Young refugees and locals living under the same roof: intercultural communal living as a catalyst for refugees’ integration in European urban communities?

In various European cities urban authorities and local stakeholders are exploring ways to tackle challenges arising from recent refugee flows. A central concern is the social integration of refugees: how to connect this particular category of newcomers – and especially the most vulnerable ones – durably with local communities? In this article, we discuss an urban programme that offers young unaccompanied refugees (aged 17–23) cohabitation with young locals (aged 20–30) during a period of one to 2 years in Antwerp (Belgium) in small-scale collective housing units. The programme’s assumption is that this mixed, intercultural communal living will promote regular, informal and meaningful social encounters between refugees and locals, which in turn will strengthen the independence and social inclusion of the young refugees. In this article, we investigate the opportunities top-down organized intercultural communal living creates for refugee integration. We draw on interviews and observations collected among locals and refugees living together to gain insights into both groups of participants’ experiences with collective living and the actual social dynamics emerging in such a setting. Our findings suggest that intercultural communal living can be conceptualized as an environment where various informal forms of social support and mutual learning emerge. As such, we contribute to the conceptualization of the impact of intercultural communal living on newcomer integration.


Introduction
Traditionally, migrant integration has largely been a purview of the nation state, as ideas about how to integrate often correlated strongly with ideas about national identity and the national community (Penninx & Garcés-Mascareñas, 2016). However, local governments perceive integration issues differently compared to the national or supranational levels, and prioritize pragmatic approaches to integration over ideological ones (Poppelaars & Scholten, 2008). Recently, local governments and cities in particular have become more entrepreneurial in these domains, developing their own integration philosophies and policies (Scholten & Penninx, 2016). Some even consider European cities' different take on diversity policies, which is characterized by proximity and pragmatism, as 'a policy rebellion' of cities against the state domination of the last decades (Zapata-Barrero, 2017).
This 'local turn' in migrant integration policy making became particularly pronounced in the wake of the 2015 Syrian 'refugee crisis', when local governments became engaged in finding appropriate solutions for some of the issues raised by the increased influx of asylum seekers, such as their need for education, housing and support. In Antwerp, the largest city in Belgium's Dutch-speaking region, counting 520,000 inhabitants, this was also the case. With the support of the European Regional Development Fund, in 2016, a group of local stakeholders led by Antwerp's municipal authorities launched a three-year holistic support programme combining housing, a buddy programme, integrated professional guidance and psychological counselling. 1 More particularly, young adult unaccompanied refugees 2 were offered the chance to live for one up to two years in affordable accommodation, shared with a Dutch-speaking young adult who would be considered their 'buddy'. While unaccompanied minor refugees have a right to live in publicly-funded reception facilities until they reach the age of 18, 3 from that moment onwards they are responsible for finding and paying for their own housing. If they have acquired a legal protection status, at the age of 18 their right to reception is replaced by a right to social benefits from the public social welfare system, which are meant to cover basic living costs (European Migration Network [EMN], 2009). However, due to structural problems related to the local private rental market 4 as well as long waiting lists in social housing, 5 refugees (and more broadly, non-EU newcomers) face enormous difficulties to find decent and appropriate housing, especially in larger cities like Antwerp. It is within this context that it was decided to make housing a central focus of the program. In addition to the provision of housing and a local 'buddy', the refugees also received support from a team of professionals (social workers, psychotherapists) to help them find an appropriate school or work trajectory and to improve their independence and general well-being. 6 In this article, we focus on the aspect of communal living between refugees and their 'buddies'.
Conceptually, the programme under study infuses the notion of organized befriending into the practice of communal living in collective housing. In befriending programmes, sometimes referred to as 'buddy systems', people without adequate support systems are matched with volunteers who act as a friend and offer support and friendship for a determined period of time (Hagard & Blickem, 1987). Furthermore, the volunteer may also take on the position of a coach or a mentor (Van Robaeys & Lyssens-Danneboom, 2016). 1 The programme is coordinated by the municipal Public Centre for Social Welfare; other stakeholders have expertise in outreaching youth-work, adult education, volunteer support, training and counselling for newcomers, and diagnostic and therapeutic support for migrant and refugee children and families. 2 Young adult unaccompanied refugees' refers to unaccompanied minors who turned 18 recently, and who are legally protected by the Belgian state through either a refugee status or a status of subsidiary protection. Six out of ten of the refugee participants are Afghan, while others come primarily from Eritrea, Syria and Somalia. 3 Due to Belgium's federalized structurein which all matters related to migration and asylum, including the reception of asylum seekers, is a federal responsibility while matters related to integration and (youth) welfare are the responsibility of the different communities -young refugees who are granted legal protection while they are still minors usually move from a federal reception centre to a reception facility under the authority of one of the communities. Consequently, it is not uncommon for young refugees to have lived in various reception centres; often enough in different areas or even different regions of the country. 4 For instance, discrimination against social benefit holder and ethnic minorities is widespread on the private housing market in Flanders (Interfederaal Gelijkekansencentrum, 2014) 5 In Antwerp, where below 10 % of the city's patrimonium is social housing, in 2018 waiting lists ranged from four up to 8 years, depending on the urban district. 6 The project's design and approach is explained in the report of Mahieu and Ravn (2017).
