“NORMAL CITIZENS” VERSUS “ROWDIES” Police categorisations of protesters in Germany

The paper investigates police perceptions of protesters. Based on group discussions with riot police and interviews with high ranking officers, six domains are analysed as dimensions of a risk constellation contributing to the emergence of an enemy image of the protester. The findings suggest that labels describing the “police counterpart” often express distance and opposition (1) and depoliticise demonstrations (2). Furthermore, formal (3) and informal (4) categorisations of protesters as well as the perception of indications of threat in policing practice (5) are examined. Bipolar patterns of classification of protesters were found to be influential. Classifications are partly based in the law and partly in particularistic and normative subcultural attributions of legitimacy which police transfer into their organisational interpretive schemata distinguishing between legality/illegality. For explanatory means the study utilises perspectives of organisational sociology as well as the cultural distance between the police and the protesters (6). This is further elaborated using the social figure of the “normal citizen”, in which specific police conceptions of normality are condensed and which serves as a threshold for the perception of deviant protesters. Besides the implications for theory of democracy of the analysed clichés and enemy images the findings conclusively suggest that the distanced to hostile relationship between the police and some protesters does not merely represent a pedagogical or “practical” problem of the police, but is the expression of a certain conflict structure. In this structure organisational and individual factors on the side of the police as well as their actual conflict experience at demonstrations converge.


Introduction
Protest movements pursue the goal of inducing, preventing or reversing fundamental social change. A certain confrontational momentum in relation to the societal status quo and the prevailing hegemonic values and practices is therefore inherent to protest actors. The police, on the other hand, take on the task of "enforcing and maintaining the government's continuously precarious monopoly of legitimate physical authority", thereby safeguarding "conflict-laden social conditions without appearing as a conflicting party themselves" (Belina, 2014, own translation). 1 This illusion of neutrality corresponds largely with the police's self-concept as a legalistically programmed organisation (Winter, 1997: 19) and with the respective views of many officers. In practice, the conflicts social movements are concerned with often shift to the constellation of police vs. protesters. The police, in the form of riot police units, then take on the role of a conflict party. 2 They turn into a representative or a "substitute object" (Kniesel and Behrendes, 1996: 291;Winter, 1998b: 2) -which may even be one of their central functions (Busch et al., 1988: 458 ff.). And as a moral enterprise (Winter, 1997: 15), police in these situations quite naturally also act on the basis of their own norms.
Different authors have put forth arguments that protest policing is determined by stereotypes and clichés about the "police counterpart" and that police decisions in the realms of sociological discretion or actual discretion 3 are, among others, influenced by these perceptions. At that, minority protestors that are classified as antagonistic, radical or marginal are said to be disproportionately subjected to drastic police measures (for an overview cf. Earl, Soule, and McCarthy, 2003). Likewise, certain forms of protest seem to be increasingly confronted with authoritarian policing styles despite the co-existence of long-term liberalisation trends in protest policing. 4 Based on the assumption that police knowledge carries great weight as a regulative framework for shaping police practice, the paper attempts to answer the question of which views and categorisations have a formative influence on the German riot police's view of protesters. 5 In so doing, special attention will be paid to the question of whether decidedly political criteria, such as political antagonism, play a role. The objective of this analysis is to extract different dimensions of a "risk constellation" that impacts on the emergence of a police enemy image regarding protesters. This focus on risk constellations (Backes, Dollase, and Heitmeyer, 1998) hence will not serve the purpose of conceiving all possible variants of police perceptions regarding protesters, but instead extract those aspects that contribute to biased, imbalanced or negative perceptions of and respective policing styles for at least certain groups of protesters.
The research questions will be answered on the basis of group discussions and interviews with riot police as well as with experts and complementary field "NORMAL CITIZENS" VERSUS "ROWDIES" 11 SOCIOLOGIA, PROBLEMAS E PRÁTICAS, n.º 92, 2020, pp. 9-36. DOI:10.7458/SPP20209214197 2 For an analysis of such a process of mutual alienation and escalation into hostility see Malthaner, Teune, and Ullrich (2018). 3 Sociological discretion encompasses all courses of action in police work that cannot exclusively be traced back to statutory provisions; cf. among others Skolnick (1966), Wilson (1968), Feest and Blankenburg (1972), and Brusten (1971). 4 This observation was made primarily with regards to the global justice movement (Gillham and Noakes, 2007;della Porta and Reiter, 2010; for differences between the US and Europe: Stott, Scothern, and Gorringe, 2013;Wood, 2014) and to contemporary protests against the multiple crisis of capitalism (Kretschmann, 2014). 5 The restriction to knowledge as an orientation for action is important. Nassauer (2015Nassauer ( , 2016, among others, showed that the actual choice of a policing style is determined by a set of different, among them short-termed, interactional conditions and not dominantly by decisions made in the planning phase. Consequently, the knowledge-focused analysis only captures a part of the relevant factors that have an impact on police actions in concreto. Institutional and situational factors can, of course, only be addressed with regards to their representations in the interviewees' knowledge and interpretations here.
observations from demonstrations taken from the research project "ViDemo" (Ullrich, 2014(Ullrich, , 2018. The current state of research and the methodology will be presented before addressing the interviewees' perceptions of themselves and of protesters. Detailed analyses consider police designations of their "counterpart", depoliticising perceptions of demonstrators, formal and informal categorisations of protesters and police indications of threat. For explanatory purposes, the study utilises perspectives of organisational sociology as well as the cultural distance between the police and the protesters. This distance culminates in the figure of the "normal citizen", in which specific police conceptions of normality are condensed and which serves as a threshold for the perception of more transgressive protesters. The conclusion reveals that the distanced to hostile relationship between the police and some protesters does not solely constitute a pedagogical or practical problem of police training but is in fact an expression of a conflict structure. In this structure, organisational and individual factors on the side of the police as well as their actual conflict experience at demonstrations converge and fuel each other.

