Turn beginning is a crucial location for establishing speakership and gaining recipiency. It is one of the most challenging tasks for an individual learner. This study used a corpus-based approach to investigate how Chinese EFL learners employ multimodal resources to successfully execute turn beginnings, with a specific focus on potential variances between intermediate and advanced proficiency levels. The findings revealed a hierarchy in the allocation of linguistic resources, with feedback being the most prevalent, followed by demonstrative, connective, and syntactic relevance. This suggested that learners prioritized feedback on the previous turn at the outset of a turn, often serving an interactional function to convey acknowledgement and recognition. Interestingly, intermediate learner group used more demonstrative to provide a cue for utterance-interpretation. However, the advanced learner group preferred to establish logical relevance between each turn through the use of connective. In terms of non-linguistic resources, learners predominantly employed gaze and head movement to project turn beginning intention, demonstrating their capacity to harness multimodal resources to a certain extent. Moreover, the advanced learner group’s increased use of hand gesture suggested that as language proficiency improved, learners were more inclined to engage the hearer’s attention at the commencement of turns. This study contributes to the scholarly understanding of multimodal turn-taking by elucidating the distinct characteristics that learners exhibit when initiating conversational turns. It offers insights into how learners navigate the complex interplay of linguistic and non-linguistic resources in the pursuit of effective communication.
Article
Multimodal Turn Beginning in Chinese English-as-a-foreign Language Learners’ Conversation
https://doi.org/10.21203/rs.3.rs-5237641/v1
This work is licensed under a CC BY 4.0 License
Version 1
posted
You are reading this latest preprint version
Turn beginning
Multimodal resources
Chinese EFL learners
Corpus-based
Turn-taking consists of three essential positions: turn beginning, turn holding and turn completion (Sacks, Schegloff & Jefferson, 1974; Li, 2019). Among these, turn beginning is particularly vital for both speakers and hearers, as it plays dual roles: projection and connection (Schegloff, 1996). Specifically, the beginning position of turn can project a multi-unit turn and provide a glimpse into the subsequent discourse content. It is also frequently employed to establish lexical, grammatical or prosodic connections with the preceding turn, fulling interactional, discourse structuring and cognitive function (Haselow 2019). In English, the commencement of a new turn is typically marked by a large number of brief words or discourse markers, such as oh, so, and, yeah, and others. For example, word oh is utilized to forge lexical connections, indicating a shift of the knowledge state and fulling an interactional function.
However, as for turn beginnings in the context of L2 interactions, it has been noted that being able to begin a turn is a potentially challenging task for L2 speakers (Carroll, 2004). Relevant studies have shown that Chinese EFL learners encounter many turn-taking problems in interaction. For example, they are incompetent in understanding and judgement of the turn relevance position (TRP), resulting in long silences and monologues in conversations and the loss of interactive communication features (Zheng & Luo, 2014). The research mentioned above, however, mostly relies on oral tests (Greer & Potter, 2008) and examines turn beginning at the linguistic structures, ignoring the multimodality of it (Deppermann, 2013; Thörle, 2016). Lee (2017) found that learners typically lean forward and gaze at the current speaker, accompanied by gesture extension, to signal their intention to obtain a turn. Yet, most of studies employ methodology of conversation analysis to examine the ways in which multimodal resources are deployed for interactional work at the beginnings of turns, very few of them have examined the use of multimodal resources by EFL learners at turn beginning by adopting corpus-based approach, especially the Chinese EFL learners.
To address this lacuna, the current study investigated turn beginning features in Chinese EFL learners’ interactions, and examined the similarities and differences in the modal resources used by the intermediate and advanced learner groups, to explore its developmental pattern. It can prevent “lingual bias” (only focus on linguistic structures) from multimodal interaction perspective (Couper-Kuhlen, 2018). Meanwhile, it can widen the research scope and more plainly show the function of multiple modes served in turn-taking and social interactions (Pekarek-Doehler & Skogmyr-Marian, 2022). The ultimate purpose of this study is to reveal the general modal use rules of learners’ accomplishment of turn beginning, and provide inspiration for their interactional competence development.
2.1 Multimodal Interaction
The core tenet of interactional linguistics is that any language and language behavior in conversation exist in specific interactive context. Therefore, the structure and use of language need to be understood from social interaction, which is the inevitable result of interactive context (Fang Mei et al., 2018; Pekarek-Doehler & Berger, 2018). In daily face-to-face conversation, participants may use both linguistic and non-linguistic resources to project their communication intention, to fulfill communicative purpose and to accomplish conversational behavior. These multimodal resources are gaze, facial expression, gesture, head movement and body posture and so on. In other words, interactive communication is essentially a multimodal display (Couper-Kuhlen & Selting, 2001:1). Informed of this idea, multimodal interaction primarily refers to the comprehensive investigation of various modal resources that constitute interaction, so as to understand the nature and panorama of human interaction (Stivers & Sidnell, 2005; Li, 2019: 45).
In a conversation, each mode is a separate symbol system with equal status, which are able to express meaning in an appropriate context and accomplish communicative purposes. The realization of turn beginning includes three types of resources: lexis and syntax, prosody and body movement (Stivers & Sidnell, 2005). Lexis and syntax are the most fundamental resources for interpersonal interaction. Prosody significantly has an impact on the turn beginning and progression of a conversation, such as pitch, intensity, intonation and pause. Body movement, such as facial expression, gaze, posture, hand movement and head movement, are important resources for conversation participation and meaning expression. In order to understand how Chinese EFL learners employ multimodal resources to achieve turn beginning, this study mainly examines the distribution of lexis, syntax and body movement, leaving prosody to another paper because of its complexity. Among these, lexis and syntax are categorized into linguistic resources and body movement into non-linguistic resources.
2.2 Multimodal resources and turn beginning
2.2.1 Linguistic resources
Scholars have studied linguistic feature at turn beginning in native English speakers or German speakers’ conversations (Li, 2003; Tao, 2003; Deppermann, 2013; Koops & Lohmann, 2015; Haselow, 2019). Turn-initiators in English are lexical in nature and take the shape of discourse markers, such as and, oh, well, yeah, yes, and so forth (Tao, 2003). Li (2003) classified turn beginning strategies in English conversation into the following three linguistic categories: using small particles (hmm, yeah); providing feedback to the speaker’s questions; and completing or supplementing the previous turn. This classification is helpful and can serve as a reference to divide the turn beginning types in EFL learners’ conversation. Moreover, Deppermann (2013) divided linguistic means of turn beginning into three categories: feedback, connective and syntactic relevance. Among them, feedback alludes to response to the previous turn, establishing interactional function and showing their respect to the speakers’ views. For example, the use of yeah/yes implies hearer’s approval of the previous speaking content; Oh denotes the change of its knowledge state. Connective refers to the phenomenon in which a speaker utilizes the initial element at turn beginning to establish a connection with the preceding turn, to facilitate effective communication and ensure a smooth flow of dialogue. For instance, the relationship between the current and previous turn might be shown by the use of connectors like and and but (complement or contrastive). Syntactic relevance indicates that the syntactic structure of the current turn is based on the previous turn, and two of them are closely related in content and meaning. Koops and Lohmann (2015) proposed an “optimal sequencing hierarchy for DM sequences”, which is as follows: oh > well > and > so > or > but > because > then > you know > now > I mean. Haselow (2019) made a classification of functional domains of discourse markers, which are interaction, discourse structure and cognition.
