Spirals of Signification? A Corpus Linguistic Analysis of the German Doping Discourse

Abstract

Based on the idea that the political success of the anti-doping movement might be reflected in an increasing moralization of media discourses, the article traces long-term trends in the German doping discourse. Thus, a unique text corpus covering the period between 1950 and 2009 is analyzed using a corpus linguistic (CL) approach. It is shown that attention for doping has heavily increased and that doping has been marked as a persistent and widespread problem subject to permanent efforts of mitigation. However, there is little evidence for more alarmist moralizations. While quantitative CL techniques proved useful for tracing long-term changes in language use, assessing the role of media within the political economy of the moral regulation of doping requires a more ambitious mix of quantitative and qualitative methods. Nevertheless, the use of CL by scholars of sport communication is recommended, as CL methods are able to process large amounts of digitized data and are quite flexible in theoretical terms.

Keywords

discourse analysis corpus linguistics doping Germany moral regulation

Introduction

Although doping appears to be a global phenomenon endemic to high-performance sport, Germany occupies a special place in the history of doping due to the existence of a large-scale clandestine doping program in East Germany during the Cold War, which included doping of minors without their notice (Franke & Berendonk, 1997). The revelations about these practices, which were largely unknown until German reunification in 1990, facilitated the political process resulting in the creation of the World Anti-Doping Agency (WADA; cf. T. M. Hunt, 2011).

By now, tighter anti-doping efforts have become a central issue in German sport policy making. Recently, the German government presented a legislative draft proposing imprisonment of doping athletes (Federal Government, 2015). The use of criminal law marks a significant policy shift from the noninterventionist approach, which had dominated West German sport policy making before (Meier, Reinold, & Rose, 2012).

Research presented here interprets these policy changes as an example of moral regulation (A. Hunt, 1999) and asks whether this rise in moral regulation is reflected in a moralization of media discourse. In order to provide a tentative answer, a unique text corpus covering the period between 1950 and 2009 is analyzed using corpus linguistic (CL) methods. Thus, a second aim of the article is to explore potentials of CL for research on sport communication.

Moral Regulation, Moralizations, and Spirals of Signification

Anti-Doping Policies and Moral Regulation

The persistent efforts to mitigate performance enhancement in sport can be interpreted as a project of moral regulation, “which involve[s] practices whereby some social agents problematize some aspect of the conduct, values or culture of others on moral grounds and seek to impose regulation upon them” (A. Hunt, 1999, p. ix). “Doping” does not refer to a transhistorical entity or a clear-cut harmful syndrome but to a complex social and institutional construction advanced by moral entrepreneurs (Goode, 2011, p. 11). Thus, what has been defined as representing unfair doping has considerably changed over time. Early anti-doping activists did not only oppose nutrition supplements but characterized even applause as some kind of “psychological doping” (Heitan, 1931). In the 1950s, many sport bodies resorted to “moral bans” demanding an ethical reflection on performance enhancement (Reinold & Meier, 2012). After these “regulations” failed, enumerative lists of forbidden substances were adopted, which primarily included stimulants against which amateurs had objected since the end of the 19th century (Gleaves & Llewellyn, 2013). However, innovations in performance enhancement required permanently amending these lists (Houlihan, 2002; T. M. Hunt, 2011).

The justifications for anti-doping regulations are problematic and inconsistent (Goode, 2011). Although doping is endemic and widespread in certain high-performance sports, in particular professional cycling (Brewer, 2002), a public health perspective would suggest to focus on mitigating substance use in popular sports, as a much larger population is affected. Moreover, given substantial health risks and disparities in competitiveness in high-performance sport, the common ethical justifications for anti-doping, that is, athletes’ health and fairness, defy a thorough philosophical scrutiny (Kayser, Mauron, & Miah, 2007). Therefore, doping regulations have to be understood as “the product of the dominant values and attitudes within the contemporary sport community rather than the product of abstract principle” (Houlihan, 2002, p. 147). Nevertheless, the anti-doping movement increasingly radicalized over time shifting doping from a health issue to “a serious moral crisis” (Dimeo, 2007, p. 93).

A. Hunt (1999) has characterized moral regulation as distinct political and discursive practicing involving “the deployment of distinctively moral discourses which construct a moralized subject and an object or target which is acted upon by means of moralizing practices” (p. 6). Accordingly, research presented here asks whether the unquestionable political success of the anti-doping movement is reflected in an increasing moralization of media discourses on doping.

The question how media coverage and projects of moral regulation are related has been most influentially addressed in moral panic research (Cohen, 2002; Goode & Ben-Yehuda, 1994, 2009; Hall, Critcher, Jefferson, Clarke, & Roberts, 1978). Notwithstanding the controversies surrounding the moral panic framework (David, Rohloff, Petley, & Hughes, 2011), the concept is still useful as heuristic tool (Critcher, 2008): First, the debates about theoretical and empirical shortcomings are relevant for discussing the limitations of the approach applied here. Second, the framework has inspired an empirical application by linguists to public discourse, which seems most promising when it comes to developing a methodological approach for longitudinal studies of public moralization.

Moral Panics and Media Discourse

The moral panic concept has been originally introduced to the sociological debate by Cohen (2002) in 1972 who claimed that:

Societies appear to be prone, every now and then, to periods of moral panic. A condition, episode, person or group of persons emerges to become defined as a threat to societal values and interests; its nature is presented in a stylised and stereotypical fashion by the mass media [ . . ] (p. 1)

The framework was rooted in ideas on deviance construction as presented by symbolic interactionists (Critcher, 2008) and suggested a division between the mainstream society and the outsiders, deviants, and law-abiding citizens (Goode & Ben-Yehuda, 2009, p. 33). Following Critcher (2002, 2009), Cohen’s (2002) model can be characterized as “processual” suggesting a sequence of emergence and naming of a new problem, sensationalized, and stereotyped treatment; unanimous moral condemnation of the threat; support of activist pressure groups and accredited experts; and call for an institution of legislative measures. In contrast, Goode and Ben-Yehuda (1994) have presented an attributional model, which identified five criteria for diagnosing a moral panic: a collective concern about a supposed threat, hostility toward the persons responsible for the threat or the wrongdoing, a collective consensus with respect to the threat and the deviant group, a disproportional concern and hostility toward the perceived threat, and the volatile character of the panic episode (Goode, 2011; Goode & Ben-Yehuda, 1994).

