Polylingual language use, framing and entextualization in digital discourse: Pseudonyms and ‘Signatures’ on two Finnish online football forums

Samu Kytölä
University of Jyväskylä

Cite this article Open article PDF

Abstract

1. Introduction

This article is about web discussion forums, a multi-authored sub-format of digital discourse which, predating what we now label as ‘social media’ (Leppänen et al. 2014), emerged around the turn of the millennium. [1] Web forums (‘bulletin boards’, ‘message boards’) may involve notable manifestations of multilingualism (Androutsopoulos 2007; Hinrichs 2006; Kytölä 2013); and such is indeed the case for the largest Finnish football discussion forums on the Web, Futisforum and Futisforum2.org (henceforth FF2), which can be analytically conceived of as layered contact zones for several languages, varieties, styles and genres (Kytölä 2012a, 2012b, 2013; Kytölä & Androutsopoulos 2012). [2] This article seeks to address two interrelated layers of polylingual (Jørgensen 2008; Jørgensen et al. 2011) language use in a ‘framing position’ (Androutsopoulos 2012) identified in two major Finland-based football forums during my long observation and data collection periods. [3] First, I explore forum members’ self-appointed pseudonyms (‘nicknames’, ‘nicks’), and then I move on to analyze the affordance of accompanying one’s postings with optional ‘Signatures’ (‘sigs’) (FF2 only) that can be made conspicuous below a member’s actual forum messages (see Kytölä 2013 for an analysis of further two member profile affordances). With the help of tentative categorization and quantitative analyses of language (variety) distributions in these ‘framing’ subspaces and selective discourse-analytic, interpretative ‘excursions’ behind those polylingual items, I outline a ‘road map’ for the visibility and positions of different languages and varieties vis-à-vis the forums’ subspaces (cf. the analyses of interactive forum discussions in Kytölä 2012a, 2012b, 2013). Section 3 below includes a brief ethical consideration of such excursions into public or semi-public forum member profiles.

This paper argues that language choice and polylingual language use in the ‘bits and pieces’ of these forums’ subspaces can have an important local social meaning: manifestations of languages or their varieties – whether drawn from readily available source texts or innovatively crafted by forum members – align the members and the subsequent interactive discourse with global and local flows (Pennycook 2007) of football culture. As we shall see, resources from the Finnish language and intertextuality between ‘Finland-based’ sources are also part of that global-local constellation (Kytölä in press). I suggest that these ‘framing devices’, allegedly marginal and mostly ignored in the sociolinguistics of digital communication so far (but see Androutsopoulos 2012), can be seen as affordances for semi-public or public self-representation, performance and construction of virtual reputation, and that locally salient social positioning and (dis)identification (Leppänen et al. 2014) are achieved through them.

The epistemology behind the recent notion ‘polylingual’ (Jørgensen 2008; Jørgensen et al. 2011) seeks to eschew the problematic a priori assumption of the inherent countability of languages, suggesting in its stead the analysis of speech (or here, ways of writing) with the unit of ‘features’, not bounded ‘languages’. [4] This approach is particularly suited to the type of language use explored in this paper, which consists of short phrases, ‘bits-and-pieces’ of linguistic resources for the purposes of sociability and identification. Closely related to this, the implications of this type of digital discourse for Rampton’s (2005; orig. 1995) theory of (language) crossing are discussed in the concluding section. However, I will also use the more general and widespread term ‘multilingual(ism)’, both to refer to the source literature and, as part of the exercise at hand, also to measure the number of items or occurrences in each variety (but see my reflection and critique on these issues below).

On the basis of ethnographically accrued knowledge (Kytölä 2012a, 2012b, 2013; Kytölä & Androutsopoulos 2012), I conceptualize the two Futisforums as a loose conglomeration of intertwined ‘communities of practice’, groups of people who are in regular interaction and who share an interest or a passion for a particular set of phenomena (Lave & Wenger 1991; Wenger 1998). The concept, although first coined for the description of collective learning processes, has gained increasing applicability with the rise of online communities, whether or not their main function is collaborative learning. The polylingual language use within the loose community here can, then, be seen as one of its broad practices, within which polylingual language use as a ‘framing device’ is a prominent and distinctive subpractice explored in this article.

While it is evident, even at a short glance, that ‘Finnish-speaking Finns’ constitute the default target audience of the two Futisforums, their members have, during the forums’ life spans, developed various polylingual and polyphonic (Bakhtin 1984; orig. 1963) practices. Polylingual language use is layered in different textual and discoursal positions of the overall forum discourse, such as pseudonyms, subspaces within members profiles, such as virtual ‘locations’, discussion topic headings and sub-headings, quotations from other online media, or language alternation of different types in the actual interactive ‘one-to-many’ discussion threads (Kytölä 2012a, 2012b, 2013). These layers display a wide and multi-faceted range of multilingual practices in tandem with the conspicuously ‘Finnish-by-default’ informational and interpersonal discourse (see also Androutsopoulos 2011, 2012). This paper discusses the extent to which one can measure the visibility and distribution of different languages or varieties in the discourse positions of key forum members’ pseudonyms and ‘Signatures’. Language (variety) choice and visibility are discussed vis-à-vis issues of socio-cultural identification and performance in this online context.

I have elsewhere discussed some of the theoretical and methodological preliminaries pertaining to the study of multilingualism in web forums (Kytölä & Androutsopoulos 2012; Kytölä 2012a; Kytölä 2012b). In this strand of research, in alignment with Androutsopoulos’s (2007, 2008, 2011, 2012) insights, I suggest that understanding multilingualism in digital discourse, and in the ubiquitous web forums more specifically, requires a combination of discourse-analytical, micro-sociolinguistic, virtual-ethnographic and quantitative approaches measuring the frequencies and distributions of identifiable languages or varieties, whether predetermined etically by the researcher, or extracted in a more emic, data-driven way – or, ideally, both. Moreover, I have conducted detailed qualitative explorations of a number of discussion threads (‘topics’) from the two Futisforums in order to document and establish how linguistic diversity and language alternation takes shape in situ, and what types of metalinguistic, attitudinal and ideological indexicalities become attached to usages of different languages and their mixes (Kytölä 2012b, 2013).

In contrast to those qualitatively grounded works, this paper takes a step toward a more quantitative approach by tapping into a somewhat ignored, peripheral and marginalized subspace of multilingual online language use. Androutsopoulos (2012) points out that, both offline and online, there are various ways in which English is organized ‘on top’ of the core discourse at hand, as a “discourse strategy, a way of using semiotic resources in discourse” (Androutsopoulos 2012: 215). With the prepositional phrase ‘on top’, Androutsopoulos simultaneously evokes the connotations ‘additionally’ and ‘above or in a visually prominent position’. Androutsopoulos diverges in his epistemological stance from the ‘Anglicisms’ approach long hegemonic in linguistics – not least in Germany, the nation state he uses as an example, but also in Finland (Sajavaara 1983; Leppänen et al. 2011). Drawing from Goffman’s (1986; orig. 1974) influential concept of ‘frames’, about which more below, Androutsopoulos conceptualizes these multilingual usages as ‘framing devices’, organizing them into ‘heading’, ‘bracketing’, and ‘naming’ and proposing for them the general designation ‘(English) on top’.

For purposes of conciseness and due to the total size of the categorizable data samples, I mostly exclude the issue of ‘heading’ here and leave the discussion of it elsewhere. (However, see the screenshots from the forums in the Appendices, which serve to show that multilingual practices pertaining to discussion topic headings abound in the Finnish Futisforums). I will first explore the practice of ‘naming’ oneself in a community of practice online, such as the two Futisforums, from the point of view of Jørgensen’s (2008) polylingualism. I highlight that language choice in such a practice cannot be viewed separately from socio-cultural factors and other dimensions of (dis)identification. Then I move on to discuss another framing device, which seems salient in the format of the web forum in general – while perhaps slightly more peripheral to my Futisforum data – the use of ‘Signatures’ (often called ‘sigs’ in computer-mediated communication). This comes closest to Androutsopoulos’s ‘bracketing’, although perhaps different from the most typical communicative ‘bracketings’ due to the usual time elapse compared to the actual posting. The older Futisforum always disallowed signatures completely, while the younger FF2 allows a ‘modest’, restricted text signature to be displayed under each message posted by a member. Unlike the other affordances (Kytölä 2013), signatures are visible only to registered members and only after a deliberate ‘opt-in’ change of one’s personal settings, rendering them in this empirical case slightly more peripheral than the ubiquitous pseudonyms. However, their specific types of polylingualism are no less interesting or illustrative.

To give readers an idea how the discoursal subspaces in focus here appear in the forums’ context, Appendices 4 and 5 contain screenshot examples from both forums with polylingual language use. The appendices illustrate how the different affordances of pseudonym, ‘Location’, ‘Favourite team’ and ‘Signature’ look when embedded in the messages (but for the temporal and topical flow of the discussion threads, see the forums online). To complement the experiential phenomenology in this respect, Appendix 3 shows how they appear to other members in member profiles, whilst Appendix 2 illustrates the user’s personal interface, i.e. how they can be managed. Below are two schematic charts showing the language choices and affordances within the ‘El Clásico’ message and the ‘Hellas’ message screenshots of Appendices 4 and 5, respectively (see the models in Androutsopoulos 2007, 2011, 2012). These charts demonstrate heading, bracketing and naming ‘in action’. The former model comes from Futisforum, which uses the phpBB software, while the latter comes from FF2, which uses the Simple Machines Forum software. These software predetermine a major part of the design; see also Appendices 4 and 5 for the differences between the two forums in focus. Tables 1 and 2 are schematic diagrams that illustrate the division of a posting into parts (subspaces; see Appendices 4 and 5).

2. Framing, heading, naming and bracketing: Polylingualism in web forums’ sub-spaces

The origin of the socio-psychological concept of frame is credited to Gregory Bateson (1972: 177–193; orig. 1955), who established that frames are metacommunicative cognitive states that define the level of abstraction of the ongoing social action (e.g. degree of seriousness; whether a given action is ‘real’ or ‘play’). Goffman’s reiteration of ‘frame’has gained even more cross-disciplinary influence in scholarship. [5] For him, framesare abstractions that help actors in the social world to effectively organize their experience, make plausible interpretations of others’ social action, and further act accordingly (Goffman 1986: 1–39; orig. 1974). While much of what goes on in the social world of encounters is, for Goffman, based on tacit knowledge and the general reliability of more or less familiar frames, they can also be broken, shifted, modified and made incongruent in situ, resulting in important and everyday communicative functions that we might call ‘misinterpretation’, ‘mockery’, ‘humour’, ‘play’, ‘irony’ and such – which are themselves types of frames for humans with experience of interaction; this was a point made already by Bateson (1972 [1955]).

Bateson’s and Goffman’s insightful notions of framing have accumulated decades of influence and reference in linguistics, sociology, media studies and other fields of research (Tannen 1993; Benford & Snow 2000); different lines of inquiry and individual researchers have found their own ways to apply  them (Tannen 1993; Scheufele 1999; Benford & Snow 2000). I will leave this discussion here, and turn instead to a later application of framing that is the most relevant for my present purposes, namely, Kress & van Leeuwen’s (1996, 2001) ‘visual grammar’ and the related field of multimodal discourse analysis (Iedema 2003; Scollon & Scollon 2003; LeVine & Scollon 2004; Norris & Jones 2005; Baldry & Thibault 2006), which take issues of visuality, prominence, layout, graphics, and design seriously for our understanding of situated, multisemiotic discourses. In this field of inquiry, framing can retain its ‘Goffmanian’ function as a cognitively understood conception that helps in the organization of experience, while acquiring an additional dimension from the organization and layout of visual discourse (comparable to frames of windows, mirrors, paintings, etc.). A heading on a website, for instance, can work simultaneously as a visually motivated framing device (often positioned above a ‘core text’) and as a contextualization cue to organize experience: reading a heading can ‘tune’ the reader to the topic and register of the core text to follow. Alternatively, seeing an online pseudonym ‘X’ in the left-hand margin of an interactive discussion forum can discourse-structurally frame the message in the middle as ‘written and submitted by X’; at the same time, readers’ possible knowledge about previous events connected to that ‘X’ may tune them to read and interpret the message in one way rather than another. Along these lines, I see the notion of frame, derived from Bateson and Goffman (and Tannen 1993), as compatible with the visual grammar and multimodal discourse analysis in the way of Kress & van Leuuwen, the Scollons and others (see also Androutsopoulos 2012). In contemporary forms of digital discourse, it can be argued that ‘framing devices’ visually occur in margins, edges, tops and bottoms to frame any given core ‘texts’ on a discourse-structural level and function as Goffmanian frames, aids with which to organize experience (writing, reading, producing and consuming digital discourse). The latter, particularly with regard to polylingual language use, is closely related to Gumperz’s (1982) ‘contextualization cues’, which can be potentially given and taken up with the act of language choice or a socio-cultural reference (of which more later).

In the end, the salience of frames in this conceptualization may be far from marginal; for example, think of websites’ headings or denotations of authors in interactive digital discourses. For many readers and participants, headings give important cues as to whether to continue reading or not, or guide how to read the rest; similarly, the interpretation of interactive digital discourse is routinely affected by who is writing what, after whom, in response to whom, and so on. Indeed, in an attempt for a holistic enterprise of ‘mediated discourse analysis’ (MDA), Scollon (2001), LeVine & Scollon (2004) and Norris & Jones (2005) have legitimately highlighted the a priori theoretical and epistemological equality and potential salience of all modalities, modes, and mediating means. [6] It is with these insights in mind that I explore the relative discourse positions of the Futisforums’ pseudonyms and signatures, visually aside or below the core contributions in forum messages. In other words, they are in visually prominent subspaces, yet they can be regarded as somewhat marginalized (‘in the margin’) and peripheral (“on spatially peripheral, yet pragmatically important elements of complex textual units”; Androutsopoulos 2012: 228) in comparison to the actual messages and discussion threads (and their headings). Thus, the application of ‘frames’ in this paper extends to two overlapping planes: that of ‘surrounding’ and organizing the visual discourse structure, and that of conveying social meaning, giving contextualization cues and points of (dis)identification within the multi-faceted football (sub)cultures at play in the community of practice. With the help of their visual positioning and layout, the different framing devices analysed below filter the core interaction towards certain plausible interpretations about the interlocutors’ stances and away from others.

