The Role of the Internet in Shaping Transnational Couple Formation in the Kurdish Diaspora

Drawing on twenty-one in-depth interviews, this article discusses the Internet’s role in the forma­ tion of transnational marriages among migrant Kurds who live in Finland. In contrast to what is presented in the European media, my findings suggest that transnational couple formation among migrants should be seen as highly diverse and more than just practices that maintain and preserve “traditional” marriage customs. Transnational online dating practices make visible how young adult Kurds actively engage in partner formation and spousal selection. Online dating enables in­ dividual autonomy by widening the circle of potential partners outside familial circles and offers a private social space in which people can create relationships on their own terms and evade social monitoring and possibly harmful rumors.


Introduction
In the field of migration studies, it has been ty pical for researchers to make a distinction between differ ent ty pes of cross-border marriages: transnational versus non-transnational, intra-ethnic versus in terethnic. In definitions of cross-border marriage, the notions of transnational and non-transnational are usually separated by the logics of how marriages are created across nation states (see also Körber & Merkel 2014). Transnational marriage is understood here primarily as a phenomenon in which migrants marr y partners from their families' country of ori gin or from the same cultural or ethnic background within the diaspora 1 (see Williams 2010Williams , 2012Charsley 2012). Multicultural or interethnic cross border marriages are not taken into closer discus sion here, since their social and cultural contexts are considerably different.
A clear, univocal defi nition of the concept of transnational marriage is still missing from the re search literature, and the term is used in numerous ways. However, many scholars study ing marriage migration agree that cross-border marriage alone does not signal transnationalism. 2 In order to quali f y as transnational, there should be elements of pre existing and relatively enduring transnational con nections (e.g. kinship ties, social networks, visits or reciprocity across borders), a sense of transnational or diasporic community, and a shared identity. This transnational sense of community may con tain a shared imagination of an original homeland which in different ways affects migrants' and their descendants' social practices and informs their rela tionship to both their country of residence and their country of origin (see Williams 2010 : 99-101; also Vertovec [1999]2010, 2009 on transnational social morpholog y).
Most of the research on transnational marriages within Europe, both qualitative and quantitative, has concentrated on fairly few migrant groups, which have a longer labor and marriage migration history to Europe, particularly Muslim migrants from Pakistan, Turkey 3 and Morocco (e.g., Lievens 1999 ;Shaw 2001;Straßburger 2004 ;Beck-Gerns heim 2007;Schmidt 2008 ;Timmerman 2006 ;van Kerckem & van der Bracht 2013;Carol, Ersanelli & Wagner 2014 ;Liversage 2014 ; for an exception, see Schmidt & Jakobsen 2004). These studies offer im portant insights into partner selection, motives and strategies for transnational marriages, gendered power structures in marriage migrant families, and the unexpected or unwanted consequences of state legislation on marriage migration and migrants' everyday lives. However, marriages within other refugee communities, including that of the Kurds, comprise a relatively unexplored field (Williams 2010 : 141). 4 In addition, little attention has been paid to the role of the Internet in this process (for excep tions see Schmidt 2008 ;Kibria 2012). 5 Ethnological research into transnational online dating practices reveals a much more diverse picture of the marriage practices of migrants than those de picted in the European media. In public discourses and in media images, Muslim migrants' marriage customs have of ten been depicted as traditional, monolithic and oppressively patriarchal, while the marriages of ethnic Norwegians, Swedes, Finns and Danes have appeared modern and liberal, based on individual autonomy and mutual love (see, e.g., An dreassen 2005 ; Keskinen 2009). Transnational mar riages are in political and public debates also seen as an "easy" route for migrants originating in the Global South to migrate to Europe. While discus sions in other Nordic countries have concentrated on issues of migrants' transnational marriages, in Finland the focus has until recently been on families formed through marriages between Finnish citizens and spouses from the Global South (Leinonen & Pel lander 2014). Issues such as forced marriages in Nor way and "honor killings" in Sweden have attracted much attention and media visibilit y in recent years (Bredal 2005), and Finnish media has followed these trends with some delay. For example in the spring of 2014, Finns could watch news on the Internet and a documentary film shown on television related to forced transnational marriages and honor-related violence experienced by Kurdish women. 6 At the core of the stigmatization and stereoty p ing of transnational marriage is a notion of essen tial cultural difference, especially between Muslim migrants' marriage patterns and European/Nordic/ Western marriage practices (Kibria 2012). From this perspective, the practice of transnational marriage stands in contrast to the prevailing values of mar riage and partner selection in contemporar y Nordic societies. In these cultural representations of the "other" there is little space left for individual sub jectivity, agency or resistance. Examining transna tional marriage practices through a focus on online dating, on the other hand, provides a different angle on the issue of transnational matchmaking.
