We Strike, Therefore We Are? A Twitter Analysis of Feminist Identity in the Context of #DayWithoutAWoman

Abstract: In this article, we explore the collective identity of feminist activists as expressed on Twitter in the context of "Day Without A Woman." We conceptualize collective feminist identity by drawing upon literature on identity, feminism, and social movements. We expected to find a politically-defined group boundary around supporters of "Day Without A Woman." Using the online tool Netlytic, we collected tweets posted from accounts in Washington D.C., New York City, and Los Angeles. In a preliminary step, we performed a word count analysis and coded frequent words within the collected tweets into categories of meaning. Based on these categories, we drew a sub-sample of tweets, which we scrutinized in-depth using discourse analysis. Through this qualitative analysis, we show that the group boundary of the supporters of "Day Without A Woman" is defined by the common denominator of their negative relation to Donald TRUMP. While the supporters stress the relevance of feminist claims, barriers to identifying as a feminist seem to persist, as reflected in those whom we call "flexi-feminists." The boundary between supporters and non-supporters of "Day Without A Woman" hence seems to broaden from a line to a space which can be occupied without complete group entry. In this space and beyond, supporters express feminist identity through an "us" versus "him" logic.

Key words: collective identity; contemporary feminism; social movement; Twitter analysis; critical discourse analysis; U.S. presidential elections

On January 21st, 2017, an estimated 470,000 people joined the Women's March in Washington D.C. (WALLACE & PARLAPIANO, 2017)—an upheaval which transcended state, national and continental borders (SMITH-SPARK, 2017; WILSON, 2017). Amassing the uprising's momentum, on March 8—International Women's Day—Women's Marches organizers called for a strike of women in both paid and unpaid work. This day was referred to as Day Without A Woman. Those who could not strike were urged to avoid shopping at all businesses not owned by women or members of minority groups, or wear red in solidarity (REDDEN, 2017). Be it by marching or striking that people stood up for women's rights, there seems to be little debate about their protest being a way of expressing objections to Donald TRUMP (see for example MOORE, 2017), who had been inaugurated as president of the United States of America (U.S.) the day before the Women's Marches took place. [1]

In this article, we look at the group of supporters of the protest. More precisely, we analyze feminist identity as an expression of collective identity within "Day Without A Woman"—a current large-scale social phenomenon which has been called a "valuable display [...] of political muscle" (CHIRA, ABRAMS & ROGERS, 2017, §10). As a social movement, feminism inevitably raises the question who "we" are as feminists. This "we" is contested, as the ideas encompassed as part of feminism are interpreted differently among those concerned with them (ARONSON, 2011; HERCUS, 2005; KELLY, 2015). Yet social mobilizations like "Day Without A Woman" rely on a sense of groupness. We address this duality by investigating how appeals to collective identity are made in the context of "Day Without A Woman." With this aim in mind, our research was guided by the following question: Which accounts of collective identity did feminist activists express in the context of "Day Without A Woman" and the aftermath of Women's Marches? By attempting to answer this question, we shed light on the relationship of feminist activists to society, and the significance of the election of Trump for this relationship. [2]

The theoretical contribution that we aim to make lies within debates on the contested nature of feminist identity (ARONSON, 2011; HERCUS, 2005; KELLY, 2015), which has been characterized by phenomena like "fence sitting" where so-called "fence sitters" (p.521) endorse feminist ideas, yet are reluctant to identify as feminists (ARONSON, 2011). In the article, we shed new light on such challenges of feminist identity in a specific socio-political context: the U.S. after the election of TRUMP. Conceptualizing identities as fluid depending on political, historical and social conditions (ELLEMERS, SPEARS & DOOSJE, 2002; HOWARD, 2000), we put forward the working assumption that the election of TRUMP facilitated identification with the feminist movement. We explore manifestations of a theorized shift in the group boundary around contemporary feminists following the election of TRUMP. In this way, we directly address the importance that issues of identity have had for feminism (BRANAMAN, 2011; GONZÁLEZ, 2008; McLAREN, 1999; STONE, 1999) and extend the understanding of feminist identity as societally embedded. Besides these thematically bound contributions, we make a methodological contribution to the literature by combining methods from big data research with an in-depth qualitative analysis. [3]

We have structured the article as follows. First, we develop a conceptual framework of feminist identity as embedded in feminism, identity, and social movements (Section 2 up to 2.3). Based on this framework, we formulate our expectations of shifted group boundaries of contemporary feminists in the context of "Day Without A Woman" (Section 2.4). We then outline how we collected and analyzed tweets posted around March 8 from the U.S.-cities Washington D.C., New York City, and Los Angeles (Section 3). In a preliminary step of data analysis, we used word counts and coded frequent words into categories of meaning. We drew a sub-sample from the categories "in-group reference," "political reference," and "solidarity," which we then analyzed qualitatively using discourse analysis. As we elaborate next (Section 4), an important element of feminist identity within the analyzed tweets seems to be a politically defined group boundary, which is more specifically created through a positioning towards TRUMP and which takes the form of a multidimensional space rather than a line. Before concluding with the idea of a fluid collective identity and group boundary of contemporary feminists, we point to limitations of our research (Section 5). [4]

Feminism as a social movement establishes a frame for feminist identity. More precisely, feminist identity, understood as a collective identity, can be thought of as a composite of the meanings that feminism comprises. Approaches to identity illuminate the conceptualization of feminist identity as a collective one. By encompassing a community, feminism lays the foundation for social identities. Feminists' attachment to this community allows for the formation of a collective identity. Here, we explore this attachment in the context of "Day Without A Woman," which can be considered an episode within feminism. In the following, we develop the concepts of feminism, identity, and social movement in more depth, spelling out how they feed into feminist identity as displayed in Figure 1.

Figure 1: Graphical representation of the conceptual framework surrounding feminist identity [5]

Feminism is expressed in both socio-political and academic forms that may overlap, yet can be distinguished from one another (see Figure 1). In this article, we will focus on feminism understood as a social movement, emphasizing equality (ARONSON, 2011), autonomy, and freedom (FERGUSON, 2007). It addresses issues like the empowerment of women in a male-dominated society (CROSSLEY, TAYLOR, WHITTIER & PELAK, 2011), sexuality, and bodies (HOOKS, 1990). It also critically reflects upon the concepts of sex and gender (BUTLER, 1990). The political dimension of feminism becomes even clearer by viewing it through the lens of identity politics (HEKMAN, 2013, Chapter 1; ZIVI, 2004). Identity politics suggests that "one's identity is taken (and defined) as a political point of departure" (ALCOFF, 1988, p.431). For feminism, this means that identifying as a "woman" or part of another group whose status feminism aims to promote, can inspire political standpoints. [6]

Largely relying on developments in the U.S., three waves of Western feminism can be identified. In the first wave, which formed in the 19th century, equal rights for women were emphasized, arguing in terms of human identity (BRANAMAN, 2011). In the societal climate of leftist and civil rights activism (CROSSLEY et al., 2011), prevalent during the 1960s and 1970s, different streams of feminism developed into what has been referred to as second-wave feminism. This second-wave feminism embodied liberal, Marxist, socialist and radical perspectives. Although to different degrees and in different ways, it focused on issues such as the advancement of women in the occupational context, women's unpaid work, male domination, and gendered socialization processes (BRANAMAN, 2011). Its achievements for women notwithstanding, second-wave feminism has been criticized for taking white middle-class women's perspectives (FERGUSON, 2007; JAIN, 2012). Out of this criticism, feminists began to emphasize differences between women based on ethnicity, social class, nationality, and sexuality during the mid-1980s (BOBEL, 2010; GÓMEZ-SÁNCHEZ & MARTÍN-SEVILLANO, 2006). With this, intersectionality became an important concern in feminism (WHITTIER, 2006), contributing to the rise of Black feminism (CARASTATHIS, 2016, Chapter 1) and to the current third wave of feminism with its focus on multiple perspectives and an inclusive approach (CROSSLEY et al., 2011). Nevertheless, criticism of feminism for taking white middle-class perspectives persists. Moreover, self-labeled feminist approaches which widely ignore structural barriers that women face (EDDY & WARD, 2017), such as Sheryl SANDBERG's (2013) call on women to "lean in" remain prominent. Such liberal approaches tend to supplant identity dimensions with a focus on self-fulfillment in terms of exploiting opportunities for economic advancement, perpetuating a capitalist logic (ASCHOFF, 2015, Chapter 1). There is a tension between identity-focused and liberal approaches within feminism where the latter seem to represent socially privileged women. Addressing this tension, Nancy FRASER (1997, Chapter 1) has argued that issues of redistribution remain relevant and are intertwined with issues of identity-based recognition. The work of bell HOOKS, for example, connects questions of social injustice to the feminist debate by advocating an intersectional approach (e.g., HOOKS, 2000, Chapter 7). [7]

Some scholars prefer referring to these recent developments in feminism as "postmodern feminism" (BRANAMAN, 2011, p.33), emphasizing their critical outlook on previous versions of feminism or "contemporary feminism" (REGER, 2008, p.101), stressing their fragmentary nature (EVANS, 2015, Chapter 1). We follow scholars in adopting the term "contemporary feminism" for describing recent feminist developments, being hesitant to claim sufficient distance from an ongoing social development for considering it as one "wave." Hence, contemporary feminism is used here as reference point for discussing issues relating to feminist identity. [8]

The use of social media platforms, such as Twitter, by feminists has been referred to as a fourth wave of feminism (EVANS & CHAMBERLAIN, 2014). Following this categorization, the current analysis with its focus on expressions of feminist identity on Twitter would also fall under the fourth wave. However, in line with CROSSLEY et al. (2011), we regard social media as a new dimension of contemporary feminism rather than marking the beginning of a new wave. Therefore, we understand tweets in the context of "Day Without A Woman" as a manifestation of contemporary feminism. [9]