However, few studies have been conducted to explore turn beginning features in second language learners’ conversations (Gardner, 2007; Thörle, 2016; García-García, 2021). Scarcely studies focus on the repeated and disfluent turn starts (Schegloff, E. A. 1987), which was further identified as the purposeful strategies used by second language learners rather than the lack of ability (Gardner, 2007). The findings revealed three primary functions of the deliberate repetition and disfluency in initiating turns. Firstly, they served to confirm that the turn commencement is achieved smoothly, devoid of interruptions. Secondly, they were instrumental in securing the hearers’ focused attention to the speaker’s message, as emphasized by Olsher (2004). Lastly, they were crucial for establishing visual recipiency (Carroll, 2004; Lam, 2021). Li & Xu (2023) analyzed the sequence of 240 turn-initial discourse makers in Chinese EFL learners’ conversations. Their study yielded that Chinese EFL learners exhibit similar sequencing preferences to native English speakers, indicating their proficiency in acquiring the fundamental discourse marker functions of facilitating interpersonal interaction, structuring discourse and marking meta-cognitive stance. Konzett-Firth (2023) focused on the development of responsive turn beginnings in L2 French peer interactions. The results showed that broken starts and a heavy reliance on sequential or linguistic affordances occurred at early stages. Later, the focal learner started using routinized expressions to deal with turn beginnings in increasingly less predictable sequential environments. Over time, they developed more flexibility and mobilized a greater variety of resources in managing uptake and projections in their responsive turns.
Previous studies have revealed the linguistics categories, sequencing hierarchy and functional domains of turn beginnings, providing some insights to the present study. However, whether Chinese EFL learners’ turn beginning features in accordance with results of these studies still needs to explore, which will enrich the existing research.
2.2.2 Non-linguistic resources
The majority of scholarly work on non-linguistic resources at turn beginnings has concentrated on an array of interactive communication contexts, employing the conversation analysis method to explore scenarios ranging from everyday natural dialogues to institutional interactions and second language conversations.
In the realm of everyday natural conversations, research has unveiled that participants utilize bodily movements to indicate their intention to initiate a turn. Specifically, in naturalistic Chinese interactions, individuals harness a repertoire of non-verbal cues to convey both their active participation and impending speakership. These include, but are not limited to, gaze, gesture, facial expression, and body posture, all of which serve as subtle yet powerful indicators of their communicative intent (Li, 2019). Besides, participants generally form mutual gaze at turn beginnings (Goodwin, 1980). In Ilokano interaction in the Philippines, participants usually employed a specific facial expression (the facial expression accompanying the vowel [a]) to indicate that they are about to take turns and start to speaking. This subtle facial expression was called “physical turn-entry devices” by Streeck and Hartge (1992: 143). According to Kaukomaa et al. (2013), a smile could also indicate the upcoming turn and the shift of participants’ emotional states from neutral or serious to positive and amusing. Also, Kaukomaa et al. (2014) found that when participants frown at the pre-beginning of turns, it indicates the turn starts. Meanwhile, it also projects the current speaker’s words deviating from what they had anticipated. Apart from facial expression, gesture also serves as a signal of turn beginnings and could also simultaneously convey the content of the turns. Specifically, in order to indicate their turn beginning intentions in storytelling, participants will raise up the palm of one hand and use the index finger of the other hand pointing to it before beginning of their turns (Streeck & Hartge, 1992). Streeck (2009) also referred gestures that can “foreshadow” the turn beginnings as forward-gesturing.
Within the context of institutional interactions, bodily movements emerge as pivotal resources, serving critical interactional functions. For example, in the intimate setting of a small-scale, three-person work meeting, participants often deploy pointing gestures as a preface to initiating a turn, thereby signaling their imminent speakership (Mondada, 2007). In the medium-sized round table meetings, participants often lean forward and gaze at the current speaker to signal their turn beginning intentions (Ford & Stick, 2012). Moreover, in the doctor-patient interaction, it usually results in the starts of doctor’s turn when the patient’s body is turning to, with eyes simultaneously gazing at the doctor (Heath, 1986: 27–28). Mortensen (2009) observed that speakers strategically position themselves within the visual field of potential interlocutors, thereby prompting the initiation of a conversational turn. This behavior underscores the subtle yet powerful influence of body movement in orchestrating the flow of dialogue.
In terms of second language conversations, Sahlstrom (2002) found that second language learners typically indicate their turn-taking intentions by hand-raising. Specifically, this strategy can be used not only to project the hearer’s desire to select himself or herself as the next speaker, but also to successfully acquire the speakership without overlapping or interrupting in the current speaker’s turn while it is in progress (Mondada, 2007; Streeck, 2009). Lee (2017) discovered that learners’ self-selection turns frequently begin with hand forwarding gestures. In addition, learners often project turn start intentions by leaning forward and gazing at the current speaker. Therefore, the significance of non-verbal turn-beginning devices in the orchestration of turn-taking is paramount and cannot be overstated.
However, previous studies still have some shortcomings. Firstly, the number of research samples is insufficient, and the methods used are often limited to introspection or conversation analysis, the absence of corpus analysis of dialogues prevents us from discerning the general rules of turn beginnings in learners’ interactions; Secondly, the focus of many studies has been on the turn beginning characteristics of a single group, neglecting to compare learners at different proficiency levels to explore the developmental trajectory of these characteristics; Thirdly, while the preponderance of learners’ conversations captured in the oral test context does offer insights into their distinct interactional traits, the lack of observing interactions in authentic natural settings hinders a comprehensive and profound exploration of the conversational attributes of learners.
Thus, in order to address these deficiencies, the present study increase the sample size and utilize corpus analysis to provide valuable insights into the patterns of recourses used by Chinese EFL learners at turn beginnings. Additionally, a comparative analysis of turn beginning characteristics among learners at the intermediate and advanced proficiency levels is also conducted. This study could enhance our comprehension of the multimodal characteristics inherent in learners’ turn beginnings and enriches the extensive scholarly discourse on quantitative facets of it.
