1. Top of page
  2. Abstract

ABSTRACT:  This paper examines the use of English in computer-mediated discourse, and considers the extent to which the traditional varieties-based approach to the study of global English can provide adequate analytic tools for describing the actuality of communicative encounters conducted on the Internet. Starting from the contention that languages as discrete entities are a problematic sociolinguistic category, the paper addresses the question of how the study of global English may be refined to accommodate the type of English-related communicative phenomena found in many online encounters. Drawing upon examples of communication between Thai speakers via social networking and instant messaging services, the paper outlines the complexity of English-related forms in this genre of online interaction, and considers the metatheoretical questions this complexity poses for the discipline of world Englishes in terms of how best to describe and categorise such phenomena.


  1. Top of page
  2. Abstract

This paper examines the extent to which the ontology of language that is posited in mainstream sociolinguistics, as well as analytic approaches to the study of linguistic interaction which adopt this ontology, can provide adequate conceptual means for describing the actuality of many of the communicative encounters that occur via online technologies, and in particular those involving English. It addresses the question of how the academic study of worldwide English, which, in its formative stages as autonomous discipline traditionally took a predominantly variety-centric approach towards the modelling of the language's global spread (Seargeant, forthcoming), is being refined to accommodate the type of English-related communicative phenomena increasingly practised in online encounters. In pursuing these issues, the intention is both to contribute to explorations of this aspect of the nature and use of English in the world today (e.g. Androutsopolous 2007, forthcoming; Blommaert 2011), and also to offer metatheoretical reflection on how the discipline can best describe and categorise such phenomena depending on the purposes it is aiming to pursue.1

The starting point for this investigation is the contention made by a number of scholars in recent years that languages as discrete entities are a problematic sociolinguistic category (e.g. Mühlhäusler 2000; Pennycook 2007; García 2009). In his conclusion to a discussion on the relationship between globalisation and the English language, Alastair Pennycook (2010: 121) suggests that:

Perhaps it is time to question the very notions that underpin our assumptions about languages … to ask whether the ways we name and describe languages as separate entities, the ways we view bi- and multilingualism, are based on 20th century epistemologies that can no longer be used to describe the use of languages in a globalizing world.

The argument he is pursuing here is that the meanings of many of the key categories that are used to describe linguistic behaviour – categories such as ‘languages’, ‘varieties’ and ‘dialects’– are a product of a particular set of events in Western intellectual history. As such they are not scientifically ‘real’ so much as historically determined, and thus their unproblematised use in modern sociolinguistics or applied linguistics (or indeed in any subfield within the language sciences which takes a social view of language) can be an impediment to or complicating factor in the development of a nuanced theoretical description of the actuality of people's everyday language practices.

The argument revolves around the premise that the concept of discrete languages is a consequence of the trend in 18th century European political philosophy to promote the nation state as the principal politico-cultural unit (Anderson 1983). One element of this worldview was that the language practices of a community were promoted as an essential aspect of its cultural (and thus political) identity and, given that the principal unit of community was the nation state, so the key unit in which languages were measured was one which was coextensive with the boundaries of the nation state. This led to the ideology of idealised ‘national languages’, which were identified with a particular standard that was codified in dictionaries and grammar books (Milroy and Milroy 1999; Blommaert 2008a), and this ideology underpinned the theoretical thinking which developed in the 20th century into the modern discipline of linguistics. And although sociolinguistic research today regularly acknowledges the idealised nature of such categories (Swann 2007), they are nevertheless deeply embedded in both the popular perception of the existence of language practices around the world, and in the ontological presuppositions of much academic research into language and society (Seargeant 2010a).

There is a further stage in this argument, though. For while the category of discrete languages – and the conceptualisations of particular named languages – are, in the above sense, historical products, and thus cannot be taken to unproblematically represent actual linguistic phenomena, they do nevertheless reflect a dominant ideological interpretation of the way in which language operates in society. In other words, these categories represent both linguistic phenomena (albeit not in an absolute word-to-world isomorphic relationship) and a particular aspect of social organisation (based around the notion of a national community). The argument continues, however, that this framework of social organisation is no longer valid in the way it was during the 19th and 20th centuries (Blommaert 2009), and thus the categories which have been inherited from the ideological climate of that time are inappropriate on two fronts: both as scientific representations of linguistic phenomena; and as social scientific representations of the role that language plays in the life of a vast number of people in 21st century, globalised societies.

The limitations of these categories are especially evident in areas of the social world where change is at its greatest and, as a consequence, language practices are rapidly evolving. Two prime instances of this are: (1) the global spread of English and the diverse linguistic practices that have resulted from it; and (2) the linguistic practices of computer mediated discourse (CMD). In both these areas, however, academic study has, at least until recently, primarily analysed linguistic practice using a varieties-based approach which operates with the basic ontological assumptions that developed during the 19th and 20th centuries. In this paper, we examine communication which occurs at the intersection of these two areas – that is to say, English-related linguistic interaction which takes place over the Internet – and explore the extent to which a revised theoretical approach is becoming necessary for the description and explanation of the linguistic phenomena displayed in such interaction. Drawing upon examples from communication between Thai speakers via a social networking site and instant messaging service, the paper aims to outline the complexity in the use of English-related forms in online interaction, and in so doing, address the following two questions: what does the use of English on the Internet indicate about English in the world today? and what implications does it have for the models that have been traditionally used for analysing the nature and use of English in diverse world communities? As such, the paper has a primarily metatheoretical purpose, though with the added intention that the analysis of the chosen data (which is illustrative rather than substantive in the paper) can identify a cluster of issues that would benefit from further in-depth empirical research.


  1. Top of page
  2. Abstract

In the late 1970s and the 1980s, the innovation of referring to ‘Englishes’ instead of ‘English’ did much to break down the myth of a single monolithic English language, and thus better reflect the ‘sociolinguistic realities’ of postcolonial territories where the language now exists as an established part of the linguistic ecology (Kachru 1991). The title of Tom McArthur's 1998 book, The English Languages, reflects this move, foregrounding as it does the multiplex nature of the language, while Kachru's highly influential Three Circles of English model talks in terms of ‘indigenised varieties’ (Kachru 1992). In both cases, the ascription of the status of a discrete ‘language’ or ‘variety’ to the distinctive usage patterns of diverse global communities did much to legitimise this usage. The act of naming the language, and the categories used as part of this act, were thus a central concern in the way that English-related linguistic practices were conceptualised (Seargeant 2010b), and helped to effect a shift in perception within the discipline about the status and validity of different English usages around the globe.

More recently, the paradigm which replaced the monocentric view of the language with this pluricentric one has itself been criticised for being too ‘varieties’ based. Scholars such as Bruthiaux (2003), Pennycook (2007) and Park and Wee (2009) note that, paradoxically, the ‘world Englishes’ approach uses the same type of categories (albeit not the exact same categories) as previous approaches to the study of the language. For example, the Three Circles model deals with language primarily at the level of the nation state, thus necessarily overlooking much of the variation that occurs within countries or regions. Furthermore, in actual practice, people often mix English with other languages in an ad hoc manner, adding English-related words and phrases while nominally speaking other languages in a way which reflects transnational cultural flows (Pennycook 2007). With English now having a global status, this phenomenon occurs with ever greater frequency, and disrupts the neat categories such as Indian English or Singaporean English with which Kachru's model works. The call that accompanies this critique is for what might be called a ‘post-varieties’ approach: an analytic apparatus that is sensitive to the dynamic communicative practices which use English-related forms and connotations as one part of a wider semiotic repertoire.

