1In the Global South, and more precisely in Brazil, scholars interested in the interplay between language and social life have long argued in support of an avowedly political research agenda. This approach stems from the view that our discursive practices are not neutral, but rather “involve ideological and political choices (whether these are intentional or not), inflected with relations of power which produce different effects on our social world” (Fabrício, 2006, p. 48). This understanding concerns not only the socio discursive practices we investigate, but also the knowledge we produce in doing so. Accepting that our research projects are therefore unavoidably subjective — guided by the researcher’s point of view of what merits investigation, as well as the ways in which this should be conducted — enables us to be more open about the political agendas which underpin our investigations and how these mould the production of knowledge. Based on this understanding — that research endeavours aimed at furthering understanding of social life simultaneously contribute to its construction — efforts have been made to design anti-hegemonic research projects focused on the reduction of human suffering.
2The politically charged research agenda outlined thus far benefited from government policies implemented in Brazil in the early 2000s which led to both the widening of university access and subsequent widening of academic debate through the inclusion of a more diverse student body. However, the more recent rise, as well as the continued legacy, of populist far-right governments in the region points to a growing resistance to such a turn in the tides. The rolling back of hard-won rights, as well as outright attacks on the value of scientific knowledge (particularly that which does not produce short term benefits) has led researchers to reflect on new ways to communicate the validity of socially committed scientific investigations, particularly to non-specialist audiences. In light of such a backlash, proponents such as Luiz Paulo da Moita Lopes (2006) and Branca Falabella Fabrício (2006), amongst others, have further advanced an “undisciplined” approach to proceedings ; in other words, an approach that is at once transdisciplinary and transgressive. This implies not only enabling detailed microanalysis of discursive practice to be bolstered by methodological tools inspired by other fields of knowledge, but to think about new ways of producing and validating scientific knowledge which may further intelligibility of social situations.
3What characterises feminist studies in Latin America and more specifically in Brazil is that they take place in a region marred by the extermination of indigenous populations and colonial practices of enslavement, whose scars are still very much apparent today. The legacies of these brutalities are manifest in everyday life, from the normalisation of social privileges for certain groups (largely those considered white and of European origin), to the denial of basic human rights for others (overridingly racialized populations). Given the scope of social injustices which tend to intersect in the region, Latin American feminisms thus emphasise issues of gender, race, sexuality, class and generation (Matos, 2010) and tend to address the struggles of minority groups more broadly speaking, though researchers may choose to foreground gender where relevant. In addition to the theoretical insights they may afford, studies which “go south” therefore frequently tackle social problems specific to the region where they are able to produce a concrete effect on historic inequalities : from healthcare (Ostermann et al., 2020 ; Ostermann, 2021) and domestic violence (Ostermann, 2003) to traffic violence (Orton & Biar, 2020, 2021) and the “war against drugs” (Albuquerque, 2017 ; Araújo et al., 2020), examining their gendered aspects and the ways in which these are concurrently shaped by other social markers.
4Moreover, based on our understanding that white male norms have long dominated what may be taken as “universal”, we adopt the Brazilian feminist philosopher Djamila Ribeiro’s concept of “place of speech” (2017) to reflect on not only what it means to produce feminist knowledge in the Global South, but the way in which our own roles as white, female researchers become entwined in this process. Similar to the notion of “feminist standpoint” (Harding, 1986), Ribeiro’s construct refers to the idea that speech is necessarily conveyed from an ideological position stemming from large-scale structures. The idea posited by the author is that, in addition to gender, other social markers, such as race and sexuality, intersect and shape the way in which these groups experience the world and thus the discursive context from which they speak.
5In order to reflect on the significance of this for our field, we draw on examples from recent feminist investigations developed in Brazil, focusing on those which favour the study of narrative practice, viewed here as fertile ground for the examination of the co-construction of meaning, in particular the negotiation of social identities and their ascribed symbolic value. By taking the view that research practice may be viewed as a type of narrative practice in itself, we aim to shed light on some of the backstage proceedings (Goffman, 1956) which preceded the publication of these studies.
6As such, the next section outlines our view of narrative research and what we believe narrative analysis may contribute to feminist investigations, as well as some of the epistemological implications of such an approach. Following this, we outline two feminist studies recently conducted in Brazil in order to reflect on the issues raised thus far.
7Inauguration of the field of Narrative Analysis, (which in Brazil is understood as a branch of Undisciplined Applied Linguistics in the sense we have laid out above), tends to be attributed to the sociolinguist William Labov (see Labov & Waletsky, 1967 ; Labov, 1972) who conceived of narrative as “a method of recapitulating past experience” (Labov, 1972, p. 359-60). As part of what Alexandra Georgakopoulou (2006) has termed the “second narrative turn”, scholars such as Charlotte Linde (1993) began to pay greater attention to the sociocultural values and interactional practices involved in the production of discourse, understanding narrative rather as a situated and dialogical way of reconstructing past events. For authors such as Deborah Schriffin (1984), storytelling thus functions as a way of imposing order, coherence and meaning on what Richard Bauman (1986) has deemed “narrated events”, constituting the foremost means by which narrators not only articulate worldviews, but invite others to ratify these during “narrative events”, in other words, the interactions which unfold during storytelling (Bauman, 1986). For scholars associated with this period, the temporal organisation of experience is therefore regarded as an interpretive practice, which occurs from a retrospective standpoint, rather than being pre-discursive – what Elliott Mishler (2006) dubbed the “double arrow of time”, as stories are fashioned from the narrators’ present perspective.
