“I Became a Lioness”

Theoretical Framework

Feminist identity falls within the domain of social identity theory (SIT; e.g., Tajfel & Turner, 1979). According to SIT, individuals develop a social identity through an understanding that they belong to a meaningful social group (Tajfel, 1982). Thus, their group membership becomes part of their self-concept. For example, a woman may develop a social identity as a woman through an understanding that other women share similar experiences. By contrast, a politicized identity involves the additional knowledge of power differentials among social groups and leads to engagement against such socially constructed hierarchies (Simon & Klandermans, 2001). Feminist identity, then, could be categorized as a politicized identity rather than a social identity (e.g., Cole & Stewart, 1996; Hunter & Sellers, 1998; Myaskovsky & Wittig, 1997), while gender identity may be a social identity and not a politicized one. Research has shown that some predictors of feminist identity are the same for women of color and White women (e.g., exposure to feminist ideas and to gender discrimination) but that some are different. For example, “traditional” gender ideology seems particularly important to feminist identification for White women (with a strong negative relation), but not for women of color (no relation; see Cole & Zucker, 2007; Reid & Purcell, 2004). The distinction between social and politicized identities has been theorized to be important to engagement in collective action, although both social and politicized identities have been shown to be associated with collective action; thus, both should continue to be investigated (e.g., van Zomeren, Postmes, & Spears, 2008).

Gurin and colleagues (e.g., Gurin, 1985; Gurin, Miller, & Gurin, 1980; Gurin & Townsend, 1986) examined issues related to both social and politicized identities in their work on gender consciousness. They theorized that gender consciousness occurs when women become aware that they belong to a social group with other women, develop feelings of closeness to other women, become cognizant of gender hierarchies of power, and understand inequality as a social rather than an individual process. Thus, their notion of gender consciousness begins with gender identity (a social identity) and broadens to include aspects of a politicized consciousness (awareness of inequality and gendered power relations; see Rhodebeck, 1996).

Since feminist identity is an important predictor of feminist collective action, it is also useful to examine how one develops a feminist identity, something not addressed in research using SIT. Downing and Roush (1985) proposed one influential model of feminist identity development. They postulated that there were five stages to feminist identity development: Women start at the default position of passive acceptance, in which they accept traditional gender roles and do not seek to challenge gender inequality. Individuals who are at this stage view the current gender hierarchy as beneficial. Second, women transition to revelation, in which they discover that there is gender inequality and feel anger and guilt about their subordinated position and how society accepts and supports it. Third, women enter the embeddedness-emanation stage where they seek out the company of like-minded women and begin to strengthen their new identity. Fourth, they move to the synthesis stage in which they start to analyze people and situations on individual rather than stereotypical bases. Downing and Roush (1985) suggest that women claim feminist identities in this stage because they begin to accept oppression-based explanations of sexism and they can understand how individual situations (e.g., being paid less than male colleagues or getting catcalled while walking home) relate to oppression in this stage. Fifth, according to this model, women end in the active commitment stage in which they actively challenge gender inequality.

Downing and Roush’s (1985) stage theory, though clearly useful for some purposes, has been criticized for having many limitations (see, e.g., Erchull et al., 2009; Liss & Erchull, 2010; Zucker, 2004). One major criticism that it shares with many stage theories is that it outlines only one potential pathway to feminist identity and postulates a single, invariant sequence. Research has shown that even when all of the stages occur, they do not always occur in the same order (Fischer et al., 2000; Liss & Erchull, 2010; Zucker, 2004). In addition, the stage model was developed using the experiences of White feminist women from the 1970s who first became feminists during the second wave of feminism (Zucker, 2004). And all of the stages do not always apply—either to younger women who grew up during a time in which feminism already existed, and thus may not have ever passively accepted sexism, or to women of color whose experiences with the intersections of racism and sexism do not fit neatly into the model (Horne, Mathews, Detrie, Burke, & Cook, 2001; Moradi, Subich, & Phillips, 2002; Zucker, 2004). Finally, Downing and Roush’s (1985) model may describe a process of development of feminist identity for some women, but it does not provide an account of the contextual factors that prompt either initial consideration of feminist ideas or their later adoption.

To address some of these issues, Henley and colleagues (Henley, Meng, O’Brien, McCarthy, & Sockloskie, 1998; Henley, Spalding, & Kosta, 2000) emphasized adoption of different feminist ideologies rather than stages. Their view was that there are multiple forms of feminism that are associated with different feminist theories, and individuals may use different processes to develop feminist identity depending on the form of feminist thinking they most accept. For instance, the womanist perspective takes into account intersectional experiences of oppression reported by Black women, which are not accounted for in Downing and Roush’s model (Hudson-Weems, 1989). Henley and colleagues also included three items for each subscale that assess “feminist behaviors,” a few of which focus on activism. The measure as a whole describes different forms of feminist identity, based on particular ideological commitments, which may describe U.S. feminists fairly well. It is compatible with our approach in emphasizing that there are multiple pathways to feminist identity among those who become activists on behalf of women’s issues. On the other hand, we did not focus on the particular ideological perspectives that women of color may have in their feminist identity and activism, and thus we did not use the scale in the current study. Moreover, this measure provides no insight into the contextual factors motivating adoption of particular ideologies.

Predictors of Feminist Identity

There are multiple predictors of feminist identity that we discuss in this section: feminist education, social relationships, experience with gender-based injustice, and exposure to violence against girls and women. These predictors will later be used as a starting point in identifying potential pathways to feminist identity.

Feminist education, or exposure to feminist ideas and arguments, has been one of the most well-documented predictors of feminist identity in the psychological literature (e.g., Bargad & Hyde, 1991; Findlen, 1995; Henderson-King & Stewart, 1994; Horne et al., 2001; Marine & Lewis, 2014). Education sometimes arises in formal settings, such as colleges and universities, and sometimes in informal settings in which individuals read feminist texts on their own or with a group of people. Most of the literature focuses on the former, perhaps because most researchers conduct their studies within higher education institutions and they often have easy access to participants who are taking classes that cover feminist topics.

