Are they nomads, travellers or Roma? An analysis of the multiple effects of naming assemblages

What is the difference between the terms ‘ Roma ’ , ‘ gypsies ’ , ‘ nomads ’ and ‘ Travellers ’ ? These are a few of the names that are used to refer to the Roma minority in scholarly research, political speeches and the media. Most of the Romani studies literature on Roma labels and the state ’ s categorisation underscores how these often derogatory denominations re ﬂ ect the widespread stigmatisation of these people and, in turn, perpetuate regimes of exclusion and segregation. However, this literature implicitly conceives of language as purely functional to exclusion, overlooking the ways in which the construction and use of these labels have also created the conditions for the emergence of practices of resistance. This limitation is mainly due to the fact that these works follow a Foucauldian approach, which tends to overemphasise the importance of dominant discourses subjecting the individual, and to downplay the presence of generative and creative practices. I suggest integrating this approach with the notion of ‘ assemblage ’ as developed by Deleuze and Guattari, which entails both ordering and territorialising dynamics together with destabilising moves. By adopting this lens, the paper discusses the effects of two different Roma naming assemblages: on the one hand, the glossary published by the Council of Europe (CoE) that carefully de ﬁ nes and differentiates all the terms used for the Roma, and, on the other, the French and Italian governments ’ discourses that ambiguously lump together all these different denominations. Although at ﬁ rst sight it may appear that the latter bolsters discriminatory and segregating policies, while the former supports more inclusionary measures, by drawing on policy-documents analysis and in-depth interviews with pro-Roma advocacy group members, I show that both these naming assemblages actually produce exclusionary as well as resisting effects.


Introduction
The ancient Latins used to say that the name is an omen of one's destiny. This is no exception for the Roma people, 1 whose different and often offensive denominations have perpetuated their exclusion and marginalisation in history. Several scholars of Romani studies underscored the crucial role played by labelling practices and the state's categorieslike 'nomads', 'Travellers' or 'gypsies'in the legitimisation of segregation and confinement. However, little has been said about how these names create the conditions for practices of resistance, thus running the risk of reproducing an image of the Roma as passive objects of the state's discourses.
This paper problematises the literature on the names of the Roma, which is often too concerned with the ways in which naming practices strategically serve marginalising policies. Although I do not deny these discriminatory effects, I also think that such a focus might lead to overlooking the unexpected and creative resistance practices that emerge within these very discourses. By drawing on the notion of 'assemblage' (Deleuze and Guattari 1987), whereby discursive articulations produce heterogeneous and multiple results, I compare the discourses on the names of the Roma of the Council of Europe and of the French and Italian governments in order to show how the construction and the use of a name does not have unidirectional and straightforward effects. The argument is developed in three sections: the first and the second discuss the limitations of the literature on the Roma names and introduce the notion of assemblage; the third part aims to show the multiple effects of Roma names and draws on fieldwork I carried out in Italy in 2013 and France in 2014, during which I conducted an analysis of policy documents and in-depth interviews with pro-Roma advocacy group members.
While this paper does not indicate the right name for the Roma, it investigates the consequences of the use of these names. There are no right or wrong names, only different effectsand often these effects are unexpected. This paper aims to contribute to the growing literature on assemblage thinking in geographical and urban research (Anderson and McFarlane 2011;Lancione 2013;Legg 2011;McFarlane 2011) by connecting it to the research on the marginalisation of the Roma, in order to show how Roma naming practices play an active role both in processes of policymaking and resistance.
