Abstract
This study of a South Asian community in the midwestern USA examines what bystanders do when they witness an incident of intimate partner violence (IPV). Because of ethical and safety constraints, in lieu of observation in a natural setting, data were collected at a Peerformance, a peer-led IPV prevention program, using the forum theatre method introduced in Augusto Boal’s Theatre of the Oppressed. Event attendees were invited to respond to an IPV incident enacted by peer educators in which a controlling husband’s behavior escalates to the point of suggesting physical violence. Using a grounded theory approach, we analyzed the videotaped bystander actions while applying pertinent aspects of visual analysis. Event attendees responded in variety of ways, exploring and/or de-escalating the situation, providing information, and encouraging the couple to resolve their conflict and/or seek outside help. They expressed empathy, support, and (dis)agreement with the husband and the wife. Their actions encompassed a number of dimensions: introductory, investigative, supportive/empathic, informational, instructional, instrumental, and confrontational. Participants’ vastly varied responses to IPV within the same scenario suggest difficulty in developing a generic IPV bystander program. The findings also suggest possible directions for developing bystander programs to address IPV. In lieu of prescribing and/or proscribing specific actions in a given risk situation, bystander training can help participants build their repertoire of actions while collectively examining the strengths and limitations of various actions in their sociocultural context. Such bystander training is complex and difficult to manualize; it will involve reflexive and critical discussions and learning and trying out various possible alternatives. A Peerformance is a site of action, a realistic and familiar milieu where community members can imagine and try out interventions that may be different from the norm of indifference and non-intervention. It can open up new possibilities for addressing IPV in diverse communities.
Keywords
Bystander Approach to IPV Prevention
Intimate partner violence (IPV) is the most common form of violence against women, affecting one-third of women who have ever been in a relationship worldwide (World Health Organization, 2013). In addition to measures focusing on victims/survivors and perpetrators, bystander approaches have increasingly been promoted to prevent IPV. Over the years, an increasing number of bystander programs have been developed and evaluated, mostly on college campuses and high schools in the USA; among the small number of community-based programs and studies, few have been conducted in communities of color (Yoshihama & Tolman, 2015). This study in a growing immigrant community contributes to the field by directly examining what individuals in a community do when they witness an IPV incident enacted in a Peerformance, a community-based theatre program conducted by trained peer educators.
The Bystander Approach to Addressing IPV
Contrary to a popular (mis)belief that IPV happens behind closed doors, an intimate partner often perpetrates various abusive acts towards their partner in the presence of others. Studies estimate that about one-third of IPV incidents occur in the presence of others (Planty, 2002; Taylor et al., 2019). Family members, friends, and neighbors who witness, overhear, or are otherwise aware of the occurrence of IPV can make a difference in the lives of people affected by IPV by intervening in some way as bystanders. There are other reasons for promoting bystander approaches to address IPV. Contact with outside agencies, such as the police, raises fear and anxiety in many communities, especially immigrant and refugee communities, due to the recent surge in the anti-immigrant and anti-refugee policies and sentiments (Kalhan, 2014; Kerwin, 2018). The possibility of seeking help from shelters and counseling programs might be unfamiliar to immigrants and refugees or antithetical to their cultural beliefs, and many assistance programs lack linguistic capability to serve people with limited English proficiency (Dabby & Yoshihama, 2020; Huang, 2020; Yoshihama, 2009). These socio-legal, sociopolitical, sociocultural, and linguistic factors make the bystander approach to IPV in immigrant communities even more critical and suitable.
Many bystander training programs promote skill development and engage participants in identifying and challenging their attitudes and societal norms that support violence against women. The Mentors in Violence Prevention Program pioneered this approach with college athletes during the early 1990s and has since been adopted by colleges, high schools, the military, and other institutions (Katz, 1995, 2018). Grounded in social justice and transformative education, bystander programs not only equip individuals to take action but also aim to transform societal norms (Katz, 2018). Over the years, the number and scope of bystander programs implemented to address violence against women have noticeably increased, particularly on college campuses and in high schools (Banyard & Moynihan, 2011; Coker et al., 2017; McMahon et al., 2014; Moynihan et al., 2011; Potter et al., 2009). For example, Bringing in the Bystander “teaches bystanders safe, personal ways to intervene in situations of sexual and IPV” (Moynihan et al., 2011, p. 706) by providing basic information about gender-based violence, facilitated discussion, role-plays, and other interactive activities. Green Dot teaches the three D’s of action: direct, delegate, distract (Coker et al., 2017). Coaching Boys Into Men (Miller et al., 2013) focuses on adolescents on sports teams. These programs are effective in enhancing bystander-related knowledge, attitudes, beliefs, and behaviors (see Jouriles et al., 2018 for review).
