Short-term ethnography and women’s voices: insights from fieldwork with Roma communities

ABSTRACT

In this paper, we explore the potential of short-term ethnography in a collaborative and participatory research project on experiences with and perceptions of migration from fieldwork with Roma groups in north-western Transylvania. We discuss particularly the challenges we encountered in engaging Roma women in terms of existing cultural and gender norms evident in some Roma communities. We argue that, in the context of working with women in patriarchal communities, short-term ethnography offers researchers a unique opportunity to (a) find and take advantage of safe spaces in which to talk to women and (b) manage sensitively access and rapport. In this study, ethnography gave insights into alternative ways of engagement in the research process based on categorisations that emphasise alternative aspects of social identity (Roma ‘mothers’ vs. Roma ‘women’) as an effective means of breaking down some of the barriers to participation. Finally, we argue that short-term ethnography can be used successfully in collaborative and participatory research projects alongside traditional ethnographic work to explore the perspectives of mobile and vulnerable or marginalised groups including, but not limited to, patriarchal group contexts.

KEYWORDS:

• Short-term ethnography
• women’s voices
• mobilities
• vulnerable or marginalised groups
• Roma

In this paper, we address the challenges we encountered in engaging Roma women in a collaborative and participatory research project of intra-European migration with Roma groups. We had a specific concern in mind to address the value of short-term ethnography for exploring gender and women’s voices (or lack of) in Roma communities in the context of sensitive negotiation of access and rapport. We start by discussing the importance of ethnography in studies of migration before turning to reviewing some of the dilemmas and challenges in research with mobile and vulnerable or marginalised groups. We then move to a discussion of traditional ethnographies with Roma people followed by a description and reflexive analysis of our fieldwork and unique advantages of short-term ethnography. We conclude with a discussion of the added value of short-term ethnography in collaborative and participatory research projects for giving a voice to women in patriarchal group contexts with significant consequences for participation and engagement.

Ethnography and migration studies

Ethnography is firmly embedded in the methodological toolkit of migration researchers. The case for the centrality of ethnography in transnational migration has been made on several occasions (e.g. Glick-Schiller 2003). Some researchers have argued that participation observation is an essential mode of enquiry in exploring international migration (Boccagni and Schrooten 2018), whereas others have emphasised the importance of other ethnographic methods and techniques (Paerregard 2008; Fernandez-Kelly 2013).

Ethnography is particularly suited to researching people flows across borders because it is ‘reflexive and flexible, adaptable to changing research conditions and to evolving research questions’ (Mahler and Pessar 2006, 32). Its ‘suppleness’ (Mahler and Pessar 2006, 32) makes it an appealing choice for researchers. Ethnographies can take several forms – from longitudinal and multi-sited (Punch 2012) to comparative ethnographies (Fitzgerald 2006; Levitt and Jaworsky 2007).¹ Variants of ethnography have been used to explore links between migration and globalisation (Wilding 2007; Fitzgerald 2006; Paerregard 2008), migration and precarity (Tervonen and Enache 2017) or migration and children’s experiences (Enache 2018).

Irrespective of label, however, ethnographic methods in ethnographies of migration are mostly suited for revealing the richness of migrants’ stories (see Lawson 2000). When the focus is on exploring people’s stories, ethnography is usually supported by or embedded in the use of traditional social science research methods (e.g. interviews, life story interviews). This is particularly the case for ethnographies of destitute groups, marginalised minorities, and forced and undocumented migrants (e.g. Khosravi 2010; Tsianos and Karakayali 2010). Ethnography is particularly useful whenever there is a substantial element of participation and observation. As O’Reilly (2012) argues, the advantages of ethnography come especially into view when researchers engage with the ‘dialectic relationship between participating and observing’ (4). In that sense, ethnography is a more effective method than more traditional narrative and interview-based approaches because it can accommodate different styles of embodied practice and rapport that can take the form of ‘opportunistic chats, questions that arise on the spur of the moment, one-to-one in-depth interviews, group interviews, and all sorts of ways asking questions and learning about people that fall in between’ (O’Reilly 2012, 4).

It is this blend of participation and observation, and the different ways of building rapport and research engagement, that make ethnography particularly suitable for working with vulnerable groups. A crucial aspect of ethnography is participation in the lives of the people themselves, understanding and experiencing the world as group insiders do whilst being keenly aware of the ‘myriad limitations associated with humans studying other human lives’ (O’Reilly 2012, 17).

Ethnography usually takes the form of prolonged engagement with groups (Markkanen 2018). Long-term ethnography is the preferred approach to understanding mobilities and mobile lives. We propose here an alternative approach based on short-term ethnography (Pink and Morgan 2013) as a way to shift our reflections on mobilities of vulnerable groups beyond casual observation, stereotypical views, and popular images of migration. In the case of this study, this is an approach that also paints a far more complex picture of how to engage women, and subvert patriarchy, when talking about individual and group experiences of migration. Short-term ethnography offers researchers a unique opportunity to gain insight into ways of managing sensitively and effectively access and rapport, and how to find and take advantage of safe spaces in which to talk to women.