While these definitions have been criticized for being too generic (David et al., 2011; Thompson & Greek, 2012), in particular the idea of disproportionate response poses severe epistemological problems, since moral panic researchers adhere to a social constructionist perspective according to which any social problem is socially constructed (Garland, 2008; Partington, 2010; Ungar, 2001). In the present context, it is more relevant that Cohen’s (2002) framework suggests that media represent key agents of moral panics (Critcher, 2008).

Cohen (2002) pursued an interest-oriented explanation according to which moral panics are fabricated by elites, but the media serve to exaggerate the scope of perceived threats and to construct deviant groups. Hall, Critcher, Jefferson, Clarke, and Roberts (1978) took on the idea of “deviance amplification” but characterized media more strictly as only a secondary disseminator of primary definitions of capitalist elites (Hier, 2008). In contrast, Goode and Ben-Yehuda (1994) deny that manipulating agents are needed for initiating moral panics and stress their grassroots character, but they perceive media’s role as passive.

These ideas inspired a vibrant research, which has stressed the need for more complex concepts. First, scholars have pointed to the fact that the idea of broad-scale moral panics involving a cohesive or monolithic public response is questionable. The media landscape has become more pluralistic and fragmented. Moreover, counterpublics have been established (Garland, 2008; McRobbie & Thornton, 1995). There are also “important differences between types of media: local and national, press and television, upmarket and downmarket” (Critcher, 2008, p. 1136). Furthermore, the moral panic concept ignored that media have their own agenda (Critcher, 2002) although they are “primed for panic” (Critcher, 2008, p. 1129), as sensationalist reporting on deviance fits into media routines and normal practices of newsmaking (Critcher, 2008; Denham, 2008). Finally, the moral panic framework “implies erroneously that the audience believes and acts upon the messages it receives from the media” (Critcher, 2008, p. 1137). What is needed, in contrast, is awareness for audience reactions and interpretations (David et al., 2011).

As the moral panic framework relies on simplifying assumptions, scholars have stressed the need to place the framework into broader theoretical perspectives and to develop a more complex political economy of moral regulation in which the relations between power, material interests, and culture are explored (Critcher, 2009, 2011). Thus, Ungar (2001) has argued that social anxieties and fears should be better studied by building on Beck’s (1992) concept of the risk society in which safety becomes a central social and political goal. However, this account follows the proposal of Hier (2008, 2011) to view moral panics as volatile manifestations of the ongoing project of moral regulation. Moralizations are interpreted as “a routine feature of social life” involving “myriad discourses, symbols, feelings, actors, and truth-claims [ … ] in which certain people/groups are subjected to the practices and processes involved in moralization” (Hier, 2008, p. 181). Accordingly, Hier (2008) emphasizes the multifaceted relationships among diverse social agents, fragmented media, representation, and reality. Moreover, the analyses have to be extended beyond episodes in order to assess the broader foundations that give rise to and sustain ongoing processes of moral regulation and result in resonance among audience members (Hier, 2011; Hier, Lett, Walby, & Smith, 2011).

These ideas are relevant because they imply that empirical investigations focusing solely on media discourses do not allow examining the broader political economy of the moral regulation. However, studying media discourse remains crucial for understanding moral regulation as “most contemporary political strategies are media strategies” (Hier, 2008, p. 177). Thus, the idea that moral regulation involves “distinctively moral discourses” (A. Hunt, 1999, p. 7) implies that attempts to moralize a certain social practice should be reflected in media discourse. In order to trace such moralizations, the account presented here proposes to make use of CL.

In contrast, previous German research on media coverage of doping has employed different techniques of content analysis. Scholars built mainly on the influential account of Bette and Schimank (2006), who claimed that media either ignore doping or that media prefer moralizing and personalizing coverage neglecting the deeper social causes of doping. Actually, scholars found some support for these ideas (Ihle & Scharf, 2007; Schneider, 2012). However, while journalists have admitted ignoring doping (Kautz, 2010; Steinbrecher, 2009), the coverage of doping has increased and became more critical over time (Dresen, 2010; Philipp, 2002). To some extent, comparing these works with the results derived here allows examining the validity of CL techniques.

Detecting Moralizations in Language Patterns

The use of linguistic methods has become increasingly popular among scholars rooted in critical discourse analysis (cf. Baker, 2010b). CL uses corpora, that is, large, principled collections of naturally occurring texts which are stored in electronic form. For detecting linguistic patterns, CL relies on computer-assisted techniques in order to handle the large amount of data (Conrad, 2002, pp. 77–78). The popularity of CL techniques results from the fact that CL methods are able to uncover ideologies and to detect evidence for disadvantage. Furthermore, CL does not represent a theory but a method (Gabrielatos & Baker, 2008). However, many critical discourse analyses employing CL techniques rely on a specific premise concerning the impact of discourse on the audience, which can be characterized as “incremental effect of discourse” (Baker, 2006, p. 13). Accordingly, exposure over time to repeated associations or concepts will act as a kind of mental trigger for the audience (Baker, 2010a). Yet, as CL is still not linked to a powerful and cognitively valid theory (Baker, 2010a), the account presented here avoids strong assumptions concerning the impact of moralized discourse on media audiences and operates on the much more modest premise that language patterns reflect social practices as well as social values and views of the world (Bubenhofer, 2009).

In order to trace the moralization of media discourse on doping, the use of CL methods applied here follows McEnery’s (2006) suggestion on how to detect moral panics in linguistic patterns. Thus, once more, the moral panic framework proves its usefulness as a tool for starting an analysis (Critcher, 2008, p. 1138). Building on moral panic research (Hall & Jefferson, 1976; Hall et al., 1978), McEnery (2006) first notes that discourses surrounding moral panics are obsessive, moralistic, and alarmist. Moreover, Hall and Jefferson (1976) have claimed that moral panics use “spirals of signification” (p. 77) to increase a perceived threat. In particular, this is achieved by linking different activities “in a process of signification as to implicitly or explicitly draw parallels between them” (Hall et al., 1978, p. 223). Therefore, McEnery (2006) has argued that “spirals of signification” represent the key linguistic devices to moral panics. They are characterized by a “process of convergence” during which associations between a certain form of a perceived threat and other, not necessarily related deviant behaviors, are drawn (McEnery, 2006, pp. 91–92). Based on these ideas, research presented here asks to what extent media discourse on doping in Germany has been shaped by increasing moralistic and alarmist language use and spirals of signification. However, in contrast to McEnery (2006), the focus is on lasting changes in language patterns.

