“I want to speak like the other people”: second language learning as a virtuous spiral for migrant women?

This article contributes to scholarship on migrant women’s second language (L2) education in North America and Europe. Questioning reductionist understandings of the relationship between female migrants, their receiving communities and L2 education, the authors consider existing literature as well as their own qualitative work to investigate the challenges, opportunities and agency of migrant women. Weaving together and thematically presenting previous scholarship and qualitative data from interviews, participant observations and classroom recordings from a mixed-gender L2 adult migrant classroom in Austria and an all-women L2 migrant classroom in the United States, they trouble conceptualisations which position women primarily as passive recipients of education and in need of emancipation, while simultaneously elevating communities as agentic providers of these. Specifically, the authors emphasise that (1) L2 proficiency is not a guarantee for migrant women’s social inclusion or socioeconomic advancement; (2) migrant women’s complex challenges and agency need to be recognised and addressed; and (3) all involved in L2 education of migrant women do well to become learners of their own experiences of oppression, including their complicity in it.

to read and write for the first time in her life, hoping to bridge the linguistic (and likely also social) divide she is experiencing between herself and "the other people", the locals (Christiane, interview, November 2016).
The list of different kinds of support systems which Hadiah and her family have been able to tap in Austria is long, ranging from language courses and accommodation to special needs services and social networks. The support the family has been receivjng is the result of a collaborative humanitarian effort, mostly carried out by local volunteers, who have built a high-quality refugee aid centre which offers language and literacy courses, tutoring, housing support, clothing and household supplies, and numerous locally relevant activities such a biking classes. This support system is, quite literally, vital, and has enabled Hadiah's and other families to live a life in relative peace and stability.
When we consider refugee families' post-migration life situations in welcoming communities, it is tempting to describe them in a binary of givers and receivers, with refugees on the receiving end and local communities on the giving end. Without diminishing the important work of volunteers, this article aims to deconstruct this binary, especially in the context of migrant women's L2 education. We take a step back to pose two research questions: • What does L2 learning have to offer migrant women, and is L2 proficiency automatically an advantage for them? • Challenges for migrant women L2 learners: What are the obstacles and desires of learning to "speak like the other people"?
We, the authors of this article, explore these questions by weaving together and thematically presenting existing literature and our own qualitative research. Before we do so, we briefly outline the academic and socio-political context that warrants them.

Why focus on migrant women as L2 learners?
As part of the refugee movement that was triggered by war, persecution and economic depression in several Middle Eastern, Southeast Asian and North African countries, a small percentage of displaced people, a recorded 1,046,599, arrived in Europe in 2015 (IOM 2016).
Of these, 88,912 sought asylum in Austria that year, making Austria the fourth-largest receiving country in the EU (BMI 2015). The large number of male refugees, who comprised 72.33% of asylum seekers in 2015, has resulted in migrant men receiving a considerable amount of attention in research and, albeit often negative coverage, in media and public discourse. While there is definitely a necessity to challenge xenophobic discourses around male migrants, we also believe it is time to focus on migrant women. Although there is a growing awareness of migrant women's needs and assets beyond the context of family reunification, for instance as members of the labour market (Kofman 2003), research is still needed which investigates and improves the experience of migrant women as they become language learners in receiving countries.
Two main arguments that illustrate the precarious situation of migrant L2 learners fuel our focus on this population. First, many receiving countries have responded to migration by developing, enforcing or changing language (learning) policies. As in Austria, governments and other administrative entities may require a particular level of language proficiency e.g. for citizenship applications, most university programmes, and many administrative processes involving non-EU nationals (e.g. application for a residence permit). This is especially problematic when such policies coexist with a dearth of L2 courses, as has been the case in Austria. The conditional combination of these kinds of requirements and barriers has the effect of trapping migrants in precarious and marginalised circumstances which can become normal and perpetual.
3 Second, once arrived in a new country, migrants often not only have to but indeed want to learn the socially dominant language as soon as possible. Besides the ability to communicate, they might hope to acquire, among other things, social integration and equality, financial stability, and professional advancement through the language which (ostensibly) holds power and privilege. These expectations are reinforced by an environment which fails to recognise discriminating (e.g. racist and sexist) policies and practices as reasons for exclusion, marginalisation and downward mobility, and often do not hold up. For example, data from Austria compiled by the Organisation for Economic Co-operation and Development (OECD) show that disadvantages in the labour market exist for migrants (OECD 2017) and are especially pronounced for migrant women from low-income countries (OECD 2012).