Current state of research
Current and systematic empirical investigations on police categorisations of protesters are not available for the German context. The only substantial analyses for Germany are older investigations by Willems et al. (1988: 153 ff., based on group discussion from 1983) and Winter (1998b: 5 ff., interviews and document analyses from the early nineties). Both reveal that police officers lack knowledge of the protesters' objectives, arguments and ideological backgrounds. Activists are subdivided by police into either "peaceful protesters" or "hooligans prepared to use violence". In so doing, the former are characterised as informed, open to discussions and oriented towards the cause; their being affected immediately by this cause is assumed to be their motivation. The "anarchists", "rowdies" or "hooligans", in turn, are said not to be communicative, but young, malleable and interested in trouble rather than in substantial claims. They are further said to be directed by "string pullers" (Willems et al., 1988: 154) and to travel from demonstration to demonstration. Occasionally, a third category of protesters is coined -protesters who will use violence as an expedient, should peaceful protest not serve its purpose. This "intermediate category" (Winter, 1998b: 6) challenges the dominant binary image only partly whilst still being based on the same distinction. Stereotypical outer appearances enable the police to make a distinction of main groups. Means and objectives are considered as one by police evaluations of legitimacy. The root of violence is constantly located within the ranks of the protesters, who the police suspect to be similarly organised in a tactical manner as is true for themselves (Winter, 1998b: 6-10). Accordingly, police officers in a survey advocated uncompromising action against protesters more than the general population (Lamnek, 1988).The relevant international literature comprises studies based on interviews and protest observations such as P. A. J. Waddington's (1994: 107 ff.) for Great Britain and della Porta and Reiter's (1998: 24 ff.) for Italy. They draw a picture quite similar to the German studies, while additionally mentioning specific actors and organisational characteristics that police attributions of legitimacy are tied to. The police perspective has established social movements with clear structures and responsibilities as having an interest in peaceful protest. The same is viewed as not true for loosely associated, non-institutionalised protesters who have no interest in peaceful protest. In the distinction between "good" and "bad" (e.g. "ordinary decent protesters" and "professional protesters", della Porta and Reiter, 1998: 24;della Porta, Fillieule and Reiter, 1998) the predictability of protesters and their willingness to communicate play a significant role; the "bad' protesters" unpredictability constitutes a problem for the police. These stereotypical distinctions are based on both regulatory guidelines (official instructions, operation concepts) and concrete experience. The more recent literature adds evidence for such binary oppositions, with peaceful workers and trade unions against violent hooligans, anti-globalists, thugs, troublemakers or anarchists (Baker, 2012(Baker, : 66, 2014aConway, 2003: 515;Wahlström, 2007: 391). Further research is of a rather cursory nature or mainly expresses general experience on the part of researchers and on-site personnel. Behrendes notes that police are often guided by worst case scenarios. The focus on violent protesters coins the police's perception of operations, even though they are aware that the vast majority of protests proceed in a lawful and peaceful manner (e.g. Behrendes, 2016;cf. Kniesel and Behrendes, 1996). The respective understandings of one's role as well as stereotypical enemy images mutually aggravate the interaction between protesters and police. These can lead to inappropriate "precautionary moves", posing a potential threat to the freedom of assembly (Aden, 2016: 9;. What precisely do we know about the consequences of such categorization processes? Multiple authors observe a greater show of force and increasingly repressive and potentially escalating tactical measures against negatively connoted groups of protesters (Eggert et al., 2016;D. Waddington, 2007: 35;Willems et al., 1988: 162;Winter, 1998b: 19). Baker, for instance, was able to show for Australia that the labelling of protesters decided on whether a dialogue with them was established or not (Baker, 2012(Baker, , 2014b. Della Porta, Fillieule and Reiter view this dialogue as a "reward" given out by the police to groups labelled as "good" in a comparative study on protest policing in France and Italy (della Porta, Fillieule and Reiter, 1998). Antagonistic (Fernandez, 2008: 8) and transgressive (Gillham and Noakes, 2007) protesters in particular seem to be increasingly in the sights of the police. 6 In comparison to the majority of demonstrations, they are more often confronted with weapons (pepper spray, tasers, long range acoustic devices, flash grenades, rubber bullets) and barricades (fences, assigned protesting zones, kettling, custody, cf. Wood, 2014: 26). The few systematic studies suggest that weak (low media presence or inner weakness) and radical protesters or those with a reputation with the police were affected by more police presence and tougher policing (Earl, Soule and McCarthy, 2003;Eggert et al., 2016;Wisler and Giugni, 1999).

Design and method
Empirically, the study relies on group discussions and expert interviews as well as ethnographical field research in three federal states of Germany.
Eleven group discussions were conducted with police officers between 2011 and 2016, most of them in 2015. The majority of participating officers were employed in mobile squads of the riot police or Master's degree police students with prior riot police experience. A large proportion of five groups consisted of officers genuinely responsible for video documentation or otherwise concerned with the topic (video documentation instructors, members of technical units or "conservation of evidence and arrest units [BFE]"). 7 Group discussions were chosen as a research strategy in order to mirror the situation of police operations (which are group situations), thereby allowing for the reconstruction of the meanings and norms of the respective groups reporting their own field practices and interpretations. Secondly, group discussions are expected to initiate self-supporting dynamics of narration, hence breaking up articulatory inhibitions, predominantly when it comes to uncomfortable topics such as illegal behaviour. An additional eight individual interviews were conducted with representatives of the police union as well as, more notably, with senior police officers involved in the police escorting of demonstrations in their functions as supervisors or staff members.
Anumber of characteristics of the police, particularly its high degree of organisational closure, have influenced the collection of data. That induces certain restrictions constitutive for most research on police (Fox and Lundman, 1974;Lundman and Fox, 1978;Reichertz, 2003: 414 ff.;Reiner, 2010;Rogers, 2014). The difficulties in accessing the field caused by police gatekeeping, for instance, incurred certain trade-offs in contrastive sampling (for details see Ullrich, 2019). To counter this, supplementary sources were used to triangulate perspectives on the police from the inside with external views on their professional practice. These include three expert interviews as well as a multitude of shorter background conversations with demonstration observants, journalists, researchers and politicians. Further sources comprise field protocols of ethnographical observations on demonstrations. All verbal sources were transcribed, anonymised and analysed using MAXQDA. The material was coded with open (abductive) coding procedures. The presentation in the paper is typifying and aims to extract risk constellations as well as dominant tendencies among riot police.
A further restriction needs to be discussed with regards to the validity of the findings. The research project "ViDemo" uses data focused to a larger extent on questions beyond the categorisations of protesters, which is why such matters were pursued more systematically during data collection. However, the qualitative data contain a number of insightful verbalisations of police views of demonstrators, many of which were uttered en passant and as a matter of course. The project's emphasis on video documentation may also have directed the participants' attention towards breaches of the law or violence to a greater extent than would have been the case in a general interview about demonstrations. 8 However, a plethora of aspects were taken up on in the conversations, thus ensuring that conflicts and problematic situations did not remain the sole topics of conversation.

