Abstract
Marginalization in decision-making discourse results in disempowerment of the marginalized and detracts from the efficacy of participatory decision making. In ESL contexts, it is usually associated with English proficiency. But this view ignores the influence of preferences for different participation styles, an understanding of which is essential for the development of effective pedagogical remedies to the problem of marginalization. The present study addresses this gap by investigating discourse participation and marginalization from a participation styles perspective. Findings reveal that marginalization resulted from a failure to adopt turn-taking strategies associated with dominant participation styles. Implications for pedagogy are discussed.
Many organizations across the world employ participatory decision making, diffusing decision-making power among their members for different purposes in different situations. In Japan, for example, big decisions tend to be made after extensive behind-the-scenes discussions or formal intraorganizational consultations between stakeholders (Brislin & Nab, 2004; Misumi, 1984). Participatory decision making is also common in the United States and northwestern Europe, where it typically involves a formal group discourse stage such as a meeting (Sagie & Aycan, 2003). Participation in such decision-making discourses can be seen by participants in value terms, as the right to exercise a degree of power in an egalitarian, professional society (Baraldi, 2013), and such discourses can be genuine forums for debate where “subordinates have to be persuaded on the basis of arguments” (Noorderhaven et al., 2007, p. 1354).
Participatory decision making is based on the rationale that exploiting the resources of all participants produces the most abundant options and thus the best possible decisions (Yates & de Oliveira, 2016). It follows, therefore, that a diversity of perspectives among participants is a strength and one that must be exploited through inclusivity (Maznevski, 1994). However, decision-making discourse sometimes lacks inclusivity in intercultural contexts where English is used as a lingua franca (hereafter ELF), and in some cases, certain individuals or groups are marginalized (e.g., Lockwood & Forey, 2016; Tanaka, 2008; Walker & Aritz, 2014). Marginalization from decision-making discourse is consequential because it leaves marginalized individuals disempowered, may be perceived as relating to their professional competence (Takino, 2020; Tanaka, 2008), and detracts from the overall efficacy of participatory decision making (Maznevski, 1994; Sagie & Aycan, 2003).
Yet despite the stakes, the problem of discourse marginalization has not been a serious focus of research or pedagogy. Historically, it has been seen as related to English proficiency (e.g., Cook, 1989) and as a result, ESL pedagogical approaches have focused on encouraging learners to adopt participation norms based on the intuitions of (often, native-speaker) instructors and textbook writers (e.g., Kern, 2009; Young, 2018). But this ignores other ways that participation can be achieved and discriminates against those who prefer different discourse participation styles. For English to be seen as an inclusive global professional lingua franca, a pedagogical approach to discourse marginalization is needed that takes into account a diversity of participation preferences in decision-making discourses. The present study contributes to the development of such an approach by investigating what happens when diverse preferences for participation styles come together in ELF decision-making discourse.
Participation Styles
Participation in group discourse is achieved through the mechanism of turn taking, a coherent theory of which was first postulated by Sacks et al. (1974) and is still the standard reference model. The Sacks model explains how turn taking constrains participation because people cannot intelligibly talk at the same time. They must take turns, which are made up of units of speech at the end of which is a transitional-relevant-place (TRP) where a transition to a new speaker may occur. Participation is dependent on these transitions, the two mechanisms for which are current speaker selects next (select-other) and next speaker self-selection. The way that people participate in a particular instance of group discourse depends on two fundamental values about turn taking with which they approach discourse: the autonomy they feel they have to self-select, and the way they balance their personal goals for the discourse with those of the group. These values are influenced not only by the context of the discourse and their perceived roles within it, but also by their perceptions of turn-taking norms derived from their past experiences in big and small cultural contexts. In other words, how people participate depends on their previous sociocultural experiences of discourse participation. Based on these two fundamentals, autonomy and goals, there are four theoretical participation styles: a collaborative style, a competitive style, a hierarchical style, and a cooperative style (Figure 1).
Features of participation styles at turn transitions.
Different turn-taking strategies are associated with each style. If autonomous self-selection behavior is the norm, there will tend to be more overlaps in the discourse because the more freely people self-select, the greater the chance there is that they will speak simultaneously, either intentionally or inadvertently (Baraldi, 2013). Overlaps are collaborative when interruptions are made to help coconstruct an idea or a sequence of discourse and reflect an orientation to communal rather than personal discourse goals (Wolfartsberger, 2011; Zhu & Boxer, 2021). Overlaps are competitive when interruptions are made in the pursuit of personal goals, to exert influence for example. If such interruptions are resisted, a discursive contest for the speakership occurs (Coates, 2004). On the other hand, if autonomous self-selection behavior is not a discourse norm, fewer overlaps will occur at transitions and elsewhere. In such cases and where participants prioritize their personal discourse goals, they are likely to self-select, but deferentially, leaving gaps at transitions, avoiding interruptions, and ceding the floor quickly when overlaps occur. They effectively cooperate to ensure that everyone can speak as much as they wish (Baxter, 2014; Tsuchiya, 2016; Yamada, 1990). Where participants prioritize the discourse goals of the group, a facilitator may manage everyone’s turn-taking through the select-other mechanism to ensure that those group goals are achieved in an efficient manner. Such hierarchical participation managers can be officially appointed or may emerge organically during the discourse.
