The stranger was now standing near to them, – almost so near that he might hear their words. Burgo, perceiving this, walked up to him, and, speaking in bad French, desired him to leave them. “Don't you see that I have a friend with me?”
“Oh! a friend,” said the man, answering in bad English. “Perhaps de friend can advance moneys?”
“Misther, Misther!” said the man in a whisper.
“What do you want of me?” asked Mr Palliser, in French.
Then the man spoke in French, also. “Has he got any money? Have you given him any money?” (Trollope 1938 : 76, 454, 457)
1When the feckless Burgo Fitzgerald and the exemplary Plantagenet Palliser went to Baden in the 1850s (in Anthony Trollope’s novel Can you forgive her?), they did not attempt to speak German with a ‘stranger’. What happened is related in the epigraph. There is a given, conventionalised lingua franca, used for speaking to strangers, in this case French. Speakers who appear not to know this language well will be addressed in what appears to be their first language if possible, but as long as both sides show that they know the lingua franca, it is the preferred option. Lingua franca use has always needed negotiation and adaptation and each situation is different.
2At present, English is most often used as a default in this way all over the world. This development evokes a variety of reactions, one of which is expressed by Jean-François Dehecq, the managing director of the Franco-German pharmaceutical Sanofi-Aventis group as follows:
Dans une réunion, c’est du cerveau des gens dont on a besoin. Si vous les obligez à parler anglais, les Anglo-Saxons arrivent avec 100 % de leurs capacités, les gens qui parlent très bien, avec 50 %, et la majorité, avec 10 %. À vouloir tous être anglo-saxons, il ne faut pas s’étonner que ce soient les Anglo-saxons qui gagnent. (Barbotin 2004)
3Dehecq focuses on the faster processing one enjoys in one’s first language. His view of languages is that they have owners – their native speakers – and that their owners use them much more easily and skilfully than outsiders. Using a foreign language takes brain capacity and slows reactions to an unacceptable degree. This is one of the attitudes that underlie arguments that the spread of English is linguistic imperialism, imposing disadvantage on speakers of other languages. The argument is that an accident of birth gives an insuperable advantage.
4An alternative emphasis comes from advocates of BELF, Business English as a Lingua Franca, that is English used for business among speakers for most of whom it is not a first language:
On the whole, our survey respondents did not feel that they were more successful in their communication with NSs than NNSs. Neither did they feel that NS-like pronunciation was an essential element in effective communication. Similarly, the interviewees hardly ever associated English with any specific native speaker model or with a national culture or its values such as the United Kingdom, the United States, or Australia; rather, the majority saw it as global and neutral […]. [What matters is] getting the facts right, making the discourse clear, and “making the recipient feel good.” (Kankaanranta & Planken 2010: 395-396)
5Kankaanranta and Planken emphasise the effectiveness and relative neutrality of lingua franca communication. They put forward a view in which English is a common good, open to everyone, and presuppose that non-natives can have fully adequate mastery. This is a classic liberal argument: what is important is that the resource should be equally available to everyone, no matter whether everyone can equally make use of it. It recalls the remark ascribed to Sir James Mathew (1830-1908): “In England, justice is open to all – like the Ritz Hotel”. The point of the remark is that you have to have money (or linguistic capital, in the present case) to make use of the opportunities offered by a liberal system, but it also makes the point that in such a system people are not excluded for other reasons, like being Roma or Jewish. No accident of birth restricts one in the liberal vision of lingua franca use, only the linguistic ability that is differentially easy to acquire for different people.
6The two quotations do not of course contradict one another, but place the emphasis differently. Dehecq feels that he is at a disadvantage communicating with natives in English; Kankaanranta and Planken’s informants do not feel that it is more difficult to communicate with non-natives than with natives. The two positions seem incommensurable and the argument to be put forward here is that both are inadequate because they essentialise the native-non-native distinction and make generalisations about situations that are not comparable. The situations in which English is used across language barriers are very diverse and so are the participants, so both blanket condemnations of the spread of English as linguistic imperialism, or celebrations of it as a brave new centreless world of English as a lingua franca are inappropriate.
