Jill Vaughan

Arnhem Land, northern Australia, is a noted ‘hotspot’ for linguistic diversity – the region is home to many languages representing several diverse language families. Although a significant genetic boundary bisects the region, the languages form a Sprachbund evidencing grammatical, phonological and lexical convergence. Individual language repertoires in Arnhem Land are large, typically drawing on between three to six local languages, as well as English, a range of contact varieties, and local auxiliary sign language.
Data from Arnhem Land has periodically been used to exemplify a range of supposed ‘exceptionalisms’ (Koch 2014) which have been argued to characterise Australian languages more generally, such as high levels of lexical diversity, intensive borrowing (as well as unique motivations for it) (cf. Bowern & Atkinson 2012; Evans 2005; Harvey 2011), and ‘extraordinary’ levels of variation (Garde 2008). This paper considers two seemingly incongruous claims about the operation of variation and change in the region. First, it has been demonstrated that extensive and long-term grammatical, phonological and lexical diffusion has occurred (Heath 1978, 1981). It is suggested that this has been scaffolded by particular multilingual practices in the region, an apparent absence of social pressures to maintain linguistic distinction, and cultural/demographic factors such as small language groups and extensive cross-linguistic marriage (Heath 1978: 142-3, 1981: 365). Second, fine-grained systems of variation have been documented in many languages of Arnhem Land. Variation may be dimensionalised along diverse lines, including indexing patriclan (Garde 2008), moiety (Green 2003, Morphy 1977) and geographical oppositions (e.g. coastal vs. inland (McKay 2000)). Such variation is thought to be created and maintained by strong ideologies which value linguistic diversity, and possibly via the conscious promotion of particular variants (Evans 2003: 30; Garde 2008). Considered together, then, it would appear that Arnhem Land simultaneously demonstrates both significant convergence as well as systematic diversification/differentiation (cf. Gumperz and Wilson 1971, and Rumsey in prep on similar processes in the northern Kimberley region).
In this paper I consider evidence from north-central Arnhem Land to investigate the operation and interaction of these competing processes. Drawing on analyses of new data from Burarra, as well as existing documentation of Gurr-goni and Djinang (e.g. Green & Nimbadja 2015, Waters 1989), I explore patterns of convergence and systematic variation in these neighbouring varieties. I argue that both processes are indeed able to coexist, and that they demonstrably do so in the language ecology of this region. Reworking Labov’s (e.g. 1972) framework for categorising variables (stereotypes|markers|indicators), I suggest that differing levels of ‘salience’ are key to understanding how these apparently contradictory processes interact within a larger system. Finally, this paper will highlight the necessity of looking beyond the boundaries of single-language systems to broader language ecologies in understanding the functioning of variation, and its role in language change.

Linguistic complexity and linguistic diversity are not evenly distributed around the globe. In fact, hotspots of linguistic complexity and linguistic diversity often coincide. For example, some areas of the Amazon basin, Papua New Guinea and Northern Australia have a very large number of language families, and these areas are also home to languages with very high levels of linguistic complexity. Languages in areas like Arnhem Land (Northern Australia) have been claimed to have many phenomena that are low in functionality and also difficult to learn—such as extensive nominal classification (Sands 1995).
Communities that favour complexity are sometimes idealised as ‘small, isolated communities’ (Trudgill 2011, Dahl 2004) but communities in Arnhem Land are anything but isolated. The small number of speakers per language is misleading—language itself does not delimit social groups. In fact, people live in multilingual communities and social interaction is intense and occurs across an extensive network. Marriages, material and ceremonial exchange, and informal fostering create strong social ties and facilitate frequent interaction between people with a broad range of linguistic affiliations. Many small languages, from four language families, are used within a speech community characterised by ‘small-scale multilingualism’ (Evans 2010; Lüpke 2016).
Small-scale multilingualism is a term used to describe communities in which people are highly multilingual, but where there are no lingua francas or polyglossia. That is, local languages cannot be assigned unique domains of use, nor can patterns of language use be explained by placing languages on some hierarchy of social status. Instead, people may use different languages with different people, across similar contexts. The interlocutors, their linguistic repertoires, and their socio-linguistic affiliations are the main determinants of language choice in interaction (Singer & Harris 2016).
Explanations for the growth of linguistic complexity either attribute it to isolation and low-contact or to contact involving child bilingualism (Dahl 2004). Nichols (1992) identifies areas with many language families as a context that favours complexification because “diversity among neighbouring languages fosters complexity in each language” (Nichols 1992:192). Nichols’ model predicts that high linguistic diversity should coincide with high linguistic complexity. This predicts the situation in Arnhem Land, but we still do not know the nature of language contact on the ground and how this fosters diversity and complexity.
Research in Arnhem Land so far has reported a range of diverse language practices, including, for example, both code-switching and linguistic purism in different contexts (Vaughan and Carew 2016), receptive multilingualism (Singer and Harris 2016, Rehbein et al. 2012) and ongoing learning of new languages throughout the life span. This flexibility seems to be balanced by allegiance to the patrilect (father’s language), creating what Stanford (2009) calls ‘communities of descent’. In certain contexts, such as formal speeches or child- language socialisation, the affiliation with the patrilect comes to the fore, a bulwark against the multitudinous options for multilingual practices observed in other contexts.
This talk will present analyses of new data collected on language use and language ideologies in Arnhem Land, northern Australia. This includes corpus data from recordings of multilingual interactions as well as ethnographic data from interviews and participant observation. Drawing on these analyses, we propose a model of small-scale multilingualism in Arnhem Land that demonstrates how language practices and ideologies in this area support complexification and diversification. Juxtaposing the study of multilingual practices with research on language ideologies enables greater insights into the nature of language contact
as well as the social conditions that constrain these interactions, and prevent any one language spreading at the expense of the others (cf. Epps forthcoming).