Police labels for protesters
Police labelling of protesters can be considered a very basic part of everyday "protest diagnoses" (Winter, 1998a: 188). In the transcripts, demonstrators are very frequently dubbed with the almost terminological jargon word "police counterpart" (polizeiliches Gegenüber). Such labels indicate that the designation of protesters often times draws on words signalling a relationship of distance, opposition or even hostility (figure 1). Quantitatively speaking, neutral labels and those indicating distance are predominant, however, terms indicating an adversarial or hostile relation were used in four group discussions and two interviews and remained uncontested therein.
Every now and then, clearly derogatory terms were used, too. In a group discussion, the increased social divide and polarisation along with the associated work load for the police through protests and conflicts were criticised. One officer added they had never met "a reasonably intelligent human being" at a demonstration. The officer is further supported by another who had the impression that this lack of intelligence manifested itself in the protesters' banners, about "ninety percent of which may be misspelt".
Pejorative and defamatory labels equally appear almost naturally in descriptions that do not incorporate violent or tense conflict situations. They are mentioned en passant, without further explanation and denote protesters that are met with obvious aversion. An interviewee, for example, describes the police's reaction to the protesters' inquiries on why they are being recorded on video as follows: Look, we don't have to explain our measures to each and every nonentity who asks us something because that way, we wouldn't be able to do our job. [034_GD_Pol] Even if other circumstantial aspects such as stress or practical concerns have an impact on the decision not to communicate, a basic attitude seems to be at work here. This attitude is supported by similar depictions from other interviewees as well as through the field observations of demonstrations. The latter frequently revealed dismissive and "NORMAL CITIZENS" VERSUS "ROWDIES" 15 SOCIOLOGIA, PROBLEMAS E PRÁTICAS, n.º 92, 2020, pp. 9-36. DOI:10.7458/SPP20209214197 8 A variety of videos that showed, among others, police cameras, tense situations, filming protesters, conflicts between police and protesters etc. served as stimuli. The discussion was then opened by asking for comment on the stimuli as well as raising the question of whether the videos corresponded to actual professional situations as experienced by the participants.
impolite styles of communication on the part of the police (even in situations with no signs of stress), which became evident as a distinction in the handling of perceived participants (distanced and impolite interaction) as opposed to perceived passers-by [friendly interaction; cf. among others 042 & 074_FP_Demo]. 9 Explicitly appreciative or positive labels for protesters and demonstrations played a minor role if any. One interviewee occasionally asserts a "nice, diverse spectrum on the civic side, too, where there are the totally normal citizens" [053_Int_Pol]. Generally positive affirmations of the right of assembly for democracy do exist; however, these are of a rather formulaic nature and represent the organisation's perspective. The few appreciative statements are not related to the demands of the demonstrations; rather, they refer to a demonstration's peaceful course or to the "sincerity" and "friendliness" of the participants (such as the gratitude expressed by them to the police).

Depoliticisation
The pejorative formulations testify to a perspective that depoliticises demonstrations themselves as well as the objectives of the degraded groups. This mode of operation allows for a delegitimisation of such gatherings without overstepping the boundaries of the binding legal framework, which defines the right of assembly as a legally protected good. An example is given by a riot police officer, who also holds instructional functions, in a group discussion:

Figure 1
Police labels for participants of a demonstration (without derivations and similar forms; individual quotes are presented in quotation marks; the most common labels are printed in bold in all figures throughout the paper) 9 Thanks to the GODIAC project, a police investigation on protest and strategies of policing in different European countries (Rikspolisstyrelsen, 2013), we can draw on a detailed account of how police officers treat certain groups differently based on outer appearances. Aleaked draft of the project's final report (Source: euro-police.noblogs.org/files/2013/11/GODIAC_Wien_2011.pdf; known to the author) contains a passage that was later removed. It indicates that certain people (e.g. tourists, individual passers-by, cyclists) were allowed to pass a police cordon at the site of a demonstration, if they were able to prove their motives to the police's reasonable satisfaction. Others were denied this right of way with no clear reason. Some, such as members of recognisable ethnic minorities, were not even heard.
Back in the days, I still had the feeling that protesters were following a cause -sometimes we even talked; but today -what's it called? -nothing but hooligan tourism. Even little boys are only interested in throwing stones. That's a veritable decline in values. No respect! [024_GD_Pol] The person quoted above equally tells of peaceful demonstrations and positively connoted (peaceful) groups of protesters on other occasions. However, as the quote indicates, the confrontation with unpleasant protesters shapes the basic image that at least parts of the police have of demonstrations -even if objectively speaking, the majority of protests are characterised by an absence of violence. 10 The formulaic appreciation of demonstrations serves the purpose of maintaining an organisational image (Schauseite) but is shifted off to the past. The contemporary impressions equate protests with unpolitical violence that is lacking both substance of content and respect. 11 A similar differentiation is the separation of serious and legitimate protesters from those considered unpolitical and illegitimate. Examples of this depoliticisation can equally be found in systems of formal classification (see below: Formal classifications), as is the case in Hamburg, where protesters pertaining to the red category of threat to public safety are characterised as having no interest in "sober protest claims" but would "prioritise violent conflicts". 12 "NORMAL CITIZENS" VERSUS "ROWDIES" 17 SOCIOLOGIA, PROBLEMAS E PRÁTICAS, n.º 92, 2020, pp. 9-36. DOI:10.7458/SPP20209214197

Examples of depoliticising statements about protesters and demonstrations
-"The problem consists of those […] seeking trouble only." -"There are certain parts of the protesters, I personally think, whose intentions do not revolve around expressing their opinion." -"I don't actually consider them demonstrations anymore. They have become more of a mass event […] because it's more like at football matches." -"To me, these are events where people go to let off steam." -"The last Love Parade [Zug der Liebe -a demonstration that took place in Berlin in 2015 and was accompanied by electronic music] was no political event after all!" -"They organise some kind of events about women-lesbians-lycra -I really don't have any sympathy for that." -"a pure happening thing"

