A comparison of spoken and signed lingua franca communication: the case of English as a lingua franca (ELF) and International Sign (IS) English lingua

: This article explores similarities between English as a lingua franca (ELF) and International Sign (IS), two lingua franca phenomena which in the last decades have been subject to increasing, albeit independent, linguistic research. In contrast to spoken intercultural communication, in which English often represents a shared resource that speakers from different linguacultural backgrounds draw on, in the visual-gestural modality no specific sign language has yet gained such global reach. Instead, in many international contexts IS is used: a lingua franca that can be more or less conventionalized and that is not based on one particular sign language. IS use depends on the communicative situation, in which signers ﬂ exibly and creatively use different signs from natural sign languages as well as iconic elements and gestures. Despite overt formal differ-ences between ELF and IS, when focusing on the actual communication process, rather than the forms that result from it, the two lingua franca phenomena share many similarities. In fact, both ELF and IS are variable communicative means that get situationally adapted by speakers and signers on the basis of different resources they have at their disposal. Similar discussions about the dif ﬁ culty of conceptualizing ELF and IS, about the role of multilingual resources, and about interaction processes at play can thus be found in both ELF and IS liter-ature. This insight opens up new possibilities for researchers in the two ﬁ elds to mutually bene ﬁ t from the study of lingua franca communication in the other modality, which prompts the need for a cross-modal collaboration between ELF and IS researchers.

A comparison of spoken and signed lingua franca communication: the case of English as a lingua franca (ELF) and International Sign (IS) Ein Vergleich zwischen gesprochener und gebärdeter Lingua franca Kommunikation: der Fall von English as a lingua franca (ELF) und International Sign (IS) https://doi.org/10. 1515/jelf-2021-2058 Abstract: This article explores similarities between English as a lingua franca (ELF) and International Sign (IS), two lingua franca phenomena which in the last decades have been subject to increasing, albeit independent, linguistic research. In contrast to spoken intercultural communication, in which English often represents a shared resource that speakers from different linguacultural backgrounds draw on, in the visual-gestural modality no specific sign language has yet gained such global reach. Instead, in many international contexts IS is used: a lingua franca that can be more or less conventionalized and that is not based on one particular sign language. IS use depends on the communicative situation, in which signers flexibly and creatively use different signs from natural sign languages as well as iconic elements and gestures. Despite overt formal differences between ELF and IS, when focusing on the actual communication process, rather than the forms that result from it, the two lingua franca phenomena share many similarities. In fact, both ELF and IS are variable communicative means that get situationally adapted by speakers and signers on the basis of different resources they have at their disposal. Similar discussions about the difficulty of conceptualizing ELF and IS, about the role of multilingual resources, and about interaction processes at play can thus be found in both ELF and IS literature. This insight opens up new possibilities for researchers in the two fields to mutually benefit from the study of lingua franca communication in the other modality, which prompts the need for a cross-modal collaboration between ELF and IS researchers.

Introduction: two lingua franca phenomena under investigation
When people from different linguacultural backgrounds communicate, they do so by use of an auxiliary linguistic means, a so-called 'lingua franca' (Samarin 1987: 371). Through the course of history, various kinds of linguistic means such as pidgins, full-fledged languages and artificially created ones have performed lingua franca function in different places and among different people (Meierkord and Knapp 2002: 9-14). Recently, one lingua franca has particularly stood out, having reached truly global dimensions and being used in all manner of infinitely varied contexts and settings: English as a lingua franca (ELF). In fact, given that English has been learned as an additional language by a vast number of people, it often represents a common shared resource in all kinds of intercultural settings. This widespread use of English has given rise to a vibrant and fast-growing field in applied linguistics, which has offered valuable insights into ELF as a multilingual phenomenon where users exploit all kinds of resources in line with their communicative purposes (e.g . Cogo 2016;Hülmbauer 2013;Jenkins 2015;Pitzl 2018a;Seidlhofer 2011). Even if "the term lingua franca has an immediate verbal connotation" (Kellett Bidoli 2014: 101), probably because its origins lie in spoken language (Meierkord and Knapp 2002: 9), intercultural communication also takes place in the visualgestural modality, namely in sign language. 1 Just like spoken languages, also sign languages have developed independently in various deaf communities around the world, they differ between one another and thus communication between signers from different linguacultural backgrounds likewise takes place through the use of a lingua franca. However, different from ELF speakers, who predominantly draw on recognizably English resources, signers are able to communicate with each other without necessarily adhering to any specific common sign language. This is illustrated in the following anecdotal account: Standing in the lobby of a hotel in Washington, DC, waiting to check in among a crowd of deaf people from many different countries Anja Hiddinga's ten-year-old deaf son became friendly with a man from Egypt. They chatted along, although the little one did not know any Egyptian Sign Language or American Sign Language (ASL), and it was not Sign Language of the Netherlands they were conversing in either. In communicating back and forth, the conversation seemed to develop quite easily. (Crasborn and Hiddinga 2015: 59) 2 How is it possible that a Dutch child and an Egyptian man managed to chat along without drawing on any specific sign language? And if it was not one of the above-mentioned sign languages they used, then what was it? What we see in this passage is one example of how signers from different linguacultural backgrounds are able to bridge language barriers through the use of what has now widely come to be known as International Sign (IS) (or international sign): "a flexible and highly context-dependent form of communication […] in which shared elements from [the signers'] own sign languages and elements of shared spoken languages are combined with pantomimic elements" (Hiddinga and Crasborn 2011: 483). IS is not a specific pre-established linguistic system, but rather a set of strategies that are employed to communicate in international settings and so "[t]he use of IS is variable and dependent on the geographical, political, social, cultural and linguistic contexts in which it occurs and the backgrounds and intentions of the people who use it" (Kusters 2020: 53). In some contexts, such as at regularly occurring deaf events and gatherings, IS is more conventionalized Whynot 2016b); in situations when people meet for the first time, shared ground has to be negotiated in situ. This latter, less conventionalized, form of IS communication, which to date has only been investigated in a few experimental studies (see Byun et al. 2018;Zeshan 2015;Zeshan and Webster 2020), is also called 'cross-signing' (Zeshan 2015). 3 It has been argued that such a way of communicating is only feasible in the visual-gestural modality, which lends itself well for more iconic representations (Byun et al. 2018: 330;Woll 1990: 118). Also, there seem to be considerable structural similarities between different sign languages, especially in terms of (iconic) grammatical constructions, that can facilitate IS communication (Supalla and Webb 1995). Some researchers have thus claimed that IS represents a unique linguistic phenomenon without a real spoken counterpart (Bradford et al. 2020;Kusters 2019). IS is indeed different from spoken lingua franca phenomena such as ELF, which can only be used because its speakers acquired English at some point in their lives. However, ELF is not a "purely English" phenomenon, or an English variety adopted and used for international purposes. Rather, ELF is a dynamic, multilingual and situationally dependent language use where all kinds of resources are exploited to achieve different communicative purposes (Seidlhofer 2011: 111) and this is where we can actually see interesting parallels to IS.