In Maningrida, northern Australia, codeswitching is a commonplace phenomenon within a complex of both longstanding and more recent multilingual practices. Fourteen Indigenous languages representing three language families are spoken among 2500 people, alongside increasing use of English and contact varieties (such as Kriol, spoken further south in the Maningrida region, and right across northern Australia). Individual linguistic repertoires are typically large, but strong ideologies exist dictating rights/responsibilities around language ownership and use. A variety of code-mixing practices is observable between local Indigenous languages, and is now also widespread between local languages and English. Codeswitching has been a feature of the longstanding stable ‘egalitarian’ multilingual ecology of the region (Singer & Harris 2016), yet the practice is also symptomatic of a changing local language ecology, shaped by the large-scale incursion of English and implicated in the emergence of a local urban koine.
In this paper I consider whether general predictions about the nature and functioning of codeswitching account for practices in the Maningrida context, and for patterns attested across the diverse multilingual contexts of northern Australia more generally. I consider: (i) what patterns characterise (intraclausal) codeswitching practices between different Australian languages in the region, as opposed to codeswitching between an Australian language and Kriol or English? What (if any) generalisations can be made?; and (ii) are the distinctions observable accounted for by general predictions and constraints from dominant theoretical frameworks, or by the typological congruence of the languages implicated? Need we look beyond these factors to explain the patterns?
First I survey historical and contemporary descriptions of multilingual practices and codeswitching in the region. Little published work addresses codeswitching between Indigenous languages only, so I draw together available accounts from northern Australia (e.g. Coleman n.d.; Evans 2010; Haviland 1982; O’Keeffe 2016; Sutton 1978) to identify any tendencies, constraints and particularities. I consider to what extent observable patterns may be accounted for by established general principles, and whether regional language ideologies may also shape these outcomes of long-term contact. These data are compared with data from work on contemporary codeswitching between Indigenous languages, Kriol and/or English (e.g. Meakins 2012; McConvell 2002; O’Shannessy 2012) and I explore possible reasons for divergences in these practices. To this discussion I contribute data from multilingual interactions in Maningrida, where codeswitching is evident both between local languages and also with English. Results indicate that general predictions, along with the effects of typological congruence, account for many observable tendencies in the data (e.g. switch-sites). However other factors, such as constraints exerted by local ideologies of multilingualism and linguistic purism, as well as shifting socio-interactional goals, may help account for certain distinct patterns in the Maningrida data at least (e.g. lower incidence of certain types of codeswitching between Indigenous languages, despite higher typological congruence and shared multilingual repertoires). Data from this context present an opportunity to test theories and to explore the impact of social-psychological pressures of divergence and convergence in cultural and linguistic practice.

In Maningrida (north-central Arnhem Land), fourteen languages representing three language families are spoken alongside English and Kriol. This linguistic diversity reflects long-standing practices of multilingualism and diversification akin to ‘small-scale’ or ‘egalitarian’ multilingualism, and now accommodates more recent linguistic arrivals, the impact of newer patterns of mobility, and the reification of certain local languages in community institutions. Using a corpus of recent naturalistic language use and linguistic biography interviews, this paper considers how local valorisations of multilingualism and diversity contrast with the bilingualism-as-deficit discourses common in certain community and governmental institutions.