Research design: Corpus and Methods

Methods

CL encompasses a great variety of approaches for studying language use (Conrad, 2002, p. 75). Here, a methodologist or “corpus-as-method” stance is adopted according to which CL techniques “can be applied in different fields of language study, and within different theoretical frameworks” (Hardie & McEnery, 2010, p. 386). CL methods avoid the pitfalls of many qualitative approaches, that is, arbitrary selection of texts and a small number of analyzed text segments. In contrast, CL allows for processing large amounts of text data and enables a higher degree of objectivity. CL can also pinpoint areas of interest for further/closer analysis by identifying emerging patterns and leading to the examination of concordance (Gabrielatos & Baker, 2008, p. 6; cf. Bachmann, 2011).

Here, basic CL methods are employed. Thus, the development of the word family “doping” is traced in order to analyze alarmist language use. Then, quantifications are explored. As stressed by several scholars, certain quantifications and the use of metaphors for natural disasters, such as, “floods” or “water,” symbolize a loss of control over a social issue. This can result in a particular negative representation of outsiders (Baker & McEnery, 2005; Gabriolatos & Baker, 2008) and create a sense for the urgency of remedies (KhosraviNik, 2010). Moreover, colligations around “und” (and) are examined. The basic idea is that colligations equal two evils and therefore serve to amplify perceived threats (McEnery, 2006). Furthermore, collocations are determined. Collocations are defined as the above-chance frequent co-occurrence of two words within a predetermined span on either side of the word under investigation (the node). According to CL scholars, collocates of a word contribute to its meaning and allow for its semantic analysis (cf. Baker et al., 2008; Gabrielatos & Baker, 2008). Here, we use the mutual information (MI) score (Baker, 2010b; McEnery, 2006). Applying the MI score enables to analyze collocates regarding their “strength of relationship” (Baker, 2010b, p. 24). Words that have a statistically significant (MI score over 3) relationship with the word doping are of interest because those words most substantially indicate “subtle meanings and connotations” (Baker, 2010b, p. 25).

Finally, based on the idea that the revelations about the East German clandestine doping program might have represented a major shock resulting in substantial changes in language use, the corpus has been partitioned into a prereunification (1950–1989) and postreunification subcorpus (1990–2009), and keywords for both subcorpora were determined. In CL, keyness is defined as the statistically significantly higher frequency of particular words or clusters in the corpus in comparison with another corpus. The purpose of keyword analysis is to determine the “aboutness” of a text or homogeneous corpus (Baker et al., 2008).

The Corpus

The analyzed corpus consists of 1,147 articles of the two weekly publications Der Spiegel and Die Zeit from the period between January 1950 and December 2009. This selection followed a cascade model of communication (cf. Entman, 2004, pp. 9–10) according to which elite media serve as opinion leaders. Spiegel and Zeit belong to the small circle of upmarket fact-corrected media, which are consumed by German elites (Jarren & Donges, 2006, p. 185). A pragmatic reason for choosing these two publications was that they were already digitized and could be easily accessed online. However, the exclusion of tabloids implies a lower degree of moralization. Furthermore, the sample is insofar biased as both, Spiegel and Zeit, represent rather left-liberal periodicals, and the German “new left” has been highly critical of high-performance sport (Bette, 2010, p. 62). Moreover, as suggested by Denham’s (2004) work on nationalism in doping coverage, there is a potential anti-East German bias because East German publications were not included, as they were subject to strict censorship under communism and do not belong to the circle of elite media.

From the two chosen publications, all articles containing the lexeme doping were retrieved. Consequently, the primary corpus represents a complete sample of the doping-related articles published and consists of a total of 1,349,218 text words. On average, an article includes 1,176 words. The sampling potentially excludes articles on performance enhancement in sport not employing the lexeme. However, measurement distortion appears limited, as the reception of the term doping in German was strongly related to questionable methods of performance enhancement. The term migrated to Germany via equestrian sports at the beginning of the 20th century. In human sports, the term doping remained vague but was used to refer to any ethically questionable practice providing competitive advantage (Beckmanns Sport Lexikon, 1933; Graf von Norman, 1928; Grimm & Grimm, 1983).

Historical Context

As a period of 60 years is covered, a short account of historical context is needed for interpreting the empirical results.

1950s

The Soviet Union’s first Olympic participation in 1952 triggered the “sporting cold war” (Dimeo, 2007, p. 75), which facilitated the use of anabolic steroids. Sport policy in divided Germany took a different trajectory. Communist East Germany followed the Soviets and used sport as a tool to gain diplomatic recognition and to prove communism’s supremacy (Balbier, 2007). The West Germans tried to act as Olympic model students and adopted a moral ban on doping after a scandal in 1952 (Reinold & Meier, 2012).

1960s

The death of Danish cyclist Knud Enemark Jensen at the Rome Olympics of 1960 forced the International Olympic Committee (IOC) to introduce doping controls against stimulants. Also, public authorities, such as, the Council of Europe, started addressing doping. Nevertheless, anabolic steroids rapidly spread as they could not be detected (T. M. Hunt, 2011). In West Germany, anti-doping legislation was discussed after cyclist Tom Simpson and a German professional boxer died from amphetamine use in 1967 (Meier et al., 2012). However, political elites pressurized on the West German sport movement to improve competitiveness with the East Germans at the upcoming Munich Olympics of 1972 (Balbier, 2007).

1970s

East Germany became a sporting superpower. The communist party implemented a centralized large-scale doping program effectively circumventing doping controls in 1974 (Franke & Berendonk, 1997). In West Germany, debates about anabolic steroids and questionable practices during the Montreal Games of 1976 forced the West German sport movement to adopt an anti-doping declaration in June 1977 (Meier & Reinold, 2013).

1980s

The 1980s saw increasing political distrust in the commitment of the sport bodies to mitigate doping. Eventually, the positive doping tests of Ben Johnson at the 1988 Seoul Olympics catalyzed more determined anti-doping efforts (Houlihan, 2002; T. M. Hunt, 2011). In West Germany, the death of heptathlete Brigit Dressel in 1987 revealed persistent doping practices (Singler & Treutlein, 2000) and facilitated implementation of out-of-competition tests (Meier et al., 2012).

1990s

German reunification in the 1990 ended the sporting cold war but confronted the reunified sport movement with East Germany’s doping legacy. Moreover, sprint star Katrin Krabbe became suspected of doping in 1992. In 1999, the affair around middle-distance runner Dieter Baumann and anti-doping activists caused public controversies about liability for doping test results. After the Festina doping scandal during the 1998 Tour de France revealed the failure of incumbent anti-doping policies, public authorities enforced the creation of WADA (Hanstad, Smith, & Waddington, 2008).