Considering the label "feminist" as a basis of collective identity presupposes some sense of community. We consider feminism to represent such a community in constituting a social movement (ibid.) as displayed in Figure 1, where social movements are understood as "organizations, submerged networks, and ideologically structured challenges to a variety of different institutional authorities" (TAYLOR, KIMPORT, DYKE & ANDERSEN, 2009, p.867). According to Nancy WHITTIER (2002), social movements—and therefore feminism, we argue—can be analyzed by looking at three levels: structure, strategies and collective action, and meaning. Meaning, in this framework, is understood as composed of collective identity and discourse. We will focus on collective identity within meaning (see Figure 1), taking an inside-out perspective. [10]

WHITTIER (2002) stresses the mutual constitution of internal structure and meaning. She attributes dynamics in meaning within the feminist movement to cultural, political, and economic changes. In "Dynamics of Contention," Doug McADAM, Sidney TARROW and Charles TILLY (2001) identify mechanisms of social movements, which they claim occur across contexts, but produce different outcomes depending on those contexts. One such mechanism is identity shift, describing how new identity boundaries form in a changing political context. Such a shift of feminist identity is what we expect to occur in the period after the U.S. presidential elections, and we further suggest that the socio-political frame influences towards which issues people position themselves and how they do so (see Section 2.4). [11]

McADAM et al. (2001, p.309) situate mechanisms like identity shift within what they call "episodes," which serve as units of observation. Taking up their approach, contemporary feminism can be viewed as an episode. When breaking this episode further down, the period after the election of TRUMP emerges as a lower-level episode. By putting "Day Without A Woman" as a one-day episode at the center, this analysis is even further narrowed down. As a one-day episode within the larger scale episodes, "Day Without A Woman" is connected to the overall movement. We do not claim that an identity shift happened on "Day Without A Woman" specifically, but that such a shift is embedded within the larger-scale episode in the longer period after the U.S. presidential elections. Jacquelien STEKELENBURG and Conny ROGGEBAND (2013) suggest that while globalization, individuation and diversification, and virtualization make new forms of interaction and organization possible, older forms of social movements have remained important. Through the current analysis, we take up this development by using online expressions of activists which are a rather new phenomenon, but which remain connected to protest and strike as an older form of activism. It is at this intersection that we investigate collective identity within a particular political frame. [12]

Identity has been discussed by various academic disciplines and at various levels of abstraction, making it impossible to capture all relevant facets of the concept (HOWARD, 2000). Post-structuralist perspectives stress the construction of identity through discourse (HAN, 2011a). Social psychological perspectives, on the other hand, focus on the specific conditions and processes through which identities emerge. Despite such differences, there is some consensus across approaches. One important consensual point is the dynamic character of identities (McGARY & McGRATH, 2013, Section 3). From a postmodern point of view this means that "authenticity is not determined by being true to oneself for all time, but rather being true to self in context or self in relationship" (HOWARD, 2000, p.387). In an attempt to recognize such a dynamic character of identity, feminist researchers have proposed metaphors of identity, such as "mestiza" (ANZALDIA, 1987), "nomade" (BRADOTTI, 2004), and "cyborg" (HARAWAY, 1995; all cited in GONZÁLEZ, 2008, p.22). While being aware of these critical views of identity, we consider a deconstruction of the concept to be at odds with tapping into relevant self-work that individuals engage in. We view self-work and its implications for social groups as important by themselves, even given their internal inconsistencies. We therefore conceptualize identity as a dynamic structure which is highly social in its process of construction and which serves the purpose of giving meaning to the self. [13]

More precisely, we focus on social identities (BREWER, 2001) (see Figure 1). We follow HOWARD (2000) in regarding them as derived from membership in social groups as opposed to personal identities, which are determined by distinctive characteristics of the self. Within the umbrella category of social identities, we focus on the subcategory of collective identities (see Figure 1) which can be understood as "an individual's cognitive, moral, and emotional connection with a broader community, category, practice, or institution" (POLLETTA & JASPER, 2001, p.285). Here, collective identity is applied to feminism as has been suggested in social movement research (BERNSTEIN, 2008; FLESHER FOMINAYA, 2010). Moreover, Twitter constitutes a distinct online environment which has been linked to expressions of identity in general (HAN, 2011b; HOWARD, 2000), and to the construction of "public selves" in particular (HARRIS, 2011, p.531), making it especially salient for the study of expressions of feminist identity (for a discussion of identity in online spaces, see GOODINGS, 2011). [14]

It has been argued that even though identity questions were only explicitly addressed in feminism in the 1970s (CROSSLEY et al., 2011), implicitly identity has always been central to feminism (GONZÁLEZ, 2008). Feminism aims to advance the position of women in society. Inevitably, this brings up questions regarding the boundaries around which this specific group of women should be defined. These questions are further complicated by the critical evaluation of gender identities (BUTLER, 1990). Feminism might even appear to be losing its ground if the category of women is de-constructed altogether (BERNSTEIN, 2008). Some have thus argued for strategic essentialism (see for example COATES, 2012), seeing the category of women as crucial for reaching feminist objectives. From this perspective, it seems most appropriate for feminism to continue to rely on women as a category. In this way, feminism as a social movement paves the way for a group-based identity, which we look at in the following through tweets written by feminist activists. The anchoring point for the collective identity under scrutiny are those who identify as women and the community of feminist activists who may or may not identify as women. Contemporary feminism rethinks perspectives on identity through an internal re-organization, using gender as one dimension among many along which identities are defined. Put differently, in questioning the notion of a "unified woman" (CROSSLEY et al., 2011, p.502), contemporary feminism seems to offer a multiplicity of feminist identities. This sheds new light onto collective identity of feminists, as it redefines whom the group represents. [15]

Claiming collective identity by labeling oneself a feminist seems to be complex. A range of researchers suggest that negative stereotypes are attached to feminists (CROSSLEY, 2010; ROBNETT, ANDERSON & HUNTER, 2012), which can hinder identification as a feminist (ROY, WEIBUST & MILLER, 2007). While anti-feminist public discourse has been observed during second-wave feminism, the perception of feminism as unnecessary has above all been linked to contemporary feminism (ARONSON, 2011), thereby enabling negative associations with the label "feminist." In other studies, authors found predominantly positive and neutral associations with the label "feminist" (TWENGE & ZUCKER, 1999). They stress that assumptions about others' negative associations with the label "feminist" can hinder self-labeling as such. Similarly, Amanda B. BREEN and Andrew KARPINSKI (2008) point to the discrepancy between predominantly positive associations that people have with feminists and the reluctance to identify as a feminist. Other factors, such as exposure to feminist ideas (HERCUS, 2005) and the experience of sexism (CROSSLEY, 2010) have also been connected to identification as a feminist. ARONSON (2011, p.521) sheds light onto the connection between feminist ideas and identity by coining the term "fence sitter" to describe those who embrace some feminist principles, reject others, and identify neither as a feminist, nor as a non-feminist. Alyssa N. ZUCKER (2004, p.427) develops a similar categorization in calling women "egalitarians" when they endorse gender equality while not labeling themselves as feminists. [16]

Taken together, researchers within this stream suggest that identification as a feminist does not only depend on one's experiences and beliefs, but also on assumptions about how feminists are seen by society. Insights from the social maintenance model (TESSER, 1988) on people's motivation to hold positive views of the self support this reasoning: to the extent that the self is seen as part of society, identifying with a group that is assumed to trigger negative associations stands in contrast to this motivation of aiming to maintain a positive view of the self. As group boundaries within society change, re-evaluations of this group are possible (COHRS & KESSLER, 2013), thus potentially reducing the incongruence between a positive view of the self and identifying with this group. In other words, applied to our analysis, the apparent contradiction between aspiring to be evaluated positively but identifying with a group that is evaluated negatively diminishes as feminists are perceived more positively. Without making any claims in terms of causal mechanisms, we argue that TRUMP winning the U.S. presidential elections triggered a re-formation of group boundaries of feminist activists, potentially reducing the threshold to identify as a feminist. [17]

We take the notion of identities as fluid and changing based on time, with specific political and historical (HOWARD, 2000) as well as social (ELLEMERS et al., 2002) conditions as a starting point. Our claim is that this fluidity enables identity expressions along political dimensions after the election of TRUMP. We argue that such expressions are manifested under the umbrella of contemporary feminism. In turn, this facilitates identification with feminism in two ways. [18]

ELLEMERS et al. (2002) identify two dimensions along which social identities can be classified: threat to the social group (no threat versus individual-directed threat versus threat to the group) and group commitment (low versus high). We argue that the political situation in the U.S. can be seen as imposing a group-directed threat, given TRUMP's sexist ways of talking about women, as they became evident in the "Locker Room Talk" recordings in October 2016, and his policy announcements, such as the planned restriction on health measures ensuring women's reproductive health. The election of TRUMP might make a large number of women increasingly experience disadvantage based on their gender. With this, the perceived irrelevance of feminist ideas to one's life, which has been found to be a barrier to identifying as a feminist (ARONSON, 2011), becomes less applicable. We argue that this is how the political context in the U.S. facilitates identification as a feminist. As an alternative hypothesis, it is also conceivable that, as a result of intensified experience of disadvantage, women more strongly display social identities attached to other more positively evaluated social groups. Regardless of this being imaginable on the individual level, we argue that the facilitated identification as a feminist remains a key reaction to the election of TRUMP, which is discernible on the collective level. [19]

Regardless of expected policy outputs, the election of TRUMP automatically created the group of non-supporters of TRUMP. This is reflected, for instance, in socio-cultural ways through the use of the hashtag "#NotMyPresident." We build our argument partly based on the umbrella function that contemporary feminism has served by making claims on behalf of different disadvantaged social groups defined along various dimensions, such as sexuality and ethnicity. This understanding of contemporary feminism provides a frame in which the political dimension can operate as a new unifying factor. We argue that political orientation emerges as a salient group boundary between feminist activists and the rest of society, resulting in an "us" versus "them" division between supporters and non-supporters (see Figure 2). [20]

The implications of this are two-fold. First, we assume that the prominence of individuals' political orientation broadens the group of those engaging in feminist activism. This, we further propose, automatically changes how feminist activists as a group relate to other social groups and society at large, since new members also belong to other social groups. Such changes in how a social group compares to other groups are a strong determinant of how the group is evaluated, as ELLEMERS et al. (2002) suggest. Hence, we presume that a re-evaluation of feminists as a social group became possible after the election of TRUMP. We assume that this re-evaluation facilitates identification with the label "feminist," which has previously been perceived in more negative ways. We regard such identification to be for instance reflected in using the hashtag "#DayWithoutAWoman." Second,—and this will be our main expectation—we expect TRUMP non-supporters to be represented in the tweets we analyze. While their group boundary is defined through political distinctions, we also expect them to be subsumed under the umbrella of feminist activism.