7After more than a decade of discussion of English as a lingua franca (ELF), a consensus seems to have been reached that ELF is not a variety like American or Indian English, or even ‘French English’ with shared code features, but rather a set of pragmatic practices and attitudes to accuracy that make communication in non-native English effective. “Any situation where people who have different L1s get together and communicate with each other through English is an ELF situation.” (Björkman 2011).1 There are of course deviances from native English that are shared by many non-natives, but in principle each speaker in an ELF situation is using their own variety (their interlanguage, to use a term rejected by most writers on ELF) for effective communication with speakers of other varieties.
8Since it is defined in terms of the proportion of non-native-speaking participants in discourse in English, the ELF notion depends on the idea of a native speaker of a language. Despite the vast controversy over the term, I will define native-speaker status as continuous oral use of the language in question since acquisition in early infancy. This is a definition that allows for someone being a native speaker of two languages but uses “non-natives” for everyone else including people who have learnt through the medium of English throughout their schooling but have another home language. The reason for using this definition is that it seems to be what underlies assertions that as high a proportion of interaction in English as 80% (Kankaanranta 2008) is among non-natives, and thus to be part of the rationale for saying that English is not predominantly used by its native speakers. It would be hard to achieve such a figure if interactions among Anglophone Asians and Africans were excluded. If lingua franca situations are situations in which non-natives use English with one another, then inter-Indian interaction, for example, is an instance of English as a lingua franca. It does not seem to be prototypical, however, nor do the participants seem prototypical non-natives. One could say of such interaction what Kankaanranta and Planken (2010) say of BELF, but it is hard to see why it should be different in any qualitative way from interactions among native speakers, so long as we accept the conventional position that the English of proficient Indians is a stable variety like British or North American.
9It is argued here that ELF situations like those described and advocated by Kankaanranta and Planken and foreign-language situations like that described and deplored by Dehecq are of several different kinds, each of which requires different types of adaptation of participants. Hence a more differentiated classification than one based merely on native-speakerhood is needed. Three parameters describing the speakers as individuals can be considered: English linguistic proficiency, communicative ability viewed as transferable across languages, and personal language repertoire.
10English proficiency (the linguistic component of communicative competence, Canale & Swain 1980, Bachman 1990) can be regarded as composed of accuracy and fluency. Accuracy is normally defined as native-like use of forms, and one of the innovations associated with the ELF notion is to say that it is not particularly important so long as communication is unambiguous. What Trudgill (2002: 92) calls ‘afunctional grammatical categories’, like verb agreement, or the fine details of count and noncount nouns in English, do not matter much in production, it is argued (for example Louhiala-Salminen et al. 2005, Ehrenreich 2010, Mauranen et al. 2010), if it is not one’s intention to be assimilated to a native-speaker community. Functional but obscure distinctions inferable from context such as the difference between pick the fruit (harvest from the tree), pick off the (rotten) fruit (from the tree and discard),and pick up the fruit (from the ground) cannot be important either, because even if speakers know them, they cannot be sure that the interlocutors do, and if those who know the distinction hear a form in use, they cannot be sure that the speaker intends the meaning natives would ascribe to it. Similarly, the types of discoursal norms due to mother-tongue grammatical habits investigated by Lambert (2006) would be irrelevant in such a situation.
