The Federal Republic of Germany, 1949-89

Empirical analyses of the sociopolitical development of the Federal Republic Germany (FRG) which have accumulated over the last decade or so have frequently been summarized in comparative perspective such that Germany has now joined the ranks of 'normal' Western democracies, with all of the positive and negative connotations of 'normality'. This may well be true, but in historical perspective, this claimed normality must be put into question. There are ample examples at hand to show that views on and perceptions of the Federal Republic from the outside are still immensely shaped by the 1933-45 period and that Germany therefore is not just one in the group of 25 or so nations under liberal-democratic rule. But serious doubts against the normality assumption can also be raised from the inside perspective. For one, Germany as a nation is still divided into two states existing under vastly differing political-institutional arrangements, with uncertain prospects with regard to the future relationship between the two states, a fact which creates an element of structural political instability in the heartlands of central Europe. Secondly, the well-known Historikerstreit' the historical uniqueness and the logic of scientific explanation the Nazi reign has - disguised in the appearance of a scholarly debate - brought back to public attention the question whether West Germany can still be characterized by the Unfdhigkeit zu trauern - the inability to mourn - which Alexander and Margarete Mitscherlich in a 1967 book had claimed existed with respect to the German way of dealing with the 1933-45 period of Nazi reign and terror.

In retrospect, it must have been the interaction of biographical factors -
at the age of ten in 1945,1 had been forced to grab some of the essence of
what a war is all about - the development of my own intellectual interests

---

* The 1989 Stein Rokkan Lecture, Bergen, Norway, May 1989.

and the exposure to leading scholars in the social sciences which have brought me to focus my scientific concerns on the conditions favoring and threatening the survival of liberal-democratic regimes. It is in this context that I will address the role of political violence. Although as a comparatively-orientedsocial I am primarily interested in generalizable statements, I will, in the context of this symposium, concentrate empirically on the West German 'case¹.

Political Violence

Gewalt in the German language bears two different connotations which for analytical reasons have to be kept separate: Gewalt in the sense of power (potestas, pouvoir), and in the sense of physical force (vis, violence) (Neidhardt 1986, 114-115; Wassermann 1989, 83-84). It is the notion of potestas which is in the core of Max Weber's definition of the modern (rational) state when he speaks of a 'state as that type of political community which within a defined territory . . . successfully claims the monopoly of physical coercion' (Weber 1964, 1043). Obviously, physical force is only one of the state's means of coercion, and one at that which will usually only be applied as ultima ratio. It is part of the social contract that physical violence is normatively excluded from the societal repertoire of sociopolitical discourse by citizens, organized interest groups and other types of social aggregation.

It may well be in the best interest of citizens freely to give up the resort to physical force in the case of conflict and to transfer that right to the state. Political philosophy has shown, however, that this notion, as developed, for example, by John Locke, has historically evolved at the time of economic take-off when people became increasingly unequal in terms of individual property ownership, and property owners therefore required a state to protect their rights (Ernst 1986, 95). The most important aspect here is to pose the question as to why all citizens alike should transfer their recourse to physical force to the state, or, in other words, to raise the problem of the bases of legitimacy of the modern state.

It may be surprising to recall that the state monopoly on physical force came into being in predemocratic times and has survived encompassing sociopolitical changes virtually untouched, with one major exception pertainingto source of legitimation (Wassermann 1989, 84). It is now the constitutional democratic state with an institutional structure at least normatively wedded to the concept of the separation of powers which regulates in very specific ways, through laws and legal procedures, the conditions under which the democratic state is permitted to exercise its monopoly on physical force against its citizens. The instances where this

happens are usually regarded as extreme, therefore commanding the highest conceivable amount of attention by the public and, as core representativesof public in the non-political sphere, the system of mass communication.

Obviously, the way in which the state exercises its monopoly of physical force against the citizens is only one important element - albeit an important one - of the legitimacy of the modern democratic state. There is no common theoretical or normative basis available defining a priori the boundaries of what belongs to the state and what is/should be beyond its reach. The debate on the limits of the welfare state is a good case in point here. Whatever the boundaries are at a given historical point in time will, however, provide citizens with the yardsticks to evaluate the political structures, processes and outputs and to arrive at an assessment of how legitimate the system in question is.

Three brief references to central elements of the legitimation process in modern liberal democracies need to be made here. Firstly, obvious as it seems, the legitimacy of a given polity is based on the free consent of those governed: unlimited political coercion as functional equivalent to consent is on a priori grounds not available.

Secondly, innate to the logic of the democratic system are institutionalized of creating, voicing and channeling political dissatisfaction. transformation into action is accomplished by the political authorities or, in mid- to long-range perspective, by changes in the political coloring of political authorities. Authorities represent the lowest level of legitimacy objects. It is this institutionalized process of potential changes in political authorities which leaves higher-level objects - the political regime and the political community (Easton 1965,1975) - most likely intact as buffers between levels of legitimacy objects (Fuchs 1989; Westle 1989).

Thirdly, the extent to which this buffering is successful depends on a variety of factors, like the quality of system outputs, inclusiveness of the system vis-å-vis newly emerging interests (Powell Jr. 1982), strength of identification with the regime and political community, and so on. Paul Sniderman (1981) has pointed to the fact that the 1980s can be characterized by the eclipse of alternate images to the democratic order, thereby providing an additional element of political stability on the legitimacy object level of the political regime.

