LITERATURE REVIEW

Political discourse is defined by the nature of participants, type of activity, and context in which it takes place (van Dijk Reference van Dijk1997). In parliamentary debates, the participants are elected representatives who engage in a formal discussion of turn-taking where one speaker is followed by another. However, this orderly turn-taking structure is often disrupted by spontaneous responses of fellow members of parliaments such as laughter, heckling, or interruptions (Illie Reference Ilie, Tracy and Sandel2015). Studies exploring legislative speechmaking along gendered lines are less common and typically explore the issue from two main angles: (1) when do women speak and (2) how often do women speak. Regarding the former, women more often than men engage in and initiate debates about women's issues (Pearson and Dancey Reference Pearson and Dancey2011 on the US Congress; Chaney Reference Chaney2006 and Reference Chaney2008 on Welsh Assembly, 2011 on the Scottish Assembly; Clayton et al. Reference Clayton, Josefsson and Wang2017 on Ugandan parliament; Piscopo Reference Piscopo2011 on the Argentinian parliament). Further, Bäck and Debus (Reference Bäck and Debus2016) find that women speak more often on soft policy issues than hard ones (see also Bäck et al. Reference Bäck, Debus and Müller2014 on the Swedish parliament and Fraune Reference Fraune2016 on energy policy-making in Germany and the United States). In terms of speaking frequency, results are mixed: some studies find that gender makes no differences in how often women speak on the floor (see Wang Reference Wang2014 on Ugandan parliament or Murray Reference Murray2010 on French parliament) while others find that women speak less frequently than men (Bäck and Debus Reference Bäck and Debus2016 on European parliaments or Bäck et al. Reference Bäck, Debus and Müller2014 on the Swedish parliament). In contrast, Funk and Taylor-Robinson (Reference Funk and Taylor-Robinson2014) find that women speak more often than men on committees in the Costa Rican parliament when women make up more than 40% of committee members.

While these studies illustrate that gender matters for the behavior of members of parliament, a major gap in these studies is that they only compare men's behavior on the floor to women's behavior rather than exploring floor interactions between men and women. Interruptions are a regular part of parliamentary debate which ‘serves as signals of the existence of a “real” - and fruitful - debate’ (Carbó, Reference Carbó1992, 37). Interruptions occur when they violate the turn-taking structure of parliamentary debate, i.e. when representatives start talking before the speaker has finished his or her speech (Eecke and Fernandez Reference Eecke and Fernández2016). The ability to hold the floor is a valuable political resource that increases the power of individual representatives; interruptions are attempts by others to absorb this political resource to increase their own power (Shaw Reference Shaw2000).

Despite the relative importance of parliamentary speech in representational acts, few studies have studied interruptions in terms of gender interaction effects (Illie 2010; Shenhav Reference Shenhav2008). One shortcoming of existing studies is that these studies focus on the interrupter only. For example, studies analyze how frequently men and women interrupt parliamentary speakers (see for example Edelsky and Adams Reference Edelsky and Adams1990; Shaw Reference Shaw2000, Reference Shaw and Baxter2006) or they compare interruption patterns by men and women; for example, studies have found that men are more likely to interrupt than women and that men interrupt women more often than other men (see Eecke and Fernandez Reference Eecke and Fernández2016 for an overview of the literature; the same findings have also been found for federal judges in the United States, see Patton Reference Dana Patton2017 or Jacobi and Schweers Reference Jacobi and Schweers2017).

To understand gendered patterns of speech interruption, we need to remember that women representatives do not operate in a vacuum; they exist in a political space that for centuries has been dominated by men and has been formed by rules rooted in masculine identities and behavior. In politics, women are not only outsiders to the political space but also potential disturbers of political power equilibria. Thus, a rise of women in politics might elicit negative reactions against women in politics (Bjarnegard Reference Bjarnegard2013; Sanbonmatsu Reference Sanbonmatsu2008) and women's presence in politics potentially can lead to resistance from male representatives for (1) violating gender roles and (2) for challenging gender hierarchies in an institution where men's qualifications and competence are hardly questioned and male dominance is seen as the norm (Bjarnegard Reference Bjarnegard2013; Murray Reference Murray2014). This assumption is backed up by social psychology and masculinity studies which have shown that men often react negatively or even aggressively to challenges of social gender hierarchies or to threats to their masculine identity, especially if women challenge dominant gender roles and stereotypes.

