The use of ‘I mean’ as a discourse particle in ELF business communication

(1)

The use of ‘I mean’ as a discourse particle in

ELF business communication

Bruno Lewis

s2243490

Faculty of Humanities, Master Linguistics Program

Leiden University, The Netherlands 2020

Supervisor : Dr. Felix Ameka Second reader : Dr. Maarten Kossman

(2)

(3)

iii

Acknowledgements

I am greatly thankful to my supervisor, Dr. Felix Ameka, not only for his constant availability, advice and good humour throughout this entire peculiar process, but in a much broader sense for having taught me immensely since the beginning of my master studies.

I want to deeply thank to Natacha Boric for her endless empathy and patience, and for sharing her knowledge with me every time I needed to discuss my ideas and doubts.

Last but never least, I owe this achievement to my parents, for their love, even across the distance, keeps driving me to fulfil my dreams.

(4)

iv

Abstract

Linguistic literature addressing the roles of discourse particles in English Lingua Franca (ELF) settings currently remains scarce. Research has shown that ELF communication presents different rules and structures than native English conversations. This study aims to explore the use of the particle I mean in ELF business communication. Building on general work in the subject, it asks two main questions: What are the functions of I mean as a DP used by ELF speakers in business settings? To what extent is I mean employed to pre-empt communicative misunderstanding? Based on the review of previous findings and the analysis of empirical data from a corpus of ELF business meetings, this thesis demonstrates the multifunctionality of I mean executing 15 functions. The results indicate that speakers in such contexts stay engaged in discourse, anticipating any potential source of misunderstanding. On that basis, it is recommended that further research should concentrate in the study of the influence situations involving hierarchical structures like business meetings have on the behaviour of I mean.

(5)

v

List of Tables

Table 3. 1. Domains of recorded interactions (VOICE, 2013) ... 20

Table 3. 2. Speech event types within PB (VOICE, 2013) ... 20

Table 4. 1. Frequency of I mean by meeting ... 23

Table 4. 2. Positions of I mean by meeting ... 23

Table 4. 3. I mean instances with pauses ... 24

Table 4. 4. Co-occurrences with other DPs ... 25

Table 4. 5. Co-occurrences by positions and pauses of I mean ... 26

Table 4. 6. Functions of I mean by positions and pauses ... 28

List of Figures

Figure 4. 1. Frequency of I mean by gender per meeting ... 25

Figure 4. 2. Distribution of co-occurrences per I mean function ... 38

(8)

viii

List of abbreviations

CL Corpus Linguistics

DM Discourse Marker

DP Discourse Particle

ELF English as a Lingua Franca

ELFA English as a Lingua Franca in Academic Settings IC Intercultural Communication

Mtg Meeting

PB Professional Business PDE Present-day English

PM Pragmatic Marker

(9)

ix

Transcription notations

Not all mark-up conventions are cited in this document. In order to facilitate the reading, I considered as the simplest solution to share the official mark-ups from the corpus I used. The readers can find the totality of notations in the following link: VOICE Transcription Conventions.

e:r lengthened sound

(.) brief pause

(1) (2) longer pause, timed in seconds

@ laughter

<1>words</1> overlapping speech <un>xxx</un> unintelligible speech

{words} translation of non-English speech

(word) word fragments, words or phrases which cannot

be reliably identified

CAPS emphasis of a syllable or a word

= continuation/completion of speech by another

speaker without pauses

. follows words spoken with falling intonation

<fast> words </fast> fast voice <soft> words </soft> soft voice <slow> words </slow> slow mode

(10)

1

Chapter I: Introduction

1.1 Background

Human communication is puzzlingly captivating. We know for a fact that humans did not invent the first system of communication, since other animals possess this ability as well. What is about language then that makes it so special? According to Duranti (2005), “what makes language so precious to human species is, namely, its ability to function in context as an instrument of both reflection and action upon the world. So-called ‘cognitive models’ rely on the assumption that it is possible…to account for human behaviour by means of context-independent rules” (p. 18).

We have come a long way in understanding more about the depths of human communication and its relation with language. Between the nineteenth and twentieth century, much of the study on language was already part of a larger domain of research digging more deeply “into the cultural origins of human populations” (Gumperz, 1982, p. 9). In this sense, the consideration of ‘culture’ as a significant variable was essential for pioneering works in the field of the ethnography of speaking focusing on the role of individual communities around its social units (speech community, speech situation, speech event, speech act) (see Hymes, 1964, 1974, 2005); on the cultural cognitive assumptions working as determinants of speech behaviour (Goffman, 1964); and in the relationship of language, social identity and the usage of discourse strategies in communication (Gumperz, 1982).

In the study of language use, the success of verbal interactions between two (or more) interlocutors is constrained by what Hymes (1972) named ‘communicative competence’:

Linguistic competence is understood as concerned with the tacit knowledge of language structure, that is, knowledge that is commonly not conscious or available for spontaneous report, but necessarily implicit in what the (ideal) speaker-listener has to say…We have then to account for the fact that a normal child acquires knowledge of sentences, not only as grammatical, but also as appropriate. He or she acquires competence as to when to speak, when not and as to what to talk about with whom, when, where, in what manner. (pp. 271-277) Thus, grammatically correct sentences are not sufficient to account for communication processes. Any action has to generate a subsequent reaction for communication to take place, and even prior to that—for such conversational circumstances to produce—participants are required to possess knowledge that surpasses the barrier of grammar skills in order “to decode short isolated messages” (Gumperz, 1982, p. 2).

As social animals, communication is present in all aspects of our lives. To some authors, “the development of…culture is made possible through communication, and it is through communication that culture is transmitted from one generation to another” (Gudykunst et. al, 2002, p. 3). Others, like Levinson (2006), differentiate between language as a means for communication and for interactional intelligence, where the latter makes the former possible, since for language to evolve “there was

(11)

2

already a need, that is, for communication interaction” (p. 42). To Levinson, ‘interaction’ is the main ingredient for culture to develop and propagate. He named it the human ‘interaction engine’:

humans are endowed with a set of cognitive abilities and behavioural dispositions that synergistically work together to endow human face-to-face interaction with certain special qualities. I call these elements collectively the human ‘interaction engine’. (p. 44)

Nonetheless, human interaction is far from being a cut-clear process. When people interact, their utterances might not always match the underlying intentions. What a listener interprets from our expressions will not necessarily align with our initial aims. This accumulative process in discourse where speakers gather prior knowledge and other information is known as ‘common ground’ and it implies the people’s “sum of…mutual, common, or joint knowledge, beliefs, and suppositions” (Clark, 1996, p. 93). Although we share similar cultural values of the society we inhabit, communicating our beliefs and world views with others represents a challenge. Frequently, we fail to make ourselves understood.