In the second half of this paper we will examine data that illustrate this type of interaction and consider its implications for analytic accounts of the use of English on the Internet. First though, to determine what a ‘post-varieties’ approach might comprise, it is necessary to examine what constitutes a ‘variety’, and why and how the term is used as an analytic concept within the discipline.

What is most commonly assumed in speaking of a variety is that there is a recognisable system of linguistic features which can be associated with a community of speakers. Swann et al. (2004: 324), for example, define a variety as ‘A linguistic system used by a certain group of speakers or in certain social contexts’. Crystal (1995: 408) offers a similar definition: ‘A term used […] to refer to any system of linguistic expression whose use is governed by situational variables’.2Swann et al. (2004: 324) continue their definition by noting that ‘“Variety” is often used as an alternative to dialect and language, and can be a useful way of circumventing the difficulty of making a clear distinction between the two on linguistic grounds’. This draws out the way that the difference between a ‘language’, a ‘dialect’ and a ‘variety’ is as much a political as a linguistic one, and that to refer to something as a language rather than a dialect is usually thought to afford it more respect (Lippi-Green 1997). Political considerations aside, however, both variety and language share the notion of a distinguishable system, and it is this that constitutes the core of their definition.

This is what one might call a Saussurean approach to linguistic analysis, in that it projects a synchronic system onto the use of language. The codification of the language is then premised on what Harris (1981) calls the ‘fixed code’ view of language, instantiating the idea of the synchronic system by inscribing it in grammars and dictionaries. And despite provisos about the abstracted or ‘fictional’ nature of ‘varieties’ (e.g. Algeo 1991), this framework of conceptual categorisation underpins several other sociolinguistic concepts such as mixing, borrowing, and hybrid languages, where the notion of discrete systems is the conceptual foundation for the metaphors used to describe particular linguistic phenomena. As Blommaert (2010) argues, for the sociolinguist, the danger becomes inadvertently reifying this key conceptual category by employing it in an uncritical fashion. For language planners, on the other hand, reification can often be the goal. For example, national language academies or national dictionary projects will purposefully attempt to circumscribe a particular usage which can then be promoted as distinctive of the national community (Seargeant 2010c).

In discussing the use of ‘variety’ as a key conceptual category in sociolinguistics, what we are dealing with is a question of how term and phenomenon relate and the purposes for which the term is used. The complexities of the relationship between term and phenomenon can usefully be analysed by means of what Searle (1985), building on the work of J. L. Austin and G. E. M. Anscombe, calls ‘direction of fit’. The sociolinguist is aiming for a ‘word-to-world’ direction of fit, whereby the term employed (the ‘word’) most satisfactorily (or usefully) represents the actuality of the phenomenon (the ‘world’), and can thus be used as an accurate conceptual category for analysis. The question thus becomes whether there exist within the flux of experiential reality natural kinds which correspond to the concept referenced by the term ‘variety’. The language planner, on the other hand, at least in cases such as that of the national language academy or national dictionary project, is attempting to temper the diversity and dynamism within the language practices of a community, and thus achieve a ‘world-to-word’ direction of fit; he or she is aiming to make language practice adapt to the normative regulations codified in the dictionary or other planning apparatus. To put it another way, for the sociolinguist, the term ‘variety’ or ‘language’ is of optimum value if it accurately corresponds to the actuality of the sociolinguistic phenomena so that it can be used to analyse these phenomena; for the language planner or policy maker, the term is useful in the way it represents an ideal which practice can aim towards for the purposes of social cohesion or other political or cultural ends (e.g. contributing to an ideology of national unity).

In practice, of course, a clear-cut dichotomy between the two opposing directions of fit is not apparent in the way a language is perceived in society, either in academic contexts or in the popular imagination. As noted above, a varieties-based approach to language does not simply aim to represent the actuality of sociolinguistic practices, but also operates as an element in the mechanism of social organisation. As such, the use of the term ‘variety’ will likely have a combination of both directions of fit simultaneously, with the balance between the two being dependent on who is using the term in what context and for what purpose. For example, in a language policy context, the world-to-word direction of fit for the purposes of social cohesion may be to the fore, yet the word-to-world direction of fit aiming at adequately representing the linguistic phenomenon will also be a factor.

In language sciences which are predominantly descriptive one would expect the phenomenal-representation function to be key, and any other function to be an inconvenient (albeit to some extent inevitable) inheritance of the history of the term. This is not always the case, however. In critical linguistics – a linguistics which examines and critiques the relationship between language and (implicit) political agendas – the social cohesion or social construct aspect is also important. For example, the Three Circles model aims both to reflect the ‘sociolinguistic realities’ of English in the world (in a somewhat generalised way), but in so doing it also intends to change perceptions about English in the world and to legitimate non-native language practices. It does this by postulating discrete varieties which are, in a sense, idealisations around which language planners and policy makers within Outer or Expanding Circle contexts can rally.

What, then, are the possible problems with a variety-centric approach? Or to put it another way, in what contexts is the category of the variety analytically limiting rather than productive? One key issue here is the extent to which the (scientific or policy) discourse and the phenomena do fit, and whether the conceptual terminology and categorisation is prejudicing the scope of research and analysis. This is certainly the conclusion that Pennycook (2007: 137) comes to from his analysis of global hip-hop:

The mixed codes of the street, the hypermixes of hip-hop, pose a threat to the linguistic, cultural and political stability urged by national language policies and wished into place by frameworks of linguistic analysis that posit separate and enumerable languages.

In other words, in Pennycook's judgement the terminology which both national language policies and mainstream linguistics uses does not fit with the phenomena that he has researched, at least in terms of word-to-world direction of fit. And this lack of fit thus suggests the need for a revised analytic toolkit. In our analysis of the data below, we will consider the extent to which the limitations that Pennycook identifies in relation to his data are valid also for the range of English-related linguistic practices that occur via the Internet, and how, from a theoretical point of view, a conceptual vocabulary derived from a variety-centric approach operates within this context. As with Pennycook's focus on hip-hop, the use of English on the Internet occurs in contexts in which new communicative genres are being established, and communicative practices are adapting accordingly. For this reason it offers compelling data for sociolinguistic analysis of the diverse ways in which English is being used around the world. Before we move to the data, however, let us first consider how research into online communication has mostly negotiated these conceptual issues up until now.


  1. Top of page
  2. Abstract

In recent years a growing body of research has focused on the description of what Herring (2001) calls ‘computer-mediated discourse’ (CMD), a term she distinguishes from computer-mediated communication (CMC) through its focus on ‘language and language use’. CMD can have a wider remit, but research has, in practice, largely been limited to text-based interaction (rather than voice or image) which is mediated by computers. Although CMD is largely described in the literature in terms of interaction between discrete ‘varieties’ and ‘languages’, the same research also reveals a range of dynamic and varied practices and a creative mix of linguistic resources.