11Fluency is not much discussed in ELF circles, but seems to be a more important element in proficiency when carrying out complex tasks. It is related to automaticity of lexical access and parsing (Paradis 2009, Anderson 1983). The better automatised such processing is, the less mental capacity it takes. Highly proficient speakers and those who have learnt the language young – and therefore native speakers – have better automatised processing (Paradis 2009). This accounts for Dehecq’s view that only half his brain power is available in English, not that he may be at a loss when choosing the right tense. It is frequently observed that writing and reading are slower in a second language, even in one that is very well known (Flowerdew 1999, Shaw & McMillion 2008). Kankaanranta and Planken (2010) observe that “In written genres […] the majority of the interviewees noted that […] there was more time to contemplate how to write something down or what a word or an expression in writing meant.” (Kankaanranta & Planken (2010: 388) The same is true of speaking; recent work has shown that highly proficient second language users speak 20% more slowly in L2 English than in L1 Danish (Thøgersen & Airey 2011).The time spent choosing the right word, Dehecq would argue, would be better spent more productively.
12Communicative ability (textual, strategic and rhetorical “competence”) in general and in a particular field involves capacities like structuring texts and utterances appropriately, judging the audience’s level of understanding and adjusting to it, using appropriate rhetorical strategies for the given audience and situation, and many other capacities arising from experience, training, and character. Probably particularly relevant in the present context is the sort of mindfulness advocated by Gudykunst (1983) for intercultural communication. Above a certain proficiency threshold some of these capacities, and the underlying ability to acquire them, are transferable from one language to another (Shaw 1991, Cumming 1989 for writing, Bernhardt 2005 for reading; Ehrenreich 2010 implies this for spoken skills). This type of ability varies widely amongst individuals and it is often argued that multilingual and intercultural experience, for example in using English in lingua franca situations, contributes to it. Ehrenreich (2010) finds that business users of English as a lingua franca generally value communicative ability much higher than English proficiency (above a quite high threshold, obviously), and it is this position that underlies most arguments that lingua franca communication is unproblematic despite deficiencies in accuracy.
13The speaker’s individual language repertoire is, I suggest, also an important element in understanding lingua franca and foreign language interaction. Two aspects can be considered. One is the availability of what Cook (1995) calls multi-competence – simply knowing more than one language, with its accompanying cognitive and communicative-proficiency advantages. In the type of situations under discussion, this distinguishes monolingual Anglophones from all other participants including of course multilingual Anglophones with English as their first language. The other is the availability of a stronger language for the task in hand. Some speakers almost always do a given task in a given language. Americans and other Anglos, and Indians, Filipinos, Nigerians and other post-colonial English users, are used to performing most business tasks only in English. Therefore, whatever the quality of their performance, they ascribe it to their communicative ability not to their linguistic proficiency in English. (They do this, I would argue, even if in fact their communication is impaired by the quality of their English, because they have no point of comparison). They are not less confident or fluent than in other situations they have experienced in the same domain. Speakers of other major languages, by contrast, are used to performing nearly all business tasks in L1, and when they have to perform one in L2, they are struck by how much more difficult it is than it would be in L1. They ascribe this difficulty to lack of fluency or accuracy in English, (even if in fact their communication is – also – impaired by poor communication skills). Even experienced users say that “English-medium interactions are […] more tiring and also tak[e] longer than those in one’s mother tongue, […]” (Ehrenreich 2010: 421) and the reason why they can say this is that they have experience of performing the task in L1.
14Speakers of small languages like Finnish, Swedish, or Dutch are in an intermediate position. They have more and more experience of performing some tasks in English, and in fact perform some of these tasks more often in English than in their own language. The frequency of performance improves fluency, but a separate point is that limited experience of using L1 for certain tasks and extensive experience of L2 means that the L2 experience is normalised and less disturbing. Dehecq compares negotiating in English with negotiating in his first language, French, and naturally notices that he is less fluent in English. Kankaanranta and Planken’s informants (2010) are predominantly Finnish and therefore have no actual experience of international negotiations in their first language; they have grown used to doing them in English.
15The individual characteristics just discussed interact with the social characteristics of the situations and participants to produce a rich diversity of types of interaction which challenge simple categorisations. In the rest of this article I shall attempt to describe some communication situations from the literature and show how they differ from one another in these various dimensions.