Many academic institutions both in the United States and in some other countries now offer Women’s Studies courses. Participants who took a broad range of Women’s and Gender Studies courses were more likely to identify as feminists at the end of their courses compared to at the beginning (Bargad & Hyde, 1991; Eisele & Stake, 2008; Henderson-King & Stewart, 1994). While Eisele and Stake (2008) chose to use participants as their own controls using repeated measures, others such as Henderson-King and Stewart (1994) and Yoder, Fischer, Kahn, and Groden (2007) included an additional control group of participants taking unrelated courses. In the study by Yoder and colleagues (2007), those who took the Psychology of Women course were more likely to identify as feminists by the end of the semester than those who took Introductory Psychology. In addition, the increase in feminist identification for those in the Psychology of Women course was statistically significant above and beyond students’ initial feminist identification at the beginning of the course. Women’s Studies courses, then, seem to increase feminist identification even for those who already identify as feminists when the course begins.

While courses in Women’s Studies often serve as catalysts to feminist identity, not all individuals have access to such courses or choose to take them. Therefore, it is important to include informal education in studies of predictors of feminist identity. While there is not much research that investigates the role of informal feminist education, some studies have included measures of the more general factor of “feminist exposure.” Both Myaskovsky and Wittig (1997) and Leaper and Arias (2011) found that exposure to feminist ideas was an important factor in acceptance of a feminist self-label. In addition, a family environment that provides engagement with a variety of opinions and perspectives has been shown to predict feminist self-identification (Wolff & Munley, 2012). While it is unclear from these studies exactly how participants come to be exposed to feminism, one potential source is informal education.

One benefit of formal feminist education is that students can interact with their peers and instructors whom they may view as feminist role models. This interpersonal contact may provide support that mitigates potential negative effects of association with the women’s movement, such as negative stereotypes about feminists (Marine & Lewis, 2014; Reid & Purcell, 2004). Furthermore, having the opportunity to debate and deliberate about feminism, and to build communities with other feminists in Women’s Studies and in community programs, may lead many individuals to self-label as feminists (Marine & Lewis, 2014). Reid and Purcell (2004) also found that those who reported perceiving higher levels of having a common fate with women (i.e., those who reported stronger beliefs that they shared experiences and social pressures with other women) were more likely to use a feminist self-label.

One way to increase perceptions of a common fate with other women is through participation in feminist consciousness-raising groups. While specific consciousness-raising groups are less popular today than they were during the second wave of feminism, women still engage in consciousness raising in many settings today (Joel & Yarimi, 2014; McGirr & Sullivan, 2017; Sowards & Renegar, 2004). Consciousness-raising discussions provide women an opportunity to discuss their gender-linked experiences (of violence, of sexism and harassment, and of motherhood) at work and at home and to discover that other women share similar experiences (Kravetz, 1978; McGirr & Sullivan, 2017). In turn, group consciousness has been shown to increase participation in collective action, and one way in which it may do so is through feminist identity (Duncan, 1999). We need research to increase our understanding of the relation between exposure to social relationships and feminist identity development.

Knowledge from experience or an observation that a system which privileges men and disadvantages women is unjust also may lead to a feminist identity. This includes development of an understanding that women face sexist discrimination. Using a correlational design, Moradi and Subich (2002) established that there was a positive association between perceived sexism and feminist identity development. Of course, the causal direction cannot be determined by this research. Using a regression analysis, Myaskovsky and Wittig (1997) found that participants with more knowledge of discrimination (however acquired) were more likely to identify as feminist than those with less. Renzetti (1987) also found that those who had personal experiences with gender discrimination were more likely to identify themselves as feminists. We need clarification of the causal path between perceptions of discrimination and feminist identity development.

Although we have used the term gender-based injustice, the focus of injustice does not need to be exclusively on gender. Having an understanding of injustice that takes into account the intersecting nature of various social identity categories can also be important. Myaskovsky and Wittig (1997) included a subsample of African American women and discovered that racial identification and beliefs in the compatibility between racial and feminist identity were both unique predictors of feminist identity for women of color. Thus, Greenwood’s (2008) concept of intersectional consciousness, or an awareness of intersecting axes of power, can also be important to feminist identity development.

While the experience of, or knowledge about, violence against women and girls could fall within the larger category of gender-based injustice, exposure to violence may have a unique effect on feminist identity development. Being the recipient of violent acts, such as gender-based harassment and gendered violence (operationalized as any of the following: physical abuse, verbal abuse, and rape), has been shown to predict support for feminism (Buschman & Lenart, 1996). Furthermore, support for feminism was not a function of having overall negative experiences, as more general negative experiences, such as receiving a low grade in a course, did not affect it. It is unclear whether or not an increase in support for feminism would lead some to adopt a feminist self-label. Duncan (1999) provided some evidence that this is likely to be the case. She found that experiences with sexual harassment at work or school predicted feminist consciousness, a precursor to feminist identification (Henderson-King & Stewart, 1997). While this is a start, it would be useful to further understand how and when gendered violence influences feminist identity.

Taken together, the above predictors of feminist identity suggest that both positive and negative experiences can motivate individuals to adopt a feminist self-label. Further, it is not clear whether any of these experiences may be sufficient to produce that identity or whether individuals must experience several or even all of them.

Current Study

In the current study, we examined the roles of four previously identified predictors of feminist identity and tried to identify any new themes that emerged in the narratives of the lives of women activists from five countries. Because much of the psychological research has been based only on U.S. samples, we used a mixed methods approach with a qualitative analysis of themes derived both from the literature and inductively from the interviews themselves, along with a quantitative analysis of how these themes relate to one another. This strategy allowed us to examine feminist identity development in a diverse sample of women from different cultural contexts representing a range of intersections of social identities within those contexts. All data used in this study are available upon request.