2000s

The German doping debate was still focused on East German state doping, as the main trial against officials took place in 2000 and convinced the federal parliament to compensate doping victims in 2006. The WADA Code of 2003 harmonized international anti-doping stipulations in an unprecedented manner. However, the affair involving Spanish sport physician Eufemiano Fuentes in 2006 revealed the scientific and global character of doping networks. Germany was affected as the cycling team Telekom and Tour de France winner Jan Ulrich were implicated.

Results

Figure 1 displays the development of doping-related contributions in Zeit and Spiegel in relation to all sport-related contributions.¹ The figure illustrates the heavily increased importance of sport in these two weekly publications, which do not primarily focus on reporting sport results. While only 50 articles in Zeit and Spiegel were related to sport in 1950, the number had risen to 600 in 2009, which indicates an increase by the factor 12. Unmistakably, this trend points to the growing cultural significance of sport. On the other hand, the number of doping-related articles increased even more. While the word doping appeared only in 1 article in 1950, 87 articles included doping in 2009. The peak number of 102 articles was reached in 2007, which implies that readers of Zeit and Spiegel were confronted with doping in almost every issue.

Figure 1.

Texts including “doping” in Zeit and Spiegel: 1950–2009.

Thus, doping has made a career from a rather marginal subject to a permanent issue. It seems that the frequency of doping-related articles shows a cyclical pattern, which appears to be related to the Olympic Games. In addition, peaks seem to correspond with some incisive doping scandals (1987: Dressel; 1988: Johnson; 1991: Revelations on East-German doping; 1998: Festina scandal; and 2007: Jan Ullrich scandal).

In order to trace trends in the use of the word family of doping, all composites consisting of doping and another constituent found in the corpus were analyzed. For example, the composite “Dopingsünder” consists of the two constituents: “Doping” and “Sünder” (sinner). First, the development of the number of composites per decade has been traced. Then, composites with the prefix “anti-,” occasionalisms or ad hoc composites, and composite adjectives were given particular attention. While the absolute numbers are of interest, the relative tokens (per hundred words [phw]) for a 100,000-word corpus were calculated in order to allow for comparisons (Perkuhn, Keibel, & Kupietz, 2012). The analyses reveal substantial changes in language usage over time (cf. Table 1).

Table 1.

Frequencies of Tokens.

Indicator	Measure	1950–1959	1960–1969	1970–1979	1980–1989	1990–1999	2000–2009
Total	Token	45	242	213	641	2.039	3.425
	phw	377.3	548.2	279.9	381.1	476.6	551.5
Anti-	Token	1	5	1	12	67	353
	phw	8.4	11.3	1.3	7.1	15.7	56.8
Occasionalisms	Token	17	31	49	118	304	267
	phw	142.5	70.2	64.4	70.2	71.1	43.0
Adjectives	Token	0	0	1	3	32	61
	phw	0	0	1.3	1.8	7.5	9.8

Note. phw = per hundred words.

The absolute number of doping composites has steadily and substantially increased. Maybe due to major scandals at the international and national level and the fact that the IOC had to acknowledge doping as Olympic problem, the 1960s represented a first step toward a more intense debate but the public attention for doping has increasingly grown since the 1980s. Furthermore, the public discourse about doping has assumed a more elaborated character. The usage of adjectives including doping as constituent has substantially increased (by the factor seven) and resulted in the creation of new words, such as, “dopingfrei” (free of doping). Additionally, the use of composites including the prefix “anti,” such as “Antidopinggesetz” (anti-doping law), has steadily increased since the 1970s. Again, the phw values show an increase by the factor 7 from the 1950s (phw = 8.4) to the 2000s (phw = 56.8). The increase reflects the growing influence and visibility of the anti-doping movement and its institutionalization. Finally, the relative number of occasionalisms decreases over time, and the trend becomes most evident with the transition from the 1990s to the 2000s. The increase corresponds with the creation of WADA, which facilitated a global harmonization of anti-doping stipulations and organizations, which might have served to institutionalize language use. While occasionalisms generally indicate linguistic productivity and the social need for exchange about a certain subject, the decreasing number of occasionalisms found suggests that the growing intensity of the public doping debate has resulted in the establishment of certain composites within public language use (cf. Rose, 2013).

Results for the most frequent composites of doping demonstrate again the high productivity of doping as token frequencies heavily increased. However, the list of most frequent composites remained relatively stable. The most visible changes appeared in the 2000s, since composites referring to institutional changes in anti-doping occupy the list of most frequent composites.

Research on moral panics has placed particular emphasis on how problems and deviants are named. Accordingly, the use of natural disaster metaphors for naming doping has been traced (cf. Figure 2).

Figure 2.

Composites referring to doping as social problem.

The problem naming employs from the 1980s on metaphors for natural disasters, epidemic diseases, and pollution, indicating a perceived loss of social control. However, more prevalent is the presence of composites referring to the clandestine East German doping system, which hints to the public shock about the revelations. In order to trace whether language use for naming involved actors has become more alarmist, composites referring to actors involved in doping were coded whether they referred to the religious or the criminal domain (Figure 3).

Figure 3.

Composites referring to doping athletes.

Since the 1990s, the notion of “doping victims” or “doping damaged” appeared. This language use is again closely related to East German doping practices, where minors were administered performance enhancing substances without their knowledge or informed consent. The composite “doping sinner” (Dopingsünder) appeared first in the 1960s and has been used with an increasing frequency. However, the relative frequency of the token is highest during the 1980s before it decreases continuously. The following quote illustrates the use of the composite:

Nahezu alle großen Radrennfahrer, die Deutschen Rudi Altig und Dietrich Thurau ebenso wie Hinaults Vorgänger Anquetil und Merckx, wurden als Dopingsünder ertappt. (Prügel im Endkampf, 1985; bold emphasis added)

Almost all great road cyclist, the Germans Rudi Altig and Dietrich Thurau as well as Hinault’s predecessor Anquetil and Merckx, have been caught as doping sinners. (Translation)

The use of the apparently religious metaphor doping sinner has been noticed before (Gamper, 2000; Hauser, 2011; Schnyder, 2000). However, the sinner metaphors represent a rather conventional metaphorical pattern in German journalism as indicated by the following quote:

Es hilft aber, wenn man schon mal Parksünder, Temposünder, Steuersünder oder wenigstens Kaloriensünder war. (Zippert, 2005)

It helps if you have already been a parking sinner, speed sinner, tax sinner or calories sinner. (Translation)

As moral panics are supposed to involve exaggerated figures on the magnitude and spread of deviant behavior, the entire text corpus was retrieved for phrases indicating claims to doping by athletes. Thus, we first looked for phrases involving “dopen*” (are doping) or “gedop*” (are doped or have been doped). Then, the phrases were manually coded. Three categories were distinguished. “Generalizations” refer to all claims providing highly general statements on the spread of doping:

Alle Radrennfahrer dopen. (Hacke, 2005)

All bike racers are doping. (Translation)

“Magnitude” refers to statements stressing the systematic character of doping:

In der DDR wurde systematisch und menschenverachtend gedopt, in der Bundesrepublik nur menschenverachtend. (Bürer & Klawitter, 1998)

In the GDR it has been systematically and inhumanely doped, in the FRG only inhumanely. (Translation)

Finally, specific quantifications provide specific numbers:

99 Prozent dopen, und 100 Prozent haben die Gelegenheit dazu. (Die Kontrollen sind ineffizient, 1998)

99 percent are doping, and 100 percent have the opportunity to do so. (Translation)

Figure 4 provides evidence that quantifications are a recent phenomenon becoming more prevalent since the 1990s. This indicates the use of a generalizing frame to doping in sport according to which this form of deviant behavior is endemic to high-performance sport and widespread.

Figure 4.

Quantifications of doping as social problem. Results of “dopen*” (are doping) and “gedop*” (are doped or have been doped) have been merged.

In order to trace how doping has been categorized by equating doping with other forms of deviance, we examined colligations, that is, phrases “Doping und *” (doping and *) and “* und Doping” (* and doping). Among others, the corpus included the following colligations:

“Doping und Kindesmißbrauch” (Doping and child abuse)

“Doping und Wettbetrug” (Doping and betting fraud)

“Stasi und Doping” (Stasi and doping)

“Drogensucht und Doping” (Drug addiction and doping)

After all colligations were retrieved, there were coded manually. However, while the number of colligations increased since the 1990s, no clear patterns emerged (for details cf. Supplemental Table A.2).

CL scholars assume that collocates of a word contribute to its meaning and allow for its semantic analysis. To calculate collocations properly, a span of three words left and three words right of the lexeme doping was predetermined. In addition, only collocates that have a combined frequency higher than 20 were considered to be essential due to the rather small size of the corpus. Thus, collocates with low frequencies and simultaneous high MI score, which are of insufficient interpretational value to this study, were excluded.

Table 2 displays the 30 highest ranked collocates according to the MI scores. The higher the score the higher is the probability that the respective words “usually occur together and rarely occur apart” (Baker, 2010b, p. 24) within the corpus. Furthermore, a high MI score indicates a strong contextual relationship, which facilitates semantic interpretations of the word under scrutiny. Most of the high-ranked collocates of doping relate to a “fighting against” character. Collocates, such as “anti” (anti), “kontrolle(n)” (control(s)), “fahnder” (investigator), and “kampf” (fight), are strongly connoted with attempts to mitigate doping. Approximately 57% of the listed collocates can be subsumed under the category “fighting against” or “opposing” doping. Thus, doping in most contexts is denoted with processes and actions that are directed toward its erasure. While other listed collocates hint to specific doping cases (“radsport” [cycling] or “telekom” [telekom]), to state doping (“stasi” [stasi] or “system” [system]), these collocates only insufficiently foster semantic generalizations of doping.

Table 2.

Collocations for “Doping.”

Rank	Freq	Freq (Left)	Freq (Right)	Stat (MI)	Collocate
1	21	21	0	8.28488	systematisches [systematic]
2	404	403	1	8.24334	anti [anti]
3	133	0	133	8.14049	agentur [agency]
4	56	3	53	7.11495	gesetz [law]
5	23	0	23	6.85052	experte [expert]
6	114	91	23	6.82627	kampf [fight]
7	74	65	9	6.64784	thema [issue]
8	24	0	24	6.36649	nada [national anti-doping agency]
9	30	1	29	6.33478	wada [WADA]
10	49	1	48	6.30564	kommission [commission]
11	35	31	4	5.91751	nationalen [national]
12	28	2	26	5.86804	mitteln [substances]
13	52	2	50	5.85315	kontrollen [controls]
14	37	0	37	5.84500	tests [tests]
15	32	2	30	5.82217	kontrolle [control]
16	20	4	16	5.79226	liste [list]
17	21	3	18	5.46113	fahnder [investigator]
18	28	3	25	5.34942	system [system]
19	24	9	15	5.34659	donike [Donike]
20	26	22	4	5.24817	kölner [cologne]
21	194	182	12	5.17199	gegen [against]
22	21	1	20	5.16718	vergangenheit [past]
23	20	5	15	5.10346	experten [experts]
24	22	5	17	4.92976	manfred [Manfred]
25	46	4	42	4.81205	mittel [substance]
26	32	10	22	4.78487	radsport [cycling]
27	26	10	16	4.73303	stasi [Stasi]
28	20	6	14	4.47033	problem [problem]
29	34	5	29	4.45646	team [team]
30	24	4	20	4.30440	telekom [Telekom]

Note. Freq = frequencies; Stat = statistics.

As referred earlier, in order to trace whether the revelations about East German doping left a decisive impact on media discourse, the corpus was split into two subcorpora, that is, a 1950–1989 prereunification corpus and a 1990–2009 postreunification corpus, and key words were determined (Table 3).

Table 3.

Keyword Analyses for Two Subcorpora.