In Maningrida, north central Arnhem Land, fourteen Indigenous languages are spoken, alongside English and a range of contact varieties, in a community of 2500 people. Interaction is characterised by high levels of individual multilingualism, as well as receptive multilingualism and code-mixing practices (Carew 2016; Elwell 1982; Singer & Harris 2016). Local language ideologies dictate that variation is dimensionalised according to patrilineal land-holding units, into what might roughly be called ‘dialects’, with linguistic boundaries “couched principally in the idiom of local geography” (Sutton 1991).
Work on dialects in contact has largely been concerned with monolingual or migrant communities in western contexts, where it has been (arguably) feasible to approach dialects as varieties of some superordinate language, and to consider linguistic accommodation to be the primary/only driver in shaping new varieties (Labov 2001, Trudgill 2008). While several important exceptions exist (e.g. Gumperz & Wilson 1971, Stanford 2007), many questions remain regarding variation in multilingual, non-western or Indigenous contexts. In this talk, I consider the case of variation and dialect contact within Burarra, the language with the largest speaker community in Maningrida. Burarra is commonly described by speakers as consisting of four major ‘dialects’: An-barra, Martay, Maringa (not always distinguished as a distinct dialect) and Gun-nartpa (often considered a distinct language), with each label referencing groups of clan estates further east in Arnhem Land. Variation within Burarra is chiefly lexical, with further more minor phonological and syntactic differences also evident. Against this backdrop, a newer variety of Burarra has emerged in the urban context of Maningrida, a product of dialect levelling, partially stabilised code-switching with English, and the privileging of shared cross-linguistic lexical stock in the region. While this newer variety is now the first language for some, the traditional ‘dialects’ still co-exist with it and are central identity categories, reproduced through “communities of descent” (Stanford 2009).
Using recent interactional and elicited language data, as well as interviews and more general ethnographic work in the community, I discuss the extent to which the extant and socially salient variation within the Burarra linguistic space can be said to be reflexes of the various attested Burarra cultural categories, and the means by which these broader categories are performed interactionally. These patterns of variation are contextualised within the ‘total semiotic fact’ (Nakassis 2016, following Silverstein 1985) of cultural and social identity practice in the Maningrida region. Ultimately I will argue that the nature and functioning of variation in the region issues a challenge to assumptions of indexicality and lectal coherence (Guy & Hiskens 2016) in the dialectology literature, and point to the need to problematise the role of accommodation in language contact more generally.

In Maningrida, northern Australia, codeswitching is a commonplace phenomenon within a complex of both longstanding and more recent multilingual practices. Fourteen Indigenous languages representing three language families are spoken among 2500 people, alongside increasing use of English and contact varieties (such as Kriol, spoken further south in the Maningrida region, and right across northern Australia). Individual linguistic repertoires are typically large, but strong ideologies exist dictating rights/responsibilities around language ownership and use. A variety of code-mixing practices is observable between local Indigenous languages, and is now also widespread between local languages and English. Codeswitching has been a feature of the longstanding stable ‘egalitarian’ multilingual ecology of the region (Singer & Harris 2016), yet the practice is also symptomatic of a changing local language ecology, shaped by the large-scale incursion of English and implicated in the emergence of a local urban koine.
In this paper I consider whether general predictions about the nature and functioning of codeswitching account for practices in the Maningrida context, and for patterns attested across the diverse multilingual contexts of northern Australia more generally. I consider: (i) what patterns characterise (intraclausal) codeswitching practices between different Australian languages in the region, as opposed to codeswitching between an Australian language and Kriol or English? What (if any) generalisations can be made?; and (ii) are the distinctions observable accounted for by general predictions and constraints from dominant theoretical frameworks, or by the typological congruence of the languages implicated? Need we look beyond these factors to explain the patterns?
First I survey historical and contemporary descriptions of multilingual practices and codeswitching in the region. Little published work addresses codeswitching between Indigenous languages only, so I draw together available accounts from northern Australia (e.g. Coleman n.d.; Evans 2010; Haviland 1982; O’Keeffe 2016; Sutton 1978) to identify any tendencies, constraints and particularities. I consider to what extent observable patterns may be accounted for by established general principles, and whether regional language ideologies may also shape these outcomes of long-term contact. These data are compared with data from work on contemporary codeswitching between Indigenous languages, Kriol and/or English (e.g. Meakins 2012; McConvell 2002; O’Shannessy 2012) and I explore possible reasons for divergences in these practices. To this discussion I contribute data from multilingual interactions in Maningrida, where codeswitching is evident both between local languages and also with English. Results indicate that general predictions, along with the effects of typological congruence, account for many observable tendencies in the data (e.g. switch-sites). However other factors, such as constraints exerted by local ideologies of multilingualism and linguistic purism, as well as shifting socio-interactional goals, may help account for certain distinct patterns in the Maningrida data at least (e.g. lower incidence of certain types of codeswitching between Indigenous languages, despite higher typological congruence and shared multilingual repertoires). Data from this context present an opportunity to test theories and to explore the impact of social-psychological pressures of divergence and convergence in cultural and linguistic practice.

Work on dialects in contact has largely been concerned with monolingual or migrant communities in western contexts, where it has been (arguably) feasible to approach dialects as varieties of some superordinate language, and to consider linguistic accommodation to be the primary/only driver in shaping new varieties (Labov 2001, Trudgill 2008). While several important exceptions exist (e.g. Gumperz & Wilson 1971, Stanford 2007), many questions remain regarding variation in multilingual, non-western or Indigenous contexts. I consider a case of dialect contact in Maningrida, Arnhem Land, where approximately 14 languages are spoken in a community of 2500 people. Interaction is characterised by high levels of individual multilingualism and code-mixing practices. Variation is traditionally dimensionalised according to land-holding units, into what might roughly be called ‘dialects’, with linguistic boundaries “couched principally in the idiom of local geography” (Sutton 1991). Within this highly multilingual context, three dialects of the Burarra language have contributed to an ‘urban’ koine. While this newer variety is now the first language for some, the traditional ‘dialects’ still co-exist with it and are central identity categories, reproduced through “communities of descent” (Stanford 2009). Using interactional and elicited language data, as well as interviews and more general ethnographic work in the community, I describe the nature and interpretation of current dialectal variation within the Burarra linguistic space. I further consider the continuity of practices and ideologies driving diversification and the maintenance of difference in this context, and demonstrate the need to problematise the role of accommodation in language contact more generally.

