Territorial bounds to virtual space: transnational online and offline networks of Iranian and Turkish–Kurdish immigrants in the Netherlands MATTHIJS VAN DEN BOS AND LIZA NELL


In this article we question a central trope of transnationalism and new media – deterritorialization – and its application to border crossing Internet usage by Iranian and Turkish–Kurdish migrants in the Netherlands. Their Internet usage indicates the extent to which territoriality channels these groups’ online practice. We found Dutch–Iranian sites reflected correspondingly sparse offline community networks and state boundaries moulded their transnational ties, while regionally specific transnational dynamics were evident in Turkish–Kurdish website surfing. These cases indicate that transnationalism and new media need not broaden or dissolve geographical identity or connectivity, but may reinforce it. Finally, we address the relations of territoriality with generation (first and second) and network medium (web forums versus conventional sites). Whereas first-generation migrants’ life online often reveals extensions of offline networks, the online practice of the second generation frequently reflects these networks in subtler ways, forming partially sovereign online communities that pivot on hyphenated identities. However, the relations of generation and network medium differ for Turkish Kurds and Iranians in the Netherlands. Even in transnational social spaces, place continues to count. (Kivisto 2001: 571) Is deterritorialization the ultimate symbol of globalization? Do its academic proponents have a strong case when arguing ‘that geographical space has ... ceased to be the master index for social and cultural processes’ (Wardle 2002: 494)? Cases abound of deterritorialized nation-states stretching beyond their boundaries, with leaderships claiming their dispersed populations as citizens (Basch et al. 1994). But a recent study of Haitians’ transnational Internet use serves as a reminder of the unabated importance of place-based social ties for sustained – as opposed to merely ephemeral – interaction (Parham 2004). In a similar vein, we enquire in this article whether the adjective ‘deterritorialized’ actually captures migrants’ Internet networking. For many analysts of the contemporary era, transmigrants and the Internet symbolize the latest stages of globalization. Globalization in their view refers to cross-border Matthijs van den Bos and Liza Nell 202 © 2006 The Author(s) exchanges and to transnational movements of people, goods and ideas across the earth (Guarnizo et al. 2002: 1). Transnational migration and Internet communication are also seen as instances of a (post-)modernity that has come to replace an earlier modern condition. The older modernity connoted massive, unidirectional migratory streams and the communications revolutions of the late nineteenth and early twentieth centuries. In the latest theorization of contemporary global flows, transnational migration and online communication also imply deterritorialization. Temporal and spatial boundaries are said to have dominated previously globalized social settings predicated on ‘where and when’, while a different logic is stated to inform current globalization. Castells (1996) believes a culture of virtuality governs global flows in the network society. The network society reflects the post-industrial information age, in which ‘timeless time’ and a ‘space of flows’ transcend prior industrial time and place regimes (Uimonen 2003: 277). Within it, the nation-state, more than any other territorial principle, suffers from dilution: ‘[O]verall the new state [in the information age] is not any longer a nation-state. The state in the information age is a network state’ (Castells 2000: 14). We address territoriality by investigating whether online interaction reflects the dynamics of a disembodied network society or, alternatively, offline contexts with territorial referents to locality, region and nationality. Ideal-type deterritorialized networks would be characterized by completely random node and tie distributions, and not many observers would describe social facts in these terms. Real-world deterritorializing networks, however, reveal a significant blurring only of territorial boundaries, and this process features prominently in many social analyses. In the Dutch–Iranian case, deterritorializing networks would preclude hyperlinks whose destinations are biased toward other Dutch–Iranian websites over any other European web address. In the Turkish–Kurdish case, deterritorializing networks would imply a substantial number of websites lacking clear territorial referents as to hosting location, or references to place of origin or settlement in web content. We argue, however, that Dutch–Iranian sites reflect offline community networks that are correspondingly sparse and decentralized. Moreover, national