The Making of Tamazgha in France: Territorialities of an Amazigh Diaspora-Assemblage

ABSTRACT This article uses an account of the Amazigh/Berber (pl. Imazighen) diaspora in France to demonstrate how approaching diaspora using assemblage thinking can help us to understand its emergent territorialities. It argues that the discourses and practices of association members frame the spaces of ‘Tamazgha’ (the ‘land of the Imazighen’), at the same time as framing the internal and external boundaries of the diaspora itself. Using assemblage thinking alongside theoretical tools from social movement theory and the study of nationalism, this article underlines how the diaspora comes into being through emergent relationships between human and non-human actors. Drawing on recent fieldwork, it outlines how ideas of Amazigh nationhood have developed and continue to develop in the process of this diaspora-assemblage, rather than simply arising from either a putative traditional ‘homeland’, state institutions, or being the invention of a purely intellectual or political project. Beginning with the ideal of a unified pan-Amazigh imaginative geography of ‘Tamazgha’, the Amazigh diaspora’s territorialities are now becoming more fragmented and regionally specific. As the Amazigh diaspora-assemblage develops within wider geopolitical assemblages, its new nationalisms not only challenge the existing state-territory configuration in North Africa, but also signal a changing sense of group identity for diaspora Imazighen themselves.

diaspora; hundreds of associations and other organisations group together thousands of individuals who are united by a common interest in the preservation and promotion of Amazighité, that is, "the collective consciousness of Amazigh being" (Almasude 2014, 132). Indigenous to North Africa and distinct from Arabic, Amazigh language (Tamazight) and culture are under threat in the eyes of the diaspora's leaders, and are portrayed as such on the world stage. The discourses and practices of association members frame the spaces of 'Tamazgha' (the 'land of the Imazighen'), at the same time as framing the internal and external boundaries of the diaspora itself. Using assemblage thinking alongside theoretical tools from social movement theory and the study of nationalism, this article underlines how the diaspora comes into being through emergent relationships between human and non-human actors. Drawing on recent fieldwork, it outlines how ideas of Amazigh nationhood have developed and continue to develop in the process of this diaspora-assemblage, rather than simply arising from either a putative traditional 'homeland', state institutions, or being the invention of a purely intellectual or political project.
This diaspora-assemblage was the subject of mixed-method qualitative research including participant observation in Berber/Amazigh association activities, online ethnography using social media and activist websites, analysis of secondary documentation and literature, and interviews with over 60 Berber/Amazigh association leaders and activists in Paris and Marseille, between 2015 and 2019. Ethical approval was obtained from the Cambridge University Department of Geography Ethics Review Group prior to fieldwork. The locus of the research project was cultural and political associations, as the sites and moments where diaspora is articulated, rather than focusing on individuals, groups or communities. Conservative estimates put the number of Amazigh people living in France at over 1 million (Gordon and Grimes 2005), with smaller but nonetheless significant Amazigh diaspora populations elsewhere (most Chleuh and Riffian diaspora activity is based in the Netherlands and Belgium). In France, however, Kabyle-dominated Amazigh associations have become a staple of local associative life (see Figure 1 for an overview of Amazigh sub-groups). Politically engaged Kabyle performers such as Idir, Lounis Aït Menguellet and more recently Allaoua have attracted crowds of thousands to France's biggest venues, and Paris-based Berbère Radio Télévision (BRTV), started in 2000, broadcasts to all of North Africa, Europe, and Canada. These different structures and events are elements of an emergent Amazigh diaspora -an assemblage of people, organisations and institutions that defines itself by its articulation of Amazighité in contrast to, above or alongside other markers of identity, be they religious (Muslim), national (Algerian/Moroccan/French), or political (Left/Right).
Underpinning this article is a theoretical approach to diaspora and nation as contingent and processual, reliant on the constant emergence of geopolitical assemblages. As Sean Carter stressed, "given the increased capacity for diasporas to 'act at a distance' [. . .] we need to reconsider the ways that we think about the nation and its territorialities, as well as diaspora and its territorialities" (2005,61). Whilst diaspora and nation are closely related, they are not homologous. The Amazigh diaspora and national consciousness have developed in tandem historically through scholar-activism, and continue to be constituted through an assemblage of people, organisations, ideologies and materials that frame and populate a shared sense of Amazighité.
After establishing the theoretical grounding of this research, the article examines the specifics of the diaspora's territorialities, and how they are changing as emergent nationalism(s) progress from scholarly and literary work of an intellectual elite to the practices and politics of popular mobilisation. Beginning with a strong ideal of a unified pan-Amazigh imaginative geography of 'Tamazgha', the Amazigh diaspora's territorialities are now becoming more fragmented and regionally specific, focused on 'Kabylia', 'Chawiland' or the 'Rif' as emergent national territories. As the Amazigh diaspora matures and responds to geopolitical developments in North African states and globally, its new nationalisms not only challenge the existing state-territory configuration in North Africa but also signal a changing sense of group identity for diaspora Imazighen themselves.