8Yet further scholars in Narrative Analysis have placed increased emphasis on the creative processes involved in storytelling, suggesting that narrated worlds in and of themselves may not be pre-discursive, rather storytelling is a way of constructing or bringing these to life (see, for example Langellier, 2001 ; Järvinen, 2003). For some contemporary scholars, narrative may thus be seen as performative action (e.g., Georgakopoulou, 2006 ; De Fina & Georgakopoulou, 2008 ; De Fina, 2021), a position which Georgakopoulou (2006, p. 128) has deemed the “third narrative turn”. Narrative scholars associated with the third turn are thus interested in talk and what people achieve through talk, granting greater “centrality to speaker’s active engagement in the construction process of narratives” (Bamberg, 1997, p. 341), rather than viewing narrative as a form of accessing past experiences, or people’s reflections on these, which are considered empirically inaccessible. This shift has further led to a focus on the observation of micro, fragmented and/or non hegemonic social practices, redirecting attention beyond narratives of personal experience to include shared and vicarious experiences, for example. Drawing on such practices, researchers are able to simultaneously examine the ways in which narrators position themselves with respect to narrated events, the narrative event here and now, and the master narratives which position them, as these worlds are drawn together and become visible in the process of storytelling (Bamberg, 2006, p. 144-5).
9The evolution in narrative study outlined here — from “recapitulation” to “reconstruction” to “construction”, leads us to revisit the view put forward by Mishler (2006) that narratives are not stable, rather they are readjusted to match the narrator’s current world views, as perspectives fluctuate over time and interpretations of experience follow suit. Increased feminist literacy, for example, may prompt us to rename past experiences — at once considered unproblematic — as social injustices or aggressions. By observing narratives subjected to multiple retellings Kim Schoofs and Dorien Van de Mieroop (2019) have similarly observed the ways in which stories may be reoriented to fit in with shifts in “master narratives” in order to convey identities and values which are spatially and temporally relevant. Worldwide movements such as “Me too”, or in Brazil “meu primeiro assédio“ for example, have helped pave the way for master narratives regarding what constitutes consent, to be slowly reinscribed. As Michael Bamberg (2006, p. 143) asks : “How do new interpretations come about ? Are they based on reflections or are they emerging in and through talk ?” It is by focusing on talk and what is accomplished as narrative meanings are negotiated in interaction that we aim to further understanding of how new feminist readings of experience may emerge.
10Since the narrative analysis in which we engage in Brazil subscribes to the social and political agenda of what we have been referring to as “Undisciplined Applied Linguistics” (Moita Lopes, 2006 ; Fabrício, 2006), it’s worth reiterating that our research design aims to afford greater visibility to minority groups by examining social practices in which labels associated with those who find themselves on the fringes of the globalised system are produced and/or contested. Our means of doing Narrative Analysis thus incorporates reflections from Brazilian scholars, such as Milton Santos (2000) and Conceição Evaristo (2017), on the need to foreground the everyday experiences of minority groups. Though some scholars of narrative analysis in parts of the Global North may be largely interested in advancing theory (e.g., Almeida, 2021 ; Matwick & Matwick, 2014 ; Norrick, 2020), both the contemporary political situation and the geopolitical situation in which we find ourselves have led us to view narrative as a tool which may be used to garner understanding of social injustices related to markers such as gender, race, class and sexuality, amongst others.
11Such an ethical commitment to social transformation follows on the heels of Ana Lúcia Souza’s (2009) concept of “re-existence” and means paying attention to stories “less told” as conveyed by their very protagonists. The type of Narrative Analysis we have been developing in Brazil (see Biar et al, 2021) therefore aims to put these oft-silenced tellers and their voices — long silenced by colonial and patriarchal systems of oppression — at the centre of our research. Nevertheless, should we wish to comprehend their experiences and inscribe new meanings as we examine their narrative practices, we must continue to challenge not only the power structures which perpetuate the North/South divide, but also those which characterise our own research proceedings.