Formal classifications
Classifications primarily result from institutional regulations. One essential distinction regarding the classification of protesters that may have an impact on policing results from the law 13 and from judicature, 14 as well as from currently prominent socio-psychological concepts of policing: 15 peacefulness/non-peacefulness (Bundesministerium des Innern, 2004: 81). The exact definition of peacefulness/ non-peacefulness, however, is both juridically and practically precarious (this is true for the actus reus itself but especially for the preventive, i.e. the prefactual attribution of "inclination") and therefore subject to discussion. The traffic-signal-classification that is common among some police authorities constitutes an attempt at a formal operationalisation that largely corresponds to the A-B-C categorisation used in football policing. At that, "green" stands for peaceful protesters who are following a cause and are willing to communicate; "yellow" stands for potentially violent and uncooperative protesters and "red" represents violent protesters in the case of which sober protest claims do not play a role (see footnote 13). The known, formalised traffic-signal-classifications, use varying ascriptive features. The Lower Saxon operation concept LEO "LEINE", for instance, uses personal characteristics, behavioural characteristics and apparel characteristics. The latter criterion in particular demonstrates the role of extremely vague and cliché characteristics, using "dark or black" clothing "typical of the scene" (in contrast to "civil" clothes "adapted to the weather") as an indication of potential threat to public safety.
Even in states or units where there is no formal traffic-signal-classification, a content-wise identical three-stage categorisation is used, though it sometimes uses different terms (e.g. "peaceful" -"interference-prone" -"pre-pared to use violence", 032_INT_Pol). The operational application of such schemes, however, varies. Police officers from different police authorities report this procedure to be generally accepted and common, while others distance themselves explicitly, claiming it to be too simplistic and unsuitable (and therefore not practiced).
On the one hand, these classifications are used situationally in order to evaluate certain groups, situations and scenarios of interaction at demonstrations; on the other hand, they are equally drawn upon during the planning phase. 16 In so doing, Article eight of the German constitution guarantees the right to freely assemble "peacefully and without arms". 14 The Brokdorf resolution of the German Federal Court is of special importance for protest policing. It obliges police to act in a friendly and cooperative manner towards peaceful protesters, even if "disturbers" are present in the protest. More communicative and managerial approaches to protest policing (since the eighties) as well as more selective policing styles relying on the (attempt of) differentiation between peaceful and non-peaceful can be traced back to this ruling. 15 For example the "Elaborated Social Identity Model" (cf. Reicher et al., 2004;Rikspolisstyrelsen, 2013). 16 In extreme cases like the "Welcome to Hell" demonstration against the G20 in Hamburg 2017 (Malthaner, Teune and Ullrich, 2018: 36) police thwart their legal obligations through illegitimate generalization: the expectations of some "disturbers" is then used to classify entire demonstrations.
they shape a horizon of expectations that is not coined by actual events or actions. The classification of the overall demonstration additionally uses an assessment of the expected "disruptive quality" (e.g. on an eight-point scale in one federal state).

Implicit and informal classifications
The organisationally determined categories are very broad and not based on unequivocal rules of allocation. They can therefore be considered a mere option for schematisation. It remains to be clarified how these patterns are substantiated in practice, meaning which actual signals are decisive in the allocation process to a certain category. In addition, the question of whether other distinctions of protester types are relevant to police needs to be raised. In a multitude of sequences, examples of positive and negative protester types are mentioned, or rather characteristics and displays of behaviour that are classified accordingly. These examples support predominantly bipolar oppositions. Trade unions, the DGB (the Confederation of German Trade Unions), "normal protesters" or "normal citizens" serve as collective symbols of the affirmed, which are contrasted with the uncooperative, radical disturbers of the "black bloc", who are violent or at least willing to use violence.
Similar examples can be found in Zimmer (2014: 88), where a leading police officer in an interview sequence on sit-ins distinguishes between per se "subversively acting" members of the autonomous scene and those who choose civil disobedience "out of a personal motive". Another leading police officer from the Lower Saxon riot police in a similar fashion talks about an "autonomous scene primarily concerned with violence" that -to him -does not represent "protesters in the proper meaning of the word" (Zimmer 2014: 96).
The function of such categorical binaries is to transfer the actually made distinction legitimate/illegitimate into the relevant categories of criminal and police law as well as police organisational processes. This allows officers to escape moral dilemmas that result from conflicts with own political views or from conflicts between legitimacy and legality. The strict sorting makes real complexity manageable within the distinctions typical of the organisation. It is a juridification of the social. The tendency towards "bipolar perspectives" (Kniesel and Behrendes, 1996: 303) is especially momentous as it equally enables police to integrate ambivalent or legally contested phenomena. The equalisation of violence and non-violent rule breaches (civil disobedience) by police, already criticised in 1996 (ibid.), is equally present in the current study's data. Participants of a group discussion, for instance, showed themselves appalled by the fact that the then presiding officer of the German parliament, Wolfgang Thierse, had participated in a sit-in against a radical right-wing march. The discussants understood this as an example of their perception that even the top political level did no longer set an example of basic values (namely law-abidance, in this case the respecting of neo-Nazi demonstration rights).

Indications: actual and political prognoses of threat
The sublegal assessments depicted above Implicit and informal classifications gain relevance when they start to induce actions, i.e. when the protest diagnosis (co-)determines how certain groups are treated. This is all the more true as many police officers state that their general "experience", more precisely a certain "intuition", or a "gut feeling" as well as a "healthy sense of guilt" determine their actions. A discussant words this most concisely as a disposal of an "inner prognosis of danger". And "quite naturally, one's senses will be sharpened" if this prognosis is high. It functions as a potential, which is equally explained in another group discussion: 17 But to bring it back to your question of whether we go in with an increased potential of danger, well, if it's a right-wing against left-wing demonstration, you do go in differently as compared to ten green activists who fight the destruction of the sunflower, right. […] Or if I go to a sixth division football match [of two hostile teams], that is, you do go into work differently as a police officer than if it's about the sunflower. [056_GD_Pol] The determination of threat potential operates with the basal distinction between violence expected/not expected. Which factors de facto function as signals for the likelihood of violence remains to be analysed. In the following, I will assess sequences that include such indicators which justify police action from their point of view in more detail. Such sequences occurred in group discussions and interviews, for example whenever someone reported on the legal requirements of videography (or more generally speaking the readiness of resources such as cameras). Interviewers