A major theoretical and empirical anchoring point for this paper is the research on digital discourse by Androutsopoulos (2011, 2012; personal communication), who proposes the deliberately ambiguous term ‘on top’ to cover several kinds of (mainly written and visual) multilingual language use both offline and online. Driven by empirical observations and data sampling in mostly Germany-based contexts, Androutsopoulos’s discussion is delimited to ‘English on top’; however, a similar framework can be fruitfully applied to other language varieties in similar textual and visual discourse positions. Androutsopoulos (2012: 210) argues that

‘English on top’ can be understood as a discourse strategy in which English features are used in addition to (‘on top’ of) the predominant national language, in specific textual positions and for specific discourse functions […] a process in which English is an additional code, always used next to the dominant or base code of the discourse under consideration, in our case German. At the same time ‘English on top’ is positioned in specific ways, including positions of more salience or visibility than the main code.

Similarly, by no means limited to online or even written language use, Heller (2007) has proposed an understanding of language as resources “organized in ways that make sense under specific social conditions” and circulating “in unequal ways in social networks and discursive spaces”. Within this view, it is not bounded or whole languages (ideological idealizations) but concrete linguistic resources – ‘features’ in Jørgensen’s (2008) framework of polylingualism – which have meaning and value that “are socially constructed within the constraints of social organizational processes, under specific historical conditions” (Heller 2007: 1–2). Androutsopoulos adopts this important insight in an exploration on how social actors in various media “use English resources in specific contexts […] not as a fixed variety or register, but as a discourse strategy, a way of using semiotic resources in discourse” (2012: 215). These dispositions by Heller and Androutsopoulos (and Blommaert 2010: 39–47) have guided my present approach; moreover, they are also compatible with the particular kinds of visually and discourse-organizationally motivated language use in the practices of naming and bracketing (framing) in the primary data of this study.

To serve his  line of inquiry, Androutsopoulos makes a tentative typology of this family of discourse functions into ‘heading, bracketing and naming’, arguing that while occasionally acknowledged in the literature on advertising (Martin 2007; Kelly-Holmes 2005, Piller 2001) or headlining in print media, these remain overlooked and under-researched. Yet, when taken in aggregate, they can constitute a considerable proportion of a given instance of media discourse (Androutsopoulos 2012: 216). In the case of web forum discourse, how much weight is emically given to naming and bracketing probably depends on the individual forum or community of practice. In the case of the Futisforums here, I would suggest, on the basis of my persistent observations, that they are relatively salient discourse functions with clear affordances for polylingual language use, and thus worth exploring from a sociolinguistic standpoint. As Androutsopoulos (2012: 221) points out with regard to language alternation, “the introductory and concluding chunks surrounding a contribution such as an email or forum post have been found to use a different code than the main text (Hinrichs 2006: 92–101; Androutsopoulos 2004, 2006)”. Reviewing a sample of written discourses across genres, Androutsopoulos regards ‘heading’ as the most researched function of these framing devices, with the warranted notion that different types of ‘heading’ are usually “set apart from their body text by multimodal means, including typeface, size, colour and combinations thereof” (2012: 216). Of the three framing devices, Androutsopoulos credits ‘bracketing’ mainly with the canonical work of Goffman (1986: 251–252), who distinguished conventional ‘opening and closing temporal brackets’ and ‘bounding spatial brackets’ for face-to-face spoken interaction. As suggested above, the closest equivalents to bracketing in the digital discourse format of the web forum are distinguishable openings and endings of message threads or single messages, along with the potential affordance of ‘Signatures’ (but see my caveat on the time lapse above). Finally, the third discourse function Androutsopoulos includes in his argument on the multilingualism of framing devices is ‘naming’:

Media names serve to identify an actor, organisation or a media product, and while they are of course also integrated into a stretch of text, I am primarily interested here in their function as emblems, which are set off from their surrounding discourse by virtue of their position and typography (e.g. on a newspaper masthead or the top of a periodical cover). Names are similar to headlines in terms of their textual placement and multimodal treatment, but differ from these in that they designate social and institutional identities instead of heading textual units (Androutsopoulos 2012: 224).

Delimiting his discussion to usages of English, Androutsopoulos suggests that his selection of ‘English on top’ items are frequently “intertextual, i.e. quotations from other texts; formulaic, i.e. routine expressions with a specific communicative function; and originate in specific varieties and styles of English (Androutsopoulos 2012: 227, italics original)”. To some extent these chime with the data in focus here: particularly in the case of resources drawn from Finnish and English, the intra-linguistic variation of style becomes evident in the analysis. For other ‘major languages’, the Standard variety is adhered to most of the time, but variety and style switching occur too. The formulaic aspect of the language use in these framing devices is also apparent: the forum members need to understand only short extracts, a limited number of features from ‘a language’ in order to display the competence needed in their profiles. Inferences about ‘real’ competences cannot be easily drawn, as chunks and formulae can be culled from the internet in an open-ended manner. The intertextuality feature, then, is also very prominent in the categories analysed here. However, in lieu of the term ‘intertextuality’, by now ambiguous in its rhizomatic histories of usage, I prefer to align with the less trodden research path of entextualization (Bauman & Briggs 1990; Blommaert 2005, 46–48). According to Bauman & Briggs, briefly, entextualization is

the process of rendering discourse extractable, of making a stretch of linguistic production into a unit – a text – that can be lifted out of its interactional setting […] such that the resultant text carries elements of its history of use within it. (Bauman & Briggs 1990: 73; italics in the original)

That entextualization (and resemiotization) factor is at its most salient in my discussion of ‘Signatures’, where entextualization from earlier text sources and origins is made prominent; however, the phenomenon is arguably also manifest in pseudonyms, ‘Locations’ and ‘Favourite teams’ (for the latter mentioned two, see Kytölä 2013). [7]

Scholars preoccupied with multilingual written discourse (Hinrichs 2006; Leppänen 2007; Leppänen & Peuronen 2012; Sebba 2012; chapters in Sebba et al. 2012) have attempted to apply the theoretical grids of code-switching research; I have used some of its central tenets for some of my interactional Futisforum data (Kytölä 2012b; Kytölä 2014). In this respect Androutsopoulos (2007, 2011, 2012) suggests a epistemological and methodological breakdown of types of CMC discourse: for him the code-switching toolkit is best applied only to interactionally organized CMC, while other types of multilingual CMC benefit from other theoretical and descriptive frameworks (such as visual grammar or multisemiotic discourse analysis). [8] I agree with him in that

the reason is not just the mere difference between spoken and written discourse, but rather the textual and discursive discontinuity between ‘on top’ elements and the textual backdrop to which they relate and against which they gain their pragmatic meaning. From the three discourse functions considered here, only bracketing has obvious functional similarities to classic conversational code-switching. (Androutsopoulos 2012: 230)

This article, then, is positioned as an expansion and a modification of Androutsopoulos’s initiative. On the one hand, I narrow down my analysis from Androutsopoulos’s discussion spanning several media and genres to focus solely on the format of web discussion forums (the two Futisforums with which I am relatively well conversant). Moreover, I leave the discussion of headings mostly for another occasion (but see the visual samples in Appendix 5 to illustrate topic headings ‘at work’ in connection with topic messages). My extension, on the other hand, from Androutsopoulos’s opening operates on two planes. First, I operate with larger samples to be able to suggest quantitatively determined patterns in the polylingualism of the ‘framing devices’; and second, I combine ethnographically accrued insight (see Androutsopoulos 2008; Kytölä & Androutsopoulos 2012) in the interpretation of polylingual language use in pseudonyms and signatures. Furthermore, despite the hegemony of English as a world language appropriated for a myriad of uses across the globe, and certainly the foreign language most prominent in many Finnish contexts such as these Futisforums, I have no a priori reason to limit the discussion to English. We will, however, inevitably locate the significance of English(es) as a pool of versatile resources in our findings. [9]

3. Ethical considerations

I discuss the ethical preliminaries of the sociolinguistic study of web forums elsewhere in more detail (Kytölä 2013), and hence offer just a brief synopsis here. With the quantitative part of this paper, the masses of items in each analysable category ‘tell their story’, and the ethical caveats are somewhat less restrictive than I would consider them to be in qualitatively oriented studies of interaction threads (Kytölä & Androutsopoulos 2012; Kytölä 2012b, 2013, 2014). Pseudonyms per se I regard as fully open information accessible to anyone, and they have (at least at a certain point) been chosen by their bearers as a locus of public self-representation. Technically, FF2 requires registration before one can browse the full memberlist, but registration is open and free to all, and a similar dataset of pseudonyms could be easily obtained by browsing openly available discussion topics from multiple subforums. Unlike many other web forums I familiarized myself during my research projects, ‘Signatures’ (here only on FF2) are found a little bit ‘deeper’ in the discourse, slightly more peripheral, as they are wholly invisible to non-registered web navigators, and only visible to members as an ‘opt-in’ choice in their personal settings. Slightly less than half of what I classified as ‘hard-core members’ displayed a signature, let alone more casual writers, among whom it demanded time and effort to find any signatures. What seems specific to the signatures manifest in FF2 is that they are often direct quotes from within the same forum, highlighting and recycling either a dead-on point, or alternatively a ridiculous, ‘silly’ statement that another member has written. I hesitated over whether signatures from the latter category (i.e., ridiculing other members) should be included in the analysis, but at the end, my decision was firm: these are examples of public, semi-open discourse created for display and publicity, and it is therefore legitimate to rephrase and recontextualize them in a research project that generally respects and appreciates the communities of practice in question. [10] I have been in computer-mediated contact with two moderators of FF2 and in face-to-face contact with several members of the two forums about my research, receiving largely positive feedback. In this paper, I do not probe more deeply into individual members’ profiles or message histories, although after several years of immersion in the forums, I cannot help having knowledge about some part of them. I will deploy this ethnographically accrued background knowledge whenever the interpretation of a certain multilingual usage or feature might be enhanced by it. Overall, the affordances in member profiles seem exactly the acts of identity that the members appear to be willing to highlight, perform, and share further – loci of public performance. There is no space here for conducting correlational studies or juxtapositions between or across the features analysed, although as the sampling criteria show, most members whose ‘Locations’ (Kytölä 2013) are in the analysable sample are also contained in the pseudonyms sample (Futisforum), and the same also goes for ‘Favourites’ and, to some extent, ‘Signatures’ (FF2) (Kytölä 2013).

It has to be acknowledged here that I have two nicknames (a common practice encouraged in the older forum and not strongly discouraged in FF2 either) in each of the two forums, one I registered for Futisforum around 2004–2005 as a football enthusiast without any research intention, and another one I registered for a researcher’s identity. I used my researcher pseudonym ‘Smaug’ for the data collection; moreover, I used it for accessing the non-football sub-forums hidden from non-members as well as for writing to the forum. Neither of these nicknames is anywhere near the top 200 samples used for this paper. Nor are my precious few, relatively matter-of-fact forum writings quoted even once in the signatures: that honour seems to be justifiably left for the more prolific, better-known, funnier, more entertaining members than the present author.

4. The primary data: Cluster samples and ‘purposeful’ selections

Tables 3 and 4 below show a numerical breakdown of the data collected for this analysis; the tables are followed by a description of the data sampling processes to make them more transparent for the reader. In theory, following this description, the research design could be replicable with the same web forums by any researcher to yield possibly changed samples and partly different results; however, without familiarizing oneself with these communities of practice, and more generally, the history and the present of football, it would be more difficult to integrate the ethnographically accrued insight into the forums’ experiential life-worlds with such a replicated analysis.

The criterion I decided to apply to the sample was always to include the 200 most active writers, as I regard these as the ‘hard core’ of the overall forum membership (community of practice). Although I had agreed on samples of 200 items at an early stage of designing this paper, I had several options on how to obtain the 200 items. One option would have been to conduct a completely non-skewed, random sample, e.g. to take every 200th (depending on the exact quotient of the division) member from the entire memberlist (approx. 40,000 members per forum), listed alphabetically, or by date of registration (see Appendices 1 and 2 for further clarification), or by the number of messages posted. Another option would have been to pick 200 consecutively registered nicknames, which would have yielded, for instance, all new pseudonyms registered during a given two or three months. Such a ‘cluster’ selection would not have made sense in the alphabetical listing (e.g. from ‘klaas-jan’ to ‘Koliotshik’) since confining the selection to specific initial letters would have profoundly influenced the language distribution due to the different phonotactics and orthographies of different languages. For instance, an alphabetical sorting illustrates that most ‘Locations’ beginning with the letter ‘ö’ are Swedish (Öja, Örebro, Österbotte, Övermark), and most beginning with ‘ä’ are Finnish (Ähtäri, Äänekoski), and such samples would inevitably be skewed for my purposes (in Finnish and Swedish, letters ä and ö come last in the alphabet, and not together with a and o, respectively. [11] In turn, the selection of the key ‘hard-core’ members for an exploration of quantitative distributions and visuality has an additional advantage: combining the forums’ general statistics and the memberlist statistics show that despite the overwhelming number of registered users (approx. 40,000), the 200 most active writers actually produce a major part of the overall forum discourse in each forum.

One key aim in my overall research project (see Kytölä 2012a, 2012b, 2013) has been to document the range and extent of the sociolinguistic diversity of Finnish football discourse. In order to more fully achieve this with regard to the discourse functions in focus here, I complemented the primary samples of 200 items with non-statistically motivated ‘purposeful’ selections of 30 pseudonyms (per forum) and 10 ‘Signatures’ (FF2) (and 25 Futisforum’s ‘Locations’ and 35 FF2’s ‘Favourite teams’ for Kytölä 2013). Tentative pilot samples were conducted during 2008–2012 for use in seminars and conference presentations, while the final samples were conducted in March–July 2012. Whenever I have observed or made a fieldnote that a user had edited these affordances of their profile between my earlier samplings and the final one, I have used the situation as per 6 July 2012. Of the most general statistical sampling methods, this is thus closest to cluster sampling, as I gave up the idea of conducting a random or systematic sample. [12] Each analysable category relies on the notion of ‘heavy users’ or ‘hard core’ members of the communities of practice; consequently, the ‘Signatures’ in the selection derive in part from the same FF2 members that are included in the pseudonym sample.