Although these data do not distinguish between L2 learners and German-speaking migrants, it can be assumed that many foreign-born migrants from low-income countries do not have the German proficiency required for many administrative, professional and social processes.
Generally speaking, many migrants become language learners -whether voluntarily, reluctantly or enthusiastically -and thus often find themselves facing an unanticipated set of challenges, especially in view of the fact that social capital and civil rights are tied to (not always easily accessible) linguistic resources. Therefore, shedding a light on migrants' L2 learning identities and experiences is a critical step towards fostering more welcoming and inclusive societies for incoming migrants.
A plethora of academic and non-academic literature addresses issues of language and gender (for introductions and reviews see for example Holmes andMeyerhoff 2003 or Mills andMullany 2011), which may give the impression that everything has been said and additions to this line of scholarships would be redundant.
As authors of this article, our response to this is threefold. First, the field of second language acquisition (SLA) has long turned a deaf ear to gender issues .
Although, as a field, we have been catching up, there is still a dearth of research examining female language learning in its own right, rather than evaluating it against a male (implied) norm and thus perpetuating female speech and language as deviant from this norm . Second, despite increased interest and scholarship on migrant issues and 4 language learning, migrant women are still underrepresented in academic literature. We contend that this very heterogeneous group deserves attention because of their unique needs and assets. Third, the absence of literature at the intersection of all three -gender, migration and language learning -is striking. This is especially true for feminist scholarship, which has been largely absent from debates on migrant second language (L2) education.

The virtuous spiral of L2 proficiency
When immigrants enter a new country, learning the socially dominant language often becomes a key task, and not without reason. A study by Alastair Ager and Alison Strang, which was commissioned to identify indicators of integration of refugees in the UK, has produced a long list of the benefits of second language (L2) proficiency for refugees (Ager and Strang 2004). Their participants' proficiency in English was related to better chances of finding employment, being accepted into higher education programmes, building more positive social relations with cultural insiders, having a better understanding of the host culture and being more inclined to adopt it, and being healthier, and more satisfied with their professional situations. In short, proficiency in the societal dominant language has been shown to play a big role in refugees' successful integration and their overall well-being (ibid.).
Based on Ager and Strang's work, Linda Morrice notes that, language is a key to opening doors to much wider social benefits. Migrants with better language proficiency enter a virtuous spiral building social, cultural and educational capital, gaining experience in the workplace and generally experiencing better integration outcomes and well-being (Morrice 2016, para 2).
Although the positive effects that L2 proficiency can bring about cannot be denied, identifying L2 proficiency as an entry point into a "virtuous spiral" paints an overly simplistic picture of learning and integration processes. Hans-Jürgen Krumm and Verena Plutzar have noted such tendencies to exaggerate the role of L2 proficiency at the expense of the 5 complexity, heterogeneity and diversity of interacting factors which shape migrants' lives and well-being. Relatedly, they lament a narrowing of the integration discourse in Europe, which has shifted the focus from a holistic promotion of multilingualism and multiculturalism, including maintenance and development of migrants' first cultures and languages, towards more assimilative (linguistic) practices and policies, some of which present second language acquisition as "'guarantee' for successful integration" (Krumm and Plutzar 2008, p. 5).
In the case of migrant women, a group which is of course highly heterogeneous, a simplified equating of integration and language proficiency can surface in combination with a desire to educate in order to emancipate, especially when women emigrate from countries where the status of women and women's rights are (or at least are perceived to be) lower than

Methodology
The two qualitative case-studies we report on in this article are situated in an interpretive paradigm which views reality, including educational environments and research processes, as socially constructed. Thus, we (the authors) understand our contribution to this special issue 2 as being part of and reporting on meaning-making processes involving both the researchers and participants of these case studies and, further, as shaped by the identities and positionalities of all involved. This paper knits together existing literature with findings from our (the authors') own qualitative data collection, presenting these together according to shared themes.
