‘I speak small’: unequal Englishes and transnational identities among Ghanaian migrants

ABSTRACT This paper investigates language ideologies involving various non-standard English-language practices among homeless Ghanaian migrants, and explores how these interplay with transnational identity management in Catalonia, a non-English-speaking bilingual society. Through a 6-month multi-site ethnography of three case-study informants which included recorded interviews and spontaneous interactions, I explore how migrants engage with various pluralisations of local and global English in reported encounters with other migrants and local residents, and I show that they share ambivalent positionings towards them. They generally present themselves as speaking ‘small’ or ‘no’ English, in acts of linguistic delegitimisation whereby they inhabit marginalised, de-skilled pan-African identities. However, on other occasions, they position themselves as ‘better’ English speakers than local populations who sanction ‘outer-circle’ English forms, in acts of self-legitimisation whereby they vindicate their ‘native speakerhood’ condition, constitutive of educated, cosmopolitan identities revolving around ‘Ghanaianness’. I conclude that these sociolinguistic comportments speak of migrants’ linguistic marginalisation. They uncover ways in which situated forms of identity categorisation linked to the censorship of socioeconomically-stratified English varieties shape, and are shaped by, hegemonic monolingual ideologies and societal normativities concerning ‘English standardness’ which dictate who count as legitimate transnational citizens in the Southern European societies of the twenty-first century.

This diversity of people has motivated a growing body of research within socially-committed interpretive humanities disciplines (see, e.g. compilations in Canagarajah, 2017;Duchêne, Moyer, & Roberts, 2013). Among linguistic anthropologists, critical sociolinguistic ethnographers, discourse analysts and narrative practitioners working within the field of transnational migrations, particular emphasis has been placed on language and identity; more specifically, on how language practices and ideologies interplay with the ways in which current migrant networks manage, inhabit, and/or resist social identity categorisations when they negotiate their place in resident societies, in the urban geographies of the twenty-first century (see, e.g. Baynham, 2005;De Fina, 2003;Lanza, 2012;Relaño-Pastor, 2010).
In this paper, I understand language as practice and as ideology (Heller, 2007); that is, as communicative practices in which we get organised in society in everyday life, and as indexes of the norms, attitudes, judgments, etc., Which govern collective and individual sociolinguistic comportments (Schieffelin, Woolard, & Kroskrity, 1998). Likewise, I conceptualise identity as social categorisation practices mediated through, and constituted in, situated communicative events. I follow a line of research which envisions transnational populations' identities as hybrid and fluid, rather than as 'fixed' or ascribed to a single place of origin (see De Fina, 2016;Woolard & Frekko, 2013). I approach these re-presentations of the Self as emerging and materialising in 'multilingua francas' (Makoni & Pennycook, 2012, p. 449). These are non-orthodox multilingual practices based on translinguistic communicative resources which consist of non-standard, inextricable amalgamations of linguistic codes from local and distant contexts -'repositories' of mobile populations' socialisation experiences (De Fina & Perrino, 2013;Dovchin, Sultana, & Pennycook, 2016;Jacquemet, 2005Jacquemet, , 2010. From this perspective, transnational migrants' language and identities challenge nativist conceptions of language which link linguistic codes to given homogeneously imagined monocultural, monolingual territories or 'ethnicities' (Sabaté i Dalmau, 2014). Despite their counterhegemonic, transgressive nature, migrants' multilingualisms are silenced and sanctioned, on being considered 'non-quite-languages' (Gal, 2006, p. 15) used by 'incompetent', 'language-less' people (Blommaert, Collins, & Slembrouck, 2005, p. 213). In this sense, migrants' languages and identities are inserted into local, nationstate and supra-state language 'regimes' (Kroskrity, 2000), including institutional language policies and mundane norms and societal monolingual/monoglossic mindsets, which foster particular standard uses of dominant lingua francas as well as of 'official' state languages as a precondition for accessing citizenship and for attaining 'proper' personhood legitimacy.