Competing glossaries: the exclusionary effects of the names of the Roma If one were to look for the correct names of the Roma people, the Romani studies literature would not be of much help. Multiple denominations referring to these groups can, indeed, be found in scholarly works, for example Gypsies and other itinerant groups (Lucassen et al. 1998), The traveller-gypsies (Okely 1983) and Roma and gypsy-travellers (Bancroft 2005). Although the emergence of a discourse on the inclusion of the Roma minority in the last 20 years made the non-derogatory endonym 'Roma' prevail, there are still several other terms used to refer to these groups. Due to the plurality of terms used, in 2006 the Council of Europe (CoE) published on its website a regularly updated glossary listing all these different designations. For instance, 'Roma' is used to refer to an Eastern European minority speaking Romani language, while 'Travellers' are defined as itinerant groups, not Romani-speaking and mainly living in the UK. There are other groups that speak languages with some influence of Romani, such as the 'Yanish' in Switzerland, the 'Sinti' (also called 'Manouches') originally from Germany and the 'Kal e' in Spain. The official CoE terminology has also varied in the last 40 years, including names such as 'Gypsies and other travellers', 'nomads', 'Rroma (Gypsies)' and 'Roma'. As shown by these examples, there is a proliferation of names and definitions around these groups that are today collectively called by the CoE 'Roma'indeed the official term until 2010 was 'Roma and Travellers'. This term has been also adopted by the Italian National Anti-Racial Discrimination Office that wrote the National Strategy for the Inclusion of Roma, Sinti and Caminanti Communities, in which the term 'nomad' is criticised for being derogatory and not corresponding to the real habits of the Roma groups.
However, the CoE glossary is just one way of making sense of the different names of the Roma. There are other naming practices that, contrary to the CoE, conflate these terms, using them as synonyms. For example, during a French parliamentary meeting about domestic security held on 15 November 2002, Nicholas Sarkozy, Minister of the Interior at that time, stated that '"Nomads", "French Travellers", "Roma", names are not important'. The interchangeable use of these terms characterises also the discourse on the Roma in Italy, where the political and legal frameworks and the media employ the terms 'nomads', 'Roma' and 'Sinti' as synonyms and in an ambiguous way. At the legal level, for instance, the Nomads Emergency Decree, adopted in 2008 and mainly targeting the Roma communities, lumped together the terms 'nomads' and 'immigrants'.
The CoE glossary and Sarkozy's statement epitomise two different conceptualisations of the naming of the Roma: while the former draws clear distinctions between different names, the latter lumps them together as synonyms. The question underpinning both the CoE glossary and Sarkozy's statement is the identification of a correct terminology. However, as argued by Puar (2007) in her analysis of the naming of the Sikh's turban, knowing the right terms does not necessarily assure a direct change in policies or attitudes towards stigmatised groups and it is, therefore, misleading to focus on the question of the right terminology. In this paper I argue that the difference between these two perspectives is not about correctness but can be found in the different consequences of the use of these names.
In Italy and France, the scholarly debate on the naming of the Roma has been mainly concerned with the construction and deployment of categories like 'gypsies' and 'nomads', and how these have legitimised exclusionary policies, perpetuating the long-lasting discrimination of these peoples. The first works on the history of the Roma aimed to uncover the unknown aspects of these stigmatised groups, problematising the use of the general term 'gypsy' (Hancock 1987). Scholars highlighted that most terms mirrored the historical contempt towards these groups, for instance the term 'gypsy', coming from 'Egyptian' and stressing their perception as foreigners (Bancroft 2005;Li egeois 1994), and the Italian 'zingari' and French 'gitans' rooted in the Greek word Athinganos or Atsikanos, meaning untouchables (Robert 2007). These studies underscored how the rejection of these groups translated into words, formalising their exclusion. Li egeois (1980) Area 2016 doi: 10.1111/area.12273 points out that the names of these groups are mainly heteronyms used in derogatory manners, like 'Bohemians', that did not have a specific meaning but was applied to everyone considered to be against a state's authority.