In Maningrida, a remote, largely Indigenous, community in northern Australia, code-switching is a commonplace phenomenon that sits within a complex of both longstanding and more recent multilingual practices. Approximately fourteen languages are spoken among Maningrida's 2500 inhabitants, and interaction is characterised by high levels of individual multilingualism and various code-mixing practices. While a unified communilect does not exist, convergences are observable in terms of shared community language features. The task of describing the nature and functions of code-switching in such a context is complicated by a number of factors: code-switching has been described as a feature of stable 'egalitarian' multilingualism in Australia, and as such may constitute a reflex of the maintenance of linguistic boundaries, shaped by broader ideologies privileging linguistic diversity and distinctiveness. However, it can also be a practice symptomatic of a changing local language ecology, a precursor to language shift or the emergence of a contact variety. Further complicating factors in Maningrida include the presence of shared lexical items which can blur the lines between borrowing and code-switching, and the nature of dialectal variation in the region, which may rely partly on the indexical values of 'shibboleth' forms. This paper will focus in particular on Burarra-English code-switching in Maningrida, which is implicated, along with dialect-levelling processes and shared cross-linguistic stock, in the formation of an urban koine variety of Burarra. We draw on long-term linguistic and ethnographic fieldwork in the community to describe the shape and functions of this form of partially-conventionalised code-switching, as well as constraints (both grammatical and social) on its use. Burarra-English code-switching within this context provides an opportunity to explore competing social-psychological pressures of divergence and convergence in cultural and linguistic practice. As such, we demonstrate what this particular site of linguistic practice can reveal about the complexities of code-switching on the ground.

In remote northeast Arnhem Land, children grow up in a highly multilingual and multidialectal environment. Yirrkala is a community of approximately 700 people where most speak a form of Yolŋu Matha, a pan-dialectal cover term encapsulating approximately 30 Indigenous language varieties. Clan groups each have their own varieties (as well as their own homelands, totems, ceremonies, songs etc.) which, while in linguistic terms might be considered dialects of Yolŋu Matha, are typically understood by speakers to be distinct languages (Wilkinson 1991: 5). These clanlects are mutually intelligible, distinguished largely by grammatical morphology and some minor, though highly salient, lexical differences (Amery 1993), and shaped by a kind of long-term ausbau process (Garde 2008). Speakers of at least 18 of these clanlects currently reside at Yirrkala.
Traditionally, children learn their mother’s and grandmother’s varieties at a young age. As they become young adults, they begin learning their father’s language, their own clan variety. The learning of these clanlects is a crucial step in learning about rom (ceremonial law/customs) and culture, and becoming strong in Yolŋu identity. Children learn to identify an interlocutor’s clan affiliation by their speech, which helps them determine their relationship with others.
This paper explores how children and young adults understand and deploy dialectal variation within the Yolŋu Matha linguistic space, and examines how linguistic coherence is created among multiple variable phenomena (Guy and Hinskens 2016). We consider how these practices are impacted by the current language ecology of Yirrkala where a further, newer Yolŋu Matha variety (not a clanlect), the koiné language Dhuwaya, is now spoken by all community members, regardless of clan affiliations. Relying on linguistic and ethnographic fieldwork, as well as targeted interviews in the community, we attempt to elucidate the practices and ideologies that shape the role and nature of variation in this context.