12As we have been outlining thus far, we see storytelling as a locus for the study of ways in which social actors (re)establish norms and values which govern specific communities of practice, by taking evaluative positions in relation to narrated characters, objects and actions and negotiating their identities in interaction with others. With this in mind, our research tends to prioritise the analysis of alternative narratives told by unconventional narrators, which we consider a productive point of departure from which to interrogate typically hegemonic or “master” narratives. Our choice to focus on stories told by those less privileged by gendered norms may help to garner understanding of the ways in which such actors are not only constrained by, but simultaneously reconfigure these norms as they struggle for recognition in the discursive domain. Moreover, we find that foregrounding gender as an axis of analysis enables new readings of data to emerge. This stems from the assumption that research proceedings entail a form of narrative practice in themselves, as research texts are woven together from otherwise isolated observations and selected data until they form a coherent whole.
13The field of Anthropology has long argued that research texts should be understood as literary genres, in which lexical choices, the use of metaphor and so forth all play a role in the representation of the other (Clifford & Marcus, 1986). The story each tells is thus one of an exponential number of possible versions. Should we adopt this view, narrative research dissemination can be viewed as a type of metanarrative practice. Moreover, in our efforts to make sense of the stories we analyse, the stories we produce are frequently interwoven with our own experiences and impressions as participant observers and, in a sense, could be regarded as narratives of vicarious experience, a genre termed “cross-fictional” by Maria Hatavara and Jamila Mildorf (2017). Though research texts do not typically include the features of mind representation associated with fictionality, the literary processes necessarily involved in their production may enable our stories to “travel across the factual or fictional divide” (Browse & Hatavara, 2019, p. 336).
14So, what does this mean for our research ? For those who subscribe to the idea that our only access to social life is through daily interaction and the meanings negotiated in these interactions, research procedures often involve some type of field work or (auto)ethnographic reflection to aid the interpretation of discursive data. Though the multiple perspectives lent by these traditions allow data to be viewed from multiple angles and may thus contribute to the robustness of a study, what may be seen by some as the researcher’s necessary involvement in the field (which merits analysis in itself), may yet be seen by others as undesirable and even verging on the self-indulgent.
15As narrative scholars in Brazil, we have attempted to give greater emphasis to the way in which scholars interact with the communities investigated, based on our acknowledgement that this involvement plays a role in the production of knowledge and that our research contexts tend to be marked by asymmetrical relationships, both economically and epistemically speaking. In a sense, we consider research activity to be inherently (auto)ethnographic, since the knowledge we produce will always depend on our “place of speech” (Ribeiro, 2017). Rather being seen as undesirable, we regard this subjectivity as inextricably bound up with what we are able to learn about the other (Coelho, 2016, p. 10). Ribeiro’s (2017) concept enables us to carefully consider what this means for our investigations, in particular the aforementioned asymmetrical relationships between researchers and participants. By taking this into account as we progress through the interpretation of data and the construction of our metanarratives, we are able to move closer to the construction of a polyvocal text.
16As we have been pointing out, what interests us about the stories we analyse is not how closely they represent events which may or may not have taken place in the past (Riessman, 1993), but rather what they tell us about competing worldviews and the way in which social norms and values are interactionally negotiated at the moment of storytelling, or during the “narrative event”. The narratives produced by researchers are similarly negotiated, edited and subsequently rearticulated at fundamental junctures, be this with peer reviewers, supervisors, colleagues, conference attendees, research participants or at public outreach events, amongst others. We can begin to think of these moments in which they are held up to scrutiny by others as “narrative events” themselves, in which new meanings may emerge. In other words, by presenting our work to different audiences we engage in narrative practice and the meanings produced must therefore be negotiated. In a Labovian sense, the “point” of the story must thus be accepted by these interlocutors who may question its coherence. Such moments of interrogation theoretically contribute to polyvocal texts (as per Bakhtin, 1981), in which the multiple voices of colleagues, supervisors, reviewers, conference attendees, research participants and so forth become contemplated, enabling the phenomenon studied to be considered from more nuanced angles and a more comprehensive understanding to be reached.
17Research practice is therefore a dialogical activity, whose findings are probed at these various junctures. Or as Mishler (1990, p. 422) puts it : “knowledge is validated within a community of scientists as they come to share non problematic and useful ways of thinking about and solving problems”. As Catherine Riessman (1993, p. 64-8) further points out, this means that validity may be assessed by examining the degree to which alternative interpretations are considered and research participants actively involved in the production of knowledge, as well as the texts’ interpretative coherence and pragmatic use i.e., to what extent a study is future oriented and may become the basis for others’ work. This may involve, for example, transparency regarding the journey to interpretation, as well as enabling third party access to primary data.
18Having said that, there is still much work to be done should we wish our texts to engage more fully with a range of differing perspectives. Frequently our opinions and interpretations may still appear disguised as facts and certainties, excluding alternative explanations by means of rhetorical devices which silence dissent. The construction of coherence requires editing and the choices underlying this produce both ethical and epistemological repercussions, as Richard Bauman and Charles Briggs (1990) have warned. We need only look to the theatrical metaphors of Erving Goffman (1956) — who suggested that the performance which is social life frequently masks the many layers of preparation which precede it — should we wish to consider the merits of discussing what goes on backstage. During the research process, rather than progressing towards illumination in an orderly fashion, investigators invariably question, edit, erase and rewrite their own reports and findings, which will later be subject to further scrutiny. Instead of “black boxing” (Latour, 1999) this human process behind the production of knowledge, greater transparency regarding the networks involved, represents a necessary epistemological shift should we wish to advocate for the legitimacy of our findings (Riessman, 1993). This may mean, for example, incorporating details of the trials and tribulations involved in a text’s construction and the various narrative events which punctuate this, which may be viewed as indispensable “turning points” (Mishler, 2006) in the building of unforced consensus on a given subject. This article represents one such attempt to do just this.