It remains an open and empirical question as to what extent claims abounding in the 1970s that the legitimacy of the liberal-democratic state were on the wane are warranted. Clearly, it lies in the logic of the democratic political order that a weakening of the state monopoly on political violence resulting in an increase in the level of political violence by social forces (citizens, social organizations, ad hoc groups) must be regarded as an indication of strain and a corollary as well as an indication of an ongoing

process of delegitimization of the political system. However, without welldevelopedtheories political transition we lack the important information regarding what levels of political violence by the state and/or societal actors can be regarded as quasi-normal and which as threats to system stability - stability being conceptualized here as change in the politico-constitutional order. Nevertheless, a quick glance back at the sunset of the Weimar Republic will easily bring back to mind the intense and frequently violent struggle between forces from the extreme left and the extreme right which preceded and helped to pave the way for Hilter's surge to power in 1933. But there is also more contemporary reason for concern. News about fierce encounters in West Berlin between demonstrators and the police on 1 May 1989, leaving roughly 300 policemen injured, have traveled around the world, and this is only the last in a long series of violent political clashes which have now plagued the Federal Republic for some time.

Political Violence in Historical Perspective

The path-breaking scholarly venture of the late Stein Rokkan (Rokkan 1973, 1974, 1980; Flora 1981; Tilly 1984, 129-143) to develop conceptual maps of Europe helping to understand the historical conditions behind the varied processes of European nation building between the seventeenth and nineteenth centuries also foresaw a place for political violence as an important element in these processes, although to the best of my knowledge this place has not been systematically developed by Rokkan. One reason for that may have been that the politically motivated collective use of physical force between people and/or groups and against people by the state or its functional equivalents and vice versa was so widespread, so normal, that it did not command special attention until acts of political involvement became more thoroughly embedded in a system of law, binding both governors and governed. Therefore, political violence was increasingly replaced by other, more peaceful types of political action as the process of enfranchisement and the establishment of equal political rights for all citizens continued.

A second, related reason may have been that, taking the intellectual emphasis of Stein Rokkan on analyzing nation building in historical terms, the extent of political violence as a variable was simply not significant enough for him in his quest better to understand the conditions determining the specific path of the various European nations into political statehood.

One piece where he seems to address the question of political violence in a more systematic fashion is his effort to reconcile his own thinking (Rokkan 1974) with the work by Albert Hirschman on Exit, Voice and Loyalty (Hirschman 1970). With the establishment of national states with

firm external boundaries, the option of physical exit for dissatisfied or disadvantaged persons became less and less easily available. As a consequence,and by the process of enfranchisement, politics turned more and more into, on the one hånd, creating loyalty and, on the other hånd, dealing with voice. Here, degree and length of territorial identity, continuity in representative organizations of interest intermediation and the existence of strong centers were among those factors Rokkan felt were most conducive to the orderly, peaceful transition to voice politics.

Rokkan ended his tentative analysis by concluding that the polities with the least legitimate territorial centers - the German Reich, Italy, Austria and Spain - were exactly those where the road to the politics of voice was accompanied with the highest levels of civil violence.

The politics of voice in the process of democratization, so Stein Rokkan argued, had to overcome barriers of legitimation, of incorporation, of representation and of executive privilege of parliaments, and it may well be worthwhile to recall that only in 1971 did Swiss women become entitled to vote in national elections, and that quite a few countries waited until the 1920s until full and equal participation in the vote was institutionalized. But for the liberal-democratic polities under scrutiny it can be maintained that the legal and ideological foundations of political equality have been firmly established for some time now. Thus, the question needs to be raised in the light of ongoing political violence in these states whether a basically similar logic can be applied for the explanation of political violence in the highly developed liberal democracies as it could be at times where the states where political violence occurred were operating under vastly different and political conditions, as we have seen.

Definitional Problems and Theories of Political Violence

I have as yet consciously abstained from the thorny problem of defining violence. In general, we can easily agree that in the social sciences definitions synthetic constructs embedded in a specific theoretical approach to any given phenomenon to be studied. Thus, there seems little point in arguing about the Tightness or wrongness of any particular definition. Definitions, however, are extremely consequential in that they structure the world to be empirically assessed and analyzed. Therefore, the definitional cannot be completely avoided here.

Violence is a term that, at least in the German scholarly as well as in the
public and political debate, has revealed substantial ambiguities. It is
therefore useful to refer to some of these ambiguities.

Generally, there seems to be agreement in the literature that violence is

any use of physical force against material objects and/or persons. Charles Tilly (1978, 174) sticks to this concept of violence which he calls intermediate,in to a narrow concept involving additionally the legitimacydimension, to a broad concept which seems basically equivalent to the Galtung (1969) notion of violence which has become broadly known as 'structural violence'. Political violence, as a more specific case in the general violence class, usually extends into the goal dimension - violent action is directed toward achieving political goals - and into the object dimension - targets of political violence are collective or individual political actors (the government, a specific public official and so on). Prototypically, such a definition reads as follows: 'Collective political violence involves destructive attacks by groups within a political community against its regime, authorities, or policies' (Eckstein 1980, 137; derived from Gurr 1970, 3-4).

It would hardly be worth dwelling on these definitional matters any further, were it not for the Galtung notion of structural violence which has by now gained such wide currency and certainly has had a substantial impact on the German debate on violence. He argues in the context of peace research that peace is the absence of violence, and that - in line with his theoretical impetus - he needs a wide definition of violence, namely

. . . that violence is present when human beings are being influenced so that their actual somatic and mental realizations are below their potential realizations. Violence is ... the cause of the difference between the potential and the actual, between what could have been and what is, violence is that which increases the distance between the potential and the

actual, and that which impedes the decrease of this distance (Galtung 1969, 168).