Figure 3
Police assessment of protest actors and their characteristics (representation as in figure 1) 17 Likewise, international research assumes that considerations of risk and pre-knowledge largely determine the selection of resources (Eggert et al., 2016).
additionally followed up on such-like suggestions, asking about whether certain nameable groups or other things were the trigger of "higher alert" situations. Indications of imminent threat are especially interesting for research on police practice as indications are underdetermined by law, and often prospective (and sometimes perhaps projective). The decision to act is therefore contingent. This distinguishes an indication from concrete (factual) misconduct that implies an obligation to act. 18 The different indications can be allocated on a continuum that is determined by the degree of concretion and therefore by its legal justifiability (the categories are subsequently printed in italics).
The indication most clearly legitimised by law is that of concrete criminal acts or administrative offences. The covering of the face by protesters (forbidden on demos in Germany), for instance, is taken as a present fact or at least as a preparation for imminent criminal acts. Concrete expectations of violence, the indications of which can be announcements on the internet, police knowledge of and experience with certain organisers or incidents on preceding events are also judicially unambiguous; this applies particularly to demonstrations with counter-demonstrations. The dimensions of the demonstration are also relevant as big dimensions and complexity signal a certain interference-prone character of the event to the police. Spatial characteristics also play a role (e.g. if symbolic locations or city districts that are somehow antagonistic to the demonstration are passed by). A further relevant category is the protesters' behaviour. While actions perceived as preparations for criminal acts (such as the covering of the face and the grabbing of stones, see above) are mentioned as signals of danger, so are totally legal practices such as closed ranks, raised banners or a countdown culminating in a burst of speed by a group or the front of the demonstration ("escalation signs", Nassauer, 2015). The behaviour of the organisers is equally of importance, with a lack of willingness to cooperate being perceived as unruliness. 19 A variety of personal or group characteristics was also mentioned. While unconstitutional symbols are expected at Kurdish or neo-Nazi demonstrations, green activists for example stand for an expectation of peacefulness.
Indicators that are related to topics and the estimation of their radical nature are barely covered by juridical discretion but constitute an expression of actual or sociological discretion (Arzt and Ullrich, 2016). Protest topics mentioned in this sense include anti-nuclear politics, anti-fascism, the Israeli-Palestinian conflict, industrial disputes and Islamism. Political antagonism in these topics is understood as increased likelihood of militancy. In the planning phase, for instance, the operational management raises the question of: "What do left-wing, and partly extreme left-wing forces do?". Other expressions typical of this kind of danger diagnosis are "radicality" or humans who are "against this corrupt capitalist system". The way in which content-related assessments have an influence on protest diagnoses, "NORMAL CITIZENS" VERSUS "ROWDIES" 21 SOCIOLOGIA, PROBLEMAS E PRÁTICAS, n.º 92, 2020, pp. 9-36. DOI:10.7458/SPP20209214197

18
In German law there is a strict obligation to prosecute each crime or offence which comes to the knowledge of the police, although in reality this principle is impossible to accomplish. 19 It is worth noting that preparatory meetings can make the course of events less problematic, however, can also be used to impose restrictive requirements on organizers (Aden, 2016: 14 In this case, the topic of industrial disputes equally functions as a signal (among others) for potential threat. Another sequence of the same group discussion further clarifies this logic. The reverse argument (i.e. referring to causes that are no threat) elaborates that a demonstration as such "is in its fundamental concern, this one was about tuition fees, in its fundamental concern not dangerous" (ibid. emphasis added). The example of the "Blockupy" protests against the European Central Bank in Frankfurt in 2012 pointedly shows how such normative attributions of legitimacy work in the translation of protest claims and their perceived degree of antagonism into categories of threat and corresponding police interventions. The police excessively imposed exclusion orders for the entire city centre of Frankfurt, which, however, did not eventually last in court. And yet, arriving protesters in buses from Berlin still received exclusion orders using the equally worded justification of "anti-capitalism". Due to the prevalence of such practices in police practice, the Federal Constitutional Court recently had to state that police interventions could not be legitimised solely with the hostile attitudes of the participants towards the state (1 BvR 3279/14, Arzt and Ullrich, 2016: 49).
The politically judging binaries of "dangerous" and "non-dangerous" topics point to a strong abstraction of concrete suspicions. The regularly mentioned (police-internal) collective symbols such as the Confederation of German Trade Unions on the one hand and autonomists or the black bloc on the other symbolise this abstraction and the related binary distinction between legitimate and illegitimate protest actors. 20 The logic of this attribution of legitimacy culminates in the frequently addressed figure of the normal citizen, whose legitimate representatives in the streets are -from a police point of view -trade unions or "normal demonstrators".

Explaining the social distance
Political and normative assessments of certain protesters or groups of protesters as well as their concerns can equally be found throughout the interviews, also in sequences that do not thematise prognoses of threat. One interviewee, for instance, generally sees nothing but "stone casters" and "hooligan tourism" in current demonstrations. Unusual forms of protest from a police point of view at "some sort of stupid demo" become a "fidgeting [with] some kind of umbrellas" and the militant demeanour of the "five hundred loons" is nothing but "letting off steam". In addition to the obviously pointed and profound aversion of some police officers and the depoliticising enmity they have against the "opponent side", an overarching pattern of hegemonic social distance becomes apparent with regards to political demonstrations. This pattern has become a genuine part of police culture among riot police. Protest and police seem to represent two cultures that are extremely foreign to each other. I will subsequently examine the aspects of police culture and cop occupational culture that are contrary or antagonistic to wide sections of protest culture, drawing upon different findings, including insufficient police knowledge of and experience with protest, demonstrations as the source of straining experiences and the normative distance towards dissidence.