Touching on questions of gender, generation and individual autonomy, in this article I discuss the Internet's role in shaping transnational dating and couple formation among young migrant Kurds who live in Finland. In global marriage markets, Infor mation and Communications Technologies (ICTs) have a quite recent role in enabling contact between individuals (and families) across nation states, as well as in creating new ways of searching for and finding a spouse, for example via various online dat ing websites (e.g., Constable 2003;Venables 2008 ;Kibria 2012). However, the role of the Internet is not always a straightforward one, and neither are the at titudes towards cross-border marriages created via the Internet. The Internet as an intermediar y in the creation of marital unions may arouse suspicion regarding motives in the eyes of immigration au thorities and among migrant families themselves. In this article I examine 1) which features characterize eThnologia euroPaea 46:1

Research Data and Methods
I base my discussion on empirical materials gathered largely in 2011 and 2012. My material consists of the matic in-depth interviews and informal discussions with single and transnationally married Kurds who live in Finland. All of them have moved to Finland from Iran, Iraq or Turkey when they were children, adolescents or young adults. Eleven of them consider themselves either religious or secular Muslims, and eight of them self-identif y as agnostics, non-reli gious or as belonging to a religious minority group. I have conducted twenty-one inter views with nine teen persons, both couples and individuals who had either married or dated transnationally. I have inter viewed nine informants two or three times, and have maintained contact with some of them continuously during the research process via email conversations, phone calls and regular visits. 7 My approach to the inter view material is social constructionist, in other words, I examine how peo ple talk about, conceptualize and verbalize lived and experienced social life as collaboratively (re)con structed by the interviewee and the interviewer in interview narratives. This is not to say that reality outside the human mind does not exist, but rather that understanding of that reality is constructed lo cally, in particular situations and in interaction with others. I also view what is said in interviews in terms of social action and performance. As Paul Atkinson and Amanda Coffey (2001) put it, "actions, we ar gue, are understandable because they can be talked about." According to Atkinson and Coffey, narrated accounts, including those told during interviews, are actions. In recognizing the performative qualities of social life and narration the researcher avoids juxta posing talk and events as if they occupied different spheres of meaning (Atkinson & Coffey 2001: 801; see also R iessman 1993). Every speech act can thus be seen as a performative act, which brings into be ing the social systems of a community through ut terance. Normative values and social practices sur rounding marriage and gender, for example, enter into personal narratives and discourses.
For my methodolog y, I utilize the narrative ethnography approach, exploring narrated experi ences and practices in their socio-cultural and dis cursive but also in their experiential and material contexts (Gubrium & Holstein 2008). I focus on one particular case at greater length. The advantage of concentrating on a few cases is that fragmentation and depersonalization are avoided, which could come about when a researcher makes cross-cutting thematic analyses of qualitative material (Riessman 1993). This strategy enables the researcher to scruti nize the context, particularities, possible contradic tions and situated nature of narratives, discourses and practices. Avoiding depersonalization is an ethi cal decision as much as an analytical strateg y: inter view citations that are detached from their contexts easily represent those under study as faceless and powerless.

Transnational Marriages among the Kurdish Population in Finland
Transnational marriage immigration is a more re cent phenomenon in Finland than it is in many oth er Western European countries. In recent decades, it has involved primarily immigrants from the Middle East, the Indian continent and Northern Africa (see Martikainen 2007;Säävälä 2013;Häkkinen 2014). As a matter of fact, Kurdish migration to Finland reflects recent Finnish migration history in general: in the 1980s, Finland went from being a country of outward emigration to a receiving country for im migrants due to Finnish returnees from the former Soviet Union/Russia. At the beginning of the 1990s, the number of immigrant refugees (e.g., Somalis, Kurds and former Yugoslavians) began to rise in Finland (Söderling & Korkiasaari 1998). Like most of my inter viewees or their parents, many Kurds be gan to arrive in Finland at the end of the 1980s and the beginning of the 1990s either as quota refugees with their families or as asylum seekers, and more recently also as reunited family members (Wahl beck 1999 (Statistics Finland 2012). This means that the question of mar riage will become relevant for a considerable group of individuals in the near future as the children of Kurdish migrants approach marital age. It also ap pears inevitable that transnational marriage will play a part in future marriages among Kurds who live in Finland, since there exists significant pressure to marry a person from the same ethnic background and the number of individuals in Finland who could be considered potential spouses (who belong to the same age group, speak the same Kurdish dialect, share the same stance toward religion, have similar socio-economic backgrounds, share family/social contacts etc.) is ver y limited (about spousal prefer ences among Kurds in Finland, see Häk kinen 2014).