19Our first example is drawn from a study conducted by the first author of this paper, under the supervision of the second (Orton & Biar, 2020 ; Orton & Biar, 2021). The project in question aimed to examine the ways in which contemporary social movements produce politics/resistance and allegedly horizontality, by investigating a specific group : bicycle advocates in Rio de Janeiro. In other words, the investigator sought to understand how the much-lauded concept of horizontality played out within the group and what participants were doing (or not doing) in their discursive practices to create more symmetrical relationships, based on the assumption that they are also participants in a society which is organised hierarchically across different social axes.
20The author chose to focus on a movement in which she herself is a participant, enabling her to draw on her experiences in the group’s activities in order to assist with data readings. In 2016 she additionally began recording debates held by the group and these recordings formed the study’s corpus. Though gender did not feature as a line of inquiry in the project’s initial outline, on listening back to the recordings, transcribing data, and sharing these transcriptions with members of a research group interested in narrative and social life (NAVIS), gender became increasingly visible as a possible axis of analysis. This is not to say that other social asymmetries are/were not at play in the group’s interactions, rather to say that gendered dynamics appeared to loom large when examined by a white middle-class female researcher and colleagues with largely similar backgrounds. It’s also worth mentioning that bicycle advocacy in Brazil generally draws a homogeneous crowd of middle-class cisgender white men, even though the overwhelming majority of those who adopt bicycles as a means of transport in Brazil do so out of financial necessity. Though the absence of a more diverse range of participants from bicycle advocacy in Brazil merits investigation and discussion, the choice to focus on gender as a vector of analysis similarly stemmed from the fact that women do participate in bicycle advocacy in Brazil, despite representing a minority of bicyclists in Brazil, thus enabling empirical observation of the meaning-making practices in which this particular minority group engaged during the discussions recorded.
21These preliminary insights regarding gender further catalysed new readings of interactions within the advocacy group itself, both of the audio recordings and those recorded on paper as part of the (auto)ethnographic project, as the researcher was able to reinterpret past experience from this newly acquired critical standpoint (Mishler, 2006). Should we accept the idea that scientific knowledge is constantly evolving rather than being immutable, open discussions regarding this process need no longer be seen as a threat to the validity of our investigations. Rather, such reassessment of experience is imperative for historical injustices to be identified and reframed, enabling a “radical temporal reorientation of knowledge” (Miyazaki, 2004, p. 5) to gain momentum and social change to become tangible.
22Nevertheless, these feminist readings were not shared by all those with whom initial findings were shared. Conference participants from other fields of study, for example, failed to identify gender in the interactions presented, questioning such an interpretation. A further hurdle emerged during what may be deemed a “narrative event” in which a panel of specialists in language and gender scrutinised the narrative of gender presented at a preliminary evaluation of the work, calling the researcher’s epistemic authority into question. This “turning point” (Mishler, 2006) in the narrative made one thing clear : in order to pinpoint exactly how gender seemed to be made relevant in the interactions new theoretical tools were required, as a focus solely on features of turn-by-turn speech, such as overlapping talk, failed to convincingly account for this. Gender rarely appeared explicitly in the debates recorded and yet nuanced discursive phenomena seemed to point to gendered dynamics at work. As a result, the investigator chose to centre the analysis on the alternative narratives told by women during the discussions recorded, with a view to better understanding how they negotiated and held the floor and how the meanings raised in their stories were co-constructed by other participants. This meant scouring the data for such stories, since most participants were male, and these participants tended to dominate talk. To analyse the stories selected, the author drew on theories of identity construction (Bucholtz & Hall, 2004, 2005) ; narrative roles (Ochs & Taylor, 1995) ; constructed dialogue (Tannen, 1989), verbal hygiene (Cameron, 2006) and code switching (Fuller, 2009).
23These constructs enabled the researcher to observe the ways in which female participants navigated the currents of a social movement in which the performance of symbolically masculine norms appeared necessary for members to accrue epistemic authority (Levon & Ye, 2020) in the eyes of their peers and in which female participants were simultaneously still seen as “interlopers” (Eckert, 1998). As a result, it seemed that when female participants vied for status and attempted to share their gendered experiences of commuting by bicycle, discursive micro aggressions occurred, be these practices of verbal hygiene (Cameron, 2006), or problematisation of their narrative practice (Ochs & Taylor, 1995), leading to the gendered issues they raised frequently being side-lined as these members found themselves trapped in what Robin Lakoff (1990) described as a “double-bind”.