We are concerned here with, on the one hand, the potential of ethnography in engaging with mobilities and the experiences of vulnerable or marginalised groups, and, on the other hand, with participation in, and observation of, the lives of people that can profitably go beyond the boundaries and specific configurations in the research process, especially those that enhance voice and the inclusion of underrepresented groups (i.e. women). In this study, we are primarily engaged as participant observers and researchers. However, one of us (the second author of this paper) has also been closely involved as participant in and observer of the lives of Roma people and Roma communities for over 20 years as a woman and co-director of a community interest company.

Before turning to ethnographic studies on Roma people and our discussion of short-term ethnography, we first review collaborative and participatory approaches with vulnerable groups in research.

Working with vulnerable or marginalised groups in research

‘Vulnerable' and ‘vulnerability’ are terms that are widely used and understood in health and social care settings to denote patients or service users whose capacity to maintain independent living or to make autonomous decisions are compromised (for example, as a result of physical or mental illness or disability). ‘Vulnerability’ and ‘vulnerable groups’ have also been adopted in research ethics and governance to denote participants who require enhanced ethical procedures to ensure they are protected in research (from harm, from risk) and understand the meaning of consent as well as how to withdraw from participation.

Some people may be ‘vulnerable’ as a result of ‘innate’ challenges or impairments (see Larkin 2009) or as a consequence of social or structural issues such as poverty, homelessness and so on that mean they are socially excluded or marginalised. In many respects, from the point of view of those defined as ‘vulnerable’, their ‘vulnerability’ is neither fixed nor necessarily dependent on objective assessments or measures, rather it depends on their own perceptions of their own condition (or circumstances) at any given time and this can be subject to change. As Larkin (2009, 1) acknowledges, the meaning of ‘vulnerability’ ‘varies according to the context in which it is used’, and thus a precise definition remains elusive. It is for these reasons that the concept of vulnerability is a mutable and often contested terms in research, even though ethically and for research governance purposes the concept remains valid and extant; regardless of the terminology or language used, measures need to be in place to protect those who participate in research – from exploitation, risk and/or harm – as a consequence of taking part and of research processes themselves (research design or methodologies, for example, that may or may not be suitable or appropriate). Such measures are also needed to protect researchers and their institutions, and it is important to recognise that some people are more at risk of harm or exploitation than others.

The need to identify and support ‘vulnerable’ participants in research with clear regard to robust ethical policies and procedures, also helps researchers and participants alike recognise and understand the processes by which some people in society are discriminated against or stigmatised as ‘other’, as well as the effects of these prejudices on and in their lives. In short, acknowledging that some people or groups of people are ‘vulnerable’ helps us to understand the ways in which they are marginalised in and from ‘mainstream’ society. While, as Steel argues (2001, 2) some groups may have very different interpretations of marginalisation, including their own – he cites the example of Travellers in the UK who may consider the NHS as ‘marginal to their culture’ – it is also clear how and why some groups in society become ‘outsiders’ or ‘othered’ and experience the adverse consequences of this. A 2019 report by the UK charity Friends, Families and Travellers (Sweeney and Worrall 2019) for example, found that nomadic Gypsies and Travellers in the UK were turned away from GP surgeries for registration and treatment, despite NHS guidance that treatment should be offered even when people have no fixed address or proof of ID.

We can also then understand how vulnerable or marginalised groups will require methods of inquiry in research that acknowledge these issues and are sufficiently ‘bottom up’ and flexible to enhance inclusion and participant ‘voice’ (see Aldridge 2015). As participants in research, Roma people require such consideration, not just because of their geographical as well as social, cultural and political segregation from society where this occurs, but also because, as is widely recognised, they experience the effects of historical and contemporaneous discrimination as a ‘unique form’ of prejudice (Kende, Hadarics, and Lášticová 2017, 12; see also Tileagă 2015). For Roma women, these prejudices are enhanced by a legacy of patriarchy within their own culture as well as outside of it (Neaga 2016); as such, they also require methodological approaches in research that facilitate inclusion on their own terms and in ways that delve beneath the surface of their lives and experiences as part of an ‘outsider/othered’ community and culture.

The use of short-term ethnography in this project with Roma women prioritised their agentic voices in research contexts that enabled them to tell their stories as Roma women and as mothers in their own right; that is, outside the boundaries of systemic and culturally prescribed marginalisation and vulnerability.

Ethnography with Roma people

We, as many other researchers interested in real-world problems, also turned to ethnography for the flexibility that it offered. We did not need to be convinced of the value of ethnography in researching real-world problems – we were in agreement with the ethos of ethnography as a social science research method.