Insufficient experience with protest
Only few of the interviewees (data from 29 respondents is available) were experienced participants of demonstrations. Two exceptions among the number of interviewed individuals can be referred to as inclined to demonstrations because they have taken a stand for a cause at a demonstration from political conviction once in a while. A few other interviewees indicated that they have attended a demonstration here and there, however, the topics largely revolved around police work (police union activism). Occasionally, they reported on onetime participation in demonstrations in the past (such as student demonstrations against the Gulf War 1991 or turnaround demonstrations in the GDR in 1989) or -more out of curiosity -in single-issue-demonstrations (e.g. on environmental issues). Political commitment to fundamental social change (not restricted to single issues) was not mentioned. Something which is much more typical of riot police officers was mentioned by a respondent who then obtained broad approval from the group: 21 I've got better things to spend my private time on than attending a demonstration. Or rather, it has become the same [strain] since there are demonstrations or football At that, even research closely linked to the police emphasises that the topic of the protest is not decisive when it comes to violence (Schreiber and Adang, 2008 Only a tiny minority of respondents shows an interest in the political questions addressed by demonstrations. This is confirmed by Zimmer (2014: 59, 98), who interviewed a leading police officer. The latter reports on a briefing whose purpose it was to give the participating police officers an understanding of the highly explosive political subject as well as the historical background of a big demonstration. The two hundred mobile squad leaders, however, showed no interest but instead, asked for "two guidelines and off we go". This description is in line with the general estimation that knowledge of protests, protesters, and social movements does not play a role in police training (Zimmer, 2014). It is therefore not only certain protesters but demonstrations in general that remain rather foreign to police officers. In principle, these findings point to the continued significance of Weiß' findings (1992: 38) who concludes that "any form of active political commitment […] [is] suspicious" to police officers.

Demonstrations as a nuisance or burden
As indicated by Willems et al. (1988;cf. also Aden, 2016: 9;Nassauer, 2011: 209;1988: 55), parts of the riot police officers refer to their work as a burden as it includes on-the-job (as in police operations) trouble and in-the-job (through bureaucratic and accountability consequences) trouble (P. Waddington, 1994). Besides the daily routine at the office, police officers frequently face tense situations in the context of demonstrations and football policing (both of which resemble one another from the officers' perspective). Common "on-the-job" experiences comprise psycho-social strain, which can be caused by tiring hours of standing around in the ardent sun, 22 but even more so because of a high density of operations or because of conflictual and complex operations including situations of high mental pressure. During such operations, police officers become the targets of criticism, insults and sometimes physical violence. This causes them to feel like "foot mats" who "put their head on the block" for others. Moreover, as has been repeatedly pointed out, riot police operations often take place in the evenings and on weekends, which is a 22 Nassauer (2016) shows that such unfavourable framework conditions (inadequate supplies, accommodation, technological issues, paucity of information, etc.) have an impact on patterns of interaction, thus contributing to dynamics encouraging violence at demonstrations. threat to leisure time, especially for families. Further strain for officers results from the general unruliness of protesters and from protesting tactics that overwhelm the police' professional interpretation patterns (the confrontational and ironic Rebel Clowns Army as well as non-violent disobedience or blockading techniques were mentioned here). In these situations, police officers are apparently subjected to a serious conflict between their (ordered) habitus and the (disordered) protest structure, which is particularly triggered by transgressive protesters. Interview sequences on video surveillance also revealed the police officers'fear of counter surveillance by protesters, which could potentially lead to public denunciations or individual harassment. Despite the fact that most descriptions of such incidents were circulating standard narratives and were not based on concrete own experiences (or those of actual colleagues), such-like statements underline that police officers view themselves as (at least potential) victims at demonstrations. This position is further reinforced by the frequent public criticism that police officers are confronted with and that they perceive as generally unjustified (cf. Frevel and Behr, 2015;Ullrich, 2019).
Some interviewees reported on an additional overload caused by the increased number of right-wing protests and police operations in the context of the so-called refugee crisis at the time of inquiry. This needs to be taken into account, too, when attempting to understand how the image of the "stupid demo" comes into being. Each and every complication -whether it be a criminal offence or simply a task that evolves from an unexpected blockading action -implies an extra effort for officers in the form of video sightings, desk work, perhaps court hearings and disciplinary checks following suspected misconduct and offence reports -in short: in-the-job trouble. All of this, however, does not constitute an evident part of the professional (self) image that especially the distanced police officers hold, owing to the low degree of appreciation towards demonstrations. Consequently, a probably quite large proportion of riot police perceive demonstrations as an evitable additional burden, not only in specific situation but in an established conflictual interaction structure (Wahlström, 2011). 23 Cultural distance from protesters: the figure of the "normal citizen" 24 A lack of knowledge or contact, a schematised perception and negative experiences or expectations all contribute to the distance between the police and the protesters. It is therefore necessary to include a more basic level in the analysis -a level which represents the sheer cultural incompatibility of (riot) police with some activist "NORMAL CITIZENS" VERSUS "ROWDIES" 25 SOCIOLOGIA, PROBLEMAS E PRÁTICAS, n.º 92, 2020, pp. 9-36. DOI:10.7458/SPP20209214197