In the second half of the 20th century, policymakers started considering nomadism as the main feature of the Roma groups (Simhandl 2009). Piasere (2005) maintains that the discourse on nomadism bolstered policies of spatial confinement and segregation and van Baar (2011) also argues that the framing of the Roma within the terms of nomadism actually 'nomadized' the Roma as an effect. Drawing on Piasere, Sigona (2011) shows that the naming of the Roma goes hand in hand with historical regimes of exclusion, containment and assimilation. Referring to the Roma as 'nomads' justified their exclusion and confinement in camps in Italy instead of their inclusion as war refugees during the Balkan Wars (Sigona 2003(Sigona 2005. Similarly to Sigona, Picker and Roccheggiani (2013) read the definition of the Roma as 'nomads' as the formalisation of their abnormality. From the mid-1990s this discourse on nomadism was replaced at a European level by a discourse on the Roma as a minority to be included (Simhandl 2009). Vermeersch (2012) points out that the terminological shift from 'nomads' to 'Roma' is the achievement of European pro-Roma advocacy groups, but it also entails a new type of exclusion of the Roma from national narratives and a potential tension between a Roma elite of advocates and grassroots organisations.
These analyses on the discourses and names of the Roma highlight how the Roma have historically been called by heteronyms, which reflected the negative attitudes towards them, while at the same time reinforcing their perceived otherness and subsequent regimes of spatial confinement, assimilation and more recent inclusionary programmes. These contributions mainly focus on the exclusionary effects of these naming practices, but little has been said about other types of outcomes of these discursive articulations. As I argue in the next section, this limitation is due to the fact that most of the literature on Roma labelling practices and categorisation is Foucault-inspired. In order to widen the study of the effects of the use of these names, I suggest integrating this approach with the work of the French philosophers Deleuze and Guattari.
Naming and state's categorisation: from Foucauldian apparatus to Deleuzian assemblage Stewart (2013) surveys three different approaches to Romani studies in the last 40 years and notices that research focusing on labelling practices and the state's categorisation is mainly characterised by a Foucauldian approach. Sigona's earlier works (2003Sigona's earlier works ( 2005 employ the concept of 'regime of truth' in order to explain how labelling practices are intertwined with regimes of power, while his later paper (2011) underscores that the label 'nomads' contributes to the formation of different kinds of disciplining regimes. Simhandl (2009) conducts a Foucauldian discursive analysis of the use of the term 'gypsy' and 'Roma' at the European Union level in order to reveal the hidden assumptions of the social construction of these categories. Finally, Picker and Roccheggiani (2013) elaborate on the Foucauldian notion of 'switch point' in order to show how the psychosocial notion of 'nomadness' was enabled to work in the political-administrative field.
While adopting a Foucauldian approach can be extremely fruitful to understanding how the knowledge production and categorisation of the Roma co-constitute specific regimes of power, it also presents some limitations because it risks providing a flattened account of labelling practices that are reduced to a form of topdown categorisation without really accounting for the different effects and resisting practices they might spur. Although Foucault never contended that only dominant discourses matter, nor did the scholars mentioned above, geographical research engaging with Foucaudian discourse analysis nonetheless tends to overemphasise processes of subjection, i.e. a dominant discourse that subjects the individual (Gibson-Graham 2000). Limitations of a Foucauldian approach are also pointed out by Thrift (2007), who argues that Foucault-inspired research underplays the role of humans and overestimates dominant discourses, being too concerned with social and spatial order. Legg (2011), in agreement with Thrift, argues that studies employing the Foucauldian notion of apparatus tend to stress order rather than generative aspects. For example, in the case of the Roma, focusing on the power/knowledge apparatus, whereby the representation and categorisation of Roma as 'nomads' entrenches different forms of either coercive or disciplining power, risks reducing the notion of knowledge as essentially oppressive and subjugating.