Unsurprisingly, one of the socioeconomically and politically powerful lingua francas which gets most frequently mobilised (i.e. relocalised, appropriated) by migrants in intercultural encounters is English (Canagarajah, 2013;Pennycook, 2012;Tupas & Rubdy, 2015), whose global spread and imposition, particularly in former UK and US colonies, has been widely attested (see, e.g. Fairclough, 2006;Pennycook, 1994Pennycook, , 2007Phillipson, 1992). In this sense, most migrants' multilingual practices are mediated in and through English pluralisations, and most transnational identities involve ideological (non)-engagement with socioeconomically de/valued translocal forms of this language. I refer to these pluralisations of English emanating in migrant identities as 'unequal Englishes' (Tupas, 2001, p. 81) in order to problematise the perpetuation of the exclusionary hegemony of 'inner-circle' English varieties and of their prestige for those who speak it, stressing the idea that non-orthodox English forms 'are all linguistically equal but [that] their political legitimacies are uneven' (Tupas & Rubdy, 2015, p. 3). This approach is particularly helpful for the exploration of the (re)-production of situated forms of social distinction, difference and, ultimately, inequality among native and non-native English-using migrants, particularly in contexts of extreme precariousness, in peripheral urban geographies of twenty-first century Southern European societies such as the one presented below.

The present study
The aim of this paper is to explore migrants' ideologies around multilingual practices involving a diversity of de/valued forms of local/global English and to understand how these interplay with English-mediated transnational identity management. I do so through the analysis of three case-study informants consisting of three homeless Ghanaian men who lived in a public-transport bench in a town called Igualada. This was located an hour away from Barcelona City, in Catalonia. Catalonia is a bilingual society of about 7.5 million inhabitants (Idescat, 2016) located in North-eastern Spain where a majority nation-state language, Spanish, coexists with a minority national language, Catalan. 1 Concerning foreign languages, Catalonia is officially non-English-speaking: the teaching of English is relatively new and its use as a lingua franca is scarce, when compared to other European regions (Eurobarometer, 2012). The methodology employed consisted of a multi-site ethnography of this small network which included participant observation, audio-recorded narrative interviews and spontaneous interactions (see below).
The analysis is organised as follows. Firstly, I provide a rationale of the informants' translinguistic English practices, frequently involving language resources in Ashanti and Arabic. I then analyse how they positioned themselves with respect to the ideological conceptions and socioeconomic legitimacies assigned to these various local/global English forms (and to their speakers) in their resident society, and I show that they shared seemingly ambivalent positionings towards them. I first focus on how informants generally presented themselves as speaking 'small' or 'no' English. I approach this sociolinguistic comportments as acts of 'self-decapitalisation' (Martín-Rojo, 2010); that is, as acts of linguistic delegitimisation of one's language resources which embedded what was dismissively constructed as 'black English' 2 (i.e. postcolonial, 'outer-circle' English) into a macro marginalised migrant identity linked to a stereotyped social image of African foreigners as powerless, uneducated persons. I then analyse how, and why, on other occasions, informants positioned themselves as 'better' English speakers than locals in town, who tended to foster dominant prestigious ('inner-circle') accents only, and who systematically sanctioned hybrid, reterritorialised English varieties, with a monolingual/monoglossic mindset. I show that they did so in acts of linguistic self-capitalisation or self-legitimisation whereby they vindicated their 'native speakerhood' condition and claimed 'ownership' of the language, constitutive of a distinctive identity which included literate, cosmopolitan 'Ghanaianness,' in the same discursive space. In the last part of the analysis, I argue, first, that migrants sought to attain a certain degree of social agency (i.e. an authoritative voice, or linguistic empowerment; see Giddens, 1984) by demarcating their bench as a Ghanaian space through the use of Ashanti in combination with translinguistic English, in front of other non-English-speaking migrants with whom they competed for transnational resources (like job opportunities in the informal economy or food). In this sense, I try to focus on ideologies on pluralised English forms in situated communicative events which are meaningful and relevant for the informants themselves. I claim that this highlights the importance of approaching ideologies of language practices involving Englishes from a participant-oriented perspective, understanding individuals as key actors in social contestation and change (Pujolar & O'Rourke, 2016).
I conclude that the informants' ambivalent attitudes index both 'linguistic insecurity' and 'linguistic affirmation/assertiveness' (Canagarajah, 2013, p. 4) concerning the use of English with other migrant networks as well as with local populations (including the researcher). I suggest that this reveals how migrants voiced, and coped with, the censoring of their multilingual resources mediated in/through English, in resident societies. I argue that such devaluation propels the linguistic marginalisation of these populations in ideology and in actual practice, particularly their 'de-languaging' and 'de-skilling' (Allan, 2013, p. 58), where a command of Spanish as the nation-state language of 'integration' is a must (see BOE, 2015), and where linguistic hybridity and 'accent' are penalised (Codó & Garrido, 2014). Overall, the findings contribute to an understanding of how situated forms of socioeconomic differentiation and inequality materialised in the devaluations of non-elite English varieties ultimately shape, and are shaped by, exclusionary language mindsets engrained in neoliberal global sociolinguistic orders that regulate who count as 'proper' migrant English speakers and citizenship-deserving, transnational Selves.