In regard to each category or forum subspace, I conducted a tentative division of items into ‘languages’ (cf. the critique in Jørgensen 2008; Makoni & Pennycook 2007) and, whenever this was motivated by the data, more subtle varieties (particularly ‘colloquial Finnish’ and ‘mock-English’). Each category yielded slightly different problems for straightforward classification. With pseudonyms, the nicknames chosen by the prolific Futisforums’ members, were often ‘existing proper names’ (see Appendix 1). This, of course, suggests that the target of choice for a person registering first time is, first and foremost, an existing persona (most often an individual football player) rather than a particular language, variety, nationality, or ‘culture’. For instance, choosing ‘Thierry Henry’ as one’s pseudonym is arguably more of an identification with a particular football star than ‘choosing French’, let alone ‘using French’. Similarly, choosing ‘Hidegkuti’ is a homage to the great 1950s player Nándor Hidegkuti rather than ‘using Hungarian’. But I do not consider this to be an obstacle to this enquiry; quite the contrary, it is in fact illuminating to reflect on these ‘naming’ choices and acknowledge the complexity of situated ‘minimal’ language use, languaging, as situated acts of social identification. Following Le Page & Tabouret-Keller (1985), ‘acts of identity’ are performed in every speech act by any individual, and in so doing, they give off facets of their sense of social and ethnic solidarity or difference (see also Blommaert 2005: 203–211). So, after all, naming oneself ‘Thierry Henry’ or ‘Hidegkuti’ remain acts of identification with French and Hungarian (respectively) football ‘cultures’, if only second-order. This particular kind of language-use-as-social-action, adopting a pseudonym, is a socially meaningful activity: it is done with language (albeit ‘bits-and-pieces’), and it is done with a conscious or unconscious language choice. It is different from many other subtypes of language use, but it is a fundamental prerequisite for this type of web forum communication and discourse in the first place. However, we should not necessarily presuppose a high degree of identity work in pseudonyms; this seems rather a matter for further investigation.

With ‘Signatures’ (Appendix 5), the investigation revealed that the majority of these were actually quotes of one type or another. In the Bakhtinian terms of double-voicing (Bakhtin 1984), earlier phrases, clauses or utterances or other social actors are recycled, rephrased, entextualized (Bauman & Briggs 1990) to serve a new function in a new context of situated language use. In that sense, these members’ signatures are no more language choice in its purest sense than the ‘naming’ discussed above. Yet, neither should this empirical observation be seen as a serious limitation of the present research design; on the contrary, this research finding, too, sheds light on the nature of this kind of late modern 21st-century digital communication. Indeed, in the purest Bakhtinian terms, no instance of language use is ever idiosyncratic or new at all, but we always rephrase and recycle earlier utterances in a new context (Bakhtin 1984). For web forum signatures-as-quotes, this double-voicing effect is highly explicit and overt. I am still inclined to see it as ‘language choice’, ‘language alternation’ embedded within the larger framework of the web forum interaction. For instance, choosing a signature in Standard written Russian (with Cyrillic script),

“Улыбайтесь, люди любят идиотов.”, [13]

or in Catalan,

“Menja bé, caga fort i no tinguis por a la mort.”, [14]

does not necessarily presuppose high competence in Russian or Catalan, respectively, these days as more or less anything can be readily copied from the vast mass of online sources, but it does imply that Russian/Catalan are in some respect meaningful resources for these persons, these web personae, in this position, for this delimited discourse function.

Within each of the four categories, some items ended up as linguistically ‘universal’ or non-categorizable; hence data-driven categories such as ‘universal’, ‘no identifiable variety’ or ‘acronyms and abbreviations’ were crafted to account for these items. However, the samples of 200 items in each category turned out to be large enough to guarantee satisfactory reliability and generalizability. [15] It is not possible to give an exact estimate of the error margin for the results in the categories, since there was no predetermined or closed categorization. Instead, new data-driven categories were always allowed until the last (200th) item in the sample. The selection was, moreover, biased towards the ‘heavy users’, the core of the community. With each category, I followed the entire population (approx. 43,000 in Futisforum, 35,000 in FF2) of the member profiles further, going through 100–1200 additional items, depending on the category. These rounds yielded the qualitative observation that a satisfactory saturation point had already clearly been achieved at the 200th item, as the same categories (typically: Standard Finnish, colloquial Finnish, English, universal or unidentifiable) kept recurring in roughly the same ratios, as did items distinguishable as other ‘languages’ that ended up in my ‘purposeful’ selections. The purpose of the latter was to unearth possible languages or more specific varieties that did not occur, or occurred only 1–2 times, in the ‘top 200’ samples. We should not exaggerate or over-represent the frequencies of such ‘rarer’ languages. The purpose of this part of the exercise was, rather, to acknowledge and appreciate that these also are potentially and factually relevant resources that can be drawn upon by Finnish football enthusiasts in respective acts of identity.

In connection with each set of quantitative results, I will make a few selected ‘dives’ into the related/adjacent discourses, such as brief reviews of the individual members’ topic preferences, expertise areas, or language uses I have encountered elsewhere during my immersion in the present communities of practice. However, the scope of this paper does not allow us to go any deeper into another interesting research design: that of following the message histories of individual members to identify web personae with multilingual preferences (e.g. code-switching or crossing behaviour), or whether certain pseudonyms are prolific initiators of discussions in a particular language or variety. The available technical functions of both forums (memberlisting, internal and external search engines) were deployed to some extent (see also Kytölä 2012a; Androutsopoulos 2008: 6–7). Overall, with this research design it may remain difficult to tap into the reasons, motivations or stages of consideration behind the name choices; instead, the large samples combined with the members’ very useful metapragmatic reflexivity (self-reports) about their name choices, along with my ethnographically accrued knowledge and insights into the two Futisforums, yield relatively reliable quantitative overviews of the self-naming patterns.

5. Polylingualism in the Futisforums’ framing positions

I will now turn to analyzing my samples of 200 items in each subspace. Following the chronology from the point of the social action of the forum members themselves, I will start with the pseudonyms (as adopting one is the first action that a new aspiring forum member will take). A sample of 200 nicknames from each forum and smaller ‘purposeful’ samples are tentatively classified as to their language or variety choice, along with the challenges to classification. Remarks are made about the usage of ‘already proper names’ for pseudonyms, as well as orthographic play such as punctuation or numerals. Secondly, I will move to the sample of 200 ‘Signatures’ from FF2, complemented by 10 sociolinguistically interesting picks from (less prolific) members.

All of the framing devices analysed in my study (see also Kytölä 2013) allow affordances for representation and performance potentially broader and more detailed than just the nickname; this ‘performance factor’ in the pseudonyms can be salient too, but is perhaps contestable. As specified above in the introductory sections, these ‘framing devices’ work on two planes: first, they frame the discourse structure by means of layout and visual positioning; and second, they frame the organization of experience, i.e. the interpretation and plausible readings that one connects to the performance of certain pseudonyms (see Androutsopoulos 2012). To illustrate in more detail the range of such ‘acts of identity’ (Le Page & Tabouret-Keller 1985; Blommaert 2005: 203–211) as well as the creative (almost) public performance that can be spotted in these ‘marginal’ but very potentially relevant subspaces of the Futisforums, selections are made out of the samples of 200 for a closer investigation.

5.1 Pseudonyms (‘nicks’)

Pseudonyms (‘aliases’, ‘screen names’, emically labelled ‘nicks’, also ‘nicknames’ in some CMC scholarship; see Lawson 2003, Stommel 2007), are an essential part of communication via web forums: in the communities of practice thus co-constructed, pseudonyms become “the locus of a certain representation, reputation and respect” (Kytölä 2012a: 117). Indeed, “naming is – at least potentially for many online actors if not universally – a salient act of identity” (Kytölä 2012a: 117). In that respect, the forum pseudonyms ‘do framing’ both in the discourse-structural sense and in the experience-organizational sense defined above. In her discussion of a forum for eating disorders, Stommel (2007) points out that CMC nicknames are emblematic and significant in that they represent or construct facets of identity. Moreover, Lawson (2003: 82) concludes in her discussion on ethical issues that “one’s online nickname is more than a simple identifier for the person behind the screen, it is a vital part of our ‘cyberself’”. An early quantitative study on the topic is Bechar-Israeli (1995), who formulated a fine-grained data-driven typology of 14 classes based on the denotational meaning of a large sample (n=260) of Internet Relay Chat (IRC) pseudonyms; however, in that study, language choice remains a side issue despite apparent language alternation between at least English and Hebrew (plus possibly Yiddish, although that is not explicitly mentioned).

A number of discussion topics from the two Futisforums themselves were deployed as auxiliary qualitative data for this article. Common to all of them is that they explicitly deal with the choice of the nickname and the underlying reasons or motivations for it. None of the topics were opened by myself; in fact, I was not even aware of any of them bar one (it was opened by my colleague Riikka Turtiainen) when I first began crafting this type of research design. [16] Each of these topics is framed so that the forumists openly tell the stories behind their nicknames. However, it soon becomes evident that many such ‘stories’ are to be understood as completely non-serious, jocular, not to be taken at face-value. Here is a list of these auxiliary data, the discussion threads overtly dealing with the choice of the pseudonym (all last accessed 24 June 2012). I have made rough, literal English translations of these topic headings and also of other data samples appearing in the analysis in the case of languages such as Finnish, Swedish, Polish or Chinese.

Nimimerkkienne tarina? [The story of your pseudonyms?] 179 messages between 30.12.2003 and 31.07.2005   Edistäkää tutkimusta, kiitos! [Please help this research!] 79 messages between 28.01.2004 and 31.01.2004   Nikkien merkitys käyttäytymiseesi [The influence of the nicks on your behaviour (of reading their posts?)] 42 messages between 16.03.2005 and 17.03.2005   Nimimerkkisi tausta? [The background of your pseudonym?] 79 messages between 17.03.2005 and 20.03.2005   Millä perustein valitsit nimimerkkisi? [On what grounds did you choose your pseudonym?] 461 messages between 26.02.2007 and 29.11.2010

Ironically, the last message in one of these topics (‘The influence of the nicks’) is as follows.

This was six months before I had any budding notion to do a PhD on the topic, and around four years before I saw its potential for the purpose of this particular line of enquiry. One of these topics, moreover, contains an illuminating debate between a few nicknames about the worthwhileness of the entire nickname research: one forumist argues that research about Futisforum users per se is acceptable and interesting, but that the nicknames do not have anything to do with it. Other users argue that some nicknames clearly seem to mean a lot to their bearers (‘Liverbird’, ‘Le God’…) while others hardly so (‘JJO’, ‘Tsei Tsei’…). What can be extracted from this emic insight is that pseudonyms can at least potentially be meaningful, but they can also be relatively insignificant. The discussion in one of these ‘metapragmatically reflexive’ (Blommaert & Rampton 2011: 8–10; Kytölä 2013) threads suggests that not all members even want to see the nicknames when routinely browsing the threads; this is an interesting point that I, too, have often pondered, although to verify it would require a completely new research design drawing from Human-Computer Interaction. Quite a few users argue that their choice of nickname is indeed meaningful (an act of identity) and well-considered. Let us, however, take it as a legitimate additional finding that nicknames might not always mean much to their bearers. This is another good reminder to us that meanings and functions of discourse should be investigated, not presupposed (Hymes 1996: 3–16; Blommaert & Rampton 2011). It turned out that many of the top 200 authors in each selection had actually participated in these meta-discussions about the origins of their nicknames, the majority of them in a relatively serious way. These emic insights provide illuminating information on a socioculturally interesting topic. For the present aims, quantitative in the first place, however, this is ‘enough of a story’, and we can focus on the language/variety distribution in more detail, placing less emphasis on the cognitive, experiential or motivational factors behind the choices.

5.1.1 Futisforum

The memberlist of Futisforum – online in its current format 1999–2014, but de facto largely forsaken since 2009 or so – can be freely viewed online. [17] As the number of registered nicknames totals more than 44,000, the publicly browsable memberlist is divided into approx. 1,100 subpages, each displaying the information on 40 nicknames at a time. To obtain 200 pseudonyms from Futisforum, I cluster sampled 200 ‘heavy posters’, i.e. 200 members with the greatest number of posts in total. Spam robots and doubles were manually excluded. Generally known ‘side-nicks’ of single persons were, however, not excluded, as side-nicks are an integral part of these communities of practice. The samples were conducted on 30 March 2012, although the top 200 list of Futisforum has stayed almost unchanged since around 2008–2009, when activity in Futisforum had decreased drastically due to malfunctioning and the success of the younger FF2. The updated numbers of posts by the members in the ‘top 200’ sample range between 2,800–29,400 (Futisforum) and 4,700–33,400 (FF2). Instead of selecting a totally random sample from the entire population of screen names, I opted for the most active writers, because for me they represent the core of the forum community better than passive or non-writing members (‘lurkers’ in the CMC literature; e.g. Baym 2000: 144–147). The rejection of a random sample or an alphabetical cluster sample was discussed above. A totally random alphabetical sample of 200 names would make sense, however; it should therefore divide the entire list of 40,000 names into clusters of 200 and sample every 200th name, then manually excluding spam robots and possibly other suspicious names, for instance, nicknames with no message posting history at all if so preferred. A sample of the 200 most active writers can be accrued with this procedure: From the memberlist main page, I selected “Select sort method: Total posts” and “Order: Descending”. The returned top 11 at the time of the final sampling were distorted due to spam robots and thus manually excluded; I therefore start from the 12th item and close the sample with number 212 (Mestori). (One out of the 201 was a double occurrence for some reason.) Alternative ways of viewing member listings can be seen in Appendix 1, but now we move to the tentative classification (Tables 5 and 6) of the conducted sample of 200 Futisforum nicknames by language or variety.

It can clearly be seen that Finnish, the native language to most (but not all) writers, as well as English, the foreign language most widely known to most Finns, dominate the choice of a nickname (52% in total, 58.5% if parts of mixed-variety names are included). Typical examples of Finnish pseudonyms are ‘Jääpala’ (‘ice cube’) and ‘makkonen’ (a surname); typical examples of the English ones include ‘Gunner’ and ‘The Wolf’. Names not specific to any distinguishable variety (e.g. ‘Jupiter’, ‘Ben’) and all kinds of acronyms and abbreviations (e.g. ‘1x2’, ‘asd’) comprise 23.5% of the sample. This leaves a significant 24.5% of the nicknames in other languages than English or Finnish, suggesting a range of polylingual language use far beyond conversational talk or most genres of web writing in Finland-based contexts. Indeed, it can be tentatively proposed that naming as a framing device is a practice highly prone to polylingual language use.