3.1 Research questions
(1) What linguistic features are present as Chinese EFL learners initiate turns? What are the commonalities and divergences in these linguistic features between the intermediate and advanced proficiency levels?
(2) What non-linguistic features are present at the onset of Chinese EFL learners’ turns? What are the similarities and differences in these non-linguistic features between the intermediate and advanced proficiency groups?
3.2 Participants
The participants in this study are undergraduate and graduate students from a normal university in China. A total of 84 participants were enrolled in this study, comprising 44 undergraduate and 40 graduate learners. The average age of undergraduates is 18.07 years old (SD = 0.76; Range = 16–20 years old); Their average duration of English study year is 9.77 years (SD = 2.20; Range = 5–15 years); Their average college entrance examination score is 131.68 points (SD = 7.40; Range = 108–143 points). The average age of graduates is 23 years old (SD = 1.48; Range = 21–27), their average English study year is 13.63 years (SD = 2.23; Range = 10–18 years), and their average TEM-8 score is 70.53 points (SD = 4.92; Range = 65–79 points). Compared with undergraduates, graduates have passed TEM-8 and those who can reach 60 points are identified as advanced EFL learners in China. Consequently, this study delineates the proficiency levels by categorizing undergraduates as intermediate proficiency group and graduates as advanced proficiency group.
3.3 Data collection
The data used in this study were video-recorded casual conversations among Chinese EFL learners during an extracurricular English corner event. Recruitment of participants took place in September 2020, at which time informed consent was secured from each participant. They volunteered to participate with great zeal. Prior to the recording, learners were not made aware of the objectives and specific instructions of the study. The researcher merely informed that the content and process of their conversation would be recorded, emphasizing the importance of behaving as authentically and naturally as possible. The data collection procedure is outlined as follows.
Firstly, the researcher required learners to pair up in accordance with their own preferences, which can alleviate the tension and anxiety during the recording and improve the quantity and quality of conversational output. Secondly, the topic discussion mode was adopted. Ten non-limited topics were provided, such as friends, travel and traditional Chinese festivals, which were freely chosen by learners and had been delivered to them in advance for preparation. To facilitate comprehensive interaction among learners, the topic serves merely as a guiding framework for discussion, with its content being intentionally unrestricted and open to interpretation. Thirdly, learners’ upper body parts were recorded by a close set-up video so as to concentrate on the analysis of gaze, gesture, head movement, facial expression and body posture. Lastly, the recording process consisted of three phases: preparation, warm-up and formal recording. In the preparation phase, the researcher set up the video and audio-recorded equipment in advance, made the scene atmosphere active to help learners gradually adapt to the recording setting. In the warm-up phase, the researcher withdrew from the scene, leaving only two learners, who introduced their interests or family members to each other to prepare for the topic discussion. The transition from warm-up to formal recording phase is seamless, enabling learners to autonomously determine the moment to delve into topic discussion, which commences upon one participant’s introduction of the discussion topic. Data garnered during the warm-up phase is intentionally omitted from subsequent analysis, focusing the examination solely on the topic discussion content.
Overall, the duration of topic discussions ranged from 5 to 15 minutes for intermediate proficiency group and extended to 13 to 18 minutes for advanced proficiency group. The cumulative conversation time for the two groups was 202 minutes and 295 minutes, respectively, summing up to a total of 497 minutes. The number of turns produced by intermediate and advanced proficiency groups were 822 and 959, respectively, culminating in a combined total of 1,781 turns. In terms of word count, intermediate and advanced proficiency groups contributed 20,448 and 31,759 words, respectively, amassing a grand total of 52,207 words.
3.4 Data transcription and annotation
The principle of “truthful transcription” is followed in this study when converting audio recordings into text version in accordance with Jefferson (2004). It involves accurately transcribing each participant’s original sentences, avoiding mistakes, and making no modifications of the words, phrases, sentence patterns, and so forth to maximize the original appearance of the corpus. The linguistic features at turn beginnings are annotated using the AnnoTool software by focusing on lexical and syntactic elements. This software allows researchers to modify annotation dimensions based on their own needs. The non-linguistic features are examined using the multimodal annotation software ELAN, focusing on body movement elements. It can integrate and display various categories of multiple modes on a window interface and operating platform. Researchers can clearly understand the contents of annotations at a glance, facilitate observation and analysis of the interaction between modes, and realize unified retrieval of multiple annotation files.
In order to ensure the consistency and accuracy of multimodal annotation, this study conducted three rounds of proofreading and annotation. After the first round of annotating is complete, the second round and the third round would be carried out one month and three months later. The internal reliability of annotation could be improved over these three rounds of annotation and eventually reached more than 85%.
3.5 Data analysis
Following the annotation of data, the researcher conducted a frequency analysis of both linguistic and non-linguistic resources and documented the results in an Excel spreadsheet. To discern significant disparities among various strata and groups, a Log-likelihood ratio (LL) test was employed for comparative analysis. Among them, the significance level is Sig (p) Value<0.05 with critical value 3.84; the significance level is Sig. (p) Value<0.01 with critical value 6.63; the significance level is Sig. (p) Value<0.001 with critical value 10.83.
This part reported the linguistic and non-linguistic traits at turn beginnings in Chinese EFL learners’ conversations.
4.1 Linguistic features
Deppermann (2013) classified the linguistic means of turn beginnings into three categories: feedback, connective and syntactic relevance. Haselow (2019) made a classification of functional domains of discourse markers into interaction, discourse structure and cognition. Based on their classification, along with the observation and analysis of learners’ multimodal data, this study classified the linguistic resources into four categories with its relevant functional domains. Table 4.1 described the detailed classification of them.