Method

The Global Feminisms Project (GFP)

In 2002, members of the University of Michigan Women’s Studies community initiated the GFP, which sought to expand the capacity of faculty and students to address global women’s issues (Lal, McGuire, Stewart, Zaborowska, & Pas, 2010). Four nations were initially included in the project: China, India, Poland, and the United States; international collaborators were identified at each of the sites. Site-specific project teams were responsible for carrying out the interviews, including deciding whom to interview, what the interview format would be, and the logistics of conducting the interviews. Later, Nicaragua was added as a fifth site with the same conditions (see Grabe & Dutt, 2015, for more detailed information about the Nicaraguan site). While the local project teams maintained control over the interview process, a basic structure was consistent across all sites: After giving informed consent for the interview and its posting on a website that included their names and photos, interviewees discussed their personal histories including where and how they grew up, their work backgrounds and participation in women’s studies or organizational activism, and their relations with feminism/women’s movements and global forms of activism and scholarship. All interviews were videotaped and transcribed by native speakers in the original language, translated and transcribed by multilingual speakers into English if necessary, subtitled in English based on the English transcripts, and made available as online videos and transcripts on the project website (https://globalfeminisms.umich.edu; Lal et al., 2010). Participants were provided in advance with information indicating that the goal of the project was to provide on the website both viewable digital versions of the interviews as films and transcripts for use in scholarly research and teaching; they formally consented to these specific uses of their names, images, and interview responses and also provided photos and other information posted on the website on request.

Participants

Fifty-three interviews had been conducted and transcribed and were available on the Global Feminisms website by the time of the present study, and all are still available today; in addition, two new sites have been added for Brazil and Russia, bringing the total number of interviews to 77 (https://globalfeminisms.umich.edu/). Interviews used in the current study had been conducted in five countries: China (n = 10), India (n = 12), Nicaragua (n = 11), Poland (n = 10), and the United States (n = 10; two of the U.S. interviews included two participants; thus, there were 12 individual participants from the United States). Representatives of organizations within each of these countries selected participants and conducted the interviews. Of these participants, one was excluded from analyses because the interviewer did not ask questions about the participant’s personal history. In addition, six participants were excluded because they did not identify as feminists and our goal was to understand different pathways to taking on the identity “feminist.” One final participant was excluded because she claimed to have always been a feminist; thus, it was impossible to determine what made her a feminist if she has always been one.

After eliminating these eight participants, we used a total sample of 45 interviews for analyses. The final total included 9 women from China, 10 from India, 10 from Nicaragua, 10 from Poland, and 6 from the United States. Participants’ ages ranged from 26 to 88 at the time of the interview; age information was not available for five participants. Four participants reported self-identifying as lesbians. Other demographic information was not consistently available since it was not asked for directly.

Procedure

The interviews were coded using Braun and Clarke’s (2013) method of thematic analysis. In this kind of analysis, the coders must be aware of the goals of the thematic analysis in order to identify relevant themes. Coders noted potential themes during an initial reading of all 45 interviews; thus, two of the themes emerged based on that reading (activism and emotion), and four had been previously identified (education, gender-based injustice, violence, and social relationships).

As a first step in coding, members of a gender and personality research group, composed of two faculty and six graduate students who study these issues, discussed the ideas about all of the potential themes; these were eventually grouped into six themes. Definitions and examples of each theme are provided in Table 1. These themes were described in a codebook developed by the first author that included descriptions and examples of the themes and was shared with independent coders. The authors then coded a selection of the interviews at the beginning and middle stages of the coding process to establish inter-rater reliability. Since we were interested in whether or not each of the themes was described by participants, rather than how many times each occurred, coders carried out presence/absence coding for each of the themes. We did not want to suggest that discussing something more times meant that it was more meaningful to the participants. For example, because violence against women might be more difficult to talk about than education, one reference to violence might be more significant to a participant than mentioning education multiple times. In any given passage or account, more than one theme might be present and could be coded. Inter-rater reliability was very good; κ ranged from .94 to .97. Category agreement (agreement about which themes were present or absent also called percent agreement) ranged from .95 to .98. Any disagreements between coders were discussed and resolved before moving on to additional coding.

OPEN IN VIEWER

Table 1.

Descriptions of Qualitative Themes.

Themes	Description	Example
Education	Formal or informal education that includes reading feminist texts or studying feminist topics	“People put the Autobiography of Malcolm X and The Black Woman, by…Cade…Toni Cade Bambara in my hand, and it was like a universe had opened up for me.”
Social relationships	Meaningful relationships with women, feminists, or people involved in social change movements	“Because our mother was very independent and strong, she also wanted us all to be independent and strong. Therefore, in this sense, the idea that a woman should be independent and have self-awareness emerged in me when I was very young.”
Gender-based injustice	Observing, describing, or understanding an experience as due to inequality between different genders	“Because I was married they feared I would be a bad influence in my new job…but there were also some men who already married. They did not have any problems.”
Violence	Observing or experiencing violence against women or understanding this violence as systemic	“But when I was growing up, when my father came home I would get scared. He will quarrel with my mother and beat her.”
Activism	Organizational and non-organizational participation in activism prior to identifying as a feminist	“I worked to establish women’s organizations in colleges, cultural organizations, and the academy of science.”
Emotion	Positive or negative emotional experiences related to women’s or girls’ experiences	“I felt really happy and very lucky, since before the establishment of the new China very few Naxi women had the opportunity to go to college.”

To code the themes associated with identifying with feminism, it was first necessary to estimate when in the course of their lives, each of these participants began to identify with feminism. Some participants were asked about their feminist identity development directly, but most were not, and individuals’ accounts were not linear; they described events that took place at different times throughout the interviews. Therefore, for all participants, we decided to code the entire interview rather than any particular question(s) about feminist identity. Many participants stated very explicitly in the interview that a particular event or experience led them to identify as a feminist (n = 29). For these participants, events that occurred prior to that one event were eligible for coding if the interviewee described that experience as formative of her identification with feminism. One who was quite explicit was Agnieszka Graff (2005) from Poland:

My feminism, as a deep intellectual and emotional experience, came at the moment when I was already in college, and it was after reading Virginia Woolf, among others, for the class taught by this…her name was I think Michelle Barrett, and she was a gender studies professor at Amherst. (p. 8)

Here, Graff states clearly when she began identifying as a feminist. Therefore, events that occurred prior to Michelle Barrett’s class could be coded for themes that the participant tied to her developing identity. Interviews similar to Graff’s were the clearest to code, as it was easy to determine whether or not a particular experience described could have contributed to accepting a feminist identity.