How common are transnational marriages among Kurds? According to statistics, two thirds (63%) of marriages within the Kurdish-speaking population in Finland registered in 2000 -2012 were transna tional. 8 The proportion of transnational marriages is the same for both men and women, and it has remained relatively unchanged during this time pe riod (Statistics Finland 2013). Statistics do not tell us much about the nature of these marriages. Trans national marriages described in my ethnographic material were highly diverse. They varied from self initiated /self-organized 9 to assisted and arranged marriages, and there were marriages both among kin (between cousins, second cousins or other dis tant relatives) and outside of the kin group. Marriage practices could also var y within the same family: some siblings had married a Finn, others a co-ethnic in Finland, a Kurd from "home" or from the dias pora. Similarly, some of the marriages within one family could be characterized as arranged and some as self-initiated/self-organized. W hat is noticeable is that semi-arranged and arranged marriages among interviewees were not always consanguineous, and conversely, some self-initiated marriages took place within the kin group.
Transnational marriage practices and ideas of marriage undoubtedly reflect changes taking place not only within diasporic communities but also in countries of origin. There have been evident gen erational changes in marriage practices and ideals (see, e.g., Hart 2007; Kibria 2012 ; Grabolle-Ҫeliker 2013: 175 ; van Kerckem & van der Bracht 2013): one common stor y I heard about interviewees' parents' marriages was that they were married or betrothed to a person whom they did not know, or whom they had no possibility to meet before the wedding. Some of them were also married off at a very young age. 10 Nowadays, the variety of ways in which young Kurds find their spouse seems to be more diverse than in their parent's generation, also in cases of arranged marriages. The role of parents in arranged marriages may actually be nothing more than suggesting a suit able partner, with the rest depending on their son's or daughter's own interest in the suggested person.
Previous research literature has also shown that higher rates of transnational marriages exist among migrants and their children who come from coun tries with a large Muslim majority population (see, e.g., Lievens 1999  13 gious or secular Muslims, but also take place among other Kurdish religious groups and atheist Kurds. In the case of Kurds, according to my interview data, the ethnicity of the spouse is at least as important as membership in the same religious group or sharing the same stance on religion (see also Straßburger 2002 : 215). The point I seek to make here is that the form of online dating discussed in this article is not necessarily attached to the question of religion or shared Muslim identity as such, but that trans national online dating practices are complex and multi-dimensiona l.

Online Dating in the Context of Transnational Marriage Formation
The mediatization 12 of life worlds and societies has an impact at the global level on transnational net work and partner formation, for example via differ ent Internet practices. In addition to being widely used for maintaining ties with relatives, friends and larger social and virtual communities that might be scattered around the world (e.g., Vertovec 2009 ;Mainsah 2010), the Internet is also used for main taining transnational contact with potential mar riage partners, especially in situations in which a personal link already exists or has been created via shared social (kinship) networks. In such cases, the Internet serves merely as a tool for communicating across nation-state borders alongside other ty pes of information and communication technologies such as the mobile phone. This was common among my interviewees regardless of whether marriages were self-initiated or arranged. This article, however, focuses not on the Inter net's role in maintaining relationships, but in initi ating first contact between potential spouses. Here, the distinction needs to be made between 1) online dating websites, such as Muslima.com, SingleKurds. com or Match.com, which are designed particularly for searching for a partner, and 2) social networking sites such as Facebook, (e.g., Turkish, Kurdish) chat rooms, blogs and online game environments, which can serve to create diverse forms of contact between individuals, and not only for matchmaking (see also Daneback 2006 ;Whitty, Baker & Inman 2007).
In her compa rat ive study of Ba ngladeshi Muslims in the USA and Great Britain, Na zli Kibria (2012) found t hat on line dat ing sites were used by her inter v iewees and t heir fa milies in order to fi nd a suit able par tner from t heir countr y of orig in. Search ing for a spouse v ia matchma k ing websites was a new form of arra nging transnationa l marriages that was more a fa mily project tha n merely an ind i v idua l init iative (see Kibria 2012). For exa mple, t he online dat i ng site SingleMuslim.com, developed to help search for a simi lar-minded Muslim par tner, encourages "you to involve your fami ly -t heir sup port a nd g uida nce w il l be invaluable and impor ta nt in build ing your ow n fami ly." On the same website it is a lso mentioned t hat "[y]ou ca n a lso reg ister on beha lf of your son, daughter, brot her, sister or a ny other fa mi ly member." In t his contex t, online dating sites are adjusted to meet t he current marriage practices of a pa r ticula r socia l group, a nd by including parents in the process, t hey become more acceptable in the eyes of pa rents. However, online dat ing sites a lso simu ltaneously transform matchma k ing prac t ices, not least by increasing t he indiv idua l autonomy of t he younger generat ion to influence spousa l select ion while reducing pa rents' role i n choosing a spouse for t heir child (see Kibria 2012).