links are overrepresented when compared with transnational links (including European ones), the distribution of which follows a national pattern. National, regional and locally specific dynamics come to the fore in Turkish–Kurdish website surfing, which connects Turkish Kurds in the Netherlands who identify with specific nations, regions and localities in their (claimed) country of origin, with Turkish–Kurdish communities in third countries and in Turkey. In other words, transnational networks and new media need not broaden or dissolve territoriality, but may reinforce it. Second, we examine whether territoriality varies according to geographical scale only, or whether it involves substantial levels of differentiation as well. We argue that while local, regional and national spaces generally mediate Turkish–Kurdish and Iranian Internet practice in the Netherlands, other factors – immigrants’ generation and network medium, or site category (web forums versus conventional sites) – alter the particulars of territoriality. First-generation migrants’ life online reveals fairly straightforward extensions of offline networks; second-generation individuals are Territorial bounds to virtual space © 2006 The Author(s) 203 more likely to reflect these networks in increasingly subtle ways. The latter may host partially sovereign online communities that pivot on hyphenated identities. Finally, we examine whether generation correlates with differential Internet practice on conventional sites and online forums respectively. Comparing Iranians and Turkish Kurds in the Netherlands Diaspora and transnational migration studies have developed in many directions but relatively few have sought to ground their work in empirical comparative research. The case of two ethnic communities in one country of settlement – Iranians and Turkish Kurds in the Netherlands – allows us to investigate the roles played by territoriality – persistent ties to a specific territory – in migrants’ Internet usage. Several traits shared by these groups – size and country of origin, beside country of settlement – lead us to project similarities in territorial aspects of their Internet use. The Netherlands has been an important destination for Iranians. In the period 1997–2001, more Iranians applied for asylum in the Netherlands than in Sweden or France (Hessels 2004). The Dutch–Iranian community now comprises 28,522 people (CBS StatLine 2005); Kurds from Turkey are registered as Turks – the largest immigrant group in the Netherlands. Although it is not known exactly how many of them identify themselves as Kurds, it is estimated that around 50,000 Kurds from Turkey live in the Netherlands (ROB 2001: 34). These figures indicate that Iranians and Turkish Kurds are small communities compared with other immigrant communities such as Turks in general (357,911), Surinamese (328,312), Moroccans (314,699), and Antilleans (129,721) (CBS StatLine 2005). Second, both these groups constitute diaspora groups par excellence, in that they are barred from direct political participation in their homeland, or do not have a homeland. For Iranians, territorial origin applies to an existing nation-state; for Turkish Kurds, a projected nation-state. Migration from Iran to the Netherlands was largely related to the 1978/9 Islamic Revolution, while economic migrants have arrived in large numbers since the 1990s. In the case of Turkish Kurds, migration motives have been more varied, although a political event – the 1980 military coup in Turkey – was fundamental. Kurdish immigrants from Turkey arrived in the Netherlands a decade after Turkish labour immigrants. Kurds were underrepresented in the first wave of labour migration in the 1960s, as recruitment mainly took place in western and central parts of Turkey (van Bruinessen 1999). This changed in the 1970s, when labour was increasingly recruited from eastern Turkey. The country of settlement has provided both Iranians and Turkish Kurds with opportunities for transnational political activism. Several political incidents, among them the takeover and ransacking of the diplomatic offices of the Islamic Republic in The Hague and the beating of the ambassador by members of the radical leftwing Fedayin in 1984 (Sreberny-Mohammadi and Mohammadi 1987), have highlighted the Iranian presence in the Netherlands. Covered by the international media and provoking outrage on the part of Turkish authorities, Turkish Kurds established a parliament in exile in the Netherlands in 1995. Matthijs van den Bos and Liza Nell 204 © 2006 The Author(s) Methods The advantage of our comparative approach is the specification of levels of difference and commonality in processes of territorial identification or dilution, allowing us to theorize explanatory factors and their scope. The difficulty in our comparison lies in findings that stem from largely unrelated research projects, which