People hold preferences for participation styles that are formed by a complexity of sociocultural influences. Their preferences represent the participation norms of communities they identify with filtered through their personalities. Research has shown, for example, that gender identity can be indicative of participation style (Coates, 2004). Baxter found that the participation style of her male subjects differed from her female ones, who “enacted leadership in a more distributed, coconstructed way so that everyone had a voice” (2014, p. 446). Another indicator may be national culture. For example, a higher tolerance for intrusive interruptions has been observed in decision-making discourse in American (Maznevski, 1994; Yamada, 1997), British (Baraldi, 2013), and some northern European contexts (Henningsen & Henningsen, 2020; Noorderhaven et al., 2007) in contrast to Japanese (Noorderhaven et al., 2007) and Chinese contexts (Ding, 2006; Morita et al., 2011).
The East-West cultural contrast in participation styles has been noted by a number of researchers. Yamada’s (1990) comparison of American and Japanese bank manager meetings was an early study that reported longer transition gaps and more equal turn lengths in Japanese meeting discourse compared to American meeting discourse. At about the same time, Murata (1994) made a similar distinction about transition gaps in Japanese ESL discourse compared to British English discourse. Tsuchiya (2016) reported more interruptions in the ESL discourse of a Japanese-Libyan group than in a Japanese-Venezuelan and Japanese-only groups. Fewer overlaps were recorded in Asian ESL meetings in contrast to native-speaker American meetings by Walker and Aritz (2014), and Du-Babcock and Tanaka described the turn-taking behavior of Japanese and Hong Kong Chinese as “mutually supportive and inclusive” (2016, p. 13).
When people with diverse stylistic preferences create discourse together for the first time, a set of turn-taking norms must be “negotiated” so that discourse can proceed efficiently. Conflicts between stylistic preferences are resolved discursively in what has been described as a “discursive struggle” (Fairhurst & Connaughton, 2014, p. 17). Stylistic “consensus” is achieved through acts of discursive leadership and followership through which the ground rules of participation are established (Baxter, 2014; Clifton, 2012; Du-Babcock & Tanaka, 2016; Walker & Aritz, 2014). However, emergent participation norms may not facilitate the participation of all participants. Some may be unable, or unwilling, to use the turn-taking strategies associated with dominant styles. Walker and Aritz found such a discrepancy between the discourse participation rates of Asians and Americans, with the American speakers dominating discourse through their greater use of certain turn-taking strategies: “by saying more, speaking more frequently, and taking longer turns” (2014, p. 95).
Causes of Marginalization and Constraints on Participation Behavior
Understanding processes of marginalization involves understanding all ways that participation can be constrained. Yet, historically, the marginalization of ESL speakers has been attributed to problems of English proficiency, and instruction based on the intuitions of native-speaker teachers or textbook writers (Cook, 1989; Dörnyei & Thurrel, 1994). Viewing marginalization as a problem solely of English proficiency results in instructors treating people with different stylistic preferences for participation as learners yet to develop correct behaviors. Such a pedagogical approach conflicts with current views of ELF as a tool for pragmatic communication across cultural barriers (e.g., Deterding, 2013; Ehrenreich, 2010; Rogerson-Revell, 2010) and is incompatible with a view of English as the language of global business, rather than acculturalization.
There are turn-taking strategies that do depend on English proficiency. This is most clearly the case for strategies associated with collaborative and competitive styles where fluency allows for the quick prediction of TRPs required to take minimal-gap turns and protect turns from interruptions (Levinson & Torreira, 2015). There are also, however, strategies less dependent on fluency, such as emphasis, which can be used in collaborative and competitive styles to mitigate relative fluency disadvantages. Presupposing that language proficiency is the only cause of marginalization also assumes that people are willing to adapt their turn-taking behavior, they are just not able to do so. This assumption ignores the moral element to participation styles. People take turns in certain ways, not because it is the only way they can behave but because they believe it is the right way to behave (Brown & Levinson, 1987).
