Lingua franca communication research
Lingua franca communication differs from other forms of intercultural communication such as native/non-native communication and communication via a professional or non-professional interpreter. Participants in lingua franca conversations are representatives of their individual mother cultures. Hence, they have their individual cultural backgrounds regarding communicative norms and standards. We will, therefore, expect interferences from the different mother tongues. At the same time, speakers have to a certain degree acquired the norms of either British (BrE) or American English (AmE) when learning the language. Thus, at least three but sometimes even more cultures are involved in lingua franca communication. Unless the speakers are familiar with the others' mother tongues, the amount of different cultures interacting in these situations demands that speakers cope with the unexpected, by having to apply imperfect knowledge of and competence in the language they use (cf. Knapp 1991). The resulting level of insecurity experienced by the participants has the effect of making them establish a unique set of rules for interaction that may be referred to as an inter-culture, a "culture constructed in cultural contact" (cf. Koole and tenThije 1994) or as a lingua franca culture, which is reflected in specific linguistic characteristics. These characteristics are apparently not influences of the speakers' mother tongues and will be discussed at a later point in this paper. At the same time, the speakers must, in most cases, be regarded as learners of the language they use as a lingua franca. Their communicative behaviour is not only a reflection of cultural norms, but it also represents the individual stages of their interlanguage5with its specific characteristics as well as the results of adaptation to the interlocutors.
Lingua franca or non-native/non-native communication has basically been studied from two different perspectives. Firth (1990 and 1996) and Gramkow Andersen (1993) analysed business telephone conversations between speakers of different European mother tongues, taking an interactional approach and focussing on the way participants cooperate to achieve the goal of their conversation. Others have approached lingua franca conversations as interactions between learners. Schwartz (1980) and Varonis/Gass (1995) investigated the negotiation of meaning between non-native speakers of English with different linguistic backgrounds. Yule (1990) studied the management of verbal conflict among Indian, Chinese and Korean students interacting in lingua franca English. Meeuwis (1994) and Meierkord (1996 and 1998) provide analyses of the discourse features of lingua franca small talk.
The above-mentioned studies yield interesting results and offer important insights. As a basic finding, they all stress the cooperative nature of lingua franca communication. Being in clear contrast to the findings of e.g. Thomas (1983) who emphasized the pragmatic problems encountered by non-native speakers when interacting with native speakers of English, cooperation among non-native speakers manifests itself e.g. in collaborative overlap and joint construction of what is usually called a turn. If our aim is to arrive at valid generalizations regarding the different varieties of lingua franca usage, quantitative analyses of larger copora are needed. A frame of reference, which is capable of dealing with the data's heterogeneity will need to be applied to the data. Categories useful for the analysis of lingua franca talk-in-interaction have been proposed by both Conversation Analysis and Discourse Analysis. However, as neither of these models have been designed to fit lingua franca data, modifications seem to be necessary and will be suggested below. Apart from modified tools for analysis, lingua franca data also requires a differenciated interpretation of the results produced by data analyses, taking into account both the intercultural as well as interlanguage aspects of this variety of English.
The following discussions are based on tape-recorded, naturally occurring, face-to-face group conversations. The data was collected in a student hall of residence for overseas students in Great Britain and comprise 23 conversations of a total of 13.5 hours. The speakers participating in the conversations are aged roughly between 20 and 30. They are of both sexes, speak 17 different mother tongues and include both less competent and more competent speakers. The corpus, thus, is very heterogeneous, but is, nevertheless, representative of the situations which involve lingua franca communication.
Data analysis in lingua franca communication research
Lingua franca communication implies the mingling of different cultures and the associated communicative norms that apply within these cultures. Discourse produced in lingua franca English has its specific characteristics, and these make it difficult to apply existing categories proposed by Discourse and Conversation Analysis (CA), which both had originally been developed for interactions between native English and American speakers.
Below a short account of the most central unit of analysis in CA, the turn, will be given. This unit requires an investigation into the applicability of the model to lingua franca talk-in-interaction and to a discussion of necessary modifications.
The concept of turn
Conversation Analysis developed from approaches by a number of American sociologists in reaction to the quantitative methodology applied in their field. Sacks, Schegloff and Jefferson (1974) applied ethnomethodological methods to spoken conversation. Analysing a corpus of informal spoken discourse they arrived at the conclusion that turn-taking is the essential characteristic that distinguishes conversations from monologic speech. Rules which seem to govern the turn-taking process were identified together with transition-relevance-places at which speaker change was found to occur, but the central concept, that of the 'turn', has remained only vaguely defined. It is the way simultaneous speech and pauses have been included in these definitions, that is of interest when we are dealing with lingua franca discourse, because both occur in a specific way in lingua franca discourse, as will be explained below.
Schegloff (1968) claimed that participants orient themselves towards the rule 'one party at a time'. Any violations of this rule would be classified as 'noticed events' by the participants of a conversation, and that these violations would result in the application of repair mechanisms. Similarly, overlap was characterised by Sacks et al. (1974) as a turn-taking error and hence as being a violation of turn taking requiring repair. Oreström (1983) admits that "speaker-shift is seldom, if ever, an entirely smooth process and the interactants generally try to see to it that the transitions from one speaker to another take place in a non-abrupt manner and they therefore try to avoid simultaneous speech and interruption."
What is central to these early statements is the fact that overlapping speech is regarded as being erroneous and a violation of some rule. Even though this argument is still prevalent in many recent discussions of the term 'turn', the existence of unproblematic overlap has also been considered. McCarthy (1991), for example, states that "speakers predict one another's utterances and often complete them for them, or overlap with them as they complete", and Langford (1994) interprets this kind of overlap as displaying "close attention and support". Yule (1996) adds a further aspect – the collaborative use of overlap:
For many (often younger) speakers, overlapped talk appears to function like an expression of solidarity or closeness in expressing similar opinions or values. [...], the effect of the overlapping talk creates a feeling of two voices collaborating as one, in harmony.
Even though the authors acknowledge cooperative overlap, they do not refer to it as being used to jointly build up a collaborative turn. Immediately related to the concept of turn is the distinction of participants' roles into speaker and hearer, which assigns to the hearer only those passive activities which support the speaker. Schegloff, however, has recently (1996) claimed that participants jointly create talk-in-interaction, and as a result he labels them co-participants. As I shall demonstrate further below, this must necessarily lead towards a re-definition of turn as a jointly completed unit of conversation, which will also have to include a discussion of the term back-channel, i.e. those utterances that are usually being considered to be the hearer's contributions.