For the remaining participants (n = 16), however, it was unclear exactly when they became feminists. To proceed with coding, we assumed they were feminists by the time they began working in feminist organizations or conducting feminist research. Although this method could potentially lead us to code experiences that actually occurred after the participants had identified with the term feminist, we adopted this method to be consistent across interviews and because it ensured there was a specific point after which they had taken on an explicit identity for us to use as a marker. We were also unlikely to miss themes, since it was probable that participants were feminists by the time they began work in feminist organizations. Some examples include Mangai (India), for whom we coded events that occurred prior to her work with an all-women’s theater group; Yamileth Mejía (Nicaragua), who worked with the Network of Women against Violence; and Rabab Abdulhadi (United States), who founded the Union of Palestinian Women’s Association with a group of friends (Abdulhadi, 2004; Mangai, 2003; Mejía, 2011). While none of these women described a specific moment when they became a feminist, it seemed safe to assume that they were self-identified as feminists by the time when they began working with these feminist organizations. Any events that happened prior to that time were eligible for coding.

Results

We begin with a discussion of the qualitative analysis of themes in the interviews, and we will end the discussion with the results of our quantitative analysis using alternating least squares analysis (ALSCAL). First, we will discuss our qualitative analysis of six themes, four of which were identified through the literature and were discussed above: education, social relationships, gender-based injustice, and gendered violence. Two additional predictors of feminist identity development were discovered through inductive readings of the data: activism and emotion. Our quantitative analysis uncovered relations between the predictors showing which were most likely to co-occur or to form distinct pathways to feminist identification.

Qualitative Analysis

We coded each interview for the six themes identified above. All of the interviews contained at least two themes. Seven included only two themes, 13 contained three, 12 referred to four, 11 included five, and 2 interviewees mentioned all six themes as sources of their feminist identification. We next describe each of these themes with examples from the interviews.

The first theme identified in the coding was education (n = 27; 60%). This included both formal (in school or university) and informal (reading feminist texts on one’s own) education. Sofía Montenegro (2011), from Nicaragua, described how she learned English from reading feminist texts on her own,

And I began to read, with my half English; by the way, that’s the way I learned English…from reading with a dictionary and reading the feminists. So, I learned English with the feminists and with a dictionary on the side…. I was reading, the first book which struck me…was Born Female. (p. 5)

Montenegro read feminist texts to help herself learn English after her family moved to Florida. This was her first exposure to the feminist movement.

An example of how formal education influenced feminist identification came from Lata P.M.’s (2004; India) interview. She said,

I was in B.A. and they told us about Tarun Shanti Sena under the leadership of Jayaprakash Narayan and I heard about class struggle for the first time. And then there were many issues and they told us that Jayaprakashji spoke about a total revolution and this movement is seven-folded; it has seven dimensions like education, women, etc. I felt very happy women, caste, education, and labour were not neglected. It combined all social questions, and also class struggle. I understood it later after reading. After ‘78 I got interested. After listening to Shubhmurti, I decided to join the J. P. (Jayaprakash Narayan) movement. (p. 7)

Lata learned about activist Jayaprakash Narayan in a class she took while earning her bachelor’s degree. She understood his message to be inclusive and decided to get involved in the J.P. movement. Her educational exposure to gender and related social identities helped her to later identify as a feminist.

Social relationships was another common theme identified in the GFP interviews (n = 39; 87%). Social relationships was coded as present if participants described meaningful relationships with other women or feminists as leading them to identification with feminism. Cathy Cohen (2004), a U.S. activist and professor, demonstrated this theme,

So, long before I ever knew what a feminist was, heard the word feminist, I think I kind of started to develop feminist politics from my mother and my father and my sister and my brother. It’s not surprising to me now that my brother’s a second-grade teacher and actively involved in the union in the Toledo public schools. My sister works in a factory and has really struggled with folks in the factory to establish a union there. That unfortunately, is now kind of taken over by white leadership that is very hostile to workers of color. But I think we all were taught repeatedly, you must have a commitment to community, and you must kind of work to kind of transform people’s lives including your own life. So, for me I kind of start that political learning back in Toledo. (p. 5)

Cohen described her relationship with her family as integral to her later feminist identity development. They taught her how to be political in a way that supported herself and her community. Cohen later applied this understanding to a feminist context. Barbara Limanowska (2005), from Poland, provided another example of this theme,

In reality, my first awareness…first conscious contact with feminism was most likely through Ewa Franu, my friend from college…Ewa was from Warsaw and she was often going there and then coming back to college in Poznań, and she’d bring me some written stuff. And she was educating me. And it was a kind of…learning from somebody, learning through an intermediary, you know, because Ewa herself was the kind of person who was getting involved in all of this, and she was kind of passing it on to me, both information and literature. (p. 4)

Here, it is clear that Limanowska’s friend, Ewa, was instrumental in Limanowska’s identification with feminism. Ewa brought her written material and taught her about feminism. She was both a peer and a role model whom Limanowska admired.

Gender-based injustice also emerged as a theme. Nearly all (n = 41; 91%) participants indicated that this was important to their development of feminist identity. This theme included an understanding of inequality between men and women or boys and girls as social rather than individual. Participants often described experiences of unfairness in gendered terms. Ge Youli (2002), from China, provides a particularly salient example of this theme.

I still remember the incident that was the most extreme example of being treated differently because I was a girl. One time we all went out together. My grandmother said, “Why don’t we go out for ice cream!” which, during winter for us kids, was a rare treat. I excitedly went with them. To my surprise, my grandmother bought my brother an ice cream, but bought me a four cent Popsicle. I was really angry, so I asked her why I did not get an ice cream, too. At the time an ice cream cost about twelve cents while a Popsicle cost about three to four cents. She just said that of course he should have a rich, creamy ice cream because he was a boy. I will never be able to forget those kinds of feelings of being treated unfairly. (p. 6)

Ge’s experience exemplifies the gender-based injustice theme. Although she expected an ice cream, like the one her brother received, she was given a Popsicle. This was an egregiously unfair occurrence to her as a child, and it was clear that it was gendered; her grandmother told her outright that her brother got an ice cream because he was a boy. Although she was young at the time, Ge recognized this as an injustice that had a formative influence on her worldview.

Anna Lipowska-Teutsch (2005) provided another example of gender-based injustice. She described her grandfather’s different relationships with her and her brother as they were growing up.