Ethnographers attempt to understand the realities of a particular cultural group through deep, persistent and prolonged engagement within a natural setting. Credible ethnographic studies require researchers to go below the surface, and break through the pleasantries, and this requires a high level of access as well as the ability to build trust and rapport (O’Leary 2005, 158)

We were also aware that, as Hammersley and Atkinson suggest, ‘the boundaries around ethnography are necessarily unclear’ (1995, 2) and constrained by time, people, and context.

Translocal Roma mobilities in Europe are now well documented (see, inter alia, Grill 2012, 2017; Tileagă, Aldridge, Lumsden, et al. 2019). It is also well-documented that the transnational lives of Roma pose challenges for researchers as well as politicians in many European countries. What Markkanen (2018) calls ‘sensitive, in-depth, long-term ethnography’ (90) may be indeed appropriate across a variety of contexts and locales. However, it may not suit specific research goals especially those related to eliciting underrepresented voices and experiences.

We started our own incursion into the question of intra-European Roma migration with a keen awareness of some of the methodological challenges of researching ‘the simultaneity of transnational flows and their effects on those who stay behind as well as those who move’ (Levitt and Jaworsky 2007, 143). We were aware that ‘spatial scales, the cultural-historical particularity of places, and the global nature of what flows through them produce different kinds of transnational social fields, or arenas with different clusters of transnational activities’ (Levitt and Jaworsky 2007, 144). However, limited funding did not permit a transnational approach.

We wanted, nevertheless, to ensure ‘embodied encounters’ (Markkanen 2018) with Roma families by working in situ, but, more importantly, we wanted to give everyone in those families a voice. For us, the study of the whole community configuration, as Benedict ([1935] 1961) would have perhaps put it, was paramount and near impossible to achieve when one of the subgroups (women) were excluded from participation by virtue of operative cultural and gender norms evident in some Roma communities.

Okely (2017) documents how, in traditional ethnographic long-term fieldwork, both access and navigating the setting is accomplished with the help of community ‘intermediaries’. The position of the researcher is that of the ‘stranger’ who inserts herself and lives with a community for a particular period of time.²

I was able to live for periods amounting to ten months in a caravan on four Gypsy camps and two months in nearby lodgings in south-east England between 1970 and 1972. In addition, I spent several weeks in northern England visiting Gypsies, many of whom lived in horse-drawn waggons. The main period of fieldwork was supplemented in 1974–5 by several months living near the camps, being visited by or visiting Travellers. (Okely 1983, 39)

Okely’s ethnographic work with Gypsies drew upon what she called ‘political and ethical loyalty’, paralleled by an ethical attitude of not ‘betraying’ the community (including not disclosing the actual location of the community itself). Okely’s ethical and political stance was perhaps unique at the time as most of the early anthropological work on Gypsies in Britain was ‘revealing not for ‘facts,’ but insights into outsiders’ stereotypes and racial profiling' (Okely 2017, 66).

Short-term ethnography

We were familiar, albeit not in full agreement, with the contention that in postsocialist countries, ‘the western model of anthropological fieldwork is still not a common way of doing research’ (Budilová and Jakoubek 2009, 7). We were less concerned with the debates in the ethnographic literature concerning eastern Europe. Our concern was with how to resolve some of the challenges and dilemmas that come from the study of cultures, and people, as ‘articulated wholes’ (Benedict [1935] 1961, 34).

How to manage, and think differently and sensitively about access and rapport, and inclusiveness, was our central concern – a matter of repeated and ‘intensive excursions’ into people’s lives, what Pink and Morgan’s (2013) call short-term ethnography. We explore some of the challenges that such repeated excursions produce. We argue that short-term ethnographies are not primarily characterised by their temporal nature. They are rather a route to producing alternative ways of knowing about people and their lives and practices that take into account the various normative and subjective ramifications made relevant by each encounter with people. Short-term or closer variant ethnographies³ have been used successfully in health research contexts (Cruz and Higginbottom 2013), visual (Pink 2007) and sensory (Pink 2009) ethnography as well as interactive systems design (Millen 2000) and engineering (Pink, Dainty, and Tutt 2013).

As insights from studies conducted by Pink and associates have shown, short-term ethnographies reveal novel anthropological ways of seeing. Ethnographic practice of this type evolves through repeated dialogues with cultural times and spaces, and assumptions, of people and communities in specific circumstances and historical periods, and as revealed in the ‘practice of everyday life’ (O’Reilly 2012, 6). Short-term ethnographies explore local, relational, dynamic phenomena, and more defined elements of cultural practices. Unlike traditional ethnography where one may not have prior experience of the setting, people, and/or context, short-term ethnographies rely on accumulated knowledge from academic and/or cultural sources, and prior personal experiences with the setting, people and context.