These programmes have been used for a variety of groups, such as the elderly, people with suicidal thoughts, children and newcomers, with the objectives of enhancing quality of life, alleviating social isolation and loneliness, providing role models and improving mental health of the help receivers (Behnia, 2007). However, usually the volunteer and help receiver spend only limited time together (Behnia, 2007;Van Robaeys & Lyssens-Danneboom, 2016); it is uncommon that they live together in collective housing. In that sense, the housing component marks a fundamental difference with traditional befriending programmes. Sven Lager, the founder of Refugio Sharehause Berlin, a collective living and working community uniting refugees and Germans, argues why living together profoundly alters the foundations of a relationship: People here [in Refugio Sharehause Berlin] don't just see each other to go swimming together, or to share a meal.
[…] In a way, whilst living together, we share the backstage of our lives. There is nowhere to hide. When you greet each other on the way to the kitchen in the morning, dreams are still looming in the eyes. This purity creates an authentic and deep connection (Heimcollectief, 2018).
The notion of communal living refers to the cohabitation of non-family members in collective housing, to be defined as 'housing that features spaces and facilities for joint use by all residents who also maintain their own individual household' (Franck & Ahrentzen, 1989, p. 3). Originally, communal living was a grassroots phenomenon, but today it has been adopted into the mainstream and is delivered through bottom-up as well as topdown processes (Williams, 2005). The programme discussed in this paper is a good example of how urban authorities aim at achieving social goals (social connectedness between locals and refugees) through the adoption of a particular (communal) housing policy. Its rationale is that by living together, meaningful, regular social encounters between refugees and their 'buddies' will take place and that this in turn will boost the young refugees' social integration and self-reliance in Belgian society. As such, a basic premise of the programme is interculturalism, namely the idea that boosting social interaction between people of different cultural backgrounds can avoid social exclusion and segregation and restore social cohesion. The intercultural policy paradigm aims to foster communication and relations among people with different backgroundsincluding nationals. It therefore focusses more on creating bonds than on stressing differences (Zapata- Barrero, 2016;2017). The paradigm does not preclude rights-and duties-based approaches to diversity, rather 'interculturalism begins then when the multicultural and national civic policies have developed all their potential, not instead of them, against them or before them ( … ) Without a certain degree of recognition of rights and fulfilment of duties, contacts can become difficult.' (Zapata-Barrero, 2017, p. 9). Indeed, in the programme under study, fostering intercultural contacts is complemented by a rights-and-duties approach towards newcomers, which puts forth the goals of learning the language, respecting civic duties and participating in the labour market. This is not surprising, as the civic integration perspective dominates current policy views on integration in the Dutch-speaking part of Belgium. 7 The programme under study raises important questions: what evidence can be found supporting the programme designers' assumption that intercultural communal livingas a radical form of intercultural contact taking place not in the public space but in the private onecontributes to refugee integration (Mahieu & Ravn, 2017)? And if evidence can be found, how can we define these dynamics contributing to integration?
In the following sections, we will first briefly review existing knowledge on cohousing and integration. Then, we explain the programme design, drawing the attention to the strategy for matching refugee-buddy duos and the layout of the collective housing. Next, we present our research methodology. The main section of the article presents an empirically grounded analysis of the social dynamics observed.

Collective housing & newcomer integration
The approach to communal living as a potential solution for a wide array of societal issues, such as the decline in intergenerational solidarity, difficult work-life balance (especially for women), suburban alienation, social isolation and environmental issues, is not new (Jonckheere, Kums, Maelstaf, & Maes, 2010;Williams, 2005). 'Supportive' and 'nurturing' cohousing communities in particular are thought to promote social inclusion and the development of social capital (Franck & Ahrentzen, 1989;Fromm, 1991;Norwood & Smith, 1995). With a hint of nostalgia, proponents of communal living describe relationships between inhabitants as similar to 'ties between villagers'; considering the village as the ideal type of a setting where solidarity, cooperation and all types of support flourish (Jonckheere et al., 2010).
Characteristic for most collective housing, however, is the homogeneity of its inhabitant community in terms of socio-economic class, race, education, attitudes and so forth (see e.g. Williams, 2005). Sometimes, this results directly from practical constraints, for instance, the generally high financial threshold often precludes participation in collective housing for lower socio-economic classes. However, on a deeper level, sharing the same values is considered a prerequisite to successful communal living. Cohousing communities are commonly composed of people sharing a similar living standard and common priorities (e.g. environmental concerns). The homogeneity of community members, in turn, reinforces social interaction among community members. Consequently, understandings of communal living are primarily based on the study of social interaction grounded in the similarity of community members.
Considering our case, the question arises how inhabitant diversity in communal living affects the social dynamics of communal living. Moreover, we want to explore in what ways social dynamics in intercultural communal living promote newcomer integration. As such, our analysis adds a new perspective to the above-mentioned broader discussion on the merits (and demerits) of communal living for society.