The Geopolitics of Nationalism and Diaspora
Scholarly uses of the term 'diaspora' range widely across and often within disciplines (Brubaker 2005). Two common uses of the term -diaspora as a reified generality, a metaphor for cultural difference and hybridity that lies outside the dominant cultural narrative (e.g. Brah 1996;Hall 1990) and diaspora as the population of a given nation-state residing outside its borders (e.g. Collyer and King 2015;Gamlen 2008;Koinova 2010) -are inadequate to describe the Amazigh diaspora. It has structures and processes grounded in place and in materiality, beyond merely a non-essentialised narrative of individual identity, but it is not defined and structured by an Amazigh state as such a state does not exist. Studying its geopolitics provides us with an opportunity to further explore an approach to diaspora using assemblage thinking. Assemblage thinking has become widely used in human geography in recent years, and whilst heterogenous in its usages and applications, takes as a central commitment a recognition of the complexity of human/non-human, nature/culture, and material/semiotic formations without reducing them to micro-or macroscale abstractions such as 'the State' or 'the Market' (Weir 2020). For some time, geographers have viewed diaspora as process, wherein the boundaries of nation, state, community and identity are (de)constructed (Mavroudi 2007). When conceptualised in this way, the making of an Amazigh diaspora can be analysed as part of a wider geopolitical context constantly undergoing an emergent process of assemblage (Dittmer 2014). As conceived by Barrineau (2015), "diaspora-as-assemblage" allows for an analysis of moments and places where the diaspora is constituted as a result of relations between human and nonhuman actors. Within assemblage thinking's "ontological requirement of immanence" (DeLanda 2016, 18), human and non-human actors are understood to continuously produce the identity of a community through their day-to-day interactions, such that "entities are never fully actualized within any of the relations that constitute an assemblage" (Anderson et al. 2012, 179). What this literature has so far failed to do is to link this approach to diaspora to emergent national territorialities beyond state structures and geopolitical technologies.
Because assemblage thinking eschews distinctions between types of social relation, the relations between actors that constitute a diaspora are not conceptually different to those that constitute a nation. In practice, as Brubaker has argued, diaspora should be treated as: a category of practice, project, claim and stance, rather than as a bounded group. The 'groupness' of putative diasporas, like that of putative 'nations', is precisely what is at stake. (2005,13) What assemblage thinking adds to this approach is a recognition of the complexity of relations between human and non-human actors, which this article will unpack through its focus on the key moments, processes, and practices that have made and continue to make the contemporary Amazigh diaspora. In doing so, it explains how these have developed Amazigh 'groupness' and territoriality within the geopolitics of diaspora and nation. In contrast with much of the work to date on geopolitical assemblages, this research does not foreground the agency of the state or state institutions, but rather consciously turns its attention to diaspora structures.
The article's theoretical basis for approaching the nation in this way lies with Miroslav Hroch's theorisation of the early phases of national movements, which foregrounds the role of intellectual work on the standardisation of minority languages and the development of literary culture (both key elements of Amazigh diasporic practice) in the making of nations (1985,2007). Basing his argument on the development of central European nationalist movements in the twentieth century, Hroch describes three 'phases': Phase A, where a small group of intellectuals devote themselves to scholarly enquiry into the language, history and 'traditional' culture and so on, of a non-dominant ethnic group; Phase B, where a new range of activists emerge, beginning to agitate for their compatriots to join them in forming a nation based on the scholarly enquiry of phase A; and phase C, where a majority of the population respond to the patriotic call and form a mass movement for national sovereignty. This teleological approach provides a framework within which the intellectual, socio-linguistic and socio-political development of Amazigh nationalism in its early stages can be analysed and understood (Maddy-Weitzman 2012). It does not rely on a pre-conceived national essence or ethnie (Smith 1986), or reduce nationalism to the politics of the state (Gellner 1983), but rather allows the nation-forming projects of the Amazigh movement, still in their early stages, to be understood as part of the diaspora-assemblage.
However, the ways in which nationalism itself changes territorialities, or develops in space itself, are not the focus of Hroch's analysis. His historical examples are of linguistic minorities that were also regional minorities, and so his framework assumes that these processes of national development take place territorialised within a preconceived idea of contiguous national territory. However, territoriality understood as a "system of mediated relations that determine the field of social geography, that is relationships of exteriority and alterity" (Raffestin 1986, 94), 1 is not the putative spatial property of a given entity occupying territory but an emergent process that can take place a distance from the territory in question, such as in diaspora.
Other historical examples of national movements that developed beyond the geographical borders of 'the nation' are, of course, numerous. Indeed, a prime example is that of the Algerian national movement, which was formed, supported and disputed by the diaspora in France from its very beginnings, including by the diaspora's many Amazigh members (Stora 1992). Another is the emergence of Arab nationalism(s) in diaspora the early twentieth century in relation to racialised legal structures, which facilitated access to a passport for cross-border travel to 'Syrian' subjects but not 'Turks' as the Ottoman Empire disintegrated, for example (Fahrenthold 2020). More recently, Kurdish political activism in the vast European diaspora has been compared with the Amazigh case (Bengio and Maddy-Weitzman 2013), highlighting the problems of transnational organisation both within diaspora and the putative homeland fragmented between states. In order to further underpin the conceptual framework of the diasporaassemblage in relation to the nation, however, I draw on the work of two geographers on Irish nationalism around the turn of the twentieth century. First, Adrian Mulligan (2002) calls for greater academic attention to what he calls the 'transnational roots of nationalism', and demonstrates the Irish-American diaspora's key role in ideologically and materially constituting Irish nationalism in nineteenth century. This was through an ongoing process of negotiation and articulation, most visible in the 'interstitial spaces' of diaspora where the naturalness of the nation was called into question, and through various material relations; the transatlantic telegraph, steamships, and newspapers intensifying connectivity and fostering a strong sense of national identity in the diaspora. Secondly, Gerry Kearns's writing on the different 'discourses of nationalism' (2003) held by nationalist leaders on the eve of Irish independence highlights the way that ideas of Irish nationhood mapped the nation differently; the historical nation was not the same as the religious nation, the political, or the linguistic nation. He demonstrates that each of these nationalisms in reality had parochial, at best regional dimensions, and were open to the charge of not capturing "the essence of the people of the territorial unit corresponding to the nation" (2003,219). The strength of the religious discourse of nationalism ended up determining, to a great extent, the territorial partition of the Island of Ireland at the moment of Irish independence.