Keywords 1950–1989				Keywords 1990–2009
#	Frequency	Keyness	Token	#	Frequency	Keyness	Token
1	105	184.503	mexiko [Mexico]	1	825	321.826	epo [EPO]
2	66	169.256	moses [Moses]	2	646	314.548	ullrich* [Ullrich]
3	127	153.701	daume* [Daume]	3	451	227.939	telekom [Telekom]
4	75	148.341	ostblock [Eastern block]	4	386	184.171	euro [Euro]
5	124	146.032	bundesrepublik [Federal republic]	5	362	172.170	stasi [Stasi]
6	75	139.860	thurau* [Thurau]	6	287	145.052	krabbe* [Krabbe]
7	73	134.777	dressel* [Dressel]	7	271	136.965	baumann* [Baumann]
8	211	129.116	professor [professor]	8	475	132.504	anti [anti-]
9	53	127.379	udssr [USSR]	9	363	130.236	jan* [Jan]
10	45	125.801	neufeld [Neufeld]	10	3418	124.264	spiegel [Spiegel]
11	126	115.814	seoul [Seoul]	11	244	123.319	armstrong* [Armstrong]
12	45	114.003	sowjet [Soviet]	12	229	115.738	jaksche* [Jacksche]
13	37	110.943	anabolica [anabolic steroids]	13	227	114.727	franke* [Franke]
14	43	108.270	breitner [Breitner]	14	235	108.846	peking [Bejing]
15	66	105.643	Bundesdeutschen [West German]	15	332	103.458	china [China]
16	56	104.933	toni [Toni]	16	530	87.120	team [team]
17	103	99.637	klümper* [Klümper]	17	169	85.414	fuentes* [Fuentes]
18	39	96.841	sowjets [Soviets]	18	167	84.403	katrin* [Katrin]
19	48	95.656	antibiotika [antibiotics]	19	404	83.276	radsport [cycling]
20	52	94.784	sowjetischen [Soviet]	20	160	80.865	sydney [Sydney]

Note. The left section of the table uses the subcorpus of 1950–1989 as primary corpus and the subcorpus of 1990–2009 as reference corpus and the right section of the table uses the subcorpus of 1990–2009 as primary corpus and the sub-corpus of 1950–1989 as reference corpus.

The results defy the idea that East German doping had a lasting impact on the aboutness of the doping discourse. In contrast, the high number of proper names of actors involved in doping (marked with *) indicates the strong personalized character of the doping debate and reflects the most notorious doping scandals in the two periods. Only one key word—“stasi” (East German secret service)—is clearly related to East German doping practices. Surprisingly, the subcorpus 1950–1989 includes more key words indicating that doping was highly associated with communist countries (ostblock [Eastern block], udssr [USSR], and sowjet [Soviet]).

Discussion and Conclusion

Moral Regulation and Changes in Language Use

The central question addressed here has been whether the political success of the anti-doping movement is reflected in an increasing moralization of media discourses on doping as indicated by moralistic and alarmist language use. The findings are, however, complex.

In accordance with previous research, it could be demonstrated that attention for doping has heavily increased over the period examined. However, increased media attention might not only reflect the success of the anti-doping movement’s project of moral regulation but also changed media agendas. As investigative journalism faces cutbacks due to the newspaper crises, even upmarket publications might resort to sensationalist and scandalizing journalism (Critcher, 2011). This idea is supported by the substantially higher attention given to more recent doping scandals. Moreover, some of the increased media attention appears to be related to the legacy of East German doping and the subsequent attempts to deal with it.

Ultimately, the idea that the political success of the anti-doping movement is reflected in a substantially moralized media discourse received only limited support. There is evidence for alarmist language associating doping with natural disasters. Usually, such metaphors indicate a loss of social control and urgency of remedies. Yet, the idea of a spirals of signification according to which doping becomes equated with other social evils was not supported. However, composites and collocations indicate that the increasingly strict moral regulation of doping has left its impact on language use, as doping has been marked as a persistent and widespread problem subject to permanent efforts of mitigation, which reflects social realities insofar correctly as stricter anti-doping efforts did not erase doping but induced technology races between performance enhancement and test methods. Again, changes in language patterns could indicate as much successful moralization as changes in social practices resulting from the institutionalization of anti-doping policies. Accordingly, it can be concluded that the long-term project of moral regulation of doping has left its impact on language use. However, the approach applied contributes little to understanding the dynamics between social problems, moral entrepreneurs, and media coverage.

Limitations of the Current Approach

First of all, the limitations result from the sample used because tabloids seem to be more prone to resort to moralizations. Moreover, the text sample was rather small restricting the opportunity to derive statistical significant results and to identify triggering events. However, as CL methods seemed to support previous findings on media coverage of doping as personalized and focused on scandals, they might be able to substitute for more costly and time-consuming methods of hand coding. Nevertheless, the use of purely quantitative CL methods poses substantial limitations.

While CL allows tracing long-term changes in language use, the methods applied do not allow drawing conclusions on causal relationships. For understanding the political economy of the moral regulation of doping in a fine-grained process, studies on volatile panic episodes are indispensable. Studying the complex interactions of diverse social agents, the media, representations, and social reality requires a mix of quantitative and qualitative methods in the tradition of moral panic research. As Denham (2004) has demonstrated, media can serve as agenda setters of anti-doping policies. However, sport officials have also managed to effectively resist public pressures (Meier et al., 2012). Accordingly, more fine-grained research on changing media agendas and political opportunity structures is needed. Above that, the enormous linguistic productivity of doping invites speculations why the subject resonates so well in media discourse. However, for understanding why performance enhancement figures so strong in capitalist societies, different methods addressing the broader ideological foundations are needed. Finally, CL does not allow for inferences on audience reactions. Thus, marking doping as an endemic and widespread problem might not per se invoke moral outrage or distrust.

Perspectives for Future Research

Notwithstanding these limitations, CL might offer scholars of sport communication new perspectives, given the fact that almost all forms of print media publications are now available in digital form. Moreover, CL methods are quite flexible in theoretical terms. Thus, CL techniques can offer new perspectives for studying sport-specific language or the use of sport metaphors (Segrave, 2000; Shields & Bredemeier, 2011). As the works of Bubenhofer on language use in Alpinist publications demonstrate, CL techniques reveal the professionalization of sport journalism and changes in language use point also to changes in motivation of mountaineers (Bubenhofer & Schröter, 2012). Ultimately, CL approaches open new perspectives for doing comparative research, as CL provides cross-linguistic techniques (Taylor, 2014).

Footnotes

Declaration of Conflicting Interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: Research presented here was partially supported by a grant from the Federal Institute for Sport Science (Bundesinstitut für Sportwissenschaft – BISp) for the project ‘Doping in Germany from 1950 until 2009’ [grant number IIA1-081903A/09-12].

Note

Supplemental Material

The online materials [appendices/data supplements/etc.] are available at .

References

Bachmann

(2011). Civil partnership—“gay marriage in all but name”: A corpus-driven analysis of discourses of same-sex relationships in the UK Parliament. Corpora, 6, 77–105.

Baker

(2006). Using Corpora in discourse analysis. London, England: Continuum.

Baker

(2010a). Representations of Islam in British broadsheet and tabloid newspapers 1999–2005. Journal of Language and Politics, 9, 310–338.