Research Gaps: Bystander Programs in Communities of Color
This study addresses a number of major gaps in program development and research on the bystander approach to IPV. The first gap is in community-based universal prevention programs. Despite the call for a broader scope, research, thus far, has focused on the development and evaluation of programs in school settings. The second gap concerns program focus. Many gender-based violence bystander programs focus on sexual violence and dating violence. Community-based programs addressing violence in long-term, marital, cohabiting, and various other relationship contexts common in the general population are scarce.
Another gap is the paucity of prevention programs for populations and communities of color; this gap is even greater for the growing immigrant communities in the USA with few exceptions (Belknap et al., 2013; Yoshihama & Tolman, 2015; Yoshihama et al., 2012). While IPV knows no national or cultural borders, the ways people respond to incidents of IPV are embedded in sociocultural contexts (Harris et al., 2005; Yoshihama, 2001, 2002, 2009). Perpetrators, survivors/victims, and bystanders experience, give meaning to, act and react drawing on their lived experiences; how specific actions of a bystander are perceived and received is intrinsically tied to sociocultural norms and expectations (Brown et al., 2014; Hammock et al., 2020). A specific bystander action considered to be effective in one population group may not necessarily be so in a different group, creating challenges in developing bystander programs to suit all communities.
Few studies of bystander programs addressing campus sexual assault have examined racial variations; they have found that Asian students report fewer intervention opportunities, lower intent to intervene (Hoxmeier et al., 2020), and having intervened less frequently in some situations than White students (Hoxmeier et al., 2018). While findings from studies of campus sexual assault cannot be extrapolated to how community members might react to IPV, these findings speak loudly to the need to pay attention to sociocultural variation in bystander actions. Community-based studies of IPV, albeit not specifically about the bystander approach, point to a high degree of tolerance for IPV among Asian population groups (Klein et al., 1997). Immigrants and refugees account for the largest share of the US population increase in the past half-century and will continue to do so in the coming 50 years; Asians are the fastest-growing immigrant group in the USA (Cohn, 2015). Their greater tolerance for IPV makes developing effective IPV prevention for this group urgent.
Back to Basics: Understanding What Bystanders Do
As discussed above, research on the effectiveness of bystander programs in high schools and college campuses has flourished over the last two decades (see McMahon, Palmer, et al., 2015). For the most part, these studies have used pre/post-test designs to examine changes in knowledge, attitudes, and beliefs about gender-based violence and gender, as well as whether program participation is associated with greater intention to intervene and/or actual intervention in specific risk situations (Banyard et al., 2014; Jouriles et al., 2018; McMahon et al., 2014). However, there remains a lack of qualitative research on “if, when, and how students intervene” (Hoxmeier et al., 2017, p. 3, emphasis added), which is crucial for designing tailored bystander training program.
There is also a need to understand the experience of bystander intervention among people in communities beyond the campus. For example, one recent community-based study found that the presence of bystanders during an IPV incident was associated with higher likelihood of injury to victims, greater disruption in their routines, and poorer mental health (Taylor et al., 2019); however, there was no investigation into what exactly bystanders did in these instances. In order to develop valid measures at the community level, it is critical to first understand how bystanders intervene. Going back to basics, the current study does just that. Focusing on an Asian community, this study examines what people do when they witness an IPV incident.
Research Context: South Asian Communities in the USA
This study is part of an ongoing, larger participatory action research effort to develop socioculturally effective approaches to IPV in the fastest-growing immigrant population group, Asians, in an urban midwestern metro area of the USA (see Yoshihama & Tolman, 2015). Among diverse Asian groups, this study focuses on South Asian populations, those with ancestry in Bangladesh, Bhutan, India, Nepal, Pakistan, Sri Lanka, and the Maldives, which grew by 47.6% between 2011 and 2019 (United States Census Bureau, 2020).
Studies have found a relatively high prevalence of IPV in South Asian communities in the USA—from 19.5% to 40.8%—mainly using non-probability samples recruited via community outreach or community-based organizations (CBOs) and assistance programs (see Yoshihama et al., 2020 for review). Our previous study found that over half of respondents knew of someone affected by IPV (Yoshihama & Tolman, 2015). Studies have also elucidated many barriers to help-seeking, especially from outside agencies, among South Asian survivors of IPV: lack of socioculturally and linguistically relevant programs, lack of familiarity with the socio-legal system, stigma, isolation, and culturally rooted values and norms (Abraham, 2005; Rai et al., in press). The high prevalence of IPV, frequent knowledge of its occurrence in social networks, and a multitude of barriers to help-seeking make the bystander approach valuable for South Asian communities. Yet socioculturally relevant IPV bystander programs in this community remain limited. Our study is part of an ongoing effort to fill this gap.