Context and participants
At the time when this project started, Igualada, the capital of a central Catalan county, had about 40 thousand inhabitants, 14.7% of whom consisting of foreign residents (the percentage of 'foreigners' in Catalonia as a whole was then 15.7%). The first largest migrant group consisted of people born in the African continent (6.49% of the town's population), the Ghanaians under study being the second largest subgroup after the Moroccans, with 112 people, mostly single men aged between 35-44 (Ajuntament d'Igualada, 2012).
The three informants of this research project, Alfred, Benedito and Paul (pseudonyms) were, respectively, an English teacher, an accountant, and a schooled cocoa farmer in their forties who were born in an urban town and two rural villages near Sunyani, the capital of Brong Ahafo, the second largest province in Ghana (West Africa), characterised by twentyfirst century mass emigration (Pierre, 2012). Between 2000-2001, escaping from violence among Muslims in their region (as detailed in Tsikata & Seini, 2004, p. 26), and trying to protect their transnational family income and find better employment chances in Europe, informants moved to Southern Spain and started working in agriculture. During that period, their mobility trajectories included frequent visits to their relatives in Ghana and in other parts of Europe (like Italy and the Netherlands). Later on, the three moved to Catalonia, pursuing socioeconomic improvement, informed by other Ghanaian acquaintances which had followed similar mobility paths. Benedito and Paul settled in Barcelona City, and Alfred moved to Lleida (Northern Catalonia) to pick fruit. They reported having had a difficult time in these places, for which, between 2004 and 2007, they decided to move to a smaller yet well-connected town where they expected to work in the industry sector. Igualada was their choice because it then was one of the biggest textile industries in Catalonia and the first tanning market of the Iberian Peninsula (Ajuntament d'Igualada, 2013). The three met there for the first time. Alfred started working in a tannery; Benedito, in the biggest foundry; and Paul, in the construction sector, and they all obtained a temporary residence visa.
In 2010, Igualada was struck by the Spanish economic recession, linked to an economic crisis of global reach. As a consequence of this, the leather and tanning industry collapsed, and the region experienced the highest percentage of employment loss in Catalonia as a whole, the most affected by it being foreign labour workers, whose unemployment rate reached 37.1% (Galí Izard & Vallès, 2010)when in Catalonia as a whole it was 22% (Comissió Obrera, 2011, p. 16). Informants became unemployed and started working in the informal economy, selling scrap from garbage containers, and begging in the car park of a peripheral supermarket, while they kept in touch with the temporary-work agencies with which they had previously found employment. None of them was receiving any severance pay at the time of the fieldwork. Cáritas, the official confederation of charities of the Spanish Catholic Church, provided them with washroom facilities, food and clothes. Their transnational mobilities had become very limited (none of them had visited Ghana since 2008), because with their non-permanent visas they could no longer travel freely to other parts of Europe. By the end of the fieldwork, they could not pay for a shared rented room anymore and became totally unsheltered. Then, they decided to take refuge on the bench of an open-air public transport area located on the outskirts (in front of the supermarket and the car park aforementioned), where they lived under precarious conditions (they developed serious stomach, lung and heart problems). This bench became their 'public in private' socialisation placetheir space of 'meetingness' (Urry, 2007, p. 68).

Methods and data
The data was gathered by means of a 6-month multi-site network ethnography of the small Ghanaian network under study (I went into the field at least three times a week during different times of the day from July 2012 to January 2013, and then, intermittently, until November 2014). This consisted of active participant observation of the informants on their bench and of several 'co-ethnographic visits' (Convey & O'Brien, 2012, p. 339) to the particular socialisation places in Igualada that they mentioned at different stages of the fieldwork, all located at a 20-/30-minute walk from one another (these places included, e.g. the mosque, Cáritas office and the temporary-work agencies). My objective in embedding this mobile ethnography into the spaces that were made salient by informants was to turn the research into an informant-oriented project, which further helped me to establish rapport with them (for the details on this methodology see [author]).