Newcomer integration is commonly understood as a two-way process involving immigrants and the receiving society, the latter comprising a legal-political (the state), socio-economic (the market) and cultural-religious dimension (the nation) (Penninx & Garcés-Mascareñas, 2016). Consequently, when investigating newcomer integration, 'the question is not only what immigrants do, with whom they interact, and how they identify themselves, but as much whether they are accepted and how they are positioned in each of [those] three dimensions.' (Penninx & Garcés-Mascareñas, 2016, p. 14).. Immigrant integration should also be conceptualized as a non-linear and multi-faceted process; including among others the evolution of migrants' structural position in society (with regard to rights, employment, health, education, housing), growing social connections between migrants and the local population, and the acquisition of the language and cultural knowledge of the receiving society (Ager & Strang, 2008).
Empirical evidence with respect to the social effects of intercultural communal living is, however, scarce. On a more general level, Williams (2005) suggests that heterogeneity in communal living community can be more difficult but at the same time creates opportunities, as it results in more diversity in terms of the resources residents can offer each other. However, Williams primarily refers to diversity in terms of household type (e.g. families with children, singles and older couples) and affluence. In our case other forms of diversity are at stake; including ethno-cultural, religious and linguistic diversity. While examples of intercultural communal living involving native citizens and newcomers are sparse (Jonckheere et al., 2010), there are a few recent examples (see e.g. Refugio Sharehaus in Berlin (GE), Grand Hotel Cosmopolis in Augsburg (GE), Startblok Riekerhaven in Amsterdam (NL)). However, so far scholars have paid limited attention to these initiatives, one notable exception being Cziske and Huisman (2018). In their study on Startblok Riekerhaven, they found that the 50/50 mix of young refugees and Dutch starters in a collaborative housing engenders the formation of social bonding across ethnic and cultural backgrounds by virtue of belonging to the same age group and household type (i.e., young single people). However, their explorative study focuses on the formation of social ties only, not considering other aspects relevant to newcomer integration such as language and cultural learning and social support. As a result, they fail to provide a broader reflection on how intercultural communal living affects newcomer integration.

Project design: organization of the mixed communal living
A central feature of the programme under study is that the project designers deliberately aim at a variety in background characteristics among the communal living residents, with the aim of nurturing bonds between refugees and locals in particular. As such, mixed communal living is employed as a top-down-oriented strategy reflecting the interculturalist policy paradigm. Before moving into their accommodation, candidate-refugees and candidate-buddies go through an intake procedure, where they are screened (among others, in terms of eligibility and motivation) and matched with one another by the project team and allocated to a specific accommodation.
With regard to the eligibility criteria, it should be noted that besides formal criteria (e.g. having a link to Antwerp, age requirement), refugees are screened during an intake procedure in which a psychotherapist assesses to what extent the youngsters suffer from trauma, chronic stress, 'frozen' grief, etc. Those with very severe psychological issues (e.g. depression, paranoia), problematic social behaviour (e.g. aggression) or a criminal record may not enter the project. Stakeholders argue that their inclusion would overburden the refugee's flatmate, make communal living impossible, and would not contribute to reaching the project's goals. In addition, refugees were required to have sufficient proficiency in Dutch, as it was feared a that a too low Dutch language proficiency would negatively affect the quality and quantity of social interactions between the refugees and their flatmates (Ravn, Mahieu, Belloni, & Timmerman, 2020).
The buddies' motivation to participate in the project was mainly dominated by altruistic motives such as a general urge to 'contribute to a better world' and concern about the situation of refugees. In addition, many were eager to get to know someone with a different ethno-cultural background. While other more material motivations (e.g. to find a decent, affordable flat in Antwerp) were also present, these were usually not decisive in the buddies' decision to participate in the project. Differently and as expected, for refugees, their primary motivation to enter the project was their pressing need for housing. This does not mean that other motives were absent, though; many refugees also expressed a clear interest in making local friends and improving their Dutch language skills (Mahieu, Van Raemdonck, & Clycq, 2019;Ravn et al., 2018).
The system of organized befriending is structured around 'matched duos' of one local and one refugee. Diversity is the key principle in matching, as Dutch-speaking locals are always paired with young unaccompanied refugees. The 'matched duos' live in four different types of collective housing, throughout the city of Antwerp: two-bedroom apartments, four-bedroom houses, a student flat with 12 bedrooms, and finally, a cohousing site that has been built as part of the project and contains 16 two-bedroom modular units and communal areas open to all inhabitants. If four people cohabitate, the two refugees are usually of different ethno-cultural background and mother tongue, further increasing the inhabitant group's diversity. Furthermore, in matching decisions, certain personal preferences were taken into account: for instance, if the refugee preferred to live with someone of the same gender, this was respected. 8 In addition, the vulnerabilities and social skills of all candidates were considered. Overall, the deliberate matching procedure by the project partners aimed at avoiding possible problems that could arise in the communal living.
Prior to moving in together, matched duos met each other during a 'matching activity' organized by the project partners. Refugees and buddies were free to decline the proposal, but this did not happen very often. In case either the refugee or buddy left the project, the remaining participant was matched with someone else. In total, 81 refugees lived together with 77 buddies.