This article aims at illustrating that ELF and IS researchers, despite having studied lingua franca communication in different language modalities (i.e. the oral-auditory and the visual-gestural one), have made similar observations, which have led to similar discussions in the respective fields. This article therefore represents a first attempt to relate two research fields that have hitherto been separate and so it shall be seen as a first step in the direction of a possible future collaboration between IS and ELF researchers. 4 2 Language or variety? The conceptualization of IS and ELF At the beginning of this comparison, it is necessary to explore a fundamental issue that has been addressed by researchers in both lingua franca research fields: the question of how to conceptualize their subjects under investigation. In fact, although it can be said that ELF and IS are lingua francas, in that they are used to overcome communication barriers between people who do not share the same first language (Samarin 1987: 371), finding suitable categories for ELF and IS seems to present more of a challenge and has drawn attention to the limitations of traditional concepts in linguistics. As already anticipated in the introduction, ELF is not Standard English or any other native variety of English distributed and reproduced "beyond their native boundaries" (Samarin 1987: 371). Rather, ELF is shaped and customized on the basis of multilingual resources by its users all over the globe and, more often than not, in the absence of L1 speakers of English. This means that L1 speakers of English are not the "rightful owners" of the lingua franca and that ELF should also be studied as a legitimate use of language in its own right, in detachment from any native-speaker oriented judgments (e.g . Jenkins 2000;Seidlhofer 2001;Widdowson 1997). Just as ELF is not a learned and reproduced version of English as a native language (ENL), ELF researchers now generally agree that ELF is not an independent variety of English either (e.g. Jenkins 2015;Kalocsai 2014;Pitzl 2018b;Seidlhofer 2011;Widdowson 2015), given that the lingua franca is not used in a stable speech community and so there is no stable, fixed set of linguistic forms that can be identified in all contexts in which ELF is used. Rather, a key feature of ELF is "an inherent diversity" (Firth 2009: 162), which not only arises from the ELF speakers' varying degree of proficiency in English, or from the specific linguistic properties of ELF speakers' L1s (Mauranen 2012), but foremost from the situationally variable use of all kinds of linguistic resources in line with the requirements of the interaction (Hülmbauer 2009(Hülmbauer , 2013Seidlhofer 2009;Widdowson 2015). As Widdowson (2015) describes it: What we see in ELF is the pragmatic process of communication live, in action, laid bare, so to speakopen to observation if only we can rid ourselves of our preconceptions based on too much familiarity with the form that this process takes in particular languages or varieties of language in particular established linguacultural communities. (Widdowson 2015: 367, original emphasis) Thus, in order to understand how ELF communication unfolds, many ELF scholars (e.g. Cogo 2016; Pitzl 2018a; Seidlhofer 2009) have moved away from rigidly focusing on formal linguistic features, whether they might constitute a possible "ELF variety" or not, and instead have focused on the actual communication process and on variation as it fulfills the specific needs of the interlocutors in a given situation.
An important concept that helps explain this variable and creative language use is the 'virtual language,' originally formulated by Widdowson (1997) and later also taken up by other ELF researchers (e.g. Hülmbauer 2013; Klötzl and Swoboda 2020;Seidlhofer 2011;Vettorel 2014). The virtual language can be defined as a "resource for making meaning immanent in the language which simply has not hitherto been encoded and so is not, so to speak, given official recognition" (Widdowson 1997(Widdowson : 138, 2020. Whenever people learn a language, they automatically also acquire these specific virtual underlying principles, which allow for the creation of new forms that have not yet been attested and may thus be labeled as "non-conformities." Widdowson's concept not only explains why ELF users manage to understand one another even if they "violate" conventionally established rules, but it also helps overcome the impact of Standard English ideology on the way that non-standard English use is perceived, i.e. as something "deviant" from ENL norms. In fact, considering that there is an underlying virtual potential of English (as there is with any language), it becomes evident that not only ELF, but also what we consider the standard variety of English and in fact, any kind of English variety, are all varied virtual language manifestations. Whereas the virtual language mainly refers to the recognizably English part in ELF, it has to be acknowledged that this virtuality in ELF is "exponentially extended through the resources available from its speakers' repertoires" (Hülmbauer 2013: 53). Consequently, multilingualism and ELF should not be treated as two separate fields of study, but they are essentially and inevitably intertwined (Hülmbauer and Seidlhofer 2013;Jenkins 2015), an aspect that will be further elaborated in Section 3.
Looking at IS communication now and considering that the signed lingua franca emerges from the interplay of different (sign) languages interactants know and the employment of iconic elements (Hiddinga and Crasborn 2011: 483), it seems that IS signers do not exploit the virtual capacities of one particular sign language. However, it has been found that in some contexts, IS is strongly influenced by American Sign Language (ASL) (see . This is perceived by many signers as a threat to the development of IS as a "free" and "inclusive" lingua franca, and so many signers actively strive towards minimizing the usage of ASL in intercultural communication situations (Kusters 2020: 65). Considering this, it might be interesting to evaluate and discuss in future if and in how far Widdowson's (1997) concept of the virtual language might be applicable to IS after all, particularly in those intercultural signed interactions in which there is a clear predominance of one sign language, such as ASL (see also Kusters 2021).