Meaning production is subject to interaction, but interaction is equally dependent on the cultural differences shaping different individuals (Wierzbicka, 2003). How we are influenced is deeply tangled to the cultural representations of the community we inhabit, which are exchanged through language by the members of same and different sociocultural groups (House, 2007). Thus, the notion of common ground gains special importance when we realize most communication nowadays takes place in uncommon grounds. ‘Intercultural’ speakers lie in the middle of communication processes happening between cultures, while carrying along the way their own values and beliefs.

As a social phenomenon, language is indisputably linked with culture. In order to be correctly understood, the meaning of linguistic elements are dependent on the cultural contexts (House, 2007). In that line, Blommaert & Verschuerern (1991) argue “this is exactly why it is so difficult to identify issues of intercultural and international communication. They are located at the level of interaction between members of such ever-shifting entities intermediate between humanity and individual human beings: communities, cultures, ethnic groups, nations” (p.2).

Nevertheless, languages have resources to face emerging issues and manage interactions. Interlocutors are endowed with various linguistic devices they can opt for at a discourse level. What is more, the language choices can also relate to the suppositions about the partner’s knowledge (Jucker & Smith, 1998). For instance, if considering planning strategies, one aspect that comes to mind is the dimension of ‘real time’ that speakers must deal with while preparing their utterances in conversation (Östman, 1981). In order to hold the floor, speakers may use discourse particles (i.e. like, well) working as pause-fillers, hesitation markers, etc (ibid, 1998).

In this thesis, I am mainly interested in the study of discourse particles (henceforth DPs) as strategic discursive tools. DPs have been observed in natural occurring conversations and were found to be used by speakers in order to integrate the utterance to their environment, with the aim of enhancing the common ground between speakers and hearers (Jucker & Smith, 1998). Our knowledge

(12)

3

on the functioning of these conversational elements is the result of the 70s research outburst aiming to deepen the field of discourse analysis. As stated by Östman (1981):

Before the increasing interest in sociolinguistics and discourse analysis…it was tacitly assumed that most pragmatic expressions and pragmatic particles were random hesitation phenomena, or emotional outcries having little to do with people’s inherent knowledge about their language. Such ‘pseudo-linguistic’ devices were regarded as the prime examples of why a competence-performance distinction was needed: the pragmatic particles being regarded as competence-performance errors. (p. 15)

Östman’s quote refers to a currently left-behind-view on DPs in communication. Increasingly, we observe the importance of studying the use of DPs in native settings as well as understanding their distinctness to non-native settings, as it is the case, for example, with English as Lingua Franca (ELF) conversations. While native English is formally defined by its norms, ELF is rather demarcated by its use in intercultural communication. The ‘world Englishes’ suggest an open door for linguistic investigation and DPs lie at the core of such opportunities. How ELF speakers incorporate these elements into speech reveals facets about their communicative competence, as well as how they conceive, interpret and apply them to attain communicative goals. Lastly, empirical evidence is becoming the norm in ELF research. With the replacement of anecdotical data with online corpora, discourse and conversation analysis has become more accessible to researchers.

1.2 Rationale of the study

English as a Lingua Franca (henceforth ELF) is considered as a type of intercultural communication. Research on ELF communication has been approached from different branches in linguistics, but in research on particles is fairly recent, with most of the accumulated knowledge accounting for observations based on native English speakers. Current investigation on the functions of particles in ELF communication remains scarce. The present study deals with a particular scenario of ELF communication, namely, professional business where English acts as the language for negotiation. In the following sections I briefly highlight the significance of DPs in ELF, as well as for business communication in intercultural contexts.

1.2.1 Why study discourse particles in ELF?

DPs abound in everyday communication. Their high frequency in discourse is indisputable. Neglecting these elements and their role in language would be unreasonable. Although DPs have suffered from bad reputation, stigmatised as unnecessary fillers in language and related to the lack of proficiency, contrasting views on particles also consider them vital for maintaining the discourse natural and comprehensible. That is to say, “[if] such markers are omitted, the discourse is grammatically acceptable, but would be judged “unnatural”, “awkward”, “disjointed”, “impolite”, “unfriendly”, or “dogmatic” within the communicative context” (Brinton, 1996, pp. 35-36).

In ELF communication, the diversity of spoken English and the participants’ different cultural norms can posit challenges that interfere with the achievement of mutual comprehension. In such

(13)

4

cases, speakers display several tactics in order to prevent or solve situations of misunderstanding (Kaur, 2011b). Furthermore, DPs have also been associated with the development of ELF speakers’ pragmatic competence as indicators of intersubjectivity and connectivity (House, 2013). ELF and native English discourse differ in their use of DPs because of the nature of these communicative situations. Although empirical research on the relation between DPs and ELF communication is increasing, most studies have focused on phenomena like turn-taking, topic management and/or the construction of utterances (House, 2009). Hence, research opportunities are wide and necessary. 1.2.2 Why study ELF business communication?

During the past three decades we have witnessed a rapid increase in intercultural communication as a result of factors such as globalization, the improvement of telecommunication technologies, namely, the internet and the development of our means for transportation. People with a vast diversity of languages and cultural backgrounds are interacting in ways that were previously unthinkable. Research on intercultural communication has never been of greater relevance as growing international contacts demand for higher intercultural competence (Kaur, 2011b). However, this has led to the common belief that communication in intercultural settings is unstable due to the differences in shared values, norms and views from people holding different upbringings (Gudykunst et. al, 2002).

Conversely, recent studies on intercultural communication in ELF have shown how instances of miscommunication and misunderstanding are not as common as expected since speakers remain actively engaged in the discussion deploying various strategies to achieve mutual intelligibility (Mauranen, 2006; see Mustajoki, 2017). For instance, and in an overlapping manner, speakers in intercultural business situations have been proved (1) to accept others participants’ discourse as understandable, ignoring the abnormalities causing confusion (Firth, 1996), and (2) to insist in the reformulation of the utterances with the aim of achieving the interactional goals (Tsuchiya & Handford, 2014). As the language par excellence for intercultural encounters, ELF presents an opportunity to shed light on and examine the sources of problematic talk, as well as the strategies adopted by its interactants.