23
A comparison with other professions would certainly be insightful here. Psychotherapists, for instance, equally have to endure straining experiences with their clients. However, psychotherapists -as opposed to police officers -are trained not to reproach the person provoking their strain but to perceive it as a signal for the client's deeper problems. This may help them cope with and integrate such mental pressure -a procedure that is often supported by supervision in the therapeutic domain as opposed to police service. 24 The word used is "Normalbürger", resembling something between a "normal citizen" and a "norm citizen". milieus (most notably left-wing, anti-authoritarian and transgressive activists). This profound difference is marked by a number of visible characteristics, such as the emphasis on masculinity, strength, military snappishness in police conduct or sexist posters as seen in a police conference room during field work (cf. also Behr, 2008). The difference is also present in the discussion protocols as the reiterated but repeatedly disappointed desire to represent authority and to gain respect. The afore-mentioned distance will subsequently be analysed starting from a discussion sequence which allows for a reconstruction of key elements pertaining to the riot police's normative system of reference. In the following extract, the discussant refers to the conservation and documentation of evidence units' camcorders, which used to be very common but have come to be gradually replaced as they were not particularly suitable, according to the discussant's opinion: Well, technology-wise and with regards to the resolution, that was really, I mean threehundred-Euro-cameras, that was… every father of a family probably has a better one at home. [022_GD_Pol] The term "father of a family" is particularly insightful for this context. It is a condensation of the taken-for-granted, normative orientations that govern the discussion group and that constitutes a communicative, symbolic enactment in the sense of Manning (1997). This sentence discloses an entire world. In this world, there is the typical middle-class (nuclear) family. The father is in possession of a camcorder, which probably serves to capture holiday events and family celebrations. This father is most likely also the breadwinner of the family. Another discussant later says accordingly that he takes on the effort of working at a demonstration while his "wife and the kids are waiting at home". The father of a family or the family's breadwinner as a normative reference value equally appears in other situations, e.g. as a crime victim deserving sympathy, as "every police officer has a family at home, maybe even kids". This small sequence emphasises in its uncontested implicitness and in addition to the many other indicators as well as in agreement with opinion polls 25 that 25 The few surveys at hand (which were conducted exemplarily in single police bodies) showed that items referring to family and social order obtained high values of agreement, whereas political commitment, sexual freedom and alternative lifestyles as well as general social leanings were rejected more often (Remke, 2002;Weiß, 1992). According to Wehr (1994), young police officers showed a higher tendency to think in categories of order and norm as well as having a stronger preference for job and security as opposed to personal development when compared with other people of their age. The hypothesis of particularly conservative values of police officers, however, could not be confirmed (Wehr, 1994: S. 80). A survey conducted by Liebl (2003) nevertheless postulated a stronger, legalist attitude and a predominant conformity of police officers with the criminal politics of the conservative party (CDU). An older, similar poll of Munich citizens and police officers equally pointed to a such-like "law-and-order" attitude of police officers towards demonstrations. More often than not, they demand no restraints on the part of the police; quite to the contrary: they deem the behaviour of the police appropriate or even too considerate but never too harsh. While the judgement of police by citizens depends largely on their preference for a certain political party, police officers judge "rather on the basis of professional, positional and stereotypical beliefs regarding their role" (Lamnek, 1988: 219).
riot police along with its members constitute a hotbed of conventionality and upright narrow-mindedness, which is characterised by normalistic, especially masculine and heteronormative, family-oriented values of a petit bourgeois ideal world that does not leave any room for deviance. In the words of a police researcher: "Non-conformists leave or are advised to do so" (Behr, 2006(Behr, : 104, cf. 2008; on the police self-legitimisation as citizen-oriented guardians of norms see Winter, 1997: 18). 26 One may deem this interpretation exaggerated as it represents a mere typecasting that in reality is accompanied by a great deal of variance (on the margins of police culture). Nevertheless, this interpretation uncovers aspects that are more or less taken for granted, not problematised or criticised in the discussionsin a nutshell: the implicit norms. This analysis becomes all the more clear when compared to the afore-mentioned descriptions of the counter image of police conceptions of legitimacy: anti-authoritarian autonomists radically opposing state, religion, police, capitalism and the like. They do not wear "civic apparel", as described by police guidelines and stereotypes of cop occupational culture, but clothes "typical of the scene". They may live in occupied houses without fixed traditional family structures, perhaps in open, polyamorous or homosexual relationships. They have no palate for order, as perceived by police officers, or authorities and their demonstrations express all of this through specific protest repertoires. Hence, the organisational imperative of the police at demonstrations to maintain an "order" that in fact cannot be defined (and is thus implicitly oriented towards hegemonic ideas) goes along with attitudes of officers, whose prime objective it is to "safeguard and preserve authority" (see also Behr, 2008;Feest and Blankenburg, 1972: 70;Fekjaer, Petersson and Thomassen, 2014;Loftus, 2010). 27 The rigid and hierarchical protest ideals (an orderly procession with a responsible leader) which underlie the German laws concerning demonstrations support and reinforce this cultural distance, but they do not do justice to the present-day diversity of protest (Aden, 2016;Kniesel and Behrendes, 1996: 295). There is an obvious historic continuity of distance and partial enmity between the police and the un-civic objects of their interventions. The communists of the 1950s and the Free German Youth (FDJ), followed by the Beat concert goers or "the students" -all of which were politically and culturally suspicious to the police (Kleinknecht and Sturm, 2004;Kniesel and Behrendes, 1996: 299 ff.) -started a tradition that is still applicable to today's perception of demonstrations by the police. The normative point of reference of the underlying attributions of legitimacy is the "normal citizen", who does not need to be defined due to its implicitness and naturalness. The frequency of phrases using the word "normal" and other, synonymous concepts such as "the general population" underlines the constitutive implicitness of these norm orientations. Findings indicate that police are guided by mainstream attributions of legitimacy D. Waddington, 2007: 30 ff.;Winter, 1997). 27 It can be assumed that the distinction between unpleasant groups of fans (often called "would-be fans") and the police ideal of a happy family represents a similar structure.

The protester as an enemy image?
The presented findings from the field of protest policing again illustrate a truism of organisational sociology: the belief that the police be a neutral actor without own interests and be driven by justice and law may be common in the police legitimising rhetoric, but proves to be a fiction, although a necessary fiction due to the legal regulations and the bureaucratic structure of the police. In order to fully understand police practice, much more than the law has to be taken into account. On the occasion of demonstrations, especially those in which the police play a more active role than merely redirecting the traffic in the background, they can take on the role of a conflicting party. It is as this conflicting party that the police are often times perceived by the protesters and comes face to face with them. This role, however, is not exerted as a mere substitute for the actual opponent, but as a normative actor in its own right, as a moral enterprise at the service of their version of order.
The summary of the findings reveals various factors which contribute to a negative perspective on protesters or groups of protesters. All in all, they can be understood as knowledge-related and attitude-oriented dimensions within a risk constellation (Backes, Dollase and Heitmeyer, 1998). The individual elements can take effect independently or amplify each other. It goes without saying that they do not paint a profound picture of all existing perspectives as not all dimensions may be equally pronounced for all police officers or police units at any given time. The hegemonic pattern is nonetheless consolidated by the fact that the concerned enemy images and stereotypes remained largely uncontested in the discussion, even if they were not promoted by all discussants, and because they are stabilised by the police organisational structure and culture.
The views that riot police officers have of demonstrations are influenced by a lack of knowledge regarding social movements, schematic perceptions of protesters and by a hegemonic and intuitive social distance that is rooted in the values of the police as an organisation and the attitudes of its members. Affirmative references to protest(er)s are rare. At best, the right of assembly is formally appreciated, even if only for a part of the demonstrations. The notorious assessment formulated by external as well as police-internal reformers over the past decades that protest policing is determined by clichés of protesters, sometimes explicit antagonism, a lack of understanding for protesters and partly rigorous legalism along the lines of raison d'état, still holds true. This may no longer be true for the entire police, and probably to a lesser degree (further research may want to quantify this), however, the basic findings of previous research can be reconfirmed. 28 28 Peter Ullrich SOCIOLOGIA, PROBLEMAS E PRÁTICAS, n.º 92, 2020, pp. 9-36. DOI:10.7458/SPP20209214197