Rather than focusing on nationalism as a political ideology, I incorporate it as part of a conceptual toolkit for understanding the diaspora-assemblage as an emergent social whole. In so doing I wish to avoid the spatial boundedness and teleological determinism of Hroch's schema whilst nonetheless accounting for the relative lack of engagement with social processes and practices in Mulligan's and Kearns's frameworks. In the study of the Amazigh movement, the diaspora is widely recognised as an important influence (Aïtel 2013;Silverstein 2013), a space of comparative freedom for political action, and a source of moral, political and financial support. Some have directly studied the ways in which the Amazigh diaspora negotiate their place in France in relation to the identity politics of the Amazigh movement 'back home' (Collyer 2008;Dirèche-Slimani 1997;Silverstein 2007), but none have specifically examined diaspora's re-territorialisations of 'home', and their effects on the constitution of Amazigh national identities.

Assembling the Diaspora, Assembling the Nation
The infrastructure of scholar-activism -the movement of intellectuals, the exchange of political ideas, the production, publication and circulation of knowledge -has long been at the core of the Amazigh diaspora-assemblage. The diaspora was key to the elaboration of Algerian nationalism in the first half of the twentieth century, but following the ostracisation of Berberism from the Algerian national project (known as the 'Crise Berbériste') and the marginalisation of Berber speakers in the early years of Algerian independence (Bouaziz 2012), it became an "indispensable location for the forging of modern Amazigh identity" (Bengio and Maddy-Weitzman 2013, 71). The Paris-based Académie Berbère revived and popularised the use of the ancient Tifinagh script (Chaker 1985), championed the use of the ethnonym 'Amazigh', established the Amazigh flag and Ж as identity symbols (Bessaoud 2000). It also published a print journal Imazighen, copies of which circulated amongst thousands of Imazighen in France before making their way to Algeria and Morocco (Aïtel 2013), to a readership starved of material to read in their mother tongue. For Amazigh activists and intellectuals engaging in academic and literary work on and in Tamazight, the (mostly Kabyle) diaspora in France afforded privileged spaces of meeting, collaboration, and 'memory work' (Maddy-Weitzman 2007), which, using Hroch's terminology, can be identified as phase A nationalism. Its tracts are frequently cited as a key formative inspiration for today's Amazigh diaspora leaders, and continue to be circulated within Amazigh associations.
Although it was dissolved in 1978, the Académie set a precedent for later forms of scholar-activism led from the diaspora: student members of the Groupe d'Études Berbères, including many linked to the exiled Algerian Front des Forces Socialistes (FFS), were later influential in leading the Berber Spring (Hirèche 2010); in the early 1990s, a group of students from INALCO formed the association Tamazgha which in 1994-5 was pivotal to the establishment of the World Amazigh Congress (CMA), the first transnational Amazigh NGO to explicitly include delegate representatives from across North Africa's Amazigh regions; in the late 1990s a group of Kabyle intellectuals, calling themselves the Groupe de Paris, met in a series of conferences where they debated and reasoned the basis for Kabyle autonomy that Ferhat Mehenni would act upon in 2001, forming the Movement for the Self-Determination of Kabylia (MAK). Today, these scholar-activist networks are not only influential within the transnational Amazigh movement, but play an active role in the assemblage of the Amazigh diaspora.
This active role includes the production and transmission of knowledge and cultural materials, the primary basis upon which the diaspora constitutes itself. The diaspora is highly fragmented, resistant to the elaboration of federative structures, and so this intellectual and cultural work is key to appealing to and maintaining individuals and groups within the social whole. Fundamentally, Amazigh association members are producers and consumers of Amazigh history, sociology and linguistics alongside music, poetry, theatre and literature. A large proportion of academic work on Amazigh issues is carried out in French universities and in the country's libraries and archives, which contain an unparalleled body of documents and sources gathered from across North Africa. Specialists in Amazighité, or berbérisants as they are known, engage in practices of knowledge production, curating, ordering and interpreting bodies of evidence and artefacts relating to Amazigh culture, language and history. In the Amazigh cultural associations, their work is interlinked with the cultural production of writers, singers, poets and artists to form the basis of a developing national culture. This is not to suggest that Amazighité is merely an invention of a clique of nostalgic North African intellectuals, but rather that the diaspora has been, and continues to be, a privileged space for its development and articulation along Mulligan's line of argument. On the one hand, the diaspora-assemblage is not interrupted by the intrusions of state institutions, its members arrested or its activities curtailed. On the other, the interstitial space of diaspora throws the very question of what it means 'to be Berber' into sharp relief, where the answer is not self-evident. Furthermore, university study has incentivised and supported diaspora scholar-activists to develop and define responses to this question that have repercussions for the diaspora's territoriality.
Today an influential scholar in Amazigh studies, Professor Chaker was among the first of these diaspora scholar-activists to gain his doctorate in Paris in 1973: All the berbérisants, whether Algerian or Moroccan, know each other in my generation. I studied in Paris or Aix with these people. [. . .] Just at the beginning of the '70s I was quite alone, but from '75 there were more and more. I was on their [PhD] juries, because I was around ten years ahead. [. . .] Fundamental questions for the future of the language were debated even at this time, about notation, standardisation. (Salem Chaker, Interview 24th July 2016) Chaker now oversees work on the Encyclopédie Berbère, an interdisciplinary, transnational collaborative work supported by the University of Aix-Marseille and INALCO started in 1984 by Gabriel Camps, which aims "to gather, collate and catalogue knowledge on the Berber world that was previously dispersed and fragmented". 2 In our interview, he explained that; Lines of communication, ease of travel, finance for academic activity without censorship by the state, and the politically 'neutral ground' of a French university allows Chaker and his colleagues to pull together and draw on the expertise of berbérisants from across North Africa. Terming the work "a sort of recuperation of history", Chaker is aware of the power that he and his colleagues have to establish an authoritative account of the Amazigh world. In Kearns's terms, they are developing the basis for discourses of the historical, the ethnic and the linguistic Amazigh nation, spatialised accordingly. In Hrochian terms, the diaspora is the primary location of Phase A Amazigh nationalism. For example, Chaker explained that although Islamic history was rarely included in the Encyclopédie Berbère, its authors have included an entry on Rahmaniyyah, one of the Muslim brotherhoods of pre-colonial North Africa, despite the encyclopaedia's policy of disentangling (and omitting) Islamic 'religion' from Amazigh 'culture'. The reason he gave was that the Rahmaniyyah was territorially concentrated on Kabylia, and so was intrinsically, primarily, Amazigh. Therefore, through its linking of scholars, artefacts and archival materials, this work is evidence of the Amazigh diasporaassemblage, cataloguing and categorising Amazighité and its territorialities.
