Functions of Discourse Markers in Nonnative English Speech: The Case of Arab English Speakers

Abstract

This study examines the use and functions of discourse markers (DMs) in nonnative English speech produced by Arab English speakers. Four DMs (and, but, so, y’know) are analysed based on two theoretical frameworks: Schiffrin’s (1987) framework of functions of DMs and Schourup’s (1999) characterisation of DMs. Semi-structured interviews were conducted with 10 Arab English speakers. The findings show clear patterns in the participants’ use of DMs. The marker and is mainly used for its ideational function, that is, for connecting ideas, events, and positions to keep the discourse together, while pragmatic functions, such as continuing an action or turn organisation, are less represented. But is mainly used for its ideational function, indicating contrastive ideas; less frequently, it is used for such pragmatic functions as returning to a topic, making disclaimers, reclaims, or showing functional contrasts. In contrast, so demonstrates a broader range of functions; while it occasionally marks results at the ideational level, it is extensively utilised pragmatically for marking claims, compliance, requests, and topic transitions, as well as managing turn initiation and adjacency pairs. y’know is used for such pragmatic functions as organising shared knowledge, signalling significant information or disapproval in stories, and appealing. Overall, the findings in the study suggest that, in the narrative register, when using DMs, Arab English speakers rely chiefly on their ideational functions. In contrast, their pragmatic functions are used much less, except for so and y’know, which also show more diversified functions.

1. Introduction

Language acquisition is a complex process that involves not only mastering vocabulary and grammar but also developing the ability to use language effectively in real communication. Second Language Acquisition (SLA) refers to the process through which individuals learn a language other than their native one (Gass et al., 2020). Unlike first language (L1) acquisition, which occurs naturally through immersion from early childhood, SLA often takes place in structured learning environments and is influenced by cognitive, social, and linguistic factors (Gass et al., 2020; Lightbown & Spada, 2020). One crucial aspect of SLA is discourse competence, which involves the ability to organise language coherently and appropriately in communication (Celce-Murcia, 2007). One of the challenges faced by L2 learners of English is the development of discourse competence. This is basically the result of a lack of natural exposure to the L2 in childhood. Nonnative English speakers (NNESs) experience many problems that cause foreignness in their speech. Such problems relate to different aspects of language, such as phonology, syntax, and pragmatics. While native speakers can effortlessly communicate in a natural, spontaneous, and contextually appropriate manner, smoothly sustaining the flow of conversation, nonnative speakers often lack this degree of naturalness—even when they make deliberate efforts to achieve native-like performance (Götz, 2013).

The smoothness of speech by native speakers is attributed to many factors, including the use of discourse markers (DMs), self-repairs, hesitation fillers, and the ability to use formulaic expressions (Erman, 2007; Hasselgren, 2002; Müller, 2005). In addition, native speakers have certain temporary conditions in their speech, such as the length of their turns, the rate of pauses, the rate of speech, and mastering the intonation (Gut, 2009; Wells, 2006), whereas nonnative speakers often lack a full grasp of these features even in advanced stages. In classroom contexts, one factor that shapes Arab learners’ experience is the role of Arabic in instruction (Alzobidy et al., 2024). Al-Shahwan and Ahmad (2024), for example, found that university students valued their instructors’ use of Arabic as a medium of instruction in translation courses, as it supported comprehension and eased the cognitive demands of learning through English.

It is important to note that the native speaker (NS) vs. nonnative speaker (NNS) distinction has been problematised in recent applied linguistics scholarship (e.g., Alghazo & Zidan, 2019; Clymer et al., 2020). While these labels can reinforce deficit perspectives, we use them here in a limited and descriptive sense, primarily for comparability with earlier studies on discourse markers (e.g., Müller, 2005; Fung & Carter, 2007; Liu, 2017). Where possible, we also refer more specifically to “Arab English speakers” to highlight the learners’ linguistic and cultural background rather than framing them solely in terms of nonnativeness.