Access was granted after I had been observing the informants for a year, on my way to the bus station, where we could have short conversations, too. I introduced myself as a Catalan English 'teacher' wanting to investigate migrants' languages in town, and I always told them what I wanted to know and why. They were totally unimpressed by the university certificates with the project information, and fruitful cooperation, followed by verbal informed consent to participate in the study, was not granted until they were convincedand sawthat I did not work for the town hall or for any NGO, because they feared both. 3 Since I had no command of any African languages, I introduced myself in Catalan, and then in English and in Spanish, too. I chose Catalan following the idea that not addressing migrants in the local language was an exclusionary 'Othering practice' (Barth, 1969) that prevented them from learning the language which opened the doors to the local economy, and which indexed membership and belonging to Igualada. This was a marked sociolinguistic comportment, for it has been attested that local populations switch from Catalan to Spanish automatically when addressing 'foreigners', and that migrants, at the same time, expect locals not to use 'their' code with them (but to employ Spanish instead), fostering a complex Catalan/non-Catalan ethnolinguistic boundary (see Woolard, 2006). For all these reasons, the informants associated my choice of Catalan with a 'Catalan' ethnolinguistic identity.
The fact that I made frequent use of English was considered a marked sociolinguistic comportment, too. This was so because local people are expected not to command English 'well enough' so as to use it as a lingua franca with foreignersas outlined above, the common language to be used between locals and migrants is Spanish, conceived of as the 'language of integration' indexing a 'right to naturalisation' and 'proper citizenship behaviour' (Pujolar, 2007). I believe that the choice of English worked to my favour in that (1) it allowed the researcher and the researched to have a distinct 'wecode' with which to interact (as opposed to what happened with the Moroccan populations, for instance, with whom I used local languages only) and that (2) it gave them a voice as legitimate English speakers who could tell their story in non-standard Englishes.
The data collection process was as follows. Over six months, I recorded narrative interviews, here understood as negotiated, reflective and transformative communicative events (De Fina & Perrino, 2011), on the following intertwined narrative themes: (1) geographic im/mobility; (2) un/employment, up/downward economic mobility and professional stagnation; (3) non-legality statuses and (non)-citizenship rights; (4) social relationships and identity ascriptions among themselves and with other migrants and (5) de/legitimised multilingual resources. I asked them to conduct these interviews in their preferred languages, which in the end consisted of English and Spanish, with extensive code-switching, as seen in the analysis.
Finally, the data also comprised a series of spontaneous interactions (mostly salutations and chitchat) between the informants and other migrant men from Senegal, Morocco and Kashmir, which took place in Spanish, English, Arabic and/or Ashanti (all recordings lasted for about 145 min), as well as archival documents, reports and visual materials such as hand-written notes. For the purposes of this paper, I chose to analyse five excerpts broaching narrative themes (4) and (5). These excerpts were selected on the basis of their use for the aims of this paper, which consisted of: (a) illustrating ideologies on multilingual repertoires and translinguistic practices involving non-elite Englishes, and (b) exploring narratives broadly concerning English-mediated transnational identity as linked to situations of social categorisation, difference and inequality (I provide an analysis of the other narrative themes in [author]).

Analysis: Ghanaians' transnational identities and unequal Englishes at play
In this section, I first offer a brief description of the multilingual resources into which informants' local/global English forms got inserted and materialised in actual practice. I then analyse the attitudes that they displayed towards them when they explicitly delegitimised non-orthodox English forms in public, on the bench. I argue that these intertwined with presentations of the Self which drew on, and relocalised, a circulating pan-African identity (a well-known 'macro' social categorisation of the 'black foreigner'), on the peripheries of Catalan urban towns. Finally, I analyse acts of linguistic self-empowerment whereby, by contrast, informants legitimised their English and make prevail their 'native speakerhood' condition in this language to present themselves as 'better' English users than locals and other migrants, which triggered the self-ascriptions of transnational identities revolving around modern 'Ghanaianness.' .

Non-standard multilingualism resources interplaying with devalued Englishes
Concerning multilingual repertoires mediated through various English varieties, Alfred, Benedito and Paul employed the most prestigious and the most widely spoken variety of what in 1950 was labelled as 'the Akan language' (Bodomo, 1996;Kropp Dakubu, 2015[1988), Ashanti, used as a lingua franca among themselvesthey also commanded other Ghanaian languages and many of the other eight Akan language forms, such as Akyem. 4 Ashanti was of crucial importance on the bench, since it demarcated that zone as a 'Ghanaian' space which welcomed and provided resources for transnational subsistence to 'the other blacks' (as informants called them). The Senegalese and Nigerian men who came by the bench to access food, cigarettes and advice on legality issues, for instance, greeted informants with the Ashanti salutation 'bone nnim' (literally, 'no problem'), before having their conversations in Spanish (with the former) and in English (with the latter), showing deference towards them.