However, most of the time, if a name is adopted from an identifiable other language than the dominant two (Finnish, English), they are already proper names per se, most frequently names of football players (‘Xabi Alonso’, ‘Vavá’, ‘di Stefano’, ‘clichy’, ‘Shevchenko’, ‘Hidegkuti’, ‘Nakata’), or proper names denoting football clubs, such as ‘Del Atletico’, ‘Lazio’, or ‘Blaugrana’. It should be noted that in cases like these, a member probably picks up a footballer he likes, rather than choosing ‘a language’. [18] Yet we should acknowledge the heuristics where an admired, famous football player that a new football forum member allegedly identifies with or wants to be associated with is already part of a local and global football culture; therefore with a name, and an identification attached to that name, a forum member also potentially ascribes to himself a connoisseur’s identity, along with possible expertise on a more specific football culture (represented by that particular name). This is by no means to suggest that one player first and foremost should represent an independent country or a nationality: quite the contrary, many football players and idols have attachments and affiliations to multiple countries, multiple ethnicities or multiple languages. However, rather than drawing direct equation marks, I will, for the sake of a feasible quantitative analysis, tentatively identify certain names with certain nationalities and certain languages; e.g. Italian–Italy; Portuguese – Brazil or Portugal; German–Germany (or possibly Austria or Switzerland), etc..

When a name indexical of a particular nationality is chosen, it is at least potentially or partially an identification with (or towards) a particular nationality or country; for instance, picking ‘Hidegkuti’ cannot be presupposed to be a direct or explicit ‘Hungarophile’ act of identity, but it is at least a partial identification with the socio-historical era when Hungary was a great football country (i.e., the 1950s). Similarly, picking ‘Nakata’ is, if not a direct or explicit ‘Japanophile’ identification, at least a partial adherence to the socio-historical era when the best Japanese players (such as Hidekoshi Nakata) could make it in big European leagues (i.e., the late 1990s and the early 21st century). It might well be that the identification goes no deeper at all: the bearer of the name ‘Hidegkuti’ just likes Nándor Hidegkuti, or the bearer of the name ‘Nakata’ just adores Hidekoshi Nakata, while these admirations, fandoms or affiliations do not cover Hungarian or Japanese football at large. But there has been, at the point of picking the name, at least a loose affiliation, identification or adherence to the values these players represent. For this sub-study, interviewing the screen personae was excluded, but see above for the discussion of the helpful meta-threads within the forums on the choices of pseudonyms (cf. Turtiainen 2004).

Swedish is perhaps surprisingly low in the quantitative analysis, given its position in Finland as an official language and by far the most prominent minority language. [19] Some Futisforum nicknames are explicitly framed as Swedish-minded, but they were not contained in the top 200 sample due to their low number of posted messages. These include ‘kalvholmsträske’, ‘Mackan’, ‘per-erik’, and ‘Danne’. I identified markedly more writers who wrote at times in Swedish or frequently quoted Swedish sources, but these members, in turn, do not have Swedish screen names. Yet other users simply seem to have a Swedish nickname, but little or no affiliation with Sweden or Swedish-speaking football culture. These include ‘Vanda’, ‘Citronslemskivling’, ‘Lars-Peter’ and ‘Johan Näs’. In the abovementioned ‘auxiliary’ data, the forumists give explicit explanations to their choice of screen name in Futisforum. Below is an example concerning a prolific writer with a Bulgarian player’s name.

And below is another example documenting one of a Brazil enthusiast’s first messages that soon became a popular meme for years, carried over to FF2 in 2006. (Vavá was a legendary Brazilian player.)

Mixed-variety or code-switched pseudonyms form 5% of the selection, including the author with the greatest number of posts, ‘vili bin ali muhammad’ (Finnish + pseudo-Arabic?). Other examples are ‘Le God’ [20] (French + English), ‘Jere=mies’ (German + Finnish [21]) and ‘PetRescue’ (English + Romanian [22]); clearly humour and word-play seem to be functions for this intra-pseudonym code-switching. To document yet more variety in this polylingual practice of screen names, I compiled 30 more purposeful picks from Futisforum’s top 1000 writers. The entire list is not included here, but interesting additions to the top 200 list include ‘Stig Tøfting’ (Danish; note the highly iconic letter of the alphabet ø), ‘Nisi optimum’ (Latin; from the English club Everton FC’s slogan) and ‘Tukeem Abdullah Jafer’ (first and foremost Arabic; I found no direct source for this pseudonym, but the member is a noteworthy expert on Italian football).

5.1.2 Futisforum2.org

The memberlist of FF2, containing approx. 35,000 members,can be viewed by registered members only. [23] This usage of the memberlist for research purposes is ethically sound, since most of the entire forum content – along with the nicknames attached to the messages – is freely available. Screenshots to illustrate the memberlist view are found in Appendix 1. Now we move to the tentative classification of the sample of 200 FF2 nicknames by language or variety (Tables 7 and 8). This sample was taken 30 March 2012, when the forum was 6 years and 9 days old. 48 nicknames were omitted from this FF2 sample, since they were already part of the Futisforum sample of 200 names above. However, it is clear from the insights accumulated during ethnographic observation that not every prolific member has carried the same nickname across the two forums. I refrain here from digging deeper into the relationships between different pseudonyms.

Taken together, the two samples of 200 pseudonyms are sufficient to show the overall trends in self-naming. Finnish is by far the most popular resource for naming, and while it was challenging to come up with explicit criteria for the division between Standard and non-Standard/stylized Finnish, it can be seen that the names containing Finnish fall roughly equally into these two categories. Finnish names include recycled, entextualized items such as ‘Satupekka’, which was the ‘folk nickname’ for the Summer Olympics 1952 hero Emil Zátopek, the Czech runner. ‘Satu’ in Finnish ambiguously denotes both ‘a fairy tale’ and a girl’s name, while ‘Pekka’ is a male name, a very common, even an iconic one in Finnish. Hence, ‘Satupekka’ can evoke the impression of a ‘female+male’, or alternatively, ‘tale-teller Pekka’. For those of us familiar with the history of sports, it primarily evokes the notion of the Czech runner (Czechoslovakia) and the particularly warm reception he was given by Finns in the only Olympics held in Finland. [24]

The proportion of English names (22.5%) is notable and predictable considering the position of English in the sociolinguistics of Finland, and indeed in Finnish people’s mindsets, competences, and leisure interests. The range of pseudonyms drawing on any kind of English is notable in that only few of the English names are readily available players’ names or other proper names (this is in contrast to e.g. Italian, Spanish and Portuguese names). Instead, generic-looking names (‘Lisbie’, ‘Joey’, ‘timmy’) are ample, and so are common nouns (‘ref’, ‘nubster’) or non-nouns (‘outdraw’, ‘Nasty’, ‘Royal’). Moreover, mock-English, pseudo-English and stylized English are used for naming in idiosyncratic, creative ways (‘AllanMesörs’ pro ‘Allan Measures’, ‘Fenley Stastus’ pro ‘Stanley Festus’, ‘Coulijoe’ pro ‘Joe Cole’, ‘Thouni’ pro ‘Tony/Toni’). English is, I would argue, so versatile and ubiquitous, so well-known and mundane among these generations of Finns (see Leppänen et al. 2011) that its emblematic or indexical force has become debatable. Arguably some nicknames seem yet to draw heavily from rich sociocultural sources, whether football such as ‘Fin_Gooner’ (Arsenal) or ‘Newton Heath’ (Manchester), or popular music such as ‘Vinnie Stigma’ or ‘Ricky Rocket’, or the cultural domains of literature, cinema or television (‘Horselover Fat’, ‘travis bickle’). Along with Finnish, English is a language that is used in many kinds of modifications, stylizations and jokes (see ‘Fenley Stastus’ and ‘AllanMesörs’ above). Resources drawn from these two languages, Finnish and English, are also the most frequent element in mixed and hybrid names, such as ‘Tarkkatravis‘ (‘tarkka’ can mean e.g. ‘exact’, ‘pedantic’) or ‘Kanuuna-Keen’ (adopted from the Finnish translation of the football comic Buster).

The frequencies of the categories ‘universal or no identifiable language’ (‘Samson’, ‘Orion’) and ‘abbreviations and acronyms’ (e.g. ‘V’, ‘#10’) were exactly the same across the two Futisforums, 10.5% and 8% respectively. Waterproof classification of every name was impossible here, but the rough classification gives us a general picture of the phenomenology of naming: it seems a legitimate and frequently used strategy to choose abbreviations or ‘universals’ for one’s screen name.

As in Futisforum, in FF2, too,Swedish continues to score relatively low in the quantitative analysis. Of the six items, only ‘grönvit’ and to some extent ‘Hisingen’ seem choices where the Swedish football culture is foregrounded, while the other four seem ‘just names’. Some more nicknames are explicitly framed as Swedish-minded, but they were not contained in the top 200 sample. These include ‘älskar_grönvitt’, ‘Saaben’, ‘Avspark’, ‘mjanders’, and ‘En observatör’. Some other users simply seem to have as Swedish nickname, but little or no affiliation with Sweden or Swedish-speaking football culture (these include e.g. ‘Pelle_och_Kalle’). After Finnish, English and universal or non-identifiable (e.g. ‘kluru’, ‘uuba’, ‘Kuku’) names, Spanish was by far the most frequent resource for naming (13 items here). In addition to Spain, Spanish names can stem from Latin American sources, such as Argentina or even Honduras. Notably general Spanish phrases that, to my best knowledge, are not football-related were used several times, and more here than in the Futisforum sample (where screen names dating mostly from the period 1999–2005). This may reflect both the recent rise of the popularity of the Spanish language (Hall 2007; Pöyhönen 2009) and that of Spanish, and even Latin American, football in Finland. Common names originating from Spanish in this sample include ‘zulo’, ‘El_Cohete’ and ‘la bromista’, whilst Catalan is represented by ‘Caganer’, the ethno-cultural figure. Italian and Portuguese continue to be the next most common resources, despite the facts that German football also has a large following and that German is the second most studied foreign language in Finnish schools. Italian and Portuguese nicknames are still mostly ‘existing proper names’ adopted by way of entextualization to the forumists’ screen names. Exceptions are ‘ForzaInter’ (a slogan, a cheer) and ‘Incrível’ (an adjective). It should also be noted that the Portuguese ending -inho seems productive and seems to match well the morpho-phonology of Finnish names (e.g. ‘Samuldinho’). A similar hybrid effect can be created with -eldo (e.g. ‘paveldo’), probably first and foremost an Italian or Spanish suffix to most users. Yet another frequently used suffix for apparently jocular purposes is the Finnishized -naattori, adopted most likely from Latin-English words such as ‘terminator’.

Germany continues to be a major football centre and a major source of interest for Finnish Futisforumists, but German nicknames are relatively rare. Of the four items in the sample, ‘Dieter Bohlen’ seems to me the most iconically and explicitly Germanophile choice; the name belongs to a pop singer (in the duo Modern Talking), not a football player. ‘Jörg’ is arguably another one, and familiarizing oneself with the German football topics in FF2 will unearth many more outside the top 200 sample. Also the auxiliary purposeful sample I collected contains two more German nicknames. French is more or less absent from this sample, although ‘Arpin Lusène’ is adapted from Arsène Lupin, a character in French crime stories. [25] Actually the adaptation and entextualization was first done by the artist Don Rosa in Donald Duck, where the contortion ‘Arpin Lusène’ first appeared.

Russian is contained in three names (1.5%), one of which, ‘Boris Pugo’ (a Soviet minister from Latvia), can be interpreted as a Sovietophile choice. The same notable and prolific Futisforumist has used several nicknames, many of which are related to the Soviet Union or the Communism era in other countries. So in this case, a clear connection and indexical link can be made with Soviet (albeit also Latvian) history. This person knows some Russian and overtly sympathizes with, for example, Russia and Serbia, often even provocatively. He also keeps a popular football blog in Finnish, which is often cross-referred to in FF2. It is very frequent that ‘Boris Pugo’ is called by one of his other nicknames in the history of the forums.

It can be argued that many Finns’ attitudes towards Russia (and to some extent, Russians) are largely negative. This is partly due to the historical conflicts between Finland and Russia (or Soviet Union), particularly the Winter War (1939–40) and the Continuation War (1941–44), the conflicts within World War II in which Finland participated. (Another historical friction between the two nations stems from Finland’s time (1809–1917) as a Grand Duchy in the Russian Empire.) Overall, despite continuously close trading and political connections to Soviet Union and Russia, Finland has looked more actively towards Western Europe, particularly the other Nordic countries, in terms of cultural and political identity; the same can be argued to apply to Finnish people’s travelling preferences and cultural contacts at large. Further discussion of Fenno-Russian relations – or their impact on football fandoms – falls outside my scope here, but in this light, it is relatively understandable and predictable that Russophiles are not a big group in the Futisforums. However, Russian football has its active followers, and some users identify themselves as Finnish-Russian or foremost Russian (another prominent linguistic minority in Finland).

Croatian football, then, is mostly discussed and followed in the connection of well-known Croat players, or Croatia’s national team in any of the big competitions. Here the sole Croatian name that made it to the sample is ‘Viduka’, and in fact it denotes an Australian player with a Croat father. Overall, football from the Balkans or countries of former Yugoslavia is perhaps under-represented in Futisforums and certainly even more so in other Finnish media, considering that Croatia in particular has had a relatively successful recent history in football, not to mention Serbian and Bosnia-Herzegovinan star players in top clubs. Among the Greek usages, the single pseudo-/mock-Hellenic item deserves brief attention. Anyone with rudimentary contact with Greek names can see that ‘Khalervos Kendopoulos’ is a pseudo-Hellenic name, with its suffices -os and -poulos. Here, ‘Khalervos’ is derived from the Finnish male name Kalervo, which denotes Kalervo Kummola, an influential ice-hockey association leader as well as a member of Tampere’s city council. [26] Kummola is both widely disliked and ridiculed in the Futisforums due to his influential character, his arrogant public performance in general, and, relevantly here, his efforts to promote ice-hockey, which is regarded as the ultimate ‘enemy sport’ by many Finnish football fans. ‘Kendopoulos’, furthermore, contains the insiders’ derogatory term for ice-hockey, ‘kendo’ (with the implied reference to fighting with sticks). Hence, a mocking name such as this has two references to ice-hockey. I cannot interpret where the usage of Greek comes from, but for the purposes of this quantitatively inclined exercise, we can argue that Greek is a resource that one can potentially draw on in order to create a jocular effect. However, Greece’s 2004 European championship in football seems little reflected in the forumists’ naming practices, although there is continued interest in Greek football in the forums, too (see below for the discussion on signatures).