This study concerns itself with an exploration of how English DMs figure in nonnative English speech. The study explores how NNESs use DMs by analysing a corpus of natural speech by NNESs to provide a glimpse into how DMs can be best used to develop naturalness and spontaneity in their speech. DMs are expressions that show the relationship between parts of the discourse (Fraser, 1996). Their use contributes to discourse coherence, the way the participants connect forms, meanings, and actions to produce a thorough content of what is uttered (Schiffrin, 1987). Other functions of DMs include the arrangement and organisation of written and spoken discourse parts (O’Keeffe et al., 2011). The use of DMs is an aspect that reflects the pragmatic competence of the speaker. Native speakers of a language acquire this pragmatic competence in childhood, while nonnative speakers find it challenging to develop this competence at a later stage.

Many studies have investigated the use of DMs in the speech of both native English speakers (NESs) and NNESs to find similarities or differences in the use of DMs (e.g., Fung & Carter, 2007; Aijmer, 2011; Aşık & Cephe, 2013; Müller, 2005; Gilquin, 2008; Liu, 2017). Based on the findings of these studies, many factors were found to contribute to the use of DMs, such as age, gender, and oral proficiency. However, the verbal use of DMs by Arab speakers of English remains an area that has rarely been explored. Although the verbal use of DMs by Arab speakers of English has occasionally been examined (e.g., Kurdi, 2008; Algouzi, 2014, 2021; Al-Yaari, 2013), such studies remain limited in scope, often focusing on single markers, single national contexts, or classroom-based discourse. By contrast, extensive research exists on other learner groups, such as German (Müller, 2005), Chinese (Liu, 2017), Thai (Arya, 2020), and Finnish learners (Nikula, 1993), which underscores the relative scarcity of work on Arab learners. The originality of the present study lies in its systematic analysis of four discourse markers (and, but, so, you know) across data from multiple Arab countries, drawing on spontaneous, narrative-style interviews rather than controlled classroom interactions. This design makes it possible to uncover both the broader trends in Arab NNES discourse (e.g., reliance on ideational functions) and the more nuanced pragmatic uses of so and you know that have not been previously documented. In doing so, the study contributes new insights into the underexplored area of discourse marker use in Arab English, thereby extending the global conversation on DM use in second language acquisition.

This study concerned itself with exploring the use of DMs that occur in the speech of Arab speakers of English to identify the functions they achieve in their speech. The study seeks to answer the following questions:

(1)

What are the functions of DMs used by Arab speakers of English?

(2)

What similarities and/or differences can be observed in the use of discourse markers by Arab speakers of English when interpreted in light of patterns reported for native English speakers in prior studies?

While the primary focus of this study is on how Arab NNESs employ discourse markers in spontaneous speech, references to NES studies are included only as a comparative backdrop. In other words, this is not a direct comparative study with NES participants, but rather an exploration of NNES usage interpreted in light of established NES findings reported in the literature. While RQ 2 references similarities and/or differences between Arab speakers of English and native English speakers, it is important to note that the present study does not directly compare original data from NESs. Instead, NES usage patterns reported in prior research (e.g., Schiffrin, 1987; Müller, 2005; Fung & Carter, 2007; Liu, 2017) serve as a comparative backdrop against which the Arab speakers’ discourse marker use is interpreted.

Drawing on previous research, the present study advances several hypotheses. First, it is expected that Arab NNESs will rely more heavily on the ideational functions of discourse markers—particularly and and but—than on their pragmatic uses (Fung & Carter, 2007; Liu, 2017). Second, in line with findings from Arab EFL contexts (e.g., Algouzi, 2014, 2021; Kurdi, 2008), it is hypothesised that so and you know will demonstrate a broader range of pragmatic functions, such as signalling transitions, organising shared knowledge, or managing turn-taking. Finally, based on comparative evidence from NES and NNES studies (Schiffrin, 1987; Müller, 2005), it is hypothesised that the overall functional range of DMs among Arab NNESs will be narrower than that observed among NESs, reflecting learners’ greater reliance on ideational rather than interpersonal discourse functions.