16The first group of cases are devoted to confirming that many speakers of English from post-colonial environments are only technically non-native, and hence that their interaction does not raise the issues of varied competence associated with ELF and foreign-language interaction. For all practical purposes, they have appropriated English and their relation to the language is the same as a native speaker’s. I could cite the very popular Nigerian films in English addressed to Nigerian and African audiences with various mother tongues. While the communication here seems to meet the definition of English in a lingua franca situation, it does not seem to differ in any significant way from that in American films addressed to Anglo audiences. Another example of appropriation is the rise of outsourced call centres in the Philippines and India. Lockwood et al. (2008) found that Filipino call-centre workers typically have excellent English and a good knowledge of the culture and environment of their American clients but (of course) do not ‘sound like Americans’ – they sound like internationally comprehensible speakers of Filipino English, and are accepted by their clients as such. Lockwood et al. found the main weakness to be in listening comprehension, but we do not know whether this is actually a second-language problem or whether similar weakness would be found in ‘native’ call centre operatives.
17Beyene et al. (2009) carried out a study of communication among software production teams in a German multinational with English as the company language. The teams were based in, and staffed from the US, Germany and India. The authors observe:
As expected, we saw few language challenges in the India-US teams. Although there were occasional issues with understanding accents and word choices, there were relatively few language related tensions in these teams. (2009: 28)
18The Americans and Indians were alike linguistically in several ways (even though it might be claimed that the Indians were “culturally” more different from the Americans than the Germans). Both groups had had extensive exposure to English since childhood, leading to high proficiency, language confidence, and fluency. Neither group had had experience of doing software production in a language in which they were more confident than English, so they had no sense of discomfort or constraint. In this context, neither had the option of reversion to another language, since the language they are likely all to speak well is English. This example again shows that ‘Outer-Circle’ (Kachru 1983) elite members are indistinguishable from native speakers within domains where they are English-dominant. It is relevant that Ehrenreich’s informants (2010) identified native speaker English and Indian English as among the most difficult varieties to understand. That is, speakers of stable varieties with well-established conventions speak faster, because they have more automatised processing, and subtler agreed conventions of meaning. (Of course it is also true that an unfamiliar accent is difficult to understand, which may make Indian or Scottish speakers particularly difficult to understand).
19The Beyene et al. (2009) study also looked at another type of situation with a potential language barrier. They examined teams with German as L1, and they observe that “Germany-US and Germany-India teams reported language challenges at about the same rate”. There were fairly frequent occasions where the Germans gave up the struggle with English and reverted to L1. This was perceived as ‘rude’ or sinister by the Indians and Americans, who were aware that power ultimately lay with headquarters in Germany. Unlike the other two groups, the Germans had an alternative language in which they felt more comfortable, and they were unhappy about having to use their weaker language. For the Indians and Americans, however, German constituted a ‘secret language’ in which they suspected plots were being formulated. The Germans were a minority in an English-dominant majority, and the evidence is that in these types of context the confident users dominate and even impose norms about the use of other languages. Different problems arise where linguistic and institutional powers are aligned and, as in this case, where they are in conflict.
20A case in which linguistic power was not counter-balanced by institutional power is reported by Conoscenti (2004), who analysed the interaction in a two-month international diplomatic simulation using English as a medium of communication. The interaction was dominated throughout, and increasingly as it progressed, in message frequency, length of messages and topic control by the 50% of participants who were native speakers of English. The power differential and a focus on ‘language ownership’ of the sort experienced by Dehecq was also suggested by the presence of compliments from mother-tongue English users to non-natives on the quality of their English.
21These examples show that where English is a foreign language for some participants and not for many others, then, inequality and tension seem likely. Proponents of the view that English is an effective lingua franca regard these as exceptional cases and basically think the confident users are misusing their power (thus largely agreeing with Dehecq). They anticipate a more egalitarian environment where everyone uses English but almost everyone has another mother tongue – the lingua franca situation.