| Types | General functions | Examples | Excerpts in learners’ conversations |
|---|---|---|---|
| Feedback | Interaction: managing speaker-addressee interaction, usually used as back-channeling signal, expressing approval and recognition | yes, yeah, ok, well, oh, ah, no, right, exactly, etc. | Bai: hum Recently uh I think my study is so hard and I was tired. uhm (0.7) How about you? Zhu: (0.5) hum Yes uh because the: uhm college entrance examination is approaching. |
| Connective | Discourse structure: indicating the type of (local, global) relationship between discourse units or between a discourse unit and implied meanings or inferences | and, because, but, or, although, also, if, etc. | Chang: …have you ever heard of the expression that good good study, day day up. Do you think it is right? Qu: (0.3) Yeah I think it is right. Chang: (0.8) hum But actually this expression have a big (0.5) mistake in English grammar. |
| Demonstrative | Cognition: providing a cue for utterance-interpretation | this (is), that, these, you (know), I (think/have/like), we (have), our, it, etc. | Song: uh So I think if the human society goes on as as hum as this as this hum our human will completely lose us uh lose our emotion and virtue. Han: (0.8) This is incredible! But after all, it is just a fictional movie. |
| Syntactic relevance | Interaction and Cognition: organizing turn-taking and interaction, and making inferences. | syntactically build on prior talk (elliptical sentences) | Zhao: (1.0) hum (1.0) He is very cute[hum]. That means (0.9) I am a member of ET. Chen: (0.5) ah ET (0.4)? What do you mean about it? Zhao: (1.0) It just (0.3) aliens. Chen: (0.7) aliens. |
Among them, feedback type refers to begin turns by providing feedback on the previous speaker’s turn (usually expressing approval and recognition), serving interactional function. For example, Yes in the excerpt is used by Zhu as a turn-acceptance word to confirm the content of the previous turn that she also thinks study is so hard and she was tired, achieving interaction with Bai. Connective type means that learners primarily employ connectors to establish and highlight the logical connection of the current and previous turn, to achieve discourse structuring purpose. For example, but in the excerpt used by Chang is to indicate the contrast or comparison relationship of her opinion towards the maxis. Demonstrative type mainly includes the use of personal pronouns, deixis and articles. The purpose of it is to identify the content of the previous turn or learners’ own stance, and the main concern is the beginning of own turn. It serves cognition function, to provide a cue for utterance-interpretation. For instance, this in the excerpt used by Han is to refer to the content of the previous turn, and she make a comment on “this”, which is incredible, to provide a cue for the following utterance-interpretation. Syntactic relevance type refers to that the current turn is closely related to previous turn in both structure and content, presenting as elliptical sentences. In other words, the current turn gets muddled and challenging to interpret if it was cut off with the preceding turn. It serves a dual purpose in both interaction and cognition, highlighting the importance of turn-taking organization and facilitating inferencing processes. For example, Chen in the excerpt use “ah ET?” as an elliptical sentence to replace “you are a member of ET?”, and “aliens” as an elliptical sentence to replace “It refers to aliens”.
Based on the aforementioned classification criteria and framework, this study analyzed the linguistic resources employed by learners, and examined the similarities and differences between intermediate and advanced learner group, as presented in Table 4.2.
Table 4.2 Frequency of linguistic resources
Overall, feedback was used as primary resources, with a total of 620 cases, accounting for 55.7%; it was followed by demonstrative (233 cases; 20.9%), connective (156 cases; 14.1%), and syntactic relevance (104 cases; 9.3%). The test showed that feedback significantly outperformed connective, demonstrative and syntactic relevance (LL = 296.93, p = 0.000; LL = 182.16, p = 0.000; LL = 407.78, p = 0.000); Demonstrative significantly outperformed connective and syntactic relevance (LL = 15.34, p = 0.000; LL = 50.66, p = 0.000); Connective significantly outperformed syntactic relevance (LL = 10.47, p = 0.001). Thus, the linguistic traits were: feedback > demonstrative > connective > syntactic relevance. It indicated that learners prioritized interactional function, and then cognitive function, and then discourse structuring function at turn beginnings.
Seeing from different language proficiency, in terms of type, feedback was the most frequently used resources in intermediate learner group, with 301 times and accounting for 54.1%. It was followed by demonstrative, connective and syntactic relevance, and their use frequency were 132 times (23.7%), 77 times (13.8%) and 47 times (8.4%), respectively. The advanced learner group showed the same trend with intermediate group. Specifically, feedback was the most dominant aspect, followed by demonstrative, connective and syntactic relevance. The use frequency of them was 319 times (57.4%), 101 times (18.2%), 79 times (14.2%) and 57 times (10.2%), respectively. The results revealed that learners preferred to give feedback on the previous turn to serve interactional function at turn beginnings despite different language proficiency.
However, intermediate learner group used more demonstrative than advanced group (23.7%>18.2%), while advanced learner group used more feedback, connective and syntactic relevance resources than intermediate learner group (57.4%>54.1%; 14.2%>13.8%; 10.2%>8.4%). However, LL tests showed that no significant differences existed in the use frequency of feedback, connective and syntactic relevance between the two groups (LL = 0.56, p = 0.456; LL = 0.03, p = 0.864; LL = 0.98, p = 0.322), while intermediate learner group used significantly more demonstrative than advanced group (LL = 4.08, p = 0.043). it was found that intermediate learner group used demonstrative words, such as it, that, you, to establish connection between the previous and current turns. They also used some demonstrative words to directly start their turn like I, me, we, reflecting that they concentrated more on the content of self-turn. In contrast, advanced group tended to interact with previous turns, and focused on establishing a logic relevance of the current and previous turn through connectors like and, but, because.
To conclude, overall, the distribution characteristics of lexical resources are: feedback > demonstrative > connective > syntactic relevance. The test showed that intermediate learner group used more demonstrative to provide a cue for utterance-interpretation and fulfill cognitive function. The advanced learner group emphasized establishment of the logical, syntactic, and content relevance of turns through use of the feedback, connective or syntactic relevance, fulfilling interaction and discourse structuring function.
4.2 Non-linguistic features
This study examined learners’ bodily movement at pre- and beginning stages of turns. The annotation domain spanned from pre-beginning to completion of the first turn construction unit (i.e. TCU), encompassing six layers: gaze, mutual gaze, facial expression, gesture, head movement and body posture. Approximately 42% data were extracted from conversations of the intermediate and advanced learner group (238 cases in each group, totaling 476 cases), with annotations and counts of their bodily movement resources conducted. The results are presented in Table 4.3.
Table 4.3 Frequency of non-linguistic features
In general, there were similarities and differences in the use rate of bodily movement at pre- and beginning stages of turns. The top two highly utilized modes were gaze and head movement, while modes such as facial expression, hand movement and body posture were less frequently used. The main difference lied in the ranking of these modes. At pre-beginning stage, gaze mode was ranked first followed by head movement. Facial expression fallen in third place, body posture in fourth, and hand movement had the lowest utilization rate. LL test showed that the use rate of the gaze mode (specifically, gaze at) significantly outperformed other modes such as head movement, facial expression, body posture and hand movement (LL = 48.58, p = 0.000; LL = 219.34, p = 0.000; LL = 222.16, p = 0.000; LL = 258.85, p = 0.000). Similarly, head movement was used more frequently than facial expression, body posture and hand movement (LL = 67.14, p = 0.000; LL = 68.84, p = 0.000; LL = 110.25, p = 0.000). Facial expression and body posture modes showed a higher utilization rate than hand movement (LL = 6.36, p = 0.012). However, the use rate of those modes showed no significant difference (LL = 0.01, p = 0.913). In summary, the five modes employed by learners at pre-beginnings stage were as follows: gaze > head movement > facial expression ≈ body posture > hand movement. This indicates that learners relied heavily on gaze and head movement to signal their turn-beginning intentions, while also incorporating facial expression, body posture, and hand movement.