In this paper we report on the design and implementation of an interview-based study into caregiver attitudes and beliefs about child-language acquisition in two remote Aboriginal Australian communities. Language socialisation plays a crucial role in children's cognitive, social and cultural development, and the child-caregiver relationship is a fundamental site in this process. As in any study of language socialisation, a key focus is to gain insight into the underlying values and beliefs that govern social interaction within a given community. Identifying these can be difficult for cultural-insiders because they must 'denaturalise' normal behaviour, as well as for cultural outsiders who must do away with their own preconceived, deep-seated assumptions about child development and caregiver/child interactions. Adopting an approach that avoids these twin pitfalls, then, can be challenging. This paper discusses the development of an appropriate methodology for undertaking a language socialisation study with sensitivity to cultural and individual differences in two communities in Aboriginal Australia. We focus in particular on interviews targeting parents' and caregivers' attitudes and beliefs about child-language acquisition and development; both shared community ideologies, and individuals' beliefs and values guiding language socialisation routines. The two communities included in our study represent very different language ecologies, but together exemplify the sociolinguistic variation that typifies Aboriginal Australia. The first is Barunga, a small community in Central Northern Australia where Kriol is the predominant language spoken. The second is Maningrida, a community of around 2000 people in Arnhem Land where around fourteen distinct languages are spoken, with most individuals able to speak between two and six (Elwell 1982; Handelsmann 1996). Both communities featured have recently undergone/are currently undergoing distinct configurations of language loss, and they differ significantly in the levels of multilingualism observable. We argue that the method should include open-ended questions, designed to elicit narrative-like descriptions, as well as questions and prompts to direct the interview along specific themes. To identify the most meaningful themes we considered features found in language socialisation literature from various cultural backgrounds (e.g. Ochs and Sheiffelin 1984; Kulick and Schieffelin 2004), themes arising from literature on child development in Aboriginal Australia (e.g. Hamilton 1981; Philpott 2003; Byers et al 2012; Kruske et al 2012) and, crucially, we relied on the outcomes of local community collaboration in designing the study. The resulting interview structure ranges across four main themes: (i) the 'development of learning/milestones'; (ii) multi-party vs. dyadic input and competence; (iii) accommodation to the infant, (verbal and environmental); and (iv) autonomy of the infant. As well as discussing the development of our interview methodology, we present initial findings from the interviews themselves. In particular, we discuss the implications of the findings for the ongoing maintenance of creoles and endangered languages as home languages. We explore insights into local family language policies HOLM 2016 Abstracts – Oral Presentations

In this paper we report on the design and implementation of an interview-based study into caregiver attitudes and beliefs about child language acquisition in two remote Aboriginal Australian communities.
Language socialisation plays a crucial role in children’s cognitive, social and cultural development, and the child-caregiver relationship is a fundamental site in this process. As in any study of language socialisation, a key focus is to gain insight into the underlying values and beliefs that govern social interaction within a given community (Ochs & Sheiffelin 1984).
The two Arnhemland communities included in our study
The first is Barunga, a small community where a variety of Kriol is the predominant language spoken. The second is Maningrida, a community of around 2000 people where fourteen distinct languages are spoken, with most individuals able to speak between two and six (Elwell 1982; Handelsmann 1996).
The interview structure ranges across four main themes: (i) the ‘development of learning/milestones’; (ii) multi-party vs. dyadic input and competence; (iii) accommodation to the infant, verbal and environmental; and (iv) autonomy of the infant. These themes were identified from Language Socialisation literature (e.g. Kulick and Schieffelin 2004) and literature on child development in Aboriginal Australia (e.g. Philpott 2003; Hamilton 1981; Kruske et al 2012; Byers et al 2012; Lowell et al 1996).
Thus far the interviews have indicated that the stages of childhood development can be characterized in locally salient ways, for example beibiwan ‘baby’; biganini ‘small child’ and bigwanbigwan ‘grown/adult young person’ at Barunga. These terms are defined in terms of physical development as well as the child’s capabilities, such as clear speech, being able to care for ones self, as well as specific cultural milestones such as Initiation Ceremonies. We found that interview questions worked best when specific children from target age range were referred to elicit developmental stage e.g. “Would you use baby talk to ‘children like John?’ (children aged 2-3)”. We also found that questions that contained inherent comparison or judgments of other people were not appropriate, for example the following two questions were deleted after workshopping with co-researchers in the community:
1. What kinds of things do mums and dads have to think about that people without kids don’t need to think about?
represent very different
language ecologies, but together exemplify the sociolinguistic variation that typifies Aboriginal
Australia.
  220
2. Are some kids better talkers than others? What are they like?
We also found that some of the most successful questions were those that elicited spontaneous hypothetical (or real) narratives e.g. “When you see a baby (1-2 yrs old) John’s age do something dangerous what would you say to them? Can you give an example story?”
As well as discussing the development of our interview methodology, we present findings from the interviews themselves and contextualise this study as a starting point for an ongoing broader language socialisation study in both these communities.

Multilingual practices, such as code-switching/mixing and receptive multilingualism, enable speakers to draw upon the range of codes within their linguistic repertoire to construct linguistic and social meaning in interaction. This is particularly the case in Aboriginal Australia, one of the most multilingual regions in the world, where traditional languages continue to be spoken alongside newer contact languages (e.g. Kriol and mixed languages), Aboriginal English and Standard Australian English (see, e.g., Koch & Nordlinger 2014). As such, speakers in these settings have many codes to choose from in their linguistic meaning-making.
This paper investigates the social and pragmatic functions of the linguistic codes of Maningrida, Arnhem Land, a predominantly Aboriginal region of Australia’s Northern Territory. Maningrida has extremely high levels of multilingualism due to its role as a linguistic and socio-political nexus of many distinct groups from further afield in Arnhem Land. Indeed, Maningrida is in some ways a microcosm of traditional Arnhem Land affiliations and hostilities, many of which are still apparent in ongoing interactions and attitudes. At last report (in 1996), fourteen distinct languages are spoken there, with most people able to speak between two and six (Elwell 1982; Handelsmann 1996).
In focusing particularly on the practices of code-switching and mixing, we investigate which languages are currently being used in Maningrida, how distinct codes are drawn upon in interaction, and the particular social meanings they may have for speakers. In particular, we consider a small set of multilingual texts collected from both private and public domains in the community in order to elucidate indexical properties and grammatical constraints (where relevant) of code-switching/mixing as practiced in Maningrida. Close examination of the choice and strategic use of codes in interaction is combined with data from qualitative interviews with speakers (regarding language ideologies and practices), as well as broader ethnographic work in the community, to elucidate the indexicalities of the codes in question (e.g. Bucholtz & Hall 2008). This provides a targeted means of investigating how social meanings are constructed more generally.
These findings are considered within the context of existing work on code-switching/mixing both in Aboriginal Australia (e.g. McConvell 2008; Meakins 2012) and beyond. This study further contributes to understanding the enduring role of ‘traditional’ linguistic practices that contribute to ongoing stable multilingualism in Aboriginal Australia, as well as the inextricable link between language, land and identity in this region, wherein sociolinguistic variation is dimensionalised into the geographical and social cosmos (Evans 2013). These extant yet traditional practices now interact with newer cultural and linguistic pressures, such as the increasing use of Standard Australian English, in a variety of ways, and within this context code-switching/mixing can be a symptom of language in a state of change but can also be characteristic of an enduring multilingual language ecology.