- 2 Although Bahasa Malaysia is the dominant language, personal communications from Malaysian colleagu (...)
22A striking example is that of the Association of South East Asian Nations (ASEAN). Of the ten member states, four have English as an everyday element in national life: Malaysia,2 Singapore, Brunei and the Philippines, and six function intranationally mainly in the national language, though three – Vietnam, Laos and Cambodia – are members of the Organisation internationale de la Francophonie.
23The language paragraph (Article 34) of the ASEAN charter simply reads ‘The working language of ASEAN shall be English’. ASEAN delegates are thus all technically non-natives who have adopted English as a neutral lingua franca. ASEAN is often presented as a contrast to the EU in that English is not seen as British or American and its status is accepted in a matter-of-fact way. Kirkpatrick (2008) shows that speakers of nativised varieties deliberately avoid localisms and aim at an internationally comprehensible environment, thus not imposing their norms as technical native speakers might. But personal communications, from Lao people close to the centre of power in particular, suggest that the poorer non-Anglophone countries struggle to find delegates linguistically able to cope. The examples presented by Kirkpatrick (2008) also show more successful communication and possibly more active participation by more fluent delegates. One might note that this is not necessarily a matter of official national language policy: the richer non-Anglophone countries have elite groups whose members may well have been educated through English and who in any case are increasingly educating their children through that medium, ensuring confidence and fluency, if not familiarity with the particular task. Participants are not really taking part on equal terms.
24In the terms proposed here, ASEAN is not a true lingua franca situation because delegates from four of its members have the advantages in fluency and confidence of those whose everyday language for professional purposes is English, alone or alongside another language, while many of the others are aware that they could perform the task better in L1. Though this second group doubtless gradually becomes more confident and fluent, it cannot really be regarded as taking part in the interaction on equal terms.
25In a true lingua franca situation all participants are using a language which is not the one they would use every day for comparable tasks. But not all such situations result in effective egalitarian interaction, even where language proficiency is appropriate. At the borders of Fortress Europe, immigration officials have the thankless task of preventing the wretched of the earth from entering the EU, as described by Guido (2004). In this extract, Italian social welfare officers are interviewing Nigerian would-be immigrants. The Italians have effective school English, the Nigerians are fluent in a pidgin-influenced variety. Guido quotes interactions like this:
A: Who helped you to escape from Nigeria?
B: The car come pick us and I bin drive for Niger hhh the awa yanish them em bin break hhh for the uranium mine dem hhh for one year
A: Did they did they make you to work in the mines who
B: Yeah hh the mine them em bin give the money for go awa away for Agadez... Here one truck come drive for the desert... After two days the sand bin make us >wakawaka< for the sun, make Libya border no see os
A: The border, eh? Hhh You had no documents, eh? (2004:133)
26 This is a lingua franca situation, and the participants have reasonable proficiency, but there is no effective communication. The Italians want to know who is organising the human trafficking, the Nigerians want to express their appalling experiences. The stakes are very high here, so both sides really want successful communication. Nevertheless, the interaction is unsuccessful because the participants have very different purposes, and their backgrounds of experience and education are too far apart. The differences between school English and pidgin make things worse, as Guido notes.
27One type of situation in which English as a lingua franca seems to be as effective and egalitarian as one would hope is in low-stakes social interaction, that is interaction where the intended outcome is merely pleasant social contact, and nothing much depends on success in such interaction for the interactants. Meierkord (2000) reports successful social interaction in student-accommodation kitchens among speakers with very different language backgrounds and competence.