Turning to Finland’s Nordic neighbour, to and from which there is relatively much cultural and social mobility, and towards which the attitudes are generally positive, there were two Norwegian names in the sample. Of these, ‘OGS’ (Ole Gunnar Solskjær) could optionally have been classified as an abbreviation. As Solskjær was a prolific player in Manchester United (one of the top three English clubs in terms of its Finnish fan base), this name choice is probably better explained through Manchester United than any affiliation with Norway. However, ‘Grorudil’ is an almost overt reference to the club Grorud IL, also manifest in that member’s activity in discussions about Norwegian football. Norway is relatively close to Finland, and to some extent a model for Finnish football (at least in the recent past); Norwegian football is discussed somewhat actively by a small group of forumists, particularly with respect to Finnish players earning their salaries in Norway, but there seems to be no ‘Norwegophile’ culture that would be reflected in naming. [27]

The sample contains only one Chinese name, and indeed this member ‘peng.xuefeng’ identifies himself as an expert on China and Chinese. Even his ‘Favourite team’ slot has the name of the film Crouching Tiger, Hidden Dragon, spelled in traditional Chinese. [28] Some more members have a Chinese slogan or ‘favourite’ displayed on the forum profile. China is by no means one of the major global football hubs, but it is not surprising that this aspect of the current 21st-century wave of globalization also becomes entextualized in some way on a football forum. Indeed, many Finns live or work or travel in China, or study (Mandarin) Chinese, or do other things associated with China; and as FF2 is a good-sized (albeit skewed) sample of Finns, the China connection predictably becomes visible, if never prominent there.

To sample and document some more variety in the polylingual practice of screen names I compiled 30 purposeful picks from FF2’s top 1000 writers (but outside the actual sample of top 200 writers). The same categories reoccurred when going through the usernames between names listed between 250–1000: Finnish, ‘stylized Finnish’, English, ‘stylized English’, real footballers’ names, and every now and then names derived from Italian, Spanish, or Portuguese (such as the suffix -inho). Abbreviations and acronyms kept reoccurring with about the same frequency as in the top 200 cluster samples. The entire list is not included here, but new categories include Turkish (‘Iskender’, ‘yaman’), Estonian (‘Kange kaitsja’, ‘tamme auto’) and Sindarin, one of the Elven languages invented by the linguist and author J. R. R. Tolkien (pseudonym ‘Arakorn’ is stylized from ‘Aragorn’). [29] Furthermore, one stylized screen name deserves a mention here: the name of the Nigerian musician, Fela Kuti (to my best knowledge a Yoruba name), is stylized into the pseudonym ‘Fela Hutikuti’, where the latter part is a compound of two colloquial Finnish nouns, huti (‘a miss’) and kuti (‘a shot), thus resulting in a rhyme and the jocular meaning ‘missed shot’.

In the above, we have seen a nuanced picture of the sociolinguistic diversity of the practice of adopting screen names in a Finland-based online community of practice. We should not, however, presuppose any direct connection between ‘language choice’ for these nicknames and corresponding ‘cultures’ or nations of interest. Such connections should be investigated (Hymes 1996; Blommaert & Rampton 2011), not predefined. However, clear signs of such connections between languages and cultural spheres can be spotted: for example, the Finnish screen name ‘suur-saksa’ (‘Großdeutschland‘, ‘Greater Germany’) actually writes mostly about German football, and ‘Blaugrana’ (denoting the blue-and-red colours of FC Barcelona) is a prolific writer on FC Barcelona, the world-famous Catalan club, and he displays an ability to cite and interpret source texts in Catalan. Overall, even with the help of the qualitative data from the discussion topics that I used as auxiliary material and where the screen name choices were explicitly discussed, it remains a challenge to make absolute generalizations about the motivations for choosing particular nicknames from particular languages or origins. The main purpose, however, of this line of analysis was to map, document and interpret the variety and the diversity, which can arguably be seen both in the sample of the 200 heavy writers and complemented with the purposefully picked 30 additional nicknames. When I extended the analysis to the more comprehensive memberlist of 1,000 names, the same patterns continued to reoccur.

Having looked at the language choice and alternation in the practice of naming, as well as its connections to larger socio-cultural flows such as the mediated representations of football players and their apparent significance to these (mainly) Finnish football enthusiasts, I will now turn to another textual-visual-discursive practice prone to polylingual language use and cultural references, namely ‘Signatures’.

5.2 Polylingual framing in web forum discourse: ‘Signatures’ in Futisforum2

Now I turn to an analysis of polylingual language use in the modifiable affordance of ‘Signature’ (FF2) from the point of view of polylingual language use. The sample size is again 200 signatures for the quantitative analysis of frequencies; just 12 more signatures suffice here to portray the linguistic diversity of that framing (bracketing) device (see Kytölä 2013 for Futisforum’s ‘Locations’ and FF2’s ‘Favourite teams’).

Contrary to Futisforum during my most active observation and data collection period (2005–2008), signatures are allowed and visible in FF2; even on this newer forum they are exclusively visible for registered users, and only if set as an ‘opt-in’ choice in one’s personal preferences. Signatures can thus be taken as peripheral, not a central ingredient in the overall discourse, as they are not visible to all users; they seem not to convey ‘obligatory’ contextualization cues. Indeed, one metapragmatically reflexive sequence of discussion from 2007 suggests that there has been overt resistance in the older Futisforum to signatures and avatars (avatars are small pictures in one’s member profile, another CMC framing device not found in either Futisforum).

After this message, the more experienced members take a laconic stance against such ‘follies’ as avatars and signatures, suggesting that they disrupt the flow of forum discussions. One experienced member does this implicitly by posting an ironic short one-liner message, accompanied with what could be identified as a ‘manually crafted avatar’ (nothing less than an iconic picture of the Finnish football legend Jari Litmanen), and a manually crafted signature (the complete lyrics of John Lennon’s Imagine, one of the most quoted pop songs in history). That posting clearly takes an ironic stance against avatars and signatures, ridiculing the ‘worst kind of clichés’ that can appear in those subspaces of the web forum. Meanwhile (May 2007), signatures were, however, already being supported by the now thriving FF2. Avatars have not been in use in either of these two major Finnish football forums, but they are allowed and eagerly deployed in Paitsio.com, the third largest forum of the time of my research.

The maximum length of the ‘Signature’ affordance in FF2 is 300 characters, which can apparently be divided into up to as many lines as one wishes. From an ethical point of view, and from my ethnographically accrued insight into the forum’s practices, I regard them as public discourse that can safely be further entextualized here for new metapragmatic (research) purposes. I do not connect signatures to any detailed life-stories of the respective pseudonyms here, but I may comment on the profile of the user in cases where this is relevant to the topic of multilingualism (e.g. if a signature entirely in German was written by a Germanophile forumist).

Smaller tentative samples have been obtained for pilot and seminar purposes over the years, but this final sample of 200 signatures was obtained by going through the memberlist up to the 655th most active user; around 30% of the core members had opted to display a signature (as per 29 June 2012). The 655 ‘heavy posters’ from the total of approx. 36,000 members comprise around 2% of the entire membership. The occurrence of signatures became notably less frequent as I went down the memberlist in the order of ‘total number of posts’. This purely qualitative observation might suggest that it is the hard-core members, the most prolific writers, who are the most likely to have a signature in the first place. A more rigorous sample would be needed to actually verify this tentative observation.

Table 9 shows a summary of FF2 ‘sigs’ from the point of view of language choice and distribution.

Again, it was not self-evident how to draw the division between ‘Finnish’ and ‘colloquial Finnish’, but I have relied on four principles as well as my (native speaker’s) intuition. [30] My final divisions are still best taken as tentative. Moreover, there was a need for the idiosyncratic ‘Forumese’ category for items that were often Finnish, often colloquial Finnish, but clearly contained ‘insider’ words or phrases, and thus fully understandable by those familiar with the two Futisforums. When a signature contained a hyperlink, I included it in one of the language categories, as far as this could readily be read from the URL address. When this could not be detected from the words contained in the URL (e.g. YouTube links), I classified them as ‘hyperlinks/URL only’, without taking into account the contents of the hyperlink (although I did follow them all at least once). When one ‘Signature’ clearly contained two or more varieties, I gave each variety a decimal fraction for the quantitative comparison (ranging from 0.2 to 0.5). Thus, even a long, multi-part item counts here as 1.0 in order to retain the advantages brought by the sample size of 200.

Of this sample of 200 Signatures, around 170 (depending on the criteria) were comprised of quotes from earlier sources. This is a prime example of entextualization (Bauman & Briggs 1990; Blommaert 2005: 46–48) and double-voicing (Bakhtin 1984): earlier phrases, clauses or utterances from other social actors are recycled and rephrased, and the texts acquire new social meanings in new contexts of use. Most of the earlier sources were found within the same forum, FF2; I assume that this deviates from the practice in many other forums. Many of these were, in fact, overtly indicated so that a hyperlink leads from the signature to the earlier source within the forum. There seem to be two very common broad social meanings for quoting an earlier excerpt within FF2: either a point made by someone else is so apt and on-target, or alternatively, so ridiculous or exuberant, that it deserves to be highlighted and recycled in a ‘frozen’ Signature.

Moving on to the sociolinguistic diversity of this ‘Signature’ affordance, by far the most common identifiable language variety is Standard Finnish (47%). For instance:

Colloquial Finnish is the second most common variety (26%). An illustrative example is this:

The items I classified here as ‘Forumese’ (5%) contained one or more features that seem to me not directly comprehensible for a non-Futisforumist native speaker of Finnish. These included the ironic use of the common noun homo ‘homosexual, gay’ in a manner I interpret as specific to the Futisforums’ style, for instance:

‘Forumese’ could in principle include other canonical languages than Finnish but in this sample they were identifiable as Forumese and Finnish. Moreover, the item I classified in the ‘onomatopoeia’ category (‘MYÖÄÄÄÄÄÄÄÄÄRRRRGHHH!’) is probably recognizable as Finnish. Thus, signatures identifiable as any variety based on Finnish totalled 79%.

Of languages other than Finnish, once again and predictably, English (16%) was by far the most usual choice for the Signature. These could be divided into more fine-grained categories such as Standard English, non-Standard colloquial English, and mock-English (deliberately ‘bad’ English or ‘Finnishized’ English). Examples of Standard English are here (italics as in the original):

Examples of Signatures in non-Standard English and perhaps mock-English are here:

2.5% of the sample were not identifiable as any language: they were either hyperlinks, numerals (‘30’), or universal abbreviations (‘R.I.P.’, an idiom which interestingly seems to be usable in several framing devices). The rest (here only 3%) were identifiable as other major football languages. German comprised 0.8%, 1 entire signature and half of another one:

Swedish, Spanish and Catalan had 0.5% (1 signature each) as follows:

One signature, by a prolific Finnish-Turkish writer, was divided between Italian and Turkish and a hyperlink (emphasis original):

On the basis of this sample and distributions discussed above, it seems that Finnish and English are by far the most favoured varieties for FF2’s signature function, while the other major football languages occur here, too, but are not particularly large in numbers. When they occur, such usages are most often retraceable to a quote from elsewhere – my application of double-voicing from Bakhtin (1984), and entextualization, Bauman & Briggs (1990) – and to the other counterparts in the user’s member profile: Favourite team, message history, public interests, and also at times the pseudonym. Thus, a prolific Germanophile or Italophile, for instance, may well display that cultural preference in the form of a language choice in all possible discourse spaces, or in many of them. This turned out to be true also with the abovementioned members who displayed profile items in Catalan and Polish.

I made 10 purposeful picks from the memberlist’s top 200–1600 (June 2012) to document further linguistic diversity in this practice (Table 10 below). The number is lower than with the purposeful picks in the other categories, and I had to go through as many as 1600 members (approx. 4% out of approx. 36,000 members in total) because signatures became increasingly less frequent when exploring the memberlist in the descending order of ‘total number of posts’.

The languages that emerged as resources for ‘Signatures’ in this purposeful sample are Greek (see also Appendix 5), Serbian (in Latin script), Latin and Icelandic (the latter two even in the same signature). It is interesting to note that as many as 170 (85%) signatures fulfilled the criteria of quotes. These were overt or covert, signalled and framed in various different ways: through the forum software’s quoting function, with the help of double quotation marks, with a hyphen and the originator of the quote, or with italics. A very prominent subcategory of quotes here, perhaps even idiosyncratic to FF2, seems to be quoting another message from the same forum, whilst at the same time indicating and embedding the exact link to where the original occurred. This is a practice that I actually saw emerge during my observation periods; it has also elicited overt meta-commentary in multiple places around the forum. This raises the question: what kind of ‘language choice’ is it when one quotes another source? We can suggest that such recycling and echoing earlier sayings requires understanding the original, if not the communicative competence of creating similar sentences oneself. All this makes an intriguing case of entextualization, the process of “making a stretch of linguistic production into a unit – a text – that can be lifted out of its interactional setting […] such that the resultant text carries elements of its history of use within it” (Bauman & Briggs 1990: 73). An alternative reading of this phenomenon can be done through the lens of double-voicing (Bakhtin 1984): existing and heard-before phrases and utterances of other social actors are constantly recycled and reformulated to the extent that no language use is ever completely innovative or idiosyncratic at all.