Social role theory argues that there is an assumed link between what individuals do and what their personal characteristics are (Eagly and Karau Reference Eagly and Karau2002; Rudman et al Reference Rudman2012). When it comes to gender attributes, women are considered to have communal skills, e.g. being warm, kind, or supportive, while men are associated with agentic skills, which typically means being assertive, independent, competitive etc. Thus, society and individuals share certain assumptions concerning what the appropriate social roles are for both women and men. Gender stereotypes build on these shared social gender roles and can be divided into descriptive gender stereotypes, i.e. how women are thought to be, and prescriptive gender stereotypes, i.e. how women should act (Burgess and Borgida Reference Burgess and Borgida1999). In a broad sense, communal skills are linked to lower social status and roles as homemakers and caretakers while agentic skills are linked to higher status social roles such as political leadership. Of course, women (and men for that matter) can exhibit female attributes and skills. In fact, studies show that leaders display a mix of both masculine and feminine attributes depending on the context (Grebelsky-Lichtman Reference Grebelsky-Lichtman2017; Baxter Reference Baxter and Madsen2017; Mullany and Young Reference Mullany, Young and Vine2018). However, this does not negate the fact that women are interlopers in the political space. Further, studies on community of practice show that masculine spaces, such as politics, tend to adopt masculine behavioral norms, making it more likely that women will display stereotypical masculine behavior in parliament (Shaw Reference Shaw2013; Cameron and Shaw Reference Cameron and Shaw2016). For example, Jones (Reference Jones2016) show that Hillary Clinton's language became more masculine the higher she climbed the political ladder.

Status incongruity hypothesis in social role theory holds that resistance against women occurs when women act contrary to gender stereotypes and social role expectations as this is perceived as an attempt to undermine the social gender hierarchy (Rudman et al Reference Rudman2012). Because both men and women are (consciously or not) aware of social roles and gender role expectations, this affects their behavior. For example, Brescroll (2011) in an analysis of US Senate transcripts found that powerful male Senators speak longer while powerful female Senators did not speak longer than less powerful ones. Drawing on social role theory, the author concludes that powerful women limit their speaking time as talkative women are perceived as incompetent and unfit for leadership. Similarly, Bäck and Debus Reference Bäck and Debus2016 suggest that social role theory potentially explains why women take the floor less frequently than men: when women take the floor, they visibly find themselves in a gender-incongruent role which makes it more likely that they are judged as being incompetent. Finally, past studies have shown that societies where traditional gender roles dominate have fewer women in politics (see Norris and Inglehart Reference Norris and Inglehart2001; Paxton and Kunovich Reference Paxton and Kunovich2003).

Masculinity studies argue that the sense of a masculine identity depends on the maintenance of traditional gendered social hierarchies and divisions of labor where men are socially superior to women and where men occupy the public sphere, i.e. politics and business, whereas women remain confined to the private sphere (Cheng Reference Cheng2007). The political space in particular, has been constructed by men for men and has been maintained by the marginalization of women from politics (Kimmel Reference Kimmel1993). When these social hierarchies and gendered divisions of labor are violated, it can open the door to resistance against women in politics ranging from defensive and discriminatory behavior, e.g. displays of prejudiced behavior, sexist jokes, the sexualization of women, benevolent sexism, and even sexual harassment (Dahl et al. Reference Dahl, Vescio and Weaver2015; O'Connor et al. Reference O'Connor, Ford and Banos2017; Netchaeva et al Reference Netchaeva, Kouchaki and Sheppard2015). Indeed, past studies have shown that male representatives react negatively to female representatives when the number of women in politics increases: for example, Kathlene (Reference Kathlene1994) finds that increases in the number of women in US state legislatures lead to men becoming ‘more verbally aggressive and controlling of the [committee] hearing’ (560); she argues that despite a growing presence in the legislatures, women grew more silenced and sidelined in the policy making process. Grey (Reference Grey2006) observed that in New Zealand parliamentary debates saw an increase in statements expressing hostility towards feminism as the number of women in the New Zealand parliament grew. Further, Grey finds that the open hostility towards women's issues has ‘made female parliamentarians (…) reluctant to speak for women in parliamentary debates’ (499). These findings correspond with studies in social role theory which have shown that agentic women are more likely to be subject to harassment in male dominated professions, such as politics, where they are more likely to be subject to interpersonal attacks than attacks based on their competence (Burgess and Borgida Reference Burgess and Borgida1999).