The influence of ASL on IS is also closely tied to the question regarding the very nature of the signed lingua franca: Kusters has in fact found that "when the use of ASL in IS is discussed, people often treat IS as a conventional language and spontaneously estimate (and disagree on) percentages of ASL in IS" (Kusters 2020: 52), which is also the reason why Kusters approaches IS as a "contact language which is subject to language contact (with ASL)" (Kusters 2020: 52, original emphasis). Other researchers have however claimed that, from a linguistic point of view, similar to what has been discussed in the ELF literature, IS is too variable to be categorized as a language (in the traditional sense) (see Hansen 2016;Mesch 2010).
The discussion of whether IS should be considered a language was particularly prominent in 2003 and 2007 at the General Assembly of the World Federation of the Deaf (WFD), when some participants claimed "that more research was needed before [IS] could be regarded as a language" (Mesch 2010: 3). On that note, sign language researchers and interpreters from all over the world were asked to fill out a questionnaire about how they would define and conceptualize IS (Mesch 2010: 3). The results revealed that most of the respondents objected to the signed lingua franca being considered a language, given that, as one respondent noted, IS "is a hybrid code that is improvised in particular times and places, between particular participants. This temporary usage means the form of IS is too variable or unpredictable to be named 'a language' in the sense of a conventional system (even though conventional languages do contain internal variation)" (Mesch 2010: 5).
According to the passage above, IS does not qualify as a language for the same reason that ELF does not classify as a variety: a lack of stability, particularly "in relation to a specific regional, historical, and social area" (Hansen 2016: 17). And although in some recurrent contexts, such as at conferences, sport events, or education camps, a more conventionalized IS is used as an official lingua franca, IS in these settings is not a fixed and standardized code either (Hiddinga and Crasborn 2011: 494-495). For example, IS interpreters at conferences likewise have to flexibly and spontaneously employ different signs and gestures that audience members will most likely understand (see e.g. Locker McKee and Napier 2002; Whynot 2016b).
Researchers have therefore suggested that IS can be best classified as a "pidgin" (Locker McKee and Napier 2002), or, in initial interactions, an "emerging jargon" (Zeshan 2015), given that the lingua franca emerges from the signers' resources in contact. However, in contrast to spoken pidgins (e.g. Mühlhäusler 1986), IS does not take place in one specific relatively stable community, but rather is shaped in all kinds of situations, between signers from different language backgrounds who communicate for varying periods of time. Consequently, different linguistic forms are exploited to communicate in these intercultural settings and so it is questionable if pidgin might be the right term for this variable, unstable use in the first place (Hansen 2016: 19). Others have proposed that IS bears similarities to "koinés", which are the result of language contact between mutually intelligible dialects (Supalla and Webb 1995). The authors argue that this label might be well applicable to IS, as the participants' native sign languages in their study, mainly North American and European ones, depict quite similar linguistic structures (Supalla and Webb 1995: 348). However, given that not all IS interactions involve historically related sign languages, the concept of koiné does not generally fit for defining IS either (Hansen 2016: 19-20).
In essence, it can be said that, "although the nomenclature may lead to such an assumption" (Whynot 2016a: 35), a standard of IS does not exist, as the forms that can be observed substantially differ in each and every IS signing situation. It is this "rampant variation" (Green 2014: 446) that has caused considerable difficulties in defining IS on the basis of its formal characteristics and thus in classifying it as a language, a pidgin or a koiné.
This "rampant variation" seems to be a common property of ELF and IS, as they are not (always) used among stable group configurations and consequently are not classifiable on the basis of recurrent formal features. It is therefore not possible to "squeeze" ELF and IS into linguistic categories that assume a certain stability of linguistic forms, which has ultimately led to "conceptual challenges that we are faced with in the age of globalization" (Seidlhofer 2011: 81). Dewey (2009) sees the difficulties in conceptualizing ELF in the traditional structural linguistics way with its "tendency towards an objectification of language" (Dewey 2009: 67). According to this conventional view, language is seen as an "autonomous and reified system, exteriorized from any speakers who might actually produce it" (Dewey 2009: 67). Hansen (2016: 22), when talking about IS, also suggests that linguists should adopt a wider perspective on human communication and not only regard language as a product on the basis of its formal characteristics, but as a dynamic process whereby language resources are appropriated and adapted to the needs of the interlocutors in a given situation.
In sum, it can be said that well-established linguistic categories to describe language are not suitable to account for the lingua franca phenomena we can currently witness on a global scale: IS and ELF are not languages, pidgins or varieties; they are global lingua francas that emerge, are shaped and used for communicative purposes by signers and speakers from different linguacultural backgrounds. And this leads to another aspect that has been discussed in both ELF and IS research: the fact that lingua franca communication unfolds through the exploitation of all kinds of linguistic resources speakers and signers have at their disposal.

An interplay of different linguistic resources
As pointed out in the section above, the functional use of ELF and IS comes along with an abundance of different linguistic forms that can be observed as lingua franca users draw on and flexibly adapt their resources for communicative effect. This use of multilingual resources is particularly visible in initial IS interactions between people who share very little from the outset, i.e. in cross-signing (Zeshan 2015). Particularly relevant in this regard is the long-term project Multilingual Behaviors in Sign Language Users (MULTISIGN) (see Zeshan and Webster 2020), aimed at studying sign language communication in various multilingual settings, focusing also on the development of an emerging jargon in interactions between signers from different language backgrounds observed over the course of several weeks. 5 Zeshan (2015), who studied these initial interactions, offers a very vivid metaphor of the different meaning-making resources at play: "Just like in an orchestra, where players of all instruments are present at all times and monitoring what is going on, ready to join in at the right moment, the multilingual-multimodal capabilities are always available to be integrated into utterances, either as a 'solo' or in combination with other elements" (Zeshan 2015: 232).