explains the discrepancy in numbers of respondents and other features of the data. The two data sets, however, are sufficiently similar for us to explore a unitary set of questions. Some 72 Dutch–Iranian websites and 21 Turkish–Kurdish respondents provided us with the core data. These data sets render comparable information on the Internet usage of both migrant groups. Dutch–Iranian websites were accessed in 2003 and 2004 through manually snowballing the web on keywords such as ‘Iran’ and ‘Nederland’. The Google results and randomly discovered online locations provided ‘Startpages’ with useful links to Dutch–Iranian sites. Each link was then checked to the point where no new hyperlinks could be found, that is, until saturation was reached. A fairly comprehensive sample was obtained by presenting the list to Dutch–Iranians from different walks of life and asking them to add to it, thus repeating the saturation procedure. All Turkish–Kurdish interviewees were initially to be selected from a database on Social Position and Use of Facilities (SPVA-survey). The list contained a select sample of 35 heads of households from municipality registers in Amsterdam, Utrecht and The Hague. These respondents had identified themselves as ‘Kurds’ and made no objections to further interviewing. Unfortunately, non-response was high or respondents could not be reached at the addresses provided. Finally, three of the 35 cooperated; in one further case the son of one of the 35 cooperated. The other 17 respondents were approached via the researchers’ networks, Turkish, Kurdish, and Turkish–Kurdish organizations and Kurdish websites. The fact that the majority of the respondents were chosen on the basis of self-identification and snowball sampling means that the sample is biased towards highly educated, politically engaged respondents. Other researchers in the Netherlands encountered similar sampling problems, for Turkish Kurds are not registered as Kurds (see Latuheru et al. 1994). The variety of respondents selected through these various means, however – beyond those with strong self-identifications as Kurds – precludes preselection toward territorial attachments. Even though the Iranian sample involves websites and hyperlinks and the Turkish–Kurdish sample consists of interviews about individuals’ website and forum surfing – and both studies were designed for largely unrelated purposes – our data allow for comparison of territorial aspects of migrants’ Internet usage. Given that the dictum ‘you are what you link’ is widely accepted (Adamic and Adar 2001), hyperlink structures may be conceived to depict chosen identities and self-expressions of the Dutch–Iranian diaspora online. Similarly, ‘you are what you link’ compares with ‘you are what you (say you) click’ in the case of Turkish–Kurdish web users in the Netherlands. Both data sets thus provide access to similar levels of selfexpression. Territorial bounds to virtual space © 2006 The Author(s) 205 Territorial bounds Anthropological theorizations of place increasingly stress fluidity and ongoing construction in contrast to older conceptualizations of place as fixed and given (Ward 2003: 81–3). These new traits emerge, for instance, from Graham and Khosravi’s rendering of (Iranian) cyberspace: ‘[f]or a displaced people in the diaspora, cyberspace can be an alternative “territory”, where a transnational community or a virtual neighbourhood can be constructed’ (2002: 228; see also Appadurai 1996). For other authors, the concern with process paradoxically obliterates notions of place, as when we are told of the alleged ‘global social fact that ... people are chronically mobile and routinely displaced’ (Malkki 1992: 24), or uprooted ‘non-places’ are said to exemplify global ‘supermodernity’ (Augé 1995). Place gives way here to displacement, and a trade-off between place and flow is similarly evident in theories of cultures’ deterritorialization. Gupta and Ferguson, for instance, claim that ‘actual places ... become ever more blurred and indeterminate [while] ideas of culturally and ethnically distinct places become perhaps even more salient’ (1992: 8). These perspectives are, by implication, beyond situated views of place construction (Ward 2003). But our cases suggest that ‘being-in-a-local-setting’ moulds migrants’ webs of relations and attachments. That settings often imply territory is evidenced by the salience of hyphenated identities – which more often than not involve national identity markers – in discussions of post-migration ethnicity. Relocating Dutch–Iranian exile Dutch–Iranian cyberspace consisted of approximately 72 websites. Exploring their patterns of interaction revealed that Dutch–Iranians clearly favour offline physical proximity over the random node and tie distributions that ‘decentred’ new media and ‘unbounded’ transnationalism would lead one to expect. First, Dutch–Iranian