However, at turn beginning stage, head movement mode ranked first, followed by gaze mode. Among the last three modes with low utilization rate, hand movement ranked third, body posture in fourth, and facial expression with the lowest utilization rate. The use rate of head movement significantly outperformed the gaze, hand movement, body posture and facial expression (LL = 89.47, p = 0.000; LL = 133.29, p = 0.000; LL = 183.83, p = 0.000; LL = 266.04, p = 0.000); The use of gaze mode was significantly more than the use of hand movement, body posture and facial expression (LL = 4.67, p = 0.031; LL = 18.38, p = 0.000; LL = 52.91, p = 0.000); The use rate of hand movement mode significantly outperformed body posture and facial expression (LL = 4.56, p = 0.033; LL = 26.66, p = 0.000); The use rate of body posture mode significantly outperformed facial expression (LL = 8.32, p = 0.002). Therefore, the five modes learners employed at turn beginnings were as follows: head movement > gaze > hand movement > body posture > facial expression. It can be seen that learners primarily employed head movement and gaze mode, supplemented by facial expression, body posture and hand movement to accompany the beginning of turns.
Moreover, LL test showed that the use rate of gaze at and mutual gaze mode at pre- beginnings stage was significantly higher than that at turn beginnings (LL = 51.79, p = 0.000; LL = 45.86, p = 0.000); however, the use rate of head movement, hand movement and body posture significantly increased at beginnings as compared with pre-beginnings (LL = 85.28, p = 0.000; LL = 68.15, p = 0.000; LL = 15.10, p = 0.000); there was no significant difference in the use rate of facial expression mode at pre- and beginning stage (LL = 0.54, p = 0.463). It can be seen that learners’ employment of body movement had slightly changed at pre- and beginning stage of turns. Furthermore, learners primarily employed gaze mode to project their turn starts intentions, while gaze shift, head movement, hand movement and body posture accompanied with turn beginnings. Nevertheless, facial expression mode remained the same state at pre- and beginning stage.
Seeing from different language proficiency, in terms of gaze, 63.4% learners in intermediate learner group gazed at hearer at pre-beginnings, and 58% of them formed mutual gaze; Only 27.3% learners gazed at hearer and 27.3% of them formed mutual gaze at beginnings. It is evident that the utilization rate of the gaze mode in the intermediate learner group decreased. The advanced group showed the same trend. At pre-beginning stage, 62.6% learners gazed at hearer, and 61.3% of them formed mutual gaze. However, at turn beginnings, only 35.3% learners gazed at hearer, and 33.6% of them formed mutual gaze. LL test showed that there was a significant difference in the use of gaze mode at pre- and beginning stage of turns both in intermediate and advanced learner groups (LL = 35.21, p = 0.000; LL = 26.85, p = 0.000; LL = 18.38, p = 0.000; LL = 19.56, p = 0.000). However, there was no significant difference in the use rate of gaze mode between these two groups (LL = 0.01, p = 0.908; LL = 0.23, p = 0.635; LL = 2.43, p = 0.119; LL = 1.55, p = 0.212). It can be seen that the use rate of gaze mode showed the same trend despite different language proficiency. Learners preferred to use gaze away to project their turn beginning intentions.
In terms of head movement, 33.2% and 79.0% learners in intermediate learner group had head movement at pre- and beginning stage of turns. 31.9% and 71.8% learners in advanced learner groups had head movement at pre- and beginnings. LL test showed that head movement use rate was significantly increased at turn beginnings in both groups (LL = 45.82, p = 0.000; LL = 37.50, p = 0.000). However, there was no significant difference in the use rate of head movement between the two groups both at pre- and beginnings (LL = 0.06, p = 0.810; LL = 0.81, p = 0.370). It can be seen that head movement use rate showed the same trend despite different language proficiency. Specifically, head movements at turn beginnings significantly outperformed pre-beginnings, which projects and accompanies the turn starts.
In terms of facial expression, 6.7% and 8.0% intermediate learner group changed their facial expressions at pre- and beginning stage of turns, respectively. It can be seen that the use rate of facial expression mode increased slightly at beginnings. The advanced learner group showed the same trend. Specifically, 10.9% and 12.6% learners had facial expression changes at pre- and beginnings, respectively. LL test showed that both intermediate and advanced learner group had no significant difference in the use rate of facial expression at pre- and beginnings (LL = 0.26, p = 0.612; LL = 0.29, p = 0.593). There was no significant difference in the use rate of facial expression mode between the two groups (LL = 2.40, p = 0.121; LL = 2.49, p = 0.115). It can be seen that the use rate of facial expression mode showed the same trend despite different language proficiency. That is, facial expression was not an important mode to project the turn beginnings.
In terms of hand movement, 5% and 15.1% intermediate learner group have hand movement at pre- and beginning stage of turns. It can be seen that the use rate of hand movement mode increased significantly at turn beginnings. The advanced learner group showed the same trend. Specifically, 4.2% and 32.8% learners had hand movement at pre- and beginning stage of turns. LL test showed that both intermediate and advanced learner group had a significant difference in hand movement mode at pre- and beginnings of turns (LL = 12.56, p = 0.000; LL = 59.68, p = 0.000). There was no significant difference in the use rate of hand movement mode between two groups at pre-beginnings (LL = 0.18, p = 0.670). However, there was a significant difference in the use rate of hand movement mode between the two groups at turn beginnings (LL = 15.84, p = 0.000). It can be seen that the use rate of hand movement mode by learners of different language proficiency had similarities and differences. That is, the intermediate and advanced learner group used hand movement mode to project turn beginnings. But the use rate of it in advanced learner group was higher than that in intermediate learner group.
In terms of body posture, 9.7% and 19.7% intermediate learner group have body posture changes at pre- and beginnings of turns. It can be seen that the use rate of body posture mode increased significantly at turn beginnings. The advanced learner group showed the same trend. Specifically, 7.6% and 15.5% learners had body posture changes at pre- and beginnings. LL test showed that both intermediate and advanced learner group had a significant difference in body posture at pre- and beginnings (LL = 7.82, p = 0.000; LL = 6.70, p = 0.010). There was no significant difference in the use rate of body posture mode between the two groups (LL = 0.61, p = 0.434; LL = 0.98, p = 0.323). It can be seen that the use rate of body posture mode showed the same trend despite different language proficiency. That is, the use rate of body posture at turn beginnings significantly outperformed pre-beginning stage. Thus, the change of body posture projects and accompanies turn starts.