In order to perform a "harmonious linguistic symphony," signers can choose to directly take over a lexical form from any natural sign language they know or to produce a mix of different linguistic and communicative resources, such as pantomime, iconic signs, fingerspelling, or mouthing. An example of these multilingual-multimodal resources at play can be found in a short snippet of a first IS interaction, in which a Jordanian signer tried to inform his interlocutor from Indonesia about the exact date (26 June) of his travel to London (Zeshan 2015: 232-234). 6 To convey this meaning, he drew on signs from Jordanian Sign Language (LIU), English mouthing, writing on the palm and iconic constructions, using outstretched fingers for the respective numeral. The use of these different resources was prompted by the other interlocutor, who signaled nonunderstanding and thus initiated several repairs.
Interestingly, studies on IS in different contexts have reported that when signers drew on resources from spoken languages, these resources mostly consisted of mouthed and fingerspelled English words (Byun et al. 2020;Locker McKee and Napier 2002;Zeshan 2015). This could be explained by the fact that English also plays a vital role in some intercultural sign language settings (Hiddinga and Crasborn 2011). For example, English is the official written lingua franca at regularly occurring deaf gatherings, such as at conferences hosted by the WFD (Whynot 2016b), or at the education camp Frontrunners , where written English is used, alongside a more conventionalized IS as the signed lingua franca. Apart from these official gatherings, written English is also a means of communication among many deaf people on social media, in chats or on other internet platforms (Kellett Bidoli 2014;Oyserman 2016). It is thus plausible that English (in mouthed and fingerspelled form) may sometimes be drawn on in IS 5 Prior to the official publication of the project results (Adamou et al. 2020;Bradford 2020aBradford , 2020bBradford et al. 2020;Byun 2020;Byun et al. 2020), the cross-signing studies had in part already been discussed in some individual articles (Byun et al. 2018;Zeshan 2015Zeshan , 2017. 6 Note that in Zeshan's (2015) article modality is used differently than how it is used in this paper: it does not refer to the two channels (i.e. the oral-auditory and the visual-gestural one) in which language is produced and perceived (see Section 1). Rather it is used to refer to "several [visualgestural] channels which are simultaneously active and coordinated [in sign languages], such as the hands and arms, the facial expressions, the mouth movements derived from spoken words ('mouthings'), and head and body postures" (Zeshan 2015: 227). communication as one possible resource to reach mutual understanding. It can in fact be argued that English used in IS communication is also a kind of ELF, which to date has not received considerable attention, neither in IS nor in ELF research. This way in which English is used by signers in IS interactions clearly differs from the spoken and written ELF studied so far, as IS signers may draw on English not as a "communicative medium of choice, and often the only option" (Seidlhofer 2011: 7), but as one out of many resources to reach mutual understanding. IS signers using English can nevertheless be regarded as ELF users, as they use and flexibly adapt ELF for their own communicative purposes. This overlap between ELF and IS therefore deserves to be studied in more depth. 7 Considering the use of spoken and written ELF among hearing people, which evidently always involves English as a shared resource, one might get the impression that ELF researchers are also mostly concerned with English. The global spread of English has indeed been perceived as a threat to multilingualism and has triggered considerable concerns about language ownership, prescriptivism and the fear that English will eventually repress and substitute other, perhaps minor, languages in Europe (e.g. Phillipson 2009). 8 However, exactly the opposite is the case: given that ELF takes place among people who come from a variety of different linguacultural backgrounds, ELF researchers have endorsed the view that ELF is not only about the E in the acronym, but an inherently multilingual phenomenon where all kinds of language resources may be exploited (Cogo 2016;Hülmbauer 2009;Hülmbauer and Seidlhofer 2013;Jenkins 2015;Klimpfinger 2009;Mauranen 2012;Pitzl 2016).
In the course of ELF research, this multilingual dimension has been addressed in various manners: some have placed emphasis on the roles of code-switching (e.g. Kalocsai 2014;Klimpfinger 2009). Others have investigated cognates as an indication of the flexible, function-oriented use of all kinds of language resources (Hülmbauer 2009(Hülmbauer , 2013. Recently, in line with current trends in multilingualism studiesand in fact also in studies on IS (Adamou et al. 2020;Kusters et al. 2017) -ELF researchers (e.g. Cogo 2016) have preferred to approach lingua franca communication as a translanguaging phenomenon (García and Wei 2014). According to this view, languages are not to be perceived as demarcated, separate entities, but in their totality as one linguistic repertoire, whose bits can be 7 In my doctoral project I investigate how English mouthing, as a visual manifestation of ELF, is employed by signers in initial IS interactions (i.e. in cross-signing). The video data for my study is taken from the ongoing research project "Deaf communication without a shared language," led by Onno Crasborn at Radboud University in Nijmegen (the Netherlands). 8 These concerns resonate with the concerns about the perceived dominance of ASL in international deaf contexts as expressed by signers in Kusters' (2020) study (see Section 2).
"creatively transformed into new linguistic realities" (Cogo 2016: 5). A translingual view is also adopted by Pitzl (2016Pitzl ( , 2018aPitzl ( , 2018b, who argues that ELF speakers are equipped with their own "individual multilingual repertoires (IMRs)", which not only contain certain English resources, but in fact "all [kinds of] languages/ varieties, dialects and registers and all the links and connections between them" (Pitzl 2018a: 195).