cyberspace mirrors the offline patterns of a low level of communal organization and sparsely-knit networks. Ghorashi notes that in contrast to the United States, which has dense pockets of Iranian migrants, ‘the Netherlands has a rather ... scattered Iranian population’; unlike Los Angeles, there are ‘few Iranian group or community activities’ (2003: 10; also Hessels 2002). This shows primarily in low ‘cyberorganization density’, which refers to Dutch–Iranians’ presentation of self as seen through their number of websites relative to (ethnic) population. Dutch– Iranian cyberorganization is low compared with the density figures for non-Iranian ethnic organization, if assessed by the number of offline ethnic organizations relative to (ethnic) population. Comparing cyber with offline principles of organization yields a theoretical hierarchy. It is comparatively easier, cheaper and often more effective to manage an online presence than to set up and administer an offline organization. Thus, the density figures of the hypothetical ethnic and cyberorganization order for the latter should be far higher than for the former. However, Dutch–Iranian cyberorganization density for 2003 is calculated at 2.57 ((72/28,043)1000), in contrast with higher densities for offline ethnic organizations of Surinamese (3.00), Moroccans (3.10), Matthijs van den Bos and Liza Nell 206 © 2006 The Author(s) Turks (5.60), Antilleans (3.70) and Ghanaians (7.10) in Amsterdam (van Heelsum 2002: 5). The lower rate still holds when taking into consideration national (organization) figures for Turks (van Heelsum et al. 1999: 11; CBS Statline 2005). Given the hypothetical order of ethnic and cyberorganization, Dutch–Iranian cyberorganization appears weak. Furthermore, the previously mentioned offline features of Dutch–Iranian organization also re-emerged online in the low level of network ‘compactness’. That is, both density (the actual number of hyperlinks out of the possible total) and centralization (the extent to which the links are organized around focal points) were low in absolute terms. Density measured 0.03 (150/(7271)), indicating low overall network cohesion. A social network analysis software programme’s ‘degree’ routine (PAJEK; see Batagelj and Mrvar 2004) produced the figures of 0.16 for in-degree centralization and 0.26 for out-degree centralization. These findings indicate that few or no websites in the network were found that overshadowed the others for incoming or outgoing link numbers. These on and offline structures were also explored transnationally, yielding two sets of conclusions on hyperlinks from Dutch–Iranian websites to Iranian websites in non-Dutch settings. The first set concerns outgoing hyperlink (outlink) numbers; the second concerns outlink patterns for specific network sectors. When excluding 83 dead links, the 72 sites show 962 hyperlinks to 20 country domains (including the Netherlands). Transnational ties predominate in Dutch–Iranian outlinks, though national ties are strongly over-represented. The United States and Iran account for the most outlinks – 304 and 243 respectively – while the Netherlands ranks third – that is, disproportionately so. An average distribution of sites per country yields about 50 hyperlinks (962/20); 161 links refer to Dutch–Iranian websites – over three times the average. That Dutch outlink numbers by far exceed those for any other European destination is additionally significant given the much larger Iranian communities in Germany, France and Britain. As for outlink patterns, (trans)national links are broadly equivalent to national, Dutch sites as to their cultural, economic, political and religious numbers, with the exception of political and economic elements. That is, economic sites are more important nationally (among Dutch–Iranian sites), whereas political links are slightly more significant transnationally (among all hyperlinks). Web content, however, which shows identical ideological orientations for national and transnational political sites and links, balances the difference: many national sites are explicitly left wing, as are the majority (67 per cent) of transnational hyperlinks. In contrast, 12.5 per cent of transnational hyperlinks are nationalist and/or monarchist, 8.6 per cent are interest group or democratization oriented, 3.8 per cent are conservative Islamist, 1.9 per cent are feminist, 0.9 per cent are liberal-progressive, and 4.8 per cent represent other identities. Another correspondence between the national and transnational patterns is evident in the scarcity of religious sites and links on either level. In addition, the scarcity of religious links indicates that the similarity of national and transnational networks results from a projection of the national pattern on the transnational stage, and not the Territorial bounds to virtual space © 2006 The Author(s) 207 other way