To conclude, firstly, learners predominantly utilized gaze and head movement as their primary non-linguistic resources, with facial expressions, body posture, and hand gestures serving as secondary means of expression. At pre-beginning stage of turns, the order of preference for these modes was as follows: gaze > head movement > facial expression ≈ body posture > hand movement. At turn beginnings, the hierarchy shifted as follows: head movement > gaze > hand movement > body posture > facial expression. Secondly, it was observed that gaze shift, head movement, hand movement and body posture modes all tended to project or accompany turn starts, whereas facial expressions showed a minimal correlation with this behavior. Lastly, learners’ use rate of body movement showed the same trend despite different language proficiency. However, it is noteworthy that advanced learner group exhibited a higher use rate of hand movement at the start of turns compared to the intermediate learner group.
5.1 Linguistic features
The beginning position of a turn is of paramount importance (Kim & Kuroshima, 2013). At this juncture, certain linguistic elements typically emerge here, which is usually lexically and semantically associated with completion position of the previous turn (García-García, 2021). These elements play a dual role: first, they establish a link between the current turn and its predecessor, and second, they offer a preview of the upcoming content. It is evident that the strategic placement of linguistic resources at the beginning of a turn is essential for maintaining coherence and guiding the flow of conversation.
Firstly, this study has found that the distribution characteristics of linguistic resources can be categorized as follows: feedback > demonstrative > connective > syntactic relevance. This hierarchy suggests a certain order of importance in the utilization of linguistic resources during communication, indicating that learners prioritize interactional function, followed by cognitive function and discourse structuring function in conversations (Haselow 2019). Tao (2003) analyzed turn initiators in spoken English using corpus analysis, and found that lexical resources are the main form at turn beginnings. In addition, the feedback words account for the first place, with the highest frequency of oh and yeah. While, connective words like and and so are in second place. Demonstrative words like this and that rank third in frequency. Deppermann (2013) observed that participants consider feedback as the primary factor when starting a new turn in German native speakers’ interaction. Kim & Kuroshima (2013) identified various lexical components at the start of turns, including discourse markers, participles, feedback, connectives, appellations, adverbs, and specialized turn-beginning structures such as I think. Furthermore, Li & Xu (2023) investigated 240 turn-initial discourse makers sequence in conversations of the Chinese EFL learners. Their study yielded that Chinese EFL learners can acquire the fundamental discourse marker functions of facilitating interpersonal interaction, structuring discourse and marking meta-cognitive stance.
The results of this study are consistent with the findings of the previous study. The words at learners’ turn beginnings predominantly encompassed the lexical terms mentioned earlier, reflecting learners possessed a certain level of turn-acceptance ability and skills. It was evident that despite the participants’ diverse mother tongues (Chinese, English, and German), feedback played a central role in initiating and structuring conversation turns. Feedback served primarily an interactional function, as illustrated by the use of Yes by Zhu to acknowledge and affirm the statements of preceding speaker, Bai. This facilitated smooth communication and engagement with the interlocutor, as exemplified in the discourse between Zhu and Bai regarding the challenging nature of their studies. Moreover, other linguistic resources, such as connective and demonstrative, were employed to achieve distinct communicative functions and organize discourse effectively. Specifically, demonstrative aided in conveying cognitive functions. As shown in the examples, this in the excerpt used by Han was to refer to the content of the previous turn, and she maked a comment on “this”, which is incredible, to provide a cue for the following utterance-interpretation. Connective, on the other hand, contributed to the structuring of discourse and expressed relationships between ideas, as demonstrated by Chang’s use of but to introduce a contrasting viewpoint regarding the topic under discussion. Through the strategic deployment of these linguistic resources, participants are able to navigate the complexities of conversation and enhance the coherence and progression of dialogue. The nuanced use of feedback, connective, and demonstrative underscores the participants’ communicative competence and proficiency in managing conversational dynamics across different languages.
However, this study found that although learners used feedback as the main means despite different language proficiency, there were differences in the use of other two types of lexical resources between the intermediate and advanced learner group. Among them, the intermediate learner group tended to concentrate on the beginning of their own turns, preferring to establish lexical connection between the previous and current turns by using demonstrative. This indicated that they tended to establish cognitive function at the starts of turns. On the other hand, the advanced learner group concentrated on starting turns after interacting with the previous turn, and tended to establish the logical, syntactic and content relationships between the current and previous turn through the use of connective and syntactic relevance. The results indicated that they preferred to attain discourse structing function at turn beginnings. By analyzing these differences, we gain a better understanding of how learners at different proficiency levels utilize linguistic resources in their communication.
It is worthy to note that whether this finding applicable to other studies or other EFL learners’ interaction still needs further evidence to confirm. What is certain is that learners’ language proficiency is an important factor that affects their conversational performance. In general, interactional, cognitive and discourse structuring functions are the most important aspects of learners’ conversations. Nevertheless, the intermediate learner group tends to focus more on cognitive function, while the advanced learner group pays more attention to interactive and discourse structing function.
5.2 Non-linguistic features
Participants usually use non-linguistic resources to project their turn beginning intention at pre-beginning stage of turns, and turn beginnings usually accompanied with some changes in these multi-modes.
First of all, this study revealed that learners primarily utilized gaze and head movement, supplemented by facial expression, body posture and hand movement mode both at pre- and beginning stage of their turns. Moreover, the five types of modes used by learners at pre-beginnings were as follows: gaze > head movement > facial expression ≈ body posture > hand movement. At beginnings, these modes were presented in the following order: head movement > gaze > hand movement > body posture > facial expression. It indicated that learners had the awareness of using multi-modes to project turn beginnings intentions and started their turns despite different usage rate. The results of this study align with previous research on the conversational behaviors of native speakers. For example, in natural Chinese native speakers’ conversation, participants used multiple resources, such as gaze, gesture, facial expression and body posture, to demonstrate their engagement in interaction (Li, 2019). Yang (2011) discovered that participants in conversations obtained turns by gazing at and touching the recipients simultaneously. Mortensen (2009) similarly observed that speakers adjusted their body position within the visible range of the recipient (particularly the potential turn receiver) to induce turn starts. In second language conversations, Sahlstrom (2002) noted that in classroom interactions, learners typically showed their turn-taking intentions by raising hands. By employing these strategies while the previous speaker is still speaking, participants could project their intention to take the floor next by self-selection, thereby facilitating a smooth transition in turn-taking without overlapping or interrupting the ongoing speaker’s turn (Mondada, 2007; Streeck, 2009).