Within the tradition of empirical and descriptive research into CMD, the focus of much early work was implicitly on computer-mediated English (see, for example, Ferarra et al. 1991; Yates 1996), and there was an assumption that CMD itself could be described as a discrete and homogenous variety of English. Herring (2001; 2007) has written extensively about the initial tendency to assume that statements about a particular instance of Anglophone CMD could be generalised to all online language use. The comparative studies which tended to dominate this early research focused on the extent to which electronically-mediated usage could be compared and contrasted with non-electronic usage/varieties, often at broad levels of generality. Studies placed it ‘somewhere’ on a spectrum between writing and speech and talked about the ‘blurring’ of the traditional dichotomy between these two modes (e.g. Baron 1998; Crystal 2001). Baron (1998: 134) posed of email the questions that tended to dominate the field:

Is email a variety of speech (as many people are claiming)? What important new properties does it share with writing? Does it have emergent qualities that are unlike those typifying speech or writing?

This approach gave way to the conceptualisation of CMD as comprising what Herring (2002) calls ‘socio-technical modes’,3 according to which linguistic variables are related not to existing geographically and politically conceived language varieties, but to particular configurations of other situational variables. Herring's (2007) ‘faceted classification system’ classifies ‘socio-technical modes’ according to features of the medium such as synchronicity and size of message buffer (i.e. the number of characters permitted per message), as well as social factors such as the demographics of the participants and the nature of their relationships, the degree of anonymity, the purpose or topic of the online communication, and the language, dialect and writing system (or script) used. ‘Modes’ within this context can thus be defined not only by the electronic medium but by the myriad of different ways in which the technology is being appropriated by different user communities. The scheme seeks to operationalise the complex interplay of factors shaping online practices and, with a focus on the technological and social features of the ‘mode’, lays out factors which cut across nationally and regionally constructed varieties. Online features which span varieties of English, as well as other languages, include: informality expressed through the written representation of speech-like features; emoticons; orthographic variation including phonetic representations, contracted forms and vowel repetition to indicate emphasis, as well as unconventional punctuation (Anis 2007; Deumert and Masinyana 2008; Tagliamonte and Denis 2008). Differences are seen to arise not so much between languages or dialects, but between different ‘modes’; for example, synchronous ‘modes’ are typically more speech-like and contain more unconventional orthography than asynchronous ones (Danet and Herring 2007). Research in the field has thus moved from assumptions about the general character of an ‘internet English’ or ‘digital English’, to focused examinations of the complex of sociolinguistic factors which are regularly observed in (and thus constitute the conceptual category of) CMD.

Another salient theme in Internet research has been the extent to which one language, English, dominates over and excludes others. Surveys carried out by both marketing companies and academic researchers suggest that the World Wide Web can largely be divided up between a small number of languages, and that English at present accounts for the largest share (e.g. Nunberg 1998; Lavoie and O’Neill 2000; Internet World Stats 2011). The use of the category ‘English’ in this context is, however, not without its problems in that it again ignores the linguistic diversity which is typically subsumed by the term, whether this be the use of different regional or social varieties or emergent community-based linguistic and semiotic practices. The implicit assumption in much of this research that English is a homogenous category in turn legitimises the wider framework used in the surveys, whereby online languages are held up as discrete entities which can be quantified and compared.

Some researchers have pointed out that Internet surveys mostly exclude the often private and synchronous interactions that take place through virtual games, online chat and instant messaging (Paolillo 2007). This is significant because it is predominantly in these interactions, rather than on the public, asynchronous Web, that speakers have been observed to engage in multilingual practices such as code-switching (Fung and Carter 2007; Androutsopoulos, forthcoming; Paolillo, forthcoming), and thus their exclusion casts doubt over the extent to which linguistic diversity on the Internet is being accurately measured by Web-based surveys. There is often scant acknowledgement in such surveys of the fact that the vast majority of ‘English speakers’ around the world use it alongside or in combination with other languages and so are likely to be engaged online ‘in a mixture of languages’ (Danet and Herring 2007: 4).

As Androutsopolous (forthcoming) points out, studies into online code-switching not only extend the initial focus on English in the CMD literature, but also offer rich data for challenging precepts in the code-switching literature, including those concerning the existence of discrete linguistic systems. Several researchers into spoken code-switching (e.g. Gardner-Chloros 1995; 2009; Auer 1998; Heller 2007) have questioned the metaphor which underlies the term – that of moving between distinct codes – which, they argue, is best viewed as ‘an abstraction derived by linguists’ from the mixed language practices of multilingual speakers (Gardner-Chloros 2009: 165). As with the case of spoken code-switching, then, it would seem advisable that researchers of multilingual CMD should be wary of reifying languages as abstract entities and look instead at the ways in which users’ linguistic (and semiotic) repertoires are deployed in locally situated interactions (cf. Li 2011). However, the Internet is arguably more than simply another domain in which code-switching occurs, and not only because it ‘represents a language contact situation of unprecedented scale’ (Paolillo 2007: 424), particularly for the written mode (Androutsopolous, forthcoming). It is also a place where languages and scripts can be mixed in new ways, and where this mixing plays an important role in indexing identity given the relative absence of paralinguistic and social cues (Georgakopoulou 1997; Androutsopolous, forthcoming; Tagg and Seargeant, forthcoming).

For instance, one striking example of an emergent localised form in online language use is the inclusion of Roman numerals in ‘Colloquial Romanised Arabic’ (CRA; that is, the representation of colloquial Arabic in the Roman script). The numerals 2, 3, 5 and 7 are used to render phonemes in Arabic which are difficult to represent using the Roman alphabet (Palfreyman and Al Khalil 2007; Warschauer et al. 2007). They are selected either because their pronunciation corresponds with the target phoneme or because of their visual analogy with the corresponding Arabic letter (e.g. <5> or <7> can be used to represent the sound /x/ (as in Scottish loch)). This innovative use of CRA indicates a realignment of scripts and languages: as Warschauer et al. (2007) point out, Egyptian Arabic was, before the Internet, chiefly a spoken language and, if it appeared in written form, it did so overwhelmingly in the Arabic script. As such, they argue that its increasing online use in Romanised form appears to be challenging traditional sociolinguistic distinctions. In particular:

the advent of the Internet could be one factor, together with other socioeconomic changes (e.g. globalisation), that contributes toward a shift from traditional diglossia in Egypt to increased multilingualism, with both English (from ‘above’) and Egyptian Arabic (from ‘below’) encroaching on the traditional dominance of Classical Arabic in written communication. (Warschauer et al. 2007: 315)

CMD studies of this sort reveal an emergent, dynamic and varied discourse in which traditional usages both in terms of the form and meaning of language varieties are modified, or even subverted, and users draw creatively on a range of semiotic resources to produce new communicative practices. At the same time, however, this discourse continues for the most part to be described and analysed in the literature by means of a terminology which reflects pre-digital, modernist conceptions of language use. For example, Fung and Carter (2007) record various ways in which Hong Kong students studying in English-speaking countries mix English and Cantonese in ICQ exchanges (one-to-one chat). Although they describe the emergent discourse as a ‘hybrid’ or ‘mixture’, throughout the analysis they conceptualise the practice in terms of the adoption of one language (Cantonese) into the discourse of another (English), using concepts drawn from a varieties approach: borrowing, loan words, code-switching. As was noted earlier, the paradox here is that, despite the way in which such code-related terminology has been challenged by a number of researchers of multilingualism (e.g. Heller 2007; García 2009), the hybrid practices identified in CMD often continue to be conceptualised in terms of choices and switching between different linguistic systems.