Cultural associations often privilege the dissemination of academic research on Amazigh issues, inviting researchers at various stages in their careers to share their findings. For the past decade, a female group called Tiɣri uz _ ar, "roots of peace" have recorded, collected and performed the Amazigh songs and poems of their grandmothers' generation in Parisian associations and cafés. The Moroccan Amazigh association Tamaynut-France frequently hosts debates, inviting academic researchers to contribute, and in 2012 published an edited collection of research papers on the "Contributions of Moroccans to an emergent Franco-Berber identity" (Bouyaakoubi 2012). Dr. Farida Aït Ferroukh, a Kabyle anthropologist, has given weekly classes at the ACB on 'Berber Civilisation' for several years. She incorporated her own ethnographic and archival research into the course, which she had taught before at Paris VIII as a postgraduate module. Dr Aït Ferroukh stressed that it was important to include all parts of "the Berber world" in the course, to show that as well as exhibiting similarities it "is varied and diversified [. . .] It's a wealth. The Berber world is not only Kabyle" (Farida Aït Ferroukh, Interview, 17th June 2016).
As demonstrated in these examples, knowledge production and transmission forms part of the programme of many Amazigh associations in the diaspora. In researching, debating, and disseminating knowledge about the Amazigh world, diaspora actors develop, define and map it, continuously assembling an imagined community and imaginative geography. This Amazigh world is political, economic, cultural and social but it is also environmental, spiritual, and material. Through the associations, and now increasingly via social media (Harris 2020b), diaspora scholar-activists are uniquely positioned to engage in forging an emergent nation.
However, the question of territorial scale divides the opinions of berbérisants in the Amazigh diaspora. For some, Tamazight was and is a single language, whose regional variations (see Figure 1) are of secondary importance. For others, these regional variations should be recognised as languages in their own right. Debates over the language are very contentious in the diaspora associations -they often serve as a proxy for political debates about the diaspora's relationship with North Africa, framing its homeland geopolitics through divergent imaginative geographies.

Tamazgha and pan-Amazighité
Imaginative geographies of Tamazgha framed many of the research encounters that underpin this article. Most interviews started with each interviewee drawing a map of Tamazgha by hand. More important than what the detail of what they drew was how they responded to the request and what they said as they did so. Some respondents refused to make a mark on the coastline-only sketch of North Africa presented to them, deferring to the cartographic expertise and authority of others found in books or online. Most broadly sketched out the same outline -from Siwa in Egypt's Western Desert, to the Niger south of the Sahara, to the Canary Islands. Few reproduced the kind of border-crossing, fragmented detail of Figure 1. For almost all respondents, this was a historical territory, regrouping peoples with a common language and culture, who had begun to diverge from their common origins during the time of the ancient Numidian King Massinissa (c.150BC). Today, according to most interviewees, Tamazgha no longer exists, or is fragmented and distant, an imagined space rather than a lived territory: I asked Yazid [Ikdoumi] to draw a map of Tamazgha, which he did. He said that he was not sure if it had ever existed, that he does not have the dream of 'La Grande Berbèrie'. 'We cannot redraw history or geography to suit our dreams', he said. There is however a cultural and linguistic solidarity between Berbers, even in the diaspora. But it bothered him to draw a territorial map; 'Tamazgha is an idea, not a territory'. (Fieldnotes, 3rd April 2016) The power of this geographical idea within the Amazigh diaspora, and its transformation over time, was described in detail by other Amazigh leaders. Yella Houha, the founder of the Movement for Chawi Autonomy (MAC), explained as he sketched his map that Tamazgha was a recent political concept, ideologically bound up with pan-Amazigh discourse, and developed in the diaspora: "before the 70s, 'Amazigh' and 'Tamazgha' didn't make any sense [. . .] All that was the work of the Académie Berbère. Behind it all was a handful of Kabyles" (Yella Houha, Interview, 10th November 2015). For Kabyles during and after the Algerian War, although an awareness of a wider Berber world did exist, 'Berber' meant themselves (Temlali 2015). Defining themselves linguistically as non-Arabophones, this Berber-speaking community was regionally concentrated, sharing common speech, historical and social links (Amrouche 2009). In the diaspora, the predominately Kabyle scholar-activists of the Académie Berbère, influenced by ideas of decolonisation current on the French intellectual Left, began to insist on the use of the ethnonym 'Amazigh', meaning 'free man', arguing that a break was needed from the 'Berber' label perceived as having been imposed and carrying pejorative associations (Bessaoud 2000). Becoming 'Amazigh' was to be part of decolonising ideology, free from the legacy of French colonialism, but also the strong regional associations to Kabylia (Tilmatine 2015). Accordingly, it changed its official name to Agraw Imaziɣen in 1969 (Silverstein 2003), and through its journal Imazighen for a time dominated Amazigh print media. The imaginative geography of Amazighité gives the impression of a far larger reach and group size than the Kabyle-dominated core, and re-evokes an opposition between Arabs and Berbers across North Africa and the diaspora, the 'Berber Myth' that has its roots in colonial-era ethnography (Shatzmiller 1983), in what Merolla has argued represents a form of 'strategic essentialism' (2020). However, this essentialism was and remains something of a double-edged sword: as a Chawi speaker arriving in Paris in the 1970s, Houha was intrigued to hear that Tamazight, the language of the Berbers, was being taught at the University of Paris VIII. However, when he went to a class he realised that he understood little, and that the language was essentially Kabyle; "For me personally, the thing was that I was Chawi, but I hadn't understood 'Tamazight', so they were trying to teach me another language to my own" (Yella Houha, Interview, 10th November 2015).