around. Religious scarcity, online as well as offline, may be related to the experience of the overbearing character of Islam in Iran as a primary cause for migrants to seek refuge in the Netherlands (Hessels 2002). Patterns of Iranian migration into Germany are much more varied and have rendered a very different immigrant community. There is historical depth to the Iranian presence in Germany’s Iranian capital of Hamburg. A religious merchant colony has existed in Hamburg since the late eighteenth century (Hesse-Lehmann 2002), and the city became the national centre of Shiism (Lemmen 2000: 64). Bafekr and Leman claim that the migration of Iranian students to Germany between 1950 and 1960 was not politically motivated (1999: 95), though many instances of Iranian student activism occurred in Germany in the 1960s and 1970s (Matin-Asgari 1991). After the Islamic revolution, the ‘Islamisches Zentrum Hamburg [...] played an exceptional role as Iran’s ideological centre for the dissemination of Iranian-type Islamism among Muslims living in Western Europe’ (Grünewald 1995). One suspects that it is because of these longstanding and varied settlements that religious sites feature prominently in Iranians’ German webspace (Azimi and Brückner n.d.; Wolter 2004). Were transnational realities of the Iranian diaspora or the Internet’s alleged unboundedness to have counted here, one would have expected not only significantly fewer Dutch–Iranian hyperlinks, but also a radically higher number of religious links. But the hyperlink pattern indicates national-to-transnational directionality instead. These counterintuitive facts on the unabated importance of national hyperlink numbers and patterns bear immediate relevance to debates on transnationalism, space and place, and globalized new media. Digitally mediated globalization is often held to be to the detriment of particularistic identities such as nations (Negroponte 1995), but case studies (for example, Shichor 2003) often suggest otherwise. A global hyperlink sample establishes that ‘the organization of the world wide web conforms to some degree to traditional national borders’ (Everard 2000; Halavais 2000: 7). Virtual transnationalism, as it emerges from the above case, demonstrates a similar paradox of spatiality: the rupturing of borders in online Dutch–Iranian transnational networks is itself patterned on a particular (namely Dutch), spatial nation-state structure. The social facts underlying Dutch–Iranian migration, then, contrast with cultural connotations of exile as dislocation – as in Iranian writer Gholamhoseyn Sa’edi’s authoritative statements on the Āvāre, who connotes ultimately deracinated exiles (cf. Karimi-Hakkak 1991: 276; Naficy 1993: 229). Even though exile typically expresses a state of ‘limbo’, it is ‘situatedness’ that moulds migrants’ communication. In summary, the above findings portray Dutch–Iranian online networks as a reflection of offline community structures; they demonstrate that national hyperlinks remain relatively important; and they show that transnational links, far from alleviating spatial constraints, follow national patterns in their sectoral distribution. Diaspora dynamics extend transnational ties beyond the dual country-of-origin and -settlement matrix of analysis, but nation-state boundaries anchor Dutch–Iranian virtual interaction in territorially circumscribed physical space. Matthijs van den Bos and Liza Nell 208 © 2006 The Author(s) Turkish–Kurdish national, diaspora and transnational web surfing Turkish–Kurdish web surfing in the Netherlands reflects national, diasporic and transnational activities of offline communities rather than unbounded virtual communication. Moreover, Turkish–Kurdish online and offline activities demonstrate ethnic, political and geographic diversity – as opposed to broad, unitary identities – in the country of settlement, the diaspora and the (ancestral) country of origin. The core data on territorial bounds to Kurdish–Turkish web surfing consist of 21 interviews with Turkish immigrants in the Netherlands identifying themselves as Kurdish, conducted in spring 2004. The sample includes twelve first-generation, six ‘in between’, and three second-generation Kurds. Seventeen of the 21 claimed to use the Internet frequently. Only 14 professed to visit websites relating to the (ancestral) country of origin. These 14 respondents, and the 25 websites they mentioned, are covered here. The national, diasporic and transnational reach of websites will be examined through their (1) hosting countries, (2) target groups, and (3) announcements of offline events, such as demonstrations and festivities. Twenty of the 25 mentioned sites are hosted outside Turkey. Seven of these are national ‘immigrant’ sites – hosted in the Netherlands and targeting Kurds living there. The remaining 13 sites may be considered ‘diaspora’ sites hosted in European