In addition to behavior, participation is also constrained by the qualities of the situation, such as the genre of the discourse and the discursive task. Virtual meeting technology also restricts turn-taking behavior through both its audio and visual communication channels (Swartz et al., 2020). Next speakers cannot be selected by eye contact in video meetings and simultaneous speech can be restricted by microphone capabilities. Yet, situational constraints usually impact all participants in the same way. In addition, extratextual modes of communication such as email may be employed to mitigate a variety of communication problems including those related to participation. Lockwood and Song (2020), for example, found that even very low English proficiency levels did not prevent inclusivity in virtual team meetings between an American client and a Chinese IT services provider when a range of strategies were employed by American meeting facilitators. These included a mixture of discursive strategies (e.g., slow speech, repetition, transition gaps) and extratextual strategies (e.g., pre and post emails, predictable meeting patterns).
Discursive Leadership and Participation
As with overcoming situational constraints, overcoming conflicting preferences for participation styles is a matter for leadership. Leadership, seen from a discursive point of view, is a social and communicative phenomenon rather than a psychological trait (Clifton, 2012; Fairhurst, 2007; Walker & Aritz, 2014). Leaders, in this sense, are revealed through their discursive acts and others’ responses to them (Baxter, 2014; Clifton, 2012; Du-Babcock & Tanaka, 2016; Aritz & Walker, 2014). While this does not require an official role, the influence of leadership on participation is clearest in cases of appointed leaders who enjoy advantageous turn management rights (Baraldi, 2013; Clifton, 2012; Kangasharju, 2002) that can be exploited to promote inclusivity (Lockwood and Song 2020) or to enhance marginalization. Lockwood and Forey (2016), for example, reported how the Australian chair of virtual meetings between Australian and Indian partners marginalized his Indian colleagues by allocating more turns to Australian colleagues. Discursive leadership does not only depend on appointed rights, however. Participation is shaped through leadership acts by all participants, and all participants have the potential to influence inclusivity and marginalization.
Summary
Marginalization in decision-making discourses is consequential for marginalized individuals and their organizations but has not, so far, been a focus of either research or pedagogy in ELF contexts. The present study addresses this gap by investigating marginalization processes so that more effective pedagogical solutions for international teams may be devised. The following assumptions have been made at the outset:
Intercultural ELF decision-making discourse is a locus for discourse marginalization because of the high diversity of participation style preferences related to cultural background and the additional constraint of relative language proficiency.
The conflict between diverse preferences for participation styles is a causal factor of marginalization.
Discursive leadership is how stylistic conflicts are resolved and how inclusivity is achieved or avoided.
The aim of the present study is to shed light on how participation styles emerge in such ELF discourses and how their emergence affects discourse inclusivity and marginalization.
Methods
Methodology
Although the present study investigates intercultural discourse, this is not a cross-cultural comparative study in the big data quantitative tradition of Hofstede (e.g., 2010). The study does not seek to assign particular behaviors en bloc to particular national groups. Nor is the study focused on exploring the origin of stylistic preferences. While it is assumed that culture in both the macro (e.g., national cultures) and micro (e.g., professional communities) sense is a significant influence on the ways that people prefer to participate in discourse, it is not assumed to be the only influential factor. The focus of the study is on the discursive negotiation of diverse stylistic preferences. Thus, the most vital data are interactional data.
Previous studies have emphasized the importance of both quantitatively and qualitatively analyzing interactional data, combining techniques from Interaction Analysis (IC) and Conversation Analysis (CA) (e.g., Aritz & Walker, 2014; Du-Babcock & Tanaka, 2016; Poncini, 2007). The present study also adopts this twin analytical approach to answer two questions:
To what degree is participation achieved (or not achieved)?
How is participation achieved (or not achieved)?
In previous studies, techniques from IC have been used to shed light on rates of participation, by comparing variables such as turns, number of words, and speaking time (e.g., Aritz & Walker, 2014; Du-Babcock & Tanaka, 2016; Lockwood & Forey, 2016), and participation preferences, by comparing variables such as transition gaps (Aritz & Walker, 2014; Lockwood & Forey, 2016; Tsuchiya, 2016), and turn-transition mechanisms (Lockwood & Forey, 2016). CA has been used to study discursive leadership by identifying and examining significant interactional sequences within discourses (e.g., Baxter, 2014; Clifton, 2012). Previous CA studies have also investigated how assessments are framed (Clifton, 2012), responses made (Baxter, 2014), and the functionality of certain linguistic items such as pronouns (Poncini, 2007). To my knowledge, however, there have so far been no studies that focused on the specific processes of participation.