Whereas, frequently on ideological grounds, scholars have in large numbers with the Galtung concept, it has had little impact on empirical research on political violence - and, it seems to me, for good reasons. If one would take this concept seriously, it would cover such a wide array of phenomena that it would lose all analytical power; it would become a catchall But, in addition, as Friedhelm Neidhardt (1986) argues in reference to Niklas Luhmann (1972), such a definition would also miss the theoretically most important element of the violence phenomenon: namely, that violence as physical force is a universal language, independent of the level of cultural and social differentiation and development, cuts across all social structures, norms and values. In sum, violence is an instrument of social and political control which, because of its directness, supersedes all other instruments of control (Neidhardt 1986, 134).

As a consequence, I use the term 'political violence' always in the sense
of application of physical force by individuals and groups against other
individuals or material objects in a political context.

In passing, it should be noted that the Vergeistigung (as Neidhardt (1986,
122) calls it), of the concept of violence has made its mark also on the

West German legal system, thereby rendering even more ambivalent the association of 'violence' with physical force. This would not be particularly worth mentioning except that it is further proof of the claim that 'violence' has gained the status of a Kampfbegriff, and that terminological differences do not only reflect differing scholarly, but also differing political, views. We shall return to this point later in the paper.

There is an enormous wealth of studies dealing with political violence. These studies, on the one hand, can be differentiated according to the data bases used. A large set of analyses dwells on some variety of event data where types of violent occurrences are coded from standard sources (like the New York Times Index) according to theoretical considerations. The most widely used data base of this kind is the Yale Political Data Program with the publication of the World Handbooks of Political and Social Indicators (see for example, Taylor & Jodice 1983); these data are freely available from all academic data archives for secondary analysis. The major advantages of this data set are that the data are produced according to stable and high standards of source and coding, that they are longitudinal and that they are comparative.

As an alternative to event data, there are, of course, survey studies. There is neither the need nor the time to elaborate on these studies in greater detail. Let it suffice to say that the basic advantage of the survey data is that they are tailored to the micro level of analysis and can therefore obtain not only objective, but also subjective, not only retrospective and present, but also prospective information on political actors. Since most theories of political violence involve individual actors at least in some part, survey data are an indispensible tool for the proper empirical testing of such theories. The major disadvantages, like a mirror image to the event data, are that, in particular for reasons of cost, the sample size in surveys is usually much too small, and the studies are neither longitudinal nor comparative. An additional problem related to that of the small sample sizes is that political violence is often concentrated with specific subgroups of the general population which cannot be properly represented in national surveys because of the small numerical size of the subgroup in question and because, for obvious reasons, members of such subgroups are not the most likely ones to participate in survey research as respondents.

Of course, in many countries there are other statistics on violent events of various kinds although experts usually are very skeptical regarding the quality of these data. All data types mentioned, though, suffer from one deficiency which is particularly relevant in studying (political) violence: Violence is uninstitutionalized behavior which emerges always - with the probable exception of political terrorism which is not under scrutiny here - from the more or less complex interaction of individuals. Thus, the unit of analysis in studying political violence should mostly - or at least also - be

relationships (see also Tilly 1984, 26-33, who phrases the unit of analysisproblemin basic terms). Information on interaction sequences which result in political violence, however, are scarcely ever available. If they are available at all, then they are either on one or on a few event sequences at best, or they are rudimentary ex post reconstructions with a high degree of theoretical and empirical selectivity.

There is little point in reproducing here detailed results of studies on political violence which all in some way or another give proof of the complexity of facilitating, accelerating, decelerating and non-conducive factors involved in the occurrence of political violence. Rather, I shall follow the route paved by Harry Eckstein in 1980 and followed by Tilly in 1984 (50-59) in grouping these studies into two categories: studies wedded to the contingency paradigm, and studies wedded to the inherency paradigm. The contingency paradigm starts from the basic assumption of man and woman being peace-seeking organisms; therefore, deviations from the peace route reflect basically some kind of non-normality in society, that is they are indicative of specific detrimental social conditions, e.g. fast social change. The relative deprivation construct and its derivates are therefore in the explanatory core of violence studies in the contingency vein. By contrast, the inherency paradigm starts from the assumption of man and woman as power-striving organisms. For this paradigm, violence as an expression of power-seeking is basically normal; what needs to be understood is why it does not occur more frequently. The studies covered by this paradigm more or less belong to the family of economic, costbenefit of political choice. James de Nardo's book on Power in Numbers (1985) is a formally and particularly highly developed prototype of this approach.

I shall return to these paradigms and the related research later on. One note of caution regarding past comparative macro research on political violence is that it often embraces authoritarian and democratic types of political order, a difference which obviously calls for theoretical consideration. one study using event data similar to the Taylor-Jodice handbook is particularly worth mentioning which concentrates on contemporary the book with the same title by Bingham Powell Jr. (1982). One of the most interesting findings there with respect to political violence is that violence is less likely to occur in states which rank high on representational and consociational institutions and practices (Powell Jr. 1982, 212-225). This finding strikes a familiar chord with Stein Rokkan's above-mentioned thoughts (1974) on the historical conditions inducing collective violence in the process of democratization and may therefore give additional weight to these analyses.

If these arguments are legitimate, then they would lend support to the
inherency paradigm, because blocking of legitimate channels of influence

is supposed to force alternate ways of seeking voice on political actors (Eckstein 1980, 143). All in all, Eckstein (1980) as well as Tilly (1984) find the evidence available from research too inconclusive to warrant a clear decision in favor of either the contingency or the inherency paradigm.