Research methodology
The data we use in this article were collected within the framework of an evaluation study (see Mahieu et al., 2019;Mahieu & Ravn, 2017;Ravn et al., 2018). More specifically, we draw on semi-structured in-depth interviews and informal conversations with young adult unaccompanied refugees and local buddies participating in the project, as well as on observations during project activities. In total, between one and three interviews were conducted with 23 refugees and 17 buddies between July 2017 and February 2019. During that same period, informal conversations and observations took place on a regular basis during project activities such as matching activities, training activities for the refugees, guided meetings between buddies, and more informal project gatherings. Through their participation in project activities, the researchers could build a rapport with the buddies and particularly with the refugees before more formal interviews took place. Respondents were selected as pairs, i.e. wherever possible, both participants of a 'matched duo' were included in the sample. In order to provide a space in which buddies and refugees could talk freely about their experiences, they were always interviewed individually and by different researchers. Interviews normally took place in the research participants' houses, unless participants preferred to meet elsewhere. All interviews with the buddies were conducted in Dutch while in the case of the refugees, sometimes an interpreter was used. During interviews, topic-lists were used to guide the conversations. These topic-lists included a range of subjects such as the participants' social networks, relations with housemates, motivations to participate in the project, experiences with the project, etc.
In our analysis, we adopt an inductive approach to our research question, investigating how the refugees' and locals' experiences with intercultural communal living relate to newcomer integration.

How does intercultural communal living facilitate newcomer integration?
Contrary to traditional forms of communal living that are mostly characterized by homogeneous communities, in our case study buddy and refugee housemates have little background characteristics in common apart from the fact that they are young unmarried adults without children. 9 How does this affect the social dynamics between inhabitants? And does communal living constitute fertile ground for processes facilitating newcomer integration? As will become clear throughout this section, our data suggest that intercultural communal living can create integration opportunities: it can provide an informal supportive environment to the refugee and may support mutual learning for refugees and buddies. While support focusses somewhat more on the benefits in terms of 'here and now', learning hints more at the (potential) long-term impact of communal living. The informal nature of the social interaction is central in both dynamics. However, we also bring to the attention some of the challenges of intercultural communal living, e.g. with regard to communication.
a) Communal living as a source of social support for the refugee Mixed communal living between locals and recently-arrived refugees can be conceptualized as a setting where different types of support are readily available for the refugees. The idea of cohousing as a supportive environment echoes the notion of 'solidarity housing' (see above) as well as assumptions that diverse cohousing communities contain a more diverse pool of resources (Williams, 2005).
Frequently, buddies support their housemates through small gestures, such as giving them a ride, helping them to buy something online (for which sometimes a credit card is needed, something the refugees do not have), sharing certain household items (kitchen utensils, a desktop computer, …) or lending or donating spare furniture, a mattress and bed-linen to the refugee. Indeed, the possibility to share stuff is one of the advantages of communal living common to all communal living communities. The following quote of one buddy, however, shows how loaded tangible support can be, even if it is accepted gratefully: When he [refugee housemate] passed his exam for a very difficult course, I was really like 'Okay, now you deserve a present. Because if I were to graduate, I would also receive a gift, and you don't have anyone who can give you something, so I would love to do so. But will you allow me to do this, or not?' Because this always a difficult issue [whether he will accept this gesture] (...) After some inquiry, it was decided that he wanted shoes. (...) I have not decided yet whether I will buy a voucher, which allows him to buy shoes himself, or if we would go shopping together. But the moment where I pay at the counter would be too difficult for him, so I think that I will buy a voucher. I also started studying again, and he told me 'I also want to do this for you! When are you graduating?' (Rob, buddy) Giving and taking tangible support can evoke an uneasiness on both sides since it highlights differing (financial and social) positions. Likewise, another buddy expressed a sense of guilt when she moved out of the house after 1 year and took all her personal belongings with her, including her sofa, leaving the house rather empty somewhat to the displeasure of her refugee housemates, who were staying there. While these tangible forms of support typically happen in all communal living communities, here they are more pronounced and unidirectional because of the inequalities in financial and material resources between refugees and (most of) the buddies, and the buddies' generally strong motivation to help their housemates. While it is positive that refugees have access to these resources, it is also a double-edged sword, as it also puts them in a dependent position vis-à-vis their housemates, implying they may feel like they owe certain gratitude to the buddy.