When, in 1980, the Berber Spring erupted in Kabylia, the social movement's demands were inspired by the pan-Amazigh imaginative geography contested in the diaspora; under a pan-Amazigh flag designed in Paris, protestors called for the recognition of Tamazight as a language of the state, and Amazigh culture as Algerian culture (Hirèche 2010). As greater numbers of Amazigh from across North Africa's regions began to adhere to the Amazigh movement in the 1980s and 90s, many consciously imitating the Kabyle-dominated movement (Silverstein 2013), the Académie Berbère's imaginative geography of Tamazgha, formed through its diasporic scholar-activism, was put into practice. Phase A intellectual nationalism was translated to phase B activism and popular mobilisation 'on the ground' in multiple and often divergent ways. Eventually however, this assemblage led to a group of Paris-based activists succeeding in gathering Amazigh association representatives from across Tamazgha into a single transnational NGO, the Congrès Mondial Amazigh (CMA), in 1995.
Whereas prior Amazigh associations had been largely Berber in name and Kabyle in substance, the CMA consciously sought to be representative of the entire Amazigh world -from Siwa to the Canary Islands. Operating transnationally but based in Paris, the CMA today remains the main advocate for a single Tamazgha within the Amazigh movement. Although its constituent members usually represent a specific region or people within the Amazigh whole, the CMA outwardly represents a single 'Amazigh nation', notably when it challenges the representatives of North African states in international fora (Harris 2020a). The CMA General Secretary at the time, Belkacem Lounes, was clear in our interview: "for us it's simply that there is a territory called Tamazgha, a more or less shared language and history, and a common ideal of liberty" (Belkacem Lounes, Interview, 29th April 2016). The CMA's activists encourage cultural associations to cooperate on the basis of this shared Amazigh language. Invited to speak alongside Kabyle, Chawi and Mozabite nationalists at a Parisian Amazigh association in 2016, who as we will see purport another vision of Amazighité, Belkacem argued for the linguistic unity of Tamazgha, saying that the difficulties of inter-comprehension between Tamazight-speakers of different regions were mostly differences in accent and vocabulary, not evidence of several distinct Amazigh languages. This discourse, and the practice of linguistic inter-comprehension, is key to maintaining the imaginative geography of Tamazgha as a homogenous territory. Only by focusing on what united rather than divided them, Lounes argued, could the Imazighen come together to constitute a force that could confront the power of the North African states. Other ideologically pan-Amazigh associations which adopt similar practices, and were included in the study, include Tamazgha, Ameslay, Association Amazir d'Île-de-France and Taferka. Through such practices of inclusion of diverse Amazigh members, concurrent use of Tamazight in its various forms, and discursive adhesion to the idea of Amazigh cultural unity, these associations a part of the wider assemblage of a singular, Amazigh nation in the diaspora.
Whilst in the fora and texts of international institutions this assemblage is dominant, within the associations it remains intellectually driven, characterised by debates and scholarly presentations, and does not have a large popular base. Most associations are 'Berber' in name but in practice regroup members from a single linguistic group, most often Kabyle. As an example, a musician from Association Franco-Berbère de Pierrefitte explained that its members were "99% or 100% from Kabylia. We would have liked to have Moroccans or other Imazighen, but they join other associations, where they feel close" (Momo Kechimir, Interview, 20 th March 2016). Such associations, several of which come under the banner of the Coordination des Berbères de France (CBF), often network with other groups of different regional origins, most often in the realm of artistic performance. For example, the CBF's Kabyle-dominated centre in Drancy frequently invites Tuareg bands or Chleuh dancers to perform in their regional styles at pan-Amazigh festivals. In June 2016 a day-long festival in a Paris suburb was organised by three different Amazigh associations. A Tuareg activist and performer described such events as "important because they allow us all to share the problems we are facing [from across Tamazgha], and to fuse together more closely, to wield more force in our challenge to the states that oppress us" (Amoumene Hiadara, Interview, 1 st July 2016), exemplifying the political edge to this cultural activism. Such 'fusing together' in diasporic assemblage develops a singular sense of Amazigh groupness, territorially expressed in the imagination of Tamazgha, but reaches a wider audience, which recognises itself in the regional variations.
Such networking and collaboration takes work, time and investment on the part of association leaders and members, as regional differences in language and relationships with the state, religion, or politics mean that Amazigh associations mostly form around different regional groupings. Emphasising pan-Amazigh commonality and putting it into practice is a political choice, with consequences for the shape and scope of the nation imagined in diaspora. How, then, can we understand such a choice through assemblage thinking? Two answers emerge from the ethnographic data. Firstly, for politically connected groups such as the CBF seeking to gain political influence in France or the CMA seeking to represent Amazigh interests to North African states, casting a wide net over the Imazighen increases their assumed demographic size, strengthening the case for the direction of resources to promoting and preserving Amazigh culture. This broader umbrella also accommodates the formation of coalitions and alliances that are better able to share resources, expertise and mutual support in the name of a shared putative identity. Secondly, pan-Amazigh identity claims do not challenge the existing order of territorial sovereignty in North Africa, are instead broadly tolerated by those states' governments, and consequently are able to articulate with them. The suspicion of irredentism that has dogged assertions of Kabyle specificity since Algerian independence, for example, falls far less on groups articulating a pan-Berber or Amazigh groupness (Tilmatine 2015), widening their appeal for those who wish to avoid association with more radical elements of the Amazigh movement. The same applies for Riffians' relationships with other Moroccan Imazighen. This is because the territoriality of this pan-Amazigh ideology is very flexible, and as such Tamazgha remains a largely abstract territory or idea for many Amazigh activists, malleable to the geopolitical imagination of existing nation-states (Cornwell and Atia 2012). For example, the Tamazight teacher at the ACB, Belaïd Addi, was "sure [Tamazgha] will always exist, because now the states recognise the language" (Interview, 24th April 2016). Similarly, for the CBF's Mustapha Saadi Tamazgha is "a historical, sociological, linguistic reality" (Interview, 3rd May 2016), but not a "political" one. Such territorialities of the pan-Amazigh ideology do not challenge existing state paradigms in North Africa, but rather have been adapted to them, geographically rendered at the scale of the existing nation-states.