Short-term ethnography with Roma people opens up spaces for learning about people ‘from their own perspective’ (O’Reilly 2012, 120). Like long-term ethnography, short-term ethnography accommodates different styles of embodied practice but offers more flexibility regarding access and rapport. The advantage of short-term ethnography is that it allows researchers to unpack collaboratively and respond, promptly and directly, in situ, to the operative assumptions in communities (e.g. patriarchy) due to specific or historical circumstances. Our own repeated engagements with community gatekeepers and intermediaries enabled us to successfully push for women’s greater inclusion. Short-term ethnography gave us the opportunity to find safe spaces in which to talk to women who may not have otherwise found a safe space to talk freely. It also gave insights into ways to maximise access and inclusiveness by emphasising alternative aspects of social identity that subverted the local patriarchal order.

Designing research studies from the perspective of collaborative and participatory approaches to research practice (cf. Aldridge 2015), enhances the after-effect of these intensive excursions into the lives of people. In the context of mobilities, short-term ethnographies are good ways of combatting existing stereotypes and misconceptions about the Roma in a public and media environment where ‘we find an array of contradictory labels and definitions employed in the media and popular speech, as well as in legal documents and political resolutions’ (Matras 2014, 20). There are also, and perhaps more important, consequences for challenging stereotypes and misconceptions about the role of Roma women in their own communities. We argue here that there might be specific advantages of short-term ethnographies. This type of ethnography has the potential to become a distinguishing characteristic of emancipatory and participatory research projects especially those projects aimed at understanding women’s experiences in patriarchal contexts more generally, and Roma women specifically.

Fieldwork

In this article, we draw upon our experiences of engaging Roma women from a study on active citizenship and Roma migration post-Brexit (Tileagă, Aldridge, and Popoviciu 2019). The project was funded for one year by the Higher Education Funding Council for England (HEFCE) in early 2018. The participants of the study were 45 people from seven Roma communities: 39 adults (aged 19–65) and six young people (aged 13–18); 28 men and 17 women. The Roma communities where the fieldwork took place were: Tâmasda, Ciumeghiu, Batâr, Diosig, Tinca, Crasna and Ineu, situated in Bihor and Salaj counties, in North-Western Romania. We did not study these communities as spatially bounded units – our ethnographic incursion within and between sites was guided by Roma people themselves, especially with the involvement of or on the advice of community gatekeepers and leaders.

Participants were selected from among community members based on their experience with the process of migration to the UK. We entered the setting with a set of open-ended research questions in mind and several data collection methods. Community-led participatory reflection groups, a local cultural event and arts and crafts activities were organised with the aim of exploring experiences of migration and understanding some of the drivers of migration to the UK.⁴

\We found that the main reasons reported for choosing the UK as a target country for migration were: (a) employment opportunities, (b) a superior educational system and healthcare, (c) social benefits and welfare support, and (d) good quality infrastructure. The general perception among participants was that informal social networks made the migration process manageable by reducing the psychological as well as the material costs and the risks involved in moving to another country. Community members reported relatively high levels of satisfaction with their choice of migrating to the UK, even if our broader analysis of the overall context of migration suggested that the overwhelmingly positive perception of Roma migrants was sometimes offset by, on the one hand, the negative treatment of Roma migrants’ by non-Roma authorities, neighbours, colleagues or employers, as well as, on the other hand, poor working conditions, unfair pay and lack of protection from exploitation (for the full context, findings and analysis of drivers of Roma migration to the UK, see Tileagă, Aldridge, and Popoviciu 2019).

We discuss here the opportunities, challenges and lessons from the process of engaging Roma women in this study. All research, but especially studies using qualitative methods, depend on the participation and engagement of participants. Generally, Roma people are least likely to be engaged and represented in studies (Boneski et al. 2014; Condon et al. 2019; Jackson et al. 2017). Some of the common barriers to the engagement of people from Roma communities with research teams include geographical segregation, language challenges and a general mistrust of non-Roma people due to a history of discrimination and racist treatment by mainstream society (Condon et al. 2019; Tileagă, Aldridge, Lumsden, et al. 2019; Wanat 2008). More specifically, Roma women are ‘erased’ (Oprea 2005): their voices are absent not just from individual research projects but also from policy national policy strategies for Roma integration related to health, employment, education and housing (Popoviciu and Tileagă 2019). In addition to poverty and social exclusion, Roma women also experience overlapping forms of ethnic, racial and gender discrimination.

One of our aims was to ensure that the voices and experiences of Roma women about the process of migration to the UK were represented. Women did not spontaneously participate in our early group discussions. Male community leaders – who were also (with one exception) the gatekeepers to the communities – considered the topic of migration sensitive and wanted to ensure a degree of control over the record. Where we could go and whom who could speak with was limited. The specific locations for interactions with community members had to be decided beforehand.

Challenges

We encountered two major challenges related to (a) recruitment of participants and (b) engagement with the research process.