Figure 2: New group boundary of contemporary feminists after the election of TRUMP [21]

We take two interrelated steps in exploring accounts of collective identity expressed by feminist activists, combining a text mining approach with an in-depth analysis (similar to WALDHERR et al., 2019, though using different methods; see also WIEDEMANN, 2013, on the relationship between big data and qualitative research). A preliminary step consists in inquiring into frequent expressions through a word count analysis of a large sample of tweets. This allows a broad representation of the data. We abstract these representations by coding frequent words into categories of meaning. Based on these categories, we develop sampling criteria allowing us to select tweets that seem especially relevant to questions of collective identity. In the second and main step, we scrutinize a sub-sample of these tweets in depth using critical discourse analysis. We base our discussion of the results primarily on this second analysis. By doing so, we aim to gain rich insights into feminist identity. [22]

Pursuing the goal of capturing expressions of feminist identity, we chose to collect tweets, being accessible social media data (PURDAM & ELLIOT, 2015). We applied an intensity sampling logic aiming at information-rich cases (PATTON, 2002, pp.231-246; SCHREIER, 2017, pp.87-88). We defined intensity in terms of the number of people who participated in the Women's Marches in the U.S. on January 21st, 2017 along a location dimension. This resulted in sampling tweets from Washington D.C., New York City, and Los Angeles. In using a purposeful sample, we refrain from making claims of generalizability. While we cannot make conclusions about their prevalence in the population, we aim to grasp meaningful dynamics, as they unfold in these intense cases. [23]

Tweets were collected via a web scraper (IGNATOW & MIHALCEA, 2017, pp.39-41), namely the online tool Netlytic. This "text analyzer and social networks visualizer"¹⁾ draws upon the application programming interface (API) of Twitter and has been used for analyzing online content, such as in SANTAROSSA, COYNE, LISINSKI and WOODRUFF (2016). While Netlytic also collects meta-data, we focus on the tweets' textual body as characteristics that vary between "objects," (MANOVICH, 2015, p.16) the tweets. Since Twitter users make their text messages publicly available and especially since we do not focus on personal information, we argue that ethical concerns with this data collection are low (AMPOFO, COLLISTER, O'LOUGHLIN & CHADWICK, 2015). [24]

We assured a thematic focus of tweets on feminist activism by using appropriate search terms. After having screened Twitter manually to investigate which hashtags and words come up frequently in online expressions of current feminist activism, we created a list of search words. This resulted in the identification of eighteen hashtags which seemed to capture the available tweets purposefully for the current research (PATTON, 2002, pp.242-246) as a point of thematic redundancy seemed to have been reached in the screening process. Given the limited number of search criteria that Netlytic allows to specify, we had to choose a maximum of six hashtags to be included in the search. Two criteria guided us in narrowing down the choice. We prioritized hashtags oriented towards action taking (e.g., "#WomensStrike") and those that seemed most closely connected to the feminist movement in particular (e.g., "#IStrikeFor"). We refined the search criteria through several trial runs. For example, contrary to what we had planned initially, the handles "@womensmarch" and "@daywithoutawoman" were removed from the search. This had the purpose of reducing the error of collecting tweets that do not qualify as expressions of feminist activists (tweets mentioning these handles often questioned their goals). Even with the exclusion of these handles, this type of error cannot be ruled out completely. Based on the trial runs, we also decided to only use four hashtags: "#WomensMarch," "#DayWithoutAWoman," "#WomensStrike," and "#IStrikeFor." It may be argued that this made the search too restrictive. But in fact this did not reduce the number of tweets collected. Moreover, we argue that it introduces transparency and a focus to exploring expressions by feminist activists. For ease of analysis, the language was set to English. This means that only tweets from accounts that are set to English were retrieved. We excluded re-tweets from the search in order to keep the search as broad as possible without reducing the relevance of the information. In addition, we specified the posting time period of tweets to be collected (March 7 to March 9). In majority, this returned tweets from March 8. Lastly, we applied the location criterion based on the proceeding described on the website Thoughtfaucet, which Netlytic recommends as a resource. The full search phrases are provided in Table 1.

After having collected 1.002 to 1.008 tweets per location via web scraping following the described procedure, we cleaned the data by hand in preparation for an exploratory word count analysis—one of the most frequently used techniques in text mining (ZHAI & MASSUNG, 2016, p.68)—through the tool provided within Netlytic. The data cleaning involved removing empty objects and duplicates, merging the datasets from the different locations, as well as separating strings of words and grouping semantically similar words (such as plural and singular forms of the same word, and equivalent words used with and without a hashtag). Removing stop words, which occur at high frequencies and therefore add little to analyses (IGNATOW & MIHALCEA, 2017, p.55), was done automatically by Netlytic.³⁾ [26]

Using a basic lexicometric analysis (WIEDEMANN, 2013), we identified a total of 148 words that occurred in at least seven different tweets. Words that appeared in less than seven tweets were disregarded. We set the boundary to seven due to the comparably high difference in the number of words that were included in seven rather than six tweets. Following Graham R. GIBBS' (2007) recommendations of applying a data-driven approach informed by our research focus, we coded these words into categories of meaning (see Appendix 1). We understand the emerging categories as popular thematic dimensions of meaning, along which Twitter users represented in this sample express themselves. These categories can be summarized as referring to actions, emotions, ideas, space, time, politics, media, and social groups. In Figure 3, we show the different categories and their frequencies, meaning the added counts of all words within one category across tweets. A total of 17 words did not seem to be classifiable due to a lack of expressive meaning (e.g., "it's") or their lexical ambiguity (e.g., "social"). They constitute the category "open." We do not claim that the categories shown in Figure 3 are the only ones in which the data can be organized. Instead, we regard them as one possible representation of the data, which fits the purpose of selecting tweets for an in-depth analysis under the given research question.

The purpose of the coding was not only to represent the data in a condensed way, but also to "enable the [...] retrieval of thematically-related sections" (GIBBS, 2007, p.48). This corresponds to what Carla WILLIG (2015) refers to as one of the first steps in discourse analysis. In other words, the thematic analysis, which we based on word counts, allowed us to single out tweets that seem especially interesting with respect to collective identity and therefore the given research question. When recalling feminist identity as an individual's connection to women and feminists as communities, tweets within the category "in-group reference" seem most closely related to collective identity from a conceptual point of view. In-group, here, is understood as feminists forming a social movement as well as in terms of people on behalf of whom feminism makes claims. Tweets within the category "political reference" point towards a political dimension in the data. To gain an understanding of how this political dimension manifests itself, we decided to also include them in the in-depth analysis. In order to explore meaning that the categories could not capture, we also selected tweets containing the word "solidarity" from the category "open." We chose "solidarity" because it was the most frequent among non-categorized words, and for the purpose of the analysis seems relevant as such due to its social nature. In Figure 4, we provide an overview of the words that were coded into the categories "in-group reference" and "political reference." The share of each word within the respective category is shown. The category "solidarity" only consists of the word "solidarity" and is therefore not displayed.

Figure 4: Words within the categories "in-group reference" and "political reference." For the category "in-group reference" the 1.505 counts (equal to 86.7% of the displayed word counts) of the word "woman" are not displayed for better readability. [28]

This conceptually-informed pre-selection of tweets laid the foundation for a qualitative analysis of parts of the large dataset, building the focus of the analysis. Through this in-depth analysis, we attempted to reduce the "semantic gap" (MANOVICH, 2015) which arose in the word count analysis as words were detached from their context. The large-scale analysis thus served as a preparatory step rather than an analysis in itself. This two-step procedure follows the recommendation to complement big data analysis with more traditional qualitative methods (FELT, 2016; VIS, 2013). Due to the limited capacity of tweets that we could analyze in depth, we drew a subsample of tweets (see Appendix 2) from each of the three selected categories. While we do not aim to make claims of representativeness, for each category in focus ten tweets were selected randomly by hand so as to give some insight into the discourses that the tweets in a given category draw upon. [29]