First, we identified and reviewed previous scholarship on migrant women's L2 learning, developing common categories and themes across the literature, which informed our research questions. Then we embarked on evaluating the data we had already collected independently in our own two case studies. Original data for this study were collected at two sites: (1) a volunteer-run refugee aid centre in a small town in Lower Austria; and (2) a family literacy programme for refugee mothers and their children in the southeastern part of the United States. One of us, Nicole Pettitt, collected data at the US site and took the lead on analysing and interpreting the respective data set. The data she presents in this article are the result of an independent study, which was completed in October 2016. The other author of this article, Johanna Ennser-Kananen, had the same responsibilities for the Austrian context; however, her fieldwork continued beyond that date. As a result, this article relies mainly on the data from the US context, with the Austrian data assuming an illustrative and complementary role.
Despite the asynchronicity of our two studies, the connection we make between them is critical in providing an international perspective on an international issue and in deepening data analysis and interpretation processes, for example by drawing attention to the role and weight of contextuality for the participants' experiences and actions.

Littletown (Austria)
The Austrian refugee aid centre in Littletown (a pseudonym) is the result of a spontaneous and needs-based initiative, entirely run by local volunteers, and funded by the city and a church-based charity. Targeted language and literacy courses started in September 2015 and are run by a team of 7-8 instructors, who are, with one exception, retired teachers. A team of tutors offers additional practice and support for small groups or individuals in the "learning 7 café", which was opened as part of the centre during the summer of 2016. In order to receive official language certificates, learners take an exam in a bigger city close to Littletown. The teachers independently plan and implement courses for proficiency levels between A0 and A1 as defined by the Common European Framework of Reference for Languages (CEFR) , leading up to A2. Christiane has taken on a coordinating role within the centre. Her responsibilities 3 range from finding accommodation for refugees and planning courses and events to recruiting and training volunteers. Although language instruction is their main priority, the teachers also act as personal advisors to many of the refugees (e.g. family matters, cultural transition processes, healthcare) and are heavily involved in a variety of activities and services the centre offers. (about 20 pages of data). She also visited the centre prior to collecting data in summer 2016.

Refugee Education Center (United States)
The US research site, Refugee Education Center (REC; a pseudonym), is a family literacy centre for refugee women and their children from infancy up to the age of five. REC is located in a diverse town in the southeastern part of the US which has received refugees from more than 40 different countries since the mid-2000s and where, at the time of data collection, over 60 per cent of the town's population spoke a language other than English at home. REC was set up in conjunction with other non-profit refugee services in the late 2000s, as the surrounding area began to shift from being populated by predominantly White, focal students: three each audio; administrator: one audio); and a review of related documents.

Combining our findings from the two case studies
We organised our original data from both research sites chronologically and coded them manually. The codes were informed by our initial review of the literature as well as recurring topics from the data themselves and included "target language use", "gender norms", "resisting norms" and "making decisions". In a second step, we (the authors) identified connections between those codes and merged them into larger themes such as "future plans/aspirations", "family life and L2 learning", and "agency". None of the stories of our participants fit neatly under one theme, nor did they align with binary categories of "benefits" versus "challenges".
In the following sections, we weave together findings from our own studies with existing literature in the area of L2 education for migrant women. Our intention is to highlight the need for engaged scholarship that understands L2 education as well as anti-oppressive work as necessary parts of integration efforts. We present our findings in ways that reflect this complexity: In the first section, we challenge the overly simplistic assumption of automatic benefits of L2 learning for migrant women. Next, we outline the variety of challenges these women may face in order to critique ideologies which reframe larger societal issues into linguistic ones. To avoid "victim discourses" of migrant women, we close with a section on the agency of migrant women L2 learners.
What does L2 learning have to offer migrant women, and is L2 proficiency automatically an advantage for them?
As prior research has shown, women have related L2 proficiency to escaping from financial hardship (Gal 1978;McDonald 1994), improved self-reported health (Pottie et al. 2008), better job opportunities, and better social relations (Casimiro et al. 2007). For example, b ased on their analysis of women's participation in and beyond a six-month bilingual literacy programme for Punjabi-speaking women in Vancouver, Canada, Alister Cumming and Jaswinder Gill demonstrated the importance and positive effects of literacy and language acquisition, especially for women with main childcare responsibilities and little to no access to first language (L1) speakers. The authors comment that [t]he general value of the kind of instruction provided would appear to be as a necessary "bridging" step from non-participation in the majority society toward more formal kinds of adult language, vocational, or academic education, potentially leading to fuller social participations and personal independence (Cumming and Gill 1991a, p. 20).