Baker

(2010b). Sociolinguistics and Corpus Linguistics. Edinburgh, Scotland: University Press.

Baker

Gabrielatos

Khosravinik

Krzyzanowski

McEnery

Wodak

(2008). A useful methodological synergy? Combining critical discourse analysis and corpus linguistics to examine discourses of refugees and asylum seekers in the UK press. Discourse and Society, 19, 273–306.

Baker

McEnery

(2005). A corpus-based approach to discourses of refugees and asylum seekers in UN and newspaper texts. Journal of Language and Politics, 4, 297–226.

Balbier

U. A.

(2007). Kalter Krieg auf der Aschenbahn: Der deutsch-deutsche Sport 1950-1972: Eine politische Geschichte. Paderborn, Germany: Schöningh.

Beck

(1992). Risk society: Towards a new modernity. New Delhi: Sage

Beckmanns Sport Lexikon. (1933). Beckmanns Sport Lexikon A-Z. Leipzig, Wien: Verlagsanstalt Otto Beckmann.

10.

Bette

K. H.

(2010). Sportsoziologie. Bielefeld, Germany: Transcript.

11.

Bette

K. H.

Schimank

(2006). Die Dopingfalle: Soziologische Betrachtungen. Bielefeld, Germany: Transcript.

12.

Brewer

B. D.

(2002). Commercialization in professional cycling 1950-2001: Institutional transformations and the rationalization of doping. Sociology of Sport Journal, 19, 276–301.

13.

Bubenhofer

(2009). Sprachgebrauchsmuster. Korpuslinguistik als Methode der Diskurs- und Kulturanalyse. Berlin, Germany: De Gruyter.

14.

Bubenhofer

Schröter

(2012). Die Alpen. Sprachgebrauchsgeschichte—Korpuslinguistik—Kulturanalye. In Maitz

(Ed.), Historische Sprachwissenschaft. Erkenntnisinteressen, Grundlagenprobleme, Desiderate (pp. 263–287). Berlin, Germany and Boston, MA: de Gruyter.

15.

Bürer

Klawitter

(1998, 3 19). Seit 1990 schmückt sich der Westen mit den Sportlern aus DDR-Produktion. Ihre Schöpfer stehen nun vor Gericht. Doping macht vergeßlich. Die Zeit, 13, pp. 17–20.

16.

Cohen

(2002). Folk devils and moral panics: The creation of the mods and rockers (3rd ed.). London, England and New York, NY: Routledge.

17.

Conrad

(2002). Corpus linguistic approaches to discourse analysis. Annual Review of Applied Linguistics, 22, 75–95.

18.

Critcher

(2002). Media, government and moral panic: The politics of paedophilia in Britain 2000 -1. Journalism Studies, 3, 521–535.

19.

Critcher

(2008). Moral panic analysis: Past, present and future. Sociology Compass, 2, 1127–1144.

20.

Critcher

(2009). Widening the focus: Moral panics as moral regulation. British Journal of Criminology, 49, 17–34.

21.

Critcher

(2011). For a political economy of moral panics. Crime Media Culture, 7, 259–275.

22.

David

Rohloff

Petley

Hughes

(2011). The idea of moral panic—Ten dimensions of dispute. Crime Media Culture, 7, 215–228.

23.

Denham

B. E.

(2004). Sports Illustrated, the mainstream press and the enactment of drug policy in Major League Baseball: A study in agenda-building theory. Journalism, 5, 51–68.

24.

Denham

B. E.

(2008). Folk devils, news icons and the construction of moral panics: Heroin chic and the amplification of drug threats in contemporary society. Journalism Studies, 9, 945–961.

25.

Die Kontrollen sind ineffizient. (1998, 7 20). Der Spiegel, 30, p. 164.

26.

Dimeo

(2007). A history of drug use in sport 1876-1976: Beyond good and evil. London, England and New York, NY: Routledge.

27.

Dresen

(2010). Doping im Spitzensport als soziales Problem. Ursachen und Folgen eines gesellschaftlichen Diskurses. Wiesbaden, Germany: VS.

28.

Entman

R. M.

(2004). Projections of power. Framing news, public opinion, and U.S. foreign policy. Chicago, IL: University of Chicago Press.

29.

Federal Government. (2015). Entwurf eines Gesetzes zur Bekämpfung von Doping im Sport. Berlin, Germany: Bundesministerium der Justiz und für Verbraucherschutz.

30.

Franke

W. W.

Berendonk

(1997). Hormonal doping and androgenization of athletes: a secret program of the German Democratic Republic government. Clinical Chemistry, 43, 1262–1279.

31.

Gabrielatos

Baker

(2008). Feeling, sneaking, flooding: A corpus analysis of refugees and asylum seekers in the UK press, 1996-2005. Journal of English Linguistics, 36, 5–38.

32.

Gamper

(2000). Doping: Spitzensport als gesellschaftliches Problem. Zürich, Switzerland: NZZ-Verlag.

33.

Garland

(2008). On the concept of moral panic. Crime, Media, Culture, 4, 9–30.

34.

Gleaves

Llewellyn

(2013). Sport, drugs and amateurism: Tracing the real cultural origins of anti-doping rules in international sport. International Journal of the History of Sport. doi:10.1080/09523367.2013.831838

35.

Goode

(2011). Sports doping as deviance, anti-doping as moral panic. Copenhagen, Denmark: International Network of Humanistic Doping Research.

36.

Goode

Ben-Yehuda

(1994). Moral panics: Culture, politics, and social construction. Annual Review of Sociology, 20, 149–171.

37.

Goode

Ben-Yehuda

(2009). Moral panics: The social construction of deviance (2nd ed.). Chichester, England: Wiley-Blackwell.

38.

Graf von Norman

(Hrsg.). (1928). Deutsches Sport Lexikon. Berlin, Germany: Schwabachersche Verlagsbuchhandlung.

39.

Grimm

(Hrsg.). (1983). Deutsches Wörterbuch. Leipzig, Germany: S. Hirzel Verlag.

40.

Hacke

(2005, 10 31). Triathlon. Allahs Ausnahmeathlet. Der Spiegel, 44, 138–140.

41.

Hall

Critcher

Jefferson

Clarke

Roberts

(1978). Policing the crisis: Mugging, the state and law and order. London, England: Macmillan.

42.

Hall

Jefferson

(Eds.). (1976). Resistance through rituals: Youth Sub-cultures in post-war Britain. London, England: Hutchinson.