The Internet ca n be creatively used in the rea lization of bot h pa rents' and chi ldren's agenc y by suppor ting cultura l cont i nuit y in marriage prac tices and simultaneously by suppor t ing cu ltura l change reflect ing more heterogeneous attitudes towards marriage a nd more diverse mea ns of fi nd ing a spouse. In the Kurdish diasporic context, transnationa l online dating seems to reflect more indiv idua l ly motivated endeavors to fi nd a suitable part ner instead of being primari ly a fami ly pro ject or a new form of arra nged marriage. A g row ing nu mber of onli ne dat i ng sites a im to help users fi nd specifica lly Kurdish pa r tners, 13 a development that signa ls both a need for t his a lter nat ive to more common ways of fi nd ing a pa r tner v ia socia l net works, and t he fact that ethnicit y plays a crucia l role in par tner selection. Such dat ing sites appea r to be more secular t ha n t hose sites specifica l ly for eThnologia euroPaea 46:1 fi nd ing a Muslim part ner. For exa mple, t hey do not high lig ht religious aspects as a centra l criterion in par tner select ion. Instead, t hese websites market t heir matchma k ing ser v ices in roma ntica l ly and indiv idua l ly or iented ways : "Find a Kurd ish sou lmate a nd ex perience the love you've been dream ing about forever" (Kurdishdating.com) or "[f ]i nd your Kurdish part ner for Life, Love and Marriage" (Kurdishdate.net). These websites g uarantee the privacy and anony mit y of t heir users a nd the possi bi lit y "to meet online in a safe env ironment" (e.g., Sing lekurds.com). Many dating sites use Eng lish as a lingua fra nca, and t heir ser v ices a re clearly d i rected toward people who live in the diaspora, and are seek ing a pa r t ner w ithi n the diaspora or from a prev ious homela nd. 14 W hat ma kes social net working sites dif ferent f rom t he use of online dating sites is t hat ind i v idua ls might not a lways be actively searchi ng for a par tner whi le v isit ing these sites or web com munities, and the fi rst contact bet ween the t wo may be accidenta l. A par tner found v ia a socia l net working site ca n be a stranger, but there might a lso be a socia l link (but not necessarily a current ly ma intained one) bet ween the two, for example t he couple might be distant relatives, fa mi ly acquaint a nces, or ex-neighbors in t he countr y of orig in. It is easier to ta ke the initiat ive to start a relation ship on a socia l net work ing site t ha n it is to reg ister for a dat ing site, which means actively and openly searching for a par tner. This mig ht be t he reason why t he transnat iona l online dat ing cases (13) of which I was informed were a l l formed v ia socia l net work ing sites. Eig ht out of thirteen cases involved a young Kurdish woman who lived in Finla nd and had a relationship w it h a Kurd ish man f rom her countr y of orig in (Turkey or Iraq), a nd four cases i nvolved a relationship w it h a Kurdish ma n f rom elsewhere in the diaspora. Four women a lso moved or planned to move back to t heir home countr y a f ter ma rriage. There was also one case in which a Kurdish ma n liv ing in Finland had created a contact v ia t he Internet w it h his fema le relative who lived in t he d iaspora.

Is the Internet an Appropriate Way to Find a Suitable Partner? The Case of Bênav
Bênav, a young single woman in her early twenties, arrived from Turkey to Finland with her family when she was a teenager. In an interview, Bênav shared the fact with me that she has a Kurdish boyfriend from her hometown in Turkey, whom she met through the Internet. The young couple has been dating for a year, and has secretly been in daily contact with each other via Messenger, Facebook and cellphone, shar ing some photos as well. They have plans to marr y and fi nd a future home in their hometown in the Kurdistan area of Turkey. Most members of Bênav's boy friend's family are aware of their dating because of information leaks regarding their relationship on Facebook. Bênav's boyfriend's family in Turkey has accepted their long-distance relationship. How ever, Bênav's parents are unaware of the relationship and the question is how the couple should go about telling her parents about the relationship and their plans for marriage, since they met in a chat room. Bênav is certain that her mother would not approve of the way the couple have started their relation ship, and fears that her parents will not allow them to marr y. She is also uncertain about whether her parents will agree to her moving back to Turkey. Ac cording to Bênav, she and her boy friend must make their relationship appear socially more appropriate, which means that they must invent another stor y of how they possibly could have met each other. This could be, for example, at the wedding of a relative or common acquaintance or through some other common link between the two families. Bênav has planned to tell her parents about her interest in this man on their next trip to Turkey, since at that time there would be a better possibility for the families to meet each other and for the boyfriend 's family to make an official marriage proposal to Bênav's par ents.