6. Summary of quantitative findings

This section summarizes and cross-compares the quantitative analyses of the four subsets of different ‘framing devices’. Finnish is clearly the first language for most Futisforumists and the two Futisforums are Finnish ‘by default’ in several ways. Even more importantly, and more verifiably here, judging from the impressive message posting histories, Finnish is the language used daily by the top 200 (or top 600) writers that comprised my samples here. In addition to the general flow of the multi-authored discussion threads, this is reflected in the dominant choice of Finnish for all the framing devices discussed here: 28% in Futisforum’s nicknames, 33% in FF2’s nicknames, and 74% in FF2’s ‘Signatures’ (and 72% in Futisforum’s ‘Locations’, 57% in FF2’s ‘Favourite teams’; Kytölä 2013). Tentatively, it would seem that, of the framing devices discussed here, pseudonyms, where the percentage of Finnish items was lowest, showed the strongest tendency to occur in a non-native language. This would suggest, in everyday reasoning, that it is desirable, ‘cool’, to be named something ‘foreign’, perhaps especially in a transcultural football context. The emically grounded category ‘Forumese’ – mostly derived from resources of Finnish – occurred in each category except that of ‘primary’ pseudonyms: 5% of ‘Signatures’ (and 3% of ‘Locations, and 1% of ‘Favourite teams’) could be regarded as ‘Forumese’. Moreover, many pseudonyms outside the two top 200 samples utilized were crafted from ‘Forumese’.

In accord with the strong position of English in Finland in several domains (see Leppänen et al. 2011, Leppänen & Nikula 2007), the collective competence of (most) Finns in English, and the strong foothold of British (especially English) football in Finland since the 1950s or so (Heinonen 2005: 58–67), English is by far the most frequently used ‘foreign’ variety in these categories: approx. 25% in Futisforum’s nicknames, 25% in FF2’s nicknames, and 16% in FF2’s ‘Signatures are classifiable as English. ‘Locations’ was the only one of these framing positions where English is somewhat less prominent; this can be partly explained by the tendency to enter real-life physical locations in this profile subspace.

Other than the expectedly hegemonic ‘winner’ Finnish and the ‘runner-up’ English, the biggest European languages were conspicuously present in all samples. After England, it is Spanish, Italian and German football that currently attract most followers in Finland, but judging from the social action in these Futisforums, countries such as Sweden, Norway, Russia, the Netherlands, France, Greece, Turkey, Portugal, Brazil, Argentina and Mexico are also followed with passion. These socio-cultural interests in the international, globalized ‘football culture’, coupled with Finns’ level of knowledge of foreign languages help to explain the regular occurrence of foreign languages in this visual-textual football discourse. [46] Although the four categories (five quantitative exercises; Kytölä 2013) are not directly cross-comparable or combinable, we can see that after English, Spanish was the most frequently used ‘foreign’ language resource: 6% of the nicknames and 0.5% of ‘Signatures’ are identifiable as Spanish. Also among the resources that occurred multiple times in the samples is Italian (a language connected with the football superpower Italy): 3.5% of the nicknames and single occurrences of the other categories were Italian. My ‘purposeful’ searches found plenty more, some of which are documented above; the rest remain in my archives, and also, it may be hoped, online for a long time to come. German is still a relatively prominent foreign language in Finnish institutions of education (Hall 2007; Pöyhönen 2009) and Germany is one of the major football cultural sources passionately discussed in the two Futisforums. Yet German is not shown very prominently as a framing device in the analysed sample: only a handful of nicknames (< 2%) merit a mention here. Portuguese, another world language of the joga bonito, mainly shows here in nicknames, mostly derived from Brazilian or Portuguese players. Despite the geographical proximity, the Russian language is relatively weakly known among Finns (see above for possible historically motivated explanations); the visibility of Russian among these framing devices is not zero, but consistently around 1%. It is possible that the Cyrillic script – not readily available on Finnish keyboards – weakens the appeal of Russian as a usable language in these functions; then again, even the 1% sticks out visually and can hardly be missed by a persistent reader in the long run. A few single occurrences were found for other languages such as Catalan (apparently mainly due to FC Barcelona), the major world language French, Greek, Latin, Serbian, Mandarin Chinese (in traditional Chinese script), Japanese, Arabic (in Latin transliteration only; but for a case of Arabic on Futisforum see Kytölä 2014), Turkish, Estonian, and the geographically close Norwegian, which is to some extent understandable to those Finns who know Swedish, and would otherwise perhaps be a contender as a potential framing device.

These were the extent and limits of the diversity in the researched phenomenon as classified by languages and varieties. Perhaps surprisingly, Dutch and French are almost non-existent in these samples, apart from a few players’ and clubs’ names, although as the Netherlands and France are major football countries globally, there are active discussions about them (especially Finnish players in the Netherlands) where Dutch or French are used to some extent – mostly as quotes or as framing devices. Finally, items not identifiable as ‘language’ in any traditional sense were found in each category. This is a relevant supplementary finding and is well in line with earlier scholarship on stylization and creative language use in digital communication (e.g. Androutsopoulos 2006, 2007, 2011; Leppänen et al. 2014).

I conclude the present analysis with an example of observer bias related to the research design and method. In the pseudonyms and the other framing devices discussed above, there are numerous cultural references to popular culture: television series, cinema, novels, rock/pop groups and artists, their albums and songs. I would admit that I tended most easily to notice those that were familiar to me prior to the analysis of the samples. Some of the cultural allusions actually had to be found through a web search engine, while others caught my eye immediately due to their familiarity. [47] The latter group certainly includes the references to the Finnish progressive rock group Wigwam, whose compositions had served as inspiration for two prolific pseudonyms, ‘Cafffkaff’ and ‘Planetist’, both part of my smaller purposeful sample of 30 pseudonyms. This relatively obscure Finnish group, who had their heyday in the 1970s, is definitely not the most prominent or obvious popular-cultural origin to draw from for today’s football enthusiasts. It just appeared prominent, appealing, and worth including here from my particular life experience, from my admittedly biased position. That bias acknowledged, I hope to have done justice to the topic of this inquiry with the large samples and many hours of additional research from the standpoints both of quantitative distribution and diversity and variety.

7. Discussion

In this paper, I adopted a visual-functionally motivated sociolinguistic analysis of framing by combining the socio-psychological and communicative notion of frame (Goffman 1986; Tannen 1993) with the conceptualization of ‘framing device’ from within multimodal discourse analysis (Kress & van Leeuwen 1996, 2001). On the one hand, we have Goffman’s (1986) argument – developed in the first instance for face-to-face encounters and activities – that social actors establish interpretation frames in order to organize their experiences during communicative activities. On the other, we have the theoretical tenets of visual grammar – according to which layout, design and visual positioning matters – which involve looking, for instance, at the ‘prominence of borders’ (Kress & van Leeuwen 1996, 2001; Scollon & Scollon 2003) as framing devices in the discourse structuring of visual discourses. Combining these two, we can argue, along with Androutsopoulos (2011, 2012), for the relevance of an applied frame analysis for the purposes of digital discourse and its participants (readers and writers; ‘producers’ and ‘consumers’). [48] Framing can function in widely varying ways in different subformats of CMC; however, given the importance of visuality in the currently popular digital discourse formats, issues of layout and design are now more prominent than they were in the earlier (1980s and 1990s), more fundamentally text-based forms, owing to their greater potential and affordances for new types of framing-devices-as-contextualization-cues. Moreover, I added the notion of entextualization (Bauman & Briggs 1990) to the agenda in order to model the movement of discourse from one context to another; with the ‘end-texts’ (the momentary snapshots of the pseudonyms, signatures, etc.) carrying elements of their histories of use within them, and so potentially signalling points of affiliation or (dis)identification.

I have restricted my analysis to the format of web discussion forums, where the relevant framing devices include pseudonyms (nicknames, screen names), ‘Signatures’, and (Kytölä 2013) two more tailorable slots for member profile information: registered member’s self-appointed ‘Location’ in Futisforum and ‘Favourite team’ in FF2. Excluded from both of the published analyses were, for example, headings of forums, subforums and topics/threads, technical meta-data (such as time of posting, the number of replies/views, dates of registration, number of messages in personal message histories), avatars, as well as communicative ‘bracketing’ within messages (such as greetings, cheerings, slogans, mottos, openings or endings).

All of the positions and subspaces in the analyses were first observed qualitatively and by means of virtual ethnography (Androutsopoulos 2008; Kytölä & Androutsopoulos 2012). As these were found to contain both patterned and idiosyncratic occurrences of juxtaposed polylingual language use, I subsequently analysed them in cluster samples of 200 items from each category to produce an overview of the polylingualism of framing devices. As Androutsopoulos points out (2012: 231–232), these function as indexical of authors’ identities, their interrelations and their imagined audiences and addressees, although we should be cautious about drawing overly straightforward, simplified links between them. The notion of indexicality (e.g. Androutsopoulos 2004, 2012; Blommaert 2005, 2010) seems central to understanding the type of polylingual language use I have documented here. As we saw, selection is often primarily targeted at other, more salient points of identification than ‘language’. Yet we can tentatively propose that selecting a language serves as a pointer to “the linguistic practices of certain groups or places that are deemed important for the interpretation of the on-going discourse” (Androutsopoulos 2012: 232). The power of language choice or usage, then, “to cue frames of interpretation is therefore not just an outcome of conventional associations” (Androutsopoulos 2012) of those languages, but “results from a link to specific social contexts and practices established” (Androutsopoulos 2012) by particular styles derived from them. In this vein, by linking the polylingual language use in the raw data in the analysis to socio-cultural movements and trajectories within the global domain of football, I have attempted, in the hope of arriving at higher explanatory adequacy, to respond to Androutsopoulos’s call to reconstruct “the social worlds, discourses and texts indexed by the specific bits” (Androutsopoulos 2012) of the various languages occurring in the discourse rather than ascribing undifferentiated symbolic values for any of them. (Remember how an unmistakably Swedish pseudonym did not guarantee any affiliation with Sweden or the Swedish-language football cultures.)

While Androutsopoulos’s discussion was solely based on usages of varieties of English(es) as a framing device, we saw here that there is no reason why other languages than English – whether ‘world languages’ or not – should not function as similar pools or sets of resources for framing digital discourse. This type of emblematic, heavily indexical ‘bits-and-pieces’ (see Blommaert 2005, 2010) usage of ‘foreign languages’ need not presuppose a high competence behind them. Nor should we, of course, automatically presuppose that English is the first and best foreign/second language for all participants in a given context, such as the Finnish fans depicted here. This notion of the limited use of ‘bits-and-pieces’ of language brings us to Rampton’s (2005) influential notion of crossing, which was originally situated (at least partially) within the code-switching research strand (see Gardner-Chloros 2009). Crossing has several elements that seem to be of high relevance and applicability here as well. Certainly, the polylingual language use as a framing device and contextualization cues here is “code-alternation by people who aren’t accepted members of the group associated with the second language they employ” (Rampton 2005: 270), involving varieties that are not generally associated with their users, in turn raising “issues of social legitimacy that participants need to negotiate” (Rampton 2005: 270–271). Moreover, there are clear signs here that such polylingual language usages contain a “disjunction between speaker and code that can’t be readily accommodated as a normal part of ordinary social reality” (Rampton 2005: 272–273). While Rampton’s theory was based exclusively on face-to-face spoken interaction data, CMC sites and spaces of digital discourse seem to offer fertile ground for renewed crossing research. Indeed, Blommaert and Rampton (2011) already suggest similar paths. As Androutsopoulos (2007, 2011, 2012) argues, this type of polylingual language use does not necessarily sit easily in the tradition of research on code-switching, the occurrence of two or linguistic varieties within the same utterance or stretch of interaction (e.g. Gumperz 1982, Hinrichs 2006, Androutsopoulos 2007, Gardner-Chloros 2009). Indeed, as Sebba (2012) suggests, language alternation in the written mode, visual by default, requires a differently grounded approach from the study of spoken/auditory code-switching, ideally drawing on visual grammar, genre analysis and literacy research (see also the chapters in Sebba et al. 2012). Code-switching approaches need not be rejected completely, however.

In sum, with the help of the analysed framing devices as affordances for (almost) public self-representation, performance and reputation-building, it becomes evident that locally salient social positioning, (dis)affiliation and (dis)identification take place in the data; but importantly, nowhere near a one-to-one relationship should be directly inferred from the sociolinguistic choices made for the framing devices. Perhaps this type of heavily entextualized and multi-voiced discourse is better seen as symptomatic of late modern multisemiotic literacy or ‘semiotic agility’ (Thorne & Fischer 2012; Prior 2010) than the more two-dimensional, traditional notion of ‘competence’. This interpretation finds a historical echo and depth in Goffman’s ‘multiple footings’ (Goffman 1981; Prior 2010) and Hymes’s (1996: 25–62) notion of ‘situated communicative competence’; indeed, the ‘bits and pieces’ of language documented in this paper, whether directly entextualized or multi-voiced from a ‘more original’ source or creatively crafted from earlier manifestations of language, make perfect sense just the way they are – at least for the initiated members of the community.From a sociolinguist’s or applied linguist’s viewpoint there is, of course, no reason to underrate these digitally mediated and socially meaningful language uses as ‘truncated’ (Blommaert 2010: 103–106), or in some way minimal or inferior to ‘more holistic’ manifestations of linguistic competence.

When putting these phenomena in perspective, an important point to remember is that our fascination with new modes and formats of communication and discourse notwithstanding, there is perhaps not much inherently new here. True, web forum pseudonyms or the other categories analysed here could not, by definition, exist exactly as such prior to the invention and emergence of web forums, or indeed prior to the spread of the internet to people’s daily use. It is perhaps challenging to think about equivalents or predecessors to our forum profile categories in pre-internet written discourses or in spoken/face-to-face discourse. Clothing is certainly one multisemiotic contender here: consider the various texts that can be found on caps or t-shirts, or cultural symbols that can be found in pins, buttons or patches. Or, to keep our discussion within the domain of football, we can think about flags, emblems or tifos created, crafted and carried around by fans in embodied ways throughout the history of football supporting.