The Structure of Political Action

The process of democratization bore two core elements: on the institutional/organizational level the establishment of constitutionally embedded democratic procedures, of legal and legitimate organizations of interest intermediation, and of a competitive party system; on the individual level the installment of basic individual rights and the institutionalization of equal political participation through the vote. Beginning in the late 1960s we have witnessed, based on unprecedented economic growth and on revolutionary changes in the systems of education and mass communication, the surge of a desire to obtain for the citizenry a larger say in political matters. The Political Action Study of the early 1970s (Barnes et al. 1979; Jennings et al. 1990) was the first comparative survey project to study this surge empirically; it has covered eight nations and has had a sizable number of mostly national follow-ups (for the West German case see Infratest 1980; Schmidtchen 1983; Uehlinger 1988).

The major research questions to be answered by Political Action were whether and how the dimensionality of political participation had changed, and whether the emerging uninstitutionalized forms of political involvement were integrated in the already existing political repertoire or were forming a distinct cluster of actions directed against the logic of the representative political process as it existed in the liberal-democratic states. Political violence was included in the study but was for a long time not analyzed at all for a variety of theoretical as well as empirical-practical reasons (e.g. small number of cases with affirmative answers to the violence items).

The basic results of that study were that conventional, electorally oriented political participation was supplemented by a second dimension of uninstitutionalized,unconventional participation, and that these two dimensions were related to each other in such a fashion that for a large part of the populus a new, broadened repertory of political action emerged. Furthermore, it proved important and worthwhile to distinguish between an attitude element of political participation pertaining to the two aspects of approval of concrete types of action and of behavioral intention in the sense of basic willingness to participate, and the behavior in question itself. These findings turned out to be stable over time, surviving a replication by the Political Action Group around 1980 (Allerbeck 1980; Jennings et al. 1990; Kaase 1990a) as well as replications with other populations (Infratest

1980; Klingemann 1985) and with different instruments (Uehlinger 1988), with one exception. Further dimensional analyses revealed that it made theoretical as well as empirical sense to differentiate the dimension of uninstitutionalized political participation into the two subdimensions of legal and illegal involvement - the latter also being called civil disobedience and to add to those separately the violence dimension (Fuchs 1984; Uehlinger 1988). Analyses building on this differentiation indicated that there was a deep cleavage line between legal and illegal forms of direct participation, and that there existed a link between civil disobedience and political violence.

The Acceptance and Public Understanding of Political Violence

In a recent (early 1989) representative survey study of the West German population (14 years and older) on behalf of the Governmental Commission on the Study of Violence, respondents were asked to assess, via a semantic differential, the general meaning of violence. With a high degree of consensus, a seven-point scale West Germans felt violence to be dangerous (x = 6.47), ugly (x = 6.41), bad (x = 6.29) and unnecessary (x = 5.94). Thus, it cannot come as a surprise that in the Political Action Study, at the two points in time, of those who were 16 years and older none of the eight countries had more than 2.4 percent (x = 1.1 percent) of the respondents approving of violence against material objects, and more than 2.7 percent (x = 1.9 percent) approving of violence against persons. Violence thus seems to be highly tabood in Western democracies.

However, other elements of the semantic connotation of violence in general shed at least some slight doubts on that conclusion. There are almost as many who think of violence as being weak as there are those thinking it to be strong (x = 3.96), more think of it as exciting than think of it as boring (x = 4.42), and those who look at violence as ineffective have only a slight edge over those looking at it as effective (x = 4.51). In other words, whereas violence as such carries, in evaluative terms, a strong negative emphasis, it holds a position (not at all surprisingly) in the minds of the populus of attention and excitement that make violence an ideal object for public and media interest and thereby for instrumental meansends

The Blumenthal study (Blumenthal et al. 1972) among American white males in 1969 has revealed some of the mechanism by which objectively violent acts of political participation are exempted from the violence taboo: by simply labeling these acts as non-violent, and doing this in such a selective way that one's own violent acts and those of the groups with which

one identifies are not regarded as violence, whereas disliked non-violent acts of others and of groups looked at with hostility are regarded as violence. These labeling approaches seem to be rather universal and well-founded in psychological theory (Mummendey et al. 1982); however, they may also offer an explanation why the definitional and terminological conflict over the concept of violence plays such an important role in the political arena, at least in the Federal Republic. The more violence-prone actors succeed in that debate to weaken the precision of meaning attached to the violence term, the more likely it is that they can justify their own resort to physical force by referring to the omnipresence of the violence phenomenon.

However, since the terminological fight is not yet decided in favor of an encompassing concept of violence as all kinds of power manifestation, the problem of having to find legitimation for this largely tabood type of behavior is still to the fore.

Before addressing the legitimation topic briefly, it seems worthwhile to report that in terms of public knowledge and perceptions there is a substantial of ambiguity regarding political violence. I have not yet been able to assess the data on a respective open-ended question in the necessary detail. Nevertheless, it is apparent that only a small subset of the population possesses a conceptually well-anchored idea of political violence, whereas the majority derives its understanding mostly from concrete events, in particular from demonstrations (this is living proof of the importance of the mass media, especially television, for the way in which citizens perceive the political process). In particular, the data show that political violence is not at all exclusively related to the application of physical force for political purposes. Thus, there is no common cognitive and verbal ground on which to walk when a democratic polity like West Germany deals with the violence phenomenon. Once again, we are confronted with a case of elite politics, and when the democratic elites are not in agreement on what violence is and that and why it should be outlawed as a legal and legitimate means of political involvement, the field is opened wide for conflict.