Some Arabic was also used in an intra-group manner by informants, too, to show respect to Paul, a practicing Muslim, whom Alfred and Benedito, non-practicing Christians, always greeted with the salutation 'As-salam alaikum' ('peace be with you'). Arabic was simultaneously mobilised in an inter-group manner, but very differently, as a site of struggle whereby to negotiate competitions of linguistic legitimacies. These uncovered rivalling relationships across migrant groups, particularly between Ghanaians and Moroccans, who kept presenting themselves as 'less advantaged' and 'more in need' than 'the others', when they talked about access to Cáritas' resources. An example of this was provided to me by informant Paul and his 'acquaintance' Abdelmahid from Morocco, who always used the nickname 'A'azi' to call each othertheir conversations then followed in Spanish. This term of reference is a racist Arabic slur equivalent to 'nigger', here used ambiguously and with laughter, as a way to manage social tension between both migrant groups (for language-mediated conflictual relationships see [author]).
Against common thought, informants did know about, and understood, the Catalan language, despite the fact that they claimed not to be 'competent enough' in it, as observed in audio-recorded comments such as: 'If you speak yes it's [ok] but I can't reply you in catalán ('Catalan')' (made by Paul). They made reference to Catalan particularly when displaying their knowledge about the language and identity dynamics and the sociopolitical situation of Catalonia within Spain, in front of the 'Catalan' researcher (for example, when talking about the non-binding pro-independence referendum in Catalonia held in 2014).
Following an 'integration through state language' monolingual ideology, informants mostly used Spanish with local populations and with non-English-speaking migrants. They presented it in interviews as the (only) legitimate language of reterritorialisation, to the extent that it got inserted in the English talk mediating their interactions, as seen, e.g. in audio-recorded statements dealing with work and legality issues, such as: 'In the almacén ('warehouse') you inside room big big big big room; it's a fábrica ('factory')' (taken from Paul).
The Spanish language was also a barometer of 'integration' to be used among informants themselves and with other migrants in linguistic competitions whose aim was to see who was recognised as a law-abiding, compliant transnational Self. Paul, for instance, insisted that Malians in Igualada spoke 'little little Spanish' and presented himself as 'more enculturated' and experienced than them. In this sense, informants participated in, and actually reproduced, the sociolinguistic regime of normalcy concerning 'integration' to which they were subjected in Spain, which fostered the ideological construction and actual use of monolingual Spanish (and sanctioned hybrid multilingualism) as the resource to gain access to citizenship status and, ultimately, to citizenship rights.

Self-delegitimisation acts and postcolonial pan-Africanism
The latest official statistics report that the percentage of Ghanaians aged 15 or more who can read and write is 76.6%, a literacy rate between 10 and 35 points higher than that of Ghana's three neighbouring countries (CIA, 2016). This is due, in part, to the introduction of the policy Free and Compulsory Universal Basic Education (FCUBE), passed in 1995, which made Ghana's educational system one of the most successful systems in Sub-Saharan Africa (Akyeampong, 2009). The vast majority of Ghanaians, particularly in Brong Ahafo, speak English (called 'Ghanaian English' or GhE), since this is the only language of formal education beyond the first three years of primary school (conducted in Ashanti) and the only official language at a national level.
The informants who participated in this study were all schooled and read and wrote in Englishone of them, Alfred, was a primary school English teacher before migrating. This was a very important lingua franca for inter-group communication; for example, in intercultural encounters with Pakistani acquaintances. And yet, the first time I asked the informants about their multilingual resources, they insistently downplayed and dismissed their command of English, as shown in Excerpt 1 below.
(1) Speaking 'small' or 'no English'. In Excerpt 1, Paul, the cocoa farmer, presents himself as speaking 'no English' (lines 2 and 4). My first reaction was of surprise, because it was apparent that we were actually conversing in this language, in that interview. After my interruption with an overlap in line 5, he clarifies this and, in a paraphrase, explains that he speaks 'small English' (line 8)note that the 'lack' of command of English was interactionally emphasised with the repetition of 'no' or 'small' before providing a list of languages that he did speak, in order to answer the researcher's question concerning his multilingual repertoire. I understand this as a public act of self-delegitimisation of one's linguistic resources in English, constitutive of a further process of a presentation of the Self which included self-delanguaging, on the part of Paul, in this case.