In Finland, as in many Western countries, on line dating has become a socially accepted and even mundane way of searching for a partner (Monger 2013: 500). However, only fifteen years ago when online matchmaking was a relatively new phenom enon, there was a highly negative stigma attached to eThnologia euroPaea 46:1 online dating relationships in Finland. This stigma was connected to the imagined attributes of online daters as desperate, lonely and anti-social, based on the presumption that only people who had problems fi nding a partner in face-to-face situations would use the Internet for matchmaking. People were skeptical as to whether a relationship started in the Internet could really endure. In public discussions there were also concerns raised about the potential safety risks of online dating. However, despite negative images, people continued to get involved in online relation ships -many successfully so (see, e.g., Wahlström 2004 ;Hamari 2011). The question arises of why the Internet is per ceived to be a less appropriate means of finding a spouse especially among the older generation(s) of Kurds. Is it due to the same stigma that earlier ex isted in Finnish society towards online dating, be cause it is a new and less familiar way of looking for a partner? Or are the reasons more complicated and also connected to migration and cultural issues particularly? I suggest that the context of migration, cultural expectations regarding how young people should meet each other, and modes of interaction considered acceptable prior to marriage, all play an important role here. The dubious nature of the Internet is intertwined with contradictions that are at play in today's marriage practices, which are cur rently changing in transnational social spaces. These are in turn centrally connected to the issue of sexu ality and premarital relationships among Kurds.
Although there are considerable differences across families in the extent to which they become involved in or control their children's partner selection, mar riage is by many Kurdish families still understood to unite families and not merely individual partners. Because the reputation of the partner and his or her kin group affects the reputation of the whole family, parental approval (even a nominal one) for marriage is still important for many parents (see also Yalҫin-Heckmann 1991;Ammann 2000 ;Grabolle-Ҫeliker 2013). When considering prevailing marriage pat terns among Kurds (see, e.g., Smits & Gündüz Hoşgör 2003 ;COSIT 2005; Grabolle-Ҫeliker 2013), there seems to be no legitimate place for premarital dating, although it does exist. On the one hand, pre marital relationships are becoming more common among Kurdish youth, but at the same time, female reputation and the guarding of female virginity are still seen to be very relevant issues in Kurdish communities in Finland. Premarital relationships threaten prevailing norms of sexuality and espe cially female honor and virginity (see also Mernissi 1982). Because of this, premarital dating usually takes place in secret. Women's behavior in the domain of sexuality comes under scrutiny by the com munity especially in situations in which an ethnic group such as the Kurds has historically confronted, and still continues to confront ethnic cleansing, gen ocide and continuing threats to its existence. Thus, women's actions and sexual morality have symbolic value and they are seen to have implications for the reproduction of the entire ethnic group, their cul ture and identity over time (Yuval-Davis 1997;Nagel 2003 ;Alinia 2013). In this context, it is not a surprise that dating is a delicate issue in the Kurdish diaspora as well.
Where the issue of female virginity is especially sensitive, premarital dating is much easier to or ganize and conceal from public view, thus avoiding harmful rumors when there is a concrete distance between the two persons involved. This makes transnational dating in many ways an attractive op tion. The Internet and particularly different online media platforms such as Facebook and chat rooms not only facilitate networking among individuals and groups, but also create spheres of invisibility and concealment. In cases similar to that of Bênav's, which are characterized by confl icting expectations regarding premarital relationships across genera tions and cultures, considerable tensions can arise within the family. One way to navigate between dif fering expectations and protect family relations is concealment and various forms of deception (Liver sage 2014). In the case of Bênav, this has meant keep ing hidden from her parents the way in which she met her boyfriend and inventing a more appropri ate fi rst meeting. This makes transnational dating in many ways an attractive option.
Because the Internet allows people to engage in eThnologia euroPaea 46:1 The fact that one can potentially date anyone via the Internet makes online dating appear particu larly questionable in the eyes of parents. Especially in the context of migration, and when the potential partner is a stranger to the family, parents may have doubts about whether the partner is interested in a real marriage with their child or whether he or she is just using marriage as a means to migrate to Eu rope. 15 If the family does not have any social link to the partner's family, they may find it difficult to evaluate the background of the potential spouse and his or her family (cf. Charsley 2010). Distrust is also heightened by the fact that the online world can be relatively unfamiliar to the parents' genera tion. However, parents have their own strategies for seeking to prevent the misuse of marriage as merely a ticket to Europe : Bênav: My parents will probably ask a lot of mon ey or gold from him.