On the Legitimation of Political Violence

Between 1974 and 1980 approximately two-fifths of the West German population (16 years and older) felt that necessary changes in the society could only be brought about swiftly when violence was used. It is unfortunatethat the other countries in the Political Action Project comparableinformation not available to place the German findings into perspective. These opinions must also not be automatically interpreted as personal support for political violence, but rather as a cognitively based assessment of the democratic political process, and one that is not all that

ill-founded considering both the procedurally innate slowness of decision making in democracies and the ever-present evidence that it requires unconventional, provocative and controversial means to achieve immediate action by authorities or at least some kind of action at all as response to citizen complaints. In the above-mentioned 1989 study, 36 percent of the respondents could personally imagine circumstances under which citizens might resort to political violence. With this kind of question the net is cast wide indeed, and differentiation is required. Fortunately, the answers to the follow-up open-ended question regarding what the specific circumstances justifying violence are permit a realistic assessment of the legitimating rationales present.

Of those 36 percent of imagining circumstances of personal political violence by citizens, roughly one-half legitimizes it by referring to a situation in which citizen rights are violated, or with a necessity to counteract dictatorial regimes. This certainly is very much related to specific German history, although one can observe that the constitutional right to violent resistance against an antidemocratic coup (Article 20.4 in the Basic Law) in the present public debate is quite easily claimed, and this claim frequently fails to take into consideration the high constraints attached to this constitutional However, the other half of respondents legitimizing personal do so on the basis of discontent with the way the government handles certain policy issues or because of a lack of government responsiveness general. The responses are linked through an instrumental notion of political violence; if 'normal' procedures of influence seeking do not succeed, an escalation in the means of political expression is regarded as appropriate and obviously quite promising.

These findings are corroborated by answers to a set of close-ended questions first asked in 1980 of respondents between 16 and 35 years of age by Schmidtchen (1983; see also Uehlinger 1988) and replicated in 1989, pertaining to the willingness to engage in illegal political actions - civil disobedience - in the case of a lack of authority responsiveness. (In Table 1, numbers in parentheses refer to past participation in these actions.)

Data from these same studies on legal uninstitutionalized forms of politicalparticipation

iticalparticipationsustain the findings from the Political Action Project that a large-scale extension of the political repertory of citizens in democraticpolities taken place. This is clearly a corollary to the process of political mobilization that Dalton (1988), in his comparative analysis, has described for France, Great Britain, the USA and West Germany. But the willingness of a sizable extent to engage also in acts of illegal uninstitutionalizedparticipation questions with regard to the public perceptionof a representative democracy is all about. And this question even remains to the fore if one takes into account that, as mentioned before in the context of the Political Action Study, it is but a small percentage of respondents who report actually having engaged in the past in some of the above actions. What are the sources of this general development?

The Value and Ideological Sources of the New Politics

The famous piece by Lipset & Rokkan (1967) on the evolution of cleavages and party systems in European democracies has triggered an enormous wealth of ensuing studies. Sociopolitical changes originating in the 1960s have since then pushed into the foreground the scholarly debate regarding to what extent the Lipset-Rokkan conclusions still possess currency in the late 1980s (Dalton et al. 1984; Crewe & Denver 1985) in the light of an ongoing process of value change towards postmaterialist values (Inglehart 1971, 1977, 1989). This in parts rather fierce debate cannot be taken up here. However, certainly in the West German case (Bauer 1989; Eckert et al. 1990; Fuchs 1989; for The Netherlands see van Deth and Geurts 1989) the integration of the Green Party into the West German party system has been facilitated by a slowly but consistently mounting number of left postmaterialists which now constitute between 10 and 15 percent of the electorate. In sociostructural terms, this group can be characterized by young age, high formal education and a lack of integration into the traditionalinstitutions society (work, family/marriage, religion). In issue preference terms, left postmaterialists place an extremely high priority on peace and on the environment - almost by definition. They are strong supporters of the welfare side of the modern state and hostile to its elements of social control. Finally, in terms of modes of political action preferred, they have pushed the extension of the political repertory to its present extreme, including (at a rate of up to 50 percent) even the willingness to engage in illegal acts of political participation. While the left postmaterialistsare group by far most prone to action, it must be kept in mind that analyses show that the propensity to engage in uninstitutionalized political participation is at present strongly and almost linearly related to

both the left-right dimension and the postmaterialism-materialism dimension.By to the left postmaterialists, it is always the group of right materialists which is least prone to action (Bauer 1989; Eckert et al. 1990). And this schism carries on into the violence domain. Some 63 percent of the left postmaterialists, but only 20 percent of the right materialists can imagine circumstances legitimizing the use of violence. Some 28 percent of the left postmaterialists and 7 percent of the right materialists are willing to consider personal violence for normal political purposes in the case of a lack of response by the authorities to substantial citizen demands.

Political Violence and the Interaction Process

Attitudes towards political participation have been shown to become less relevant the closer one gets to action (Kaase 1990a). This finding emphasizes the role of the process of mobilization into action, in particular the important of network integration - the strength of weak ties, to dwell on the title of a well-known piece by Granovetter (1973; see also Pappi 1990). However, for political action to turn violent additional factors need to be brought in. Most importantly, political violence thrives on interaction and confrontation, be it between hostile groups or between a given group and political authorities, in particular the police. This is why thinking about the reduction of violent confrontations by democratic authorities needs to center so much on police strategies, measures that are under police control (Eckert et al. 1990).