In the next section, we turn to examples of such discourse from our own data in order to consider in greater depth the nature of this ‘mixed’ use of English-related forms in online encounters; and in the final section we will reflect upon the problems this usage raises for traditional varieties-based conceptualisations of language use.


  1. Top of page
  2. Abstract

The examples in this paper consist of three exchanges between a number of Thai speakers via both the status updates and comment function on the social network site Facebook, as well as MSN communications via Blackberry phones. They were collected as part of a larger body of data during a three month period from January to March 2010 from interactions between members of one of the research associates’ network of Thai-speaking online friends. These examples were selected from the larger corpus (which totals approximately 40 exchanges in all) as being representative of communication patterns amongst the network (see Seargeant et al. forthcoming, for the broader and more detailed study). The Thai elements of the exchanges are translated in parentheses after each turn, with explanatory notes added to gloss particular cultural allusions or other features of note. Informed consent was gained from all participants, and names have been anonymised.

The participants in these exchanges were located either in Thailand or the UK. They were all between the ages of 20 and 32 at the time of the research. Further specific details of their biography and relevant circumstances are given before each individual extract. In all cases, the participants have Thai as their first language, but have studied English to at least higher education level, either in Thailand or in an Inner Circle country.

The data have been analysed by identifying features of the discourse which are commonly associated either with a particular context of use, or with a particular variety or ‘mode’ (as such varieties and ‘modes’ are described in the sociolinguistics literature).4 Feature identification uses a comparative approach between different notional systems of linguistic patterning, and so reference to different codes, varieties, styles and ‘modes’ is necessary despite the fact that, as has been noted on a number of occasions, a central concern for the paper is the problematisation and complexifying of just these conceptual categories. (We will return to the seemingly paradoxical nature of this aspect of the approach, and a possible means of resolving it, in the concluding section of the paper.)

An overriding observation is that all these elements are mixed constantly in the discourse. Although the focus of the paper is primarily on the metatheoretical issues involved in the conceptualisation of linguistic interaction in this context – and thus, as noted above, the data analysis is intended as illustrative rather than substantive – some consideration has also been given to what motivates the shifts in style or code, and whether they have a particular communicative significance. Environmental factors such as the restrictions imposed by certain aspects of the technology (the lack of Thai script input on certain generations of smart phones, for example) can partially explain these switches. Further insight into this motivation has been sought by means of the second aspect of the method, the use of participant interviews.

Interviews with one of the participants who was present in all but one of the exchanges was used to supplement the textual analysis at a few selected points. She was asked to reflect on her understanding of the meaning of the conversational exchanges and of various individual utterances, as well as being asked to speculate on the motivation, in so far as she was able retrospectively to rationalise this, for the shifts in style and code. In this way the analysis has adopted what Blommaert (2008b) calls an ‘ethnographic epistemology’; that is, an examination of participant beliefs about the discourse as cultural practice as these emerge and function within their cultural context. The intention in pursuing this additional approach was to probe how the participants understood the interaction and the practices involved (i.e. how it was emically conceptualised), and use this as a point of cross-reference for the features analysis. It should be noted at this stage that, given the space limitations of the paper, the analysis foregoes detailed examination of the interactional meaning of different linguistics features. The complexity of this aspect of the interaction (including as it does the interrelation of different languages, modes, scripts and technologies) warrants further, dedicated investigation. Omission of this line of analysis is not meant to signal however that the patterns in linguistic behaviour are being interpreted as arbitrary or indiscriminate.5 One point of particular note to come from the interview data is that the informant considered that oral interactions between the same participants would invariably be conducted almost entirely in Thai, with very little use of English or English-related forms at all. This provides an interesting point of contrast for the nature of the written online communication as illustrated in the examples.

Extract 1

This first extract comes from a conversation between two female participants, Dream and Cherry, via MSN on their Blackberry phones. Neither had Thai script capabilities on their phones for this exchange, so had to use Roman script. Cherry had recently finished an MSc in Marketing (English Programme) in Thailand, where she was still residing. Dream was living in London pursuing postgraduate studies. They are discussing the state of Cherry's love life.

 1 Cherry: oh

 2 Cherry: Ken-noi ngai [How about Ken-Noi?6]

 3 Dream: Mai wai la [No way]

 4 Dream: Too young

 5 Dream: They are the same age as my students loei

 6 Cherry: shouldnt b phd …

 7 Cherry: should b undergrad …

 8 Cherry: but everything too late now

 9 Cherry: herr [<sigh>]

10 Dream: Why don't u have a bf?

11 Cherry: i have

12 Cherry: hahaha

13 Cherry: but i want exciting thing banggg

14 Dream: I can't believe. U r cute mak mak na [You are cute!]

15 Dream: Gu wa laeww [This is what I think]

16 Dream: Dee mak I will tell ur bf dee gua lol [Great. I’d better tell your boyfriend about your plan]

17 Cherry: oh

18 Cherry: no la no laaa [No!]

19 Cherry: jai rai [You are mean]

20 Dream: Eeh eeeh [Ha ha ha]

This extract illustrates a markedly ‘mixed’ discourse, in which Anglophone forms are prominent, though very closely combined with Thai and shaped by CMD and other informal conventions. Features of note are a distinctive ‘Thai English’ usage, both in the use of sentence-end discourse particles such as ‘loei’ (turn 5), ‘banggg’ (turn 13) and ‘laaa’ (turn 18) (Smyth 2002), and syntactic patterns in which auxiliary and copular verbs and articles are dropped (e.g. turns 8 and 13). In reflective comments on this exchange the interviewed participant considered this to be the default English register for her online interactions with her Thai friends, and felt that it indexed what she referred to as a ‘Thainess’, that is, a shared culturo-linguistic identity.

Other features of note are the use of several English-word contractions (e.g. turns 6, 10, 14), non-standard punctuation (e.g. the lack of capitalisation and apostrophes in turns 6 and 7 respectively), as well as orthographic play (particularly the repetition of letters) in the transliterated Thai (turns 13, 18). All these are features commonly found in CMD, and many are cited as linguistic variables in the ‘socio-technical modes’ which Herring (2002; 2007) has identified.

There is also both intersentential and intra-sentential code-mixing, as, for example, across and within turns 14–16. The interviewed participant suggested that the motivation for her switch into Thai in turn 15 was because the word ‘Gu’, which means ‘I’, is part of a very familiar register, thus making its use here not very polite. Her intention at this point was to express annoyance with Cherry, and she felt this could be conveyed more easily in Thai by exploiting these particular politeness registers which English does not possess. Elsewhere in the extract there is the mixed use of culturally-specific ways of capturing spoken interjections. For example, turns 9 and 20 are both Thai-specific, but the ‘oh’ of turn 1, on the other hand, as well as the ‘hahaha’ in turn 12, could either be (transliterated) Thai or English.

A final point of interest relates to the way the Thai language is rendered in the Roman alphabet. As there is no universal standard for transliterating Thai into the Roman script, instead people commonly use what is colloquially referred to as pasa karaoke or ‘karaoke language’. This is a form of ad hoc transliteration using approximate phonetic rendering of the Thai words, and is so called because of its resemblance to the phonetics used in Thai karaoke videos (Mayes 2009). This results in variation within the spelling of the same words (e.g. ‘loei’ in turn 5 of this extract is rendered as ‘loey’ elsewhere in the data), and innovative attempts to replicate aspects of Thai orthography – specifically elements of the tone system – in the Roman alphabet (examples of which we will look at in Extract 2).