CA depends on the clear transcription of the complexities of discourse, including its verbal and nonverbal features. Transcriptions are used to facilitate analysis and to report results. The conventions of transcription used in the present study are based on Jefferson (2004) (see Table 1).
Transcription Conventions.
Turns are indicated by a turn number (e.g., T1), followed by the speaker’s name. In the present study, two categories of utterance are not considered as full turns and are recorded without a turn number. The first category are short overlapping utterances made by someone contesting the speakership but who does not win it. The second are overlapping short one- or two-word backchannels made to signal agreement, confirmation, or some other response to what the current speaker is saying but not intended to lead to an extensive turn (here referred to as tokens). Simultaneous speech is recorded in square brackets and transcribed in one of two ways. Longer overlapping sequences are recorded across two or more lines belonging to the different speakers. Spaces may be added inside the square brackets so that the two overlapping sequences align vertically on the page. Shorter overlapping tokens are recorded in square brackets containing the name of their speaker and are placed on the same line as the current speaker at approximately the position they occurred in the discourse.
Transitions with no gaps and no overlaps (zero-second gaps) are described here as latched transitions and indicated by an equal sign. An utterance that is abruptly finished (cut off) is indicated by a hyphen at the moment of cut-off. Gaps assessed as significantly long are indicated by round parentheses containing one to three period marks depending on the length. In significant cases, the time of the gap is recorded in seconds. Vocal stress is indicated by an underline of the word in which it occurs. Where speech is unintelligible, round parentheses are left empty. All analyst commentary about the discourse is contained in double round parentheses. This category includes commentary about missing segments of the discourse as well as descriptions of significant nonverbal signals made by participants.
Although the interactional data form the basis for the analysis and take precedence over other data, it was decided to also collect participant views about participation styles using a questionnaire instrument. A similar technique has been used in previous discursive leadership studies (Du-Babcock, 2006; Walker & Aritz, 2015) and in the present study, allowing the analyst to directly ask participants about their stylistic preferences, which could then be compared with their actual behavior.
The Study Context
Previous studies of decision-making discourse have been conducted using data collected from both authentic workplace interactions (e.g., Clifton, 2012; Kangasharju, 2002; Lockwood & Forey, 2016; Lockwood & Song, 2020) and simulated scenarios (e.g., Aritz & Walker, 2014; Baraldi, 2013; Baxter, 2014; Du-Babcock & Tanaka, 2016). Collecting data from simulated decision-making contexts has the advantage of allowing researchers to exercise control over influential variables such as the diversity of participants and the nature of the decision-making task. Some studies have also collected data from virtual meetings (e.g., Du-Babcock & Tanaka, 2016; Lockwood & Forey, 2016; Lockwood & Song, 2020). Virtual meeting data offer many advantages for IC and CA analyses. A video recording of a video meeting, for example, exactly replicates the experience of the participants for the analyst, in that it captures the same visual and audio signals they experienced at the time. In addition, the playback and pause facilities of video allow for accurate transcription of complex overlapping sequences and accurate recording of quantitative variables related to participation. For these reasons, it was decided to use decision-making discourse data collected by video recording of simulated virtual interactions using the Zoom application.
Three 40-minute meetings, held virtually in Tokyo in January 2021, were selected for the present analysis. The meeting simulations formed part of an English-medium diplomatic leadership program organized under the joint auspices of the Japanese government and a public graduate university in Tokyo. The researcher, in his role as course instructor, was given access to the data, which was collected and provided by the university. Written consent was obtained from all participants in the meetings as well as the program administrators at the university prior to the commencement of the study. A decision-making scenario from a published text called A Difficult Choice! (Lloyd & Preier, 1996) was used in which participants role-play a meeting of an ice cream company board choosing a new product. Subjects were assigned one of four manager roles, an order of preference for the three candidate products A, B, and C, and a set of data supporting those preferences. At least one of each manager type was assigned to each meeting group. No facilitators were appointed.
The design of the task gave participants conflicting personal goals (to choose their preferred product) and required them to rely mostly on self-selection to participate. As a result, competitive or cooperative styles were incentivized over hierarchical and collaborative styles. This design was adopted to encourage individual acts of discursive leadership and prevent a reliance on others for participation. The members of the meeting groups were selected from eight Japanese participants and six non-Japanese participants from the Czech Republic, South Korea, India, Thailand, and Romania (Table 2). Within the constraint of that pool, gender and cultural diversity was maximized so that participation preference diversity might also be maximized. While not a first encounter in the truest sense, the participants did not know each other well, had not met face-to-face, were of a similar age (late 20s to late 30s) and social status, and therefore were unlikely to have already formed participation norms or to adopt a hierarchical style based on perceptions of seniority. All participants were current employees of government ministries or agencies in their respective countries who used English regularly at work. In addition, a high English proficiency was a requirement of their entry into the English-medium program.