1.3 Research questions

This thesis explores the use of I mean in ELF business communication (see also Fernandez-Polo, 2014) by observing data of naturally-occurring interactions gathered from a corpus of professional business meetings. In addition, it analyses the extent to which I mean helps alleviating business communication issues (see also Kaur, 2011b). Specifically, two main research questions are addressed:

1. What are the functions of I mean as a DP used by ELF speakers in business settings? 2. To what extent is I mean employed to pre-empt communicative misunderstanding? In addition, the following sub-questions are considered:

3. What is I mean’s predominant position? How does it relate to its functions?

(14)

5 in ELF business communication?

5. Do I mean co-occur with other DPs? Do different clusters affect I mean’s functions? 6. Does gender affect the recurrence of use and functions of I mean?

7. Is there evidence for language transfer by ELF speakers when using I mean? Do the other languages they speak have an equivalent translation to I mean?

1.4 Overview of the thesis

The thesis is structured as follows. Chapter 2 introduces previous literature on the two main axes for this work: discourse particles, and ELF communication. The research method, process for sample selection, tools employed as well as the difficulties encountered are then explained in Chapter 3, after which the analysis and results of the data are presented in Chapter 4. Chapter 5 discusses the findings and implications of the study, addressing the research questions in contrast to earlier research, as well as recognising its limitations. Finally, Chapter 6 outlines the key conclusions to the study and recommendations for further research.

(15)

6

Chapter II: Literature review

2.1 Discourse particles

The study of DPs has long been an intriguing subject of interest in linguistic research. While DPs gained particular attention back in the 70s, we can trace the discussion on this subject as early as in the 50s when Quirk (1955) described the use of you know, you see and well as “sharing devices and intimacy signals in our everyday talk” (p. 178). Later on, Levinson (1983) signaled the importance of these items:

there are many words and phrases in English, and no doubt most languages, that indicate the relationship between an utterance and the prior discourse. Examples are utterance-initial usages of but, therefore, in conclusion, to the contrary, still, however, anyway, well, besides, actually, all in all, so, after all, and so on. It is generally conceded that such words have at least a component of meaning that resists truth-conditional treatment…[W]hat they seem to do is indicate, often in very complex ways, just how the utterance that contains them is a response to, or a continuation of, some portion of the prior discourse. (pp. 87-88).

And almost a decade after Levinson’s statement, Wierzbicka (1991) added:

There are few aspects of any language which reflect the culture of a given speech community better than its particles. (p. 341)

While existing studies agree on the importance of DPs as items that provide hints on how to understand the organization and segmentation of discourse (Aijmer, 2002), the discussions revolve around the agreements and contentions on whether the term correctly defines and englobes its true meaning and functions.

2.1.1 Terminology and definition

Little consensus has been achieved so far regarding the terminology of DPs, making it especially difficult for newcomers to find their way in this domain. To a certain degree, this occurs because of the multifunctional nature of particles. Some of these functions have been described as “discourse connectors, turn-takers, confirmation-seekers, intimacy signals, topic-switchers, hesitation markers, boundary markers, fillers, prompters, repair markers, attitude markers, and hedging devices” (Jucker & Ziv, 1998, p. 1).

Among the variety of terms used to refer to these elements, next to the term “discourse particle” (Schourup, 1985; Aijmer, 2002; Fischer, 2006), we also find “discourse marker” (see Schiffrin, 1987; Jucker & Ziv, 1998; Fraser, 1990, 1999), “pragmatic particle” (see Östman, 1981; Foolen, 1996), “discourse operator” (see Redeker, 1991), “pragmatic marker” (see Brinton, 1996; Aijmer, 2013; Beeching, 2016), and “pragmatic expression” (see Erman, 1987). Depending on the type of research and design the terms show subtle differences from the others. In linguistic research, it is often the case that the term DP overlaps with “discourse markers” (DMs) (Schiffrin, 1987), in the sense that both work as linguistic expressions in the realm of discourse (language use) as opposed to that of the

(16)

7

sentence (language structure) (Hansen, 1998). In the present study, both terms will sometimes be used interchangeably. As Lam (2008) states about these two terms, “both refer to items which function at the discoursal level” (p. 10).

The work by Schourup (1985) considers DPs as being primarily discourse phenomena and as mediators of what he calls the “disclosure problem” (p. 2). In other words, DPs are seen to work as a bridge between the speakers’ covert thinking (what is considered as appropriate to be verbalized and what is retained from conversation) and their external behaviours (ibid):

It’s clear, then, that what we call speakers are also thinkers, with one foot in the collaborative world of talk, and the other in the internal world of their thoughts, which they may, or not, choose to display. (p. 4)

Based on this premise, Schourup suggests an analysis of DPs following a tripartite model: the

private world (what the speaker has presently in mind but has not yet disclosed); the shared world

(what is displayed in conversation by the speakers and is thus available to others); and the other world (what remains invisible to the speaker from the covert thinking of other interlocutors). The author intentionally avoids differentiating between DPs, interjections and DMs because of the existing overlap between these items. His primary focus is on the particles like, well and y’know, and how they function within the model of the speaker’s world. As he concludes, each of these particles indicate the speaker’s previous thinking that remain undisclosed to the rest. They are what he calls an “evincive”: a linguistic item that indicates that at the moment at which it is said the speaker is engaged in or has just then been engaged in, thinking; the evincive item indicates that this thinking is now occurring or has just now occurred but does not completely specify its content. (p. 14)

Another popular definition is offered by Schiffrin (1987), who describes DMs as “sequentially dependent elements which bracket units of talk” (p. 31). She further adds to her definition “a more theoretical level as members of a functional class of verbal (and non-verbal) devices which provide contextual coordinates for ongoing talk” (p. 41). Schiffrin’s meticulous work set as solid base to understand how DMs function to add coherence to the discourse. In discussing the properties of markers, she explains that their communicative force is dependent on the discourse slot in which they are employed and “the linguistic properties of the expressions used as markers” (p. 41). Schiffrin’s model of coherence in discourse is based on the idea that language occurs in a context and it is sensitive to it. Here, markers are treated as coordinators for utterances occurring within the model at different levels. First, in three structures: two non-linguistic, i.e. exchange structure (alternation roles by speakers), action structure (organization of speech acts); and one linguistic, i.e. ideational structure (propositions). Second, in two planes of conversation: participation frameworks (the interlocutors’ ways to relate to one another) and information states (the interlocutors’ knowledge management). Altogether, the five features work as the pillars for Schiffrin’s model, considering local adjacency units in discourse as well as global dimensions.