The sub-sampled tweets were analyzed using critical discourse analysis. Discourse analysis is suitable for our research questions because it allows researchers to explore multiplicities of meaning (PHILLIPS & HARDY, 2002, p.22-23) and is best applied to naturally occurring text (WILLIG, 2015). In addition, discourse analysis allows researchers to account for the context in which language is produced (TITSCHER, MEYER, WODAK & VETTER, 2000, Chapter 11). This makes it even more fitting to the purpose of exploring feminist identity, which is assumed here to be influenced by a socio-political event, namely the election of TRUMP. This specific context inspired our decision for critical discourse analysis within the broad scope of discourse analysis traditions for two reasons. Firstly, critical discourse analysis allows for a focus on distal contexts like this one (PHILLIPS & HARDY, 2002; RUIZ RUIZ, 2009, however, asserts limited attention to broader social contexts in critical discourse analysis). Secondly, the socio-political context after the election of TRUMP introduces power issues, which critical discourse analysis is able to capture (FAIRCLOUGH, 2001; KENDALL, 2007; WODAK, 2001; on the use of critical discourse analysis in a feminist context, see PAPADOPOULOS & SCHRAUBE, 2004). Within the variety of approaches towards critical discourse analysis (MEYER, 2001), we loosely follow Norman Fairclough's analytical frame as described by TITSCHER et al. (2000, Chapter 11, see also RUIZ RUIZ, 2009). Thus, we focus on descriptions (What are the text's linguistic features?), interpretations (Which type of discourses are represented in the text?), and explanations (How do these discourses relate to one another and to the social reality at hand?). [30]

In line with FAIRCLOUGH (2013), we adopt a critical realist perspective by viewing feminism as an object that exists in "reality," whilst also allowing that a subjective construction of this object (feminism) is interpreted differently by different actors. It would be desirable to more closely examine Twitter, ingrained in the public sphere, as part of this "reality." Twitter is the immediate environment which together with the public sphere forms the conditions under which the text, i.e., the tweets, are produced, distributed, and consumed. These conditions are one important dimension of discourse in Fairclough's framework (REISIGL, 2013). A close examination of them, however, does not seem feasible within the scope of the current analysis. Instead, our focus is on the text itself and its socio-political embeddedness as a dimension of discourse (ibid.). In so doing, we follow FAIRCLOUGH (2013) and consider text and context to be connected through a dialectical relationship where they are "different but not discrete" (p.4). Consequently, the construction of text and context is co-dependent. This means that, for example, the tweets—as an expression of feminist activism—are not only anchored in feminism, but also shape the constitution of feminism. We analyze tweets as one important manifestation of this text-context relationship. [31]

Following WILLIG's (2015) recommendation to first gain an impression of the material, we read through all the tweets that contain words that had been coded into the selected categories. Second, we analyzed the subsample of tweets by answering the three analytical questions taken from TITSCHER et al. (2000, Chapter 11) for each tweet. While the focus of each question is different, they build upon one another. Therefore, we attempted to connect insights across questions as well as tweets as early as possible during the process of analyzing (see also RUIZ RUIZ, 2009 for the interconnectedness of different analytical focuses). In a first round of analysis, we adopted a broad perspective and explored possible meanings. In a second step, our focus was narrower: we viewed the tweets through a collective identity lens. Through this procedure overarching themes emerged, which we discuss below. The procedure also made some tweets stand out as especially "rich" data, which we have used to suggest some overall tendencies. [32]

Through the qualitative analysis, we demonstrate how feminist values, such as equality, respect, and human dignity, are linked across time and space. This builds a common basis for feminist identity, which is further underpinned by the personal relevance of feminist claims that Twitter users express, such as the promotion of women in the workplace. Two main themes around collective identity emerge: a politically defined and personalized group boundary, as well as perspectives on the in-group that are contingent on the position of contemporary feminists in society. In the following section, we discuss these findings. [33]

Appeals to values, most often equality, but also respect and dignity recur across tweets. These values are phrased as being promoted "for" women and beyond. Adherence to equality is, on the other hand, phrased as being beneficial not exclusively for women, but for everyone ("An equal society is a better society for all of us," Tweet 22; tweet numbers refer to numbers given to tweets in the subsample, see Appendix 2). One Twitter user reports wearing clothes that disclose concern for both feminist and anti-racist ideas. Putting on both items of clothing would not be necessary, but the activist seems to purposefully want to merge the two standpoints on their body ("Wearing my Black Lives Matter red t-shirt with my Feminist AF sweatshirt tied around my waist," Tweet 10). This underlines that "Day Without A Woman," being an episode of contemporary feminism, addresses the advancement of various social groups in society, not only of women. [34]

In another tweet, the author depicts the promoted values as inherent to "women-identified people" (Tweet 21). This suggests that regardless of whether their entitlement to these values is recognized or not, their respect, dignity, and equity persists ("#DayWihtoutAWoman, a time to reflect on the respect, dignity, and equity [of] ALL women-identified people," Tweet 21). By adopting an unconventional categorization of people, it is not entirely clear what might be meant by "women-identified people." We will discuss this in some more detail in Section 4.3.1. The described appeals to values demonstrate a moral aspect of contemporary feminism. [35]

In the data, the promotion of feminist values is characterized by continuity. A connection to the past is for example established by referring to the poem "Phenomenal Woman" (Tweet 5) by the civil rights activist Maya ANGELOU from 1997. Through other tweets, it is conveyed that the claims of feminism continue to be relevant "today and every day" (Tweet 27) and "every single day" (Tweet 20). The author of yet another tweet calls for empowering women across generations ("#strongwomen, [...] may we be them, may we raise them," Tweet 2). Similarly, some tweets are clearly oriented towards action taking ("Let's do this," Tweet 10; "join us to continue the fight," Tweet 27), proposing an endurance of feminism. The feminist fight is hence connected to the past, present, and future. One tweet author presents continuous negative individual experience as evidence for the legitimacy of feminist claims, ending the underrepresentation of women in work contexts in this case ("#Istrike because I'm tired of being the only woman in the room.," Tweet 3). Apart from time, space is another dimension that feminism is described to cut across. Twitter users refer to local (e.g., "DC," Tweet 17) and global (e.g., "around the world," Tweet 22) levels that "Day Without A Woman" both, reaches and should reach. [36]

In a nutshell, within the analyzed tweets, contemporary feminism is given a moral character through values that it is being linked to while being made salient across contexts. Both may be seen as forming a strong basis for—or the result of—the connection of contemporary feminism to a community. [37]

Frequently, the political sphere is introduced by referring to TRUMP. These ways range from mentioning people that stand in association with him (e.g., "Donald Trump [a]dministration," Tweet 18) or his politics (e.g., "#TrumpGlobalGag," Tweet 17), to directly addressing him or his Twitter account ("@realdonaldtrump," e.g., Tweet 16), to borrowing features from discourse associated with him (e.g., "THE RISE OF THE NATION," Tweet 7). Often, these references to TRUMP contain elements of humor. For example, one tweet author uses irony when depicting TRUMP's environment as not implementing the feminist claim of women being represented in positions of power ("Thanks @realDonaldTrump, for observing #DayWithoutAWoman every single day," Tweet 20). By thanking TRUMP directly, TRUMP is ascribed the capacity of agency in this tweet. However, he does not use this capacity for promoting the representation of women in his environment in this description. Instead, he remains passive by "observing." In this way, TRUMP is presented as an opposing power to feminist claims through his passiveness. This not only illustrates TRUMP's salience in discussing "Day Without A Woman," but also the aversion that is expressed towards him from within the feminist movement. This finding underlines the separation between feminism and TRUMP rather than feminism and non-feminism. [38]

Another Twitter user distances themselves from women in TRUMP's environment and even devalues them ("[D]id the Trump women say or do anything about #InternationalWomensDay #DayWithoutAWoman? I wasn't paying attention to them," Tweet 14). TRUMP is listed first as a characterizing element and only within TRUMP's social environment, women are specified as a group. This implies a hierarchy of group distinctions: Being on the side of TRUMP seems to override being a woman. "Trump women," being women and therefore belonging to those on whose behalf feminism makes claims, could be regarded as part of the in-group defined through feminism. Yet a distancing from them occurs by "n[o]t paying attention to them." Going further, the Twitter user illustrates aversion to TRUMP (women) through the use of irony in their tweet. Irony is introduced by asking whether these "[said] or [did] anything" and reinforced by the unexpected answer to this question. By these means, the author of the tweet seems to criticize others for not implementing feminist ideas. At the same time, the author does not adhere to the idea of contemporary feminism as accommodating all women. We propose that in this case the aversion to TRUMP might supersede the concern for solidarity among women. An alternative interpretation could be that the tweet authors aims at pointing towards the underrepresentation or absence of women in the TRUMP administration resulting from TRUMP's sexist conduct. In either case, TRUMP is constructed as a clear point of reference located outside of feminism, further highlighting the separation between the two. [39]

Direct references to the political spectrum in terms of parties are also frequent in the sample of tweets (e.g., "#democrats," "@HouseDemocrats"). Identification with a certain political orientation or group is made most explicit by one tweet author referring to "Our @HouseDemocrats" (Tweet 15). They include themselves in the political community of Democrats while singling out its political representatives. By addressing politicians directly, the tweet authors seem to use them to gain access to the political arena. More specifically, one Twitter user seems to talk to Democratic Members of Congress about TRUMP ("@TomPerez @MariaPTCardona Wonder if Trump is wearing his ubiquitous red tie today [...]?," Tweet 12). In a figurative sense, this seems to build a bridge in the form of Democratic representatives from the tweet author to TRUMP. In our view, this illustrates felt proximity between the tweet author and the mentioned Democratic politicians on the one hand, and distance from TRUMP (who is not directly addressed) on the other. The separation between feminist activism and TRUMP thus unfolds further. [40]