countries other than the Netherlands or Turkey. Three sites are hosted in Turkey and are thus ‘nationally Turkish’; surfing these websites from the Netherlands indicates transnational activity. The host country of two websites is unknown. Immigrant websites focus on ethnic or national groups from the country of origin in the Netherlands – not merely a virtual community. Two immigrant websites are extensions of offline organizations, three are discussion forums and two are information sites. Didf.nl and Ksvn.nl are extensions of Holanda Demokratik Isçi Dernekleri Federasyonu, DIDF (Federation of Democratic Organizations of Workers from Turkey) and the Koerdische Studenten Vereniging Nederland, KVSN (Kurdish Student Association in the Netherlands) respectively. DIDF is an organization for people from Turkey in the Netherlands and thus not a Turkish organization. The organization has Turkish and Turkish–Kurdish members, and organizes debates on political issues with a Kurdish federation in the Netherlands. KVSN is a student association for Kurdish students in the Netherlands. In contrast to DIDF, it appeals to a broader ethnic group (Kurds from all countries, for example Syria, Iran or Iraq) instead of a narrower national group (Turkish Kurds). Its website announces the organization’s activities, including lectures on Kurdish topics and excursions to Roj TV, the Kurdish broadcasting service in Belgium. Kurdish forums in the Netherlands emphasize belonging to a broad Kurdish group settled in the Netherlands, while Turkish forums, as far as is known, do not emphasize ethnicity but a common national background of persons (also) settled in the Netherlands. Respondents mention three forums: Groups.msn.com/Ksvnadam, Koerdistan.nl and Turkishforum.nl. Although KSVN has a forum on its website, there Territorial bounds to virtual space © 2006 The Author(s) 209 is also a chat group on Msn.nl for members of KVSN. Koerdistan.nl – the most popular website (mentioned five times) – is a forum for Kurds in the Netherlands: its ‘netiquette’ says the site is Dutch–Kurdish and members’ input should be either Dutch or Kurdish. The site’s events section announces Kurdish cultural, social and political activities in the Netherlands such as Newroz (Kurdish New Year) celebrations. Turkishforum.nl targets ‘public opinion’ and Turkish communities worldwide. The site’s aim is twofold: to improve understanding of Turkey by informing ‘public opinion about the realities of the world with regard to Turks’, and to distribute information on Turkish communal events. It publicizes events for (formerly) national Turks whose identity may become ethnic in host countries. ‘Worldwide’ suggests that they not only focus on labour migrants and their descendants in Europe, but also professionals, for instance in the USA. The two remaining websites focusing on immigrants in the Netherlands are information sites: Turksuitgaan.nl and Turkijepagina.nl. The first advertises Turkish parties; the second contains recreational information on Turkey. Neither explicitly refers to ethnicity – the common ground is national origin and national settlement. Some 13 websites are hosted from European countries other than the Netherlands. Two are offline organizations’ websites, Halkevi.com and Kongra-Gel.com, hosted from the UK and Belgium respectively. Halkevi (‘people’s house’) is a Kurdish and Turkish community centre in London that ‘welcome[s] and provide[s] support for Kurdish and Turkish refugees’. Halkevi.com, then, is an immigrant website for ethnic and national Turks and ethnic Kurds from Turkey in London. It announces Newroz activities in London and the organization’s own activities. Kongra Gel (People’s Congress of Kurdistan, formerly known as the Kurdistan Workers Party Partiya Karkeren Kurdistan – PKK) is a Turkish–Kurdish political organization in exile in Europe aiming at political change in Turkey, such as official recognition of Kurds as an ethnic group in Turkey. The website of the organization, forbidden in Turkey, focuses on ethnic Kurds from Turkey in Europe and Turkey. Diaspora websites that are extensions of print media are directly or indirectly connected to political movements in exile. Being in exile from and aspiring to political change in Turkey, they write specifically for Kurds from Turkey – not for all Kurds. Four websites fit this profile: Sterkaciwan.com, Ozgurpolitika.com, Rizgari.com, and Birnebun.com. Birnebun.com targets a particular Turkish–Kurdish group and is the website of a magazine printed in Sweden. Bîrnebûn means ‘not forgetting’ and focuses on Kurds with ancestors deported to central Anatolia in the sixteenth and seventeenth centuries (Sykes 1908). Having lived in Anatolia for several hundred years, this group still preserves its culture and language (Kurmanci). Özgur Politika (‘Independent Politics’) is printed in Germany and covers news about the