The Amazigh diaspora-assemblage, territorialised in relation to the imaginative geography of 'Tamazgha', has therefore developed over time and through various human and non-human relationships. However, as DeLanda argues, we need to include in a realist ontology not only the processes that produce the identity of a given social whole when it is born, but also the processes that maintain its identity through time. And we must also include the downward causal influence that wholes, once constituted, can exert on their parts. (2016,18) The following section details how the various 'parts' of the Amazigh diaspora have come in the past two decades to regionalise territorial referents and claim more parochial diaspora-national identities.

Kabylisation of the Amazigh Diaspora
Yes, [the diaspora] is Kabylising. Since 2002, the Tafsut Taberkant, the Black Spring. Since the eruption of the political question of autonomy for Kabylia. We've gone from 'Berber/Amazigh cultural' associations to Franco-associations and now it's more and more Kabyle [. . .] Because there was an evolution in the discourse that has permeated the real. (Kamel Saidi, Interview 7th July 2016) An increasingly common mode of pan-Amazigh diaspora assemblage marks a shift in its imaginative geographies and territorialities. Some interviewees called this shift 'Kabylisation', capturing both the particularity of Kabyle assertions of national identity and the more general segmentation of the Amazigh movement along ethno-linguistic lines (the word Kabyle comes from the Arabic Qaba'il, meaning 'tribes' was historically applied across the supposedly 'tribal' Amazigh peoples). Kabylisation has seen Chawis, Chleuhs, Riffians, Mozabites and other Amazigh groups asserting their linguistic, cultural and social specificity within the wider Amazigh whole. It is a rescaling of groupness to more parochial spaces and regions, which Kearns argues are the territorialisations of different discourses of nationalism (2003). Kabylisation in the Amazigh diaspora hinges on changing discourses of nationalism, relative to language, history, religion and sociology. Unlike the 'balkanisation' of the Yugoslavian diaspora in the USA observed by Carter (2005), where Serbian and Croatian diasporas split over the conflict at home, this fragmentation of diaspora territoriality has not arisen out of conflicts between Amazigh groups. Rather, as phase B Amazigh activism has put phase A nationalist frameworks developed by scholar-activists into relational networks and practices, they have been rescaled along lines of redefined ethno-linguistic homogeneity, both in the diaspora and North Africa. As the ethnographic data shows, their introduction into the assemblage of assemblages that is the Amazigh movement has produced new relations of exteriority, such that many activists now see themselves as part of diasporas based exclusively on their respective regions, as well as or instead of a singular Amazigh diaspora.
During the 1990s, a new wave of Amazigh activists and intellectuals came to France from Algeria, many fleeing the civil conflict that gripped the country. Many were Kabyles who had been active in politics, and seeking refuge in France. One figure who became a prominent leader among them was Ferhat Mehenni, a Kabyle singer whose songs calling for human rights and democracy were among the first examples of 'modern Kabyle song' in the 1970s. As a member of the new Kabyle-dominated political party the Rassemblement pour la Culture et la Démocratie (RCD), he led a school boycott in Kabylia in 1995 which led to the government's formal introduction of Tamazight to the school curriculum, but fled Algeria soon after. However, when the Black Spring of 2001 drew Kabyle youths into another round of violent confrontation with the state Mehenni re-entered politics, but uniquely for Kabylia, not for Algeria. We broke with the dogma of the single Amazigh nation, speaking the same language -it's a lie to say this. Maybe at the time of [the Académie Berbère] we needed to stand together, but every time has its ideology that hides the truth. From 1926 [formation of the Etoile d'Afrique du Nord, commonly identified as the beginning of Algerian national independence movement] the Kabyles were only invested in Algeria, but [from 2001] many began to see that it was leading us nowhere, this pan-Berberism (Ferhat Mehenni, Interview 25 th May 2016) Ferhat Mehenni announced the creation of the MAK, a political movement for Kabyle national independence, which is having profound effects on the wider Amazigh diaspora today.
This structure did not emerge in a vacuum. Kabyle autonomy was already being discussed by scholars and activists in the diaspora towards the end of the 1990s. The hope that a democratic transition could occur throughout Algeria that would allow Berberophones to assert their linguistic and cultural rights, which was strong at the end of the 1980s, was fading after years of civil violence (McDougall 2017). Following the assassination of celebrated singer and protest figure Matoub Lounes in 1998, Salem Chaker wrote a column in French broadsheet Le Monde entitled "For the linguistic autonomy of Kabylia", arguing that the "objective of the Berber movement can only be the recognition of the linguistic and cultural specificity of Kabylia, in the context of a large regional autonomy" 3 (Chaker 1998, 11). As well as coordinating the Encyclopédie Berbère, Chaker was one of a group of Kabyle intellectuals based in the diaspora who debated the case for Kabyle regional autonomy in the late 1990s and early 2000s, looking to forms of regional autonomy within Europe (e.g. in Switzerland, Belgium, the UK, Spain and Italy) for inspiration. In a series of meetings, this 'Groupe de Paris' and later the 'Cercle d'étude et de réflexion sur l'autonomie de la Kabylie' considered the political, economic, and social modalities of different forms of autonomy, and theorised the form that an eventual Kabyle autonomy could take. At a time when the diaspora's associative activities were largely framed by its members as pan-Amazigh (Pouessel 2010;Silverstein 2004), these scholar-activists elaborated an alternative framing with a different territoriality.