Fraser (1990) considers DMs as a class within pragmatic markers (PMs) and differentiates them from interjections, and pragmatic expressions (such as y’know, I mean). Under Fraser’s definition,

(17)

8

DMs are “expressions…which signal a sequential relationship between the current basic message and the previous discourse” (p. 383). Following his approach, the meaning of a sentence must be split into two forms of encoded information that he describes as ‘content’ and ‘pragmatic’. While the content meaning of a sentence transmits what the speaker is trying to communicate, the pragmatic information of a sentence “signal[s]…the speaker’s communicative intentions” (p. 386). Fraser refers to these indicators as PMs. He further claims the importance of analysing DMs, not merely relying on their content meanings, but distinguishing among their core pragmatic meanings (ibid.). The following examples are adapted from Fraser’s (1991, p. 393) analysis of so:

a. Susan is married. So she is no longer single. b. John was tired. So he left early.

c. Teenage son: The Celtics have an important game today. Disinterested parent: So?

Each of the previous scenarios display so with several interpretations as a consequence of the prior context, but, as Fraser concludes, “certainly not with any narrow result sense associated with the content meaning of so” (ibid, p. 393).

A third definition comes from Redeker’s (1991) critical response to Schiffrin’s research, arguing that the latter’s “minimalistic approach to the semantics of discourse markers places too heavy a burden on the syntactic and contextual determination of marker meanings” (p. 1139). Redeker opted instead for the term “discourse operator”:

A discourse operator is a word or phrase—for instance, a conjunction, adverbial, comment clause, interjection that is uttered with the primary function of bringing to the listener’s attention a particular kind of linkage of the upcoming utterance with the immediate discourse context. (p.1168)

Redeker’s definition finds its consistence from three components (i.e. ideational structure, rhetorical structure, sequential structure) delimiting a discourse operator, as opposed to Schiffrin’s three planes of conversation for DMs. Redeker mentions “these three components are roughly equivalent to [Schiffrin’s] ideational and actional structures and an extended variant of her exchange structure” (p. 1167). While describing the two first components, the author refers to two discourse units as being ideationally and rhetorically connected. In the ideational component the relation exists only if the speaker acknowledges the existent bond between utterance and context. In the rhetorical component the focus lies on the illocutionary intentions conveyed between the two units. At last, the sequential transitions mark the “paratactic and hypotactic relations between ideationally and rhetorically only loosely adjacent discourse segments” (p. 1168). Finally, discourse operators differently influence each component. How discourse operators will aid to improve language understanding may be determined by the hearer’s knowledge and inferences (ideational), the illocutionary goal of the discourse (rhetorical), and the guidance provided to raise awareness in the hearer (sequential) (ibid.)

Moreover, Fung & Carter (2007), describe DMs as “intra-sentential and supra-sentential linguistic units which fulfil a largely non-propositional and connective functions at the level of

(18)

9

discourse” (p. 411) mainly working as signallers of the relation between utterance and context. Similarly, Aijmer (2013) refers to PMs with respect to the language users and their relation to context. What is interesting about this relation, as she highlights, is the defining property of pragmatic markers to be creative, which is evidenced in their capacity to “index the context by altering an existing context and creating a new (linguistic or social) context” (p. 6). This is useful for speakers to shift to different topics and generate responsiveness in the listener. To Aijmer, PMs are “contextual resources enabling speakers to express a number of new stances in the interaction” with the “rhetorical function [of] allowing the speaker to take up a stance of alignment or disalignment to the hearer or to what is said” (p. 15). Finally, Hansen (2006) escapes from the focus on the relation between utterances to signal instead the cognitive context:

The role of markers is, in my view, to provide instructions to the hearer on how to integrate their host utterance into a developing mental model of the discourse in such a way as to make that utterance appear optimally coherent (p. 25).

Hansen’s approach to DMs places the emphasis on the relation between the utterance and a mental model under development rather than the utterance and the context. Following a cognitive-functional framework, her analysis of French toujours revealed it works as an adverb in all its syntactic uses, but with instances of both propositional and non-propositional marking applications attached to truth-conditional meanings.

To my understanding, DPs structure discourse by rendering it clearer to the listener’s ears, to whom it might not be immediately evident. DPs are signallers of the relations between adjacent utterances to the direct context. Thus, they make the speaker’s speech more transparent and understandable to convey information.

2.1.2 Characteristics of discourse particles

Aijmer (2002) refers to DPs as difficult to analyse because of the pragmatic functions relating the interlocutors and the utterances, but at the same time serving as contextual clues to facilitate the hearer’s interpretation of the utterance. DPs represent both a challenge—for its evident multifunctionality as linguistic items—and a field of interest because of the contrast they provide on the relationship between “semantics and (discourse) pragmatics…speech and writing, and cohesion and coherence” (Hansen, 1998, p. 6).

Occasionally, conversations in English can be constituted mainly by DPs. Mind the next extract from the film “Ocean’s Twelve” (2004), when Danny approaches Basher with a question, while waiting for their train:

Danny : Can I ask you…? Do I look 50 to you? Basher : Yeah.

Danny : Really?

(19)

10

Notice how Basher’s second utterance would not change in the absence of well, I mean and you

know. In most studies, this feature of DPs is referred to as optionality. If removed from their original

position the sentence structure remains unaffected (Schiffrin, 1987). Thus, grammaticality and intelligibility are not compromised (Lam, 2008). Brinton (1996) claims this occurs because DPs “occur either outside the syntactic structure or loosely attached to it and hence have no clear grammatical function” (p. 34). However, syntactic optionality is never misinterpreted with DPs as being useless. In fact, Schourup (1999) claims that optionality is also considered as guiding the listener “toward a particular interpretation and simultaneously ruling out unintended interpretations” (pp. 231-232). This loose relation with the syntactic class provides them with great mobility to work in multiple categories including “adverbs (e.g. now, actually, anyway), coordinating and subordinating conjunctions (e.g.

and, but, because), interjections (e.g. oh, gosh, boy), verbs (e.g. say, look, see), and clauses (e.g. you see, you know)” (Schourup, 1999, p. 234).

Another defining feature of DPs is that they are said to carry no or little propositional meaning. Schiffrin (1987) called it the ‘meaning-minimalist’ view. The presence of DPs does not affect the propositional content of the utterance. To Brinton (1996), the reason why it is so difficult to translate DPs into other languages comes as a consequence of their semantic blurriness. However, Erman (1987) argues that DPs are not completely desemanticized, “although, admittedly, their original or literal meaning is only present to a certain degree” (p. 15). While this property remains important, the consensus among researchers on this component is not absolute.