Buddies sometimes also offer support by accompanying refugees to formal institutions, such as a hospital or school. For instance, when Azizullah (a refugee from Afghanistan) mentioned he would like to attend drawing classes, his buddy Sander took him to the open house of the arts school. Here, informational support is intertwined with mental support in the form of companionship or a broader sense of 'being taken care of', which relates to another type of social support: emotional support. However, while buddies want to offer a listening ear and show concern when their housemate does not seem to feel well, most of them refrain from prying too much out of a fear of being seen as an 'interrogator'. Moreover, this wait-and-see attitude is also encouraged during training sessions for buddies, where they are advised against asking direct questions about the refugees' story and past. As a result, buddies tend to leave the initiative to the refugees to talk about their personal background and issues, and focus more on building a relationship of trust. On their side, refugees sometimes refrain from sharing their worries because they do not want to burden their housemates. Azizullah, for example, does not like to talk too much about his family with his buddy -although they have a close relationship -, because he does not want to bother him with his own worries: 'When I have problems, … for instance if I don't feel so well, I don't want others to feel like that ( …) For instance, when I'm afraid, when I'm sad, if I tell this to him, maybe he would feel like that too.' (Azizullah, refugee from Afghanistan) However, buddies provide or contribute to refugees' emotional well-being in at least three other ways. First, buddies tend to see it as their role to be generally encouraging; for instance by appraising their flatmates when they accomplish something or make progress (e.g. in Dutch language proficiency). Secondand a distinctive feature of communal living -some refugees see the mere presence of flatmates as a source of support: '[W] hen you are alone it is difficult to study. If you are alone, you are more likely to start thinking about your past and [your] problems, and then you can get stressed … Now, I am happy. I am living in the moment, attending training every day.' (Asadi, refugee from Afghanistan) The excerpt above shows how 'having someone around' is perceived as a source of distraction, helping refugees to think less about their past and problems, and to focus more on their future. Suleymaan (a refugee from Somalia), too, explains how he would not want to live alone because then 'you're just by yourself the whole time' without anyone to talk to. While he and his buddy do not spend much time together, the mere fact of having another person present in the house seems to make him feel better. In larger communal living arrangements, where multiple refugees and buddies cohabitate, this companionship support also involves the other refugees: Third, when buddies observe or suspect the refugees have psychological issues, they signal this to the psychotherapist or other professional caregivers who are involved in the wider supportive programme. For instance, when Bashir, a refugee from Syria, was having a hard time and displayed signs of depression and even suicidal thoughts, his buddy Yasmina warned his social worker, thereby enabling referral to appropriate professional support.
Sometimes, social support not only draws on the refugees' housemates, but also on the buddies' broader social network. In one case, a friend of the buddy comes by weekly to offer supplemental training to a refugee struggling with French classes (a compulsory subject in regular education in Flanders). Sometimes, parents of the buddies jump in to assist refugees with small services and favours, such as repairing their bike, teaching them how to use a sewing machine, or helping out when they need special care: Once I had an accident. I drove my bike into a pole … I had an operation on my cheek and my teeth and then Lies really helped me a lot. And her father and mother also (…). They made a lot of soup for me, because for 1 month, 2 months I couldn't eat [solid food].' (Mursalin, refugee from Afghanistan) Arguably, an important motivation for the buddy's extended network to lend a helping hand is that the refugees are young and unaccompanied, due to which they are perceived as 'alone' in Belgium and therefore in need of support. This type of support shows how the buddies' supportive social networks may also extend to their housemates, however, usually this was not the case and where it was primarily the buddy him or herself who provided support. While many of the above-mentioned types of informational and emotional support may also take place in traditional, organized befriending programmes (outside a setting of communal living), the low threshold to ask for help or information is unique in this case, due to the quasi-permanent availability of the buddy. In addition, buddies often spontaneously offer help and are able to detect particular needs related to the refugees' day-to-day activities. It should be noted that particularly emotional support can be reciprocal. For instance, refugees also ask buddies about their day at work, or make tea when the buddy is ill. However, due to the set-up of the project in which the refugees' vulnerabilities are emphasized (Mahieu & Ravn, 2017), and the inequalities between flatmates (in resources, but also in their motivation to participate), it is not surprising to find that buddies offer more support to their flatmates than the other way round. Finally, it should also be stressed that the extent to which different types of support emerge depends on many factors, such as the particular needs, capacities and personalities of the flatmates and the nature of their relationship. As a result, there were also several duos between whom support asked or offered was merely minimal.

b) Communal living: room for mutual informal learning
Mixed collective housing can also be conceptualized as an environment where informal learning for both locals and refugees occurs. While this conceptualization partly overlaps with looking at communal living as a supportive environment, 10 it draws the attention to benefits that potentially last after the temporary communal living ends. In addition, it highlights mutual dynamics between locals and refugees.
Learning processes, in particular in the realm of language and cultural learning, are vital in newcomer integration since they facilitate inclusion in crucial domains, such as education, labour market and health (Ager & Strang, 2008). As we argue below, communal living carries the potential to facilitate or accelerate these and other types of learning, and as such, to complement the learning processes taking place in formal (e.g. Dutch language courses, full-time education) or non-formal (e.g. workplace internships) learning environments. Below we group the learning experiences we observed into three categories: language learning, intercultural-societal learning and shifting mutual perceptions.

The house as a safe language learning environment
A first domain of learning regards language learning. For the young refugees, the prospect of acquiring better proficiency in Dutch was an important motivation to enter the project. Even though not all buddies have Dutch as their mother tongue, they all master the language, while the refugees, who on average have been in Belgium for about 2 years before entering the project, are still in the early process of learning it. In accordance with the project aims, the language of communication between the refugees and their local flatmates is primarily Dutch. Many refugees report how after a period of living in a mixed setting, they feel more confident speaking Dutch.