Extract 2

The participants in this second extract are Dream and Tee, and they were conversing via MSN. Tee was doing a PhD in Neuroscience in Thailand at the time, and for this exchange he had access to a computer which allowed the input of Thai script. Dream, on the other hand, was writing from her phone so could not input Thai script. In the interviews during which she was asked to reflect on the interactions, Dream commented that because Tee is older than her, she felt it more appropriate to write in Thai rather than English. Dream was again London-based at the time of this exchange.

 1 Dream: Jai raii ah [You are mean]

 2 Dream: Gin rai pai ja?[What did you eat?]

 3 Tee: blueberry cheese pie

 4 Dream: Dinner nia na?[As your dinner?]

 5 Tee: inline image[Yes]

 6 Dream: Save some for me pao?[Did you save some for me or not?]

 7 Tee: inline image[No]

 8 Tee: inline image[Nok’ll get fat (Nok is Dream's nickname)]

 9 Dream: Auan kor me kon love :D [I will be loved although I am fat]

10 Tee: inline image[Really?]

11 Dream: Uhmmm

12 Dream: Young ngai mother, p’ni, p’na kor love [(My) mother, Nini (Dream's sister), Nana (Dream's

   other sister) love me anyway]

13 Tee: inline image[Yes]

14 Dream: Laew p’tee la? :P [How about Tee?]

15 Tee: inline image[What?]

16 Dream: Mai mee arai [Nothing]

17 Dream: :(

18 Dream: Cher![<a sound showing anger>]

19 Tee: inline image[What?]

20 Dream: Nok ngon [Nok disappointed]

21 Tee: inline image[What makes you disappointed?]

22 Dream: Ngon mai mee krai replies my facebook messages loei [(I’m) disappointed (because) no one replies to my Facebook messages]

23 Dream: :’(

24 Tee: inline image[What's wrong with Nok?]

25 Dream: Arai kong p’mhee [What's wrong with P’Mhee?]

26 Dream: Kor p’mhee has never replied my fb messages [P’Mhee has never replied to my Facebook messages (P’inline imageis an honorific term for one's elder)]

27 Dream: Hurrr [<a sound indicating sadness>]

28 Tee: inline image[What?]

29 Tee: inline image[What's wrong with Nok?]

30 Dream: Nok saddd

31 Dream: Nok mai dai tickets for the show [Nok didn't get tickets for the show]

32 Dream: T____T’

33 Dream: Mai mee kai son jai noooo loei :’([No one pays attention to me]

34 Tee: inline image[(I’m) tired]

35 Dream: Ok

36 Dream: Because of me lor?[Because of me?]

37 Dream: :(

38 Tee: inline image[many things]

39 Dream: ({). Hug hug naa [Let me hug you]

40 Dream: Mai disturb u laew naka [I won't disturb you any longer]

41 Dream: Bye byeee

42 Tee: inline image[Are you logging off?]

43 Dream: Kor u r busy [Because you are busy]

44 Dream: And don't wanna chat with me nee

45 Tee: inline image[a bald Nok (In Thailand there is the belief that the bald tend to be sensitive)]

Again, this is a very mixed discourse, with influences from both English and Thai, as well as a variety of CMD conventions. There are a number of similar features to Extract 1. These include contractions (e.g. turn 43); orthographic play (especially the repetition of letters) in both English and Thai (turns 30 and 33); a distinctive ‘Thai English’ with sentence-end discourse particles (e.g. turns 6 and 36); intersentential and intra-sentential code-mixing (e.g. turns 2–4, 31, 43); and culture-specific onomatopoeia (turns 18, 27), as well as examples which are shared across both cultures and languages –‘Bye byeee’ in turn 41 is an English-derived loanword in Thai, and so, from a coding point of view, could be ascribed to either language. Also in this extract there is the use of a range of different emoticons (for example turns 9, 14, 17 and 23), thus adding a multimodal dimension to the communication.

Again the significance of the switches between Thai and English has, at a general level, interpersonal significance, at least in the rationalisation that the informant gave in the interview reflecting on this data (where she saw it as a consequence of the hierarchically-structured age relationship between the two of them). This is in line with work on code-switching which has analysed the structure and motivation of switching in great detail (e.g., Myers-Scotton 1993; Deumert and Masinyana 2008). In combination with the technologically-determined constraints, this explains some of the motivations behind switches between code, script and style.

The interviewee also suggested, however, that certain of the switches were made primarily for reasons of communicative efficiency rather than being semiotically significant. In turn 31, for example, the use of the word ‘tickets’ in English avoided an ambiguity that would exist if it had been written in transliterated Thai. Written in the Roman script the word would be ‘tua’, but as this fails to represent the tone mark on the word, she thought it could easily be misinterpreted. A similar motivation lies behind the use of the English word ‘hug’ in turn 39. In Thai this would be ‘kord’ which is a word which has several homophones, so using English avoids ambiguity.

The rendering of tone marks in the pasa karaoke, although cited as a reason for using an English word in the explanation above, is a point of note in other features of the data. The honorific  inline image, which is used when addressing one's elders, has a tone mark (˙) above the letter. In this extract, Dream uses an apostrophe to represent this, as for example in p’tee. Elsewhere in the data other participants use a period instead, thus producing forms such as p.tim (the honorific plus the diminutive form of Tim). With the rendering of these features, the lack of a conventionalised system of transliteration means that the interlocutors are drawing in a semi-individualistic way on the semiotic possibilities available to them from the technology, thus creating rather than following discourse norms.

Finally, this exchange also includes one salient instance of an anomalous choice of script which is not determined by the affordances of the available technology. Tee uses Thai script throughout, except for turn 3, where he switches to Roman script for ‘blueberry cheese pie’. This is an established loan-phrase in the Thai lexis, but rather than write it in Thai, Tee has chosen to use the English form. It is unclear from the textual data alone why he has done this. It is possibly due to the fact that, despite being an established borrowing, it still retains strong foreign connotations which make English seem more appropriate.

Extract 3

The final extract is from an MSN conversation between Dream and Big. Big only studied English up to university level, so has a comparatively limited fluency. At the time of the exchange he was located in Thailand, while Dream was again in the UK. For this exchange both were in front of computers with Thai keyboards, but Big initiated the conversation in English, and they continued using the Roman script throughout. The topic of the conversation is Valentine's Day.

1 Dream: I got a bar of white choc

2 Dream: Someone put it in the envelope and wrote my name on it

3 Dream: And put it in my flat postbox

4 Big: 555 Don't eat much na [Don't eat too much]

5 Dream: I ate it up laew lol [I ate it up already lol]

6 Dream: Aroi duey [(It was) delicious]

This is the clearest case of a situation where English and the Roman script have been chosen with no obvious influence from the technology, as for both participants Thai script was an option. Dream speculated that interpersonal considerations were a motivating factor in the choice, as the use of English had a distancing effect which suited a slight strain in the relationship as it was perceived by Big. The extract is part of a longer conversation, which was begun by Big who chose, from the outset, to use English. However Dream also suggested that choice of code was motivated in part by what she termed ‘laziness’. For example, she noted that ‘laew’ in turn 5 is shorter to write than its English translation ‘already’, and likewise ‘aroi’ in turn 6 is composed of less letters than the English equivalent ‘delicious’.