In order to overcome this limitation, I suggest integrating this Foucauldian approach with the concept of assemblage developed by Deleuze and Guattari (1987). Assemblage and apparatus are not opposite concepts, but I argue that the former can enrich the latter. The Foucauldian notion of apparatus is far from being straightforward. One definition can be found in a 1977 interview in which Foucault describes a dispositif as consisting of both discursive and non-discursive elements that constantly interact with a clear strategic function of 'manipulation of relations of forces, either Area 2016 doi: 10.1111/area.12273 developing them in a particular direction, blocking them, stabilising them, utilising them, etc.' (Foucault 1980, 196). While this definition strongly focuses on strategic functions, Deleuze (1992) conceived of the Foucauldian apparatus as characterised not only by 'lines of sedimentation' but also by 'lines of "breakage" and of "fracture"' (Deleuze 1992, 159). He argued that Foucault's apparatus was predominantly understood as stratification and sedimentation because his most famous works focused on prison, torture and discipline. However, Foucault produced a series of different writings, interviews and conversations in which he opened up the discussion to 'lines leading to the present day or creativity' (Deleuze 1992, 165). For this reason, in Deleuzian terms, the concept of apparatus can be described as a particular kind of assemblage characterised by re-territorialisation, striation and governing (Legg 2011). While the literature on the Roma names mentioned above has mostly focused on the territorialising functions of the apparatus, by turning to the methodological tool of assemblage, I propose to encompass its deterritorialising effects.
The concept of assemblage refers to a composition of heterogeneous elements in 'some form of socio-spatial formation' (Anderson and McFarlane 2011, 124). Deleuze and Guattari identify two main components of an assemblage: material content and discursive expression (on the horizontal axis), and different degrees of territorialisation and re-territorialisation (on the vertical axis) (Deleuze and Guattari 1987). Deleuze and Guattari's work is especially important for this paper because they distance themselves from an ideological approach to language that links language to power in a unidirectional way (Massumi 2002). In fact, the two horizontal components are independent, while, at the same time, intervening on each other: the discursive expressions are inserted into or intervene in contents, not to represent them but to anticipate them or move them back, slow them down or speed them up, separate or combine them, delimit them in a different way. (Deleuze and Guattari 1987, 86) being a real 'action on an action' (Massumi 2002, xix). Thus, using certain names rather than others bears consequences on spatial arrangements and distributions that can be either reinforced (territorialised) or unhinged (deterritorialised). The names of the Roma cannot be reduced to a mere territorialisation of exclusionary practices, like confinement and assimilation, but should be considered part of an assemblage that enables different effects, both stabilising and de-stabilising ones, in a rhizomatic fashion. The next section shows how different discursive articulations cannot be accounted for as sheer territorialisation of power, as most of the literature on the naming of the Roma has donemore or less explicitlyby adopting a Foucauldian approach, but also imply moves of deterritorialisation and lines of escape.

Roma naming in France and Italy: between closures and new openings
Both French and Italian governments' discourses are characterised to different extents by a strong ambiguity and interchangeability of the categories 'nomads', 'Roma' and 'Travellers', which in the last 15 years have bolstered discriminatory and segregating measures targeting the Roma groups. These exclusionary policies have often been criticised by European institutions for exacerbating stigmatisation and were countered by naming strategies that carefully distinguish all these categories, avoiding their conflation into an ambiguous and threatening 'other'. Yet, as I show in this final section, both these naming strategies produce a plurality of effects that cannot be reduced to either inclusion or exclusion: while, as already acknowledged by Vermeersch (2012), the naming practices of the CoE indeed counteract discriminatory national discourses, they can at the same time reinforce divisions. The ambiguity of names not only aggravates discriminatory policies but can also spur new creative practices of resistance.