In the same tweet, the author then situates TRUMP with regard to "Day Without A Woman." The text becomes ironic through the underlying assumption that TRUMP would not intend to show "solidarity with women in #daywithoutawoman" (Tweet 12) by wearing a red tie. By phrasing this assumption as a question, TRUMP is thought of as aiming to distance himself from the movement. In turn, feminism is recognized as a movement which TRUMP opposes. Through the tweet, its author creates a situation for TRUMP in which he cannot avoid taking a position towards "Day Without A Woman." If he wears his supposedly "ubiquitous" red tie, he shows solidarity. If he chooses to wear a tie of a different color, his action can be framed as a reaction to the movement. Hence, in this tweet the Twitter user constructs a situation where the feminist movement asserts power over TRUMP. A similar logic is applied in another tweet, in which Linda SARSOUR, one of the main organizers of Women's March and "Day Without A Woman," is framed as "worries of the Donald Trump Administration" (Tweet 18). We view the authors of these tweets as upending socio-political power dynamics. This might be re-asserting the power of feminism, as activists in the movement not only fight for substantive concerns, but also against perceived obstacles to addressing these concerns, such as TRUMP. We regard such displays of strength as demonstrating self-confidence in the movement up to the extent that one of its main opposing powers is not taken seriously. This would also mean that features of the feminist movement itself are formed—and hence the construction of the collective identity attached to it occurs—through TRUMP in the context of "Day Without A Woman." [41]

While the authors of these tweets accentuate the opposition between TRUMP on the one side and feminists on the other, a different Twitter user links the two. The Twitter user takes up TRUMP's patriotic discourse and applies it to feminist ideas ("THE RISE OF THE WOMAN=THE RISE OF THE NATION," Tweet 7). The tweet author reverses TRUMP's logic by equating the goals—formulated in very general, abstract terms—that patriotism and feminism follow. On the other hand, Hilary CLINTON has also been associated with an emphasis on patriotism (LIEVEN, 2016). The tweet could therefore also be understood as confirming the affiliation of the feminist movement with TRUMP's opponent in the U.S. presidential elections. If CLINTON's appeals to patriotism are understood as a reaction to TRUMP's patriotic discourse, this tweet can be viewed as mimicking CLINTON's political move. It would then reiterate the movement's association with the Democratic community. Either way, the author of this tweet politicizes the feminist movement, drawing upon elements of the current socio-political context in the U.S. [42]

The data also contains tweets in which authors criticize claims made by the contemporary feminist movement. For example, this is evident from the use of "#tcot" (short for top conservatives on Twitter). Through the following example, we illustrate how "#tcot" is put into context. One Twitter user employs the hashtag in a row with several other politically-loaded terms and embeds this in a critique of the feminist movement ("Get real and protest the serious crimes of persecution #StopSharia #WomensMarch #MAGA [short for Make America Great Again] #tcot #ccot [short for Christian conservatives on Twitter] #RedNationRising," Tweet 31). This allows further insight into how identifying with the U.S. traditional political distinction between Republicans and Democrats is being linked to a sense of "us" versus "them" in the context of "Day Without A Woman." Here, the Twitter user identifies with the Republican side and positions themselves against "Day Without A Woman" by questioning the importance of the issues it addresses. At the same time, the Twitter user may still see themselves as a feminist pursuing other goals than those represented by "Day Without A Woman." This is in line with the expectation that identifying with a political orientation functions as a boundary between supporters and non-supporters of the contemporary feminist movement associated with actions like a "Day Without A Woman," as we outlined in Section 2.4. [43]

In sum, in the tweets we discussed in this section feminism is linked to politics, especially TRUMP's politics, and feminism and Democrats are separated from TRUMP. These findings are in line with the expected relevance of the election of TRUMP for the boundary around the group of contemporary feminists. The most prominent defining group notion even seems to be "us" versus "him" rather than "us" versus "them." [44]

As outlined above, there are different levels of feminist community. It can be differentiated between those on whose behalf feminism makes claims and those who actively engage in promoting feminist ideas. While the former are defined through the conceptualization of feminism, the formation of the latter is more dynamic. [45]

Among those groups on whose behalf feminism makes claims, women emerge as the most salient group. While tweets whose authors "#[...]strike for humans everywhere" (Tweet 32), implying a more general reference to groups that face disadvantage in society, are represented in the data, they stand out as exceptions. [46]

Even more explicitly than feminist ideas, women are seen as being connected with one another across space ("wom[e]n all over the world," Tweet 9; "women around the world," Tweet 22). Despite this spatial unity, differences between women are implicitly recognized. Stressing the validity of feminist claims for "ALL women-identified people" (Tweet 21) anticipates the questioning of a "unified woman" (p.502), which CROSSLEY et al. (2011) have outlined as an element of contemporary feminism. Yet, referring to "women-identified people" might restrict who is regarded as part of the in-group. For this Twitter user, it might not suffice to be a woman or willing to promote women's status in society. Instead, one might have to free oneself from all male-dominated influences, as is suggested from some feminist standpoints (see for example blogger PURPLE SAGE, 2015). [47]

Whether unity or differences between women are stressed, the views of women that our sample of Twitter users convey through their tweets are overwhelmingly positive. Some counter-examples, however, imply strong sexist attitudes and draw upon prejudice against women (e.g., "#daywithoutawoman significant reduction of crashes on the road!," Tweet 33). Positive views are often charged with sentiments, revealing an emotional attachment to contemporary feminism as another aspect of collective identity. On a larger scale, such positive views are reflected in the use of the hashtag "#SheInspiresMe." Another example is that women are "celebrat[ed]" (Tweet 8). Likewise, even without knowing what the poem that one Twitter user refers to is about, its title "Phenomenal Woman" (Tweet 5) elicits an image of a strong, admirable woman. Besides evoking an emotional dimension simply by referring to lyrical work, the author of this tweet takes a strong position for women. Displaying a certain degree of frustration and aggressiveness, one Twitter user goes further by saying that women are inevitably necessary for life ("[N]one of this shit exists without wom[e]n ... don[']t ever forget that," Tweet 1). We interpret this as suggesting that women's contributions are not recognized by society to the degree this tweet author would find adequate. [48]

These positive ways of viewing women stand in contrast with the experiences of being a woman that are portrayed in some tweets. One Twitter user connects the underrepresentation of women in the work context to being treated in an inhumane manner by accentuating that she is "#NotAToken" (Tweet 3). Here, negative individual experience is framed as a consequence of conditions that feminism aims to change. The author of a different tweet (Tweet 16) reflects upon another woman's agreement to being "grabb[ed by] her 'bits'" by Donald TRUMP. The agreement that the woman expresses through the slogan on her T-shirt is re-phrased as a desire, something this woman "pine[s] for." The author—a man who describes himself as an "Angry Democrat"—reacts with strong disgust, expressed by "Yuck." He interprets the woman's agreement as a consequence of not receiving enough attention—a potential result of the unfulfillment of feminist goals—and links the woman's behavior to the wider societal frame by expressing disappointment about it being possible ("So sad a woman could be so attention starved"). [49]

The implications of these tweets in which Twitter users uncover negative experiences are two-fold. Firstly, they articulate a sense of the necessity of feminism. Secondly, feminism becomes personally relevant, which is reflected in emotion-laden expressions feeding into a sense of attachment to a feminist community. [50]

Identification with feminist activists constitutes another source of building a sense of "us." In one tweet, the author clearly separates feminist activists from institutionalized state power. The author depicts the views of feminist activists as diverging between representations of state power and in-group members ("law enforcement [...] see [Women's March] as a white woman's group—and so they treat us like white women," Tweet 4). This Twitter user places themselves within the group of feminist activists and affirms that the group's members are more diverse than white women. Institutional power, on the contrary, is perceived as communicating the notion of a unified woman through its actions. Firstly, we interpret this tweet as a criticism of police violence towards people of color in the U.S. Secondly, we take the contrast that is reflected in the tweet as highlighting the importance of the perspective that the police, a representative of the state, have on the feminist movement. It suggests that as long as society does not adopt a differentiated image of who feminist activists are, one goal of feminism, namely that all women be recognized and treated equally, has not been reached. [51]

Such an awareness of societal perspectives on the in-group is also reflected in the presumption that feminists may be perceived negatively (CROSSLEY, 2010; ROBNETT et al., 2012). This is most openly expressed by tweeting "If u have any issues with that word, I have major issues with u. #feminism #feminist" (Tweet 6). Interestingly, this Twitter user includes both an ideological attachment to and a personified identification with feminism. This might exemplify that the two are intertwined yet separate, and that both feed into the construction of a boundary around the group of contemporary feminists. Similarly, the notion that it may be surprising for others to identify with feminism is reflected in other tweets. The authors of these tweets re-emphasize their identification with the feminist movement by for instance not simply identifying as feminist, but as "feminist AF [as fuck]" (Tweet 10), wearing "bright red" (Tweet 17) and "fully" (Tweet 9) supporting "Day Without A Woman." For feminist identity, this means that a barrier of identifying as a feminist persists, but can be overcome. [52]

One way of reducing the degree of identification while not eliminating it and hence reducing the barrier of association with the feminist movement is to show solidarity. The ways in which solidarity is embedded into tweets point towards a clear differentiation of and by those who support feminism from those who the movement claims to represent. [53]

Solidarity is expressed by a diverse set of Twitter users, individuals as well as groups. Among these are work teams (e.g., @BCakeNY) and less formal unions, such as young playwrights (@YPTDC). They show solidarity with women rather than identifying with contemporary feminism themselves. Similarly, a "[w]omen-[o]wned (and -run) [s]mall business" presents themselves as standing in solidarity with their "sisters" (Tweet 26). Given the reference to International Women's Day preceding this expression of solidarity, the author shows signs of a strong identification with women in their tweet. The tweet author may also more directly refer to the feminist movement: In Tweet 26, the author demonstrates some identification with feminism by borrowing a phrase from feminist discourse, namely the term "sister" which has not only been employed in discussions about feminism, such as by Nima NAGHIBI (2007) who writes about "global sisterhood," but has also been taken up by Women's March National as part of feminism to refer to marches elsewhere in the world. Further indicating support for feminism, the author of the tweet alludes to women's ability to lead a business. While these linguistic elements seem to epitomize identification with the feminist movement, solidarity is chosen as the main expressive motif. In another tweet (Tweet 27), the author seems to demonstrate solidarity from the movement's inside in a similar way. Given the declared goal of the author, "[a] statewide network to improve the economic outlook for New York women," they can as such be seen as an active feminist player. Mirroring this role, they include themselves in fighting for feminist goals ("Join us to continue the fight"). At the same time, they "stand in solidarity [with] #DayWithoutAWoman" instead of regarding themselves as an integral part of the movement around "DayWithoutAWoman," suggesting a certain separation from this specific manifestation of feminism. [54]