I argue that Paul's attitude may be tied to a dominant language ideology which conceives of Englishes that are not monoglossic 'inner-circle' varieties like GhE as faulty and non-complete. This social construction of GhE as a non-fully-fledged code is deeply rooted and widespread not only in society at large but also in some applied linguistics circles where it gets defined as 'broken English' and 'pidgin English', reinforcing the idea that it is a primitive/basic type of English (see, e.g. the language labels employed in the reference guidebook The Languages of Ghana by Kropp Dakubu, 2015[1988), despite the fact that sound evidence has long been provided that postcolonial Englishes are totally functional, legitimate codes (see, e.g. Kachru, 2006, pp. 247-250).
I suggest that Paul's presentation of the Self as having scarce or no English resources interplays with the migrants' use of the self-ascription of a broad social categorisation of African foreigners in Europe as docile, marginalised and victimised personas in need of Western 'resocialisation' and schooling (for more examples of this media-sponsored 'macro' identity see Codó & Garrido, 2014;Sabaté i Dalmau, 2014). The apparent embodiment of this identity (which does not imply internalisation of English-language non-ownership, as shown below) may be read as this network's complaints against social disadvantage and linguistic marginalisation, issued in public in front of a local, advantaged researcher. The linguistic marginalisation Paul was complaining about was observed, for instance, when his English résumé was translated into Catalan by work-agency employees who were mistrustful of the authorship of such document (and, therefore, of Paul's literacy resources).
On other occasions, though, this simplistic image of the African migrant, which draws on Africa's past colonial heritage and present-day oppression, is taken up by the same informants to vindicate a proud sentiment of shared 'pan-Africanism'a complex multivalued construct encompassing the idea of a 'broader African people' (Lake, 1995, pp. 21-22) stereotyped as 'underdeveloped' but as being more 'honest' and 'humane' than 'Europeans'. I suggest that pan-Africanism, among these particular informants, revolves around the mobilisation of two traits, 'blackness' and 'Englishness,' as seen in Excerpt 2.
(2) Pan-Africanism: Blackness and Englishness. The interview in Excerpt 2 was conducted when I was trying to find more Ghanaian informants for the study. In line 1, I tell Alfred about George, a cybercafé worker whom, I mistakenly believed, was from Ghana (line 3). Paul and Alfred together present George as a Nigerian (lines 2, 4 and 5), and I was very surprised that they knew and talked about him with such a degree of familiarity, since his cybercafé was located in the town centre, at a 20-minute walk (besides, I never saw George near the Ghanaians' bench). Informants were surprised, at the same time, that I asked such a question: For them it was obvious that they knew the members of the network of 'blacks' in town, as seen by Alfred's emphatic expression of astonishment 'ah!' (in line 7), which overlaps with Paul's explanation that they indeed knew George and that they conceived of him as 'African' because of a physical trait that they shared: his black skin complexion (line 8). Note that, in doing so, he self-attributes an 'insider knowledge' about Africans in Igualada. Paul is supported by Alfred, who repeats the same argument, in another overlap (in line 9). Alfred later provides yet another reason why they knew about George. He explains that Ghanaians and Nigerians are both English speakers, on having been born in countries where this is the only official language, appropriating an 'Englishness' trait concerning language choice (in lines 12 and 14). This is also constitutive of the sort of pan-Africanism which informants presented under an umbrella social category that they constructed as 'we the blacks' (see [author]). As we shall see, this connects with the communicative events where they drew on their 'Englishness' and, in this case, spoke proudly about their English resources, as presented below.

Self-legitimisation acts and modern 'Ghanaianness'
As outlined before, on other occasions informants presented themselves as fully-fledged English speakers, drawing on nativists conceptions of the language whereby they made prevail their 'native' speakerhood condition and 'ownership' of the language (GhE, in this case) not only to interactionally construct themselves as linguistically competent multilingual personas but also to position themselves as better English-language users than local populations, as illustrated in Excerpt 3.