My Dutch is quite a bit better than before, because before I lived alone. Now I live with people. Now I feel more confident to ask questions; for example, when we're watching TV together, then I will dare to ask what the programme is about, or what the meaning of this word is. (Yasin, refugee from Iraq) [Before], I was living in a LOI [local reception initiative for asylum seekers]. There I always spoke Pashto, never Dutch. Only in class, in school [I did]. Here I come home and I always speak Dutch. Always, with Yasin, Rob and Els. (Asadi, refugee from Afghanistan) As highlighted in the first quote, Dutch language learning is embedded in everyday domestic practices, such as watching TV. It also indicates how the television is employed as a learning tool rather than being merely a source of leisure. Menial tasks such as cooking, cleaning, doing laundry, doing groceries etc. are also common sources of conversation. In addition, as the second statement highlights, during their initial period in Belgium, refugees often have few opportunities to practice Dutchespecially colloquial Dutch with peers in an informal context.
Buddies, for their part, report that after a while they find it easier to talk with their housemate, on the one hand because their flatmate's language level has improved, but likely also because they have become more skilful in communicating with a Dutch language learner. Therefore, though the newcomer is doing most of the learning, an adaptation has also occurred on the side of the buddies, which resonates with the notion of integration being a two-way process. In addition, buddies indicate how they use an array of strategies to support their housemates' language learning process. For instance, they give positive affirmation concerning their housemates' Dutch language level, explain words they do not understand or help them to understand formal letters. Furthermore, some deliberately create 'speaking opportunities' for their newcomer flatmates, for instance by encouraging them to use the common rooms by putting facilities there (e.g. a desk with a shared computer) or by hanging around in the common room themselves (to highlight their availability for a talk). During conversations, buddies intentionally ask side-questions, in order to move beyond superficial 'how are you / I'm fine' conversations. While many of these activities clearly serve other goals too, such as establishing a relationship or building trust, creating opportunities to practice Dutch are an important driver in the social interactions.
While the language gap between refugees and their buddies opens up an array of opportunities for language learning, the downside is that communication among housemates does not always run smoothly, especially during the early stage of communal living. Particularly for conversations concerning personal matters, besides factors such as general social skills, cultural and gender differences and psychological vulnerability, language is sometimes experienced as a barrier. As a result, in most houses more in-depth conversations between buddies and refugees remain relatively rare. Communiciation difficulties were cited by project participants as a major reason why their relationships did not evolve into deeper friendships, which also reveals the limits of interculturalism as an approach.
In sum, the confluence of two characteristics defines the setting of mixed cohabitation as a language learning environment. First, interaction is taking place in an informal setting and revolves around everyday issues, including shared responsibilities in the communal living areas, contrasting with an artificial classroom setting where there is also a hierarchical relation between language learner and teacher. Importantly, when buddies have too high expectations about the refugees' language progress this turns out to be counterproductive. In one case a buddy invested much effort in language learning without it yielding the expected results, and this led to mutual frustration. This indicates how the informal and non-performance-oriented character of learning needs to be respected. Second, the nature of the buddy as a willing conversation partner for the newcomer sometimes differs from newcomers' limited or more negative experiences with native speakers in the public sphere, where people are often impatient or unwilling to talk. This illustrates how in order to realize the premises of interculturalism, it is important to invest in safe spaces for encounters between different groups.
Learning about Belgian society and culture and the newcomers' society and culture of origin A second domain of learning can be framed as intercultural and societal learning. For refugees, the everyday interaction with local housemates helps them to decipher Belgian society and its (tacit) cultural, social and other rules, norms and institutions. Many of the buddies regard their own willingness to explain Belgian society as a central part of their commitment. For instance, one buddy described himself as 'a teacher-buddy', as he often explained things that he considered to be commonly understood in Belgium, such as the right to euthanasia in case of an incurable disease. 11 However, in these interactions, he is also aware of how he is learning about his housemate's views on this topic. Another buddy has taken up the habit of explaining his own social activities (e.g. going out, going to festivals) to his refugee housemate, as a way of familiarizing him with the 'ordinary' social life of young Belgians.
While buddies are generally open to the newcomers' questions, one buddy lamented feeling like a 'Google-buddy' who was approached by his housemate mainly as an off-line search engine, and that apart from practical questions, little interaction was taking place. Others reported that, at times, it is exhausting having to explain so much, which indicates that, while learning is recognized by the inhabitants as one of the merits of mixed communal living, if the social interaction focuses on this aspect only, or does not take into account the capacity of the other housemates, it may work counterproductively.
While the project has the aim of familiarizing newcomers with Belgian society, it is clear that in practice, learning happens in different directions. For instance, newcomers also pass information on actively to other newcomers, and buddies report how they learn about the refugees' backgrounds and viewpoints.
Like language acquisition, learning about each other's culture and society is triggered by everyday practices and experiences, embedded in the shared living space, like domestic activities. For instance, buddies frequently report how they watch TV-shows with their housemates and discuss the content. For example, watching 'Married at First Sight' 12 and other dating programmes prompts conversation about relationships and different views on them. Sometimes TV programmes are deliberately chosen for this purpose, which points at intentionality in learning: in one of the houses, a popular Belgian children's programme about the human body and sexuality as well as a show about transgender people and drag queens are watched together, as the buddy wants to spark discussions about these sensitive issues.