Overall, the data shows a distinction between those standing in solidarity with and those positioning themselves within the feminist movement. Those who show solidarity may be seen as an extension of "fence sitters" (ARONSON, 2011, p.521). So-called fence sitters endorse feminist ideas but are reluctant to identify as a feminist or non-feminist. Those who show solidarity go a step further by openly supporting feminist objectives. The election of TRUMP, which imposes a threat onto women, might have facilitated this step of approaching identification as a feminist. Yet, the step is not complete, as those who stand in solidarity with women restrict themselves to showing solidarity without explicitly adopting the label "feminist." Those who show solidarity are therefore on the inside of the movement in one way and separate from it in another. A space in between "us"—the in-group of contemporary feminists—and "them" seems to exist. The boundaries around the community that feminist identity is built upon within this sample of tweets might not be a thin line but seems to broaden to a space which can be occupied without complete group entry. This space represents the overlap in endorsement of feminist ideas in the current socio-political setting by people from both the outside and the inside of the group. [55]

This research is limited in different ways, particularly with respect to the technicalities linked to the data used. One limitation inherent to Twitter analyses is that "the online [...] becomes the baseline" (VIS, 2013, p.4). This means that people who do not use Twitter are excluded. At the same time, however, Twitter is inclusive in the way that it is accessible for those who cannot engage in other forms of feminist activism, such as demonstrations, due to physical or temporal constraints. Through the collected tweets, we looked at a subset of Twitter users who are only a subset of the population of interest (BOYD & CRAWFORD, 2012), in this case contemporary feminists. Setting the language to English further limited whose tweets were retrieved. Nevertheless, since we explicitly chose an online environment which promises identity-rich expressions and provides ease of access to expressions by feminist activists , the chosen procedure seems legitimate. [56]

The research is further complicated by technical restrictions imposed by the Twitter API. The Twitter API does not give access to all public tweets, but only to a subset of them, limiting the scope of people whose expressions are analyzed (BOYD & CRAWFORD, 2012; FELT, 2016). By using Netlytic, transparency is further impeded, since it is not entirely clear how the tweets that Netlytic returns are chosen among all tweets in the Twitter API that fulfil the search criteria. This raises questions of the validity of the data. The "black box nature" (VIS, 2013, p.9) of tools like Netlytic is for instance evident from tweets of some accounts being represented several times. This produces the need to clean the data, as we did, which adds subjective elements to the analysis. To broaden the large-scale analysis, it would have been desirable to collect more, if not all tweets posted around March 8 from the different locations. This would have, however, not ruled out concerns of data error (BOYD & CRAWFORD, 2012). While it becomes clear that the data that we collected is necessarily restricted, by carefully screening the data manually and combining this with analyses inbuilt into Netlytic, we have attempted to reduce the data error. [57]

Although we have argued that ethical concerns are low in the given analysis, they should not be ignored. In analyses using data that is accessible online like this one, the author of the online content is left in the dark about their data being used for research (VIS, 2013). This introduces power issues where the researcher decides how the researched become part of the inquiry. Ways of gaining informed consent from participants in large-scale analyses, which does not yet seem feasible with the current means of big data research, need to be further investigated (BOYD & CRAWFORD, 2012). Despite these limitations, since the purpose in this analysis was not to generalize but to explore expressions, and since the large-scale analysis was only a preparatory step for an in-depth analysis, we argue that the data and procedure are adequate. [58]

Further research should investigate the located dynamics of feminist identity in constantly changing socio-political frames and follow up on how those inhabiting the space of solidarity continue to position themselves towards the feminist movement. Hypotheses based on the dynamics identified here should be tested with different samples. [59]

Moreover, taking a longitudinal approach would allow researchers to investigate the dynamics of feminist identity around a re-composition of the socio-political frame in the U.S. and elsewhere. Such an approach would complement the focus on the inside of a social movement with a more strongly pronounced macro perspective than this analysis allows for, as Rachel L. EINWOHNER, Jo REGER and Daniel J. MYERS (2008) suggest. A longitudinal approach to feminist identity in the U.S. would make the nature of the situatedness of "Day Without A Woman" as an episode within contemporary feminism increasingly clear and thus contribute to tracing the ongoing development of the feminist movement. The snapshot-like analysis of feminist identity that we carried out in this article mirrors contemporary developments in the feminist movement, such as an emphasis on differences between women. This illustrates the dynamic character of feminism. But the question remains to what extent the uncovered dynamics around feminist identity will persist in the future. Put differently, researching feminist identity longitudinally would give insight into where the feminist movement is heading and where it is coming from. More precisely, this would shed light onto whether the election of TRUMP, then understood as a discursive event and possible turning point for feminism, remains an influential landmark for feminist identity or whether it "just" constitutes/triggered a short-lasting momentum (for an analysis of spontaneously emerging right-wing mobilization in Germany contingent on socio-political events, see NAM, 2017). [60]

In another line of research, researchers should investigate expressions of feminist identity in other contexts and via other mediums, both online and offline. This would yield a more comprehensive understanding of collective identity of contemporary feminists. Further research may also include visuals to capture multiple layers and meanings (FLICK, 2014), which VIS (2013) regards as "tokens of collective identity" (p.5) in online contexts. Taking Twitter user data, such as how they describe themselves, systematically into account might further enrich the results. Due to technical restrictions and the limited frame for this article, it was, however, not possible to incorporate these additional aspects. [61]

Based on these results, we suggest that barriers to feminist identity persist while the election of TRUMP indeed seems to have at least re-characterized, if not shifted the group boundary around contemporary feminists. Our research highlights the complex implications of the contested nature of feminist identity as this identity interacts with the socio-political context. [62]

The group boundary, which defines contemporary feminists and therefore constitutes the basis for a collective identity, is strongly constructed by referring to TRUMP. Rather than the notion of "us" who support feminism and who object to TRUMP versus "them" who do not, an "us" versus "him" distinction emerges. An important common denominator among supporters of "Day Without A Woman" is negative, the positive elements of feminist identity go together with non-explicit feminist and anti-TRUMP identities. In this situation, different streams within contemporary feminism, which, in their heterogeneity, are partly opposed to one another, can act as a relatively united force. We expected the onset of this new group-defining dimension to facilitate identification as a feminist for people who previously located themselves on the outside of contemporary feminists as a social group. However, it rather seems that, while some approaching towards contemporary feminist activists takes place, identification as a feminist remains contested. At the same time, the feminist movement’s values and continuity across time and space pave the way for attachment to the community of contemporary feminists. [63]

Taken together, this calls for re-thinking the group boundary around contemporary feminists. Rather than a line it might be a space, inhabited by people who support feminist ideas, yet do not consider themselves as part of the in-group. Put differently, the boundary between the inside and the outside of contemporary feminists takes shape as an overlapping support for feminist ideas from both sides in this specific socio-political setting. This overlap, which we understand as a space of solidarity, makes it easier to approach and distance oneself from feminism, as evaluations of the group change. This speaks for the importance of the societal context for social identities. The experience of threat, produced by the socio-political climate, and the immediacy of feminist claims stemming from this threat seem to enable a closer positioning toward contemporary feminism. The election of TRUMP might have made some enter the space of solidarity, but it did not suffice for making them enter the group of feminists. Once opposition to TRUMP fades out, those who show solidarity can leave the space they have been inhabiting without crossing a group boundary. It seems that not only collective identity is fluid, but that which shape the boundary around contemporary feminists takes is as well. [64]

The mobilization around "Day Without A Woman" as analyzed here points towards a re-organization of feminist identity. Yet, it remains to be seen whether or not this mobilization, in concert with other present ones, constitutes a discursive event, understood as a turning point that will change the content and direction of the discourse it is associated with (JÄGER, 2001). Instead of a discursive event, the present mobilization might merely be an episode without lasting effect. On the one hand, other movements under the presidency of TRUMP, such as #Metoo, demonstrate that feminist mobilization continues to be strong. Cutting across different societal groups, these forms of mobilization further rearrange the societal position of contemporary feminists as a social group. From a feminist identity perspective, they can provide a low-barrier entry point to contemporary feminists, as individual experiences of sexual harassment and (bodily) threat merge into feelings of injustice on a collective level. The most recent Women's Marches, however, show some signs of fading mobilization. The main march took place on January 19, 2019 in Washington D.C. and had a remarkably lower turnout than the marches in 2017 and 2018 according to participants (WINES & STOCKMAN, 2019). While Women's Marches organizers speak of an "annual march" and the #WomenWave, controversies around Women's Marches organizers have arisen. Among others, Tamika MALLORY and Linda SARSOUR have been accused of supporting anti-Semitic thought and remarks (STOCKMAN, 2018). This accusation, as well as concerns about insufficient inclusiveness with respect to minority women (BRICE-SADDLER, 2019), seems to weaken the movement. These developments might in the end outweigh its potential to become a turning point for feminist identification which the socio-political setting clearly lays the ground for. [65]