My grandfather’s ways of loving varied. But it was weird, weird…I remember one scene, and I remember it well. We had a balcony and there was an outside gutter around it. The house was in a park. And I am in the park with my mom and grandma, and we see that my grandfather is leading my brother on the gutter, hooking his neck with his cane; he is leading him around the balcony and then back inside. And then my mom and grandma reacted: “Kazik, what did you do? Why did you do something like this?” And my grandfather responded: “Because he’s the person I love the most.” And this is…. But he did love him very much. In his own way…. They played garibaldka, this card game, with my brother. I wanted to learn how to play, but my grandfather told me: “You’re too silly for that”…My brother, who was a year and a half older, and I were attending the underground school together…I had better grades than my brother. And my grandfather…when he saw the report card, he cried. (p. 10)

Lipowska-Teutsch’s grandfather showed favoritism to her brother in various ways. Even when Lipowska-Teutsch outperformed her brother in school, she was unable to impress him. She initially took her frustration out on her brother but later came to see these experiences as based in gender inequality.

Another theme that emerged from the GFP interviews was exposure to gendered violence (n = 18; 40%). This theme indicated that the participant experienced or observed violence against women or girls by men or boys or that the participant began to understand violence against women as a social rather than an individual experience, similar to the gender-based injustice theme above. Barbara Labuda (2003), from Poland, described this theme in her interview, “And this was the case both with my father, from whom I often had to protect younger siblings physically when he was simply beating up on them, or abusing them” (p. 5). Despite being the second oldest sibling, Labuda described taking responsibility for preventing her father’s violence against her younger siblings. Although Labuda did not directly attribute her experiences of violence to her later development of a feminist identity, she explained that over time she began to understand her father’s violence through her study of Eastern philosophy, which set the stage for her feminist identity development, so it formed a crucial link in the causal chain.

Yamileth Mejía (2011), from Nicaragua, described similar familial violence growing up,

Within my own family there were many violent men, many, many violent men: my uncles, cousins, we knew that sexual abuses were being committed. Like I said, we were an average family, like any other in this country where many sexual abuses take place, as well as violence against women, battering and where the norm you must follow—regardless of what my mother and grandmother had told me—is “you have to suck it up because he is your husband” and also that “you don’t air your dirty laundry in public.” But I kept thinking we can’t just keep on being quiet about it, we can’t be silenced, and one day I decided to talk about the abuse I lived and the abuse my nieces lived. (p. 5)

Mejía described not only the violence she and her family experienced but also the evolution of her understanding of that violence. While she knew that abuses were taking place in her family, she had to resist the dominant narratives about how women should behave in the face of violence. She recognized that her family was no different from any other family, suggesting that she understood that her experiences were shared by other women. She eventually began taking feminist actions by talking about the abuses she and her family members faced.

Turning to themes that arose in the interviews but not from the literature, several participants indicated that they participated in feminist activism prior to identifying as a feminist (n = 22; 49%). Activism was coded for participants who described organizational or individual experiences of feminist activism as moving them toward feminist identification. Jarjum Ete (2005), from India, for example, discussed how she threw rocks at people who were trying to take a girl away to be forced into a child marriage.

And then I heard them say like, you know, she has to be taken away and she was crying, shouting for the help and saying, “No, I don’t want to go” and then I saw them dragging her and I remember I had picked up some pebbles and I was throwing at those people who were dragging her away, you know, that was a real kind of, you know, touching scene I still can visualize. (p. 12)

Ete was in kindergarten at the time of this event, but it clearly influenced her later in life when she began to identify with the feminist movement in India.

Mónica Baltodano (2011), Nicaragua, was another young activist who engaged in activism before identifying as a feminist. She credited her early desire to challenge inequality to the school she attended, which was exclusionary, particularly of students from lower-class backgrounds. Her understanding of inequality led her to participate in a march to free a guerilla woman who was captured.

And when I was fourteen, fifteen years old, I became involved in my first political activities, my first big action was a march to demand Doris Tijerino’s freedom, a guerilla woman who had been captured. She had been raped, humiliated, and she appeared with bruises from her capturers, with marks from the aggression that she suffered. So, we mobilized with the school to protest and demand her freedom. (p. 4)

Both Ete and Baltodano had a general understanding that their actions were intended to correct unfairness in the experiences of the girls and women they knew. Still, they were involved in activism before they fully engaged with and began to identify as feminists. These early experiences, though, helped them to contextualize and accept feminism later in life.

The final theme discussed in the GFP interviews was emotion (n = 21; 47%). Participants described both positive and negative emotions related to women’s or girls’ experiences. Loretta Ross, from the U.S., described her experience growing up with five brothers: “My family was very patriarchal too, you know, with the five boys. I remember being very resentful of my mother, because she used to make me get up and cook my brothers breakfast for school” (Ross, 2006, p. 4). Ross’s use of the descriptor “resentful” demonstrates the negative emotions attached to having to adhere to patriarchal gender roles. Specifically, she pointed out how unfair it was that she had to cook breakfast for her brothers, which they were able to eat without doing any work.

Not all examples of the emotion theme included negative emotions. He Zhonghua (2005), from China, explained that she felt happy to have been given opportunities that similar others did not receive: “Thus, I felt really happy and very lucky, since before the establishment of the new China very few Naxi [an ethnic minority in China] women had the opportunity to go to college…. So I was very lucky to be offered this opportunity to attend university” (p. 7). He’s personal happiness was related to her knowledge that she was able to take advantage of a unique opportunity that had been denied to many Naxi women.

Taken together, these themes represent experiences that contributed to the development of feminist identity. It is useful to have shown that these themes characterized the process of feminist identity development for the women’s movement activists who comprised our sample; they can therefore be seen to have preceded and motivated not only their identity as feminists, but also their commitment to lifelong activism. Our analysis verified the personal relevance in these activists’ lives of the four themes studied in the literature (education, social relationships, gender-based injustice, and gendered violence) and also identified two that have not been studied in the past (activism and emotion). It is particularly interesting that activism—usually studied as an outcome of feminist identity—was also seen by these women’s movement activists as having preceded and motivated their feminist identity. This bidirectional relation between identity and activism is an important finding of this study.

In some cases, a single event might encapsulate more than one experience. For example, exposure to a role model could include that model providing education or engagement in activism might lead an individual to exposure to violence or gender-based injustice. In order to understand more about the role of these themes in developing feminist identity, we ran analyses to determine which experiences (captured in codes of themes) co-occurred most and least frequently. We viewed this analysis as allowing us to move beyond identifying separate experiences to identifying some common pathways to feminist identity that involved multiple experiences rather than discrete occurrences.