In Austria, Christiane reported benefits for migrant women who attend her German L2 courses in Lower Austria. She believes that apart from language skills which are needed for interaction with locals as well as fellow migrants, the course participants benefit from taking a break from difficult family situations when they attend these classes, like in Hadiah's case, and having the opportunity to come into contact with other migrants or refugees. In addition, German skills and course certificates could potentially increase asylum seekers' chances to be granted asylum. Christiane further described the L2 classroom as a safe space for practising new roles and relationships (e.g. engaging in language-focused activities across gender, L1, religious and ethnic boundaries) as well as incentive for women to experiment with new identities and discover opportunities for personal development, which is often triggered by the progress they make as L2 learners (Christiane, personal communication, November 2016).  (Alcorso 1991), often related to limiting them to domestic work or low-paid jobs (Kamalkhani 2001;Kofman 2003). High L2 proficiency is often seen as a way to exit or avoid this process, or at least as a way for migrant women to be more competitive and successful in the job market. Pierrette Hondagneau-Sotelu's (2007) research adds important nuances to this notion.
Based on her study with Latina migrant nannies and cleaners in a wealthy community in the US, she pointed out that, in addition to race and ethnicity, L2 proficiency opened doors to higher-end positions for her participants. However, her work also showed the opposite to be true, for instance when employers sought to emphasise a sense of racial and linguistic segregation and superiority towards their employees and therefore preferred to hire less proficient English speakers. Thus, Hondagneau-Sotelu's work illustrates how racism, sexism and linguistic discrimination interact to undermine the "virtuous spiral" of L2 proficiency and make L2 proficiency an unreliable factor for social advancement for some migrant women (see also Piller and Pavlenko 2007).
In addition, or own research presented here points to the need to develop more 11 awareness and nuanced understandings of the changes in social standing experienced by some migrant women, which may not be tied to work or de-skilling. For instance, Sahra, a woman from South Sudan who attended courses at the Refugee Education Center (REC) in the US, went to lengths to share with her English teacher and classmates her family's status in her home community during a class activity. With language support from her teacher, Sahra described the social structure of her small agricultural home community, explaining that her grandfather had been a chief in her ethnic group, and her father and uncles important leaders, which at least partially contributed to her "value" when it was time for her to get married (e.g., she reported with a smile that her bride price included, "a lotta, lotta cows" Field notes, 13 February 2015). Although she had been married for over 20 years, Sahra carried pictures of herself as a bride in her phone, which she showed her teacher and classmates. Sahra's bridal wear included elaborate dresses, makeup and an intricate hairstyle, which some of her classmates remarked on as they had only seen Sahra in a hijab and no makeup. In response to the teacher's questions, she described the celebration, which lasted several days, and the special dishes served during her wedding. While not extravagant by many US standards, it seemed the wedding events as described by Sahra were lavish in her home community. Now in the US, Sahra, her husband, and their children have greater economic mobility and physical safety than in South Sudan: Sahra no longer fears for her own safety; they live in a large, suburban single family home; and Sahra drives her own minivan. Her husband owns his own business, which Sahra promoted to classmates and teachers by passing out his card and describing his services and "reasonable prices" (Field notes, February 13, 2015). Yet, no matter how much economic success she and her husband achieve, or how proficient in English she becomes, Sahra is living within a very different social structure in the US, and it is unlikely she will ever again experience her former social standing as the direct descendant of the most powerful individuals in her community. found that language skills were one of "multiple issues of insecurity" (Casimiro et al. 2007, p. 56). The women also worried about employment, housing, gender roles and responsibilities, and the hostile socio-political clim ate that made them feel insecure and unwelcome. For instance, the participants reported being harassed, threatened, and associated with terrorism and fanaticism on the grounds of their religion, gender and race. One woman from Iraq summed up her fears in this way: Everywhere you go, you have this constant fear that you'll be attacked or that someone will be racist to you (ibid., p. 65).
In the face of these words and, more generally, the work introduced above, it becomes clear that presenting L2 proficiency as the be-all and end-all of integration not only oversimplifies issues but also ignores the responsibilities of receiving societies to work towards becoming inclusive, safe and just environments for all migrants.