Anne : Aha. [sounds surprised]
Bênav: For me, not for them. My mother, her idea is, if he is a stranger, she will ask for lots of gold, because if this man is being serious [about the marriage], of course he will buy [gold]. If he is not serious, and if he wants to use me and wants to just come to Europe, and always parents think these kinds of things, then he won't buy [gold].
[…] That's why my mother thinks that she will ask for a lot of gold.
Anne : How much, do you think, would it be ? Bênav: Well, let's say at least twenty, thirty thou sand euros.
In this citation, Bênav is referring to the practice of bride price which has been traditionally paid to the bride's family and has been a common practice among Kurdish communities at least in rural areas of Turkey at the end of the 1990s and the beginning of the 2000s (see Smits & Gündüz-Hoşgör 2003;Er tem & Kocturk 2008). However, according to Amid Hassanpour (2001: 243), this practice is virtually absent among the Kurdish urban middle classes. In stead, the families of the marr ying couple agree on an amount of gold that the husband must pay to the wife in case of divorce (ibid. 2001: 243). It seems that Bênav is referring to the latter type by emphasizing that the money her parents might insist upon will be intended for Bênav, not for her parents. Its pur pose will be to support Bênav in case of a divorce, or merely as a deterrent to divorce. By asking for a considerably large amount of money from a poten tial spouse, parents strive to ensure that the partner comes from a wealthy family. However, the idea of bride price takes on new meanings and functions in the migration context. It is also used by parents to ensure that the partner's motives for marrying are proper: if he is only interested in Bênav for the residence permit he would obtain by marr ying her, a high bride price will make it more likely that he will consider the stratagem too costly (also Timmerman 20 06 : 135).

Indiv idual Autonomy and Partner Choice in Online Matchmaking
When examining individual autonomy in the choice of partner in online dating, it is important to take into account parents' criteria for their child's future spouse. The question of individual autonomy, how eThnologia euroPaea 46:1 In Bênav's case, it seems that the scope of suitable partners is quite limited. However, it should be noted that Bênav is talking about the preferences of her parents, especially her mother, rather than any ob ligation to marry from her own kin group or from the familiar family circle. She also mentions that her parents have not discussed much about marriage with her, since they want her to focus on study ing and educating herself. In the eyes of parents, a rela tive represents a person who is a more trustworthy marriage candidate since his background and his reputation can be more easily ascertained (also Charsley 2010). There are some contradictions between Bênav's and her mother's ideas of a proper partner, particu larly their views regarding marriage to a close rela tive such as a first cousin. Neverthless, she still tries to find common ground with her mother. Bênav, unlike her sister, does not totally reject the idea of marrying a relative, but there does not seem to be any potential relative whom she might accept -the possible candidates being either total strangers or too close to her, like brothers. She is not convinced that marrying a relative would automatically lead to a good match and an enduring marriage. In her statement she makes a distinction between "love marriages" and "cousin marriages," even though in practice they can be intermingled (see Hart 2007). By making this distinction, she defines partnership and the idea of a good match somewhat differently than her parents, highlighting mutual affection over the decision to marry. On the other hand, it seems very important to her to receive her parents' approv al for her marriage. The Internet widens the circle of possibilities for fi nding a suitable partner, at least a partner who would meet Bênav's own criteria and fulfi ll her ideals. Online dating also enables partici pants to get to know each other and communicate privately in ways that may not be possible for them in face-to-face situations. What is noticeable is that the idea of love as a criterion for marriage now becomes more nuanced in her answer in the same interview: Anne : Well, what kind of things do you see as important in your future husband, what things are important to you, what makes him a suitable spouse for you? Bênav: Well, I have had boy friends before. I al ways wanted that kind of a person whom I would love. But then it was always a wrong option, be cause when I loved them, it felt like they loved less, and the respect and all feelings were less than my own. Or I felt that way, and then I wanted to end a relationship. But then this boy, he loved me and I did not love him. And he tried ver y much that I would trust him and love him and he did a lot and he always respected me very much. Even if I said something stupid he accepted it. That's why, and he was studying then and he believed in God. Then he is from [my] hometown. When I was small, and we listened to young people's talk eThnologia euroPaea 46:1 Bênav's reply to my question concerning her criteria for a suitable partner is multifaceted. I do not sug gest that this is a ty pical answer, but it does dem onstrate that the reasons for choosing a marriage partner are ver y complex, as are the questions of individual autonomy in partner selection. In this narrative, Bênav presents the decision as her own personal choice, by detailing how she had carefully considered and rationally reasoned out what kind of a person she would want to marry. She presents herself in the narrative as a person who has consid erable power over her boy friend, including the right to decide whether to continue or break up with him. She also challenges the idea that romantic love is the best or only indicator of the quality of the relation ship. In her interview, she emphasizes that mutual respect comes before romantic love. However, as her narrative reveals, her decision has not been based on merely rational choice but has also been affected by emotional considerations. Bênav's narrative is also highly nostalgic in its tone. The fact that her boyfriend feels familiar to her even though he does not belong to her kin or circle of family acquaintances is important to her. The fa miliarity she feels toward him is connected to their shared origins (being from the same village), as well as a shared world view and values. By emphasizing the (ethnic) identit y, roots, religion and culture she shares with her boyfriend, Bênav stresses not only their importance to her but also that she misses par ticular aspects of Kurdish culture, those which are "right" and "old," in other words, vanishing or vul nerable to change. These aspects of Kurdish culture were familiar to her in her youth, but are now, ac cording to her, rare among the younger generation, not only among those who live in Europe or in Fin land but also in her hometown itself.