Our first challenge was to determine from which communities to recruit participants. A major part in this decision consisted of pragmatic factors such as which communities had Roma people that migrated to the UK. In order to gain access to participants, we contacted Roma community leaders that were already familiar with the work conducted by Ruhama Foundation.⁵ With the exception of one community, who had a woman leader⁶, all other communities had men as leaders. The importance of gatekeepers in gaining access to marginalised and socially disadvantaged groups is well recognised in the literature (Boneski et al. 2014; Condon et al. 2019; Wanat 2008). However, community leaders and other gatekeepers often make paternalistic, and at times patriarchal judgements about who is best suited to participate in research (Boneski et al. 2014; Condon et al. 2019; Gheorghe 2016; Oprea 2012). In this study, with the exception of the Roma community which had a woman leader, all of our gatekeepers, which were all men and community leaders, initially helped us recruit a majority of Roma men. Due to our desire of including as many women as possible in our sample, the research team specifically asked each Roma leader to invite more women into the study group. Although, we planned to conduct all aspects of the research with the involvement of community members, rather than ‘about’ or ‘for’ community members (see Aldridge 2015), in this particular situation, the inclusion of Roma women arose from an explicit need expressed by the researchers, rather than by the people themselves. Even with this approach, and with a continuous effort at negotiating the inclusion of more women in the study, the final sample still had a majority of men (28 men, and 17 women) as participants.

After the groups were selected, prior to our data collection visits, we met with community leaders to learn about each Roma community’s history, culture, religion, traditions, and group organisation. We also asked community leaders to help us shape the questions we were asking, and to help us understand the specific cultural and social realities of the people from each Roma community. Community leaders have stressed the importance of creating a safe place for people to share experiences. Together with community members we organised all activities in Roma community venues, using classrooms, community centres or the personal homes of community members. The views of Roma people were sought, not only on migration issues, but also on the materials used during activities, data collection, and the dissemination of study findings. All gatekeepers accompanied the research team to all project activities.

As Condon et al. (2019) observed, in groups that are set apart geographically, socially and ethnically, gatekeepers can hold disproportionate influence and power over people’s engagement in activities organised in partnership with researchers that are not part of the community. Significantly, in our study, community leaders were influential in deciding who had priority in sharing their migration story. Community leaders often named the people whose migration experience had, in their view, the most relevance for the researchers’ questions. In some of the participatory reflection groups community leaders even signalled turn taking, either with hand or head gestures or verbal cues. Our goal was to facilitate the democratic expression of views on migration whilst striking a balance between managing the explicit pressure from male community leaders and being as inclusive as possible of women’s voices. With the exception of the woman leader who involved and encouraged Roma women to share their migration stories, no other gatekeeper named or pointed to the women in the groups in order to signal their turn in engaging in the conversation.

Our second challenge was thus how to involve Roma women in the proposed research activities. One major concern that we had was the ways in which we could challenge the patriarchal cultural and gender norms without risking the dissolution of our previously established rapport with community gatekeepers. Although, according to our findings, women had very similar experiences of migration to men, and they were often more knowledgeable about topics such as the educational system, healthcare and the informal support networks available in the UK for Roma migrants (Tileagă, Aldridge, Lumsden, et al. 2019), they were less likely to contribute to the discussions, unless specifically asked by a sympathetic gatekeeper or researchers themselves. We used the flexibility offered by this style of ethnography in mixed gender group discussions to elicit women’s views by asking them direct questions about their experiences of mobility. The main interest in our ethnographic investigation was to understand the perspectives of all study participants. We opted for engaged listening and facilitation which involved directing or redirecting questions at specific people, in this case, women. The act of questioning positioned us as outsiders, but as outsiders, we were able to challenge in situ the taken for granted routine conversational practices within the community that mean that women generally remaining silent. According to the European Roma Rights Centre (2000) in many Roma communities, there operates an entrenched hierarchy of power in which Roma women have very little control over the decisions taken by community leaders. In many cases, women’s roles are restricted to issues directly affecting their children and household, such as their children’s education and housework.

This state of affairs brought a couple of problems into view. First, the voices of those who had less power were inadvertently silenced, which ran counter to the objectives of the study. As Oprea (2004, 2012) argues, when studying traditional Roma communities, researchers that aim mainly to observe and not actively collect the perspectives of women will ultimately silence the voices of Roma women and other feminist allies who seek to address the patriarchal practices within these communities. According to feminist Roma authors, a traditional Roma community is described by a highly stratified patriarchal structure in which men hold the power, moral authority, social privilege, and predominate in roles of leadership (Gheorghe 2016; Vincze 2013). In groups with a history of disentitlement of women from social power, and entrenched unequal gender relations, it is important to open up safe spaces for equal representation of all views by prioritising the study of women’s experiences. It is important to note, however, that in line with Okely’s (1991) pertinent observation, we consider Roma women to be active subjects, rather than submissive objects. We believe that with the help of ethnography, points of resistance and ways of subverting patriarchy can be revealed, even if only in fragmented moments of defiance.