24Moreover, questions raised during the course of the research narrative’s construction, with respect to the researcher’s epistemic authority as a scholar of language and gender, suggested that further details regarding the group’s specific characteristics — until then largely absent from the research text — urgently required incorporation. Gender as a potential line of inquiry had much to do with such characteristics, with which outsiders had little familiarity. Cycling as a sport has long been dominated by men, or, as the British athlete Nicole Cooke put it “cycling is a sport run by men, for men.” To this day, races such as the “Tour de France” do not hold a parallel women’s competition on the same course. Women are rather reserved the role of “podium girls”, presenting men with their prizes. Cycling for transport is not remarkably different. With the exception of perhaps the Netherlands, most countries still tend to see a stark discrepancy between the number of men and women using bicycles for transport. Studies have suggested this gap tends to be linked to gender equality in the region more broadly, as well as the infrastructure available (see, for example, Singleton & Goddard, 2016). In Rio de Janeiro, where the study in question was based, a local NGO suggests that less than 4 % of those on bikes in the city centre are women. As we have mentioned elsewhere (Orton & Biar, 2021), prevailing gendered norms regarding who has the right to occupy public space mean that, in addition to traffic violence, women are subject to all manner of abuse while cycling, from catcalling to gender-based violence and assault, which may account for their diminished presence on the roads.
25In addition to such large-scale contextual information, the inclusion of specific (auto)ethnographic observations such as decision-making processes regarding routes to be taken by the group (in which performances of what Robert Lawson [2013] describes as “tough masculinity” appeared to be rewarded), or the assignment of gendered roles (such as minute taking) during debates, enabled a more fully formed account of the social context the speakers inhabit to be developed and shared with potential interlocutors (Lawson, 2013). These contextual details — otherwise inaccessible to the wider community – helped shore up the analyses, as well as simultaneously working to allay criticisms of “bias”, by juxtaposing insider (the first author of this paper) and outsider (the second author of this paper) viewpoints, thus aligning multiple, sometimes conflicting perspectives. This process similarly involved repeated rewritings of the text, with a view to producing a more dialogical narrative, acknowledging alternative interpretations of the data : both by presenting multiple possible readings of the phenomena examined and by paying attention to both syntactic and lexical choices involved in the presentation of these interpretations, with a view to avoiding categorical assertions which might communicate the idea of inarguable truths.
26By taking a new approach to presentations of the data, consensus regarding feminist readings was slowly built. Thus, by at once adhering to a feminist agenda and simultaneously revisiting the theoretical-methodological tools employed to do so, we were able to observe more nuanced functions of discursive practices such as codeswitching (Fuller, 2009) in the public sphere, the negotiation of epistemic authority during narrative events (Levon & Le, 2020 ; Bauman, 1986) and positioning in relation to conventional norms of femininity. This framework allowed the data to be seen from new angles which, supported by (auto)ethnographic details, shone new light on the movement’s gendered dynamics. By accepting that multiple readings of any given data set are possible, the choice to draw on a theoretical framework which enabled such asymmetries to become more widely visible no longer needed to be seen as problematic, but rather as a necessary ingredient as part of a political research agenda which merits discussion.
27As underscored earlier, the potential for narrative events ⎯ such as the preliminary evaluation described here ⎯ to catalyse new meanings is by no means limited to the academic sphere and attempts to share initial findings beyond the academic community provided challenges of their own. In addition to potential disputes over meaning, such outreach activities further involve rendering results intelligible to those unfamiliar not only with the analytical categories or transcription conventions employed, but on occasion fundamental assumptions of the field, such as the idea that social reality is discursively constructed.
28One such endeavour involved the sharing of findings at a national event known as “Bicicultura”. The event in question is organised by Brazilian civil society with the aim of bringing together bicyclists, activists and all those interested in democratising the urban environment with a view to exchanging knowledge and building new partnerships. Though findings from a range of academic studies were shared during the proceedings, these tended to involve the presentation of more easily digested graphs, charts and percentages, rather than attempts to account for the subtleties involved in everyday communication. On presenting a study which examined the very minutiae of daily interaction, the occasionally perplexed faces of the audience and lack of questions which followed seemed to point to the difficulty of communicating the methods and results of qualitative, interpretive discursive research to a lay audience better accustomed to hard facts and figures.