During the 'Black Spring', this alternative framing became a vector of mobilisation, as a new set of Kabyle associations and informal collectives were set up by diaspora activists to raise public awareness in France and internationally, to arrange material aid for protestors in Kabylia, and to assure their medical treatment in France. In contrast with the pan-Amazigh territoriality outlined in the previous section, this diaspora mobilisation was on the scale of Kabylia, an assumed cultural, political and territorial entity. A 'Kabyle' diaspora was evoked and shaped through the cooperation of individuals and associations emerging from within the wider Amazigh diaspora assemblageincluding the involvement of older village-based institutions operating in France since the first Kabyle immigrants arrived there (Dirèche-Slimani 1997), By the mid-2010s, this alternative framing was a central concern of the Kabyles of the Amazigh diaspora, as the MAK's call for independence sparked debate and controversy: we have never debated independence as much as we do today, and it's thanks to the upset caused by Ferhat Mehenni. It's shaking people up. It makes people laugh, people fear, but it's worth debating. (Chérif Benbouriche, Interview 18 th November 2015) Any use of the 'Kabyle' framing, regardless of the political affiliations of individual activists and associations, contributes to the distinctly Kabyle diaspora-assemblage. Whilst this assemblage should not be conflated with the MAK's project, it has certainly benefitted and been encouraged by the Kabyle nationalists. As such, explicitly "Kabyle" associations and events were frequently suspected of being aligned to the Kabyle nationalists. For example, the organisers of the Journée de la robe Kabyle in May 2016, which mobilised a range of indigenous Kabyle material cultural artefacts including both female and male dress, instruments and song (see Harris 2020a), felt the need to distance their event from the MAK; "The term 'Kabyle' is not in the pocket of the MAK. We don't have a complex about it, we're apolitical [. . .] The MAK is not in the business of profiting from every event that says 'Kabyle'. They're probably just proud that it's happening" Lounes 'Le Kabyle ', Interview, 26th May 2020 Similarly, at a cross-association planning meeting for the Semaine de la Kabylie in 2016, a week-long series of public, open-air events partly receiving municipal funding, one organiser clearly signalled that the event could have no political angle; " No RCD, no FFS, and especially no MAK" (Fieldnotes, 18th May 2016). Another organiser later told me, "If [the MAK] gain from [the Semaine de la Kabylie], good for them, but that's not our intention [. . .] Our event, in the end, is apolitical" Zoubir Ghanem, 31st May 2016. Of course, such claims to neutrality are in themselves political, but what is nonetheless significant is that the Kabyle framing, which twenty years ago, even in the diaspora, was seen as a radical political statement of opposition to the Algerian state (Dirèche-Slimani 1997), has become acceptable to the point that such an event today could claim to be politically neutral. Such events manifest the existence of Kabylia as an emergent social whole. Associations like the Parisbased Association de Jeunes Kabyles de France (AJKF), are explicit in regrouping Kabyles, and speaking and teaching Taqbaylit (Kabyle) rather than Tamazight. Moving on from the phase A scholar-activism of the previous decades, their members articulate a common-sense understanding of Kabyle groupness in their everyday practices. This popular adoption of the Kabyle frame represents a rescaling of 'homeland' for many in the Amazigh diaspora, and a shift in its diasporic territoriality -both 'up' from the individual village as mediated by family ties and village committees and 'down' from Tamazgha as mediated by the activism and intellectual productions of prior diaspora assocations and media.
Such assemblages of cultural associations, village committees and political movements within a Kabyle national framing mark a break with the prior territorialities of the Amazigh diaspora. This territorial frame is far more precise than that of Tamazgha -focusing on a single region, and in the case of the MAK calling for national self-determination at this scale. In 2015-2016 the MAK was in the process of drawing up its case for a referendum on independence, which it took as a published memorandum to the UN in September 2017 (Mehenni 2017). Part of this process, its leaders recognised, would require defining 'Kabylia' and 'Kabyles' as a nation. Research by proindependence university students in France's colonial archives produced cartographic evidence to support their historical claim that Kabylia was only annexed to Algeria by the French in 1857, being hitherto an independent territory and polity. For the MAK, this map ( Figure 2) is a piece of evidence that Kabylia was, for a time, territorially independent of Algeria, and could be again.
Maps are only part of the material culture and imaginative geography of Kabylia as a nation, of course. As Benedict Anderson put it, "changing apprehensions of space" are the necessary 'coordinate' of "changing apprehensions of time" (Anderson 2006, pxiv). Youcef Zirem's Histoire de Kabylie (2013) is an example of the popular historiographical reorientation being worked on by diaspora activists, writing Kabylia into history and territorialising it in the process. In his book, episodes of state-building or rebellion centred on the region such as the 'Kingdom of Koukou', the rebellion of 1871, or the FLN's Wilaya III during the Algerian war, are stitched together to as part of a Kabyle national narrative stretching back centuries. Following Anderson's observations, I suggest that such historical mappings are "designed to demonstrate, in [. . .] cartographic discourse, the antiquity of specific, tightly bounded territorial units" (Anderson 2006, 147). The antiquity and historical continuity of the Kabyle national territory is further documented online, through social media pages and groups based on Kabyles and Kabylia, such as 'Kabyles de Paris' or the 'National Geographic Kabyle'. These are popular places for several thousands of members to share all kinds of digital media that play a role in shaping and defining the contours of what Kabyle national identity is, and where its borders begin and end. As discussed more fully elsewhere (Harris 2020b), this online popular geopolitics is a key part of the MAK's appeal, communication strategy, and, they argue, a source of their legitimacy. Diaspora actors in this online mapping of Kabylia collectively and reflexively define a sense of groupness that reconfigures the territoriality of the Kabyle diaspora and therefore the Amazigh diaspora of which it forms a central part.