2. Theoretical Framework

This study draws upon the frameworks of Schiffrin (1987) and Schourup (1999). According to Schiffrin (1987), DMs are used as deictic, and they are also used to signal utterances to prior and/or posterior utterances and the speaker and/or hearer. Schiffrin (1987) formed a discourse model with five planes to categorise DMs’ functions: exchange structure, action structure, ideational structure, participation framework, and information state. In this model, exchange structure refers to an outcome of communication by which participants take turns repeatedly and relate these turns to each other, such as answering a question. Schiffrin also defined action structure as a function used to show how speech acts are located, not just in relation to the speakers’ identities and social context but also in relation to the actions that come before, are supposed to follow, and do follow. Schiffrin (1987) views these two planes—exchange structure and action structure—as pragmatic.

The third discourse structure is the ideational structure, which refers to the content meaning of an utterance. In contrast, exchange and action structures can be obtained using pragmatic language. The fourth function is the participation framework, which concerns the relationship between the speaker and the hearer and the relations between the speaker and the utterance. The last plane is the information state, which analyses the speaker’s knowledge or meta-knowledge (Schiffrin, 1987). In contrast to the participation framework, which strongly emphasises the speaker and hearer’s social and interactional abilities, the speaker and hearer also significantly impact their cognitive abilities. For Schiffrin, this plane is also pragmatic. According to Schiffrin (1987), each DM primarily functions on a specific plane while also operating across multiple levels. The model examines several DMs and their roles across all five planes.

Table 1. Discourse markers and their functions.

Discourse Markers	Functions
And	- Building the text; connects ideas on all levels - Marking a conclusion of events; it marks the summary of events - It has contrastive meaning - Continuing the speaker’s action; fitting speakers’ talk into interactional slot, e.g., asking or answering questions, or continuing arguments - Smoothing and organising the turn-taking
But	- Giving contrastive meaning - Marking contrastive ideas - Marking functional contrast; responses contain what was requested, and other information was not requested - Marking referential contrast - Prefacing cooperative response - Marking disclaimers - Organising the turn-taking - Marking the speaker—return to the topic of the discourse - Marking self-repair - Marking reclaims - Marking as a point-making device in defense
So	- Marking a result of an extended piece of talk - Marking an inference - Marking an action (request, compliance, claim) - Organising the turns - Initiates turns - Marking topic transitions - Marking a conclusion of the discussion; once both the hearer and the speaker recognise the conclusion, both of them can start a new topic. Marking a completion of adjacency pairs. (question/answer)
y’know	- Organising and smoothing the turns - Indicating the relations between ideas; e.g., marking background information - Expressing and negotiating the relations between participants - Organising the knowledge and meta-knowledge - Marking the speaker’s seeking of the hearer’s endorsement - Marking the speaker’s seeking to invalidate another’s actions or description - Drawing the hearer’s attention to a very important point in order to understand why a story has been told - Drawing the hearer’s attention to a character’s dissatisfaction about something in a story

outlines the functions of and, but, so, and y’know, which are used in this study.

The study also utilises Schourup’s (1999) model of identification of DMs. Schourup (1999) presents a framework of features used to characterise linguistic structures as DMs. The first feature is connectivity, which means that DMs are expressions used to show the relationship between utterances or other parts of discourse. This feature of DMs is used to differentiate DMs from other elements, such as attitudinal adverbials (sadly, unfortunately), illocutionary adverbials (frankly, confidently), and primary interjections (oops, yipes) (Schourup, 1999). The second feature is optionality, which means that if the DM is deleted from an utterance, the meaning it refers to can still be obtainable by the hearer. Although DMs are optional, Schourup (1999) asserts that they enhance speakers’ intended meaning. Non-truth conditionality is the third feature of DMs; this means that DMs do not affect the truth conditions of the meaning implied in an utterance, unlike other words that affect the truth conditions, such as manner adverbials (sadly, fortunately). The other features are less consistently considered essential for determining DM status (Schourup, 1999). DMs have weak clause associations, meaning these markers appear outside the syntactic structure or are barely attached to it (Schourup, 1999). Another feature of DMs is initiality, which refers to their prototypical placement at the beginning of an utterance. This initial position helps ensure the correct interpretation of the utterance and prevents potential misunderstanding of the intended meaning (Schourup, 1999). The orality feature of DMs is based on the claim that they generally occur in speech and are mostly attached to the formal or informal contexts in which DMs are used. The last feature is multi-categoriality, often used for a functional category that is not harmonic with the syntactic class. Thus, Schourup (1999) includes many categories to which a DM function refers: “adverbs (e.g., now, actually, anyway), coordinating and subordinating conjunctions (e.g., and, but, because), interjections (e.g., oh, gosh, boy), verbs (e.g., say, look, see), and clauses (e.g., you see, I mean, you know)” (p. 234). These characteristics of DMs show that the common DM mentioned in the literature is a syntactically optional statement that is used to relate an utterance to the one that comes right before it and has no bearing on the truth-conditions attached to the utterance it introduces (Schourup, 1999).