28
A different approach aiming at a typology of (different) police officers (instead of the risk constellations carried out here) with regards to their attitudes towards demonstrations would have to include at least three ideal types: (a) the statist type, who sees demonstrations primarily as a threat to public order and to the police risk community (especially prevalent among "street cops"); (b) the pragmatic type, who accepts the right of assembly at least on a formal, abstract The dominant system of categorisation is based on the distinction between peaceful and nonpeaceful (Kniesel and Behrendes, 1996: 303). This distinction makes it possible for the police to process the complexity of reality by transferring the latter into the juridical framework of interpretation pertaining to the organisation and by linking it to courses of action. 29 Normative distinctions between legitimate and illegitimate protesters, which are partly assessments of political opinions, outer appearances or clichés related to someone's lifestyle, are equally forced into this bipolar juridical scheme. The results of acting out sociological discretion are hence transformed into a juridical entity. This does not mean that today's police work according to the authoritarianism of the 1950s and 1960s; certain tendencies of liberalisation and democratisation are at work in protest policing, not least because of the academisation of police training or the processing of ground-breaking court decisions such as the Brokdorf  level as a valuable good without emphatically appreciating it (this is found primarily among officers of higher rank or with a background of higher education, the "management cops"); and finally (c) the critical type, situated outside of the hegemonic police culture, who shows political commitment and words criticism with regards to their own organisation (primarily police officers with higher education, often times lateral entrants or working in special fields such as training or communicative teams, etc.). Refer to Reuss-Ianni (1993) for a distinction between "street cops" and "management cops". 29 Cf. Police Service Provision 100: "The consequent separation of individuals willing to resort to violence from demonstration participants is imperative" (Bundesministerium des Innern, 2004: 81).
resolution. The findings show, however, that these tendencies do not apply to all protest groups in equal measure. Actors who are perceived as a minoritarian or as antagonistic with regards to their claims, ideology or action repertoires -and even more so if their causes are interpreted as hostile to the state or the public order or if they are the reason for an additional workload -sometimes face extremely negative expectations, which prove themselves time and again in the context of the conflict anyway. The normative basis on which such actors are denied legitimacy by far exceeds what can be justified with their criminal offences and is guided by the ideal of the "normal citizen". In the context of the theory of democracy, the latter is particularly revealing, firstly because of its extremely vague counterpart (the "abnormal citizen"?), and secondly because it points out the existence of a zone of the illegitimate (the "abnormal") in the police view, which is not at all substantiated in the law.
These police perceptions, however, are an orientation framework for planning processes that can induce actions (cf. especially Eggert et al., 2016) but are not equal to the actual practice, which, in turn, is determined also by further, e.g. contextual or situational aspects (Malthaner, 2017;Malthaner, Teune and Ullrich, 2018;Nassauer, 2016;Ullrich, Teune and Knopp, 2018). Yet, if the described explicit or implicit classifications of protesters as deviant, radical or dangerous (on the level of organisation and operational planning as well as in the interaction between police officers and protesters) take effect, they may have consequences for the people concerned. The latter may have to expect selective overpolicing and hostile treatment irrespective of their concrete actions, meaning that their protests are not treated from the "enabling" perspective of the German right of assembly but rather, from a preventative and repressive perspective of criminal and police law. With regards to the normative foundations of this classification, a combination of organisational provisions (the law, guidelines, and operational concepts), common values and attitudes of the police officers (cop culture 30 ) and actual protest experience can be assumed.
From a theory of democracy perspective, a constitutive tension between police categorisations and fundamental rights (freedom of assembly, equality before the law) can be detected. This tension becomes particularly evident in the issue of categories of threat converging with (sub)cultural norms and political classifications. An additional problem may arise from the fact that -according to many interviewees -there is no practice-oriented distinction between football policing and protest policing, while the practices of one field are transferred to the other, more legally protected field of political assemblies and demonstrations through courses of action.
In conclusion, all of this must not be misunderstood as an organisational or pedagogical problem of deficient police training. It would certainly be good to hear if future police officers were familiarised more thoroughly with the colourful world of alternative culture, dissidence and social movements. The right of assembly as a common good could also take centre stage in the context of police training and police-internal communication; and of course, the potential for conflict could be minimised by consistently respecting demonstrations, through a pragmatic style of police operations, de-escalation and improvements of the police working conditions. At least the factor of demonstrations as a nuisance or burden could be relieved through better logistics 31 and a rejection of neo-liberal concepts of new public management. More comprehensive training concepts, supervision, and more time for operation post-processing 32 could equally minimise conflict potential. Nevertheless, it cannot be denied that this conflict, which equally reveals itself on a cultural level here, is of a structural nature. If the hypotheses formulated in the present paper are true, the assumption that the police on the one hand and transgressive protesters on the other attract types of people, whose sole commonality is their great distance to each other, is not far-fetched. These groups are often confronted with each other in real situations of conflict. The actors on one side are characterised by disorder, a lack of hierarchy and structure as well as uproar against authority, while the other side is shaped by clear hierarchies, authoritarianism and cultural homogeneity. The legitimatory ambiguities of criticism, protest and (non-violent) resistance cannot be easily transferred to legalistically structured "conditional programmes" (Willems et al. 1988: 22 ff.) of problem processing as they are used by the police. The afore-mentioned problems can therefore not predominantly be considered a moral failure of the police's mission to protect the right of assembly for everyone but result to a large part from the police's function of keeping order, its organisational structure and its participation in as well as the logic of conflict situations in protest. 31 The situations, which facilitated police violence, collected by Nassauer (2016) also comprised trivial incidents such as breaks in communication due to flat batteries of mobile radio devices, causing reactions of fear among police officers. 32 Interviewees indicated several times that they perceived the group discussion as an agreeable way to reflect upon their work and talk about their concerns without any imminent pressure to act, which is usually not possible due to time constraints. In addition, many concerns and fears need to be processed or sublimated individually. Police officers are required to "put their head on the block", even if that is not part of their self-image. All of this shows an insufficient imparting of psycho-social competencies, and a lack of reflection and supervision.