When it comes to the actual manifestation of these resources, Cogo (2016Cogo ( , 2018 distinguishes between "overt" and "covert" multilingual resources. According to Cogo (2016: 4), overt multilingual resources may be visible in ELF in instances of code-switching, whereby ELF users, consciously or not, alternate between identifiable different languages in one speech event. Covert multilingual resources on the other hand "are the result of cross-linguistic or cross-cultural influences in the speakers' repertoires" (Cogo 2018: 358) and are visible in newly invented lexical forms or constructions, which are still recognizable as English. These resources are covert, as it is not easy to uncover which linguistic resources are used in which combination, as they underlie cognitive principles that cannot be observed (Cogo 2016: 3). An example of such covert resources can be found in the term reclamate, used by a Spanish ELF speaker informing her German and Spanish colleagues about a "customer's tendency to complain about payment" (Cogo 2016: 8). Cogo speculates that this word was most likely influenced by the Spanish word reclamar (Cogo 2016: 8), although such an influence can only be surmised (see also the discussion of re-emplace in Pitzl et al. (2008: 33-34)). This example well illustrates how new linguistic expressions emerge both from the inherent potential of the virtual capacities of English, as well as from the participants' other multilingual resources (Hülmbauer 2013).
Cogo's distinction between overt and covert multilingual resources is reminiscent of Zeshan's (2015) distinction between resources used in cross-signing that may appear "as a 'solo'" or "in combination with other elements" (Zeshan 2015: 232) as presented in the passage at the beginning of this section. However, in contrast to Cogo's notion of covert multilingual resources in ELF, which are still recognizable as English, the outcome of combining multilingual resources in IS may not necessarily be identified as a particular sign language (Zeshan 2015: 229), which again hints towards the fact that IS signers do not necessarily exploit the potential of only one virtual language, or in fact of any "language" at all (when using pantomime).
The different studies mentioned above give an idea of how both ELF and IS researchers have approached their lingua franca from a multilingual framework, highlighting that, as lingua franca communication takes place in culturally and linguistically rich settings, multilingual resources are always at play and may at any time be used throughout the interaction. Some ELF researchers (e.g. Klötzl and Swoboda 2020; Seidlhofer and Widdowson 2017) have however recently questioned this highlighting of the multilingual/plurilingual/translingual dimension of ELF as an exclusive defining feature of lingua franca communication, since this view seems to promote the idea that ELF is unique in its "multilingual nature" and thus different from so-called "monolingual" interactions. What ELF users perform is however not that different from what can be observed in any human interaction, namely "the exercise of a general lingual capability" (Seidlhofer and Widdowson 2017: 33). In other words, ELF communication, just like any other communication, involves people who strategically employ their resources in line with the communicative goals, i.e. it involves people engaging in "languaging" (Becker 1995).
This languaging process seems to be more easily discernable in (initial) IS interactions, in particular when the combination of different sign language resources with context-dependent and iconic elements results in the creation of novel signs which are no longer attributable to any specific sign language (Zeshan 2015: 229). In that regard, the concept of 'semiotic repertoires' (Kusters et al. 2017) could be used as an umbrella term to refer to the different meaningmaking resources signers and speakers have at their disposal. This concept would provide a suitable alternative to the notions of multilingual, or multimodal repertoires, as it encompasses all resources regardless of whether they are signed, spoken, mouthed, written, linguistic, non-linguistic or attributable to specific named languages. Indeed, given that this concept has "the potential to bridge studies of multilingualism in spoken and signed languages, gesture studies and multimodality research" (Kusters et al. 2017: 228), it might also have the potential to bridge the studies of ELF and IS.
It can thus be said that ELF and IS users make use of their capability as "linguals" (Seidlhofer and Widdowson 2017)they creatively exploit different semiotic resources and negotiate and co-construct meaning to achieve their communicative goals. And this communicative process can be observed in different constellations and groups, from transient groups who interact only once, to relatively stable groups who interact on a regular basis. In the next section it will be presented how ELF and IS researchers have described and conceptualized these groups.

From transient to stable groups
In our globalized and interconnected world, intercultural communication may occur in all kinds of domains of use and, owing to new means of technology, also in virtual spaces such as in online videogames, chatrooms or on other communication platforms. Consequently, the traditional notion of the 'speech community,' as a group who shares the same native language(s), is unsuitable for describing ELF and IS users. Especially ELF researchers have therefore looked for better alternatives to account for the sociolinguistic realities of the spoken lingua franca and their speakers (see Seidlhofer 2007). One alternative they have considered is Wenger's (1998) 'Communities of Practice (CoP).' A CoP is established not necessarily because its members share a specific native language, but because they engage in regular interaction (mutual engagement) to achieve a specific purpose (joint enterprise), which eventually results in the establishment of a shared repertoire (shared repertoire) (Wenger 1998: 73). Given that in many ELF situations interactants develop and use "linguistic and other resources which are the agreed result of internal negotiations" (Seidlhofer 2007: 314), ELF researchers have proposed that Wenger's CoP concept might indeed be suitable for describing ELF speakers (House 2003;Seidlhofer 2007).
Since the CoP concept was first applied to ELF users at the turn of the millennium, it has been widely used in ELF research as a viable alternative to the speech community concept, especially in contexts in which the same constellations of ELF speakers regularly get together to jointly fulfill a certain goal, such as in educational (Smit 2010), academic (Hynninen 2016;Kalocsai 2014), or business settings (Cogo 2016;Ehrenreich 2010). Recently, also IS researchers have proposed the applicability of Wenger's CoP concept to refer to investigated groups of signers from different sign language backgrounds, observed over a period of some weeks, as part of the MULTISIGN project (see Section 3). Given the small size of the group as well as the short time span of their existence, these groups were referred to as "micro-communities of practice (MCoP)" (Bradford 2020a(Bradford , 2020bBradford et al. 2020). These MCoPs took shape as signers participated in the research project, gradually developed their own linguistic rules and conventionsincluding in-group jokesand established them as identity markers of their group (Adamou et al. 2020: 15). Considering that "[t]he more regular contact is between a group of international deaf people, the more likely they are to develop their own stable variety" (Mesch 2010: 6), it can even be speculated that MCoPs may turn into CoPs when they spend more time together.