3. Literature Review

A consistent finding across global research is that while both native and nonnative speakers employ discourse markers, nonnative speakers typically rely more on ideational functions, whereas native speakers make greater use of pragmatic and interactional functions (e.g., Fung & Carter, 2007; Müller, 2005; Liu, 2017). For example, Müller (2005) compared the use of so, well, you know, and like by American speakers of English and German speakers of English. The researcher found that these four DMs appear in both groups, but with differences in the individual functions of each DM. For example, the researcher found that NESs use textual functions of so more than L2 speakers of English, but that there is no clear difference found in the use of interactional and ideational functions of so among native and nonnative speakers of English. Müller (2005) also found that NESs use you know five times more than German speakers. Moreover, Arya (2020) explored the use of DMs in conversations between Thai EFL learners and non-Thai English speakers, highlighting notable differences. The findings revealed that Thai learners relied heavily on a limited set of DMs, such as “and,” “OK,” “but,” and “so”. Thai learners exhibited abrupt topic shifts, suggesting limited proficiency in employing structural DMs to ensure conversational coherence. Their conversations frequently included abrupt topic changes, indicating a struggle to use structural DMs effectively for smooth and coherent communication. On the other hand, native or proficient English speakers tended to use a wider range of DMs more strategically, helping them manage conversations more naturally and convey subtle meanings.

Furthermore, Fung and Carter (2007) compared the use of DMs between native British English speakers and Hong Kong speakers of English. They found that L2 speakers of English make more use of the DMs with referential functions (and, but, because, ok, so, etc.), but L1 speakers of English make more use of DMs with interpersonal functions (yeah, really, say, sort of, I see, you see, well, right, actually, cos, you know, etc.). The study found that L1 speakers of English use the pragmatic markers, whereas L2 speakers of English are restricted to the referential use of these markers. Similarly, Liu (2017) compared the use of but and so by American English speakers and Chinese speakers of English. The researcher found that L2 speakers of English underuse the pragmatic functions of but and so and overuse the ideational function of but. The researcher suggested that the use of the ideational function of but by L2 speakers of English is due to their L1. In other words, the ideational function of but in their language is translatable, but the pragmatic functions are not. Buysse (2012) studied the use of so by NESs and Dutch speakers of English. He found that the NNESs use so more frequently than NESs. Also, Dutch speakers of English were found to use the interpersonal and textual functions of so more than NESs. Similar findings were reported by Fuller (2003), Nikula (1993), and Moreno (2001). Evidently, the findings of these studies indicated that both NESs and NNES employ DMs. However, NESs use a wider range of markers more strategically, often leveraging their interpersonal and textual functions. In contrast, NNESs often rely on a limited set of referential DMs, such as “and,” “but,” and “so,” which may result in less conversational coherence and abrupt topic shifts. These findings underscore differences in proficiency and the functional use of DMs across linguistic and cultural contexts. In addition to these pragmatic and discourse-level differences, affective factors—particularly foreign language anxiety (FLA)—can also interfere with oral performance and the natural deployment of markers in speaking tasks (Attia & Algazo, 2025).