Analysis of Patterns

Preliminary analyses

We tested the relations between participants’ demographic characteristics and their endorsements of the six themes. In particular, we looked at potential differences based on nation of origin and participants’ ages at the time of the interview using χ² and correlations, respectively (see Tables 2 and 3). Other demographic information was either unavailable or there were not enough data per cell to accurately test differences (e.g., there were only four participants who said they were lesbians in the data set). There were no significant differences in the occurrence of any themes as a function of either nation or age (ps > .05).

OPEN IN VIEWER

Table 2.

Qualitative Themes by Nation Cross-Tabulation.

Nation	Theme
	Education	Social Relationships	Gender-Based Injustice	Violence	Activism	Emotion
China	8	8	8	1	5	5
	88.9%	88.9%	88.9%	11.1%	55.6%	55.6%
India	4	9	9	5	6	4
	40.0%	90.0%	90.0%	50.0%	60.0%	40.0%
Nicaragua	5	9	10	6	6	2
	50.0%	90.0%	100.0%	60.0%	60.0%	20.0%
Poland	8	9	9	4	2	5
	80.0%	90.0%	90.0%	40.0%	20.0%	50.0%
United States	2	4	5	2	3	5
	33.3%	66.7%	83.3%	33.3%	50.0%	83.3%

Note. Number and percentage of interviews in which the qualitative theme is present for each nation. No significant differences emerged.

OPEN IN VIEWER

Table 3.

Correlations Between Qualitative Themes and Age.

Theme	Age
Education	−.23
Social relationships	.09
Gender-based injustice	−.21
Violence	.10
Activism	−.13
Emotion	−.10

Note. All ps > .05.

Quantitative analyses

Our main interest in these analyses was to determine whether co-occurrences of themes suggest that there are some common pathways to feminist identity for these feminist activists. To investigate this, we conducted analyses that mapped the co-occurrences of themes. We used multi-dimensional scaling to identify greater or lesser distance between variables as implemented in the SPSS ALSCAL procedure (Takane, Young, & de Leeuw, 1977; Young, Takane, & Lewyckyj, 1978). Like other forms of multi-dimensional scaling, ALSCAL optimizes fit to the observed proximity (distance) data (Kruskal, 1964; Takane et al., 1977) and can be used with binary codes like ours. We coded a theme as 1 if it was present or as 0 if it was absent from the interview. We computed distance matrices between each of the codes using SPSS tools for binary present/absent data. For a given interview, every combination of two themes was compared, and they were considered distant if one was present and the other was absent, or close if both were either present or absent. We fit one- and two-dimensional solutions, given the small number of themes (six) and our hypothesis that there would be separate pathways to feminism.

There are various criteria for determining the best number of dimensions to accept as meaningful in ALSCAL. Generally, the procedure is to create a scree plot of stress values from various solutions with different numbers of dimensions and look for the “elbow” (Cattell, 1966; Wish & Carroll, 1982). Lower stress values indicate better fit, and the elbow indicates where adding dimensions no longer significantly reduces stress. However, since we used qualitative data based on only six variables, we were unable to plot solutions for more than two dimensions because the number of parameters estimated would have exceeded the number of data values for solutions with more than two dimensions. Therefore, we could only create a scree plot with the stress values from the one-dimensional and two-dimensional solutions; there cannot be an elbow on a line connecting two points. Therefore, we were unable to use this criterion in determining the best number of dimensions to report. However, we were able to compare the stress values for one-dimensional and two-dimensional solutions. According to Kruskal’s (1964) criteria, the one-dimensional solution fell into the poor range of fit, with a stress value of .20 while the two-dimensional solution had substantially better fit, falling into the excellent range with a stress value of .01. This suggests that the two-dimensional solution better describes the data, although both the one-dimensional and two-dimensional solutions were interpretable. Wish and Carroll (1982) suggest two additional criteria for determining how many dimensions are best to use: interpretability and replicability. The two-dimensional solution had already been shown to fit the data better, and interpretability and replicability added additional support for the two-dimensional solution. We did not have an additional data set with which to replicate the solution. It would be ideal to do so in future studies. While we chose to use the two-dimensional solution, we did not interpret the meaning of the dimensions per se. Rather, we were more concerned with addressing proximity in the solution (i.e., which themes most often co-occurred), as this was more in line with our research goals.

A visual representation of the two-dimensional ALSCAL solution can be seen in Figure 1. Within the solution defined by the two dimensions, four quadrants including the six themes emerged, suggesting that each quadrant describes a distinct pathway to feminist identity. Since most respondents mentioned two or more of the themes (and some mentioned nearly all of them), these quadrants identify instances where co-occurrence of themes was systematic and not random.

OPEN IN VIEWER

Figure 1.

Alternating least squares analysis (ALSCAL) two-dimensional solution. ALSCAL similarity matrix of the six themes; closer distance between points represents more similarity or greater co-occurrence between themes compared to greater distances. The values on the axes represent the bipolar magnitude of the deviation from the midpoint (0) based on a multi-dimensional unfolding analysis that minimizes stress.

Quadrant 1 consisted only of the education theme described above. While education often occurs alongside other people, internal representation and integration may be necessary in order to contribute to feminist identity. One pathway to feminist identity may involve development of a cognitive understanding of feminist goals and concepts within an individual, which must be integrated internally. It appears that education is a pathway to feminist identification that can exist in combination with any of the other themes, as shown by the high co-occurrence of themes.

Quadrant 2 included two themes: social relationships and gender-based injustice. The theme for gender-based injustice (which was the most frequently observed theme) was coded when individuals mentioned personal experiences recognizing inequality in one’s own or others’ interactions as a social rather than individual phenomenon; the social relationships theme references relationships with others who have feminist beliefs. This approach depends on social experiences as encountered in two ways: through links to feminism or through gender-based injustice; it is often experienced in combination, and these two kinds of social experiences (which may or may not be combined with other experiences) in our data, seem to have commonly co-occurred to create a social pathway to feminist identification.