It seems that Wenger's CoP concept has been proven suitable for specific investigated domains of ELF and IS use; however, some ELF researchers have also recognized its limitations (Dewey 2009;Hynninen 2016;Jenkins 2015;Pitzl 2016;Smit 2010). In fact, given that larger CoPs (such as at universities or in international companies) are often heterogeneous themselves, being composed of many smaller groups of speakers that interact to a greater or lesser extent and for different periods of times, the CoP concept may require modification "to better fit the fluidity of [such] ELF settings" (Dewey 2009: 77). Apart from the more transient encounters in institutionalized contexts, ELF also takes place to a considerable extent outside these institutions, in fleeting, short-lived encounters between interlocutors who might not necessarily see each other again. In these situations, shared practices may never become stabilized through regular interactions and, consequently, the CoP concept may not be applicable.
To account for those situations in which "a group forms, speakers negotiate and interact, and the group dissolves again" (Pitzl 2018a: 21), Pitzl (2016) introduces the concept of 'Transient International Groups (TIGs).' While certain ELF-TIGs may become ELF-CoPs as they engage with each other more frequently so that certain shared resources become more fixed, other ELF-TIGs in their specific constellation of speakers may only exist in a single speech event over a certain period of time (Pitzl 2018a: 24). Thus, according to Pitzl, "exploring TIGs alongside CoPs can help us gain a more sophisticated understanding of the organizational and social dynamics that influence communication […] in many multilingual, inter/transcultural and/or inherently short-lived or unstable social contexts" (Pitzl 2019: 6).
When considering IS communication, not only Wenger's CoP concept, but also Pitzl's TIG concept may be apt to describe groups of signers using IS in unstable and ephemeral contexts. One example of such a short-lived interaction is the one between the Dutch boy and the Egyptian man as presented in the introduction. Judging from this anecdote, the two interlocutors did not make use of any particular sign language, but gradually created their own endonormative conventions by drawing on different resources they had at their disposal. The two IS users can therefore be considered a TIG, namely an IS-TIG. The formation of this particular IS-TIG was possible because both interlocutors had traveled to the same place in order to visit the international festival Deaf Way II. A concept that has been used to refer to such gatherings is the term 'deaf spaces' (Friedner and Kusters 2015). These spaces are shaped by deaf people in different locations all over the world and allow for the development and enactment of shared cultural knowledge, such as the use of sign language (Kusters and Friedner 2015: xviii).
It can be speculated that such deaf spaces represent an ideal environment for the emergence of different IS-TIGs. Taking an international conference as an example of a deaf space, IS-TIGs may form there as participants ask and are given information at the registration desk, have a conversation at a table during the conference dinner, or exchange views on a topic as they have gathered in front of a poster during a poster presentation. And even though these IS-TIGs take place in a confined deaf space where a specific form of IS is used as an official lingua franca, every TIG is characterized by their own shared norms. These emerge as signers negotiate and establish common resources in the interaction on the basis of their personal multilingual resources, which may contain already conventionalized (IS) signsbut need not. In the next section of this article I will further focus on the process of establishing shared conventions in lingua franca interactions, which has been described in a similar way in both ELF and IS literature.

The establishment of in-group norms and conventions
ELF and IS emerge as speakers and signers, who often cannot rely on common shared knowledge, explore and agree on their own shared norms to fulfill their communicative goals. This process can be particularly well observed in initial IS interactions and especially when signers do not share many sign language resources from the outset. Especially relevant in this respect is the already mentioned MULTISIGN project (Zeshan 2015;Zeshan and Webster 2020), which was aimed at tracking the development of shared norms between signers who had just met. In order to achieve this, the research team not only chose participants who did not share any common sign language, but also deliberately kept its participants isolated from each other from the moment of their arrival to the first filming of their interaction in the research lab (Bradford et al. 2020: 132). Thus, when being instructed to have a spontaneous conversation about a topic of their choice in pairs and later to perform an elicitation game, the participants slowly had to explore common ground and cooperatively co-construct meaning to fulfill the particular tasks. In order to do so, signers drew on their own multilingual-multimodal resources and employed different meaning-making strategies until a so-called shared "multilingual-multimodal space" was created (Zeshan 2015: 227). This shared space, Zeshan explains, "can be thought of metaphorically as a jointly created communicative toolkit" (Zeshan 2015: 213, my emphasis), or, as Bradford et al. (2020: 152-153) prefer, a common "feature pool," which signers gradually expanded throughout their interaction by adding, negotiating and agreeing on linguistic forms and conventions. According to Zeshan (2015: 241), the expansion of the participants' communicative toolkit typically followed a regular "IAP-sequence", where signers introduced new linguistic forms (introduce), which were adopted by the other participants (accommodate) and maintained throughout the remaining course of their interaction (persist). The accommodation phases were particularly well visible in these investigated interactions, given that signers had not yet established any shared resources and so they often identifiably converged with their interlocutors by taking over introduced linguistic forms. By repeating newly introduced signs, IS signers explicitly signaled that they had understood the linguistic form and could continue using it for the remaining course of the interaction (Zeshan 2015: 241). Thus, throughout the course of their first interaction, "[a]s participants become increasingly familiar with each other, the shared space expands and includes more and more communicative resources" (Zeshan 2015: 236).
A similar account of this conventionalization process in intercultural communication was already given by Moody (1987), who, based on his personal experiences in international signing contexts, stated that "[o]nce a new gesture is understood […] it is considered automatically established for use in the group and becomes part of the group's lexicon" (Moody 1987: 82), a development which he later termed "mini-evolution" (Moody 2002: 34). The central focus on the "group's lexicon" is tied to the claim mentioned in the introduction: while many sign languages depict similar grammatical structures that signers can take over from their known sign languages when communicating in intercultural settings (Supalla and Webb 1995), the lexicons of various sign languages have been described as "radically different" (Moody 2002: 33). Consequently, the negotiation of lexical forms, i.e. the establishment of common form-meaning relationships, seems to pose the greatest challenge in IS communication. 9 The MULTISIGN project also predominantly focused on the establishment of (abstract) lexical forms such as numerals (dates, fractions, etc.) (Zeshan 2015), color terms and signs for "male" and "female" (Bradford et al. 2020), whose negotiation was more demanding and hence overtly negotiated by signers. In addition, it has been reported that signers in these studies also developed certain in-group jokes (Bradford 2020a(Bradford , 2020b. For example, one group of signers joked around that the participant who managed to receive the highest score at the elicitation game would get a lollipop as a prize. This joke was presented through an invented iconic sign for "lollipop" and indicates that signers involved in the research activity gradually formed a distinct group identity over the course of some weeks not only through the development of lexical, but also of pragmatic conventions (Bradford 2020b: 275).