This study seeks to examine what South Asian community members do when they witness an IPV incident. In light of the ethical and logistic challenges inherent in observing an actual IPV incident, we used forum theatre (Boal, 1979), inviting community members to respond to an IPV scene enacted by peer educators at a Peerformance, discussed below.
Methods
Research Design
Forum theatre is a method set forth by Augusto Boal in Theatre of the Oppressed (Boal, 1979). Boal, inspired by the work of Brazilian educator–philosopher Paulo Freire (1970), developed Theatre of the Oppressed to raise awareness about societal oppression affecting communities and involve community members in exploring and developing possible solutions through performance. In forum theatre, the ending of a skit portraying oppression is intentionally left unsettled or inadequately resolved; audience members are then invited to replace one of the characters to try out a new action to alter the course of the skit. Actors remain in character, improvise, and respond to each intervention. Typically, multiple individuals take a turn at trying out actions, frequently building on previous actions. Audience members have a chance to witness different possibilities. Typically, facilitated, interactive discussions follow, in which audience members explore the strengths and limitations of various interventions. Forum theatre is flexible, able to be performed onstage, in classrooms, or in various community venues. Forum theatre and other methods of interactive theatre have been used in bystander programs addressing gender-based violence in various settings (Ahrens et al., 2011; McMahon, Winter, et al., 2015; Mitchell & Freitag, 2011).
Peerformance, an Application of Forum Theatre to Bystander Programs
Our community-based IPV prevention program organizes peer-led theatric presentations using forum theatre. We call them Peerformances to emphasize the central role of community members as peer educators as well as creators and actors in theatrical presentations. They are part of an ongoing effort to develop socioculturally relevant IPV prevention approaches (see Yoshihama & Tolman, 2015 for program development). Over 30 community members have served as peer educators, or peerformers, over the duration of the project. Each has undergone a series of trainings covering a wide range of topics, including gender justice, gender-based violence, prevention, community engagement, and theatrical approaches. Peerformers are male and female, youth and adults, and diverse in ethnicity, religion/faith, immigration status, and generational position. Over the years, peerformers have worked with research team members to create and perform skits, encouraged community members in the audience to try out interventions, and facilitated post-performance discussions in various community venues.
One year, a local South Asian CBO with which we have been collaborating asked us to organize an IPV interactive prevention activity for their annual fund-raising walkathon called Walk to End Domestic Violence. Widely publicized across local South Asian communities, the walkathon typically includes a consciousness-raising component that features a speaker delivering an inspirational speech and/or providing information about IPV. To make this component more interactive, we and the peerformers brainstormed possible adaptations of forum theatre to this outdoor event, a new experience for all of us. We created a much shorter script suitable for the new venue. Although IPV is not limited to heterosexual relationships, this play depicted the controlling behavior of a husband toward his wife, a choice based on peerformers’ input, our ongoing work with the community, and the organizing CBO’s suggestions. The play unfolds as follows. The wife asks her husband to stay home to watch their children so that she can attend a one-hour meeting at work. The husband insists that it is her role to take care of the children and tells her to stay home. The wife objects and points out that she rarely asks him to care for their children, emphasizing that her work meeting will be short but is very important. There is increasing intensity in their voices. Finally, the husband approaches the wife and makes a gesture indicating that he might resort to physical force.
Participants and Data Collection Procedures
The study was approved by the institutional review board of the lead investigator’s university. Held in a local public park on a weekend, this outdoor event attracted a diverse group of adults, many of them attending as couples and families, some with children. Inside the tent near the start of the walk, information tables displayed literature about IPV and brochures from local organizations. Ethnic snacks, mostly cooked by CBO members and volunteers, were served while participants stood about and mingled. According to the event organizer, approximately 135 participated in the walk.
Before the walk started, research team members introduced themselves and explained the procedures, including the use of a video camera, emphasizing the voluntary nature of participation and audience members’ rights not to participate and/or be videotaped. We also warned them that the content of the skit might be upsetting, offering them an opportunity to leave the immediate area should they not wish to view the skit. Then, the peerformers took the stage and performed the skit. At the request of walkathon organizers that the walk start right after the performance, we made adaptations to forum theatre conventions—an illustrative example of the flexibility and tailoring necessary in a highly applied community setting like this one. Unlike conventional forum theatre, where audience members are invited to go onstage and try out their interventions right after the performance, at this event, interaction between audience members and peerformers happened slightly later, during the course of the walkathon; in other words, the walkathon itself became the stage.