@Bck:
With the help of Paul (PAU), Alfred (ALF) presents himself as a legitimate English speaker, constructing, in turn, locals as having no (or scarce) command of the language, in a self-capitalisation act, in front of the researcher (RES In Excerpt 3, Alfred, the English teacher, presented his mobility trajectories, which included visits to Holland (line 1). I inquire about the language that he used there and about whether he found it 'difficult' (line 4), thinking that maybe Dutch had become part of his multilingual resources. He replies, though, that people in Holland spoke 'good English,' actually equating 'Holland' to this lingua franca, 'English' (in line 5). I then ask about the use of English in Igualada by local populations (line 6), to which Alfred replies that they do not speak it (lines 7 and 9). Paul, the cocoa farmer who had previously presented himself as having 'no' or 'scare' English resources (see Excerpt 1), answers, in an overlap, that people in town speak 'small English' (line 8), indirectly positioning himself as a 'better' English speaker than them, in this interaction. Alfred finishes the conversation by clarifying, again via repetition, that what they meant is that just a few locals have a command of this language (lines 10 and 12). 5 I claim that on having been given a legitimate voice to assess the locals' multilingual resources, these two informants gained a degree of linguistic legitimacy and empowerment. On the one hand, they conduct an act of self-legitimisation whereby they indirectly present themselves as competent English speakers; on the other hand, they dispossess locals of their 'Englishness' with authoritative voice.
Similarly, informants tended to assume that English should be the lingua franca among migrants (along with Spanish). They presented other African migrants as non-English speakers, too, and they saw this, literally, as a 'problem' for intercultural communication and socialisation, as shown in Excerpt 4 (lines 1 and 5), where Paul indirectly constructs himself as more enculturated or linguistically equipped than Malians and Senegalese migrants (line 3) (again, his claims here stand in opposition to his presentation of the Self as having none or scarce English resources, in Excerpt 1).
(4) Dispossessing 'other' African migrants of 'Englishness'. I argue that the informants' linguistic self-empowerment interplays with a projected social category tied to a proud sentiment of 'Ghanaianness.' This is an identity constitutive of modernity and Westernness that counteracts stereotypes pejoratively associated to Ghanaian migrants which include socioeconomic stagnation, rurality, illiteracy and cultural backwardness (see Pierre, 2012). In other words, linguistic legitimisation acts and modern Ghanaian identities are both mobilised to fight the image of the marginalised African foreigner presented in the section above. In the particular context under analysis, this self-ascribed social categorisation revolved first and foremost around the educational system of their country, as seen, for instance, when informants listed the world-ranked Ghanaian universities in front of the researcher, an example of which is provided in Excerpt 5.
(5) 'Ghanaianness': Educational leadership and modernity. In Excerpt 5, Benedito, the accountant, starts talking about the Kwame Nkrumah University of Science and Technology (KNUST, or Tec) in Kumasi (lines 1, 3 and 4). He emphasises its reputation in Ghana ('ask anyone'; 'they all know'; lines 3-4) and its importance for the continent (with the expression 'all Africa,' in line 6). Benedito's construction of Ghana as having attained educational leadership worldwide is also observed when he more explicitly emphasises the international character of this university, mentioning that 'the Europeans' (lines 8 and 11) -'the whole Europe' (line 17)studied there. Benedito does so by displaying his literacy practices. He took an envelope and wrote down the university acronym on it, despite the fact that the researcher had already done so in her fieldnotes, reinforcing his presentation of the Self as a schooled, cultivated persona, as part of the linguistic self-legitimisation acts that he conducted in public (he had told me that because of his homeless condition some local populations believed he could neither read nor write).
This construction of Ghana as having attained global educational prestige is, once again, interactionally achieved by means of repetitions (see, e.g. the emphasis placed in repeating the name of the country, both by Benedito and Alfred, in lines 15 and 16). Besides, during the time of the fieldwork, 'Ghanaianness' was reinforced by the informants' mobilisation of key information about Ghana's importance in the international arena, such as the fact that it was the first sub-Saharan country to gain political independence from the UK; that the former Secretary General of the United Nations Kofi A. Annan was born there (and studied at KNUST, as they noted later on in our interview), all aimed at situating this country in the global map.