Similarly to IS, where signers have to agree on their own in-group norms, also in ELF, speakers may gradually shape their own shared conventions throughout the course of their interaction. This process has for example been discussed with regard to ELF speakers' reliance on Sinclair's (1991) "open-choice principle": instead of drawing on pre-establishedand often non-transparentidiomatic expressions serving as identity markers of a specific ENL speech community, ELF speakers "tend to construct what they have to say more analytically, in a bottom-up fashion" (Seidlhofer 2009: 202). In other words, as ELF interactions mostly take place far away from ENL territorial ground, certain ENL expressions may not only be irrelevant, but even hindering for achieving mutual understanding. Consequently, ELF speakers may establish their own idiomatic expressions, relevant for their momentary interaction. Seidlhofer (2009) calls this process "online-idiomatizing," which not only helps speakers "get a particular job done" (Seidlhofer 2009: 211), but at the same time, similar to what has been observed in some IS contexts (Zeshan 2015), also establishes "a sense of in-group belonging" (Seidlhofer 2009: 209). A similar concept that describes the process of establishing shared conventions in ELF can be found in Hülmbauer's (2014) "common grounding." According to Hülmbauer, ELF speakers establish common ground by negotiating meaning and thus contribute to the establishment of an own shared repertoire, a so-called "situational resource pool" (Hülmbauer 2009(Hülmbauer : 326, 2014. This situational resource pool, which is very similar to the notion of the "communicative toolkit" (Zeshan 2015) and the "feature pool" (Bradford et al. 2020) as described in initial IS interactions, "changes as speaker constellations change" (Hülmbauer 2009: 325) and thus varies depending on the personal backgrounds of the interactants, as well as on the variable duration of their interaction.
Recently, also Pitzl (2016Pitzl ( , 2018aPitzl ( , 2018b has addressed the ad-hoc establishment of in-group norms in ELF, focusing in particular on fleeting ELF interactions, so-called ELF-TIGs (see Section 4). Pitzl (2018bPitzl ( , 2019 approaches this development from a micro-diachronic perspective, i.e. from the moment speakers start interacting to the moment they end their conversation. Based on Hülmbauer's situational resource pool, Pitzl introduces her notion of the 'Multilingual Resource Pool (MRP),' a shared space jointly created by ELF speakers as they share resources from their individual multilingual repertoires (IMRs) with the other interlocutors. The process of expanding this in-group resource pool is represented in Figure 1, which can be found in Pitzl (2018b: 35).  Figure 1 shows the development of a generic ELF interaction at different points in time (T0, T1, T2) between three speakers (S1, S2, S3) from different language backgrounds. T0 marks the moment when ELF speakers enter into communication with each other and their IMRs, consisting of both English resources and resources from other languages, "overlap to a considerable extent" (Pitzl 2016: 298). The amount of shared resources in the ELF speakers' initial MRP at T0 may increase throughout the course of the interaction, which is illustrated in the graph at T1. In this phase of the interaction, an ELF speaker, intentionally or not, shares certain multilingual resources taken from his/her own IMR, which is not contained in the other interlocutors' IMRs. Using Cogo's (2016) terms here, these introduced linguistic forms may either be overtly recognizable as forms of a particular language, through code-switching, or covert multilingual forms (see Section 3). Once the newly introduced language forms are understood by the other speech participants, they get established in the group's MRP, which gradually expands, along with the expansion of the other interlocutors' IMRs, which is represented at T2. Pitzl (2018b) stresses that this process of expanding the group's MRP "lasts as long as the same group of speakers (i.e. this TIG) communicate and interact with each other" (Pitzl 2018b: 36), whereby the phases T1 and T2 continuously alternate until the group stops communicating. From this described process, in which the ELF speakers' IMRs form an MRP, it is thus possible to identify three distinct types of "sharing" multilingual resources corresponding to the distinct phases: (T0) "what is shared," (T1) "what can be shared" and (T2) "what becomes shared" (Pitzl 2018a: 195).
It seems that the ELF researchers mentioned above (Hülmbauer 2009(Hülmbauer , 2014Pitzl 2016Pitzl , 2018bSeidlhofer 2009) have drawn attention to the very same social and linguistic dynamics IS researchers have observed in their investigated groups of signers: the use of different available linguistic resources appropriate to context and purpose of the interaction, the gradual establishment of a group's shared repertoire and along with it, the formation of a particular in-group identity. In particular, Pitzl's (2016Pitzl's ( , 2018b descriptions of the ad-hoc establishment of shared norms in ELF-TIGs bear similarities to Zeshan's (2015) findings obtained from an investigation of initial IS group interactions. In fact, both researchers describe similar phases of the process that takes place when a shared repertoire gets expanded throughout lingua franca conversations: especially noticeable is the analogy between Pitzl's (2018b) What can be shared (T1) and Zeshan's (2015) Introduce (I) and between Pitzl's (2018b) What becomes shared (T2) and Zeshan's (2015) Persist (P), phases, which repeatedly occur as long as ELF and IS users interact. Moreover, both ELF and IS researchers have focused on the development of in-group norms over the course of a limited timeframe, investigating what Pitzl (2018b: 36) terms the "micro-diachronic dimension" of lingua franca interactions.