This criterion of shared identity, culture and roots, which is central to the notion of transnation al marriage, is especially evident in Bênav's stor y, whereas the criterion of pre-existing transnational connections in the form of social ties between po tential partners is missing. In research literature on transnational marriages, the role of pre-existing social networks has usually been highlighted in the process of couple formation (see Straßburger 2002 ;Beck-Gernsheim 2007;Charsley 2012 ;Williams 2012). This is probably the most conspicuous differ ence between marriages arising from transnational online dating and other forms of transnational mar riages (especially arranged ones, but also self-organ ized marriages). Particularly in the case of trans national online dating, it is more likely that social links are missing or that their role is less important in couple formation.
The case of Bênav illustrates the complexity of transnational online dating and its connection to religiosity. It is important to note that Bênav did not find her boyfriend through religious forums on the Internet. The issue of her partner's religious af fi liation is a tricky one in Bênav's case, even if it is only one (however meaningful) attribute of a suit able partner in her opinion. Bênav considers her self a devout Muslim despite the fact that she does not cover her head with a veil or pray regularly. It is worth noting that while her mother is also a devout Muslim, her father and sister are not particularly religious according to Bênav. Because her father is not religious, the religiosity (or non-religiosity) of eThnologia euroPaea 46:1

Conclusion : Online Dating and Transna tional Marriage Practices in Transition
While there is no clear or single definition of trans national marriage, it has been common to highlight the role of existing transnational social ties between families and individuals in its formation. Especially in cases of arranged marriages, this social link can usually be traced with relative ease, but in the case of transnational marriages that have started online, the relationship is not necessarily based on a prior social link between the couple. On the other hand, if the criteria of pre-existing transnational connections is understood in a broader and more general sense, referring to visits to the country of origin or any pre existing social ties a person may have to the country of origin or within the diasporic community, then online dating and marriage meet the defi nition of transnationa l. Transnational marriages are more than just prac tices that maintain and preserve "traditional" mar riage customs among migrants. An examination of transnational dating online offers a different picture of transnational marriage than that familiar from the media in which transnational marriages are pre sented as merely arranged or forced, or as used by those seeking to immigrate to Europe. It is clear that through online dating, young adults actively engage in transnational partner-formation practices in or der to find a suitable spouse. Transnational online dating also reflects more heterogeneous attitudes toward marriage and couple formation among the younger generation of Kurds than among previous generations.
What makes transnational online dating an at tractive option for young Kurds is that it widens the circle of potential marriage partners which can be quite limited, as for example in the Finnish context: finding a partner through social networks or via social gatherings (weddings etc.) can be challeng ing in Finland due to the small number of potential partners in the same age group. In addition, Kurds are not a socially, linguistically, religiously or politi cally homogenous group in Finland (or elsewhere), which exacerbates the perceived scarcity of suitable partners.
The other feature that makes online dating an at tractive option is that the Internet offers a private social space in which subjects can create relation ships on their own terms, evading power structures, social monitoring and possibly harmful rumors (also W heeler 2006 ;Schmidt 2008 ;Pearl Kaya 2009 ;Christensen 2011). The Internet may therefore en able individual autonomy in partner formation that is not possible elsewhere.