As for Arabic speakers of English, few studies (e.g., Algouzi, 2014, 2021; Kurdi, 2008; Majeed & Hmoody, 2019; Harb et al., 2022; Rabab’ah et al., 2022) have been conducted to examine the use of DMs by Arab speakers of English. For example, Kurdi (2008) explored the use of DMs by Syrian English learners. The study found that so was the most used marker in the learners’ speech. They use so to show coherence in their discourse. In a sense, so was used to mark a cause/result relationship between utterances, to mark repeating or summarising what participants had already talked about, and it is also a marker of transitions, which is used to move to another idea that is irrelevant to ideas previously mentioned. Moreover, she found that learners also use you know generally for understanding with the interviewer. I mean is also examined in her paper, and she found that learners use I mean to find any justifications for what has been said.

Another study was conducted in the Arabic context of Saudi Arabia by Algouzi (2014), who investigated the use of DMs by Saudi English learners and British speakers. The study revealed that advanced learners of English use DMs in their discourse with different frequencies. In addition, they use like, so, and you know the most in their speech. Algouzi (2014) pointed out that the occurrence of so and you know in Saudi speech is due to their equivalents in Arabic. Algouzi also investigated the functions of DMs used by NNESs, and the study found that Saudi speakers of English use the textual and interpersonal functions of DMs. Specifically, out of 17 different functions, Saudi learners use 13 as textual and interpersonal functions. Saudi learners of English use you know more than native speakers, and this result is due to two different reasons. First, NNESs acquire this DM easily and apply it as formulaic language, which refers to multi-word or whole chunks (Wood, 2015). Second is the effect of L1 on speakers with an Arabic equivalent “?araft”; Algouzi explained that although Saudi speakers of English use the functions of DMs as NESs, the frequency of using DMs is different in that native speakers use DMs more than NNESs.

Al-Yaari (2013) is another study conducted in the Saudi context, which investigated the use of DMs by Saudi EFL learners and compared them with other EFL learners and NESs. The study found that learners of English use “and”, “but”, and “also” the most in their talk; however, their usage of DMs is less than that of native speakers. Interestingly, the study found that Saudi EFL learners use fewer DMs than other nonnative English speakers. Specifically, Spanish, Hungarian, and Chinese EFL learners use DMs more frequently than Saudi learners. However, further research may be needed to investigate whether there is a significant difference between Arab NNESs and NNESs from other linguistic backgrounds. The findings of studies in the Arabic context generally demonstrate the frequent use of DMs such as so, y’know, and like in the speech of Arabic English learners, often influenced by their equivalents in Arabic. In addition, NNESs of Arabic origin use DMs for coherence, transitions, and interpersonal functions. However, their frequency and range of use differ from those of native speakers. Factors such as formulaic language and L1 influence provide a nuanced understanding of DM usage. These findings set the stage for a deeper investigation in the current study, which explores how Arabic-speaking learners employ DMs.

Together, these studies confirm that while the use of DMs among Arab learners has received attention in written, professional, and digital contexts, spoken, narrative-based data across different Arab nationalities remain underexplored. The present study, therefore, addresses this gap by focusing on spontaneous spoken discourse and by providing a cross-national perspective on Arab learners’ use of multiple DMs. This gap matters because narrative-based spoken data provide insights into learners’ real-time pragmatic choices, which are not easily observable in written or classroom-controlled contexts. Spoken narratives capture how learners manage turn-taking, repair, and coherence in ways that written data cannot. Research on Arab learners has shown that pragmatic errors—such as inappropriate use of discourse markers or mismanagement of turn-taking—can negatively affect communicative success and lead to perceptions of non-fluency or lack of competence (e.g., Götz, 2013; Ishida, 2006). Moreover, learner motivation studies suggest that Arab learners often value communicative fluency and interactional competence as highly as grammatical accuracy (Rabab’ah, 2015). Thus, by focusing on spontaneous spoken discourse, the present study addresses an underexplored but pedagogically crucial area: how Arab learners actually deploy discourse markers in narrative interaction, where pragmatic missteps are most visible and most consequential.