Meeting Groups (pseudonyms used).
Analytical Approach
Quantitative
The main assessment of participation was made using a participation rate variable, derived for each group member by calculating a mean of two percentages: their speaking time as a percentage of the group total and their total number of turns as a percentage of the group total. An assessment of linguistic proficiency based on the notion of fluency was calculated by counting the number of content words (excluding fillers, false starts, and repetitions) in a full minute of each member’s most substantial and representative utterance. Multiple turn-transition variables were also collected. These included the number of select-other transitions, the length of transition gaps (in seconds), and variables concerning simultaneous speech, which come in different categories and have different stylistic associations.
Qualitative
The qualitative analysis of participation involved identifying key transition and overlap sequences and associating turn-taking strategies with one of the four participation styles. For example, an instance of simultaneous speech resolved through a speakership contest is indicative of a competitive style, but simultaneous speech for the purpose of discourse coconstruction or alliance building is indicative of collaboration. The same feature resolved deferentially is indicative of a cooperative style. To understand how dominant styles emerged, acts of leadership related to participation were identified. In particular, novel turn-taking behaviors and responses to them were studied to see whether those novel behaviors were validated or resisted by other participants.
A framework of five questions were used to guide the analysis:
What participation styles emerged in the meetings?
How did linguistic proficiency affect rates of participation?
What kinds of turn-taking strategies were used?
How did participation styles emerge?
How did participation styles affect participation rates?
Results
What Participation Styles Emerged in the Meetings?
The dominant participation style in Meeting 2 was cooperative, while the dominant styles of Meetings 1 and 3 were competitive (Table 3). Cooperative participation in Meeting 2 was indicated by a high frequency of lengthy transition gaps, and uncontested transitions. Competitive participation in Meeting 1 was indicated by short, often minimal-gap (latched) transitions, and a high frequency of speakership contests. These features were even more prevalent in Meeting 3, indicating the style was highly competitive.
Main participation style variables.
How Did Linguistic Proficiency Affect Rates of Participation?
A simple comparison was made between the linguistic proficiency variable and the participation rate variable (Table 4).
Participation Rates Compared to Linguistic Proficiency.
nd = no data because poor quality of Sally’s microphone prevented a reliable word count.
The comparison revealed that the participation rate correlated quite well with linguistic proficiency but with notable exceptions. In two of three meetings, the participant with the highest rate of participation also recorded the highest English proficiency score, and in all three meetings, the lowest rate of participation was recorded by the person with the lowest proficiency score. Exceptions included Takahashi in cooperative Meeting 2. Her 16.7 participation rate was only a third that of Yamada’s 45.5 despite recording an almost identical proficiency score. Park was another exception in competitive Meeting 1. He recorded a participation rate higher than Sato (18:15.9) but a much lower proficiency score (82:111). Exceptions like these indicate that linguistic proficiency did not, by itself, explain different rates of participation.
What Kinds of Turn-Taking Strategies Were Used?
Conversation analysis of transition sequences revealed different turn-taking strategies that aligned with different styles. In cooperative Meeting 2, overlaps were rare and when they did occur, they were inadvertent rather than intentional, and resolved deferentially by one party quickly ceding the floor. This process can be seen in Extract 1, when Yamada starts T31 without realizing that Takahashi has not finished T30. When the overlap is noticed, both speakers attempt to cede the floor to the other with Yamada urging “go on,” and Takahashi responding with “please” (Extract 1).
Extract 1. Cooperative Meeting 2: Resolving inadvertent overlaps with deference
((Continues)) It has potential. And that’s what we need to look at (. .)
I see, I see [I think you’re working-] (.) yeh? (.) [Go on ]
[and then- m-hm? ] [uh, yes, yes]
(.) please
(. .) I just think that one thing we have to think about is ((continues))
In competitive Meetings 1 and 2, the prevalence of speakership contests as a strategic technique of turn transition meant that those who achieved a high degree of participation did so by outcompeting others. There were three strategy types used to win or avoid such contests: (1) strategies to empower speech through emphasis, (2) strategies to minimize gap length at TRPs and thus deny space to potential competitors, and (3) strategies to get attention from current speakers and attract a select-other turn allocation (Table 5). In the competitive meetings, the skilled use of these strategies was a significant factor distinguishing high participation rates from low. They constituted part of the “powerful discursive resources” that Clifton (2012) argues are the tools of discursive leadership (p. 150).
Turn-Taking Strategies Used in Competitive Meetings 1 and 3.