Second, as Rowa-Dewar et al. (2008) noted, perceptions of researchers’ attempt to resist routine patriarchal practices within Roma groups can lead to the erosion and sometimes termination of the ongoing research relationship. In our study, the gender and ethnic power differential became evident and influenced the (lack of) participation of Roma women. Adding to the complexity of the problem was the fact that in studies on Roma migration, there is a predominance of male-dominated research (Pessar and Mahler 2003; Oprea 2012; Sanchez et al. 2013), which does not take into consideration the ways in which migratory decisions and family reorganisations are influenced by the views and decisions expressed and played out by Roma women. Even in activist circles, during panels and conferences on Roma issues, there is often a lack of Roma women panellists, despite a considerable number of Roma feminist initiatives across Europe (Gheorghe 2016; Oprea 2004, 2012; Vincze 2013).

Specific measures to engage Roma women in research activities have been suggested. One solution is to organise groups separately for women. For example, other projects which focused on dialogic feminism, created spaces for Roma women to share various experiences from their communities (Sanchez et al. 2013). According to the participants of Sanchez et al.’s (2013) project, its success in collecting shared experiences was in large part due to the creation of a space where women could talk about their perspectives ‘without their husbands’ present (Sanchez et al. 2013, 233). On the one hand, a benefit of this approach is that it offers a space for women to talk openly about their experiences in the presence of other women. On the other hand, it suggests that women need to be separated from their male-dominated communities in order to have a voice (although that needn’t always be the case as we found in this study).

Also, this type of solution to what is otherwise a complex issue, presents a few other problems. Firstly, it involves additional project costs and time pressures, as it necessitates either additional trips to Roma communities for separate meetings with Roma women, or the organising of a common dialogic space for women from different communities. Prior research has shown that engaging the most vulnerable groups who are experiencing overlapping and interdependent systems of prejudice and discrimination demand high costs in both time and effort (Condon et al. 2019; Thompson and Philips 2007). In this research project, we returned to two communities in order to have discussions only with the women. Due to financial and time restraints, we were not able to plan out such discussions in all communities. Secondly, although separate activities aimed specifically at remedying the disproportionate representation of male and female perspectives in Roma communities are valuable, they do not resolve the broader problem of the exclusion of Roma women’s voices from mixed gender groups. Ideally, Roma women need to be engaged in existing programmes and community activities, in order to make clear gains in challenging the hierarchical and patriarchal structure.

Voice and breaking down barriers to engagement

In this paper, we propose the introduction of meaningful categorisations that, in our experience, are effective in breaking down some of the barriers to engagement for Roma women. Gender and identity categories are, generally, a rich conversational resource (e.g. Stokoe and Smithson 2001).⁷ We show here how gender and identity categories can be used productively in the research process to overcome barriers to participation.

When we framed our research goals as an exploration of migration experiences of ‘Roma’ people, gatekeepers tended to recommend mostly male participants. When, during our initial discussions with community leaders and gatekeepers, we explicitly introduced the gender category of ‘women’, we partially solved the problem of including in our sample of participants both men and women (although, a majority of men remained). But, at that point, we also faced a different challenge: how to engage the women that were present in the discussions and project activities. We further narrowed our category of ‘women’ to ‘mothers’.⁸ We knew from prior studies on engaging Roma women that, in mixed gender groups, Roma women who contributed to discussions about their experiences and knowledge usually spoke from their positions as ‘mothers’ (Godlewska-Goska 2016; Melgar et al. 2011). Even prior to us introducing the category ‘mother’ as a means of facilitating the involvement of women, women themselves often, at the beginning of their turn, would preface their view by referring to themselves as ‘mothers’. Some examples include: ‘As I mother, I saw that … ’, ‘we know, as mothers … ’, ‘Being a mother, I thought that … ’. According to the majority of ethnographic studies conducted in traditional Roma communities, motherhood is seen as one of the main social roles for Roma women, and the birth of children can be, at least theoretically, a way to increase the status of women within the group (Godlewska-Goska 2016; Janky 2005; Marushiakova and Popov 2001; Vincze 2013).

For example, after a divorce – which generally takes place within the group and not in a court of law – women who have children typically receive full custody and also get to keep their house, unless their husbands or other people from the community can prove that they have broken some of the moral codes of the community, such as having premarital sex or dressing immodestly by community standards (Godlewska-Goska 2016). In traditional Roma communities, family structures are monogamous, patriarchal and typically based on a tight structure based on kinship (Godlewska-Goska 2016; Janky 2005; Oprea 2012; Vincze 2013). Nonetheless, women who had more than three children and grandchildren, and had been married to only one man had the right to participate, although not generally lead, meeting of the elders of the community. They can also have a say in community matters and give advice to younger men. In exceptional cases, Roma women can become local leaders. Such exceptions include: (a) women who gained a university degree and who in time, become invaluable in solving community problems such as access to electricity, clean water or representing the community in the city council, (b) widowed elderly women who were married to former community leaders, or the daughters of former community leaders, whose authority, is in a way an extension of the authority that their husband or father had.