Immediately after the skit, audience members were encouraged to start walking and told that they would encounter the characters again towards the end of the walk when they would have an opportunity to intervene to help the family out. Throughout the walkathon, characters from the skit (played by another set of peerformers) appeared and talked to the walkers along the path to remind them of the skit and cue them to consider possible bystander actions. For example, the wife approached the walkers, asking them to help her with her controlling husband. Separately, the husband walked up to various walkers to tell them that he needed help talking to his wife because she was not listening to him. Near the end of the walk, the original peerformers again performed the initial skit as walkers approached the finish line. Peerformers who did not take part in the skit and members of the research team encouraged the walkers to try to help the family by entering the scene as a friend or family member (e.g., a bystander). After walkers had intervened and moved on, the peerformers would regroup and the process would repeat itself, with the peerformers enacting the same skit for a new group of approaching walkers. Having walkers intervene as a new character in the scene was also a departure from traditional forum theatre. This approach was developed in close collaboration with peerformers, who explained that community members would be more willing to participate in the skit if they would not be perceived as “taking over” someone else’s role. We videotaped these bystander interventions in the scene near the finish line.
Data Analysis Approaches
We analyzed both the audible and visible conduct of the participants captured in the video. To aid our analysis, we transcribed what was audible in the video recording and also included notable non-verbal behaviors. Video recording provides rich data that preserve relevant details of situated action (Mondada, 2012); however, the richness also presents complications and methodological challenges. Analytical and methodological work on videotaped human behaviors and social interactions encompass a wide range of disciplines, including visual sociology (Knoblauch & Tuma, 2011), visual anthropology (Collier & Collier, 1986), and visual ethnography (Pink, 2001). Specific methodological choices and procedural details adopted for specific studies largely depend on research questions and overall project goals, as well as disciplinary training. We used the analytical approach of grounded theory (Charmaz, 2014; Glaser & Strauss, 1967) because it allowed for inductive, constant comparative analysis of each situated bystander interaction. We also utilized pertinent aspects of visual analysis (Goodwin, 2004; Heath et al., 2010) to examine bodily movements and facial expressions. We examined the video recordings and transcripts using Heath et al.s’ (2010) three-stage approach.
Preliminary reviews.
Three members of the research team, including the first and last authors, reviewed the video recordings multiple times separately and together to obtain a general overview of bystanders’ actions, both verbal and non-verbal. The goal of this stage was immersion. While we noted different types of interventions, we focused on obtaining a broad understanding of what happened without breaking apart the data to assess patterns.
Subsequent reviews.
Two members of the research team separately reviewed each bystander’s actions, breaking down their interactions in the scene: how they began their intervention, how and with whom they engaged, and the order of specific verbal and nonverbal actions taken. The first author met regularly with the researchers to review and verify these assessments of bystander behavior and the sequence of interactions.
Analytical reviews.
Both researchers independently reviewed video clips and associated transcripts to assess the particularities of bystanders’ specific verbal and bodily actions. To do this, they conducted initial, inductive coding of each transcript as recommended by Charmaz (2014), coding for actions, not events or themes, using gerunds. For example, when a bystander took the husband away from the wife, our initial code was “separating the couple physically,” not “physical separation” or “diffusion.” The researchers then met with the first author to review the initial codes. Then, as a group, using a process of constant comparison between coded data segments, they reworded, grouped, and regrouped the initial codes into a set of focused codes relevant to the research question (Charmaz, 2014; Glaser & Strauss, 1967). During this process, discrepancies between the two coders were also discussed and resolved as a group.
Subsequently, the research team, including the first author, engaged in focused coding of the bystander interactions. Watching the videotaped interactions, they labeled each bystander’s behavior with a relevant focused code. This process allowed the team to compare and contrast the observed bystander actions in order to characterize broader categories and their relationships, which form the basis of the results presented herein. This iterative process of constant comparison (Glaser & Strauss, 1967) was used to ensure that (a) codes were consistently applied across all bystanders and within each bystander; (b) broader categories reflected the codes that were grouped together, and (c) codes were grouped in ways that reflected their meaning.