Despite the task design, collaborative participation strategies also sometimes occurred. Collaborative strategies such as supportive backchanneling (e.g., agreement tokens) and select-other turn allocations were sometimes used in attempts to build alliances in favor of, or in opposition to, various positions (Clifton, 2012; Kangasharju, 2002). An example of this can be seen in Meeting 1 (Extract 2) where in T65, Sato declares support for Anna’s T64 assertion. In T66, Ryuta then seeks an alliance with Park in opposition to Anna and Sato by selecting Park.
Extract 2. Competitive Meeting 1: Collaborative turn taking for the purpose of alliance building
((Continues)) People who want fruit will buy fruit not ice cream=
=Yeah, exactly!
(. .) What do you think Park?
An attempt to direct participation through an exclusive use of the select-other mechanism, a hierarchical participation strategy, was made by Ota in Meeting 3. A typical example can be seen in Extract 3, T7, where Ota selects Harada as next speaker and also assigns him a topic, product A.
Extract 3. Highly competitive Meeting 3: Ota attempts to direct participation
So er, let’s first check which manager are we [Harada: Yeah] (.) OK, who’s the er Sales Manager?
((raises hand)) me
(.) Thank you (. .) and the er Marketing (.) yes, is me. ((continues))
((continues)) So, how shall we start? So, Yuki-san and I, have the same choice [Harada: We prefer Teddy bar] Teddy bar (.) A (.) which is for children. Do you have any arguments to support, can we start with the support, er, for the A? OK, thank you. So, could you start Yuki-san?
How Did Participation Styles Emerge?
Dominant participation styles began to emerge in the early stages of the meetings when novel turn-taking strategies were used and validated by other participants. Where those strategies were resisted, a discursive struggle over style ensued. Deferential self-selection was modeled by Yamada and Romesh in cooperative Meeting 2 as early as T4 and T5 (Extract 4). Yet, their appeals for other participants to self-select were ignored, suggesting that there was resistance to self-selection.
Extract 4. Cooperative Meeting 2: Establishing deferential self-selection as a turn-taking norm
So who goes first? Like, anybody has (. .)
((3s gap)) Anybody?
((4s gap)) OK, so let me go first
In Meeting 1, Anna was instrumental in making competitive transitions a norm. She was the first to model the key competitive strategy of interruption during Park’s T8 (Extract 5). Park quickly ceded the floor, but when Sato then overlapped Anna (inadvertently and unintelligibly), Anna did not cede the floor. Her “sorry” indicated that Sato should give way. She demonstrated, therefore, that it was “normal” to both interrupt and block interruptions, and by giving way, Park and Sato validated those strategies. Soon afterward, they were copying Anna. Sato interrupted to claim T11, and Park to claim T16 and T18.
Extract 5. Competitive Meeting 1: Establishing interruption as a turn-taking norm
((continues)) We can make our specialties using (. .) some (.) um (.) little bit weird, fascinate ice-cream champagne truffle bar (.) [yeah- ]
[but the]
[big companies are there because (.) oh (.) sorry] (.)
[( )]
the market is good, established already ((continues))
A discursive struggle over style was instrumental in Meeting 3 becoming highly competitive. Initially, there was little resistance to Ota’s attempt to appoint himself meeting facilitator, as when Harada took the turn Ota selected for him (Extract 6, T8), albeit with an ironic comment, “I can speak?” Initially, Ota’s facilitation was validated, as when Mara asked him for permission to change topic (Extract 6, T8).
Extract 6. Highly competitive Meeting 3: Responses to Ota’s direction of participation
((Turn ends)) So, could you start Yuki-san?
Oh, I can speak? ((All smile))
Yes, please
((Completes turn 8))
((Latched self-selection)) Er(. .) Can I say about C or we go with B first?
OK then, please, for C Mara
However, as Meeting 3 progressed, resistance to Ota’s turn management increased. This was demonstrated by an increasing number of self-selections, which were in turn resisted by Ota as in Extract 7, T27, where he blocked Mara’s attempted interruption. The struggle to self-select in the face of Ota’s attempts to control selection led to an increase in the frequency and competitiveness of speakership contests. To win those contests, more gap minimalization, more emphasis, and more attention-getting strategies were used.
Extract 7. Highly competitive Meeting 3: Interruption becoming a turn-taking norm
[The situation that con-] (.)
[Er, if if I may ]
((Holds finger up to face)) Er, can I can I can I can I, [er yeah], finish it?
[Yes ]
so, the situation can be the same ((finishes turn))
How Did Participation Styles Affect Participation Rates?