Our aim was to seek accounts of migration from both men and women from Roma communities. Once we realised that women would be in great part excluded due to the predominantly patriarchal culture within these communities, we had no choice but to find alternative ways to include women. The only way we found that we were able to get around this was to include as many women who describe themselves as ‘mothers’ as possible. We realise that this was not ideal, as it created a context where women who were not mothers were more likely to be silenced. This seemed the best way to balance our aims of collecting data that was inclusive of women, while also working successfully with communities that are often closed off to non-Roma researchers, whose views on equality and inclusion are very different. We also tried to find an acceptable way for the Roma male gatekeepers to have the women engage with our activities, without jeopardising our relationship with the community.

However, the discursive context within which women voiced their experiences as mothers continued to be one where the hierarchy between genders and the domestic patriarchal pattern of motherhood remained unchallenged. The use of identity categories by Roma groups usually reveals a highly gendered role hierarchy. As Casey (2014) notes, Gypsy women tended to use categories that either emphasised their role in the community (‘caravan wives’) or individualise them in relation to assumptions around Gypsy culture and morality (‘decent girls’) (see also Okely 1983 on the myriad of beliefs and practices that constitute gender among Gypsy women).

Recategorising Roma women as ‘mothers’ did not come without its problems. For example, asking for the opinions of ‘mothers’ rather than ‘women’, silences the voices of all Roma who aren’t mothers, either by choice or circumstance. In traditional communities, not having children can result in the public ridicule and humiliation of women, who are said to fail in performing a basic duty of their gender (Gheorghe 2016; Godlewska-Goska 2016; Vincze 2013). Research that includes the voices of mothers at the expense of the voices of women who do not fall into that category, contributes to the unequal distribution of power dynamics between genders and those within subcategories of womanhood. Moreover, the body of academic literature that looks specifically at Roma women usually places emphasis on issues that concern Roma mothers, rather than those related to the more inclusive and heterogeneous category of Roma women (see for example Gheorghe's 2016 feminist critique about the representation of Roma women in academic research).

As Okely remarks, ‘throughout Europe a Roma woman is presented as sensual, sexually provocative, and enticing’ (Okely 1983, 201). This representation is at odds with what the Gypsy patriarchy thinks of women and the role of women in the community: ‘far from being a flighty seductress, the Traveller [Gypsy] woman is burdened with many domestic duties’ (Okely 1983, 203). Also, due to still widespread anti-Roma discrimination and racism, in Roma communities, there is often a rhetoric of unity, where strong ties to the community and family somewhat compensates for a lack of ties with the broader community. For many women, this can serve as an additional impediment to speaking and acting against the sexism in their communities, and consequently, perpetuating gender inequality (Oprea 2004, 2012; Vincze 2013).

Conclusion

In this paper, we explored the potential of short-term ethnography in a collaborative and participatory research project on experiences with and perceptions of migration from fieldwork with Roma groups in north-western Transylvania. We discussed particularly the challenges we encountered in engaging Roma women in terms of existing patriarchal norms evident in some Roma communities. There are two immediate advantages of short-term ethnography with Roma people, especially Roma women.

First, short-term ethnography opens up spaces for managing access and rapport by finding and taking advantage of safe spaces in which to talk to women about their experiences of mobility. It can also be a platform to challenge intra- and extra-community power relations. As Budilová and Jakoubek (2009) argue ‘we are used to supposing that wherever the Roma live in a spatially defined locality (a settlement, a neighbourhood, or a town) they inevitably form a local community’ (10). The notion of ‘local Roma community’, although perhaps useful in some contexts, can obscure power relations especially those related to gender. The first step towards equality of perspectives on mobilities involves including the experiences of Roma women in research studies. When doing so, the recognition of privilege is important in order to challenge systematic racism and sexism. In the context of research with Roma people, both Roma and non-Roma researchers need to evaluate and reflect on the ways in which their privilege can lead them to draw boundaries around whose voices and experiences are included and whose are excluded. Care needs to be taken to include and then involve Roma women in such a way that the empowerment of some subcategories of people, such as Roma mothers, do not come at the price of the disempowerment of Roma women in general.

Second, short-term ethnography offers insights into and opportunities for the successful reframing of identity/gender categories, in this case by researchers themselves, with powerful effects and consequences for participation and engagement. Repeated incursions and engagements with community members and gatekeepers permit identity recategorisations that, in turn, shape the research process. We offered here an example of how alternative categorisations of the kind we discussed here can be introduced successfully by those that are not part of the studied group without altering essentially or fragmenting already established relationships with gatekeepers and group members. Working with exclusive identity categories and set assumptions about participant groups can present many disadvantages at any stage in the research process not least the risk of excluding those potential participants that do not fit neatly those categories. We hope that researchers designing emancipatory and collaborative studies on migration experiences will perhaps start to adopt the standpoint of short-term ethnography to explore the in situ expressions and consequences of context-dependent constructions of identity that give a voice to people that would otherwise remain silent.