Language socialisation plays a fundamental role in children’s cognitive, social and cultural development. Learning language, and being able to use it to manage the world, is critical to the healthy cognitive and linguistic development of all children. Two of the most important sites of language socialisation are the home and the school classroom. In Aboriginal Australia, children typically enter school with a wide range of linguistic codes and, for the most part, are expected to adjust to the Standard Australian English (SAE)–dominant environment of the classroom (Lowell & Devlin 1998, Rhydwen 1992). A small body of research has addressed the kinds of issues Aboriginal children subsequently face in classroom participation, in part resulting from this transition, such as the discussion around whether children are equipped to deal with the nature of questions typical of school interactions (see, e.g., Christie 1985, Malcolm 1982, Moses & Yallop 2008).
This paper presents insights from research in four Aboriginal communities: Murray Downs and Ltyentye Apurte in Central Australia, Wurrumiyanga on the Tiwi Islands, and Yirrkala in north-east Arnhem Land. These communities represent diverse language ecologies, and together exemplify the sociolinguistic variation that typifies Aboriginal Australia. We first consider the linguistic codes that children in each of these communities use at home, and how these pattern onto expected code-use in the classroom. These repertoires vary significantly across the sites due to: (i) the different traditional language spoken in each (Alyawarr, Arrernte, Modern Tiwi and Yolngu Matha respectively); (ii) the differing role and nature of Kriol in each (i.e. whether it is used at all, whether varieties are acrolectal/basilectal); and (iii) the linguistic environment of the school (the differing role of SAE, e.g. Yirrkala School has a bilingual program).
We further discuss the effect of children’s linguistic repertoires on their ability to participate effectively in the classroom, and the key issues faced by children in each community during and beyond this transition. In the early years of schooling, children worldwide acquire more developed linguistic skills (such as advanced use of pronouns, determiners and demonstratives and more complex syntactic structures) as well as increased pragmatic and metalinguistic competence. This process is complicated when there is the additional challenge of acquiring a new code. The linguistic situation faced by children in much of Aboriginal Australia is distinct, however, from what has typically been considered in the literature on early education and L2 acquisition (e.g. the experiences of immigrant children learning SAE). In contrast, the creoles and mixed languages many Aboriginal children speak at home have some features in common with SAE, and children usually have prior exposure to the standard code (Heath 1983, Siegel 1992). As such, the experiences of Aboriginal children in this regard constitute an under-researched area that this study will contribute to illuminating.
This work builds on findings from the first phase of the Aboriginal Child Language Acquisition project (ACLA1, see Simpson & Wigglesworth 2008), which detailed the complex linguistic environment children are raised in by investigating children’s language input in three Aboriginal communities (Yakanarra in the Kimberley, Dagaragu in the Victoria River District, and Tennant Creek). In combination, these two groups of studies provide a much clearer picture of how children are exposed to and learn to manage their linguistic repertoires across a variety of contexts in Aboriginal Australia, and the ways in which the first language (acrolectal vs. basilectal creoles, mixed languages, traditional languages) impacts language behaviour and integration in the school environment.