world, Turkey and Kurdish and Turkish people in Europe. Both Sterkaciwan (‘Star of Youth’) and Rizgari (‘Liberty’) are online extensions of (former) magazines of political organizations in exile. Zazaradio.de and Nefel.org target another ethnic group: Zaza. Zaza speakers have always been considered Kurds by the Turkish state and other Kurdish movements, even if not all of them identify as Kurdish themselves (van Bruinessen 1994/2005). Matthijs van den Bos and Liza Nell 210 © 2006 The Author(s) Radiozaza.de offers music and information about Zaza Kurds in Turkey and announces concerts all over Europe. The site also offers forums for Zazas from the city of Dersim (officially Tunceli) and the nearby Munzur Mountains. Here, ethnic identity is connected with a specific locality and the Turkish tradition of organizing around the city or region of origin is reflected online. Finally, five websites hosted neither from the Netherlands nor Turkey provide online news (Amude.com, Kerkuk-Kurdistan.com, Kurdishmedia.com, Kurdistanpost.com and Nasname.com). No webmaster describes the target group as ‘Turkish Kurds’ or people from ‘Northern Kurdistan’ – they refer to Kurds and ‘Kurdistan’ as a whole (including parts of Turkey, Iran, Iraq and Syria). However, the names of sites, the contents of their articles, the main language on their homepages and the languages in the toolbar make the websites more accessible and interesting for specific geographical, political or language groups. For example, Kerkuk-Kurdistan.com mostly covers news from Iraq. At the same time, it pays much attention to Swedish politics, Kurds in Sweden and announcements of Kurdish events in Sweden – where the site is hosted. The three websites hosted from Turkey (Tkp.org.tr, Welat.com and Turkuler.com) do not announce events for Kurdish or Turkish people in Europe: here, respondents ‘surf transnationally’. In sum, Turkish–Kurdish national web surfing on immigrant websites reflects settlement in the Netherlands even while the sites target a broad Kurdish ethnic group or a common (former) nationality. Surfing on diaspora sites extending offline organizations reveals ethnic Kurdish identities connected with regional or national spaces of origin. Here, being an immigrant or refugee in a (particular) European country defines common ground. Finally, in transnational surfing, common nationality provides the basis of identification and sometimes allows for ethnic identification. The fact that respondents mention several types of websites – for example, national (relating to country of settlement), diasporic and transnational varieties, indicates that territorially anchored identities are not mutually exclusive. Internet usage shows the multi-layered nature of identities among Turkish Kurds in the Netherlands. One respondent, for instance, consults the websites Özgurpolitica.com, Radiozaza.de and Koerdistan.nl, indicating identification with Turkish Kurds, Zaza Kurds from Dersim and Kurds in the Netherlands, respectively. Surfing on websites targeting different ethnic, national or political groups fulfils an informational need about ‘one’s’ specific ethnic or political groups. These identities are strongly related to one’s origins in a specific local, regional or national environment. Hosting of and surfing on Kurdish websites by immigrants in the Netherlands indicates how national boundaries, or territorially fixed authority, touch on and mould Turkish–Kurdish Internet practice. These facts of territoriality contradict widely held claims about the Internet’s unmediated ‘virtuality’. As for territorially fixed authority, home countries’ restrictive policies force diasporas to host controversial websites from abroad. Here, the Internet differs little from print media even if it is often thought of as an anonymous, unbounded ‘safe space’ for expression. Closer scrutiny confronts the observer with territorial bounds to hosting and surfing at various levels: the Internet is anything but a ‘placeless space’ Territorial bounds to virtual space © 2006 The Author(s) 211 that is characterized by flows functionally independent of locality (Castells 2000: 14). Respondents’ online activities evidence the importance of geographical, cultural and political diversity among them. Rather than amalgamating Turkish–Kurdish groups with a similar ethnic background into a homogeneous ‘cyber-nation’, the Internet provides ample room for reproducing multi-layered and distinct identities. Modes of territoriality Both Turkish–Kurdish and Iranian migrants in the Netherlands interact online in a variety of ways, within the Netherlands’ borders, transnationally between countries of origin and settlement, or in the diaspora. But all the resulting patterns of communication reflect territoriality and their spatial context(s) of production. In this section, we examine whether territoriality varies according to scale