Data Analysis Quality and Rigor
Several methods of triangulation contributed to analytic rigor. While it was difficult to conduct member-checking of the bystanders themselves given the public and transitory nature of the walkathon, we were able to debrief with the peerformers, who were also members of the community. In addition to a brief debriefing with peerformers on the day of the walk, we held several meetings with peerformers where they reviewed, reflected on, and examined the meaning of the bystander enactments; their perspectives on the sociocultural context of the bystanders’ interventions (see Yoshihama & Tolman, 2015) provided important information corroborating the present analysis of videotaped interactions. Participant observation and informal conversations with bystanders and other walkers by the first author also helped triangulate analysis (Lincoln & Guba, 1985). Credibility of results was informed by prolonged engagement and persistent observation (Lincoln & Guba, 1985). Although the data analyzed came from one event, the research team had been working and conducting research with the focal community for over 10 years, conducting IPV prevention programs, including Peerformance presentations.
Several methods were used to improve trustworthiness of the results, including use of multiple coders, regular meetings to reach coding concurrence, and memoing (Lincoln & Guba, 1985). Coders received initial training and additional on-the-job training in how to code and write memos. Throughout the coding process, the two researchers and the first author met regularly and wrote memos about conceptual issues, decisions about coding and categories, and questions and insights about analysis and interpretation. Periodically, they conferred with other researchers involved in the project and peerformers who participated in the walkathon.
Results
Passers, Observers, and Bystanders
Of about 135 people who participated in the walk, 55 individuals approached the scene and were videotaped; the rest finished the walk without being captured in the video. Not everyone, present at the scene (and videotaped), engaged in bystander action. Sixteen individuals passed without stopping, continuing toward the finish line, and 16 stopped to observe but did not intervene. The remaining 23 individuals, referred to as bystanders hereinafter, stopped at the scene and tried out an action to intervene.
In an open event like this, participants’ exact age is unknown. Thus, we used four broad age ranges—younger than 35 years, 35–49 years, 50–64 years, and 65 years and older—to classify bystanders and estimated their age based on appearance and other contextual information (e.g., walking with a cane). Participants’ gender was also assessed based on appearance. Two-thirds of the participants were female, and over two-thirds were aged 50 years and older.
Bystander Action
While each bystander intervened differently, our analysis identified certain categories of actions, such as seeking attention, exploring, de-escalating, encouraging conflict resolution and help-seeking, and expressing support. These actions fall into the following dimensions.
Introductory: Seeking attention and permission.
Some bystanders acted in a way to gain the couple’s attention, introduce themselves, and ask for permission to engage. Several bystanders used a short utterance, such as “hey” and “excuse me,” to seek the couple’s attention, while others sought explicit permission to engage: “May I interfere/bother you?” “Can I help you?” One bystander, a woman aged 50–64 years, began her action by introducing herself as a neighbor and legitimized her role and involvement: “Hey, hey, hey. I’m your neighbor and I just overheard.” Other bystanders skipped this introductory step and went right into action; approaching the couple, they expressed what they thought would be helpful, as discussed below.
Investigative: Exploring the situation.
Along the investigative dimension, bystanders observed, and sometimes explored, the situation. Many bystanders asked broad questions, such as “What’s going on?” “What’s the problem?” and “What’s happening?” Others inquired about specifics, for example, “How long has situation been going on?” Notably, following the exploration, only some bystanders explicitly acknowledged what they heard the couple say, mostly with a short utterance, such as “okay” and “okay, got it.”
Instructional.
Actions along the instructional dimension involved bystanders asking the couple to do something, such as stop yelling and seek outside help. Bystanders also acted in an instructional mode by sharing, if not imposing, their beliefs about what the couple should do (e.g., resolve conflict, work as a team). Below we describe three instructional categories.
Instructional: De-escalating the situation.
Bystanders attempted to de-escalate or diffuse the situation in various ways. Tactics ranged from brief utterances, such as “Stop!” “Hold on!” and “Calm down!” to more concrete instructions, such as “Take a break” and “Do not scream.” One bystander, a man aged 50–64 years, used multiple diffusion-oriented remarks. As he saw the husband raising his hand, he began by saying, “No, no, no. No touchy. No touchy. Keep it in words. Keep it in words.” He followed with “You guys gotta settle down.” When the situation did not calm down (enough), he gave the couple more specific instructions: “Take a few deep breaths.” “Count to 10 or 20 or a 100.” Diffusion strategies were mostly verbal, although several bystanders also physically separated the husband and wife.
Instructional: Addressing the conflict.