Again, the bi-directionality of learning processes should be noted. For instance, buddy Annabel reports how discussions about relationships are informative for her, because it made her realize that the notion of 'romantic love' as being the best basis for a relationship is not shared across different cultures. In addition, news items on the newcomers' origin countries incite newcomers to share stories about these countries. Buddies not only expand their knowledge about other cultures; their intercultural attitudes and skills are affected as well: My initial way to deal with it [different opinions] (…) used to be pedantic, whereas now, I do not have the pretention to change their perspectives. I would rather ask," Why do you think that?: (…). So [I would rather apply] a Socratic approach, in order to try to really understand why they think something. (Ruben, buddy) In addition, the refugees' attitudes also incite introspection among the buddies about their own attitudes: for instance, various buddies report being inspired by the generally welcoming attitude of their refugee housemates, for whom sharing food and offering help to people in their network are self-evident.
Learning about life in Belgium also takes place on a more practical level, e.g. in the realm of household responsibilities. For many refugees, it is the first time they are living independently, since as minors they stayed in government-funded accommodation. Just as for their native peers, many tasks and routines related to renting and maintaining a house (making small repairs, sorting the waste and putting garbage outside at the right moment, communicating with the landlord, negotiating with neighbours, dealing with a power outage, adjusting the thermostat, reading the energy meter, using the washing machine, etc.) and, more broadly, living independently (organizing cleaning and cooking, buying groceries, paying bills, etc.) are unfamiliar to them. In addition, the mostly male newcomers often grew up in social environments where household responsibilities were almost exclusively a female matter; and where other tasks (such a waste sorting) were absent or organized differently. Especially the somewhat older and/or female buddies tend to highlight how they 'educate' their younger, male housemates in the essentials of household tasks, not just by explaining them, but also by setting the example and showing for instance how to clean (e.g. what cleaning products to use). Therefore, while more difficult to pinpoint, it is obvious how the young refugees also learn by observing their housemates' household practices. This leads to a type of practical skill that is often overlooked in discussions on 'newcomer integration' but indispensable, as a lack of these skills may for example lead to troubles with homeowners and neighbours, or may even have financial consequences (e.g. fines due to incorrect garbage sorting). Typically, it is also a sort of knowledge that is not addressed in formal educational settings. However, it should be stressed that in several cases, newcomers do know quite well how to go about these issues, or in some cases, are more self-reliant than their local housemate.
With regard to administrative tasks and day-to-day planning, buddies tend to demonstrate 'how things are done here'. For instance, one buddy has introduced a 'letter-hour', a moment where they both deal with written correspondence, also as a sign that reading mail should not be postponed, especially when it concerns registered letters. More generally, buddies stress how they try to set an example and show how 'ordinary Belgians' are arranging their lives in terms of administration, planning, managing their financial budget etc.
Important here is not only the assumed cultural-societal gap (where Belgium is framed as a country with a heavy administrative workload compared to the newcomers' origin countries) but also the age gap, as buddies tend to be somewhat older than refugees (respectively 25 versus 19 years old, on average). However, some buddies also report learning from their refugee housemates' survival skills (e.g. how they manage to survive on a very limited budget), perseverance and effectiveness.
Learning about the 'other': challenging prior attitudes and prejudices Buddies and refugees generally entered the project with prior perceptions about one another, based on societal views and prejudices, but also on personal experiences. This, in turn, shapes their expectations and behaviour towards their flatmate. This dynamic is most clear with regard to the gender-culture nexus.
Since most refugees participating in the project are male (95%) and half of the buddies are female (48%), most of the houses and apartments are gender-mixed. Interview data indicate that housemates behave in a more self-conscious way in gender-mixed cohousing. Suleymaan, a refugee from Somalia, says he would act 'more crazy' if he lived with a man and would, for instance, be able to sit in the living room without a shirt on. Now that he lives with a girl, he acts more 'quiet', with 'more respect'. Similarly, female buddies report how in the communal rooms, they adapt their behaviour, 'Because I know how they [my refugee flatmates] think about certain things, like gender stuff, I notice how they see me differently from the way they see [my male buddy flatmate]'. (Els, buddy).
Being more self-conscious about their behaviour and adapting for instance the way they dress is something both the refugees and the buddies mention doing 'out of respect' for their flatmates. However, it is clear how gender-mixed communal livingand the adaptations it entails on both sidesis understood through a cultural frame. The buddy quoted above, for instance, also explained that she would probably not be as discrete in her behaviour if she were only living with her native Belgian male flatmate, which suggests how her behaviour is driven by an assumption regarding her refugee-housemates' cultural otherness, more than simply their different sex. These 'cultural differences' and diverging opinions about gender-related issues are also frequently mentioned by other buddies and refugees. While these may sometimes lead to uncomfortable situations (e.g. when a buddy's boyfriend spending the night causes clear discomfort with the refugee and the buddy therefore decides to limit the frequency of boyfriend visits) in other cases they are experienced more as a source of 'interesting discussions' and learning opportunities about perceptions and attitudes regarding gender roles. Obviously, a lot also depends on individuals' attitudes; not all refugees hold 'traditional' views, nor are all buddies equally open-minded in their opinions. Moreover, it is important to note that conflicts or difficulties because of cohousing with femaleor in some cases homosexualflatmates are mediated by the project through screening processes and a matching procedure. In order to avoid problematic social dynamics, only refugees with an open attitude were matched with female or gay buddies, while refugees with more conservative beliefs were linked to male, heterosexual buddies. A such, the project aimed at creating appropriate conditions for interpersonal contact, and mitigates the risk that 'contact zones' turn into 'conflict zones'a risk especially in those areas where tensions between communities prevail (Zapata-Barrero, 2017).