Violence emerged on its own as a pathway to feminist identity in Quadrant 3. Many of the women in the GFP directly experienced violence themselves or witnessed violence against other women they loved. These violent experiences were highly salient and often recurred, perhaps accounting for why such experiences were enough on their own, or in combination with any other theme, to motivate a woman to identify as a feminist.

Finally, the themes activism and emotion comprised Quadrant 4. Both of these themes emerged from the data inductively rather than from theory or prior research. However, research has established a relation between activism and emotion (e.g., Tausch et al., 2011; van Zomeren, Spears, Fischer, & Leach, 2004; Yzerbyt, Dumont, Wigboldus, & Gordijn, 2003). Accounts of co-occurrence of the themes of activism and emotion suggest that some individuals come to feminism through emotionally intense active engagement with social movements, alone or in combination with any other themes.

Discussion

Although the interviews for the GFP took place in five countries with unique histories, women activists from each of these countries described their journey to feminist identity in similar ways. Six themes emerged in the data as important to feminist identity development. Consistent with previous research, formal and informal education, social relationships with other women and feminists, gender-based injustice (including personal experiences of discrimination), and experiences or knowledge of gender violence all contributed to participants’ acceptance of a feminist label regardless of their country of origin (e.g., Buschman & Lenart, 1996; Henderson-King & Stewart, 1994; Myaskovsky & Wittig, 1997). In addition, two themes emerged through inductive analysis: participation in activism that benefited women and emotion related to the experiences of girls or women. The identification of these additional themes provides support for a more active and engaged path to feminist identification and points to the value of inductive analysis for identifying pathways not yet recognized. Downing and Roush (1985) suggested that there is a single pathway to feminist identity development in which individuals proceed through the same stages in a fixed order. On the other hand, Downing and Roush do consider multiple factors to feminist identity within their stages, which is consistent with the current data. We posit, however, that different factors are used by different women in a more flexible way.

In addition to the above findings of relevant themes related to feminist identity development, we were interested in exploring which of these themes came together to produce more particular pathways to feminist identity. Using an exploratory data analysis procedure, we discovered four pathways to feminist identity. The first consisted of education, the second included both social relationships and gender-based injustice, the third comprised exposure to gender-based violence, and the final quadrant included the new themes of activism and emotion. The identification of these four pathways to feminist identity further reinforces the notion that feminist identity can be achieved through a range of experiences, some of which co-occur and some of which may not.

It is important to consider why these particular pathways exist rather than others. The first pathway––education or exposure to feminist ideas––may sometimes be enough to influence the acceptance of feminist identity, as much of this education occurs during a time in individuals’ lives when they are forming their political identities. Often the first exposure to feminist texts occurs at colleges or universities, when students are in the early adulthood life stage in which identity formation tends to occur (Stewart & Healey, 1989). Therefore, other factors may not always be necessary for feminist identity development. It is equally possible that education is a facilitating factor for some women which, when combined with any other catalytic experience, produces feminist identification. The second pathway includes social relationships and gender-based injustice. These themes together call to mind feminist consciousness-raising groups in which women share their experiences of discrimination with others and often develop an understanding that their personal experiences are also “political” or reflections of unjust gender relations. Experiencing injustice on one’s own may rarely be enough to produce feminist identity, especially if individuals feel that their experience is particular to them alone rather than something that happens to many women. But social relationships can help women move from feeling helpless about gender-based injustice to feeling that feminist identification will enable activism and social change. The co-occurrence of these two themes defines a path that may also include any of the other themes.

Two other pathways arose from the data rather than the literature. Gender-based violence provided a pathway to feminist identity for some activists; that pathway may occur on its own or in combination with another theme. It is interesting that violence did not co-occur with other forms of gender-based injustice. However, the importance of violence is likely related to its intensity. Intimate partner violence, rape and sexual assault, and all other forms of violence are such significant experiences that they can produce an abrupt adoption of a feminist identity. Finally, activism and emotion occurred together, suggesting that either alone may not be sufficient to produce feminist identity and that the two together may occur in the presence of any other theme. Since an individual is unlikely to fully understand what they are trying to accomplish by participating in activism before developing a feminist identity, perhaps an intense emotional response helps define the experience as personally significant. Without this attachment, a person may attend one activist event and then decide the cause is not important or not for them. However, if participation produces powerful emotional responses, the intensity of those responses may motivate further examination of the cause and development of a feminist identity. This pathway should be of particular interest to scholars associated with SIT, who have generally examined the connection between different social identities and how social identities motivate action on behalf of social groups (see, e.g., Duncan, 1999). Many do not consider that participating in activism about (for example) a particular issue may actually lead to the development or strengthening of identification with a social group. In order to assess how prevalent this pathway is, it would be valuable to study a range of types of participants in social activism (as Duncan did in the 1999 study) and assess the degree to which they identify with various social groups before the event and after (Duncan only assessed group consciousness prior to the event).

There is no reason to assume that the organization of these themes occurs only for feminism and not for other social movements. For instance, taking a Disability Studies course can produce an anti-ableist identity that leads to participation in activism (Connor, Gabel, Gallagher, & Morton, 2008). Since education related to many social movements occurs for the first time during the same life stage, it is likely that it can affect a wide variety of politicized identities. Similarly, an individual can have an emotional connection to any movement that they participate in. Going along with someone to an LGBT pride parade and having a positive experience could potentially foster the formation of an anti-heterosexist identity later, and experiencing and sharing familiarity with discrimination based on social class could lead people to form anti-classist identities. And knowledge of and experience of violence against Black men and women have motivated many to take part in the anti-racist #BlackLivesMatter movement, seeking to end police brutality against Black individuals (Garza, 2014). It is likely that many of the participants in this movement have developed anti-racist identities since this violence has been made more visible through the media. It would be interesting to further examine the pathways identified in this study within the contexts of different social movements.

Strengths and Limitations

Analyzing women’s life narratives helps produce a new understanding of experiences that lead to feminist identification. Narrative accounts enable participants to express fully how the events in their lives contributed to development of their past and current selves. Although we cannot make causal claims from our data, we were able to identify potential pathways to feminist identity that can be empirically tested in future studies using a range of methods. For example, researchers could collect life narratives from activists in other social movements and code them for the same themes as the current study and search for other themes. Researchers also could design an experiment around any of the four quadrants identified in our study and test causal directions of the pathways to feminist identity we identified. Although causality cannot be assessed with our data, participants did describe their experiences with the six themes as occurring before they began identifying as a feminist, so it is logical that causality may occur in the expected direction. Some themes, most notably activism––which the literature has shown to be a result of feminist identity rather than a predictor––likely work bidirectionally, even if it is more often an outcome than a predictor (e.g., Liss et al., 2004; Yoder et al., 2011; Zucker, 2004).