29This impression appeared to be further corroborated by a request from a female participant involved in the study to quantify the number of occasions on which she was “interrupted” during meetings. Around the same time (c.2016), the term “manterrupting” began gaining traction in Brazil, along with the appearance of billboards advertising an app developed for use by the general public which claimed to “detect manterruption in everyday conversations” by means of voice recording. Though there is value in raising public awareness of the potential relationships between discursive phenomena and gendered dynamics, we believe researchers should strive to avoid the temptation to offer neatly packaged solutions to complex problems, which may fail to account for the subtleties of situated interaction. The analytical categories selected to examine the data in the study outlined here hinted at the potentially multifunctional nature of overlapping talk, which may be at once competitive and collaborative, whilst simultaneously enabling us to pay attention to the ways in which the use of code switching (between formal and informal registers) contributed to the negotiation and maintenance of high-status roles within the group as gendered assumptions were brought to bear on the interactions. While female participants frequently adopted informal registers during narrative practice, overlapping speech from male counterparts which exhibited semantical likeness ⎯ but tended to be uttered in a more formal or technical register ⎯ appeared to point to claims to specific knowledge (of bicycles, or urban mobility more generally) and thus greater storytelling rights in the context of bicycle advocacy.
30As previously mentioned, findings such as these ultimately rely on choices made by the researcher with whom decision making power tends to lie in terms of participant representation. This “power of representation” (Coelho, 2016) points to the shifting nature of power dynamics as overlapping roles and social markers ebb and flow during interaction (in this case, the complementary roles of a low status female participant in a bicycle advocacy group and a high-status academic researcher at a Higher Education institution). As previously mentioned, one way of potentially redressing the power imbalance between researcher and participants is by involving participants in the construction of the research narrative. In the study in question, those who appeared on the recordings selected for analysis were invited to offer input on the transcriptions, with the aim of creating a more polyvocal narrative. Nevertheless, one male participant reframed such an invitation as a request for “help”, appearing to temporarily invert power dynamics. Though this interactional manoeuvre could be attributed to a lack of familiarity with participant involvement in academic studies, gendered assumptions regarding the researcher’s ability to adequately carry out the work herself may also have been at work here.
31Nevertheless, later attempts at involving activists connected to the study suggested ways in which shared meaning might be created ; in particular, the potential for small stories to bridge the divide between the sometimes-opaque world of academia and informally organised civil society groups. This potential became particularly visible during a conversation circle arranged by a collective of bicycle messengers and small business owners involved in the world of bicycling. Though the event's proponent was one of the few female members of the collective and the meeting drew participants from beyond the strikingly masculine niche of cycle couriers, the audience on the day was predominantly male.
32On sharing results of the study by means of informal anecdotes, participants initially appeared reticent, or possibly even disinterested, glancing furtively at the screens of their devices. However, after a period, one of the attendees shared a small story concerning a recent incident while cycling as part of a group which struggled to maintain cohesion. According to the narrator, as members spread out during the ride, a female counterpart asserted her concern at being left behind and was subsequently labelled aggressive. By highlighting that the participant’s use of language might not have been interpreted as such had gendered assumptions not been brought to bear on the interactions, the storyteller thus communicated understanding of the “double bind” in which his peer found herself.
33Pedro Garcez (2001) has pointed to the ways in which second stories, by encapsulating the same evaluative point of a previous story, serve to demonstrate understanding of a story’s point and build solidarity, as this example seems to illustrate. Moreover, Mary Gergen and Kenneth Gergen (2006, p. 117) have argued that stories may be more effective in crossing “boundaries of meaning” and bringing people “into a state of mutuality” than other forms of discourse. The authors further suggest that the storytelling process fosters receptivity as it eliminates the tendency to argue against a point of view, since it is “socially difficult and even rude to directly challenge or undermine a personal life story” (Gergen & Gergen, 2006, p. 116). Stories are also easier to comprehend than abstract arguments, tend to afford tellers credibility as “witnesses” of lived experience and thus similarly foster empathy on the part of listeners who are less likely to stand outside as analysts or critics, as may be the case with arguments. For these reasons, we believe that the sharing of research narratives, as well as the alternative narratives we analyse, may be a way of harnessing the potential for socially and politically committed research to “lever human change” (Gergen & Gergen, 2006, p. 119).
34Our second example is also guided by the political agenda of Brazilian Applied Linguistics, particularly with respect to “stories less told”. This is an example that sheds light on an aspect of Brazilian public security policies that affects the routine of underprivileged women : the mass incarceration of Black men from Brazilian favelas. We now turn to the stories of women who occupy the margins of Rio de Janeiro's penitentiary system as visitors (mothers, daughters, sisters, wives) to their imprisoned relatives. Although prison contexts have been widely investigated by Brazilian social scientists — and although such contexts have given rise to an array of documentary and audio-visual fictional productions —, there are still few studies which foreground the (secondary ?) role played by women in the prison scene. In this project, the researchers were particularly interested in listening to some of these stories with a view to increasing their circulation beyond the prison's environs.
35Supervised by the second author of this paper, Natália Albuquerque (2017) set out to observe and generate data behind-the-scenes of a male penitentiary in the largely overlooked visitors’ queue. The researcher produced an ethnography of inmate visitation days when, from the early hours of the morning, large lines exclusively composed of women materialise in the vicinity of the penitentiary. Her detailed description was complemented by in-depth interviews in which collaborators narrate their life stories and outline their journeys to offer support to their institutionalised relatives.