DPs are considered to fulfil pragmatic functions of indexicality. All utterances are context-bound (Schiffrin, 1987), which implies that speakers are expected to present their utterances to the hearers in temporal and spatial terms (ibid). This characteristic of relating discourse units has also been referred to by Schourup (1999) as ‘connectivity’; that is, “either relating the current utterance to its immediate linguistic co-text, often known as local coherence, or to segments further apart, often known as global coherence” (Lam, 2008, p. 15). To Schiffrin (1987), indexicality is key to understand the reason for the use of DMs, since they “propose the contextual coordinates within which an utterance is produced and designed to be interpreted” (p. 315). In contrast, studies have also focused on the relation between DPs and the attitudinal aspect speakers express through language (Lam, 2008). Some even have considered DPs to “have an emotive, expressive function rather than a referential, denotative, or cognitive function” (Jucker & Ziv, 1998, p,3). For instance, the propositional content in Basher’s utterance carried some hesitation or uncertainty that becomes obvious only after his consecutive use of the three DPs.

The positioning of DPs is a feature that usually holds agreement between researchers. DPs are found in all positions during discourse, playing a vital role in providing information of the discourse event taking place, its progress, turnabouts and segments. Still, DPs are typically positioned as utterance-initial (Schiffrin, 1987; Fraser, 1990; Brinton, 1996; Schourup, 1999). As pointed out by Schourup (1999):

The tendency toward initiality must be understood to refer to the position of DMs in relation to the central clause elements rather than to the position of the first word in an utterance since items supposed to be DMs frequently cluster at utterance onset and elsewhere. (p. 233)

(20)

11

Some studies have considered initiality to be a defining syntactic property of DPs (see Brinton, 1996; Fung & Carter, 2007). Alternatively, Lam (2008) suggests that the tendency of utterance-initial position in DPs “is a language-specific choice rather than a universal phenomenon” (p. 20).

Furthermore, DPs are described as multi-functional. Differently from ordinary words in the language, DPs are assigned manifold pragmatic values by speakers, who in return appropriately differentiate and manipulate the DPs’ several meanings in different contexts (Aijmer, 2002). It has been suggested that multifunctionality “can be explained as a result of grammaticalization or ongoing grammaticalization” (ibid, p. 5). Moreover, Schiffrin (1987) considers a specific condition to be a marker that “[it] has to be able to operate at both local and global levels of discourse” (p. 328). The ongoing debate on the multifunctionality of DPs extends even to issues such as whether it is feasible to determine if they possess core meaning.

The extent to which non-truth-conditionality is an inherent feature of DPs remains controversial. DPs are usually not considered to contribute anything to the truth conditions of an utterance proposition (Jucker & Ziv, 1998; Schourup, 1999). For instance, Grice (1989) discussed connective devices such as but, so and therefore concluding they have meaning but remain non-truth conditional. As for like, it has been claimed that “the omission of the marker…may lead to ungrammaticality and loss of propositional meaning” (Andersen, 1998, as cited in Lam, 2008, p. 19).

Finally, the orality of DPs is often seen as a stylistic, non-mandatory property. This appears to imply that DPs mainly occur in spoken discourse. Brinton (1996) considers PMs as:

A feature of oral rather than of written discourse. The appearance of pragmatic markers is a result of the informality of oral discourse and the grammatical “fragmentation” caused by the lack of planning time, which makes the use of pragmatic markers expedient. (p. 33)

But there are no reasons to believe on their exclusion from written discourse, since similar elements are widely found here as well (i.e. moreover, consequently) (Schourup, 1999). Lastly, the increasing development of new technologies, makes the distinction between oral and written use of DPs more blurred (Lam, 2008).

2.1.3 Functions of discourse particles

The multi-functional nature of DPs reveals intricate approaches with different taxonomies. First, we consider as a central function of DPs how they organize the communication process by inserting utterances into the communicative context (Foolen, 1996). Moreover, DPs are said to function in two major domains: textual and interpersonal (Lam, 2008). According to Jucker & Ziv (1998), DPs act as “text-structuring devices (marking openings or closings of discourse units or transitions between them)” (p. 4). On an interpersonal (or global) level, particles can initiate or close discourse, marking boundaries in talk and/or sequential dependence, “to constrain the relevance of one clause to the preceding clause by making explicit the conversational implicatures relating the two clauses” (Brinton, 1996, p. 37).

(21)

12

For example, well works as a frequent turn-initial particle with the potential of assuming different meanings in different contexts, and signaling both the previous utterance or the upcoming turn (Aijmer, 2013). In contrast, really appears as a closing marker which signals the stances of speakers “towards the propositional meanings of the utterances” (Fung & Carter, 2007, p. 419). Finally, Fraser (1999) considers DMs to have procedural rather than conceptual meaning. While the latter “specifies a defining set of semantic features” (p. 944), the procedural meaning of markers indicate how the introduced section is to be understood with regard to the prior topic.

2.1.4 ‘I mean’

I mean is one of the most frequent particles occurring in English discourse (Beeching, 2016). It is

considered as a speaker-oriented DP that marks the pace of the interlocutors within the talk (Schiffrin, 1987) and belonging to a cognitive category, because it denotes the process of thinking (Fung & Carter, 2007). Among the terminology assigned to this particle we find discourse item (Schourup, 1985) and pragmatic expression (Erman, 1987).

A recent study by Beeching (2016) indicates that many of the functions of I mean derive from the two canonical senses of the verb “to mean”: to signify and to intend. The two following examples are adapted from her work, showing the difference between both senses:

What does “quintessentially” mean? (= What does “quintessentially” signify?) What do you mean to do? (= What do you intend to do?)

However, these meanings may be subject to variation in specific settings “according to how the basic corrective or clarifying core use is applicable in the particular context at hand” (Schourup, 1985, p. 108). Brinton (2008) describes particles like I mean as comment clauses composed with a formal structure of the first person I and a present-tense verb mean that have acquired pragmatic functions with both subjective and intersubjective meanings. Both components influence the uses of I

mean as a DP. While the first person centres its focus “on the speaker’s own adjustments in the

production of his/her own talk” (Schiffrin, 1987, p. 299), the predicate mean presents polysemous meaning and meta-linguistic properties (ibid.). Though, the orthographic composition of I mean does not allow adverbial elements (i.e. only) in between both words (Brinton, 2008).

As with most DPs in the literature, the functions of I mean have been identified and accepted by some scholars while excluded by others. First, I mean presents an explanatory function if the predicate mean has ideational meaning that refers to expressing concepts or enlarging the speaker’s previous ideas (Schiffrin, 1987). Second, when it is used for clarification purposes, the speaker’s intention is also displayed by the particle when the speaker considers the hearer to have missed the intended action or misinterpreted the meaning behind that action (ibid.).