While sexual harassment in politics certainly exists, it is less acceptable when it is committed in plain sight of the constituency; and parliamentary debates are typically available as visual or audio recordings. The same is true for other obvious discriminatory attacks such as sexual assault or other forms of physical violence. Thus, resistance against women in politics needs to occur in more subtle and acceptable ways. Since interruptions are institutionalized and acceptable parts of parliamentary debates, men can utilize interruptions to reassert dominance against their female colleagues. Thus, the key question this article addresses is whether manterruptions can be considered a form of symbolic violence against women in politics. To be considered a form of symbolic violence, manterruptions need to impede women's ability to fulfill their representative functions. I propose the following hypotheses, based on the existing findings in the literature discussed above. First, according to the insight provided by parliamentary discourse studies into parliamentary speech interruptions, I expect that regardless of whether the interruption is a manterruptions or not, women will be more frequently interrupted than men. Second, based on expectations articulated by social role theory, I expect that male representatives are more likely to utilize manterruptions when they interrupt women. Third, based on the assumptions of social role theory that agentic women are more likely to be subject to harassment in male dominated professions, such as politics, I expect that women are more likely to be subject to interpersonal attacks than attacks based on their competence. Finally, based on existing findings in masculine hegemonic studies, I expect that manterruptions should be more frequent if women representatives support policies that would upend male dominance in other spheres, such as corporate board quotas do in the business world.

METHODOLOGY AND CONCEPTUALIZATION

To test these hypotheses, this article analyzes gendered speech interruption patterns in the 234th session of the German Bundestag on April 13, 2013 during the 17th legislative period in Germany (2009–2013). Specifically, the analysis will focus on the Gesetz zur Förderung gleichberechtigter Teilhabe von Frauen und Männern in Führungsgremien (law for the promotion of equal participation of women and men in leadership positions) introduced by the opposition parties (the Social Democratic Party and the Green Party) to establish a legal corporate board quota in Germany. Studying instances of male interruptions in female parliamentary speeches in this context has several advantages: first, this case represents a clear instance of the substantive representation of women as the law intended to address gender discrimination in the appointment procedures to corporate boards. This allows us to understand how men react when they are confronted with women's claim-making. To determine whether men interrupt women differently when the substantive representation of women is involved, I also analyze two additional debates: one concerning a gender-neutral topic (internet governance) and one regarding a masculine topic (finance). Second, because the majority of men in the government coalition opposed the proposed bill, it allows me to study whether men used interruptions as a means to discredit women's arguments or the merit of the bill proposal itself. Third, studying a controversial debate of great national interest is in line with many other studies of parliamentary discourse; while these debates might be atypical, they often expose the quintessential features of parliamentary discourse (Bayley Reference Bayley and Bayley2004).

I define interruptions as ‘intrusions into the current speaker's turn’ (Kollock, Blumenstein, and Schwartz Reference Kollock, Blumstein and Schwartz1985, 38). In this article, I focus on verbal interruptions both institutionalized and non-institutionalized (or illegal) interruptions (Illie 2010). Typically, parliamentary rules allow for institutionalized interruptions. In the German case, institutionalized interruptions take the form of an interposed question - Zwischenfrage - which allows a member of parliament to ask a question. The speaker can reject the request for an interposed question if so desired. Non-institutionalized interruptions in the German parliament typically occur when individual members of parliament heckle the speaker or disrupt the speech with acclamations of support or opposition from the bench. I define manterruptions as unnecessary interruptions by men of women. Unnecessary is defined to mean any interruption that does not substantially add to the debate, or what Bärs (Reference Bärs2010) calls apolitical interruptions, i.e. personal insults. This is in contrast to political interruptions which substantially add to the debate and refer to the content of the speech. I define personal insults to mean any interruptions that either (a) discredit a women's expertise on the subject matter or (b) questions a women's professionalism and qualifications as a representative. Adopting this definition of manterruptions, I follow insights from previous research which show that disruptive discourse practices often ‘exhibit gender-related asymmetries in parliamentary power balances’ (Ilie Reference Ilie2013, 502) and that unparliamentary language (i.e. political insults) are based in social and cultural norms (Ilie Reference Ilie, Tracy and Sandel2015). Because parliament and politics more generally are considered a masculine domain and as previous studies, as discussed above, show, political insults are grounded in traditional gender norms and expected social gender hierarchies.