36In this case, unlike in the previous example, the researchers had a gendered perspective in mind from the outset. They are further outsiders who occupy privileged socio-economic positions in Brazilian society. By comparison, the community investigated is largely made up of Black women (aged 25 to 57) on low incomes, living in the peripheries of Rio state. Because penitentiaries are generally located in isolated areas, these women's journeys tend to begin the day before. Equipped with large bags filled with food, clothing and hygiene products not provided by the State, these visitors are often forced to spend the entire night outside, often exposed to extremely high temperatures, to ensure their entry the following day.
37Upon entering the penitentiary, these women are subjected to rigorous routine inspections the purpose of which is to intercept forbidden materials, such as drugs and weapons. These inspections entail a thorough examination of clothes and bags, as well as direct manipulation or confiscation of the food they have brought. A number of women are also made to undergo the officially abolished, but still widespread practice of “intimate searches” (in Portuguese, “revista vexatória” : shameful search). This type of search constitutes a major violation of rights, which involves stripping visitors of their clothes and examining their genitals in pursuit of hidden objects. Among other initiatives deployed to control women's bodies in the service of so-called “institutional security” is the imposition of loose, long and plain clothing, a “dress code” recognizably similar to the one prescribed by traditional evangelical churches. Such bodily performances, as we know, naturalise the idea that women are to blame for the countless instances of sexual harassment which take place in mostly male institutions.
38In short, in order to attend to needs often overlooked by the State and to support their imprisoned relatives, these women subject themselves to a degrading and exhausting routine of responsibilities. Even so, they must continually negotiate the stigma of maintaining relationships with “criminals” and, what's more, respond to a discursive logic, supported by a patriarchal system, which blames them for the fate of their partners. When asked to assess their own situation, research collaborators commonly recognize that they are “treated like animals”, but that they undergo this routine “because of love”.
39Given this context, a first point that stems from the epistemological discussions proposed in this text is that research interviews should be understood as social encounters between two subjects, and not as a mere methodological device. As per Ribeiro's previously mentioned perspective, the “mixed” (Goffman, 1963) status of these encounters bring into play different “places of speech” and should not be neglected. Here, the research scenes are best seen as interactions in which a number of social symmetries and asymmetries must be managed. As such, Albuquerque's (2017) interviews bring together differently positioned Black women : on the one hand, a young Masters’ researcher ; and, on the other, the relative of an inmate, typically a woman in a much more precarious condition. As in other meetings of this nature, differences and similarities are newly inscribed in the interaction in a way that can be productively explored in the analytical stage.
40As Albuquerque (2017) examined these interviews, her general research aim was to observe the narrative resources available to collaborators as they construct themselves. Furthermore, her analysis highlighted the emergence of vicarious narratives (Norrick, 2013) and considered how her collaborators construct their imprisoned relatives. Throughout her study, she highlighted how specific systems of coherence (Linde, 1993) were constructed by these women, discussing the notions of justice, absolution, and commitment according to which they operated.
41One of these interviews, in particular, triggered the “turning point” for this research, becoming the subject of intense discussion during numerous supervision sessions with the researcher, as well as on the occasion of her viva. It is the story of Sílvia (pseudonym), a 39-year-old, low-income housemaid who regularly visits her 57-year-old husband, accused of raping his daughter from a previous marriage. During her conversation with the researcher, Sílvia revealed that she herself had been a victim of rape as a teenager.
42Like the other interviewees, Sílvia resists the stigma attributed to those associated with criminality, animating (Goffman, 1981) a vicarious narrative that places the accused in a favourable light, all while disqualifying the formal accusations made by her stepdaughter and her husband's former wife. Unlike the other collaborators, however, she intertwines her own life story with this vicarious account. Her story starts with the following summary : “I never believed it, you know why ? ‘Cause I was raped and I know the pain of that”. Here, Sílvia suggests that she possesses double epistemic authority over her companion's innocence : an authority emanating not only from her status as a family member of the accused (Norrick, 2013), but also from her own condition as a victim/survivor of rape.
43Following the argumentative structure announced in her summary, Sílvia's narrative then proceeds to lend weight to controversial argumentative premises, among them : (i) that rapists have an aggressive personality ; (ii) that rapists do not act against blood relatives ; and (iii) that victims/survivors do not have the strength to press charges.
44In face of such a discourse, all of us — researcher, supervisor, members of the panel, all women engaged with Gender Studies — were puzzled. A number of concerns, criss-crossing different phases of the research, came to light during this encounter. As a consequence, we were prompted to pursue certain (re)writing exercises (such as this one) in order to further contemplate our ambivalence towards the data. The first issue emerged during fieldwork : the discomfort narrated by the researcher, who was unexpectedly confronted with an account of rape. According to Albuquerque (2017), the way Sílvia, with whom she had no personal relationship prior to the interview, opened up about the rape was “surprising and odd”. Using Goffmanian lexicon, extracted from her theoretical framework, she understood “such a personal and intimate” revelation about a “violent and embarrassing” episode as a threat to both their faces.