Many bystanders encouraged the couple to resolve the conflict. For example, the above-mentioned bystander who repeatedly tried to diffuse the situation also told the couple, “You guys got to talk it out and resolve it.… You gotta resolve it peacefully.… Figure it all out amongst you but no touching.” Another male bystander, aged 65 years or older, said, “Now the question of how to work it out. That’s the important thing.” When another bystander, a man aged 50–64 years, said, “Guys, you need to work it out, all right?” the husband responded, “We don’t have anything to work out,” to which this bystander replied, “Yes, you do. Please do. All right?” Another bystander, a woman aged 35–49 years, told the husband to “see if there’s some other way you can resolve this.”
Instructional: Encouraging help-seeking.
A small number of bystanders mentioned the need for outside help on the part of one or both individuals. Several bystanders made specific reference to the local South Asian CBO (the organizer of the walkathon), others mentioned the police, and another recommended that the couple talk to someone “you both trust.”
Informational: Sharing information and observation.
Along the informational dimension, bystanders offered various types of information, such as resources and procedures. For example, they explained how the local CBO provided assistance, including confidentiality principles to help ease the couple’s concerns that other people might find out about their problem. Sometimes bystanders shared their observations and thoughts. For example, a male bystander aged 65 years and older talked about more egalitarian gender relations: “Collectively we all men have to learn. Unfortunately, in our earlier days, we did not learn.”
Instrumental: Providing tangible assistance.
Along the instrumental dimension, bystanders also provided tangible assistance. For example, beyond telling the couple that they needed help, one bystander, a woman aged 50–64 years, went as far as to say, “You got to go counseling right now. I’ll take you to counseling.”
Supportive.
Bystanders’ actions also fell along the empathic and supportive dimension; they expressed empathy, provided emotional support, and sometimes expressed agreement with one or both members of the couple.
Supportive: Empathizing with the husband.
Several bystanders expressed some sort of agreement and/or empathy with the husband. For example, in response to the husband’s claim that the wife must stay home to take care of the children (“Our kid is important, and she’s the mother”), a male bystander, aged 65 years and older, said, “Because they (the children) don’t have any support, so you are right.” This bystander also downplayed the importance of the wife’s work responsibility. After the wife emphasized how important her business meeting was and that it was just an hour long, this bystander responded, “Meeting will be there tomorrow. Meeting will be there day after tomorrow. The important thing is if you do not train the kids now, when?” However, some bystanders’ expression of support for or agreement with the husband was conditional. One bystander, a man aged 65 years and older, said to the husband, “I think everything you said was ok except you have to realize that she (the wife) has certain responsibilities (at work). And she has to fulfill that.” Another bystander, a woman aged 35–49 years, said, “He has a very good point,” before stating, “But this is called abuse.”
Supportive: Supporting the wife.
A majority of bystanders expressed their support for the wife. This took various forms, such as telling the wife to go to her business meeting, saying, for example, “You have to put your foot down. No arguments. Just do exactly what you want to do.” Other bystanders directed their comments in support of the wife to the husband. For example, a female bystander, aged 35–49 years, said to the husband, “I’m on her side,” then turned to the wife and said, “This is America. You have a lot of support.”
Confrontational: Confronting the husband.
Most bystanders confronted the husband in various ways and to various degrees, frequently in tandem with supporting the wife. A female bystander, aged 35–49 years, told the husband, “You have no business pushing your own rights on her.” Another, aged 35 to 49 years, said, “You need to back off.” In response to the husband’s claim, “You’re the wife and you have to listen to me,” another male bystander, aged 65 years and older, told him, “That is exactly where the problem is. That is where we are going in the wrong direction.” Another male bystander, aged 50–64 years, said, “No, no, no. That’s wrong,” disagreeing with the husband who had said, “You’re the wife. You’re supposed to be here for the kids.”
Discussion
In this study, videotaped enactments in the context of forum theatre provided rich data for examining what people do as bystanders witnessing an IPV incident. One unique contribution of this study is its examination of bystander action to address IPV in a long-term relationship (i.e., the married couple in the script) instead of in a dating or acquaintance relationship, which tends to be the focus of existing research. Our investigation of a previously under-studied population group is another contribution to the field; this study was the first of its kind in a growing Asian community to examine what lay community members did when they witnessed an IPV incident. In response to the enacted scene involving a controlling husband, South Asian community members who attended the event responded in various ways: exploring and/or de-escalating the situation, providing information, and encouraging the couple to resolve the conflict and/or seek outside help. They expressed empathy, support, and agreement with the wife as well as with the husband. Notably, support for the husband was often conditional or strategic; bystanders empathized with the husband to “join” him, build rapport, and garner his trust. A majority confronted the husband.