In Sarkozy's statement quoted in the first section of this paper, 'nomads', 'gens du voyage' and 'Roms' strongly came together, almost melting into each other, differently from the CoE approach. The association of these categories had the effect of constructing an ambiguous threatening other, which could in turn legitimise exclusionary policies targeting these groups, like the Second Besson law adopted in 2000. This law, regulating the illegal occupation of land by sanctioning squatters that settled on unauthorised areas, was denounced as disproportionately impacting on the French Travellers communities. During the parliamentary debate on this law held on 15 November 2002, the former minister Sarkozy was strongly criticised for dismissing the difference between French Travellers and Eastern European Roma, but his phrase was not a simple mistake. It constituted a re-articulation of discursive components that could bolster future measures aiming at both French Travellers and more recently arrived Roma migrants. The early 2000s witnessed an increased migration of Eastern European Roma and this indeed led to the adoption of new policies tackling their presence in informal settlements. Eight years later, a new naming strategy emerged. In 2010 there were a series of highly mediatised violent episodes, which had allegedly implicated members of French Traveller communities and led to rioting against the police. As a result, the French government responded with a wave of evictions of Roma informal settlements and repatriations. On 30 July 2010, in Grenoble, Sarkozy presented the Declaration on Security, a speech defending the actions of the police and explicitly referring to 'various kinds of delinquents in France, including immigrants that have failed to integrate and, in particular, the Roma' (Parker 2012, 478). A press release published on 28 July 2010 condemned the behaviour of a small minority of French Travellers, while directly targeting all the Roma migrants coming from Eastern Europe. The communiqu e reported that the French President did not accept the illegal situation of 'the Roma populations arrived from Eastern Europe into the French territory' (Pr esidence de la R epublique, 2010). Contrary to eight years before, this time the Roma from Eastern Europe, and not the French Travellers, were presented as a threat to public order and associated with activities of illegal trafficking, 'shameful living conditions, child exploitation, prostitution and delinquency'. The 2010 repatriation policies formalised a change in the discursive articulation referring to the Roma people: while in 2002 Sarkozy presented these names as synonyms, in summer 2010 there was a separation of Roma and Travellers. The formal disjoining of the terms French Travellers and Roma had the effect of reinforcing the threatening ambiguity of the Roma, then subject to increasingly oppressive policies.
Although the conditions of the French Travellers have not yet improved because they are still strongly discriminated against for their way of life, the highly contested repatriations and evictions carried out in summer 2010 were mainly aimed at the Roma migrant communities (especially from Romania). Furthermore, this distinction, between nomadic French Travellers on the one hand and Eastern European Roma migrants on the other, exacerbated divisions and mutual distancing strategies between the two groups. For instance, in 2013, the president of the association France Libert e Voyage (an association of gens du voyage) stated that: We highly respect the Roma and we have to help those who are bound to live in horrible slums. But our problems are different. We reject all these misunderstandings that only add the prejudices against the Roma to all the discriminations that we are already subject to. (D epêches Tsiganes 2013, np) Similarly, a Roma activist of a French pro-Roma advocacy group said during an interview: The French Travellers, they've got nothing to do with the Roma, they can move, they're French citizens, there are specific measures for them . . . they're another thing.
These examples confirm that the distinctions made between these names in the CoE glossary, as a result of pro-Roma groups that advocated a framing of the Roma as European ethnic minority, do not necessarily imply less discrimination but can even weaken efforts to create solidarities between these two groups.
The French case reveals that both the construction of tight and ambiguous discursive formations, and the drawing of distinctions, can buttress exclusionary measures. In addition to this, the Italian case shows that also the ambiguity of tight assemblage formations can enable deterritorialisation. As mentioned at the beginning of this paper, similar to France, the Italian national and local governments adopted a naming articulation based on the conflation of the categories 'nomads', 'Roma' and 'Sinti'. This ambiguity has been observed by several scholars who have analysed the Italian context (Piasere 1999;Sigona 2011) illustrating that the segregating system of Roma camps in Italy was legitimised by a discourse that deployed the terms 'nomads', 'Roma' and 'Sinti' interchangeably. However, during my fieldwork I discovered that this ambiguous articulation has also enabled practices of resistance.