45From a micro-sociological point of view, the embarrassment suggests that a social expectation had been breached : intimacy and confession seem out of place in a research interview, especially if the exchange is conducted in public, such as in a prison visitors’ queue. Now, in hindsight, we may perhaps revisit this research meeting and consider how much race and gender came to intersect in this supposedly impersonal interaction, decisively affecting the way in which it was co-constructed. The research scene temporarily gives way to a less vertical dialogue that encourages the sharing of experience. At the same time, the rupture caused by Sílvia somehow produces the effect of reframing the “personal” as “public” (or “political”, in the terms of Carol Hanisch's 1969 essay), with significant consequences for the research report — in which the story of this particular interview became particularly prominent.
46However, that was not the most delicate issue which required our consideration. If, in the first example of this article, there seemed to be a certain compatibility between the data and the researcher's beliefs about gender, and the researcher's own ideological framework could prompt her to highlight narratives of resistance shared by underprivileged groups, this is not the case here. Sílvia's narrative engenders an ideological impasse : by defending her husband, she reinforces discourses which disqualify victims/survivors, thereby upholding patriarchal views about the origins and manifestations of sexual violence which are vehemently disputed by contemporary feminist research. Although it is not entirely unheard-of for women to call into question narratives about rape especially in contexts where a man of their acquaintance is being accused, some decisions regarding research writing needed to be made so that the participant's voice here was not simply disqualified.
47When we conducted a meta-analysis based on a different research process elsewhere (Biar & Torres, 2018), we argued that such movements call for “evocative introspection” (Anderson, 2006), defined as a form of analysis predicated on the researcher's very status as a “non-member”. A deliberate recognition of the “outsider” dimension from which researchers approach the field is required. This entails a careful investigation of our personal experiences in the field (see Ellis & Bochner, 2000), including constraints and ideological discrepancies, which must be regarded as important turning points in determining the direction of the study. In the case of Albuquerque's work, such a perspective was instrumental in allowing for a negotiation of both the researcher's and participants' expectations — including an awareness of the former's desire to produce favourable representations of the latter. This debate is, of course, inseparable from the well-known issue of how to represent subaltern identities — or, in Riessman's (2005) terms, how to present narratives generated under unequal conditions and ethical uncertainty.
48During the researchers' conversations about this data, they emphasised the importance of debating the ways in which socially committed academic research may at times romanticise peripheral knowledge, thus producing a form of colonialist condescension. We also discussed the need to theoretically address the uneven circulation of, and access to, certain discourses, such as those underpinned by feminist ideas.
49Above all, these reflections allowed us to dodge the initial impulse (and the ethical problem) of simply disregarding impertinent data, or, alternatively, of oversimplifying it and presenting it as mistaken or ill-judged. Instead, we are on our way to consolidating new understandings about how to conduct discursive research — by women and for women —, and to view such studies as a bricolage of overlapping, non-hierarchical voices. This, in short, allows us to sympathise with and better understand differences.
50Though the two examples shared here arise from strikingly different research contexts, both highlight the ethical necessity of transparency vis-a-vis the often-messy behind-the-scenes discussions which precede the publication of any study, should we wish to make claims to validity. They further point to some of the thornier issues involved more specifically in feminist investigations and potential ways in which these may be addressed as scholars negotiate the meanings of the metanarratives they produce at crucial junctures we have been referring to as narrative events.
51Our first example highlights how increased feminist literacy may be built during the research process and enable new readings of a given social situation to emerge as researchers fashion narratives from their present position (Mishler, 2006). This investigation further suggests that the nuances of some gendered discourse phenomena may require an array of theoretical-methodological tools in order to contemplate their sometimes-multifunctional nature and build consensus regarding feminist readings, particularly when investigating situations which — in the context of Brazilian society — may not be widely regarded as misogynous. Moreover, we argue that once we accept the view that multiple readings of any given data set are possible, the selection of a toolkit which enables such feminist readings to become visible need not be seen as problematic provided our political agendas are made explicit. Finally, this example enables us to observe how the exchange of small stories may be an effective way of creating shared meaning and actively involving research participants in the construction of research narratives as part of a polyvocal project.
52Our second example illustrates some of the ethical dilemmas produced by disconcerting data in feminist research and the ways in which these may be bound up with the typically asymmetric relationships between researchers and collaborators. In particular, this example points to the urgent need to consider our “place of speech” (Ribeiro, 2017) as researchers when drawing conclusions regarding data produced under unequal conditions in a society still scarred by colonial processes and where access to feminist literacy is by no means guaranteed.
53As scholars interested in narrative analysis and committed to a political agenda, we consider the sharing of these behind-the-scenes research stories “less told”, as well as the subsequent dialogue this may lead to, a fundamental part of the research process, which may in turn generate new meanings and understandings of the stories “less told” that we strive to place at the centre of our investigations.