In sum, while several researchers have documented advantages of L2 and bilingual proficiency, it is important to understand their contextuality and fluidity and, in some cases, their limits -in other words their contingency on linguistic, socio-cultural, and political contexts. As L2 educators and scholars, we need to acknowledge the ways in which this unstable nature of L2 benefits challenges simplistic or overblown expectations for migrants, especially women, as L2 learners. Thus, our answers to the question "Is L2 proficiency an advantage for migrant women?" need to be complex and nuanced. While we certainly want to promote L2 learning (and L1 learning, for that matter) that aligns with women's long-term goals and needs, we also need to provide support and opportunities that go beyond linguistic ones. This is especially true when we consider the challenges many migrant women face.
Challenges for migrant women L2 learners: What are the obstacles, desires and pitfalls of learning to "speak like the other people"?
In order to provide accurate descriptions of and effective support for migrant women's integration, it is important to understand the complex challenges many of them face when they become, or want to become, L2 learners. Before even entering a language classroom, obstacles for migrant women can occur around gaining and maintaining access to L2 instruction, which can be due to limited availability of courses or of appropriate (e.g. same-sex) courses, limited family resources, and practical matters like lack of transportation and childcare (Goldstein 1997, Higgins 2010Rida and Milton 2001), but also migrant women's constructed or imposed identities as mothers and homemakers that can be incompatible or difficult to reconcile with those of learners or students (Cumming and Gill 1991b;Pavlenko 2004a, 2004b;Kouritzin 2000).
It is important to note that some educational programmes have worked to reduce these barriers by scheduling classes during times that are more accessible to women and/or arranging ancillary supports such childcare and transportation. In addition, the Refugee However, even in the presence of appropriate course offerings, women can be effectively, explicitly or implicitly, banned from language learning, for example through an imposition of traditional gender roles by their male spouses (Kouritzin 2000) or rigid policies and practices that do not recognise their existing cultural and linguistic knowledge and exclude them from participating in formal education (Blackledge 2001).
Further, in her study in an ESL family literacy programme in the US, Julia Menard-Warwick (2004)  of women and their families. Unearthing such restrictions and creating programmes that respond to them, but also help women overcome them, is as difficult as it is necessary.
Once migrant women have beaten these odds and gained access to L2 courses, they may still find themselves excluded from learning experiences and legitimate discourses. As Bonny Norton's study with female migrants in Canada has illustrated, migrant women may face stereotypes and exclusion from majority language speakers who bar them from entering conversations in the target language and thus limit their legitimacy as L2 users. Such blatant exclusion can negatively affect migrant women's investment, identity options and linguistic progress (Norton 2000). Furthermore, Warriner's (2007) larger study in the same context showed that proficiency does not always translate into economic improvement, independence or upward mobility, even though L2 proficiency is often seen as critical by authorities as well as migrant 16 women themselves. With its goal to integrate learners into the workforce as quickly as possible (see also NIF 2016) or help them obtain their high school diploma in a short time, the adult ESL programme in Warriner's study was not able to provide the support the women needed to access jobs which would give them financial stability and independence. Personal advising fell prey to the teachers' excessive workload, and even successful graduates of the programme struggled to find jobs which paid more than minimum wages. Warriner explains that [t]he women's first-hand experiences demonstrate that proficiency in English does not necessarily confer the social, cultural, economic, or political capital necessary to achieve "substantial citizenship" (Warriner 2007, p. 355), and although many met their short-term goals, "few possibilities for long-term social advancement, economic stability, or educational opportunity" arose from successful course The existing body of research is a reminder to critically examine and address the challenges within receiving communities and societies, not merely the ones in migrant women's backgrounds.
To summarise, migrant women who are or plan to be L2 learners may find themselves facing obstacles every step of the way. Racism, sexism, neoliberalism and other interacting systems of oppression can create dynamics that keep women out of classrooms, keep them from learning, and keep them from turning their linguistic gains into financial ones. Thus, any serious attempt at educating and empowering migrant women needs to also tackle these systems of oppression, as they exist within and beyond the receiving society and migrant communities.

The agency of migrant women L2 learners
Considering the research presented thus far, it would be tempting to reduce migrant women to victims of unjust societies and communities who are being discriminated against and in need of emancipation, especially through L2 learning. Several scholars have challenged this notion by highlighting the agency of migrant women.