Bystanders’ actions encompassed various dimensions and differ from the requisite steps for bystander intervention in the widely used situational model developed by Burn (2009) and Latané & Darley (1970): noticing an incident, identifying a situation as appropriate for intervention, and taking responsibility for acting. Previous research has also identified levels, types, and dimensions of bystander action. For example, McMahon and Banyard (2012) delineated two types of bystander situations (reactive and proactive) and incorporated the dimension of timing into reactive situations, where bystanders could intervene before, during, and after the incident; they further divided the reactive situations into low versus high risk. Bowes-Sperry and O’Leary-Kelly (2005) identified two dimensions of action—immediacy (low versus high) and type (low versus high level of involvement)—and classified bystander actions into four cells in a 2 × 2 table. These formulations were conceptually derived and informed by previous research and theories, whereas the dimensions in this study were empirically derived via grounded theory analysis. Unlike an axis or axes on a table or grid, with each cell pertaining to a specific type of action, the dimensions found in this study are more fluid; these dimensions can intersect and work together or may contradict each other. These dimensions also capture a wider range of action, including those that are conventionally not considered bystander interventions, such as empathizing with the perpetrator as a way to build rapport.
Implications for Bystander Training Programs and Research
The findings of this study underscore the uniqueness and complexity of bystander interventions. Unlike most social interaction, which tends to consist of some type of a pre-engagement phase followed by some type of engagement-oriented action before delivery of an intervention, many bystanders in this study went right into the intervention. This may have been a function of the setting, where event participants were encouraged to step up and intervene, which might have made them feel as though no introductory engagement was needed. It might also have been because the situation (e.g., the husband’s gesture suggesting physical violence) called for immediate intervention. This immediacy might be a realistic reflection of what a bystander might encounter in real life, as the immediacy dimension identified by Bowes-Sperry and O’Leary-Kelly (2005) suggests. Bystander programs must prepare individuals to intervene in varied situations. This poses a challenge to development of a generic bystander program that fits all risk situations.
The finding that each person intervened in a unique manner is not surprising since different risk situations and contexts call for different actions. Studies of bystander action among college students have also reported a range of responses (McMahon et al., 2013). Yet this study also elucidated variations in how people approach and respond even to the same scenario. While illustrating the difficulty in developing a generic, community-based IPV bystander program, the findings also suggest some possible directions. In lieu of prescribing and/or proscribing specific actions to take in a given risk situation, bystander training might introduce multiple dimensions of action, combinations of which could be used in different situations. For example, when witnessing an escalating incident, a bystander might act along the instructional dimensions to try to diffuse the situation, while in another incident, actions along the investigative and/or empathic/supportive dimensions might be helpful. Bystander training could help participants build a repertoire of actions while collectively examining the strengths, limitations, and unintended consequences of these actions in their sociocultural context. Such bystander training would be complex and difficult to manualize; it would involve reflexive and critical discussions and learning and trying out various possible interventions through role-plays and enactments in the context of forum theatre.
Methodological Issues: Limitations and Possibilities
One of the study’s limitations is use of a non-representative sample, which limits the generalizability of findings. For the current study, rather than inviting individuals to come to a study site, we went to where they were, a community-based event designed to create awareness about domestic violence. Those who voluntarily attended such an event likely differed in their interest in IPV prevention and/or bystander intervention from those who did not attend. Even among attendees, the majority finished the walk without watching the Peerformance at the end of the walk. While forum theatre serves to engage community members in trying out possible actions and interactive discussion, its effectiveness is strengthened when participants reflect diverse backgrounds not limited to those already interested in the topic. That the majority of those who voluntarily attended this IPV prevention event chose not to take part in forum theatre attests to the need to create an atmosphere where participants feel comfortable trying out bystander prevention strategies. While the absence of a video camera might have aided this goal, it would have precluded the possibility of studying what we intended to examine: bystander action in a specific sociocultural context. Another potential limitation is that peerformers and research team members encouraged walkers to intervene, which may have added to the performative nature of the event, potentially altering bystanders’ intervention behaviors. The tension between research and community-based prevention remains a challenge.
This study is of an exploratory and descriptive nature. Even so, beneath its simplicity lies methodological complexity, including use of a performance coupled with videotaped visual and narrative data. Extensive methodological discussions are outside the scope of this study, but an exciting and growing body of relevant knowledge can be found in discussions of performance/performed ethnography (Alexander et al., 2005) and ethnotheatre/ethnodrama (Saldana, 2008). Here, we highlight a number of salient aspects pertaining to the use of theatre in research in tandem with the study’s limitations.