Conclusions: unequal Englishes and linguistic marginalisation materialised in 'English standardness' ideologies
In this paper, I have explored linguistic ideologies concerning the English-mediated multilingual repertoires of a network of three Ghanaian migrants and the ways in which these interplay with their transnational identity management, in a peripheral urban zone of a non-English-speaking bilingual society in southern Europe. I have done so from an informantoriented perspective which has put participants' self-reflexivity concerning their linguistic resources and communication acts at the forefront of a socially-engaged critical analysis. I have focused on their positionings towards their translinguistic practices involving English pluralisations as well as local and allochtonous codes in order to problematise essentialising constructions of languages as homogeneous bounded units 'belonging' to a particular fixed ethnicity (and territorial polity). In particular, I have focused on how the informants' English forms challenge 'outer'-'inner-circle' English-language dichotomies and de/legitimisations. I have shown that migrants' socialisation processes and 'integration' practices today are conducted through these counterhegemonic complex amalgamations of linguistic codes, though in ways that are subjected to, and in the end, get modulated or regulated by, local, nation-state and supra-state neoliberal language ideologies and dominant sociolinguistic normativities. I have argued that migrants' intercultural encounters take place through English varieties which constitute non-standard multilingua francas for socialisation across and beyond social networks, in public-transport benches which have become underexplored migrant-regulated spaces of silenced multilingualisms. These hybrid forms provide an understanding of the social meanings of non-elite languages which are frequently backgrounded but which are core in English-mediated multilingual practice, such as, for example, Ashanti, crucial, here, for migrants' gatekeeping and access to transnational subsistence resources (e.g. food, information and communication technology).
The analysis of the informants' discourses towards their own and the others' Englishes has provided an account for the ambivalent positionings that they show towards them. I have claimed that their gliding through acts of linguistic de/legitimisation in English indexes both linguistic assertiveness and insecurity, revealing how migrants voice, and cope with, the devaluing of their English-mediated non-standard varieties by other local migrant groups and by society at large (including institutions such as bureaucratic offices, NGOs or temporary-work agencies). I have called this devaluation linguistic marginalisation, and I have argued that these contradictory sociolinguistic comportments speak of these migrants' frequent de-languaging and de-skilling, which occurred not only when they were not conceived of as workforce for the tertiarised new economy but when they were positioned as non-schooled, 'illiterate' manual labourers who should command Spanish as the nation-state, locally legitimised, language (one of the informant's credentials as an English teacher, for instance, were totally ignored).
The self-legitimisation acts whereby they counteract linguistic marginalisation sheds light on the informants' degree of linguistic authority and social agency, as observed, for instance, when they defined what counts as legitimate ways of speaking and of being in the bench in linguistic competitions with other rivalling migrant networks; particularly, their use and vindication of the appropriateness of their 'outer-circle' English forms. These forms, however, reproduce traditional nativist conceptions of the language, since they are grounded on 'native speakerhood' constructions of linguistic codes and, thereby, in fact follow classic nationstate regimes of thought concerning (territorial, ethnolinguistic) 'ownerships' of languages.
When it comes to identities, migrants sometimes appropriated presentations of the Self such as the pauperised African migrant based on paternalistic conceptions of displaced migration from the 'underdeveloped' south. However, they also simultaneously inhabited pan-African social categorisations and cosmopolitan 'Ghanaianness' identities linked to 'blackness' and 'Englishness,' as well as to modernity, mobility experience, world knowledge and education.
Overall, this shows that situated forms of social distinction, difference and inequality among migrants living under precarious life conditions are entrenched in language (Piller, 2016). More specifically, it demonstrates that situations of marginalisation are linked to the censorship of transnational populations' non-standard practices and fluid identity enactments involving reterritorialised English forms. This allows us to better understand the degree to which unequal Englishes shape, and are shaped by, exclusionary sociolinguistic regimes of mind and hegemonic local and global ideologies linked to the racialising language policies and geopolitical orders which today dictate who count as legitimate English speakers and, ultimately, as citizenship-deserving Selves, in the resident societies of the twenty-first century. Notes 1. Catalan is a minority language in the sense that it has been historically, socioeconomically and politically 'minorised' (see Bastardas, 1996) today, for instance, it is not official in the European Union. 2. Inverted commas denote emic social categorisations.
3. The confidentiality of the data as well as the protection of the informants' identities were ensured by the Ethics Committee at the Autonomous University of Barcelona (file number 1818, 2012). 4. In Ghana, only 'dialects' have a name. The terms for African languages are modern inventions to meet the standards of Western variationist approaches to describe the linguistic codes of that area. The Akan language group belongs to the Volta Comoé languages, classified under three smaller clusters of 'dialects', all considered 'national', Ashanti belonging to the Central Comoé cluster (Kropp Dakubu, 2015[1988). 5. Reports suggest that Catalans have a 'medium'/ intermediate level of English, higher than the proficiency levels attributed to Italy and France, though lower than those of Northern European countries (EFSET, 2016).