The concept of agency has received considerable attention in second language acquisition, education and psychology in the past years, as scholars have grappled with describing the relationship and interaction between the individual and the social that acknowledges the power of the environment while also highlighting the individual's possibilities for choice, creativity, self-regulation and proactiveness (e.g., Bandura 2008;Bown 2009;Gao 2010;Mercer 2011). We, the authors, align with these efforts as we borrow and adapt the concept of (human) agency from Glen Elder and his work in the context of life course studies. According to him, agency highlights the fact that [w]ithin the constraints of their world, people are planful and make choices among options that construct their life course (Elder 1994, p. 6).
Even when individual choices of migrant language learners occur on smaller scales, for instance in classroom interaction, they often reflect or contribute to larger beliefs and aspirations. To us, expressing such choices, as well as rejecting choices of others, are forms of agency.
First, Daryl Gordon's (2004) study examined the language socialisation and shifts in gender identities of two Laotian immigrant women in the US. Her data from workplace (a foam factory), classroom, domestic and other (e.g. court) contexts illustrate the complexity of interactions in the women's lives and the bidirectional interaction of language learning and gender identity development. In other words, she showed how migrant women navigated a variety of complex interactions in the L2 (e.g. corresponding with legal officials or selling a car) and how their traditional gender roles sometimes promoted L2 learning rather than hindering it, as in the case of one participant whose responsibility for selling the family car spurred her to learn car and sales-related language.
Like Gordon's, Norton's well-known work (e.g. 1997Norton's well-known work (e.g. , 2000  Migrant women's agency can also stem from their L1 resources. Tara Goldstein's (1997) ethnography of 27 Portuguese-speaking female factory workers in Canada revealed that local language practices can be in contrast to those of the official policies (or ideologies) of host countries. For the women in her study, Portuguese was the language that was associated with social inclusion, access to jobs, and community and family life. Even though the women did not see upward mobility as a realistic option for themselves in their host country and their development of English was severely limited, they enjoyed a relatively secure status within the environment of their workplace. Their agency stemmed from to their ability and choice to communicate in their L1.
In our own data from Austria and the US, we also find migrant women acting with agency, particularly surrounding the identities they chose to access -or not -in their L2s.
One poignant example comes from the Refugee Education Center (REC), where teachers were asked to incorporate civics instruction into lessons with an eye towards students eventually passing the US citizenship exam. The melding of English and civics/citizenship education has been common in US adult L2 education for many decades, particularly since 2001, when the US Congress began to set aside funding for this purpose. EL-Civics, as it is called, was codified by the 2014 passage of the Workforce Innovation and Opportunities Act (WIOA; US Government 2014). A key evaluation criterion for federally funded adult ESL programmes, then, has been the number of students who became citizens during the previous reporting period which, in turn, shapes delivery of educational programming: It is common for programmes to offer English and citizenship exam preparation courses together.
US citizenship is desirable to many migrants because it promises certain rights and is 6 sometimes (problematically) associated with social inclusion. In many US adult ESL programmes, then, achieving citizenship is promoted and celebrated. One way REC teacher Joy accomplished civics integration was through in-class practice on the school's e-tablets of key questions which appear on the US citizenship exam, i.e., "The 100 Questions". 7 Frequently, when Joy wanted to move on to a new activity, it was difficult to convince the women to turn in their e-tablets, which Joy countered good-naturedly. The following excerpt is a partial transcript of class audio-recordings and field notes of one such event. Dara's resistance to pursuing US citizenship was remarkable, considering the discourses of citizenship as a desirable goal within Joy's classroom and more broadly, as well as Dara's history in the classroom. Specifically, Dara was a dedicated English learner who studied outside of class, was rarely absent, and was usually amongst the first to comply with Joy's instructions. In addition, she silently took on tasks that would benefit the class as a whole, such as washing dishes after their weekly tea break. In short, Dara did not generally resist classroom practices or discourses, but rather frequently took up a "good student" identity.
By divorcing L2 learning from citizenship in a context where these are often married, Dara challenged expectations held by her teacher, some classmates, federal policy-makers and many Americans that becoming a citizen is necessarily a goal for all who migrate to the US.