There are a couple of further points of note. The first is the use of numerals (via their phonetic reading) to represent words. In Thai, the number 5 is pronounced ha (inline image), and thus ‘555’ is used for the sound of laughing. In the next turn, Dream uses a standard English Internet contraction ‘lol’ (‘laugh out loud’). Of note here is that in Excerpt 1 turn 16, Dream used the same form (‘lol’) but in that instance it was operating as a transliterated Thai question particle. A mixed use of codes and scripts can thus result in identical forms having different meanings, namely, in a type of homograph which is the product of the convergence in a single communicative space of semiotic resources with diverse and varied provenances.


  1. Top of page
  2. Abstract

As explained in the introduction, this paper has a two-fold focus: offering a description and initial examination of a particular aspect of the diverse nature of English in the world, and discussing metatheoretical issues concerning the categorisation of this type of linguistic practice. In summary for the former of these two objectives, the examples we have looked at display a number of intriguing features about the way that English is incorporated into a wider semiotic mix of communicative practices. A first point of note is the extent to which English – or at least, English-related forms what Blommaert (2010) refers to as ‘bits’ of language – is being used, despite the participants all sharing an L1 (Thai); and the assertion by the informant that in the oral mode it is likely that Thai would be used almost exclusively. For this translocal group, geographically dispersed at the time of the interactions but with shared cultural roots and mobility patterns, the use of English appears to have become an intrinsic element of their online literacy practices, and seemingly offers a broader range of semiotic opportunities (such as the indexing of different degrees of social distance, and the flexibility to overcome the limitations of available technologies) than Thai alone would.

Another point evident here is that the English being used is not, nor does it appear to predominantly orient towards, a monolithic Inner Circle variety of the sort often used as a teaching standard. Although there are phrases which appear standard-like in form, these are embedded elements of a discourse which also displays a diverse range of non-standard, and even nonce, forms. Thus, although Thailand qualifies as an Expanding Circle country, the language practices on display do not, in toto, constitute elements of a norm-dependent variety. Conversing in an informal and unmonitored context, there appears to be little compulsion amongst these interlocutors to comply with Inner Circle norms as these are codified as teaching standards, and instead the discourse is, in both the complexion and complexity of its mixing, sui generis if viewed as a ‘variety’. As such, this can perhaps be seen as an example in which ‘vernacular practices of digital literacy’ act as a site for unregulated and individualised ‘vernacular linguistic expression’ (Androutsopolous 2011: 206), rather than one in which the participants replicate an offline vernacular variety.

Functional ascription for the choice of code is also not always immediately apparent. As noted, common sense suppositions about the shared speech community of the participants are complicated by the fact that English-related forms are used extensively. The restrictions of the technology can in part explain the choice. Where the input of Thai script is unavailable, the interlocutors have to transliterate into the Roman alphabet if they wish to write in Thai, and the natural association between the Roman script and the English language may thus lead to the use of English in some instances. Yet this is not a general rule – English forms, as well as Thai written in Roman script, occur in exchanges where these technological restrictions were not apparent, and in some instances the same speaker switched between languages and scripts within the same conversation. The indexical value of different codes and scripts – and the interpersonal meaning generated from them in specific contexts – was also given as a motivating factor behind certain choices, as were factors such as communicative clarity (the need to avoid ambiguity), convenience (or ‘laziness’) and creativity (exploiting the formal or connotative affordances of different codes and scripts). These motivations reflect those expressed by participants in other studies to explain similar language practices (e.g. Warschauer et al. 2007). However, as in other studies, it is important to recognise that these explanations constitute post hoc justifications of choices which may or may not have been evident to participants as they drew on their bilingual repertoires; and that there remain instances in which the motivation is obscure, as well as linguistic features which straddle the different languages and cannot easily be assigned to one or the other. The overall result, however, is a highly ‘mixed’ written discourse, which broadly reflects the globalised identities of the participants (in each exchange, at least one of the participants had experience of living for an extended period outside Thailand), as well as the affordances and emergent norms of the medium of communication.

As noted then, this type of language use does not fit neatly into the categories employed by the Three Circles of English model. Nor does it fit with studies quantifying linguistic practices on the Internet in terms of discrete languages. Both approaches conceptualise language usage from more of a long-range perspective, and make use of general categories (named national or transnational languages). As such they are not equipped to deal adequately with this type of data. Furthermore, the language use on display here cannot productively be categorised as a stable or emergent variety in its own right (one could not, for example, usefully talk of it as an ‘online Thai English’) because of the diverse ways in which the different codes, different styles, different scripts, and different semiotic modes combine. Herring's concept of the ‘socio-technological mode’ is helpful in highlighting the type of regularities that do occur in CMD, but the ad hoc nature of many of the instances of code-mixing, script choice and transliteration conventions, as well as the sheer extent to which mixed forms are utilised, means that it appears almost impossible to discern a core set of stable features in the corpus which could be described as systematic.

We are left with what appears to be a paradox then. We can describe the linguistic phenomena on display here by means of a terminology based around varieties – we can (mostly) identify which features belong to English or to Thai, we can match syntactic patterns with specific cultural usages and perhaps even varieties (‘Thai English’), and we can identify structures and features which are often associated with online discourse. In each case, the methodology for our analysis is premised on the logic which underlies the conception of the variety: namely, we are aiming to identify what appear to be systematic regularities that can be associated with a particular community of language users. In other words, all the features analysed above are ones commonly found – and described – in sociolinguistic studies. Yet the phenomenon as a whole – the actual discourse – cannot be subsumed under the category of a variety. It exhibits too much diversity; it does not have obviously identifiable systematic regularities and, given the variation on display it seems unlikely that one would be able to predict with consistent accuracy how shifts in style, script choice, mode and code take place. In other words it is not, in itself, an emergent ‘variety’. Rather, it is a communicative act which draws on available semiotic resources in a semi-improvised way, exhibiting certain very broad regularities in terms of the constraints of the technology and the mutual competencies and orientations of its participants, but otherwise drawing in sundry ways on features from different ‘systems’. We appear, then, to be at the intersection of what is regular (i.e. systematic) and what is free-flowing and possessed of a complexity which, in epistemic terms at least, evades being captured by a generalised conceptual terminology. And it is for this reason that the term ‘post-varieties approach’ seems appropriate. For what we are engaged in is a descriptive exercise which builds upon the logic of the varieties approach, but in doing so problematises the validity of the category as a description of a sociolinguistic phenomenon.

To an extent, studies of code-switching and code-mixing have been engaged in such an exercise for a long time, describing and explaining mixed language practices that occur in contact situations (e.g. Auer 1998; Heller 2007). Yet, as noted above, here too there has been discussion about the theoretical implications of the modernist models of language description upon which this approach is built, and recent reformulations of code-switching have attempted to provide models – or at least a metaphorically-based theoretical language – which acknowledge the fuzziness of the boundaries between languages (cf. Canagarajah's 2006 notion of ‘code-meshing’; García's 2009 notion of ‘translanguaging’) and which avoid approaching multilingual usage via a theoretical perspective grounded in assumptions about language based on monolingual practices.