But finally, let us briefly consider our first category, pseudonyms, from the point of view of the pre-internet era. It is evident that people have always had aliases for different purposes, in various spoken and written (and signed) contexts. The needs and motivations for this may have varied from artistic and performative (e.g. having an aesthetically satisfying artistic name that ‘sticks’) to intertextual and emblematic (e.g. desire to identify with a point of identification in history and stand for it), and from affective (e.g. expression of social or familial belonging) to political or practical (e.g. deception, hiding from other people or institutions). All such functions arguably abound in the social history of pseudonyms, although there is no space for a deeper investigation here. To return finally to the domain of football, a brief personal anecdote from my childhood offers an illuminating parallel. Around 1984–1986, my best friend and I were ardent football enthusiasts and each of us had our favourite national teams, clubs and players – several of each to be sure. And we played in the yards, parking lots, grounds and fields around our hometown; we imagined ourselves football stars (seen on television), or sometimes Finnish or even local players. Yet each of us had only one ‘true’ alias among our favourite players. My friend supported the Soviet Union in every sport; hence he was ‘Rinat Dasajev’, the superb goalkeeper of the Soviet national team and of FC Spartak Moscow. [49] I admired the European champions France, and the ultimate star Michel Platini above all, but I felt that being ‘Michel Platini’ would have been ‘too obvious’ and too arrogant, so I was ‘Luis Fernández’, a less outstanding but equally solid and crucial player in the legendary French midfield formation, le carré magique (note that his name is actually Spanish). Since then, I have of course admired many football players of many generations, but I doubt if I have ever again truly been one, in the sense of a personal alias, a deeper point of identification. [50]

Notes

[1] An extended earlier version of this paper is included in my doctoral dissertation (Kytölä 2013). In addition to pseudonyms and signatures, the earlier version includes an analysis of two further subspaces afforded by the personal member profiles: ‘Locations’ (Futisforum) and ‘Favourite teams’ (FF2). I am thankful to Sirpa Leppänen, Jukka Tyrkkö, Joe McVeigh and the anonymous referee for their comments and support, and to Michael Freeman for language checking. All the remaining shortcomings are solely mine. [Go back up]

[2] Currently located at http://suomifutisnet.adv1.nebula.fi/phpBB2/ and http://futisforum2.org/, respectively. The edited football website where Futisforum was formerly embedded announced its closure in March 2011, but the actual forum remains still online (as per 23 November 2014). Two other major Finnish football sites that contain multi-authored, pseudonym-based forums are Paitsio.com and Futismaailma.com; this analysis excludes them for reasons of space. [Go back up]

[3] Despite ambiguity, I have chosen to use the British and European word ‘football’ consistently throughout my current work to refer to association football, i.e. the sport most North Americans, among others, unambiguously call ‘soccer’. [Go back up]

[4] It has to be noted that the meaning of the Greek-based prefix ‘poly-‘ does not, at the end of the day, differ much from the Latin-based suffix ‘multi-‘, both of which denote ‘many’. However, it is the epistemology, critical rethinking and the general line of inquiry around Jørgensen et al.’s work that I intend to align with here; hence my preference for ‘poly-‘. It could be argued that ‘translingual(ism)’ (e.g. Pennycook 2007: 36–57) might capture the essence of these phenomena equally well. The current (ca. 2006–2014) terminology debate among scholars over ‘multi-‘, ‘poly-‘ ‘trans-‘ or ‘metro-‘ -lingualism remains, however, outside my focus here (but see Kytölä 2013). [Go back up]

[5] Goffman himself (1986: 7) acknowledges the influence of several other scholars, particularly Gregory Bateson, for the concepts ‘frame’ and ‘bracketing’, and credits himself only with their application to situational, face-to-face conversational discourse. [Go back up]

[6] See also Blommaert (2005: 1–4) for a further justification of this approach. Coupland (2009) is highly recommended as an ambitious and intriguing attempt at an integrated analysis of sociolinguistic style, performance, frames and multimodality in mass media contexts; however, digital discourse remains a side issue in his article. [Go back up]

[7] Entextualization is very closely related to the notion of resemiotization, the process of “how meaning making shifts from context to context, from practice to practice, or from one stage of a practice to the next” (Iedema 2003: 41). For a discussion on their differences and similarities as well as how they can be integrated, see Leppänen et al. (2014). The notion ‘resemiotization’, although fully potential here, would hold even more appeal when the multisemiotic aspects of the researched discourses are more prominent than in the mainly text-based analysis at hand. [Go back up]

[8] Many non-CMC linguistics scholars in the formative years of CMC have sought parallels between internet writing and spoken conversation (see, for instance, Montgomery 2008: 135–140 and the revised and improved discussion in Crystal 2006: 26–65), although both Androutsopoulos and I have major reservations against such simplified comparisons across modes that mostly ignore the dimensions of genre and register. Moreover, research on internet writing will benefit from a fuller adoption of a multisemiotic analytical framework (Kress & van Leeuwen 1996, 2001; LeVine & Scollon 2004; Norris & Jones 2005; Leppänen et al. 2014). [Go back up]

[9] In fact, my entire research project had ‘uses and functions of English’ in (Finland-based) football discourse as a point of departure, while my later move to include a larger constellation of multilingualism in the project was motivated both empirically (especially the two Futisforums) and theoretically (e.g. Heller 2007; Jørgensen 2008, Blommaert et al. 2005; Makoni & Pennycook 2007). [Go back up]

[10] I do not, however, refrain completely from critique or research into the more malevolent sides of the forums’ trajectories and discourses. A critical investigation into two cases of bullying and discrimination, for instance, is contained in Kytölä (2012b). [Go back up]

[11] Available at http://suomifutisnet.adv1.nebula.fi/phpBB2/memberlist.php?mode=location&order=DESC&start=0. [Go back up]

[12] The older Futisforum was nearly derelict by 2012 apart from a handful of active discussion topics; the situation in 2012, therefore, closely resembles that of ca. 2008–2009. [Go back up]

[13] ‘Smile, people love idiots.’ [Go back up]

[14] ‘Eat well, shit a lot and do not be afraid of death.’ [Go back up]

[15] I am grateful to statistician, PhD Kari Nissinen (University of Jyväskylä) for instances of personal communication and good advice in the sampling options. The shortcomings in the choices related to sampling, however, remain my own. [Go back up]

[16] I also benefited the very helpful questionnaire survey by Turtiainen (2004). This was an insightful early work on her path to a professional researcher. [Go back up]

[17] http://suomifutisnet.adv1.nebula.fi/phpBB2/memberlist.php. [Go back up]

[18] I use the masculine pronoun throughout the paper in order to eschew clumsy structures such as ‘her-/himself’ or ‘(s)he’. This is by no means to defend or recommend the sexist use of masculine pronoun in a generic sense. Instead, my ethnographically informed perspective on the Futisforum community strongly suggests that the majority of the 200 or 1,000 most active authors are male, or identify their screen names as male. I duly acknowledge possible female authors, and I will do my best to refer to known or alleged female authors with feminine personal pronouns. [Go back up]

[19] The proportion of Swedish speakers in the population of Finland was 5.4% (approx. 290,000 people) in 2008–2012 (Statistics Finland 2014). However, the official figures pertaining to ‘mother tongue’ do not allow for bilingual choices (which are a norm for many Swedish and Finnish speakers). [Go back up]

[20] The nickname of the great English player Matthew Le Tissier is ‘Le God’, allegedly derived from the surname and given to him by Southampton FC fans. The usage of the French definite article ‘Le’ is thus neither this Futisforumist’s linguistic strategy nor code-mixing by the fans; its origins can be traced to a real-life surname. Actually Le Tissier was born on Guernsey, which might partly explain the French surname. [Go back up]

[21] Jens Jeremies is a former German footballer, while Jere is a Finnish male name, and ‘mies’ is Finnish for ‘man’. [Go back up]

[22] Pet Rescue is the name of at least one TV series and one computer game, while Petrescu is a common Romanian family name. There are footballers named Petrescu both in Finland and in Romania. [Go back up]

[23] Most of the forum can be viewed at http://futisforum2.org/; see Appendix 1. [Go back up]

[24] Idiosyncratic names such as this can nowadays just be entered in a search engine such as the currently popular Google, which will probably return ‘the right answer’ in a second. It is a different question, however, whether many web users, or in this case Futisforum members, actually do that in case of others’ nicknames. Overt meta-talk about nicknames has surfaced at times; particularly prevalent examples are Maakuntasatraappi (‘province satrap’) and Boris Pugo (Latvian communist leader in the Soviet era). [Go back up]

[25] At the time of finishing this article, a very active new nickname ‘Le Garçon Formidable’, apparently specialized in French football, had surfaced. [Go back up]

[26] Tampere is a major city in Finland: by population it is outnumbered only by the capital city Helsinki and the city of  Espoo, which is part of the Helsinki metropolitan area. [Go back up]

[27] Actually Norway also surfaces in the two Futisforums’ humorously framed topics about winter sports and the cult sports reporter Jari Porttila, whose language use is referred to, mocked and made much fun of in the two Futisforums. [Go back up]

[28] Nowadays this is relatively easy to tentatively verify with the help of automatic online translation, such as the currently popular (but not very sophisticated) Google Translate. [Go back up]

[29] ‘Strong/stiff defender’ and ‘Tamme Car’ (the name of a football club), respectively. [Go back up]

[30] The following features have led to an item’s classification here as ‘colloquial Finnish’: a) obscenities such as ‘vittu’ (‘cunt’, ‘fuck’), ‘paska’ (‘shit’), b) contracted word forms in all word classes, such as ‘nii’ (‘so’, ‘oh’, ‘well’, ‘um’), ‘ois’ (‘would be’), ‘mun’ (‘my, mine’), ‘tost’ (‘that’, ‘it’ [pronoun in elative case]), c) lexemes clearly at home in colloquial registers and styles such as ‘saletisti’ (‘for sure’), ‘kundi’ (‘guy’, ‘dude’, ‘lad’), ‘neekeri’ (‘nigger’). Moreover, there were markedly non-Standard or colloquial sentence or clause formations even in Signatures where all the single words were more or less Standard. [Go back up]

[31] ‘Sohlo, a good name for a traditional Finnish striker, or a football kicker in general. –Nahka_Tapsa’ [Go back up]

[32] ‘I want the Zambians back here, they couldn’t lose so badly even if they were paid for it. Fuck’ [Go back up]

[33] HJK from Helsinki was the biggest football club in Finland during the first decade of the 2000s. [Go back up]

[34] ‘My girlfriend asked if Malinen is a gay’ [Go back up]

[35] From the film Withnail and I. [Go back up]

[36] After Benjamin Franklin. [Go back up]

[37] This is a quote from within the forum aimed at ridiculing the user of an unidiomatic English expression. [Go back up]

[38] A slogan of the fans of Hammarby (a Stockholm-based club), roughly translatable as “but as long as we stand here on the stands, football will never die!” [Go back up]

[39] A chant sung by Real Madrid fans in honour of Juanito, the player. [Go back up]

[40] Catalan for “Eat well, shit a lot and do not be afraid of death”. [Go back up]

[41] ‘Score goals, Jaro!’ [Go back up]

[42] ‘Smile, people love idiots.’ [Go back up]

[43] ‘Hard Bass’, the dance music genre. [Go back up]

[44] ‘Ellas.Europi.Panathinaikos’. This user’s ‘Favourite team’ is in Greek script too: “ΠΑΝΑΘΗΝΑΙΚΟΣ ΘΥΡΑ 13” (see Appendix 5). [Go back up]

[45] ‘For the Honorable Cross and Golden Freedom’. Note that Serbian could optionally be written in Cyrillic script; that script choice might in itself have ideological or other indexical social meanings. [Go back up]

[46] Swedish is de jure a domestic language, but de facto foreign to most Finns. [Go back up]

[47] Most of the time I utilized the currently popular engine Google, in the knowledge of some of its biases such as search results tailored according to the user’s habits with the help of ‘cookies’. [Go back up]

[48] For a socio-historical overview of the convergence of these two ‘roles’ in the merged role of ‘prosumer’ or ‘conducer’, see Ritzer (2010). [Go back up]

[49] Дасаев; thus Dasayev or Dasaev in Anglo transliteration. [Go back up]

[50] The same friend aptly pointed out that over the same childhood years, only 2–3 years later, we became ardent Tolkien enthusiasts; in that zeal for fantasy, play and imagination, he was Aragorn (the Ranger and King), while I was Legolas (the Elven warrior in the Fellowship of the Ring). I owe a debt to Lauri Rikala for this insight. [Go back up]

[51] http://suomifutisnet.adv1.nebula.fi/phpBB2/memberlist.php?mode=joindate&order=ASC&start=104 [Go back up]

[52] Some snippets from these early periods have been archived by Internet Archive Wayback Machine at http://wayback.archive.org/web/19970815000000*/http://soccernet.fi [Go back up]

[53] http://suomifutisnet.adv1.nebula.fi/phpBB2/memberlist.php?mode=posts&order=DESC&start=0 [Go back up]

[54] http://suomifutisnet.adv1.nebula.fi/phpBB2/memberlist.php?mode=posts&order=DESC&start=200 [Go back up]

[55] http://futisforum2.org/index.php?action=mlist;sort=registered;start=0 (requires registration) [Go back up]

[56] http://futisforum2.org/index.php?action=mlist;sort=posts;start=0 (requires registration) [Go back up]

[57] http://futisforum2.org/index.php?action=mlist;sort=posts;start=225 (requires registration) [Go back up]

[58] http://suomifutisnet.adv1.nebula.fi/phpBB2/profile.php?mode=editprofile (requires registration) [Go back up]

[59] http://futisforum2.org/index.php?action=profile;sa=forumProfile (requires registration) [Go back up]

References

Androutsopoulos, Jannis. 2004. “Non-native English and sub-cultural identities in media discourse”. Den fleirspråklege utfordringa, ed. by Helge Sandøy, Endre Brunstad, Jon-Erik Hagen & Kari Tenfjord, 83–98. Oslo: Novus.

Androutsopoulos, Jannis. 2006. “Multilingualism, diaspora, and the Internet: Codes and identities on German-based diaspora websites”. Journal of Sociolinguistics 10(4): 429–450.

Androutsopoulos, Jannis. 2007. “Language choice and code switching in German-based diasporic web forums”. The Multilingual Internet: Language, Culture, and Communication Online, ed. by Brenda Danet & Susan C. Herring, 340–361. Oxford: Oxford University Press.

Androutsopoulos, Jannis. 2008. “Potentials and limitations of discourse-centred online ethnography”. Data and Methods in Computer-mediated Discourse Analysis, special issue Language@internet, ed. by Jannis Androutsopoulos & Michael Beißwenger, 5(8). http://nbn-resolving.de/urn:nbn:de:0009-7-16100

Androutsopoulos, Jannis. 2011. “From variation to heteroglossia in the study of computer-mediated discourse”. Digital Discourse: Language in the New Media, ed. by Crispin Thurlow & Kristine Mroczek, 277–298. Oxford: Oxford University Press.