Much of the previous work on the impact of repression on collective violence has produced inconclusive results (Opp 1988a). Obviously, and quite trivially, under non-revolutionary circumstances there exists a point beyond which the repressive power of the state exercising its monopoly of physical force is so dominant that violence no longer takes place. Not only is it the case, though, that democratic states are not in a position to use repressive tactics at will and at liberal measure; there is a low threshold beyond which state violence may become counterproductive in that it delegitimizes political authorities, particularly if the goals to be achieved by citizens are regarded as highly legitimate by the populus (Neidhardt 1989). This consideration points to the important intervening factor of public opinion and the mass media. It is in line with this argument that protest strategies by social groups are frequently directed at provoking disproportionately high repressive action by the authorities. Such action has been shown empirically to have a radicalizing effect on protest participants(Opp and furthers high network integration. As mentioned before, this solidarization raises substantial problems for democratic polities which cannot easily push repression to a point where violence no longer

takes place because its costs have become too high for protestors. James de Nardo (1985, 188-228, 243-264) has shown that for each side in the confrontation process it is extremely difficult, in faet impossible, to optimize the respective calculations of the amount of violence or repression to be used. This may explaim why so frequently such confrontations run out of hånd.

Clearly, as a summarizing note on this aspect of the violence phenomenon, is evidence that especially social groups at present use political means - including violence - in a strategic fashion (Opp 1988b). If these results will stand even after further rigorous testing, this would lend strong support indeed to the inherency hypothesis for the explanation of contemporary violence in democratic polities. More emphasis along the same lines can be derived from resource-mobilization theory (Zald & McCarthy 1987) in the context of the new social movements. This theoretical explicitly starts from the assumption that relative deprivations be created by social-movement entrepreneurs, thereby once again stressing the rational-choice basis of political action. This aspect is so relevant here because the new social movements supply network resources for political action, especially for postmaterialist goals, and might establish themselves even as alternatives, and certainly as supplements, to traditional channels of political intermediation (Kaase 1990b).

Violence and the Democratic State

Let me start my concluding remarks with a note of caution. The Political Action Project, because of its comparative approach, was in a position to present its findings and conclusions on a very reliable basis; too quick a generalization from country-specific findings was not a problem. By contrast, the data on attitudes toward political violence which I have presented here stem from but one nation. However, the findings on Germany have accumulated over time are highly consistent across different data bases. Thus, the question arises as to what extent these results can indeed be generalized. Bergen, the university where Stein Rokkan has worked and has contributed so much to comparative social science, is the most proper place to suggest that more comparative research on contemporary political violence in liberal democracies gets under way.

I am coming full circle when I argue that political violence is a challenge to the democratic political order. This challenge takes various directions. One target group for concern is the citizenry which in part apparently finds it difficult to reconcile its desire for quick political action in areas deemed absolutely central for survival with the slowness, ambiguities and compromisesso of the democratic political process. The majority principle

is increasingly understood as preventing necessary change and is questioned on these grounds. Many who do so fail to understand that the combination of minority rights and majority decisions can only be equally guaranteed if embedded in the legal and procedural framework of the constitutional democratic state (Kielmansegg 1988, 97-131). Violence has no part and no role in it, except in the figure of the Gewaltmonopol of the state which is indispensable for regulating conflicts in an open society in a peaceful way.

The second target is the democratic political process and the role political authorities play in this process. The US constitution is built on the firm belief of the frailty and fallibility of human existence. Such a well-taken skepticism finds its expression in the emphasis on institutional arrangements of checks and balances which structurally limit the opportunities for citizens to transform their individual frailty into the frailty of the political system at large.

But institutions, as Samuel P. Huntington remarked some time ago (1974), are not only reflections of iron principles of political organization, but may also lose their compatibility with the way societies operate in times of encompassing social change. Exactly ten years ago Kaase and Barnes (Barnes et al. 1979, 523-536) concluded from their analysis of the Political Action data that the desire by citizens to participate more fully in the process of political decision making is deep-rooted and not reversible. They saw the problem of newly arising political inequality through selective selfinterested resourceful participation beyond the vote. They suggested these ambivalent developments should be used to the advantage of the democratic process by institutional fantasy, i.e. by fantasy with regard to inventing institutions enabling more, but nevertheless equal, political participation.

It is an open question how successful such efforts - if they come about at all - can be. Peter Graf Kielmansegg (1988, 41-74) even turns the argument around. He says that whatever institutional mechanisms for the self-determination of the people may be envisaged, they will not overcome the problem of the political division of labor, the need for representative structures of decision making.

His reflections, however, are contingent on a second principle: modern representative democracy cannot exist without political offices, the holders of which operate under legal conditions and constraint, are responsible for the common good and are subjected to regular popular control (Kielmansegg 58-64). Considering the series of political scandals the Federal Republic and other democratic states (e.g. Japan) have suffered over the last few years, it seems worthwhile to reflect more also in politicalscience than is presently done on this precondition for liberaldemocratic

Here seem to be two linkages to the problem of contemporary political

violence. If existing channels of decision making are by large parts of the population no longer deemed satisfactory, and if in the views of the citizenry the performance and style of political officials increasingly fall short of the standards of a democratic polity, then in the light of the enormous problems facing humankind these days it can surprise nobody if, as inherency theories would predict, citizens think of violence as one means of finding attention, to be heard and to get things done in their way.

Successful use of citizen violence possesses the built-in tendency to inflate its use. Such an inflation will then require more repressive measures by the state in order to raise the costs of violence for the actors. These measures, however, are extremely consequential for democratic polities because in a spiraling process they bring the state dangerously close to crossing the border from liberal to authoritarian or totalitarian rule. The consequence to be drawn from this consideration is that the political process has to be organized and legitimated in a way such that in terms of outcomes there is no success premium for political violence by citizens and social groups.