Another strength of the current study is the use of a novel method of quantitative analysis of the qualitative data. Although ALSCAL was developed many years ago, compared to other techniques, such as analysis of variance and regression, ALSCAL has not been used frequently in psychology. However, this method has several benefits for analyzing the current data and other thematic analyses. First, ALSCAL allowed us to identify which themes were most likely to co-occur in the interviews. Although all of the themes occurred frequently across the interviews, particular themes such as social relationships and gender-based injustice tended to be present in the same interviews. Using ALSCAL made finding patterns within a large and complex data set much easier. In addition, we were able to test an alternative model using ALSCAL. We tested a one-dimensional model of pathways to feminist identity, but the fit of this model was poorer than the two-dimensional model discussed in this article. However, we were unable to test more than one alternative model in the current study due to the small number of variables. A future ALSCAL study using more traditional quantitative data could address this limitation. In addition, if researchers have enough data to do so, they can use the more powerful analysis individual differences scaling (INDSCAL) and individual differences version of ALSCAL that allows the researcher to test how much each individual participant’s data differ from the overall model. Due to the small number of themes, we could not conduct INDSCAL in this study, but future studies may do so. Finally, although we did not use ALSCAL in this way, future qualitative studies can use the procedure at an earlier stage to assist with identifying themes from a set of codes. Codes that are most similar to each other might co-occur; codes that cluster together could be combined to make up larger themes.

Finally, because of the relatively small number of interviews associated with particular demographics, we were only able to test for differences between participants based on nation and age. While neither of these analyses resulted in significant differences, it is not our intention to suggest that there is a universal experience that all women go through to achieve a feminist identity. Each of the women interviewed in the GFP experienced the themes and pathways identified above within the contexts of their home countries. For instance, many of the women who were interviewed from Nicaragua participated in the Sandinista Revolution to overthrow the Somoza dictatorship in the 1960s and 1970s. Their experiences were very different from those who grew up in the United States, when the Civil Rights Movement and the second wave of feminism were underway. Thus, although all of the themes were present in all of the contexts, the implications and specific experiences of these themes were likely very different across contexts.

Practice Implications

There are several implications of the current study for education, research, and activism, which may be different based on context. Here we focus on how the data can inform practices in the United States since that is the context we, the authors, know. This is not to suggest that U.S. interventions and experiences are more important or that implications should not be discussed for other countries, but rather that we leave the drawing of conclusions about other contexts to those grounded in those locations.

One potential implication of the results of the current study is that Women’s and Gender Studies educators should approach feminism through multiple angles to appeal to a wide variety of students. While Women’s and Gender Studies classes include students who are neutral about, or even hostile to, feminism, we know that these courses tend to produce increased feminist identification overall. Moreover, Rios and Stewart (2013) argued that not only can privilege be taught through the Global Feminisms Archive, but features of intersectional invisibility can be as well. Assignments that allow students to go beyond passively reading texts to active engagement in feminism would be beneficial to some students’ development of an understanding of, and identification with, feminism. For example, while studying feminist theory, students could be asked to engage in a number of feminist actions such as participating in campus activism or interrupting or challenging someone who makes a sexist joke. This experiential learning could then be connected back to theory through a written assignment. In her recent analysis of feminist pedagogy, Fahs (2015) has argued for prioritizing activism in the Women’s Studies classroom. This assignment could also be designed as a group project to create a social component. Incorporating several of the pathways into the classroom can be beneficial for feminist identity development because it is likely that different students will use different pathways to form their identity. Adopting more pathways in the course pedagogy might maximize the number of students who would leave the classroom identifying as feminists. Similarly, including and studying pedagogies that reflect multiple pathways can be theoretically useful to gain a more thorough understanding of feminist identity development.

This is the first study, to our knowledge, to categorize some of the predictor variables typically found in the feminist identity and self-labeling literatures as features of feminist identity development. We also identified a pathway to feminism that has not been discussed in the literature, that of activism and emotion. In most cases, perhaps, identity precedes activism, but clearly in some cases participating in activism about a particular issue can solidify budding feminist thought and create a personal connection with the movement. It should be noted, though, that the current sample consisted of many of the most politically active women in their respective nations. It would be interesting to see whether the relation between activism and feminist identity holds with another less politically engaged sample of women.

Finally, activists aiming to mobilize greater participation in women’s movements by encouraging feminist identity development should recognize that strategic appeals to all four of these pathways are likely to be more broadly effective than any single approach. Thus, participation in a particular collective action will be more effective in producing a feminist identity if it is emotionally compelling to the individual. In addition, if that sort of activism occurs in combination with past education or experiences or observations of gender-based injustice or violence, it may be more likely to produce a feminist identification than any one of these experiences alone.

Conclusions

Women learn about and come to identify as feminists in a wide variety of ways that are far from mutually exclusive. We know that many women (and men) endorse values associated with feminism but refuse the label; we also know that these individuals are less likely to engage in feminist activism (Zucker, 2004). Among those who do identify with feminism, some learn about feminism intrapersonally, and some with a large group of their peers. Some individuals develop their feminist identity through emotional and active engagement, and some approach it through cognitive learning. Educators and activists seeking to expose more people to feminism in a positive way can make use of any or all of these various pathways. As more individuals accept a feminist self-label, the rate of feminist activism will likely increase effective challenges to the ever-resilient and pervasive status quo.

Supplemental Material

Supplemental Material, PWQ_42_3_Fred_Teaching_Slides - “I Became a Lioness”: Pathways to Feminist Identity Among Women’s Movement Activists

Supplemental Material, PWQ_42_3_Fred_Teaching_Slides for “I Became a Lioness”: Pathways to Feminist Identity Among Women’s Movement Activists by Jennifer K. Frederick, and Abigail J. Stewart in Psychology of Women Quarterly