28Exchange student communities have always had difficulty in forming relations with local students (Ehrenreich 2008), so being an exchange student has often meant living in a community of foreigners. Under EU language policies one would expect Erasmus student exchanges within Europe to aim to encourage plurilingualism, but when the best language of nearly everyone in this community of foreigners is English, a natural situation arises for the spread of English as a lingua franca. Anecdotal evidence suggests that this can happen even where the local language is instrumentally desirable, as in Germany, France or Spain, but it is most common in small-language universities. Our study (Shaw et al. 2009) of communication in English among exchange students in Sweden and Denmark showed a great deal of interaction across L1 groups within a ‘lingua franca-English bubble’ with little contact with the Swedish-speaking environment. Lingua franca English seems very successful as a medium in this environment and our study elicited entertaining responses like this from a Spanish exchange student after five months at Stockholm University:
Are there any people that you avoid communicating with?
Yes, Americans... and also Australians. They are very friendly but I can't understand them. (2009: 192)
29An English-educated native speaker of French Creole from Mauritius reported after a year that her English had “got worse”. What she meant was that speaking English to non-natives made her avoid expressions which she had observed to cause difficulties, and thus to use simplersyntax and less rich vocabulary. Both quotations reflect the way that lingua franca ‘communities’ develop their own, possibly somewhat simplified, speech patterns, to which English-dominant members have to adapt (like the Mauritian), or sacrifice some opportunities (like the Americans and Australians). The successful lingua franca English discussed here is low-stakes social interaction in which topics can be, and often are, abandoned if they are not generating meaningful interaction. The participants have similar ages and interests and they have time to get used to each other over the five months of their Erasmus visit abroad.
30The time element, the development of an interacting community, seems important, but James (2000) gives an example of successful social interaction that is not based on long-term familiarisation. He reports this ‘Austrian/Italian/Slovenian conversation’ from Vienna railway station:
A: I don wanna drink alcohol
B: Me too
C: I also not (2000: 22)
31 Although this is undoubtedly successful interaction in a lingua franca situation the speakers’ proficiency appears to be so low that it restricts the level and depth of communication. This is not what advocates of viewing English as a lingua franca envisage.
32The environments that are described by Kankaanranta and Planken and other proponents of ELF can be called high-stakes transactional (Brown & Yule 1983), in that they involve interaction to achieve business or academic agreement on substantive issues and success is of genuine importance to participants in terms of career or grades or other types of prestige. A community of exchange and other international students at a technical university in Stockholm was examined by Björkman (2010, 2011) in the high-stakes transactional environment of group work on engineering problems in class, and she found that interaction was effective and achieved its goals. It was also appropriate, in that communication difficulties did not lead to topic abandonment, which would have been inappropriate where the purpose was not ‘interactional’ socialisation but ‘transactional’ achievement of a shared goal. Tange and Lauring (2009) report on a Danish company which adopted English as company language and encountered little difficulty with transactional tasks. They quote:
Because of the diversity you focus more on the professional. You don’t think about from where people come but only whether they contribute their best no matter how they feel among themselves. You don’t focus on people’s mindset but on the result. Whether people get on socially or not is unimportant. (2009: 226-7)
33This seems to reflect the same effective goal-oriented communication that Björkman observed. It corresponds to Ehrenreich’s observation (2010) that lingua franca English belongs in a community of practice (Wenger 1989), a group held together by shared goals and shared practices.
34Social interaction can be instrumentally important too. Tange and Lauring actually argue that company-level communication in English may be professional but it is thin because it lacks the social dimension. This is expressed by the continuation of the quotation above, which is that speaking in L1 “allows you to draw on a broader set of keys” (2009: 227). Everyone has the experience that socialising is an area in which things function better in L1), and where ‘culture’ is important, and consequently everyone notices that they are not putting their best social self forward in a second-language situation. The exchange students in Shaw et al. (2009) had struck up many friendships across language boundaries, but they were young and could choose their friends. Socialising in a business context such as described by Tange and Lauring is a matter of getting on with colleagues or partners one has not chosen, and failure to get on with them has serious consequences: this is quite high-stakes interaction, although it is socialising. Tange and Lauring found that the introduction of English as the corporate language in the Danish-dominated company they examined resulted in the formation on a social level of language clusters and the reduction of interaction in English to the professional level. They show that:
[…] when people withdraw from gossip, small-talk and story-telling sessions, they abstain from an important social practice, which contributes to the maintenance of social norms and roles within the organisation […] (2009: 227)
35It seems as though this kind of high-stakes socialising is not easy even in a lingua franca situation. Communication in lingua franca groups that have a practical purpose is likely to be rather restricted to that purpose: socialisation in lingua franca goes best when the partners are self-chosen and creating good intragroup relations is not an ulterior aim.