Focus on the Observable
Our analysis was limited to the observable, specifically the audible and visible material captured by a video camera. We staged a skit and set up a video camera at a location where we thought could capture most, if not all, walkers heading to the finish line. Yet of the 135 participants in the walk, only 55 were captured in the video. Pauwels (2011, p. 14) argues that, as in non-visual research, in visual research “different questions and research methods necessitate different sampling strategies and data collection/production (shooting) techniques … more explorative research may benefit from more ‘opportunistic sampling’.” Nonetheless, this type of opportunistic sample limits the generalizability of the findings.
Focus on Bystander Action
Another limitation is that this study focused mostly on bystander actions and did not analyze the reactions of the couple (as improvised by the peer-educator actors). Bystander action takes place in a dynamic, evolving interaction with a couple, whose responses shape the bystander’s subsequent actions, and vice versa. Our focus in this exploratory study was to learn what bystanders do as opposed to examining the impact or effectiveness of specific bystander actions. Analytical approaches such as conversational analysis and focused ethnography might have been more suitable had we attempted the latter. Echoing Pink (2001), who points to the need for new methodologies, Stanczak (2007, p. 12) argues that “unique configurations of each individual project require correspondingly new and tailor-made methods.” We used the grounded theory approach widely applied to text-based data while incorporating certain aspects of visual analysis. This analytical approach is germane to the aim of this exploratory study.
Performance, Not a Natural Setting
A fundamental limitation of this study, by design, was its setting in the context of forum theatre; what bystanders witnessed and intervened in was staged. The extent to which the observed bystander actions reflect what event attendees would do in a real case of IPV remains unknown. To our knowledge, no research exists on the congruence between actions tried out in the context of forum theatre and conduct in “real life.” Indeed, particularly in the case of IPV, such research would not be possible for ethical reasons. A scripted, enacted play might be analogous to a vignette or a case example to which study participants are asked to respond. Compared to a vignette or case example, typically text-based, a theatrical performance provides multimodal information involving not only text and language but also gesture, bodily display, and facial expression. This multimodality can help render a hypothetical situation more realistic and relatable (Mondada, 2012). While Jouriles et al.’s program (2016, p. 74) uses virtual reality technology to “help participants suspend ‘normal’ reality and experience a ‘new’ reality,” our Peerformance strives to (re)create a familiar and realistic scene.
What enhanced the “realness” of the play in this study were the peerformers, whose phenotypical features and mannerisms were similar to event participants’. The peerformers also generated the plot, script, and characters. They drew on knowledge and experience of their daily milieu, which was shared by event participants. Engaging community members as peer-educators and actors in a play they wrote has the potential to facilitate emotional involvement and identification with the characters. According to drama theories (Kincaid, 2002), this type of identification increases the persuasiveness of the performance. Evaluation studies of health education programs have documented the positive impact of educator–participant ethnic/racial match (Traylor et al., 2010). Stanley et al. (2015, p. 128) state that theatrical work based on real life accounts and narratives “had the potential to deliver an emotional charge which contributed to authenticity and promoted imaginative identification;” as described by a youth participant, such work is “so close to home and it happens to people that you might know and … I think drama kind of conveys that a bit more.”
Paradoxically, because it is not real, a theatrical performance allows the audience to “conjure other possibilities for something that they know has already happened and therefore cannot be changed” (Prendergast & Saxton, 2015, p. 282). This possibility is critical to a study like ours, where we are encouraging people to change the course of what is happening. A Peerformance is a site of action, a realistic and familiar milieu where community members can try interventions out themselves or take a look at others’ interventions, action that is different from the norm of indifference and nonintervention. It can open up new possibilities for addressing IPV in diverse communities.
Footnotes
Acknowledgments
We wish to express our utmost gratitude to the dedicated peerformers whose enthusiasm, creativity, and compassion have made this study possible. We dedicate this article to a long-time peerformer, Mr. Prithvish Parekh. Despite his untimely passing, his commitment to ending gender-based violence lives on.
We also thank participants of the Walk to End Domestic Violence and its organizer, MAI Family Services, as well as Divya Chand, Jasleen Singh, Richard M. Tolman, and many other research team members and collaborators.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: The work was in part supported by grants from the University of Michigan Injury Prevention Center and Michigan Institute for Clinical and Health Research, as well as the Michigan Coalition Against Domestic and Sexual Violence.