Furthermore, as a pragmatic connective that occurs in fluent speech, I mean exhibits self-repairing functions. Erman (1987) considers I mean to fall into four different repair categories, namely, the repetition of a structure with no alteration, the restart or abandon of the structure to begin a

(22)

13

different construction, the insertion of a more accurate wording and the correction of the original structure. In contrast, Schiffrin (1987) differentiates between two types of self-repairs: background and replacement repairs. In background repairs, the speaker adds or modifies information to assure listener comprehension, leading “back to the discourse which had been interrupted” (p. 301). The repair usually targets the surrounding information which is available to the hearer. In contrast, replacement repairs supply different information than the background repairs in the sense that “they switch the direction of the developing discourse to that initiated by the substitution” (p. 301).

When I mean performs as a hesitation it is “fairly frequently used in initial positions along with other pause-fillers and stallers without any apparent (self-)repair or reformulation purpose” (Beeching, 2016, p. 185). Moreover, the hedging character of I mean can function as a softener of the previous statement, considered by the speaker as an “evaluative comment” (ibid., p. 189) to be mitigated. Finally, Fox Tree & Schrock (2002), claim that the particle’s functions can be discussed in terms of its basic meanings. First, I mean works as an adjuster of speech in interpersonal encounters where speakers are careful about what they say and how they say it. Secondly, speakers use I mean to monitor the understanding of other interlocutors as a sign of response that the adjustment was taken into consideration.

2.2 English as Lingua Franca (ELF)

To understand ELF, it is important to place it in perspective. That is, we need to recognize that the English diversity around the world, such as we know it, has its foundations in colonial and postcolonial history. Schneider (2007) explains how the current state of affairs with English in the lead as the world’s lingua franca, occupying a central role in international communication, is the result of a series of uncalculated events that promoted its spreading. Schneider notably refers to the colonization processes of the British Empire between the seventeenth and nineteenth centuries, subsequently complemented with the industrial revolution, and followed by the emergence of the USA as the new economic and military power in the twentieth century, wrapped up and driven by the engine of globalization.

For the past four decades the field of research on ELF has focused on its worldwide deployment and practice, settling as the widest use of contemporary English among people who do not share the same first language. As Seildhofer (2009) states:

English as a lingua franca is a fascinating object of study in that due to its extremely widespread and frequent use by speakers from a vast number of first languages it affords us the opportunity of observing language contact, variation and change happening in an intensified, accelerated fashion right before our eyes. (p. 195)

A large portion of all international communication taking place nowadays is indeed attributable to English and its function as a lingua franca, to such an extent that the number of ELF speakers overtakes that of native speakers of English (House, 2012). However, the question has been posited if due to all these English varieties, different levels of proficiency and speakers’ cultural backgrounds communication may become unintelligible.

(23)

14

According to Firth (2009), there are three overlapping areas contributing to the development of the field: “(a) language policy and language planning, (b) language teaching and pedagogy and (c) micro analyses of naturally occurring ELF conversations” (p. 162). The last area has enjoyed from special attention in ELF research during the past three decades, with abundant topics addressing matters such as professional negotiations (see, Firth, 1991, 1996; Tsuchiya & Handford, 2016), the use of DPs (see Baumgarten & House, 2010; House, 2009, 2013; Kaur, 2011b; Fernández-Polo, 2014), pragmatic strategies (see Björkman, 2011), misunderstandings (see, Mauranen, 2006; Kaur, 2011a, 2017), negotiation of meaning (see De Bartolo, 2014), language accommodation (see Drljaca Margic, 2017) and humour (see, Matsumoto, 2017). Relatedly, this thesis concentrates its analysis on naturally occurring ELF conversations.

2.2.1 Definition and Characteristics of ELF Communication

In talking about ELF, Kecskés (2019) refers to it as a slippery object, almost like a ‘soap’ for scholars, “[t]he more scholars try to ‘grab’ it, the more it escapes” (p. 17). Differently from a language or dialect, interaction in ELF communication is difficult to analyze under standardized rules or conventions. As a product of communication it is inherently variable. Firth (1996) defined ELF as “a ‘contact language’ between persons who share neither a common native tongue nor a common (national) culture, and for whom English is the chosen foreign language of communication” (p. 240).

The first and most defining characteristic of ELF is that “all participants use a common language that is not their L1” and “[w]hen they use the common language they usually rely on the variety of English” (Kecskés, 2019, p. 19). As a contact language, the essence of English acting as a lingua franca is that of allowing distinct cultural backgrounds to enter in touch with each other, pursuing conversational objectives. Furthermore, ELF speakers have been usually trained in the English language at some point in their lives (Mauranen, 2006). An additional feature can be understood as a consequence of such multicultural encounters, namely, the speakers’ constant negotiation of meaning to ensure successful communication (De Bartolo, 2014).

In addition, accommodation processes are generally an important aspect of social interactions (see Giles, 2016) and thus they also implicate ELF situations. Human communication characterizes for its cooperativity and territoriality, because on the one hand we adjust our language in order to address others, while on the other hand we safeguard our individual space and identity (Seildhofer, 2009). The results of such accommodation reflect in the strategies ELF speakers use for meaning negotiation.

Among the earliest of these studies analysing pragmatic strategies in ELF was Firth (1991, 1996). His findings on the analysis of business telephone conversations between Danish export managers and their clients showed a range of tactics adopted by speakers. The first strategy is named the ‘let-it-pass’ strategy, in which the hearer allows unclear utterances in order to avoid redundant, problematic communication and enhance common ground. The second strategy is termed ‘make-it-normal’ and is described as the normalization of the speaker’s unidiomatic usage of language (ibid.). Here, the hearer focuses on producing reformulations of the opaque usage of language. In contrast to Firth’s findings, Tsuchiya & Handford (2014) emphasise what they call the ‘not-let-it pass’ strategy

(24)

15

in professional ELF meeting, showing the Chair engaged in reformulating the discourse of other participants in order to enhance overall comprehension.

Moreover, Björkman (2011, 2014) analysed and compiled ELF pragmatic strategies in academic settings. He highlights the prioritization of communicative effectiveness by ELF speakers. Among these different tactics are the use of “backchanneling, repetition of the phrase, idiomatic avoidance, confirmation checks, self-repair, word-spelling and explicitness” (2014, p. 126). Moreover, the search for communicative effectiveness lies at the core of the characteristics defining ELF discourse and, therefore, the “pragmatic strategies can be presumed to play a profound role in the communication process and in preventing misunderstanding since significant communicative efforts are made through these strategies” (2011, p. 960).