Speech interruptions are identified and analyzed with the help of a qualitative content analysis of the verbatim parliamentary transcript of the 234th sitting of the Bundestag which featured three separate debates lasting 1.5 hours each. The main analysis is limited to one sitting of parliament to increase the likelihood that the same parliamentarians were present throughout the full sitting as well as to hold the background conditions, such as political issues and climate of the day, constant. Both institutionalized and illegal as well as verbal and non-verbal (laughter, applause, etc.) interruptions are marked in the verbatim transcripts of parliamentary debates as such and give the name and party affiliation of the interrupter and the content of the interruption.

The Bundestag is considered a competitive space where debates take on a combative tone and where parliamentary speakers are frequently interrupted and ridiculed by other members of parliament. On average interruptions occur every 1 to 1.5 minutes during a parliamentary speech (Bär Reference Bärs2010) and are used to confuse and derail the speaker. When debating polarizing issues, the level of respect towards the speaker, and thus the frequency and hurtfulness of interruptions can increase (Steiner et al. Reference Steiner, Bächtiger, Spörndli and Steenbergen2004). Over time, interruptions have become longer and often result in a mini dialogue between speaker and heckler whose purpose is to ridicule the speaker (Burkhardt Reference Burkhardt2016). Interruptions then, are not noteworthy per se. However, Burkhardt (Reference Burkhardt, Günthner and Kotthoff1992) found that women are not only more often subject to heckling than men but that they were more likely to be subject to sexist comments on their appearance or competence, i.e. manterruptions.

I coded individual interruptions noting the frequency, object (who is being interrupted), and subject of the interruption (who is doing the interruption). This allows me to understand whether women are more often interrupted by men than vice versa. Further, to assess whether the gendered interruption patterns displayed in the corporate board quota debate are typical or not, this article also analyzes gendered interruption patterns in two other parliamentary debates during the same parliamentary session: one on a gender-neutral subject (internet governance in public life) and the other one on an issue typically considered to be a masculine policy area (economic stability in the European Union - fiscal aid for Cyprus). Further, I also analyze the verbatim parliamentary transcript of an additional four parliamentary debates on the issue of corporate board quotas which occurred in the 17^th legislative period which lasted from 2009 to 2013 to gain an in-depth understanding of how gendered interruption patterns played out in the overall corporate board quota debate.

ANALYSIS & RESULTS

In total, 51 representatives spoke across all three debates; of those, 32 speakers were men and 19 were women. Table 1 shows the breakdown along each debate and shows that in both the gender-neutral topic (internet governance) and the masculine topic (fiscal policy) men made up the majority of speakers while women represented the majority of speakers on the issue of corporate board quotas.Footnote ¹

Table 1.

Number of Speakers by Sex in all Debates

Across all three debates, I recorded a total of 200 interruptions by individuals. Of these, five were legal interruptions in the form of the Zwischenfrage while the rest were illegal interruptions called out by individual members of parliaments from their seats. To test the first hypothesis that women will be more frequently interrupted than men, I count the number of times female speakers are interrupted when they hold the floor. A simple count of interruptions shows that women were interrupted less frequently than men overall but more often in the corporate board quota debate. However, because women represent fewer speakers overall, we need to weigh the number of interruptions comparatively to the number of female speakers in the overall debate. If all 200 interruptions were equally distributed across speakers, men should experience 63% of all interruptions and women should experience 37% of all interruptions. In total numbers, this means 126 interruptions for male speakers and 74 interruptions for female speakers. Yet women experienced a total of 90 interruptions rather than the expected 74. Thus, despite being fewer speakers overall, women are 17.7% more likely to be interrupted. I can thus confirm the first hypothesis: women are more frequently interrupted then men.