Androutsopoulos, Jannis. 2012. “English ‘on top’: discourse functions of English resources in the German mediascape”. Sociolinguistic Studies 6(2): 209–238.

Androutsopoulos, Jannis. Personal communication. 3–13 April 2008 in Amsterdam, London and Birmingham (UK); 26–29 May 2009 in Jyväskylä.

Bakhtin, Mikhail. 1984. Problems of Dostoevsky’s Poetics, (Originally published in Russian in 1963.) Edited and translated from Russian into English by Caryl Emerson. Minneapolis, MN: University of Minnesota Press.

Baldry, Anthony & Paul Thibault. 2006. Multimodal Transcription and Text Analysis. London: Equinox.

Bateson, Gregory. 1972 [1955]. Steps to an Ecology of Mind: Collected Essays in Anthropology, Psychiatry, Evolution, and Epistemology. Chicago, IL: University of Chicago Press.

Bauman, Richard & Charles L. Briggs. 1990. “Poetics and performance as critical perspectives on language and social life”. Annual Review of Anthropology 19: 59–88.

Baym, Nancy K. 2000. Tune In, Log On: Soaps, Fandom, and Online Community. Thousand Oaks, CA: Sage.

Bechar-Israeli, Haya. 1995. “From <Bonehead> to <cLoNehEAd>: Nicknames, play, and identity on Internet Relay Chat”. Journal of Computer-Mediated Communication 1(2). doi:10.1111/j.1083-6101.1995.tb00325.x

Benford, Robert D. & David A. Snow 2000. “Framing processes and social movements: an overview and assessment”. Annual Review of Sociology 26: 611–639.

Blommaert, Jan. 2005. Discourse: A Critical Introduction. Cambridge: Cambridge University Press.

Blommaert, Jan 2010. The Sociolinguistics of Globalization. Cambridge: Cambridge University Press.

Blommaert, Jan, James Collins & Stef Slembrouck. 2005. “Spaces of multilingualism”. Language & Communication 25(3): 197–216.

Blommaert, Jan & Ben Rampton. 2011. “Language and superdiversity”. Diversities 13(2): 3–21. http://www.unesco.org/shs/diversities/vol13/issue2/art1

Coupland, Nikolas. 2009. “The mediated performance of vernaculars”. Journal of English Linguistics 37: 284–300.

Crystal, David. 2006. Language and the Internet. Second, revised edition. Cambridge: Cambridge University Press.

Gardner-Chloros, Penelope. 2009. Code-switching. Cambridge: Cambridge University Press.

Goffman, Erving. 1981. Forms of Talk. Philadelphia, PA: University of Pennsylvania Press.

Goffman, Erving. 1986. Frame Analysis: An Essay on the Organization of Experience. Second edition. Boston, MA: Northeastern University Press.

Gumperz, John J. 1982. Discourse Strategies. Cambridge: Cambridge University Press.

Hall, Chris. 2007. “Recent developments in Finnish language education policy: a survey with particular reference to German”. German as a Foreign Language 3/2007: 1–24

Heinonen, Harri. 2005. Jalkapallon lumo: Tutkimus suomalaisesta Everton-faniudesta. Jyväskylä: Atena.

Heller, Monica. 2007. “Bilingualism as ideology and practice”. Bilingualism: A Social Approach, ed. by Monica Heller, 1–22. Basingstoke: Palgrave Macmillan.

Hinrichs, Lars. 2006. Codeswitching on the Web: English and Jamaican Creole in E-mail Communication. Amsterdam: John Benjamins.

Hymes, Dell. 1996. Ethnography, Linguistics, Narrative Inequality: Toward an Understanding of Voice. London: Taylor & Francis.

Iedema, Rick 2003. “Multimodality, resemiotization: Extending the analysis of discourse as multi-semiotic practice”. Visual Communication 2(1): 29–57.

Jørgensen, J. Normann. 2008. “Polylingual languaging around and among children and adolescents”. International Journal of Multilingualism 5(3): 161–176.

Jørgensen, J. Normann, Martha S. Karrebæk, Lian M. Madsen & Janus S. Møller. 2011. “Polylanguaging in Superdiversity”. Diversities 13(2): 23–38. http://www.unesco.org/shs/diversities/vol13/issue2/art2

Kelly-Holmes, Helen. 2005. Advertising as Multilingual Communication. Houndmills, NY: Palgrave.

Kress, Günther R. & Theo van Leeuwen. 1996. Reading Images: The Grammar of Visual Design. London: Routledge.

Kress, Günther & Theo van Leeuwen. 2001. Multimodal Discourse – The Modes and Media of Contemporary Communication. London: Arnold.

Kytölä, Samu. 2012a. “Multilingual web discussion forums: Theoretical, practical and methodological issues”. Language Mixing and Code-Switching in Writing: Approaches to Mixed-Language Written Discourse. Routledge Critical Studies in Multilingualism, ed. by Mark Sebba, Shahrzad Mahootian & Carla Jonsson, 106–127. New York: Routledge.

Kytölä, Samu. 2012b. “Peer normativity and sanctioning of linguistic resources-in-use: On non-Standard Englishes in Finnish football forums online”. Dangerous Multilingualism: Northern Perspectives on Order, Purity and Normality, ed. by Jan Blommaert, Sirpa Leppänen, Päivi Pahta & Tiina Räisänen, 228–260. Basingstoke: Palgrave Macmillan.

Kytölä, Samu 2013. Multilingual Language Use and Metapragmatic Reflexivity in Finnish Online Football Forums: A Study in the Sociolinguistics of Globalization. PhD dissertation, University of Jyväskylä. http://urn.fi/URN:ISBN:978-951-39-5132-0

Kytölä, Samu. 2014. “Negotiating multilingual discourse in a Finland-based online football forum: Metapragmatic reflexivity on intelligibility, expertise and ‘nativeness’”. Globalisering, sprogning og normativitet – online og offline Copenhagen Studies in Bilingualism 70, ed. by J. Spindler Møller & L. Malai Madsen.

Kytölä, Samu. In press. “Translocality”. Handbook of Language and Digital Communication, ed. by A. Georgakopoulou & T. Spilioti. London: Routledge.

Kytölä, Samu & Jannis Androutsopoulos. 2012. “Ethnographic perspectives on multilingual computer-mediated discourse: Insights from Finnish football forums on the Web”. Multilingualism, Discourse and Ethnography, ed. by Sheena Gardner & Marilyn Martin-Jones, 179–196. New York, NY: Routledge.

Lave, Jean & Etienne Wenger. 1991. Situated Learning: Legitimate Peripheral Participation. Cambridge: Cambridge University Press.

Lawson, Danielle. 2003. “Blurring the boundaries: Ethical considerations for online research using synchronous CMC Forums”. Readings in Virtual Research Ethics: Issues and Controversies, ed. by Elizabeth A. Buchanan, 80–100. Hershey, PA: Information Science.

Le Page, Robert B. & Andrée Tabouret-Keller. 1985. Acts of Identity: Creole-based Approaches to Language and Ethnicity. Cambridge: Cambridge University Press.

Leppänen Sirpa. 2007. “Youth language in media contexts: Insights into the functions of English in Finland”. World Englishes 26(2): 149–169.

Leppänen, Sirpa, Samu Kytölä, Henna Jousmäki, Saija Peuronen & Elina Westinen. 2014. “Entextualization and resemiotization as resources for identification in social media”. The Language of Social Media: Identity and Community on the Internet, ed. by Philip Seargeant & Caroline Tagg, 112–136. Basingstoke: Palgrave Macmillan.

Leppänen, Sirpa & Tarja Nikula. 2007. “Diverse uses of English in Finnish society: Discourse-pragmatic insights into media, educational and business contexts”. Multilingua 26(4): 333–380.

Leppänen, Sirpa & Saija Peuronen. 2012. “Multilingualism on the Internet”. The Routledge Handbook of Multilingualism., ed. by Marilyn Martin-Jones, A. Blackledge & A. Creese, 384–402. London: Routledge.

Leppänen, Sirpa, Anne Pitkänen-Huhta, Tarja Nikula, Samu Kytölä, Timo Törmäkangas, Kari Nissinen, Leila Kääntä, Tiina Räisänen, Mikko Laitinen, Päivi Pahta, Heidi Koskela, Salla Lähdesmäki & Henna Jousmäki. 2011. National Survey on the English Language in Finland: Uses, Meanings and Attitudes (Studies in Variation, Contacts and Change in English 5). Helsinki: VARIENG. http://www.helsinki.fi/varieng/series/volumes/05/

LeVine, Philip & Ron Scollon, eds. 2004. Discourse & Technology: Multimodal Discourse Analysis. Washington, DC: Georgetown University Press.

Makoni, Sinfree & Alastair Pennycook. 2007. “Disinventing and reconstituting languages”. Disinventing and Reconstituting Languages, ed. by Sinfree Makoni & Alastair Pennycook, 1–41. Clevedon: Multilingual Matters.

Martin, Elizabeth. 2007. “‘Frenglish’ for sale: Multilingual discourses for addressing today’s global consumer”. World Englishes 26(2): 170–188.

Montgomery, Martin. 2008. An Introduction to Language and Society. Third, revised edition. Abingdon: Routledge.

Norris, Sigrid & Rodney H. Jones, eds. 2005. Discourse in Action: Introducing Mediated Discourse Analysis. Abingdon: Routledge.

Pennycook, Alastair. 2007. Global Englishes and Transcultural Flows. Abingdon: Routledge.

Piller, Ingrid. 2001. “Identity constructions in multilingual advertising”. Language in Society 30(2): 153–186.

Pöyhönen, Sari. 2009. “Foreign language teaching in basic and secondary education in Finland: Current situation and future challenges”. Plurilinguismo e innovazione di sistema: sfide e ricerche curricolari in ambito nazionale e internazionale, ed. by Sandra Lucietto, 145–173. Provincia Autonoma di Trento: IPRASE del Trentino.

Prior, Paul. 2010. “Remaking IO: Semiotic remediation in the design process”. Exploring Semiotic Remediation as Discourse Practice, ed. by Paul Prior & Julie Hengst, 206–234. London: Palgrave Macmillan.

Rampton, Ben. 2005. Crossing: Language & Ethnicity Among Adolecents. Second, revised edition. Manchester: St. Jerome.

Ritzer, George. 2010. “Focusing on the prosumer: On correcting an error in the history of social theory”. Prosumer revisited: zur Aktualität einer Debatte, ed. by Birgit Blättel-Mink & Kai-Uwe Hellmann, 61–49. Wiesbaden: VS Verlag für Sozialwissenschaften.

Sajavaara, Kari. 1983. “Anglo-American influence on Finnish”. The Impact of American Culture: Proceedings of an International Seminar, ed. by Eero Kuparinen & Keijo Virtanen, 36–49. Turku: University of Turku.

Scheufele, Dietram A. 1999. “Framing as a theory of media effects”. Journal of Communication 49(4): 103–122.

Scollon, Ron. 2001. Mediated Discourse: The Nexus of Practice. London: Routledge.

Scollon, Ron & Suzie Wong Scollon. 2003. Discourses in Place: Language in the Material World. London: Routledge.

Sebba, Mark. 2012. “Researching and theorising multilingual texts”. Language Mixing and Code-switching in Writing: Approaches to Mixed-language Written Discourse, ed. by Mark Sebba, Shahrzad Mahootian & Carla Jonsson, 1–26. Routledge Critical Studies in Multilingualism 2. New York, NY: Routledge.

Sebba, Mark, Shahrzad Mahootian & Carla Jonsson, eds. 2012. Language Mixing and Code-switching in Writing: Approaches to Mixed-language Written Discourse. Routledge Critical Studies in Multilingualism 2. New York, NY: Routledge.

Statistics Finland. 2014. Population According to Language 1980–2013. http://www.stat.fi/til/vaerak/2013/vaerak_2013_2014-03-21_tau_001_en.html

Stommel, Wyke. 2007. “Mein nick bin ich! Nicknames in a German forum on eating disorders”. Journal of Computer-Mediated Communication 13(1). doi:10.1111/j.1083-6101.2007.00390.x

Tannen, Deborah. 1993. “What’s in a frame? Surface evidence for underlying expectations”. Framing in Discourse, ed. by Deborah Tannen, 14–56. Oxford: Oxford University Press.

Thorne, Steven L. & Ingrid Fischer. 2012. “Online gaming as sociable media. Le jeu en ligne, une forme de réseau social”. Apprentissage des Langues et Systèmes d’Information et de Communication 15(1). doi:10.4000/alsic.2450

Turtiainen, Riikka. 2004. Virtuaalikenttien Pelét: Huomioita Internetin jalkapalloaiheisen keskustelupalstan nimimerkeistä. Unpublished Bachelor’s thesis (‘proseminaarityö’). University of Turku, Department of Cultural Production and Landscape Studies.

Wenger, Etienne. 1998. Communities of Practice: Learning, Meaning, and Identity. Cambridge: Cambridge University Press.

Appendices

The screen views included and URL addresses quoted here date from 30 March 2012 and 6 July 2012. The final samples for the quantitative analysis were conducted on 30 March 2012.

Appendix 1: Screen views of the two Futisforum memberlists

Appendix 2: Screen views of the ‘edit profile’ dialogue window

Appendix 3: Screen views of how member profiles appear to other members

Appendix 4: Three messages from Futisforum illustrating different subspaces at work

This appendix includes views of Futisforum messages displaying both polylingual language use and the relative positions of the different subspaces to one another.

Appendix 5: Three messages from FF2 illustrating different subspaces

This appendix includes views of FF2’s messages displaying both polylingual language use and the relative positions of the different subspaces to one another. The non-changeable software meta-tools are in Finnish, embedded quoting of a previous message is part of the layout (the darker-coloured rectangle), the relatively permanent pseudonym (‘paofinlandia’, ‘uneek’, ‘Dieter Bohlen’) is in bold in top left, the tailorable ‘Favourite team’ (‘Suosikkijoukkue’) in middle left, and the signature on the bottom, is separated by a horizontal line.