While the socio-political setting has shown some power shifts, it largely remains to be a trigger for feminist mobilization. We were inspired to write this article by the societal attention that feminism has received through Women's Marches and "Day Without A Woman." This attention may be considered a success for the feminist movement. The circumstances of these protests—the election of a U.S. President who openly objectifies women—however, are the opposite of feminist success. The election of TRUMP not only undermines the practical implementation of feminist concerns in society, but also perpetuates non-feminist ideas from a culturally powerful position. Beyond that, the election has demonstrated that a large share of the U.S. population is ready to vote against or at least without priority for feminist ideas. Through our analysis, we have shown how central the election of TRUMP seems to be for defining who "we" as contemporary feminists are. In the midterm elections in November 2018, the Democrats regained seats in the House of Representatives and representation at the state level, mirroring a partial reorganization of political power within the U.S. But even with this recent development it would be naïve to claim that feminist concerns do not remain especially important under the presidency of TRUMP. Issues of equality and racism remain salient in U.S. politics, for example in discussions around Brett KAVANAUGH. TRUMP and his politics seem to have come to increasingly serve as a tangible object of political and feminist mobilization in the sense that abstract political elements were imported into socially-bound and simultaneously personal spheres of identity. TRUMP might hence remain an important reference point in the definition of the group boundary around contemporary feminists. [66]

Overall, those who we found show solidarity from neither the inside nor the outside of the feminist movement, but from in-between, might make up a considerable part of what has become visible as success of the feminist movement. Behind this success, feminist claims might stand on shaky ground, as those showing solidarity might be "flexi-feminists." This bears the risk of losing them from feminist arenas, which they had never fully entered, and would make the apparent success of the feminist movement a fragile one. However, such flexibility does not need to be negative for feminism. Quite the opposite: maybe feminism needs a bit of flexibility- and fragility-embracing outreach to gain broader societal acceptance and can leave refining itself for later. [67]

We would like to thank Audris UMEL for her support with conducting the Netlytic searches and two anonymous reviewers for their comments on the manuscript.

Alcoff, Linda (1988). Cultural feminism versus post-structuralism: The identity crisis in feminist theory. Signs, 13(3), 405-436.

Ampofo, Lawrence; Collister, Simon; O'Loughlin, Ben & Chadwick, Andrew (2015). Text mining and social media. When quantitative meets qualitative and software meets people. In Peter J. Halfpenny & Rob Procter (Eds.), Innovations in digital research methods (pp.161-192). London: Sage.

Aronson, Pamela (2011). Feminists or "postfeminists"? Young women's attitudes toward feminism and gender relations. In Verta Taylor, Nancy Whittier & Leila J. Rupp (Eds.), Feminist frontiers (pp.516-526). New York, NY: McGraw-Hill Higher Education.

Bernstein, Mary (2008). The analytic dimensions of identity: A political identity framework. In Jo Reger, Daniel J. Myers & Rachel L. Einwohner (Eds.), Identity work in social movements (pp.277-301). Minneapolis, MN: University of Minnesota Press.

Bobel, Chris (2010). New blood: Third-wave feminism and the politics of menstruation. New Brunswick, NJ: Rutgers University Press.

boyd, danah & Crawford, Kate (2012). Critical questions for big data. Communication & Society, 15(5), 662-679.

Branaman, Ann (2011). Feminism and identity. In Anthony Elliott (Ed.), Routledge handbook of identity studies (pp.30-48). New York, NY: Routledge.

Breen, Amanda B. & Karpinski, Andrew (2008). What's in a name?: Two approaches to evaluating the label feminist. Sex Roles, 58(5-6), 299-310.

Brewer, Marilynn B. (2001). The many faces of social identity: Implications for political psychology. Political Psychology, 22(1), 115-125.

Butler, Judith (1990). Gender trouble: Feminism and the subversion of identity. New York, NY: Routledge.

Carastathis, Anna (2016). Intersectionality: Origins, contestations, horizons. Lincoln, NE: University of Nebraska Press.

Coates, Jennifer (2012). Gender and discourse analysis. In James P. Gee & Michael Handford (Eds.), The Routledge handbook of discourse analysis: Part I: Approaches to discourse (pp.90-103). London: Routledge.

Cohrs, Jan Christopher & Kessler, Thomas (2013). Negative stereotypes, prejudice, and discrimination. In Agnieszka Golec de Zavala & Aleksandra Cichocka (Eds.), The social psychology of social problems: The intergroup context (pp.3-29). Basingstoke: Palgrave-Macmillan.

Crossley, Alison Dahl (2010). "When it suits me, I'm a feminist": International students negotiating feminist representations. Women's Studies International Forum, 33(2), 125-133.

Crossley, Alison Dahl; Taylor, Verta; Whittier, Nancy & Pelak, Cynthia Fabrizio (2011). Forever feminism: The persistence of the U.S. Women's Movement, 1960-2011. In Verta Taylor, Nancy Whittier & Leila J. Rupp (Eds.), Feminist frontiers (pp.498-516). New York, NY: McGraw-Hill Higher Education.

Eddy, Pamela L. & Ward, Kelly (2017). Problematizing gender in higher education: Why leaning in isn't enough. In Pamela L. Eddy, Kelly Ward & Tehmina Khwaja (Eds.), Critical approaches to women and gender in higher education (pp.13-40). New York, NY: Palgrave Macmillan.

Einwohner, Rachel L.; Reger, Jo & Myers, Daniel J. (2008). Introduction: Identity work, sameness and difference in social movements. In Jo Reger, Daniel J. Myers & Rachel L. Einwohner (Eds.), Identity work in social movements (pp.1-17), Minneapolis, MN: University of Minnesota Press.

Ellemers, Naomi; Spears, Russell & Doosje, Bertjan (2002). Self and social identity. Annual Review of Psychology, 53, 161-186.

Evans, Elizabeth (2015). The politics of third wave feminisms: Neoliberalism, intersectionality, and the state in Britain and the US. London: Palgrave Macmillan.

Evans, Elizabeth & Chamberlain, Prudence (2014). Critical waves: Exploring feminist identity, discourse and praxis in Western feminism. Social Movement Studies, 14(4), 396-409.

Fairclough, Norman (2001). Critical discourse analysis as a method in social scientific research. In Ruth Wodak & Michael Meyer (Eds.), Methods of critical discourse analysis (pp.121-138). London: Sage.

Fairclough, Norman (2013). Critical discourse analysis: The critical study of language (2nd ed.). London: Routledge.

Felt, Mylynn (2016). Social media and the social sciences: How researchers employ big data analytics. Big Data & Society, 3(1), 1-15, https://doi.org/10.1177/2053951716645828 [Date of Access: April 11, 2019].

Ferguson, Michaele L. (2007). Sharing without knowing: Collective identity in feminist and democratic theory. Hypatia, 22(4), 30-45.

Flesher Fominaya, Cristina (2010). Collective identity in social movements: Central concepts and debates. Sociology Compass, 4(6), 393-404.

Fraser, Nancy (1997). Justice interruptus: Critical reflections on the "postsocialist" condition. New York, NY: Routledge.

Gibbs, Graham R. (2007). Analyzing qualitative data: Thematic coding and categorizing. London: Sage.

Gómez Sánchez, Lucía & Martín-Sevillano, Ana Belén. (2006). Feminism and identity in political psychology. Feminism & Psychology, 16(1), 65-72.

González, María Martínez (2008). Feminist praxis challenges the identity question: Toward new collective identity metaphors. Hypatia, 23(3), 22-38.

Goodings, Lewis (2011). The dilemma of closeness and distance: A discursive analysis of wall posting in MySpace. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 12(3), Art. 16, http://dx.doi.org/10.17169/fqs-12.3.1576 [Date of Access: March 29, 2019].

Han, Sam (2011a). The fragmentation of identity: Post-structuralist and postmodern theories. In Anthony Elliott (Ed.), Routledge handbook of identity studies (pp.83-99). New York, NY: Routledge.

Han, Sam (2011b). Virtual identities: From decentered to distributed selves. In Anthony Elliott (Ed.), Routledge handbook of identity studies (pp.219-235). New York, NY: Routledge.

Harris, Anita (2011). Young women, late modern politics, and the participatory possibilities of online cultures. In Verta Taylor, Nancy Whittier & Leila J. Rupp (Eds.), Feminist frontiers (pp.527-535). New York, NY: McGraw-Hill Higher Education.

Hekman, Susan Jean (2013). Feminism, identity, and difference. New York, NY: Routledge.

Hercus, Cheryl (2005). Stepping out of line: Becoming and being a feminist. New York, NY: Routledge.

hooks, bell (2000). Feminism is for everybody: Passionate politics. Cambridge, MA: South End Press.

Howard, Judith Ann Kathleen (2000). Social psychology of identities. Annual Review of Sociology, 26, 367-393.

Ignatow, Gabe & Mihalcea, Rada F. (2017). Text mining. A guidebook for the social sciences. Los Angeles, CA: Sage.

Jäger, Siegfried (2001). Diskurs und Wissen. In Reiner Keller, Andreas Hirseland, Werner Schneider & Willy Viehöver (Eds.), Handbuch Sozialwissenschaftliche Diskursanalyse: Band I: Theorien und Methoden (pp.81-112). Wiesbaden: VS Verlag für Sozialwissenschaften.

Jain, Devaki (2012). Passion: Driving the feminist movement forward. Journal of Women's History, 24(4), 201-207.

Kelly, Maura (2015). Feminist identity, collective action, and individual resistance among contemporary U.S. feminists. Women's Studies International Forum, 48, 81-92.

Kendall, Gavin (2007). What is critical discourse analysis? Ruth Wodak in conversation with Gavin Kendall. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 8(2), Art. 29, http://dx.doi.org/10.17169/fqs-8.2.255 [Date of Access: March 28, 2019].

Lieven, Anatol (2016). Clinton and Trump: Two faces of American nationalism. Survival, 58(5), 7-22.