Table 2.

Frequency of Interruptions by Sex of Speaker and Debate

Further, we also need to consider how often female representatives interrupt speakers. If politics is still a male space as many claim, we should see men interrupting women more frequently than the other way around. Contrary to this expectation, I find that women interrupted men more often than vice versa and that men interrupt other men almost twice as often as they interrupt women.

Table 3.

Interruption Dyads by Sex in all Debates

Thus, interruptions of female speakers by male colleagues seems to be more an indicator of regular parliamentary debate behavior than a form of resistance against women.

The next question is how often men interrupt women and how often these man-woman interruption dyads classify as manterruptions (hypothesis 2). To test this hypothesis, I both analyze the count and the content of each interruption. In total, men interrupted women once in the fiscal debate, nine times in the gender-neutral debate and 26 times in the corporate board quota debate. Of these interruptions, zero were manterruptions in the fiscal debate, 44% were manterruptions in the gender-neutral debate, and 15.3% in the corporate board quota debate (see Table 4).

Table 4.

Manterruptions across all Debates

The following exchange is an example for manterruptions questioning the expertise of female representatives in the gender-neutral debate:

In this example, the male representative questions the female speaker's expertise by interjecting that data storage is also part of the issue. When the female speaker rejects his claim, the male representative doubles down and insists that it is the same issue, essentially asserting that he knows more about it than the female speaker. The questioning of expertise is particularly important because speakers are typically picked for parliamentary debates based on their knowledge about the subject. Here, the female speaker – Brigitte Zypries - was the parliamentary secretary of the Economics and Energy Department. In contrast, the interrupter – Volker Beck – was the Green's party speaker on state and religious matters. Thus, while Beck was less qualified to speak on the issue, he nevertheless questioned the expertise and refused to back down when she rejected his attack on her expertise. While, by its very nature, manterruptions’ purpose is to discredit women's voices, in the cases under consideration, manterruptions are not systemic or frequent enough to represent a form of resistance against women in politics or to prevent the substantive representation of women.

Looking at the type of manterruptions, manterruptions targeted women's professionalism or qualifications three times more often than women's expertise on a policy subject. Thus, I can confirm the third hypothesis 3 that women are more likely to be subject to interpersonal attacks than attacks based on their competence. Below is an example for a manterruptions in the corporate board quota debate questioning the professionalism of the female speaker:

This type of interruption serves to question women's professionalism by questioning the seriousness and professionalism with which she approached the subject rather than the merits of the argument presented by the female representative. A much more callous example questioning women's qualification is the exchange below which also occurred during the corporate board quota debate:

Here, a male member of parliament interrupts the female speaker right when she claims that women do not need quotas because the merit system works. In essence, he claims that she is not qualified to be a representative and that she herself in unqualified to be a representative.

Finally, I am unable to confirm the final hypothesis that manterruptions should be more frequent if women representatives support policies that would upend male dominance in other spheres. To test this hypothesis, I analyzed all instances of interruptions across all corporate board quota parliamentary debates in the 17^th legislative period which lasted from 2009 to 2013. In this particular case (and in contrast to the finding above), I find that men interrupt women more often than other men. Using the frequency of interruption as an indicator, this could mean that men react more strongly against attempts to represent women's interest which challenge traditional gender hierarchies in the public sphere. Yet when taking the ratio of female to male speakers into account, women are actually less likely to be interrupted (24% less likely) while men are more likely to be interrupted (43.4% more likely). In terms of manterruptions, I counted a total of 20 instances across all five debates. Yet, overall, manterruptions accounted for only 9.1% of all interruptions across all five parliamentary debates centered on corporate board quotas and thus, do not make up a significant part of the overall interruptions. Further, when controlling for partisanship, I find that in the German case, men never manterrupted female speakers from their own party.