43.

Hanstad

D. V.

Smith

Waddington

(2008). The establishment of the world anti-doping agency. A study of the management of organizational change and unplanned outcomes. International Review for the Sociology of Sport, 43, 227–249.

44.

Hardie

McEnery

(2010). On two traditions in corpus linguistics, and what they have in common. International Journal of Corpus Linguistics, 15, 384–394.

45.

Hauser

(2011). “Im Klub der Spitzensportler”-Medienlinguistische Beobachtungen zur kontrastiven Analyse von Dopingskandalen. In Burkow

Petersen

(Eds.), Skandale. Strukturen und Strategien öffentlicher Aufmerksamkeitserzeugung (pp. 207–226). Wiesbaden, Germany: VS Verlag für Sozialwissenschaften.

46.

Heitan. (1931). Doping im Sport. Die Leibesübungen, 4, 111.

47.

Hier

S. P.

(2008). Thinking beyond moral panic: Risk, responsibility, and the politics of moralization. Theoretical Criminology, 12, 173–190.

48.

Hier

S. P.

(2011). Tightening the focus: Moral panic, moral regulation and liberal government. British Journal of Sociology, 62, 523–541.

49.

Hier

S. P.

Lett

Walby

Smith

(2011). Beyond folk devil resistance: Linking moral panic and moral regulation. Criminology and Criminal Justice, 11, 1–18.

50.

Houlihan

(2002). Dying to win (2nd ed.). Strasbourg, France: Council of Europe.

51.

Hunt

(1999). Governing morals: A social history of moral regulation. Cambridge, England: Cambridge University Press.

52.

Hunt

T. M.

(2011). Drug games. The international olympic committee and the politics of doping, 1960-2008. Austin: University of Texas Press.

53.

Ihle

Scharf

(2007). Männer, Mythen, Medien—“Tour de France”-Berichterstattung in Deutschland. Studies in Communication Sciences, 7, 203–229.

54.

Jarren

Donges

(2006). Politische Kommunikation in der Mediengesellschaft: Eine Einführung. Wiesbaden, Germany: VS Verlag.

55.

Kautz

(2010). Blickpunkt Doping – eine explorative Studie zu Meinung und Einstellungen von Sportjournalisten im Bezug zu Doping, die Berichterstattung darüber und mögliche Lösungen des Dopingproblems. München, Germany: Lit.

56.

Kayser

Mauron

Miah

(2007). Current anti-doping policy: A critical appraisal. BMC Medical Ethics, 8, 2.

57.

KhosraviNik

(2010). The representation of refugees, asylum seekers and immigrants in British newspapers: A critical discourse analysis. Journal of Language and Politics, 9, 1–28.

58.

McEnery

(2006). The moral panic about bad language in England, 1691–1745. Journal of Historical Pragmatics, 7, 89–113.

59.

McRobbie

Thornton

S. L.

(1995). Rethinking ‘moral panic’ for multi-mediated social worlds. British Journal of Sociology, 46, 559–574.

60.

Meier

H. E.

Reinold

(2013). Performance enhancement and politicisation of high-performance sport: The West German ‘air clyster’affair of 1976. The International Journal of the History of Sport, 30, 1351–1373.

61.

Meier

H. E.

Reinold

Rose

(2012). Dopingskandale in der alten Bundesrepublik: Öffentlicher Diskurs und sportpolitische Reaktionen. Deutschland Archiv, 45, 209–239.

62.

Partington

(2010). Modern Diachronic Corpus-Assisted Discourse Studies (MD-CADS) on UK newspapers: An overview of the project. Corpora, 5, 83–108.

63.

Perkuhn

Keibel

Kupietz

(2012). Korpuslinguistik. Paderborn, Germany: Wilhelm Fink.

64.

Philipp

(2002). Die Konstruktion des medialen Dopingdiskurses. Wiesbaden, Germany: Deutscher Universitätsverlag.

65.

Prügel im Endkampf. (1985, 7 22). Der Spiegel, 30, pp. 123–125.

66.

Reinold

Meier

H. E.

(2012). Difficult adaptations to innovations in performance enhancement: ‘Dr Brustmann’s power pills’ and anti-doping in German post-war sport. Sport in History, 32, 74–104.

67.

Rose

(2013). Von der Dopingbande zur Dopingmafia—Eine linguistische Analyse des Dopingbegriffs am Beispiel von Dopingakteuren. Spectrum der Sportwissenschaften, 25, 21–43.

68.

Schneider

(2012). Analyse des medialen Dopingdiskurses exemplarisch durchgeführt an der Medienberichterstattung über, Doping-Fahnder’ Jeff Novitzky. In Asmuth

Binkelmann

(Eds.), Entgrenzungen des Machbaren? Doping zwischen Recht und Moral (pp. 203–222). Bielefeld, Germany: Transcript.

69.

Schnyder

(2000). Erfundene Ursprünge und sprechende Bilder. Historische, etymologische und metaphorologische Bemerkungen zum Dopingdiskurs. In Gamper

Mühlethaler

Reidhaar

(Eds.), Doping. Spitzensport als gesellschaftliches Problem (pp. 64–91). Zürich, Switzerland: NZZ Verlag.

70.

Segrave

J. O.

(2000). The sports metaphor in American cultural discourse. Culture, Sport, Society, 3, 48–60.

71.

Shields

Bredemeier

(2011). Contest, competition, and metaphor. Journal of the Philosophy of Sport, 38, 27–38.

72.

Singler

Treutlein

(2000). Doping im Spitzensport, Sportwissenschaftliche Analysen zur nationalen und internationalen Leistungsentwicklung. Aachen, Germany: Meyer & Meyer Verlag.

73.

Steinbrecher

(2009). Politische Partizipation in Deutschland. Studien zur Wahl- und Einstellungsforschung, Band 11. Baden-Baden, Germany: Nomos.

74.

Taylor

(2014). Investigating the representation of migrants in the UK and Italian press: A cross-linguistic corpus-assisted discourse analysis. International Journal of Corpus Linguistics, 19, 368–400.

75.

Thompson

Greek

(2012). Mods and rockers, drunken debutants, and sozzled students: Moral panic or the British silly season? SAGE Open, 2012, 1–13.

76.

Ungar

(2001). Moral panic versus the risk society: The implications of the changing sites of social anxiety. The British Journal of Sociology, 52, 271–291.

77.

Zippert

(2005, 1 5). Der Mautsünder. Die Welt. Retrieved from http://www.welt.de