only – as it does in the above cases – or may at the same time involve substantial levels of differentiation as well. While local, regional and national spaces generally mediate Turkish–Kurdish and Iranian Internet practice in the Netherlands, we consider site category and generation to chart their bearings on the particulars of territoriality. Forums allow for instant interaction between users: all that is usually required is registration. Conventional websites are often stricter about identity, with declarations of intent and about pages, and more singular in content. These differences beg the question of how individuals on the one hand and organizations on the other relate to conventional sites and forums, respectively. Given that instant interaction defines individuals’ motives for an online presence more than those of organizations, and the establishment of stable identities is more important for organizations’ online presence than for individuals, one is more likely to find forums as individuals’ spaces and conventional sites as virtual extensions of offline organizations. As for social implications of the difference for organizational links between sites and forums, one probable factor is generation. Karim writes on the effects that diasporic networks are having on nation-states, particularly in the light of the emergence of digital technologies: ‘While it appears that older immigrants prefer content exclusively related to their cultural backgrounds, younger ones seem to want a broader variety’ (1999: 15), namely they favour multi-layered identities. Similarly, the literature on the immigration process generally holds that first-generation migrants are more likely to organize in ethnic organizations, whereas the second generation are more likely to develop cross-ethnic ties (Vermeulen 2005: 33). Whereas firstgeneration migrants’ lives online may reveal clear extensions of offline networks, the second generation often reflects these networks in more subtle ways. Do secondgeneration individuals host partially sovereign online communities that pivot on hyphenated identities? Do these differences correspond with the usage of forums as opposed to conventional sites? Dutch–Iranian sites and forums The previously outlined research on Dutch–Iranian cyberspace did not differentiate between types of sites and pages. It is nevertheless likely that one deals with very Matthijs van den Bos and Liza Nell 212 © 2006 The Author(s) different social phenomena in an adolescent’s homepage and a professional organization’s web forum – for any of the reasons indicated above. Some of the largest Dutch–Iranian forums appear on Isan.nl (hosted by the Iranian Scientific Association in the Netherlands) and Iranl.com (which stands for ‘Iranl.communicate’ and is hosted by a Dutch–Iranian student group). Board members of these forums play key roles in Dutch–Iranian communal life. There are many ways to analyse the forums but the most obvious manner was to examine the percentage of messages representing an organizational link to offline settings. Both direct postings by organizations and postings on behalf or organizations by others were included. On Isan.nl on 8 March 2005, only two out of 451 messages, or 0.4 per cent, represented an organization (Club Perzie [Perzië] and Vereniging van Iraanse Vrouwen) (Figure 1). This compared with 191 out of 3675 messages, or 5.2 per cent, on Iranl.com (8–18 March). One organization (Studentfile, Iranian Students in the Netherlands) accounted for most (179) of these messages. When correcting for this and taking into account numbers of organizations instead of numbers of messages with organizational links, a lower figure results: ten out of 3675, or 0.3 per cent. These findings for organizational links on Isan.nl and Iranl.com compare with at least 25 out of 70 sites, or 35.7 per cent, for conventional sites. When checking for organizational linkages, that is, striking differences in ways of relating to offline settings appear between web forums on the one hand and average sites on the other. Dutch–Iranian forums could be seen as a natural habitat for second-generation Iranians who have moved beyond the realm of traditional migrant organizations. This assumption, however, does not stand up to scrutiny. Although age data are unavailable for message posters, circumstantial data shed light on the age factor. Some 84 per cent of the ISAN organizational board belong to the first generation; this applies to 75 per cent of ISAN members; and to 50 to 40 per cent who visit ISAN social events. Figure 1: Network medium and generation

2 Figures and Tables

Cite this paper

@inproceedings{Nell2006TerritorialBT, title={Territorial bounds to virtual space: transnational online and offline networks of Iranian and Turkish–Kurdish immigrants in the Netherlands MATTHIJS VAN DEN BOS AND LIZA NELL}, author={Liza Nell}, year={2006} }