Whilst certainly the most prominent in the France-based Amazigh diaspora, the growth of Kabyle national sentiment is not alone in the wider Amazigh movement; other regionalist and nationalist projects have developed alongside it. Chief among them are the Riffian and Azawad independence movements, whose recent mobilisations on the ground in northern Morocco and Mali respectively form part of a far longer history of anti-colonial struggles for selfdetermination. Whilst less well-represented in the France-based Amazigh diaspora, the Riffian Mouvement 18 Septembre under Fikri El Azrak was created in 2014 in the Netherlands. Other nationalist Kabyle and Chawi leaders travelled to meet them, publicising their support and "natural diplomatic ties" as fellow Amazigh. These Chawi leaders came from the Paris-based MAC (Chawi), whose handful of activists have been companions of the MAK since 2001. The MAC's founder, Yella Houha, claimed even to have come to an agreement with the MAK's leaders as to the frontier between each nation's eventual sovereign territory. Similarly, to the Kabyle case, a set of websites and social media pages have been created over the past 10-15 years, dedicated to elaborating a Chawi national sense of groupness (e.g. inumiden.com, chawinet.org, ichawiyenautrement.fr). The MAC portrays the Chawi nation as needing to defend itself both from the grip of Algerian nationalism and the 'Fus dug fusisme' ('hand in handism') of pan-Amazigh ideology (Fieldnotes, 14th June 2016). Then, in 2014, a group of Mozabite activists led by Dr. Kamaleddine Fekhar began to argue for greater autonomy for the Mzab, with the support of the MAK and the MAC expressed in communiqués posted online. Following the arrest of the majority of this group, those that claimed asylum in France were taken in by MAC and MAK activists, housed and given financial and legal help, and supported in continuing their activism on behalf of Dr. Fekhar and the Mzab under the name Izmulen. Drawing on the localised national territorialities that are becoming increasingly widespread in the Kabyle-dominated Amazigh diaspora, these autonomist movements are symptomatic of a broad change in the organisation of the Amazigh diaspora, and its imaginative geographies of Tamazgha.
Even the CMA has had to adapt its organisational structure to fit this changing territorial frame. Until 2015 the CMA membership elected individuals to represent each nation-state in North Africa. However, the abortive Azawad war of independence fought in Northern Mali from 2012 meant that the CMA's Tuareg delegates came to the CMA's 2015 general assembly arguing that for them; "there was no longer Mali, there was a state called Azawad, even if it's contested. And so the Kabyles, who were well represented at Agadir, said; 'us too, us too!'. That's how the Kabyles proposed and firmly defended the idea of not having Algerian delegates but delegates by territory of Algeria. Kabylia is not yet like Azawad, but under the circumstances it was difficult to refuse them, and the General Assembly accepted [their proposal]. So, now there is a Kabyle vicepresident, and a Mzab vice-president. There's only no Chawi vice-president because there was no Chawi representative". Belkacem Lounes, Interview, 29th April 2016 This shift in the organisation of the diaspora and its associations is happening as ideas of nationhood translate from Hrochian phase A intellectual work into phase B activism. As has been argued elsewhere (Harris 2020b), much of this nationalist activism is populist in that it opposes 'the people' to 'corrupt elites'. The linguistic, cultural, historical and geographical differences between Amazigh peoples that are downplayed by pan-Amazigh ideologues and the governments of North African states, increasingly form the basis of an emergent set of national identities formed around a denser set of social relationships, a shared language, and a parochial geography.
Whereas the pan-Amazigh vision of Tamazgha could be accommodated within the borders of the existing states of North Africa, the new nationalist claims of Amazigh independence movements directly challenge them. Safely at a distance from the states whose territories they contest, leaders in the diaspora such as Ferhat Mehenni, Yella Houha and Fikri El Azrak elaborate arguments for national independence that appeal to parochial territories and populations, situations and politics. Regardless of whether any of these political movements succeed in gaining national independence, the shape and organisation of the diaspora-assemblage has changed. The imaginative geographies through which the diaspora understands its homeland(s), and the territorialities that are their consequents, shape the groupness of the diaspora, assembled through associations, networks and practices.

Conclusion
This article has demonstrated that ideas of Amazigh nationhood have developed and continue to develop in the process of diaspora-assemblage, territorialised through material practices, moments and places that co-produce the imaginative geography of 'Tamazgha'. Through knowledge production and cultural activism related to the Tamazight language, and drawing on Amazigh cultural and political institutions, a small group of mainly Kabyle scholars and activists located in France developed an ideology of a single Amazigh nation, which in the 1970s, 80s and 90s framed the groupness and territoriality of the Amazigh diaspora. This framing enlarged the scope of Amazigh movement, strengthening the case for promoting and preserving Amazigh culture in France and in North Africa. The sovereignty of North African states was left unchallenged, and Amazigh activists rather called for the state to recognise and safeguard their language and culture. However, in the last two decades this broader assemblage has diversified along ethno-linguistic lines and the diaspora is undergoing a process of Kabylisation. The ideal of creating a standardised, singular Tamazight shared by all of Tamazgha, as evidenced by the data presented in this article, is now seen by many as an impossible dream that does not articulate with the material realities of their emergent national groupness. Instead, diaspora associations and individuals in France are increasingly likely to identify as 'Kabyle', 'Chawi' etc., whilst 'Amazigh' remains an umbrella term.
Approaching the diaspora as assemblage through a focus on structures and processes grounded in place and in materiality has allowed this analysis to link scholar activism and social mobilisation outside of the host-state/sending-state paradigm. This approach will be useful to other scholarship focusing on the territorialities of socio-political mobilisations outside the confines of the state, particularly in the postcolonial context where Indigenous peoples and stateless nations confront and subvert colonial-era state boundaries. It could equally lead to a more granular account of the geopolitics of state diaspora strategies and institutions as transnational mobility and interconnection appear to become increasingly subject to surveillance and regulation.