Based on the conceptual framework proposed by Pitzl (2016Pitzl ( , 2018b for transient ELF interactions, it can therefore be suggested that also those groups of IS signers filmed in the research lab could be classified as IS-TIGs. In this respect, the graphical representation of the development of a situational MRP in an ELF-TIG as shown in Figure 1 (Pitzl 2018b: 35), could likewise be used to illustrate the development of an in-group repertoire in a generic IS-TIG, with the abbreviations S1, S2 and S3 standing not only for speaker, but also for signer.
In that regard, it is probably not too far-fetched to imagine that ELF-TIGs and IS-TIGs could even be investigated within the same theoretical and methodological framework. For that matter, it might be important to consider Pitzl's (2018b) ideas and suggestions on how to systematically track the development of in-group norms in an ELF-TIG and evaluate whether and how this methodological path can be applied for an investigation of IS-TIGs as well. Exploring such a possibility could be a valuable topic for further research and would represent an important step to eventually be able to compare ELF and IS communication in a more systematic way and arrive at a deeper understanding of similar language dynamics at play.

Conclusion: towards a cross-modal collaboration
This article has presented the results of a first comparison between the two lingua franca phenomena English as a lingua franca (ELF) and International Sign (IS), used in intercultural spoken and sign language encounters, respectively. At first glance, these two lingua francas appear to be rather different, given that ELF always involves English as a shared resource, whereas IS is not based on one particular sign language. However, the comparison presented in this article has revealed that ELF and IS researchers have made similar observations when studying their respective lingua franca. The first similarity has been found in the struggles that both ELF and IS researchers have encountered when trying to grasp their lingua francas with traditional linguistic categories. In fact, given that on a formal level ELF and IS display great variability, ELF researchers (e.g. Seidlhofer 2011) have argued that ELF cannot be classified as a variety of English and similarly, IS researchers (e.g. Hansen 2016) have noted that IS cannot be categorized as a language, pidgin or koiné. Scholars in both fields have therefore drawn attention to the shortcomings of traditional linguistic concepts to capture these inherently dynamic and fluid lingua franca phenomena. Tied to this, a topic that has intensively been discussed in both ELF and IS research is the interplay of different multilingual resources. Similarly to IS interactions in which signers create their lingua franca on the basis of their multilingual-multimodal resources (Zeshan 2015), also in ELF overt and covert multilingual resources (Cogo 2016) other than English may at any time be used during the interaction, which renders ELF, just like IS, a multilingual phenomenon. Thus, when entering into lingua franca communication, signers and speakers draw on different kinds of resources at their disposal and shape their own in-group norms accordingly. Tied to this, an investigation into ELF and IS in both Transient International Groups (TIGs) (Pitzl 2018a(Pitzl , 2018b and in Communities of Practice (CoPs) (Wenger 1998) might help grasp the manner in which ELF and IS users interact in both transient and more stable configurations. When considering the establishment of situational in-group norms, it results that ELF and IS researchers have described a similar communication process: whereas ELF speakers expand their group-specific Multilingual Resource Pool by actively sharing resources from their Individual Multilingual Repertoires (Pitzl 2018b), IS signers create a shared multilingual-multimodal space on the basis of their individual multilingual-multimodal resources (Zeshan 2015). Given that these observations draw attention to the same process that leads to the development of shared resources in ELF and IS, it can be speculated that an investigation into how communication unfolds in different ELF and IS groups could be pursued on grounds of the same conceptual and methodological basis.
In sum, the comparison presented in this article has shown that IS signers and ELF speakers make situationally adapted use of different linguistic resources in line with purpose and context of their interaction. In response to the formal difference between ELF and IS, it can thus be said that once we shift our focus away from the product of lingua franca interactionssuch as English forms in ELFand instead look at the communication process lingua franca users engage in, then ELF and IS communication seem to unfold very similarly. Adopting a process-oriented perspective appears to be important for capturing the essence of lingua franca communication, where forms are chosen because they fulfill the purpose in the particular moment and may thus also not be relevant beyond it. This resonates with Widdowson's (2015) claim presented in Section 2 that, in order to understand how ELF communication unfolds, it is important to "rid ourselves" of the idea that ELF can be grasped by spotting linguistic forms that may eventually stabilize and instead embrace the possibility that ELF grants us to observe the "pragmatic process of communication live, in action, laid bare" (Widdowson 2015: 367). However, given that ELF is recognizable as English, it seems that it is easier to "fall back" on adopting a formal approach and find certain tendencies, especially if these forms are not in line with codified norms, and even more so if these divergent forms reoccur in other ELF interactions.
And this is where the great potential of drawing on insights from the other lingua franca research field comes in and which prompts the necessity for a cross-modal collaboration between ELF and IS researchers. In fact, it seems that in many IS contexts, particularly in initial encounters, "the pragmatic process of communication live, in action, laid bare" (Widdowson 2015: 367) can be discerned more easily. In other words, intercultural signed communication can reveal in a more "unfiltered" way how forms are shaped and used to fulfill certain functions in a specific context, regardless of whether these forms have already been attested or are in line with certain conventionalized language norms. ELF researchers would therefore benefit from the study of IS to strengthen their argument that ELF is a flexible and adaptive phenomenon that has to be conceptualized and studied "in its own right" (Seidlhofer 2011: 171), i.e. independently from ENL norms. Conversely, IS researchers might benefit from the study of ELF, especially from the various critical conceptual discussions which have shaped ELF research in the last decades. For example, Widdowson's (1997) virtual language could be of interest to IS researchers, especially if ASL continues to gain dominance in international sign language encounters (Kusters , 2021.
On the basis of the above-mentioned considerations, it is high time that ELF and IS researchers become aware that their respective object of study might have an equivalent in the other language modality and explore the possibility of carrying out interdisciplinary projects by tackling the same research questions through the coordinated use of the same research methods. Such a cross-modal collaboration would strengthen both research fields, as each would obtain new insights from the other language modality and, more importantly, it could help arrive at a deeper understanding of the general characteristics of lingua franca communication.