In summary, then, the issue at the heart of this discussion can be formulated as the following question: If significant amounts of English use around the world today resemble in some form or other the nature of this discourse (i.e. if much worldwide English usage is of a similarly ‘mixed’ nature to these examples), does the category of the ‘variety’ simply become an instrumental way of describing aspects or features of the discourse, or does it still have a phenomenological reality? One could argue that standard Englishes are, doubtless, varieties, in that they are accompanied by an apparatus within society (dictionaries, grammar books, the education system) which regulates and promotes their autonomy and systematicity. Furthermore, they exist as conceptual categories for users of languages, and thus have an ideological existence in this form. But in the case of the data we have looked at here it would seem counter-productive to begin with a varieties approach when documenting this as an example of English-related linguistic behaviour. Instead, a conceptual methodology which focuses from the outset on the diverse semiotic resources employed, and examines the way the various features respond to specific contextual influences, offers an opportunity for a more nuanced understanding of the ways in which the spread of English manifests itself in online communicative encounters in the era of globalisation.

There is one further point to append to this discussion, however. What has not so far been mentioned in the analysis is world-to-word direction of fit. This is because the concept of the category (‘variety’ or ‘language’) is not explicitly referenced or discussed by the interlocutors in this data. (The category is discussed and used by the informant in her reflection on the exchanges, but this usage is ‘outside’ the interactions themselves.7) In other words, metalinguistic issues are absent from the interactions here, and thus do not have a bearing on the metapragmatics which order the meaning-making of the communication. Yet it is in the context of world-to-word direction of fit that the concept of the variety retains much of its potency. The fact that this data does not fit with the Three Circles model does not invalidate the model if one views it in the context of critical theory, as attempting to reorder the concerns of the discipline and promote non-native usages as legitimate and worthy of dedicated research. This is an important point to stress. For while a post-varieties approach might be appropriate for the detailed sociolinguistic analysis of specific examples of data, the notion of the variety and the status this has as an element of social organisation may yet be more appropriate for the description of a community's language use in contexts where the focus is not simply on strategies of communication but also on cultural and political identity.

In conclusion, the category of the variety is not in any sense an unproblematic term. But an analysis of the components of its meaning (its insistence on systematic regularities in community language practices, its relation to strategies of social organisation) can yet help illuminate how we understand language use, even in contexts where the actual phenomena appear to be diverging, due to the influence of new technologies and changing forms of social organisation, from the post-Romantic Western notion of a world made up of a series of discrete languages.


  1. Top of page
  2. Abstract

We are very grateful to Wipapan Ngampramuan for assistance with the data collection, translation and analysis. We are also grateful to Joan Swann for providing feedback on an earlier draft of this paper.

  • 1

    We have pursued the detailed analysis of the sociolinguistic features and communicative strategies that occur in this type of online encounter in [Seargeant et al. forthcoming]. The purpose of this current paper, therefore, is primarily to use the example of this type of English-influenced communication to interrogate metatheoretical issues about the categorisation and coding for social linguistic analytic purposes of English-related linguistic practices in global contexts.

  • 2

    A slight alternative to this trend of definition comes from Trudgill (2003: 139), who lays the stress on the pragmatic usefulness of the term: ‘A neutral term used to refer to any kind of language – a dialect, accent, sociolect, style or register– that a linguist happens to want to discuss as a separate entity for some particular purpose’. In defining it in this way, he manages to avoid positing the existence of varieties as natural kinds found out there in the real world, and instead focuses on the way the term is used as part of the discursive construction of knowledge in the language sciences.

  • 3

    This usage of mode is slightly different from the traditional usage is social semiotics (Kress 2000). To distinguish between the two, the word will be placed in quotes when it is being used to refer to Herring's concept.

  • 4

    A more expanded explanation of the methodology used for the analysis of the data is to be found in Seargeant et al. (forthcoming).

  • 5

    Given the conscious decision not to focus on the interactional pragmatics of the exchanges, the participant interviews act as a supplementary rather than primary source of data.

  • 6

    Ken is a Thai actor and Noi a Thai actress. They are a married couple, and Ken is eight years Noi's junior.

  • 7

    There are two distinct questions here: do varieties/named languages exist as categories for the users of those languages (and if so, to what consequence)? and do they work as categories for the analytic description of linguistic data? The former question is an issue for folk linguistics or language ideology investigations, and for examining the ways in which beliefs about language have both social and metapragmatic consequences. The use of the category of the variety for purposes of linguistic analysis, on the other hand, need not reflect what users believe about their linguistic practices, but is instead an analytic category which maps onto how people actually use language. As a predominantly metatheoretical discussion, this paper is primarily concerned with the extent to which the notion of the variety is effective as an analytic category (i.e. for use by world Englishes studies) and, because the interactional data (although not the interview data) provides no explicit evidence that metalinguistic beliefs about different languages influenced the patterns of language choice, this has not been addressed in the paper.