2.2.2 (Mis)understanding in ELF interactions

Problematic communication in intercultural encounters may emerge as a result of poor comprehension at different linguistic levels between the interlocutors and their unwillingness to cooperate in the interaction (House et al, 2014). According to Gudykunst & Kim (2002), the cultural differences make us seeing other people (and groups) as strangers, because “they are physically close…[but] they have different values and ways of doing things” (p. 23). In particular, the notion of ‘misunderstanding’ has been broadly explored from multiple disciplines such as psychology (see Keysar & Henly 2002), second language acquisition (see Bremer, 1996; House et al., 2014), intercultural communication (see Gudykunst et. al 2002), pragmatics (see Thomas, 1983), as well as in ELF studies (see Mauranen, 2006; Kaur, 2011a, 2017).

Bremer (1996) believes misunderstanding situations develop when “the listener achieves an interpretation which makes sense to her or him–but it wasn’t the one the speaker meant” (p. 40). In intercultural communication, definitions like the one advanced by Bremer have usually been associated in preceding studies to the cultural background differences of the interlocutors, working as the main factor that prompts misunderstandings. However, recent evidence studying ELF interactions have suggested that, contrary to the intuitive belief, “it is not clear that communication breakdown is more common among second language users; the anticipation of communicative difficulty may in itself offset much of the trouble, and speakers resort to proactive strategies” (Mauranen, 2006, p. 123).

In one study, Kaur (2011a) analysed 15 hours of naturally occurring spoken interactions within an academic setup to find that, rather than culture-based comprehension problems, the sources of misunderstanding related to instances of “ambiguity, mishearing, and lack of world knowledge” of the interlocutors, where the latter case of “misunderstandings…pertain to [the] content [being shared] rather than [the] language [used]” (p. 109-112). Finally, Mustajoki (2017) goes even further to suggest misunderstanding is more common in everyday communication (EvC) than in lingua franca communication (LFC). In EvC the interlocutors are characterized for knowing each other so well that they disregard the dissimilarities in their speeches, while in LFC the participants remain actively focused in the interaction, unconcerned “with the way they speak [since] what they want is to be understood” (p. 62).

(25)

16 2.2.3 ELF, discourse particles and I mean

DPs have been claimed to play a central role in language (Aijmer, 2002). Although there is a substantial amount of research on DPs (see section 2.2), our understanding of their development and ways of use in lingua franca communication is limited (Mauranen, 2006). While native speakers seem to employ them competently, it appears to be particularly difficult for non-native English speakers (ibid.). Moreover, evidence suggest that non-native speakers tend to use DPs less than native speakers (Fung & Carter, 2007).

House (2009) suggests that ELF use of you know differs from its native use in that ELF speakers employ it to track down the progress and organization in the discourse. That is, differently from its interpersonal functions usually associated to the construction of intersubjectivity “appealing to knowledge shared between speakers and addressees…[r]ather you know is predominantly a self-serving strategy” (ibid, p. 178). In addition, the concept of ‘subjectivity’ has been defined by Benveniste (1966) as “the capacity of the speaker to posit himself as [a] subject” (p. 224) through language. Benveniste considers that the organization of language endows the speaker with a body of deictic elements to refer himself or herself as I, in the same way it allows him or her to designate others by pointing to you. Besides the deictic pronoun first person singular I, he adds:

Other classes of pronouns that share the same status depend in their turn upon three pronouns. These other classes are the indicators of deixis, the demonstratives, adverbs and adjectives, which organize the spatial and temporal relationships around the “subject” taken as referent: “this, here and now,” and their numerous correlatives “that, yesterday, last year, tomorrow,” etc. They have in common the feature of being defined only with respect to the instances of discourse in which they occur, that is, in dependence upon the I which is proclaimed in discourse. (p. 226)

Similarly, Baumgarten & House (2010) explored the collocations I think and I don’t know in ELF and native English discourse, viewing them as highly frequent, stance-marking I + verb constructions. They sought to find if the linguistic expressions differed in both types of interactions. In doing so, they define the concept of ‘stance’ as the “speaker’s…attitude[s] towards what he or she is saying…which convey [the speaker’s] personal feelings, attitudes, value judgements and assessments” (p. 1185). As such, the use of I by the speaker is at the same time his or her way to express subjectivity and to forward a stance. The findings revealed ELF speakers use the constructions

I think and I don’t know as clause structures (in their prototypical meaning), as compared to native

speakers who favour the structures with pragmaticalized meaning. Furthermore, House (2013) observed the use of three DMs, namely yes/yeah, so and okay in ELF academic consultation hours and found the users to exhibit pragmatic competence by reinterpreting the DMs for their own purposes in structuring discourse “as well as for expressing concern for their interlocutors” (p. 65).

Existing knowledge on the performance of I mean in ELF communication is rather recent. Few studies exploring this DP are available. For example, Baumgarten & House (2007) studied I mean in a context of academic consultation talk with one group of native English speakers and two groups of ELF speakers. Although the instances of I mean happened more often in ELF than L1 discourse, the results showed dissimilar patterns for both ELF groups. On the one hand, the group 1 ELF speakers

(26)

17

used the particle following its primary property of reformulation and clarification in order to enhance hearer comprehension. On the other hand, the group 2 ELF speakers revealed another related property, namely, the use of I mean serving as a focalizing device to their own speeches “or an explicit expression of a subjective evaluation” (ibid., p. 208). Likewise, the analysis of academic presentations in ELF by Kaur (2011b) proposed multiple self-repair strategies by ELF speakers to increase explicitness and secure speech comprehension. I mean served as a repairer of segments with the speaker forewarning upcoming adjustments or, as Kaur states, “what I actually want to say or am trying to say” (ibid. p. 2712). Unfortunately, her analysis is limited to the self-repair function of the DP.

Lastly, Fernández-Polo (2014) studied 34 monologic conference presentations from the ELFA corpus containing 97 000 words. In total, his findings show 48 instances of I mean performing four main roles: correcting mistakes and dysfluencies, enhancing clarity and explicitness, organizing text and marking certainty and salience (ibid.). Particularly, the function of correcting mistakes and dysfluencies revealed problematic uses of I mean, because in conference presentations the speakers are typically expected to be prepared in advance. However, the minor number of instances constrained the author to conclude with possible generalizations on his findings.

2.2.4 ELF business communication

Intercultural business communication is characteristically defined by the interlocutors not sharing the same L1 and choosing as a result a lingua franca. As mentioned before, English is a specific type of intercultural communication, characterized by its array of voices and the fact that non-native world-English speakers regularly adapt their discourse strategies and communicative styles when using English as a means of interaction (see Firth, 1996; Tsuchiya & Handford, 2014).