A shift in categorisation from ‘women’ to ‘mothers’ ensured that women were, first, included in mixed gender group discussions about mobilities, and, second, they could express their own views as ‘mothers’ (and women) in a safe space when directly invited to do so by women researchers. By managing both access and rapport, as outsiders and researchers, we were thus able, albeit temporarily, to chip away at the taken for granted power relations that stipulate that women are meant to remain silent. The ethnographic encounter proved to be the best way to balance our aims of collecting data that enhanced women’s participation, while also working successfully with communities that are often closed off to outsiders.

We propose that short-term ethnographies are a meaningful addition to fulfilling the goals of emancipatory, participatory, and inclusive approaches by addressing head-on the numerous tensions related to access, rapport, inclusion/exclusion in the research process. Positive collaborative partnerships with community gatekeepers may lead to a continuing transformation of traditional views related to what works when engaging research participants. We suggest that ethnographically driven research forms an important means to move beyond simplistic understandings of Roma migration. As other recent studies are showing (cf. Chase et al. 2020) methodological innovations in ethnographically driven research on migration can promote not only research reflexivity but also confer visibility to people that would have been otherwise hidden either by the research process itself or social arrangements that keep people in their place and reproduce the status quo.

We have shown how short-term ethnography can be used successfully in this collaborative and participatory research project to explore the perspectives of Roma women. Challenging patriarchy is difficult for many Roma women because it involves critiquing and rejecting historical and engrained hierarchical relationships that are present in their families and communities. In the words of Roma activist Sandra Selimovic (apud Meaker 2018): ‘We are not just housewives, whores or beggars’ … ‘We are businesswomen. We are very tough, strong women and role models’. Having successfully managed to include women in our research project we have, perhaps, also managed to disrupt, at least provisionally, a local ethics of power and gender that refused women participation. As one woman from our study remarked: ‘It is really good for us mothers also to have this opportunity to share our experiences’. Gender and power relations mark ethnic and inter-ethnic spaces (Schneeweis 2016). At a time when Roma women are mounting challenges to their public and private marginalisation (Meaker 2018), giving Roma women a voice is a first step towards overcoming some of the individual, group and societal obstacles that refuse women a voice.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This project has received funding from the Higher Education Funding Council for England (HEFCE) in early 2018

Notes

1 See also Xiang and Toyota (2013) for descriptions of different kinds of ethnographies as new modes of exploring transnational mobilities.

2 There were many risks and shortcomings associated with doing anthropological work with Gypsy communities in Britain and elsewhere: having to conceal the true purpose of research, trying to get financial support from government with little sympathy for nomads (see Okely’s numerous accounts of her original work with Gypsy communities in Britain).

3 There are similarities between Pink and Morgan’s notion and what other researchers call ‘focused’ (see, for example, Knoblauch 2005; Wall 2015) or ‘rapid’ (Millen 2000) ethnography.

4 Ethical approval was granted by the Loughborough University Ethics Sub-Committee and consent was received from all participants. This research project was designed and conducted in line with the Romanian law no 206/2004, concerning the Good Practice in Scientific Research, Technological development, the Romanian Law no 677/2001, notification no 9088, concerning the protection of people and the processing of personal data and the free circulation of personal data, and Innovation, and the British Psychological Society’s Code of Human Research Ethics (2014). One ethnical difficulty which arose was that not all participants were able to read or write. In each group, we asked one representative to read out aloud all of the written information handed out, including the information on the consent forms. When participants were asked to consent in writing, in order to prevent any embarrassment, all names were written in advance, and participants were told that, if they agreed with the information that was read before, they could make ‘a pen mark’ on the signature line. Examples of possible choices for ‘pen marks’ were drawn on a whiteboard or flipchart.

5 Ruhama Foundation is an international NGO with headquarters in Romania. Ruhama has over 20 years of experience of working closely with disadvantaged Roma communities and influencing policymaking on Roma issues in Romania and beyond. Ruhama Foundation was established in 1996 with the aim of improving the quality of life of people and local communities at risk of social exclusion or marginalization, especially Roma people living in Romania. Its interventions are at the grassroots level. Ruhama works primarily through partnerships with public authorities, private organizations, universities, and the civil society.

6 One of the seven communities we visited had a woman leader. This situation was exceptional, as far as Roma communities in north-western Transylvania were concerned. She was the only child of a former community leader – a man. She used to help her father with solving various social work and other administrative community problems. The community became accustomed to her having an important administrative and social role, and after her father’s passing, she became a moral leader in the community and a gatekeeper.

7 Self- and other-categorizations play an important part in the construction and negotiation of identity categories (Stokoe and Edwards 2009; Edwards 1991; Antaki and Widdicombe 1998; Merino and Tileagă 2011).

8 Out of 17 women participants, 14 described themselves as mothers.