This paper addresses the ways in which the Irish language is recruited as a means of resistance in communities in Ireland and the Irish diaspora: resistance to enforced or presumed identities, to pressures of cultural assimilation, and to prevailing and hegemonic discourses.
Almost 2 million people in the North and South of Ireland identify as Irish speakers and an estimated 70 million around the globe can claim Irish heritage. While Irish ancestry may be distant for many, the Irish language is active in numerous communities, as documented in some limited research (e.g. Ihde 1994, Ó hEadhra 1998, Noone 2012) and evidenced by the existence of cultural and language groups. Census figures indicate that over 30,000 people currently speak the language in Canada, the United States and Australia alone, yet no general account of Irish-language use in the diaspora exists. Linguistic practices within Irish communities worldwide vary widely with regard to Irish-language use, with each community subject to distinct concerns, histories and discourses, and hence with different possibilities for creating social meaning through language use. This paper is therefore located within a broader research project providing a synchronic sociology of the Irish language worldwide, and examining constructions of socio-cultural identity among those learning and using the language in different communities.
Research was carried out in 7 contrasting field sites spanning the Republic of Ireland (Galway and Dublin), Northern Ireland (Derry and Belfast), and the diaspora (Melbourne, Australia; Boston, U.S.; St. John’s, Newfoundland, Canada). Data is drawn largely from semi-structured interviews with 93 Irish-language users across the field sites, and is supplemented by questionnaires and participant observation in each. Discourse analysis (e.g. Hook 2001) of the data’s emergent themes allows for an exploration of the dominant discourses which influence language choices in these communities, as well as in distinct communities of practice within them. Positioning theory (e.g. Harré et al. 2009, Tirado & Gálvez 2007) is used to investigate how Irish speakers situate themselves within these discourses by taking up or resisting available subject positions, and thus how use of the Irish language is made meaningful as a social action in local environments.
This paper examines the ways in which resistance is enacted within discourses specific to each field site, as well as discourses that span across communities. I discuss how Irish takes on local meaning in each location: for example, providing a tool to construct oppositional identity in Newfoundland (in opposition to Newfoundlanders descended from English migrants; in opposition to a generalised Canadian identity); expressing resistance to presumed majority identities in Northern Ireland; and providing a means to symbolically resist pressures to assimilate culturally in diasporic contexts in Boston and Melbourne. I also present ways in which Irish use may be recruited more generally as an act of resistance to dominant discourses via the language’s position within counter-discourses. As such, in addition to contributing to a broader description of Irish-language communities worldwide, this paper argues for an incorporation of social constructionist approaches to identity within sociology of language, particularly with regards to elective bilinguals and in diasporic contexts.

It is estimated that 70 million people around the globe can claim Irish heritage and, although Irish ancestry may be distant for many, the Irish language is active in numerous communities worldwide. The current vitality of Irish as a minority language within Ireland has been addressed in many studies, however research on Irish-language use within diasporic communities is relatively scarce, documented in some limited research but widely evidenced by the existence of cultural and language groups. This paper reports on research conducted in Ireland (North and South) and in several communities in the Irish diaspora (in Australia, the U.S. and Canada) on constructions of socio-cultural identity among those learning and using the language.

Research was conducted through qualitative interviews with Irish-language users and participant observation in a variety of Irish-language environments. Data collected provides the basis for a synchronic sociology of the Irish language worldwide, revealing differing patterns of usage in distinct communities of practice, some regionally specific and some determined by other factors. Content analysis of the data’s emergent themes allows for an exploration of dominant competing discourses that impact on Irish and minority-language use, including discourses of ethnicity, tradition and minority. These discourses are examined in their historical and political contexts and considered as frames within which the Irish language is made meaningful as a social action in local environments.

LOLspeak is a complex and systematic reimagining of the English language. It is most often associated with the popular, productive and long-lasting internet meme ‘LOLcats’. This style of English is characterised by the simultaneous playful manipulation of multiple levels of language in order to perform an authentic ‘cat’ voice, and serves both as an entertaining in-group practice and as a cultural index which is recruited in the construction of identity.
Using community-generated web content as a corpus, we analyse some of the common speech play strategies (Sherzer 2002) used in LOLspeak, which include morphological reanalysis, atypical sentence structure and lexical playfulness. The linguistic variety that emerges from these manipulations displays collaboratively constructed norms and tendencies providing a standard which may be meaningfully adhered to or subverted by users.
Building on this, we use Bucholtz and Hall’s (e.g. 2005) interactional and ethnographic approach to linguistic analysis to examine how the speech play strategies used by participants allow for the simultaneous construction of two identities: firstly the identities of the cats that they claim to be speaking for, and secondly their own identity as savvy members of an online community of practice (Jones & Schieffelin 2009).

This paper reports on research conducted within the Irish language community in Australia on constructions of socio-cultural identity among those learning and using the language. This study forms part of a broader research project aiming to provide a synchronic account of identity in Irish language learning and usage in Ireland and the Irish diaspora.

Many studies have addressed the current vitality of Irish as a minority language within Ireland (e.g. Ó Flatharta et al. 2009, Ó Giollagáin et al. 2007, O Néill 2005), however research on the status of Irish in the Irish diaspora is relatively scarce. It is estimated that 70 million people around the globe can claim Irish heritage and, while Irish ancestry may be distant for many, the Irish language is active in numerous communities, as documented in some limited research (e.g. Ihde 1994) and evidenced by the existence of numerous cultural and language groups. Data from the last two Australian censuses show a slight increase in Irish speaker numbers from 828 in 2001 to 918 in 2006, but the status of Irish in Australia is similar to the situation Ihde noted in the U.S., that is to say characterised by small pockets of native speakers that will vanish unless replenished by new immigration, and also by larger groups of language learners who in general will not reach a high level of Irish language competence.