Inherency theory tells us that violence is part of the condition humaine and will therefore stay with us as an element of sociopolitical normality, in the sense of the Durkheimian theory of anomy, in the family, in school, in sport as well as in political life. However, the extent to which we have discovered positive affinities to civil disobedience up to the point of political violence in the Federal Republic gives reasons for concern, and this not the least in the light of the totalitarian episode in Germany which indirectly has served as an occasion for this week's meetings.

REFERENCES

Allerbeck, K. R. 1980 Politische Ungleichheit. Ein Acht-Nationen Vergleich. Opladen:
Westdeutscher Verlag.

Barnes, S. M., Kaase, M. et al. 1979. Political Action. Mass Participation in Five Western
Democracies. Beverly Hills: Sage.

Bauer, P. (1989). Der ideologische Unterbau politischer Beteiligung in der Bundesrepublik
Deutschland. Eine empirische Untersuchung politischer Überzeugungssysteme'. Dissertation,
fur Sozialwissenschaften, Universitåt Mannheim.

Blumenthal, M. D., Kahn, R. L., Andrews F. M. & Head K. B. (1972). Justifying Violence.
Attitudes of American Men. Ann Arbor: Institute for Social Research.

Crewe, I. & Denver, D. (eds.) 1985. Electoral Change in Western Democracies. Patterns and
Sources of Electoral Volatility. London: Croom Helm.

Dalton, R. J. 1988. Citizen Politics in Western Democracies. Public Opinion and Political
Parties in the United States, Great Britain, West Germany and France. Chatham, New Jersey:
Chatham House Publishers.

Dalton, R. J., Flanagan S. L. & Beck, P. A. (eds.) 1984. Electoral Change in Advanced
Industrial Democracies. Realignment or Dealignment? Princeton: Princeton University
Press.

van Deth, J. & Geurts P. A. T. M. 1989. 'Value Orientation, Left-Right Placement and
Voting', European Journal of Political Research 17, 17-34.

Easton, D. 1965. A Systems Analysis of Political Life. New York: John Wiley.

Easton, D. 1975. 'A Re-assessment of the Concept of Political Support', British Journal of
Political Science 5, 453-457.

Eckert, R., Kaase, M. & Neidhardt, F., with Willems H. 1990. Erstgutachten der Unterkommission
(Soziologie) der Unabhångigen Regierungskommission zur Verhinderung und
Bekåmpfung von Gewalt. Berlin: Duncker & Humbolt.

Eckstein, H. 1980. 'Theoretical Approaches to Explaining Collective Political Violence', in:
Gurr, T. R., (ed.), Handbook of Political Conflict. Theory and Research. New York: The
Free Press, 135-166.

Ernst, W. W. 1986. Legitimationswandel und Revolution. Studien zur neuzeitlichen Entwicklung
Rechtfertigung politischer Gewalt. Berlin: Duncker und Humblot.

Flora, P. 1981. 'Stein Rokkans Makro-Modell, der politischen Entwicklung Europas: ein
Rekonstruktionsversuch', Kolner Zeitschrift fur Soziologie und Sozialpsychologie 33, 397-436.

Fuchs, D. 1984. Die Aktionsformen der neuen sozialen Bewegungen, in Falter, J. W., Fenner,
C. & Greven M. T. (eds.), Politische Willensbildung und Interessenvermittlung. Opladen:
Westdeutscher Verlag, 621-634.

Fuchs, D. 1989. Die Unterstutzung des politischen Systems der Bundesrepublik Deutschland.
Opladen: Westdeutscher Verlag.

Galtung, J. 1969. 'Violence, Peace, and Peace Research', Journal of Peace Research 6, 167—
192.

Granovetter, M. S. 1973. 'The Strength of Weak Ties', American Journal of Sociology 78,
1360-1380.

Gurr, T. R. 1970. Why Men Rebel. Princeton: Princeton University Press.

Hirschman, A. O. 1970. Exit, Voice and Loyalty. Cambridge, Mass.: Harvard University
Press.

Huntington, S. P. 1974. 'Postindustrial Politics: How Benign Will It Be?', Comparative Politics
6, 163-191.

Infratest 1980. Politischer Protest in der Bundesrepublic Deutschland. Stuttgart: Kohlhammer.

Inglehart, R. 1971. 'The Silent Revolution, in Europe: Intergenerational Change in Postindustrial
American Political Science Review 65, 991-1017.

Inglehart, R. 1977. The Silent Revolution. Princeton: Princeton University Press.

Inglehart, R. 1990. Cultural Change. The Impact of Economic and Sociopolitical Change on
Culture and the Impact of Culture on Economics, Society and Politics in Advanced Industrial
Society. Princeton: Princeton University Press (in print).

Jennings, M. K. et al. 1990. Continuities in Political Action. A Longitudinal Study of Political
Orientation in Three Western Democracies. Berlin: Verlag Walter de Gruyter.

Kaase, M. 1990a. 'Mass Participation', in Jennings, K. M. et al. Continuities in Political
Action. A Longitudinal Study of Political Orientations in Three Western Democracies.
Berlin: Verlag Walter de Gruyter, 23-64.

Kaase, M. 1990b. 'Social Movements and Political Innovation', in Dalton, O. & Kuechler,
M. (eds.), Challenging the Political Order. Oxford: Oxford University Press (in print).

Kielmansegg, P. Graf 1988. Das Experiment der Freiheit. Stuttgart: Klett-Cotta.