Business contexts were English is used are known as Business English Lingua Franca (BELF) (Kankaanranta, 2009) and are demarcated by the cultural differences, multiple pronunciations and degrees of comprehensibility among its speakers (Tsuchiya & Handford, 2014). As a result of globalization, that is, the global mergers and company networks, the promotion and integration of English into corporate language signified that “members and employees of such entities had to be able to use ELF professionally” (Kankaanranta, 2009, p. 2).

A key aspect enfolding business communication is the recurrence for negotiation. Whether it involves explicit transactions, where the expected outcome is factual and tangible, problem-solving workgroup discussions or workplace meetings, the shared (intrinsic) constant of interrelated goals by the interlocutors remains present, which they want to combine for the profit of both (Dresemann, 2007). On the other hand, BELF speakers have been found to engage, first, in meaning negotiation (Firth, 1996) in order to succeed with their ambitions in business negotiation. In BELF communication the participants’ language proficiency is generally goal-oriented and domain-related. In other words, because of the professional framework within which these conversations take place, the speakers’ communicative competence is proficient enough to cope with the objectives and achieve results in their domain of expertise.

(27)

18

Business meetings in professional communication are of particular interest for conversation analysts. Yunick (1997) explains how, as social encounters, meetings “are a prime location for the investigation of an internal ‘corporate culture’ and of how national cultures and Englishes play into the construction of that corporate culture” (p. 628). The dynamic process of meetings to achieve goals engages all participants in the display of communicative strategies usually based in power relationships (Măda & Săftoiu, 2012). Among the interactants, the most important role is assumed by the chair of the meeting, who is in charge of setting the agenda, initiate the meeting, establish control, keep track of the progress, negotiate consensus on the purpose of the meeting with everyone, as well as making everyone feel involved in the discussion (ibid.)

Nevertheless, research in ELF business meetings has often been neglected as a result of the bias in research to study dyadic interaction over the exploration of turn-taking in larger groups (Wolfartsberger, 2011). This can be explained in terms of the complexity of polylogal organizations, where turn-taking is affected by the existing unbalance in simultaneous talk and floor-holding, leading to problematic distribution of the chances to talk among the participants and, on the other hand, giving rise to ‘schism’ or “the splitting of the interaction into various parallel conversations” (ibid, 2011, p. 166). This issue has led the exploration of ELF business meetings to remain underexplored.

(28)

19

Chapter III: Methodology and Data

This chapter discusses the methodological approach of the study. It describes the path followed for data search and selection, the reason for my choice, as well as the freeware tools employed to extract and analyse the data containing I mean instances. Finally, I list some difficulties encountered in the process.

3.1 Corpus research

Corpus linguistics (henceforth CL) was the chosen method for data collection on the use of I mean in ELF business communication. CL concentrates on the empirical analysis of language by means of computer software, allowing access to real-life spoken or written texts of recorded conversations. Corpus analysis is considered to fit within the confines of discourse-based analyses (Handford, 2016). Particularly, in the domain of ELF and intercultural communication, corpora methods have shown to be resourceful for the study of interactions both qualitatively and quantitatively. Seidlhofer (2012) commented on the intersection between CL, ELF and intercultural communication:

If we want to understand how intercultural communication works in a globalized world and how it may be improved, we need to revisit and revise established ideas about what constitutes a language or a community and how people actually draw on linguistic resources to communicate with each other in the real world. We need, in other words, to understand how ELF is used. (p.146)

The previous quote makes reference to the rise and positioning of English as the lingua franca in the present-day world. Hence, corpus analysis represents an attempt to understand the nature of ELF in order to move beyond traditional models based on native speaking contexts. One such effort is the Vienna-Oxford International Corpus of English (VOICE), working as the first structured corpus compilation of language data seizing ELF interactions. The data for this thesis was gathered from VOICE.

3.1.1 The Vienna-Oxford International Corpus of English (VOICE)

VOICE is probably the biggest corpus for the analysis of lingua franca communication. As stated on their website, VOICE (2013) is defined as a “structured collection of language data, the first computer-readable corpus capturing spoken ELF interactions of this kind”. The corpus comprises 1 million words of transcribed naturally occurring, non-scripted spoken ELF conversations from five different domains (i.e. education, leisure, professional business, professional organizational and professional research/science), equalling approximately 120 hours of transcribed speech.

An important fact, relevant for the selection of VOICE over others corpora, is that the recorded participants were advanced ELF speakers. This matters to our purposes because of the preferred emphasis placed on ‘language use’ rather than ‘language learning’. Therefore, speakers of VOICE are considered to be proficient within their domains. In addition, the project contains approximately 1 250 ELF speakers with approximately 50 different first languages with a primary focus on European ELF

(29)

20

speakers. The recorded interactions are classified into the following speech event types: interviews, press conferences, service encounters, seminar discussions, working group discussions, workshop discussions, meetings, panels, question-answer-sessions, and different kinds of conversations.

The importance of VOICE to present research on ELF interactions is explained in the fact that, as a contact language, English is used for communication between persons without the same tongue nor culture and it is, therefore, being mainly shaped by them (Kecskes, 2014).

3.1.2 Sample selection

As above-mentioned, VOICE includes interactions from five domains. Table 3.1 displays the information regarding the speech events and number of words for each particular domain:

Table 3. 1. Domains of recorded interactions (VOICE, 2013)

Domain Speech Events Words Relative in 1/100

ED (educational) 35 261 003 25.51

LE (leisure) 26 101 216 9.89

PB (professional business) 23 203 421 19.88

PO (professional organizational) 41 354 602 34.66

PR (professional research and science) 26 102 945 10.06

I opted for analysing the domain of PB (professional business), which “includes all social situations connected with activities of making, buying, selling or supplying goods or services for money” (VOICE, 2013). The PB contains five sub-segments of speech events: conversation (con), meeting (mtg), panel (pan), question-answer session (qas) and service encounter (sve). I focused on the meeting (mtg) sub-segment. A meeting is defined as “a speech event at which a clearly defined group of people meets to discuss previously specified matters” (ibid.). Table 3.2 summarizes the information regarding the PB domain:

Table 3. 2. Speech event types within PB (VOICE, 2013)

PB’s speech events Nº of Events Speakers Words

Conversation (con) 1 4 2 212

Meeting (mtg) 8 55 154 634

Panel (pan) 4 37 36 695

Question-answer session (qas) 4 43 6 684

Service encounter (sve) 6 13 3 196

The use of ‘I mean’ as a discourse particle in ELF business communication