Becoming a US citizen may have represented shifts that were in conflict with Dara's political and/or identity investments, such as shedding her identity as a woman from Burma. Although we do not know Dara's particular investments, it appears they did not include becoming a US citizen, at least at the time these data were collected. Instead, Dara showed that she had her own reasons for learning: English needed to serve the person of Dara, not the US citizen of Dara.
Relatedly, in the Austrian context, we expect Hadiah, a woman who navigated an undertaking as complicated as escaping war with three children (Christiane, interview, November 2016), to have agency that would transfer quite easily into cultural capital that is highly valued in her environment. Although, thankfully, Hadiah, ended up in a supportive community, the situation in many other receiving communities is different. There, the identification and enhancing of migrant women's agency rests on shaky ground, since it often depends on individuals or small groups of volunteers who have little support, and are undermined by public discourses and policy decisions which are driven by xenophobic and neoliberal ideologies rather than asset-based approaches to integration.
Given what we have learned from existing studies and our own research, we call for a critical evaluation of L2 education with regard to its potential for building true and sustainable agency of migrant women. There is access and linguistic legitimacy (Ennser-Kananen 2014) to be negotiated within and beyond the L2 classroom (Norton 2000), there can be agency in 21 learning the societal language for one's own purposes, or even in not learning the societal language (Goldstein 1997). Further, sometimes language learning can be a result of enacting traditional gender roles (Gordon 2004), not of overcoming them. As language educators and policymakers, we need to pay attention to these complexities, and expect additional ones, in order to serve our students and develop our profession. At the same time, we cannot turn a blind eye to existing marginalisation and oppression. We are reminded to be thoughtful and (self)-critical about emancipatory efforts and regard them as a two-way street. We need to keep asking: Who is in need of emancipation? Whom does it serve? Who decides what emancipatory efforts look like?

Feminism and empowerment in L2 education
In order to learn about the complexity of a feminist approach to L2 learning, we need to look beyond the realm of L2 education. In foreign language (FL) contexts, research exists that problematises victim discourses which surround female representations in textbooks and female language learners. For example, Jane Sunderland (2000) has emphasised the problematic nature of reducing women to oppressed victims and thereby erasing their agency.
In addition, Kristine Zentgraf has described female empowerment post-immigration and stressed its complexity and non-linearity. In interviews with 25 working and lower class urban Salvadoran women in the US, a group that tends to work in low-paid jobs without many opportunities for upward mobility, Zentgraf noted that her participants did not challenge traditional gender roles but reported a sense of independence and self-confidence in performing and negotiating them in the new environment, for example when running errands in a new environment or managing complex conversations. When she notes, in admittedly problematic language, that [t]here is no reason to expect … that women who migrate from less-developed periphery nations to core Westernised areas will automatically adopt more modern ways of life after migration (Zentgraf 2002, p. 628), 22 she offers an important reminder not to hastily impose Western standards of feminism and emancipation on migrant women who have not been socialised into them and may not self-identify as oppressed.
How can empowerment of migrant women L2 learners progress? As scholars and educators, we have to find nuanced ways of acknowledging systems of oppression, like sexism or racism, and their intersectionalities, while also highlighting the agency of migrant women, which enables them to resist, negotiate and modify gender norms. We cannot assume lack of agency and victim status of migrant women without closely and critically looking at their experiences as well as recognising our own belief systems as historically grown, constantly developing, and imperfect. Thus, a crucial part of empowerment and emancipation has to be a critical reflection of our own power and complicity in oppressive systems. Such introspection is not supposed to be a moral evaluation of individuals, but rather a process which critically evaluates how Western policies, especially throughout the past decades, have enabled dynamics which contributed to the existence of refugee movements and continue to define the life and fate of millions of displaced people and war-torn communities. For example, interventionist policies, the building of borders ("Fortress Europe"), be they 8 ideological, economic or physical, and the criminalisation of those who seek shelter from persecution, violence and economic hardship, have a long tradition in the Western world, which is rarely addressed in discussions of the so-called refugee crisis.
In addition to acknowledging our own role in the oppressive systems at play, we need to break the silence about race and understand the privilege of white males and the feminism of white females, whose experience is fundamentally different from that of women of colour, including a large part of female refugees who currently arrive and seek asylum in Europe.