Klingemann, H-D. 1985. Formen, Bestimmungsgrunde und Konsequenzen politischer
Beteiligung. Ein Forschungsbericht. Berlin: Zentralinstitut fiir Sozialwissenschaftliche
Forschung der Freien Universitåt Berlin.

Lipset, S. & Rokkan, S. 1967. 'Cleavage Structure, Party Systems, and Voter Aligments. An
Introduction', in: Lipset, S. M. & Rokkan S. (eds.), Party Systems and Voter Alignments.
New York: The Free Press, 1-64.

Luhmann, N. 1972. Rechtssoziologie Band 1 (rororo-Studium, Band 580). Reinbek bei
Hamburg: Rowohlt.

Mitscherlich, A. & Mitscherlich M. 1967. Die Unfåhigkeit zu trauern. Grundlagen kollektiven
Verhaltens. Miinchen: Piper.

Mummendey, A. et al. 1982. 'Aggressiv sind immer die anderen. Plådoyer fiir eine
sozialpsychologische Perspektive in der Aggressionsforschung', Zeitschrift fur Sozialpsychologie
174-193.

de Nardo, J. 1985. Power in Numbers. The Political Strategy of Protest and Rebellion.
Princeton: Princeton University Press.

Neidhardt, F. 1986. Gewalt - Soziale ESedeutungen und sozialwissenschaftliche Bestimmung
des Begriffs, in Bundeskriminalamt (ed.), Was ist Gewalt? Auseinandersetzungen mit einem
Begriff, Band 1. Wiesbaden, 109-147.

Neidhardt, F. 1989. 'Gewalt und Gegengewalt. Steigt die Bereitschaft zu Gewaltaktionen mit zunehmender staatlicher Kontrolle und Repression?', in Heitmeyer, W., Moller, K. & Siinker, H. (eds.). Jugend - Staat - Gewalt. Politische Sozialisation von Jugendlichen, Jugendpolitik und politische Bildung. Weinheim - Miinchen: Juventa, 233-243.

Opp, K-D. 1988a. 'Sanctions as Incentives for Political Movement Participation', Institut flir
Soziologie, Universitåt Hamburg, Mimeo.

Opp, K-D. 1988b. 'Grievances and Participation in Social Movements', American Sociological
Review, 53, 853-864.

Pappi, F. U. 1990. 'Neue soziale Bewegungen und Wahlverhalten in der Bundesrepublik', in
Kaase, M. & Klingemann, H-D. (eds.), Wahlen und Wåhler. Analysen aus Anlags der
Bundestagswahl 1987. Opladen: Westdeutscher Verlag (in print).

Powell Jr, G. B. 1982. Contemporary Democracies. Participation, Stability and Violence.
Cambridge Mass: Harvard University Press.

Rokkan, S. 1973. 'Cities, States, and Nations. A Dimensional Model for the Study of Contrasts
in Development', in Eisenstadt, S. N. & Rokkan, S. (eds.), Building States and Nations:
Models and Data Resources, Volume 1. Beverly Hills and London: Sage, 73-97.

Rokkan, S. 1974. 'Entries, Voices, Exits. Towards a Possible Generalization of the Hirschman
Model', Social Science Information 13, 39-53.

Rokkan, S. 1980. 'Eine Familie von Modellen fiir die vergleichende Geschichte Europas',
Zeitschrift fur Soziologie 9, 118-128.

Rokkan, S. with Campbell, A., Torsvik, P. & Valen, H. 1970. Citizens, Elections, Parties.
Oslo: Universitetsforlaget.

Schmidtchen, G. 1983. 'Jugend und Staat. Übergånge von der Burger- Aktivitåt zur Illegalitåt. Eine empirische Untersuchung zur Socialpsychologie der Demokratie (in Zusammenarbeit mit Hans-Martin Uehlinger)', in Matz, U. & Schmidtchen, G. (eds.), Gewalt und Legitimitåt, 4/1 der Analysen zurn Terrorismus. Opladen: Westdeutscher Verlag, 105-437.

Sniderman, P. M. 1981. A Question of Loyalty. Berkeley: University of California Press.

Taylor, C. L. & Jodice, D. A. 1983. World Handbook of Political and Social Indicators, Third
Edition, Volume 2: Political Protest and Government Change. New Haven: Yale University
Press.

Tilly, C. 1978. From Mobilization to Revolution. Reading: Addison-Wesley.

Tilly, C. 1984. Big Structures, Large Processes, Huge Comparisons. New York: Russel Sage
Foundation.

Uehlinger, H-M. 1988. Politische Partizipation in der Bundesrepublik. Opladen:
Westdeutscher Verlag.

Wasserman, R. 1989. 'Staatliches Gewaltrnonopol: Garant oder Hindernis fiir die Weiterentwicklung demokratischen Gesellschaft', in Heitmeyer, W., Moller, K. & Siinker, H. (eds.) Jugend - Staat - Gewalt. Politische Sozialisation von Jugendlichen, Jugendpolitik und politische Bildung. Weinheim - Miinchen: Juventa, 81-93.

Weber, M. 1964. Wirtschaft und Gesellschaft, Studienausgabe, 2 Bånde. Koln: Kiepenheuer
und Witsch.

Westle, B. 1989. Politische Legitimitåt Theorien, konzepte, empirische Befunde. Baden-Baden:
Nomos.

Zald, M. N. & McCarthy, J. D. (eds.) 1987. Social Movements in an Organizational Society.
Collected Essays. New Brunswick and Oxford: Transaction Books.