Secondly, learners primarily employed gaze mode to project their turn beginning intentions, as previous studies have found that speakers tended to establish mutual gaze at the start of their turns (Goodwin, 1980). During interactions at the medium-sized round table, participants often employed body movements like leaning forward and directing their gaze towards the current speaker to signal their readiness to speak next (Ford & Stick, 2012). In doctor-patient interactions, when the patient’s body faced the doctor and their eyes gazed at the doctor simultaneously, it typically cued the doctor to begin speaking (Heath, 1986: 27–28). Since gaze and head shifts frequently occur simultaneously, there is a consistency between the gaze and head movement modes at learners’ turn beginnings. This phenomenon suggests that participants from diverse cultural backgrounds adhere to specific physical and social norms, and they rely on gaze mode as the primary means to convey their turn initiation intentions.
Thirdly, gaze away, head movement, hand movement and body posture modes were accompanied with turn beginnings; facial expression mode remained in the same state at pre- and beginnings, showing little correlation with turn beginnings. However, previous studies have found that apart from the above four modes playing an important role in turn starts, facial expression is also an important projection mode. For instance, in the Ilokano interactions in the Philippines, participants often employ a specific facial expression (the one accompanying the vowel [a]) to signal their intention to speak next. This subtle facial expression has been referred to as “gestural turn-entry devices” by researchers (Streeck & Hartge, 1992: 143). Kaukomaa et al. (2013) found that smile can also indicate the upcoming turn in conversation, reflecting the change of participants’ emotional state from neutral or serious to positive and humorous. Kaukomaa et al. (2014) discovered that participants might frown before taking a turn, indicating their readiness to speak and projecting deviation of current speaker’s words from their expectations. The different findings between this study and previous studies could be attributed to variations in social and cultural contexts. Specifically, influenced by the Chinese culture, Chinese EFL learners may exhibit shyness and passivity in English conversations. In addition, their language proficiency and psychological readiness may not enable them to engage in free-flowing conversations, resulting in stiff overall performance and often expressionless facial expressions. Furthermore, learners may prioritize language accuracy and clarity in their expressions, overlooking the nuances conveyed by facial cues.
Finally, the use of hand movement mode increased at turn beginnings compared to pre-beginnings. Additionally, the use rate of hand movement in advanced learner group was higher than that in intermediate learner group at the beginning of turns. Previous research has shown that hand movement played a very essential role in turn starts. For example, gesture could project both the upcoming turn and its content. Specifically, participants often raised the palm of one hand and pointed with the index finger of the other hand towards the palm to signal their intention to speak and indicate that it is part of storytelling (Streeck & Hartge, 1992). Streeck (2009) also noted that gestures could foreshadow the beginning of turns, referring to such gestures as “forward-gesturing”. In institutional interactions, like small three-person working meetings, participants employed pointing gestures to indicate that they were about to take a turn as the next speaker (Mondada, 2007). In second language conversations, learners’ turn beginnings were often accompanied by changes in hand gestures, typically using forward hand gestures (Lee, 2017: 677). The above phenomenon suggests that advanced proficiency learners employ more strategies than intermediate learners at turn beginnings, utilizing gestures to achieve their communicative purposes apart from gaze and head movement modes.
Turn beginning is a crucial location both in terms of its interactional functions and challenges posed to learners. It is also a sensitive moment where speaker establishes speakership and gains recipiency (Ford & Stickle, 2012), to further achieve interactional, cognitive and discourse structuring function (Haselow, 2019). Additionally, “would-be-speakers” (Mondada, 2007 p. 214) are under pressure to select the right moment to intervene and act quickly. Therefore, speakers must pay close attention to their turn starts and have necessary linguistic and non-linguistic resources at their disposal. This study examines the multimodality of learners’ turn beginnings by adopting a corpus-based approach. The findings are as follows:
In terms of linguistic resources, learners’ turn beginnings were mainly based on lexical resources, supplemented by syntactic resources. The distribution characteristics showed a hierarchy: feedback > demonstrative > connective > syntactic relevance. Overall, there was no significant difference in the use rate of linguistic resources between the intermediate and advanced learner group. Moreover, Chinese EFL learners exhibited similar lexical and syntactic features at turn beginnings compared to native speakers. However, the intermediate learner group tended to focus on initiating self-turns directly, establishing lexical connections and serving cognitive function between previous and current turns with demonstrative. The advanced learner group preferred to establish logical and syntactic relevance between each turn through the use of connective, to achieve interactional function. The findings indicate that higher language proficiency levels lead to more interactive turn-taking.
In terms of bodily movement, firstly, learners primarily utilized gaze and head movement modes, with supplementary use of facial expression, body posture and hand movement modes. During the pre-beginning stage, learners prioritized the use of gaze, followed by head movement, with facial expression and body posture being of equal importance, and hand movement mode being of least priority. However, at beginning stage, learners tended to focus more on head movement, followed by gaze, hand movement, body posture, and facial expression, respectively. Furthermore, the shift of gaze, head movement, hand movement and body posture often signaled the initiation of a conversation turn, while facial expression showed minimal correlation to this. Additionally, although the overall usage of body movement remained consistent across different proficiency group, advanced learner group tended to incorporate hand movement more frequently than intermediate learner group during the beginning stage. Overall, learners’ use of bodily movement aligned closely with that of native speakers, with an emphasis on gaze and head movement. As language proficiency increased, learners also began to utilize gesture more effectively for initiating conversation turns. However, there was a noticeable lack of emphasis on facial expressions, which warrants greater attention from learners.
The research is of great significance for understanding how learners utilize multimodal resources to initiate turns, and can contribute to the existing literature on multimodal turn-taking by learners. However, although this study employed corpus data to investigate learners’ uses of multimodal resources in peer interactions, the analysis only focused on frequency of the occurrence for each modality. Future research could provide more detailed explanations of the findings by adopting conversation analysis method and could further validate the results by including additional empirical data from interactions involving other EFL learners. Furthermore, potential future studies could also delve into the prosodic features at learners’ turn beginnings for a more comprehensive analysis.
1. Competing interests
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
2. Data availability
Data will be made available on request.
3. Ethical approval
The studies involving human participants were reviewed and approved by the Professor Committee of School of Foreign Languages, Northeast Normal University. The procedures used in this study adhere to the tenets of the Declaration of Helsinki.
4. Informed consent
The participants provided their written informed consent to participate in this study in September 2020.
5. Author contributions
Mengmeng Ji makes the conception and design of the study, acquisition of data, and analysis and interpretation of data. She also drafts the article and revises it critically for important intellectual content.
Ming Chen makes analysis of data and revises it critically for important content.
- Carroll, D. (2004). Restarts in novice turn beginnings: Disfluencies or interactional achievements? In Second language conversations (pp. 201–220). London: Continuum.
- Couper-Kuhlen, E. (2018). Finding a Place for Body Movement in Grammar[J]. Research on Language and Social Interaction, (1): 22-25.