The current study investigates the role of the Irish language in the construction of diasporic identity, as well as reporting on the role of diasporic language use in revitalisation efforts in Ireland. Research is conducted primarily through questionnaire-based interviews with learners and users of the Irish language in Australia, where information is elicited regarding informants’ demographic, Irish language use, language attitudes, ethno-cultural activities, ancestry/migration history and personal identification, as well as the motivating factors in their language use. Identity in this study is viewed as a dynamic and emergent social construction, which relies on the negation and negotiation of the ‘other’ (Woodward 2002, Bucholtz and Hall 2005) and which, within the context of the diaspora, is subject to “necessary heterogeneity and diversity” (Hall 1993: 401) and heightened contingency, indeterminacy and conflict (Gilroy 1997). Ager’s (2001) model of motivation in language planning, with components of motives, attitudes and goals/needs, provides a framework for accounting for motivating factors in language use.

Results demonstrate that established narratives and discourse within the Irish diaspora that tend to privilege the migrant, and especially the Irish-speaking migrant, as white, Catholic, nationalist, straight and working-class may not reflect the reality of the demographic, and that prevalent notions of diasporic nostalgia, with attendant phases of nationalism and the fetishising and repeating of the past do not provide a complete account of factors in Irish language use. Comparisons are drawn between identity constructs identified in this study and those established in other surveys conducted in Ireland and key areas of the Irish diaspora (Boston, U.S.A.; Newfoundland, Canada).

Ager, Dennis. 2001. Motivation in language planning and language policy. Clevedon: Multilingual Matters Ltd.
Bucholtz, Mary and Kira Hall. 2005. Identity and interaction: a sociocultural linguistic approach. Discourse Studies, 7(4-5): 585-614.
Gilroy, David. 1997. Diaspora and the detours of identity. In Kathryn Woodward (ed) Identity and difference. London: Sage. 299-343
Hall, Stuart. 1993. Cultural Identity and Diaspora. In Patrick Williams and Laura Chrisman (eds) Colonial Discourse and Post-Colonial Theory: a Reader.  Hemel Hempstead: Harvester Wheatsheaf. 392-401.
Ihde, Thomas. 1994. The Irish language in the United States: a historical, sociolinguistic and applied linguistic survey. Westport, CT: Bergin & Garvey.
Ó Flatharta, Peadar, Caoilfhionn Nic Pháidín, Colin Williams, Francois Grin, and Joseph Lo Bianco. 2009. 20-Year Strategy for the Irish Language. Prepared for the Department of Community, Rural and Gaeltacht Affairs. Dublin: Fiontar, Dublin City University.
Ó Giollagáin, Conchúr , Seosamh Mac Donnacha, Fiona Ní Chualáin, Aoife Ní Shéaghdha, and Mary O’Brien. 2007. Comprehensive linguistic study of the use of Irish in the Gaeltacht : principal findings and recommendations : a research report. Prepared for The Department of Community, Rural and Gaeltacht Affairs. Dublin: National University of Ireland, Galway (Acadamh na hOllscolaíochta Gaeilge).
O’Néill, Diarmuid. 2005. Rebuilding the Celtic languages: Reversing Language Shift in the Celtic Countries. Talybont: Y Lolfa Cyn.
Woodward, Kathryn. 2002. Understanding Identity. London: Arnold.

Keywords: language endangerment, language and identity, language attitudes, Irish, diaspora

Many studies have addressed the current vitality of Irish as a minority language within Ireland, however research on Irish-language use within diasporic communities is relatively scarce. It is estimated that 70 million people around the globe can claim Irish heritage and, while Irish ancestry may be distant for many, the Irish language is active in numerous communities, as documented in some limited research (e.g. Ihde 1994) and evidenced by the existence of cultural and language groups. This paper reports on research conducted in Ireland (North and South) and in several communities in the Irish diaspora (in Australia, the U.S. and Canada) on constructions of socio-cultural identity among those learning and using the language.

Research was conducted primarily through questionnaire-based interviews with Irish-language users. Data collected provides the basis for a synchronic sociology of the Irish language worldwide, revealing differing patterns of usage in distinct communities of practice, some regionally specific and some determined by other factors. Qualitative analysis of the data’s emergent themes allows for an exploration of dominant competing discourses that impact on Irish and minority-language use, including discourses of ethnicity, tradition and minority. Positioning theory (e.g. Harré et al. 2009) provides a means to investigate how Irish speakers situate themselves by taking up or resisting subject positions within these discourses in their historical and political contexts and how use of the Irish language is made meaningful as a social action in local environments.

Ihde, Thomas. 1994. The Irish language in the United States: a historical, sociolinguistic and applied linguistic survey. Westport, CT: Bergin & Garvey.
Harré, Rom, Fathali M. Moghaddam, Tracey Pilkerton Cairnie, Daniel Rothbart and Steven R. Sabat. 2009. Recent Advances in Positioning Theory. Theory & Psychology, 19(1): 5-31.