Manovich, Lev (2015). Data sciences and digital art history. International Journal for Digital Art History, 1, 13-34, https://doi.org/10.11588/dah.2015.1.21631 [Date of Access: April 11, 2019].

McAdam, Doug; Tarrow, Sidney & Tilly, Charles (2001). Dynamics of contention. Cambridge: Cambridge University Press.

McGary, Orla & McGrath, Brian (2013). "A virtual canvas"—Designing a blog site to research young Muslims' friendships & identities. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 14(1), Art. 1, http://dx.doi.org/10.17169/fqs-14.1.1805 [Date of Access: March 29, 2019].

McLaren, Margaret A. (1999). Two feminist views on the self, identity and collective action. Hypatia, 14(1), 120-125.

Meyer, Michael (2001). Between theory, method, and politics: Positioning of the approaches to CDA. In Ruth Wodak & Michael Meyer (Eds.), Methods of critical discourse analysis (pp.14-31). London: Sage.

Naghibi, Nima (2007). Rethinking global sisterhood: Western feminism and Iran. Minneapolis, MN: University of Minnesota Press.

Nam, Sang-Hui (2017). Spontaneous mobilization and the transformation of collective formations on the internet: A case study of the PEGIDA movement. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 18(1), Art. 3, http://dx.doi.org/10.17169/fqs-18.1.2642 [Date of Access: March 28, 2019].

Papadopoulos, Dimitris & Schraube, Ernst (2004). "This world demands our attention". Ian Parker in conversation with Dimitris Papadopoulos and Ernst Schraube. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 5(3), Art. 14, http://dx.doi.org/10.17169/fqs-5.3.558 [Date of Access: March 28, 2019].

Patton, Michael Quinn (2002). Qualitative research & evaluation methods (3rd ed.). Thousand Oaks, CA: Sage.

Phillips, Nelson & Hardy, Cynthia (2002). Discourse analysis: Investigating processes of social construction. Thousand Oaks, CA: Sage.

Polletta, Francesca & Jasper, James M. (2001). Collective identity and social movements. Annual Review of Sociology, 27, 283-305.

Purdam, Kingsley & Elliot, Mary (2015). The changing social science data landscape. In Peter J. Halfpenny & Rob Procter (Eds.), Innovations in digital research methods (pp.25-58). London: Sage.

Reger, Jo (2008). Drawing identity boundaries: The creation of contemporary feminism. In Jo Reger, Daniel J. Myers & Rachel L. Einwohner (Eds.), Identity work in social movements (pp.101-120). Minneapolis, MN: University of Minnesota Press.

Reisigl, Martin (2013). Critical discourse analysis. In Robert Bayley, Richard Cameron & Ceil Lucas (Eds.), The Oxford handbook of sociolinguistics (pp.67-90). Oxford: Oxford University Press.

Robnett, Racheal D.; Anderson, Kristin J. & Hunter, L. E. (2012). Predicting feminist identity: Associations between gender-traditional attitudes, feminist stereotyping, and ethnicity. Sex Roles, 67(3-4), 143-157.

Roy, Robin E.; Weibust, Kristin S. & Miller, Carol T. (2007). Effects of stereotypes about feminists on feminist self-identification. Psychology of Women Quarterly, 31(2), 146-156.

Ruiz Ruiz, Jorge (2009). Sociological discourse analysis: Methods and logic. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 10(2), Art. 26, http://dx.doi.org/10.17169/fqs-10.2.1298 [Date of Access: March 28, 2019].

Sandberg, Sheryl (2013). Lean in: Women, work, and the will to lead. New York, NY: Alfred A. Knopf.

Santarossa, Sara; Coyne, Paige; Lisinski, Carly & Woodruff, Sarah J. (2016). #fitspo on Instagram: A mixed-methods approach using Netlytic and photo analysis, uncovering the online discussion and author/image characteristics. Journal of Health Psychology, 24(3), 376-385.

Schreier, Margrit (2017). Sampling and generalization. In Uwe Flick (Ed.), The Sage handbook of qualitative data collection (pp.84-98). London: Sage.

Stekelenburg, Jacquelien van & Roggeband, Conny (2013). Introduction: The future of social movement research. In Jacquelien van Stekelenburg, Conny Roggeband & Bert Klandermans (Eds.), The future of social movement research: Dynamics, mechanisms, and processes (pp.xi-xxii). Minneapolis, MN: University of Minnesota Press.

Stone, Lynda (1999). Experience and performance: Contrasting "identity" in feminist theorizings. Studies in Philosophy and Education, 18(5), 327-337.

Taylor, Verta; Kimport, Katrina; Dyke, Nella van & Andersen, Ellen A. (2009). Culture and mobilization: Tactical repertoires, same-sex weddings, and the impact on gay activism. American Sociological Review, 74(6), 865-890.

Tesser, Abraham (1988). Toward a self-evaluation maintenance model of social behavior. In Leonard Berkowitz (Ed.), Advances in experimental social psychology (Vol. 21, pp.181-227). New York, NY: Academic Press.

Titscher, Stefan; Meyer, Michael; Wodak, Ruth & Vetter, Eva (2000). Methods of text and discourse analysis. London: Sage.

Twenge, Jean Marie & Zucker, Alyssa N. (1999). What is a feminist? Evaluations and stereotypes in closed- and open-ended responses. Psychology of Women Quarterly, 23(3), 591-605.

Vis, Farida (2013). A critical reflection on big data: Considering APIs, researchers and tools as data makers. First Monday, 18(10), https://firstmonday.org/article/view/4878/3755 [Date of Access: March 23, 2019].

Waldherr, Annie; Wehden, Lars-Ole; Stoltenberg, Daniela; Miltner, Peter; Ostner, Sophia & Pfetsch, Barbara (2019). Inductive codebook development for content analysis: Combining automated and manual methods. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 20(1), Art. 19, http://dx.doi.org/10.17169/fqs-20.1.3058 [Date of Access: March 28, 2019].

Whittier, Nancy (2002). Meaning and structure in social movements. In David S. Meyer, Nancy Whittier & Belinda Robnett (Eds.), Social movements: Identity, culture, and the state (pp.289-307). Oxford: Oxford University Press.

Whittier, Nancy (2006). From the second to the third wave: Continuity and change in grassroots feminism. In Lee A. Banaszak (Ed.), The US women's movement in global perspective (pp.45-68). Lanham, MD: Rowman & Littlefield.

Wiedemann, Gregor (2013). Opening up to big data: Computer-assisted analysis of textual data in social sciences. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 14(2), Art. 13, http://dx.doi.org/10.17169/fqs-14.2.1949 [Date of Access: March 28, 2019].

Willig, Carla (2015). Discourse analysis. In Jonathan A. Smith (Ed.), Qualitative psychology. A practical guide to research methods (3rd ed., pp.143-167). London: Sage.

Wilson, Cherry (2017). Women's March: UK protesters join anti-Donald Trump marches. BBC Politics, January 21, http://www.bbc.com/news/uk-38700123 [Date of Access: May 1, 2017].

Wodak, Ruth (2001). What CDA is about—A summary of its history, important concepts and its developments. In Ruth Wodak & Michael Meyer (Eds.), Methods of critical discourse analysis (pp.1-13). London: Sage.

Zhai, ChengXiang & Massung, Sean (2016). Text data management and analysis: A practical introduction to information retrieval and text mining. San Rafael, CA: Morgan & Claypool.

Zivi, Karen (2004). Who or what are we? The identity crisis in feminist politics. Polity, 36(2), 323-340.

Zucker, Alyssa N. (2004). Disavowing social identities: What it means when women say, "I'm not a feminist, but ...". Psychology of Women Quarterly, 28(4), 423-435.

Lillan LOMMEL is a master of science by research student in Sociology at the University of Edinburgh, United Kingdom. Her research interests include gender, identities, feminism, women and leadership, and emotions.

School of Social and Political Sciences
University of Edinburgh
15a George Square
Edinburgh EH8 9LD, UK

Margrit SCHREIER is professor of empirical research methods at Jacobs University Bremen. In her research, she focuses on qualitative research methods and methodology, mixed methods, media psychology, sensory processing sensitivity, and health-related topics.

Jacobs University Bremen
Psychology and Methods
Campus Ring 1
28759 Bremen, Germany

Jakob FRUCHTMANN is faculty member of Bremen International Graduate School of Social Sciences and lecturer of sociology at Jacobs University Bremen. He works in the fields of political, media and economic sociology with a strong regional focus on Eastern Europe.

Jacobs University Bremen
Social Sciences & Humanities
Campus Ring 1
28759 Bremen, Germany

Lommel, Lillan; Schreier, Margrit & Fruchtmann, Jakob (2019). We Strike, Therefore We Are? A Twitter Analysis of Feminist Identity in the Context of #DayWithoutAWoman [67 paragraphs]. Forum Qualitative Sozialforschung / Forum: Qualitative Social Research, 20(2), Art. 16, http://dx.doi.org/10.17169/fqs-19.2.3229.

Forum Qualitative Sozialforschung / Forum: Qualitative Social Research (FQS)

ISSN 1438-5627

Creative Commons Attribution 4.0 International License

Location	Search phrase
Washington D.C.	-RT geocode:38.9,-77.0,20mi #womensmarch OR #womensstrike OR #daywithoutawoman OR #Istrikefor Lang:en since:2017-03-07 until:2017-03-10
New York City	-RT geocode:40.7,-74.0,20mi #womensmarch OR #womensstrike OR #daywithoutawoman OR #Istrikefor Lang:en since:2017-03-07 until:2017-03-10
Los Angeles	-RT geocode:34.1,-118.2,20mi #womensmarch OR #womensstrike OR #daywithoutawoman OR #Istrikefor Lang:en since:2017-03-07 until:2017-03-10