Once a dominant style of participation had emerged, accommodation to that style by adopting its strategies had the greatest positive impact on participation, and not doing so, the greatest negative impact. Avoiding using appropriate turn-taking strategies increased the likelihood of marginalization, and thereby exclusion from the decision-making process. In cooperative Meeting 2, participation required deferential self-selection. It was avoided, however, by Takahashi, Sally, and Yamagishi with the result that their participation rates were significantly lower than those of Yamada and Romesh (Table 6).
Comparison of Participation in Cooperative Meeting 2.
In cooperative participation styles, avoiding self-selection means being dependent on being selected by a current speaker. A lack of such other-selections is why Yamagishi was marginalized in Meeting 2. He self-selected only once and was selected by others only twice as when he was selected by Takahashi in T22, Extract 8.
Extract 8. Cooperative Meeting 2: Yamagishi’s dependence on select-other
((turn continues)) I think there is another person who is a Sales Manager (.) if that person could speak, that would be good
((raising his hand)) Hi (. .) I am the Sales manager so I would like to give a little supplementary explanation (.) and er I also recommend to put the top sales for the C product ((turn continues))
In the two competitive meetings where turn taking was more dependent on language proficiency, participation rates were also dependent on the use of suitable competitive strategies. The adoption of competitive strategies mitigated the effect of lower English proficiency on participation rates, while avoiding them enhanced its effect. Emphasis strategies, in particular, helped win speakership contests. This can be seen in a comparison of participation variables from Meeting 1 (Table 7). Despite a low proficiency rate of 82 words per minute, Park enhanced his participation rate by effectively using emphasis strategies. Through their use, he was able to win turns and keep them in the face of competition. On the other hand, Sato’s lack of emphasis strategies meant that she lost most of her speakership contests. As a result, her participation rate was low relative to her English fluency. Tanaka’s avoidance of competitive turn-taking strategies altogether enhanced the effect of her low proficiency relative to other participants, and marginalization was the result.
Comparison of Participation in Competitive Meeting 1.
How a relative proficiency disadvantage can be mitigated with competitive turn-taking strategies can be seen in Park’s contributions. In Extract 9, T16, for example, he used a range of strategies in combination to outcompete Sato and Anna for the speakership, interrupting Sato and overcoming an attempted interruption by Anna. He used repetition of a rhetorical question (an emphatic structure), “what it cost?” stressing the word “cost,” followed by eye contact, a direct look into the camera, combined with emphatic pauses to empower his words. These strategy combinations contrast with Sato, who did not use such emphasis strategies but, instead, used a softening mitigator “I think” (T15).
Extract 9. Competitive Meeting 1: Using emphasis to win speakership contests
((continues)) I think (. . .) [we need to ( ) ]
[but, what that
[You don’t ( ) ice-cream]
is (.) well (. .) ((continues))
In Meeting 3, emphasis strategies were used most effectively by Mara and Ota, who consequently had the highest rates of participation (Table 8).
Comparison of Participation in Competitive Meeting 3.
The case of Harada in Meeting 3 is interesting from a pedagogical point of view because it offers an example of how stylistic adaptations can be made as discourse progresses, and how participation rates can be increased as a result. At first, Harada avoided competitive strategies but then gradually implemented them. As a result, his participation rate improved. His turn taking can be divided into two parts with the distinction being drawn at T39 when he made his first interruption (Table 9).
Comparison of Harada’s Two Participation Stages.
In the first stage, Harada took only 16% of turns, whereas in the second stage he took 23%. How he achieved this increase is revealed by examining his turn-transition behavior before and after T39. Prior to T39, Harada achieved most turns through latching, a gap minimalization strategy. He avoided emphasis strategies during this period losing the speakership to those who used emphasis, such as Mara in Extract 10, T23.
Extract 10. Meeting 3: Mara interrupts Harada
Er, I have er same opinion as Mara. People usually never associate er
[ice ] cream with er (.) healthy-
[yes ] -They associate it with a desert.
To supplement latching, Harada then started using attention-getting strategies to elicit turn selections from current speakers. In Extract 11, for example, he called the current speaker, Mara, by name. She then selected him for T29 in preference to Vanida, who used a similar, but less direct, attention-getting technique.
Extract 11. Meeting 3: Attention-getting strategies
[Mara?, Mara? ]
[((raised hand)) OK, May I?]
Yes, Harada?
I have a piece of information from my real story ((laughter)) ((continues))
Next, Harada began to use emphasis strategies. At first, he used them only to keep turns, such as in Extract 12, T32, where he employed emphatic gestures, word stress, and emphatic qualifiers to block multiple attempted interruptions (Extract 12).