36Or is the reason rather the numerical predominance of one language group? One of Tange and Lauring’s informants said
[…] they cannot say that it is an international company and that English is the company language. It is just a Danish company with a lot of foreigners. (2009: 225)
37Even when the majority of informants have non-English-dominant multicompetence, then smooth communication at a social level requires no one language to be dominant as L1.
38Many of the successful English-as-a-lingua franca situations described in the literature have in common that participants have benefited from time to adapt and develop a code and pragmatics that function effectively. Either the participants have often performed these tasks, albeit with different interlocutors, as in the Kankaanranta & Planken’s (2010) and Ehrenreich’s (2010) studies, or they have formed a smaller more tight-knit type of community of practice like the international students of hotel management described by Smit (2009), who learnt to understand one another over a period of months despite initial difficulties with accent. As Kankaanranta and Planken note,
Overall, three contextual factors emerged as relevant in the interviewees’ conception of BELF discourse: the shared business domain of BELF use, the shared special field of expertise, and the length of relationship with the communication partner. (2010: 390)
39So I would conclude that in a situation where one or more participants are functioning in an L2 the success of the interaction, at least in terms of the comfort of the interactants, depends on the following factors, given in the order they were raised in the text above:
Linguistic proficiency both in terms of accuracy and fluency;
Level of communicative skill (in the domain in question);
Participants’ language repertoire (for the domain in question) and in particular the place of the language used in their fluency ranking for the domain;
Shared goals and purposes for the interaction;
The nature of the interaction as primarily transactional (favours lingua franca use) or interactional (only favours lingua franca use where it is purely social and not oiling transactions);
The balance of participants with different linguistic proficiency in English;
The extent of their joint membership of a community of practice;
The balance of participants with different L1s;
The time participants have had to become accustomed to one another.
40Interaction in same-language situations without language differences can be ineffective or dysfunctional, so it is no surprise that the same can happen in lingua franca situations. The point is perhaps that in lingua franca interaction the exact forms and meanings of the language used are less predictable, so interaction is a little fragile, and requires rather better conditions for success than interaction where the code used is equally shared. It is worth noting that in foreign language interactions, where outsiders are expected to use the cultural and linguistic norms of the majority, there is an illusion of predictability; L1 users suppose that the utterances of the L2 users mean what they would mean from L1 users, and misunderstanding may arise. In well-functioning lingua franca situations participants are on the lookout for the unexpected, and are likely to develop strategies for dealing with them. To use Donald Rumsfeld’s terms, the lingua franca users are dealing with known unknowns, while in a foreign language interaction there are, at least for the L1 users, many unknown unknowns.
41It is clear that a Danish company that becomes international has little choice other than to adopt English as company language, or as Ehrenreich’s informants say (2010), the only realistic option for Germans talking to Japanese people is some form of English. Widespread use of English is an inescapable fact in the present age but, as Deneire (1998, 2011) reminds us, it is also an issue of power and politics. The world is full of inescapable facts, like climate change and national indebtedness, for example, and what is important is to understand the phenomena and devise ways of ameliorating the problems they bring. The list above suggests a way of deciding where intervention and management are required and which interactions can safely be left to market forces, and hence to English, the dominant lingua franca for the moment.
I am grateful to the Editors of ASp, Marc Deneire and Beyza Björkman for generous comments which have greatly improved the article.