  1. Top of page
  2. Abstract
  • Algeo, John (1991) A mediation of the varieties of English. English Today 27, 36.
  • Anderson, Benedict (1983) Imagined Communities: Reflections on the Origin and Spread of Nationalism. London : Verso.
  • Androutsopolous, Jannis (2007) Bilingualism in the mass media and on the internet. In Monica Heller (ed) Bilingualism: A Social Approach (pp. 20730). Basingstoke : Palgrave Macmillan.
  • Androutsopolous, Jannis (2011) Localising the global on the participatory web. In Nicholas Coupland (ed.) The Handbook of Language and Globalisation (pp. 20331). Oxford : Wiley-Blackwell.
  • Androutsopolous, Jannis (forthcoming) Code-switching in computer-mediated communication. In Susan C. Herring, Deiter Stein and Tuija Viertanen (eds) Handbook of the Pragmatics of CMC. Berlin : Mouton de Gruyter.
  • Anis, Jacques (2007) Neography: unconventional spelling in French SMS text messages. In Brenda Danet and Susan C. Herring (eds) The Multilingual Internet: Language, Culture and Communication Online (pp. 87115) Oxford : Oxford University Press.
  • Auer, Peter (ed.) (1998) Code-switching in Conversation: Language, Interaction and Identity. London : Routledge.
  • Baron, Naomi (1998) Letters by phone or speech by other means: the linguistics of email. Language and Communication 18, 13370.
  • Blommaert, Jan (2008a) Artefactual ideologies and the textual production of African languages. Language & Communication 28, 291307.
  • Blommaert, Jan (2008b) Grassroots Literacy: Writing, Identity and Voice in Central Africa. London : Routledge.
  • Blommaert, Jan (2009) Language, asylum, and the national order. Current Anthropology 50, 41542.
  • Blommaert, Jan (2010) The Sociolinguistics of Globalization. Cambridge : Cambridge University Press.
  • Blommaert, Jan (2011) Emerging normativity and the politics of authenticity. Paper delivered at the MOSAIC Centre for Research on Multilingualism, University of Birmingham , 18 May.
  • Bruthiaux, Paul (2003) Squaring the circles: issues in modeling English worldwide. International Journal of Applied Linguistics 13, 15978.
  • Canagarajah, A. Suresh (2006) The place of World Englishes in composition: pluralization continued. College Composition and Communication 57, 586619.
  • Crystal, David (1995) The Cambridge Encyclopedia of the English Language. Cambridge : Cambridge University Press.
  • Crystal, David (2001) Language and the Internet. Cambridge : Cambridge University Press.
  • Danet, Brenda, and Herring, Susan (2007) Introduction: welcome to the multilingual Internet. In Brenda Danet and Susan Herring (eds), The Multilingual Internet: Language, Culture, and Communication Online (pp. 339). Oxford : Oxford University Press.
  • Deumert, Ana, and Masinyana, Sibabalwe O. (2008) Mobile language choices – the use of English and isiXhosa in text messages (SMS): evidence from a bilingual South African sample. English World-Wide 29, 11747.
  • Ferarra, Kathleen, Brunner, Hans and Whittemore, Greg (1991) Interactive written discourse as an emergent register. Written Communication 8, 833.
  • Fung, Loretta and Carter, Ronald (2007) New varieties, new creativities: ICQ and English-Cantonese e-discourse. Language and Literature 16, 34566.
  • García, Ofelia (2009) Bilingual Education in the 21st Century: A global Perspective. Chichester : Wiley-Blackwell.
  • Gardner-Chloros, Penelope (1995) Code-switching in community, regional and national repertoires: the myth of the discreteness of linguistic systems. In Lesley Milroy and Pieter Muysken (eds) One Speaker, Two Languages: Cross-disciplinary Perspectives on Code-switching (pp. 6889). Cambridge : Cambridge University Press.
  • Gardner-Chloros, Penelope (2009) Code-switching. Cambridge : Cambridge University Press.
  • Georgakopoulou, Alexandra (1997) Self-presentation and interactional alignments in e-mail discourse: the style- and code-switches of Greek messages. International Journal of Applied Linguistics 7, 14164.
  • Harris, Roy (1981) The Language Myth. London : Duckworth.
  • Heller, Monica (2007) Bilingualism: A Social Approach. Basingstoke : Palgrave Macmillan.
  • Herring, Susan (2001) Computer-mediated discourse. In Deborah Tannen, Deborah Schiffrin and Heidi Hamilton (eds), Handbook of Discourse Analysis (pp. 61234). Oxford : Blackwell.
  • Herring, Susan (2002) Computer-mediated communication on the Internet. Annual Review of Information Science and Technology 36, 10968.
  • Herring, Susan (2007) A faceted classification scheme for computer-mediated discourse. Language@Internet 1. Retrieved 25 March from
  • Internet World Stats (2011) Retrieved 25 March from
  • Kachru, Braj B. (1991) Liberation linguistics and the Quirk concern. English Today 7, 313.
  • Kachru, Braj B. (1992) Teaching world Englishes. In Braj B. Kachru (ed.), The Other Tongue: English Across Cultures, 2nd edn (pp. 35565). Urbana , IL : University of Illinois Press.
  • Kress, Gunther (2000) Multimodality. In Bill Cope and Mary Kalantzis (eds.), Multiliteracies: Literacy Learning and the Design of Social Futures (pp. 182202). London : Routledge.
  • Lavoie, Brian F. and O’Neill, Edward T. (2000) How world wide is the Web? Trends in the internationalization of Web sites. Journal of Library Administration 34, 40719.
  • Lippi-Green, Rosina (1997) English with an Accent. London : Routledge.
  • Li Wei (2011) Moment analysis and translanguaging space: discursive construction of identities by multilingual Chinese youth in Britain. Journal of Pragmatics 43, 122235.
  • Mayes, Warren Paul (2009) Unsettled post-revolutionaries in the online public sphere. Sojourn: Journal of Social Issues in Southeast Asia 24, 89121.
  • McArthur, Tom (1998) The English Languages. Cambridge : Cambridge University Press.
  • Milroy, James, and Milroy, Lesley (1999) Authority in Language: Investigating Standard English. London : Routledge.
  • Mühlhäusler, Peter (2000) Language planning and language ecology. Current Issues in Language Planning 1, 30667.
  • Myers-Scotton, Carol (1993) Duelling Languages: Grammatical Structure in Codeswitching. Oxford : Clarendon .
  • Nunberg, Geoffrey (1998) Languages in the wired world. Paper presented at La politique de la langue et la formation des nations modernes, Centre d’Etudes et Recherches Internationales de Paris, 2 October.
  • Palfreyman, David, and Al Khalil, Muhamed (2007) ‘A funky language for teenzz to use’: representing Gulf Arabic in instant messaging. In Brenda Danet and Susan Herring (eds), The Multilingual Internet: Language, Culture, and Communication Online (pp. 4363). Oxford : Oxford University Press.
  • Paolillo, John (2007) How much multilingualism? Language diversity on the Internet. In Brenda Danet and Susan Herring (eds), The Multilingual Internet: Language, Culture, and Communication Online (pp. 40830). Oxford : Oxford University Press.
  • Paolillo, John (forthcoming) ‘Conversational’ codeswitching on Usenet and Internet Relay Chat. In Susan C. Herring (ed.) Computer-Mediated Conversation. Cresskill , NJ : Hampton Press.
  • Park, Joseph Sung-Yul, and Wee, Lionel (2009) The three circles redux: a market–theoretic perspective on World Englishes. Applied Linguistics 30, 389406.
  • Pennycook, Alastair (2007) Global Englishes and Transcultural Flows. Abingdon : Routledge.
  • Pennycook, Alastair (2010) English and globalization. In Janet Maybin and Joan Swann (eds), The Routledge Companion to English language Studies (pp. 11321). Abingdon : Routledge.
  • Seargeant, Philip (2010a) The historical ontology of language, Language Sciences 32, 113.
  • Seargeant, Philip (2010b) Naming and defining in World Englishes. World Englishes 29, 99115.
  • Seargeant, Philip (2010c) Lexicography as a philosophy of language. Language Sciences 33, 110.
  • Seargeant, Philip (forthcoming) Disciplinarity and the study of world Englishes. World Englishes.
  • Seargeant, Philip, Tagg, Caroline and Ngampramuan, Wipapan (forthcoming) Linguistic hybridity in computer mediated discourse: Anglophone influence in Thai social networking interactions.
  • Searle, John R. (1985) Expression and Meaning: Studies in the Theory of Speech Acts. Cambridge : Cambridge University Press.
  • Smyth, David (2002) Thai: An Essential Grammar. London : Routledge.
  • Swann, Joan (2007) English voices. In David Graddol, Dick Leith, Joan Swann, Martin Rhys and Julia Gillen (eds), Changing English (pp. 538). Abingdon : Routledge.
  • Swann, Joan, Deumert, Ana, Lillis, Theresa, and Mesthrie, Rajend (2004) A Dictionary of Sociolinguistics. Edinburgh : Edinburgh University Press.
  • Tagg, Caroline, and Seargeant, Philip (forthcoming) Writing systems at play in Thai-English online interactions. Writing Systems Research.
  • Tagliamonte, Sali A. and Denis, Derek (2008) Linguistic ruin? Lol! Instant messaging and teen language, American Speech 83, 334.
  • Trudgill, Peter (2003) A Glossary of Sociolinguistics. Edinburgh : Edinburgh University Press.
  • Warschauer, Mark, El Said, Ghada R. and Zohry, Ayman (2007) Language choice online: globalisation and identity in Egypt. In Brenda Danet and Susan Herring (eds), The Multilingual Internet: Language, Culture, and Communication Online (pp. 30318). Oxford : Oxford University Press.
  • Yates, Simon (1996) Oral and written linguistic aspects of computer-conferencing. In Susan C. Herring (ed), Computer-mediated Communication: Linguistic, Social and Cross-cultural Perspectives (pp. 2946). Amsterdam : John Benjamins.