- Couper-Kuhlen, E & Selting M. Introducing interactional linguistics[C]. In: Selting M & Couper-Kuhlen E (eds.). Studies in Interactional Linguistics. Amsterdam/Philadelphia: John Benjamins Publishing Company, 2001.
- Deppermann, A. (2013). Turn-design at turn-beginnings: Multimodal resources to deal with tasks of turn-construction in German[J]. Journal of Pragmatics, 46 (1), 91–121.
- Fang, Mei, Li, Xianyin & Xie, Xinyang. (2018). Interactional linguistics and Chinese studies in interactional perspective[J]. Language Teaching and Research, (3):1-16.
- Ford, C. E., & Stickle, T. (2012). Securing recipiency in workplace meetings: Multimodal practices[J]. Discourse Studies, 14 (1), 11–30.
- García-García, M. (2021). Turn-initial discourse markers in L2 Spanish conversations: Insights from conversation analysis[J]. Corpus Pragmatics, 5(1), 37-61.
- Gardner, R. (2007). ’Broken’ starts: Bricolage in turn starts in second language talk. In Language learning and teaching as social inter-action (pp. 58–71). Basingstoke, UK: Palgrave MacMillan.
- Goodwin, C. (1980). Restarts, pauses, and the achievement of a state of mutual gaze at turn-beginning[J]. Sociological Inquiry, 50 (3–4), 272–302.
- Greer, T & Potter, H. (2008). Turn-taking practices in multi-party EFL oral proficiency tests[J]. Journal of Applied Linguistics, 5(3): 297-320.
- Haselow, A. (2019). Discourse marker sequences: Insights into the serial order of communicative tasks in real-time turn production[J]. Journal of Pragmatics, 146: 1-18.
- Heath, C. (1986). Body movement and speech in medical interaction[M]. Cambridge, England: Cambridge University Press.
- Heritage, J. (2013). Turn-initial position and some of its occupants[J]. Journal of Pragmatics, 57, 331–337.
- Jefferson, G. (2004). Glossary of transcript symbols with an introduction. In Conversation analysis: Studies from the first generation (pp. 13–31). Amsterdam, John Benjamins.
- Kaukomaa, T., Peräkylä, A., & Ruusuvuori, J. (2013). Turn-opening smiles: Facial expression constructing emotional transition in conversation[J]. Journal of Pragmatics, 55, 21-42.
- Kaukomaa, T., Peräkylä, A., & Ruusuvuori, J. (2014). Foreshadowing a problem: Turn-opening frowns in conversation[J]. Journal of Pragmatics, 71, 132-147.
- Kim, H. R. S., & Kuroshima, S. (2013). Turn beginnings in interaction: An introduction[J]. Journal of Pragmatics, 57, 267–273.
- Konzett-Firth, C. (2023). On the development of interactional competence in L2 French: Changes over time in responsive turn beginnings in peer interactions[J]. Linguistics and Education, 75, 101176.
- Lam, D. M. K. (2021). Don’t turn a deaf ear: A case for assessing interactive listening[J]. Applied Linguistics, 42 (4), 740–764.
- Li, Xiaoting. (2019). Multimodal interaction and multimodal analysis of Chinese interaction[J]. Language Teaching and Research, 198(4): 45-59.
- Li, Weijing & Xu, Jiajin. (2019). A Study on the Combination Rules and Pragmatic Functions of Discourse Markers in Chinese College Students’ English Speech[J]. Foreign Language World, (5): 40-48.
- Lee, J. (2017). Multimodal turn allocation in ESL peer group discussions[J]. Social Semiotics, 27 (5), 671–692.
- Li, Lu (2003). Turn-taking: The Speech Mechanism in English Conversations [J]. Journal of Shenzhen University (Humanities and Social Sciences Edition), (3): 118-121.
- Mondada, L. (2007). Multimodal resources for turn-taking: Pointing and the emergence of possible next speakers[J]. Discourse Studies, 9 (2), 194–225.
- Mortensen, K. (2009). Establishing recipiency in pre-beginning position in the second language classroom[J]. Discourse Processes, 46 (5), 491–515.
- Olsher, D. (2004). Talk and gesture: the embodied completion of sequential actions in spoken interaction[C]. In: Gardner R & Wagner J(Eds.), Second Language Conversations. Continuum, New York, 221-245.
- Pekarek Doehler, S., & Berger, E. (2018). L2 interactional competence as increased ability for context-sensitive conduct: A longitudinal study of story-openings[J]. Applied Linguistics, 39 (4), 555–578.
- Pekarek Doehler, S., & Skogmyr Marian, K. (2022). Functional diversification and progressive routinization of a multiword expression in and for social interaction: A longitudinal L2 study[J]. The Modern Language Journal, 106 (S1), 23–45.
- Sacks, H., Schegloff, E. A., & Jefferson, G. (1974). A simplest systematics for the organization of turn-taking for conversation[J]. Language, 50 (4), 696–735.
- Sahlström, F. (2002). The interactional organization of hand raising in classroom interaction[J]. Journal of Classroom Interaction, 37(2): 47-57.
- Schegloff, E. A. (1987). Recycled turn beginnings: A precise repair mechanism in conversation’s turn-taking organization. In Talk and social organization (pp. 70–85). Bristol: Multilingual Matters.
- Schegloff, E A. (1996). Turn Organization: One intersection of Grammar and Interaction[C]. In: Ochs E, Schegloff E A & Thompson S A. Interaction and Grammar. Cambridge: Cambridge University Press: 52-133.
- Stivers, T & Sidnell, J. (2005). Introduction: Multimodal Interaction[J]. Semiotica, 156(1/4): 1-20.
- Streeck, J. (2009). Forward-gesturing[J]. Discourse Processes, 46 (2–3): 161–179.
- Streeck, J & Hartge, U. (1992). Previews: Gestures at the transition place[J]. The contextualization of language, 135-157.
- Tao, H. (2003). Turn initiators in spoken English: A corpus-based approach to interaction and grammar. In Corpus Analysis (pp. 187-207). Brill.
- Thörle, B. (2016). Turn openings in L2 French. An interactional approach to discourse marker acquisition[J]. Language, Interaction and Acquisition, 7(1): 117–144.
- Waring, H Z. (2011). Learner initiatives and learning opportunities in the language classroom[J]. Classroom Discourse, (2)2: 201-218.
- Yang, P. (2011). Nonverbal aspects of turn taking in Mandarin Chinese interaction[J]. Chinese Language and Discourse, 2(1): 99-130.
- Zheng, Yan & Luo, Dazhen. (2014). Turn-taking in English classroom group discussion[J]. Journal of the Chinese Society of Education, (12):59-62.
No competing interests reported.