While prior perceptions shape the social interactions between housemates, living together engenders shifts in their ideas about the 'other' when these prior perceptions are invalidatedboth for refugees and for locals. For refugees who previously found it difficult to socialize with native Belgians, communal living seems to have altered their view on the native population and on Belgian society more broadly: Yonas: I used to think that all Belgians were racist or something. When you go outside, nobody will talk to you. In Africa, people are more open, you help each other. (…) Belgians are a bit closed. However, if you make contact, [they are] good people. (Yonas, refugee from Eritrea).
The account of buddy Els, after around 1 year of communal living, is equally telling: I'm a bit ashamed to say this, but if I were to meet Yasin and Asadi [refugee housemates from Syria and Afghanistan] on the street, I would not feel at ease. Because they have a -they do not look like criminals, but … [pauses to think] if they are wearing their leather jackets and all …. these are not men or boys I would feel at ease with, as a girl. But now, if I walk on the street, I see many of this kind of boys, sometimes I even know them as friends of my housemates, sometimes I don't. I notice that now I feel more at ease with them, because I got to know them. I know now that these are super nice boys, soft-hearted, very friendly. (...) I knew this before with my mind, but I didn't feel it. Now I really feel it [touches her heart] (Els, buddy).
However, buddies usually think the experience of mixed communal living has not fundamentally altered their viewpoint on refugees. Rather, they feel how their opinion has become much more informed or even embodied (as the above excerpt illustrates), as it is now based on their own first-hand experience. Buddies also report a larger awareness about the particular issues newcomers and refugees need to handle, such as their administrative and legal daily struggles, but also the prejudices they encounter. Buddies tend to stress how they learned how refugees are often more self-reliant than expected and were often surprised to find out that refugees may not need a lot of assistance from themdifferent from what they expect from a group that is always labeled as 'vulnerable'.
However, it should be noted that learning processes seem to depend strongly on the nature and frequency of social interaction between the housemates. Therefore, the idea that all of the above types of learning are necessarily taking place should be nuanced strongly. Our analyses indicate rather that whether and to what extent the housemates engage in learning processes depends strongly on the capabilities, availability and willingness of the buddies and refugees. The above examples come from houses where the general atmosphere was good and contact frequent. In places where the social interaction was limited or tense, the outcome was more mixed. For instance, when refugees had a very limited Dutch language proficiency, learning focussed more on language acquisition and practical routines, rather than on deeper intercultural exchange. In other cases, where housemates were rarely around together, or one or both were not interested in social interaction, learning was limited altogether. Moreover, when the experience of communal living was negative, mutual stereotypes were sometimes confirmed.

Conclusion
Intercultural, socially mixed communal livingas a radical application of the interculturalism paradigmrevitalizes older discussions on the societal benefits of communal living. As we have demonstrated, when recently arrived young unaccompanied refugees and local youth live together, this can create opportunities for newcomer integration. While others have marked the formation of social bonds between both groups as a positive outcome of communal living (Cziske & Huisman, 2018), we adopt a broader perspective on integration as a multi-layered process and highlight multiple other elements such as the acquisition of language and cultural skills. Resulting from this, we add two particular perspectives on intercultural communal living: First, our findings show that communal living may represent an environment where different forms of support are readily available, since the thresholds both for asking and offering help are very low. While this resonates with more general conceptualizations of communal housing as environments facilitating the sharing of human and material resources (Williams, 2005), we have indicated the particular forms of support emerge when refugees and local young adults live together. An essential feature of all these forms of support is that they are informal and easily available. As such, they have the potential to complement or improve access to institutionalized, formal social support for young refugees (e.g. by a social assistant, psychotherapist, etc.).
Second, we have demonstrated how intercultural communal living contains several opportunities for mutual learning, with regard to knowledge on culture and society but also to the acquisition of specific skills, such as language and household skills. A central feature here is that learning is embedded in daily social interaction in a shared environment. Drawing on the experiences of the refugees and their buddies, the communal living arrangement seems to have potential to support or complement newcomers' more formal learning processes taking place elsewhere (e.g. in Dutch language and orientation classes).
In sum, more than just being an environment where one can 'get to know new people' intercultural communal living has the ability to address some of the basic needs of young refugees in terms of learning and support. However, to what extent this potential is realized remains highly dependent on the nature of social interaction. Future research needs to focus on the factors feeding or hindering social interaction among housemates in a setting of mixed cohousing. Nonetheless, we hope the empirical findings presented in this article fuel evidence-based academic and policy discussions on the merits of mixed, intercultural communal living, and on how it offers an adequate response to societal challenges such as newcomer integration.