terrorism. ‘Why Italy’, people often ask when they look at the geography of violent