Extract 12. Meeting 3: Employing emphasis strategies to keep the speakership
((continues)) We have a good reason (.) to have a first option for A, ac[tually] data set C, is also evi[dence] for the, you know, teddy bar.
[yeah ] ((nods)) [yes ]
[Yes, ((waving pen)) I
[But, but, teddy bar is the highest cost ]
((raised finger)) (.) [May I?]
Yeah, we may know very well Ota-san is very good at risk management from our previous conversation ((laughter))
[Okay ((attempts to interrupt)) ]
[and . er . . ] ((holds up finger to Mara)) (.) teddy bar is best option because the cost is very
Finally, He used emphasis strategies to initiate and win speakership contests. In Extract 13, T56, for example, he used a persistent repetition of the word “even” and a raised finger gesture to empower his claim to the speakership over Ota, and to retain his turn in the face of an attempted interruption by Vanida.
Extract 13. Highly competitive Meeting 3: Employing emphasis strategies to interrupt
((turn continues)) Eighty percent [ of ] the young shows a willingness for
[Yeah]
chocolate (.) [and we need-] ((cedes the floor))
[and even- ] ((finger raised)) (.) yeah (.)
[and even the age even the even the ] (.) age group of ((continues turn))
[That’s for ice-cream not for chocolate]
Conclusion
Participatory decision making is undermined by discourse marginalization. The process is deprived of diverse perspectives and the marginalized are disempowered. In contexts where marginalization can be a problem, such as in ELF contexts, preventing marginalization should be a pedagogical goal for communication instructors. The findings from the present study support the hypothesis that participation rates in such contexts are influenced not only by language proficiency but also by participation styles, and that marginalization is often a product of the discursive negotiation of conflicting preferences for participation styles.
The study investigated the emergence of participation styles in decision-making discourse involving culturally diverse groups based on the assumption that members of such groups often hold conflicting stylistic preferences. Indeed, that assumption was endorsed by the diversity of stylistic preferences that were reported by participants in post-task questionnaires. While, in most cases, relative proficiency correlated well with relative participation rates, in some cases, it did not (e.g., Sato, Park, and Takahashi) and a stylistic explanation appeared more likely. In competitive-style meetings, a relative proficiency disadvantage was mitigated by the skilled use of competitive turn-taking strategies (as demonstrated by Park in Meeting 1). And in cooperative Meeting 2, relative proficiency did not explain why Sally, Yamagishi, and Takahashi, avoided self-selection. The analysis therefore suggests that while English proficiency facilitates participation it does not ensure it. The influence of participation styles must therefore be a central consideration in a pedagogical approach to discourse marginalization.
Discussion
Two possible explanations for the avoidance of turn-taking strategies that would enhance participation rates are that (1) participants are unable to adopt them because of a lack of sufficient English proficiency, and (2) participants are unwilling to adopt them for some reason unrelated to proficiency. Research in the field of pragmatics has revealed that communication behavior closely associated with sociocultural norms rather than language, an area known as sociopragmatics, is governed by deeply held values that resist change (Leech, 1983). In other words, sometimes how language is used is a moral question rather than a linguistic one and that people associate politeness and impoliteness with participation styles (Brown & Levinson, 1987).
Some of the reported evidence from questionnaires collected in the present study suggested that turn-taking behavior was influenced by stylistic preferences, while other evidence did not. Tanaka, who was marginalized from competitive Meeting 1 because she avoided competitive strategies expressed an aversion to speakership contests, writing, “even just a facial expression without voice makes me feel the saying is unsuitable for the timing.” Yamagishi, who by avoiding self-selections was marginalized from cooperative Meeting 2, expressed a preference for a hierarchical style where participants’ turns are managed by a facilitator and should “wait to be asked to speak.” On the other hand, Takahashi, who avoided self-selection in Meeting 2, expressed a contradictory preference for “competitive turn-taking” where “people speak when they feel like it.”
Pedagogical solutions to marginalization need to address causes related to both language proficiency and participation style preferences. The building of fluency and the teaching of turn-taking strategies addresses proficiency as a cause but it does not address the avoidance of certain turn-taking strategies. For that, people need to be helped to accommodate to behaviors they currently find objectionable and to develop leadership competence so that they may better influence the participation styles of their meetings. Referring to cultural role models, like Harada or Park in the present study, may be useful in such instruction. Ideally, discourse marginalization should be a concern for all stakeholders, not only those in language training programs. Native English speakers and organizations alike must realize that in the multicultural ELF meetings that are such a feature of modern globalized business, their stylistic preferences for how participation should be achieved should not hold a privileged status and inclusivity is as much their responsibility as that of ESL speakers.
Footnotes
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