In 2009 a group of Roma evicted from an informal settlement in Rome joined a political squat called Metropoliz, in the eastern periphery of the Italian capital city (Maestri 2014). This group of Romanian Roma started living in an occupied abandoned factory with other squatters and were not targeted by the policies for Roma living in informal settlements. In this caseand also in the other squats in Rome that later included Roma groupsthe Roma started mobilising together with other migrants and Italians as part of a new political subject based on their socioeconomic status of being excluded from the formal labour market and being left homeless in times of crisis. The ambiguity of the discursive articulation, mixing different terms and lacking any formal and clear definition, has enabled this group of Roma to present themselves as squatters, escaping the policies aimed at 'Romanomads'. This move to a squatter category has troubled the local administration in Rome, which now includes the 'Roma-squatters' in policies for squatters instead of accommodation policies for the 'Roma'.
The very discursive formation that sustained discriminatory practices and the segregating policies denounced by many of the scholars mentioned at the beginning of this paper has also created the conditions for a resistance against these types of exclusion. During an interview, a public official working for the City of Rome declared that by simply moving from the informal settlement to the squat, these Roma were no longer seen as 'Roma-nomads' by the municipality: [When they become squatters] they're no longer the responsibility of this office [which deals with the issues of the Roma community]. [. . .] For the moment it's a strategy that does not have a clear outcome yet. But I'm happy that they finally interface with the city not as Roma, but as people experiencing severe housing deprivation.
Since the official discourse underpinning segregating policies depicts the Roma as nomads living in self-built shacks or caravans in informal settlements because of their culture, moving to an occupied building with squatters is no longer interpreted through the lens of 'Roma-nomads'. The Italian case shows that ambiguity can be used to enhance the perceived danger of a threatening category, as argued by Puar (2007), but can also be used as a moment of creativity in which change becomes possible (Mahoney and Thelen 2010). The assemblage constraining the Roma also enabled them to escape it by presenting themselves with a different name and moving to a different space. As a member of an Italian pro-Roma and pro-migrant association articulated an interview: 'They have been considered another thing'. To escape ambiguous and oppressing assemblages, clarifying the distinction between the possibly mis-used termsas the CoE does might not be the only way.
These examples -first the lumping together of different categories and then their separationshow that single naming and labelling practices do not have straightforward effects. The CoE glossary is usually considered to be the result of the fights of pro-Roma groups that advocate Roma inclusion, while discourses constructing an ambiguous other are seen as perpetuating the stigmatisation of these groups. However, as I argue in this paper, assemblages can be both restraining and enabling. In the French case, both naming strategies have produced exclusionary effects, while in the Italian case the discourse blurring the distinction between terms has been used to develop new forms of resistance.

Conclusion
This paper originated from a dissatisfaction with the Romani studies scholarship on labels, and state categorisation, which are too focused on the exclusionary functions of the names used to refer to the Roma groups. In this paper I did not aim to provide a correct definition of the Roma or to find their right name. Rather I have looked at the effects of the different denominations of the Roma employed in France and Italy. By drawing on the notion of assemblage, I have shown that labels and categorisations cannot be reduced to a pure legitimisation of discriminatory and segregating policies, but can buttress both exclusionary regimes as well as resisting practices. This conclusion remains, however, strictly specific to the countries and contexts I have analysed. Although the Roma minority is one of the most vulnerable and discriminated in Europe, their history and current situation differ among geographical areas and states, especially between Western and Eastern European countries (see, for instance, Cret ßan and Turnock 2008). The analysis of other countries would be an important step to enrich a comparative work on the effects of naming assemblages.
The quote from Deleuze and Guattari at the start of this paper reminds us that it is important to know how things are named, as names are not inert labels that objectively indicate things. However, knowing whether something is a 'revolt' or a 'petition' does not necessarily convey one meaning, but opens up the interpretation to a multiplicity of effects that are made possible by one name. Similarly, knowing that a group is referred to as 'nomads', 'gypsies', 'travellers' or 'Roma' is, indeed, important not because the term 'nomads' is relentlessly linked to segregation, or the ambiguous use of all these terms together inevitably serves the state's exclusionary practices, but because these different names also enable strategies of deterritorialisation and unexpected effects. accepted as non-derogatory and non-discriminatory by several, both Roma and non-Roma, institutional, nongovernmental and advocacy organisations.