White feminism has been rightfully criticised for excluding women of colour and thus perpetuating their oppression (Zeillinger 2015). A feminism that serves migrant women of colour has to understand and address the intersectionalities of race and gender as a first step towards reforming itself.
Recognising the complexities and intricacies of gender, language learning, and migration is critical if we aim to work towards long-term solutions for integrating and empowering migrant women and moving our societies towards more inclusive and just ones, rather than promoting band-aid solutions that only provide surface-level change and, in reality, serve to perpetuate the status quo.

Conclusion and implications for migrant women's L2 education
Scholarship on language learning and gender has made big strides. It has moved from a positivist comparing of (sometimes essentialised) male and female learners' language in the 1970s and 1980s towards a post-structural 21st-century approach which examines the construction of learners' identities and their dynamic and complex interaction with gender within particular environments (Davis and Skilton-Sylvester 2004). These scholarly efforts need to keep moving forward, especially in these times of large migrational movements between various cultural contexts in which women's rights have been lacking in different ways and where different forms of emancipation are needed. As research, including our own presented in this article, has shown, learning a second language does not automatically set in motion a virtuous spiral which propels migrant women into spaces of economic stability, linguistic legitimacy, personal and familial emancipation, or social belonging. Although, undeniably, positive effects of L2 learning exist which potentially may enhance (and have enhanced some) migrant women's life quality, these effects are neither automatic nor reliable.
A more helpful image would depict L2 education as embedded in a larger context of holistic support for migrant women, including real possibilities for the development of new roles and identities, interest in and care for all aspects of women's lives, including the experience of war and trauma, and investment in the fight against systems of oppression which deeply affect migrant women's lives, such as racism, sexism, linguicism and neoliberalism. To inspire ideas which would move the development of the field along, we offer the following suggestions:

•
Bridging the research gap. In 1999, Patricia Pessar called for research on gender and migration: "Although there is now broad consensus that immigrant women attain some limited, albeit uneven and sometimes contradictory, benefits from migration and resettlement, we await the next wave of scholarship" (Pessar 1999, p. 586). Granted, important work has been done since then, but the intersection of gender, migration and language learning is still under-researched. Valuable studies will examine intersectionalities such as those of gender, language, race, ability, class, sexual orientation, religion and other factors. They will not stop short at documenting and describing these intersectionalities, but take steps to ensure greater and sustainable linguistic, social and material gains for migrant women. In other words, at the crossroads of gender, migration and language learning, there is a need for researcher activists, who spearhead community-based, emancipatory research and activism and act as reliable allies of migrant women.
• Offering imaginative and transformative opportunities for L2 learning. Based on our literature review and data analysis, we echo the calls which have been made for gender-equitable access to L2 learning and for learner-centred and learner-codesigned curriculum and instruction Pavlenko 2004a, 2004b) which "cente[r] on increasing the control people have over everyday working and living conditions and relationships" (Goldstein 2001, p. 93). We believe that L2 learning can be realised through various "transformative practices" (Norton and Pavlenko 2004a, p. 509), for example through activities and materials which connect students' experiences to feminist thinking and tie them to larger social discourses, and that "[s]uch practices encourage students to imagine alternative ways of being in the world and to consider a range of life trajectories" (Norton and Pavlenko 2004a, p. 509). When Deborah Cameron (2007) laments the shift in language and gender research towards identity and the accompanying depoliticisation of scholarship, we believe a turn towards legitimacy (Ennser-Kananen 2014), an inherently political and power-sensitive concept, can be helpful. Rather than investigating "who am I?", there is a need to examine "who can I legitimately be?", for example as female L2 learner, or their teacher, and then push and cross those boundaries and to expand the options of legitimate ways of being. Of course, such emancipatory transformation which is offered through education needs to be bidirectional.
• Insisting on bidirectionality. As scholars and educators explore and establish transformative L2 education, it is crucial to remember and insist on multiple levels of bidirectionalities, most importantly between learners and teachers. For example, as L2 teachers of migrant women, we become learners of our students' aspirations, histories and ways of being in the world, but also about our own cultures, biases and histories. As members of receiving societies, our complicity in oppression as well as our own oppression become lessons we can learn through the education of migrant women. Coming back to Hadiah's case, as scholars, educators and responsible citizens, we need to learn to foreground what she and women like her have to give, not only what they receive. In the end, the quality of our teaching will depend on the quality of our learning.