However, online dating does not entirely liberate persons from the social restrictions existing in the offline world. Normative expectations of the co-eth nic community as well as family and social relations in the countr y of origin are not rendered meaning less by the Internet. The case of Bênav in this article illustrates how transnational online dating is not yet widely accepted in the process of couple formation among Kurdish migrants. The Internet as a tool for finding spouses creates distrust because it is a less familiar means of seeking a partner, but also because it challenges sanctions against premarital relation ships (even when there is physical distance between the partners) and the role of social ties in partner se lection. Additionally, online dating does not entirely mitigate the risks to female reputation. Finally, on line dating raises doubts in parents' minds regarding eThnologia euroPaea 46:1 The process by which plans to marr y are realized involves either confronting or adjusting to prevail ing norms and existing power structures outside the virtual world (Schmidt 2008: 18). This means fac ing up to social pressure and expectations accord ing to which some aspects of the couple formation is inappropriate. Because of its questionable status as a meeting place for young persons, relationships created via the Internet will probably be disguised at least to some degree as relationships whose origins are more appropriate and socially acceptable. One way of by passing and adjusting to prevailing mar riage norms is to invent a social link between the couple after the fact. In the long run, successful and unsuccessful stories of transnational marriages with online origins will undoubtedly shape both posi tive and negative images of this mode of first con tact between potential marriage partners. The more successful and enduring the marriages are that were initiated via the Internet, the more socially accept able online dating and resulting marriages might become.

Notes
1 In this paper, I use the term "diaspora" to refer to di verse Kurdish communities that live outside a collec tively imagined homeland, Kurdistan. By using this term, I want to bring to the fore particular dimensions such as diasporic consciousness -the awareness of be ing part of a globally dispersed community and the desire to identif y with a Kurdish community -as well as transnational social networks and activities within the diaspora and between the country of origin and the country of settlement (see also Brah 1996 ;Wahlbeck 1999 ;Toivanen 2014). 2 In its most reduced form, marriage is defined to be transnational if a migrant living outside his or her country of origin marries a partner still residing in the same country of origin. The problem with this defini tion is that it leaves out marriages between couples who live in different countries in the diaspora. 3 Those studies that examine Turkish marriage migra tion to Europe treat migrants from Turkey as a rela tively homogenous group, and rarely address Kurds separately. This is due to the population register sys tems used in many European countries : people are not registered in the statistics according to their mother tongue (Finland is one of the few exceptions). Kurds thus comprise a vast, invisible group within Europe. Qualitative studies of marriage migrants from Turkey do contain interviews with Kurds as well. 4 There are several exceptions that partly deal with ei ther transnational marriages or spousal issues among Kurds in diaspora (see Ammann 2000 ;Gran 2007; Eli assi 2010). However, the focus of these studies is not particularly on transnational marriages. 5 There are several studies that take the role of the Inter net in global marriage markets into account (e.g., Con stable 2003; Johnson 2007; Venables 2008), but even though the online dating and marriage practices they describe take place across nation-state borders, they cannot be considered transnational per se, since such marriages are made across, rather than within, eth nic communities (e.g., Vertovec [1999]2010 ; Williams 2010). 6 The documentary film "Vuorilla olen vapaa" ("In the mountains I am free") was shown on television by the Finnish Broadcasting Company YLE on May 15, 2014. In addition, the semi-tabloid newspaper Iltalehti ran a story on May 31, 2014, of a Kurdish woman who lived in Finland and had experienced considerable social control and threats by her family and thus had to es cape from home to be able to live the life she wanted. Especially in the documentary, Kurdish women were represented as being generally oppressed, and all prob lems such as violence in the family and forced mar riages were represented unproblematically as caused by Kurdish culture. 7 All names of inter viewees used in this paper are pseu dony ms. I have also used approximations of age and have changed some minor details related to personal information in order to make it more difficult to iden tif y the participant. 8 Statistics do not reveal the number of those transna tional marriages in which a person has moved from Finland to another countr y. This means that the actual percentage of those having a transnational spouse is probably much higher than what register data can re veal. 9 I prefer to use the term "a self-initiated or self-organ ized marriage" instead of "a love marriage" as it better describes the process of finding a spouse in this context (see also Bredal 2006). 10 I was also told about "love stories," but these cases were presented as quite exceptional, something that was contrar y to the norm. 11 It has been estimated that the majority of Kurds (75%) eThnologia euroPaea 46:1  (Väliverronen 2001: 159). It also characterizes changes in practices, cultures and institutions and denotes transformations of these societies. 13 E.g., SingleKurds.com, Kurdishdating.com, Kurdish date.net (launched in August 2014, also on Facebook). 14 There are also a few sites whose target group is singles in Kurdistan, for example Mingel2.com. 15 The problem of marrying purely to get a residence per mit in Europe does not apply only to online marriages but also other types of transnational marriages, and is recognized not only by parents but also by young Kurds themselves (see also van Kerckem & van der Bracht 2013; Karlsson Minganti in this issue). 16 I asked this question just af ter Bênav spoke of the mar riages of her cousins, and of how her parents tend to indirectly express their expectations for Bênav's future spouse when talking about other people's marriages. 17 Emphasis in original.