Beyond the studies already reviewed, a wider body of research has examined the role of discourse markers in L2 contexts, highlighting both cross-linguistic variation and pragmatic development. Romero-Trillo (2002), for example, investigated pragmatic markers in Spanish learners’ English and found that learners significantly underused markers with interpersonal functions, which affected the naturalness of their speech. Similarly, Hellermann and Vergun (2007) analysed the development of discourse markers in young ESL learners in the United States and observed that markers such as and, but, and so were acquired earlier than interpersonal markers like well and you know, suggesting developmental sequencing in pragmatic competence. Ishida (2006) explored Japanese learners’ use of well and reported that while learners recognised its textual function, they struggled with its pragmatic and interactional roles. Lam (2009), in a study on Hong Kong learners, found that DMs were often used for coherence but rarely for interpersonal alignment. Collectively, these studies reinforce a common trend: nonnative speakers tend to rely on the ideational and textual functions of DMs, whereas the interpersonal and pragmatic dimensions are underdeveloped. Within the Arab EFL context, prior research has been relatively limited, often concentrating on Saudi learners (Algouzi, 2014, 2021; Al-Yaari, 2013) or Syrian learners (Kurdi, 2008), with some comparative studies including Arabic and English (Majeed & Hmoody, 2019). While these studies highlight frequent use of markers such as so, you know, and like, often influenced by Arabic equivalents, they also indicate a narrower functional range compared to native speakers. This suggests that, despite the growing international scholarship on DMs in L2 speech, there remains a notable gap in research addressing Arab English speakers more broadly across different national contexts. The present study, therefore, extends this line of inquiry by examining how Arab learners from diverse backgrounds employ DMs in spontaneous English speech, thereby providing a more nuanced understanding of their functional deployment.

4. Methodology

4.1. Participants

The participants of this study were 10 Arab speakers of English who were chosen randomly from universities and schools. All participants filled out a questionnaire online with their basic personal information. The questionnaire contains personal information, including the participant’s name, gender, age, nationality, and place of residence. It also contains some educational questions, such as the participant’s primary degree, place of study, and qualifications in the English language. The last part of the questionnaire has job information: whether the participant has a job and whether they speak with NESs during work hours. The participants were from different nationalities: Jordan, Egypt, and Palestine. All of them are native Arabic speakers and use English in their workplaces. Their age ranges between 17 and 44 years old. Eight participants are English language teachers in schools where they speak English daily with their students and NESs. The participants of this study were 10 Arab speakers of English selected from both school and university contexts. Specifically, 7 were English language teachers working in schools, while 3 were university students in English-medium programs. All participants reported advanced proficiency in English, either through holding degrees in English language/translation or through daily professional use of English as teachers. While formal standardised test scores (e.g., IELTS/TOEFL) were not collected, participants’ backgrounds and professional use of English indicate that they were capable users of the language, able to engage in extended discourse in English.

Participants were recruited through personal and professional networks at universities and schools in Jordan, Egypt, and Palestine. Within each institution, volunteers were invited to participate, and the final sample was selected based on willingness and availability. While the sample size (10 participants) is modest, it is consistent with the exploratory and qualitative nature of the study, which emphasises depth of discourse analysis over statistical generalisation.

The decision to interview 10 participants was guided by the study’s qualitative and exploratory design, where the emphasis was on generating in-depth, naturalistic data rather than statistical generalisation. Each interview lasted around 20 min and, when transcribed, contributed to a corpus of over 25,000 words, which is substantial for a qualitative discourse analysis. The inclusion of participants from three different Arab countries (Jordan, Egypt, and Palestine) was intentional, as it allowed the researchers to capture regional commonalities rather than focusing on one national context. This cross-national composition provides broader insight into Arab learners’ use of discourse markers while remaining manageable for fine-grained qualitative coding.

4.2. Data Collection

The data were collected using one-to-one semi-structured interviews with the participants. This data collection tool has enabled the interviewer to obtain more profound answers and explanations to elicit as many DMs as possible from the participants. Semi-structured interviews enable the discussion to flow naturally and interchangeably because the questions are not fixed. In addition, participants feel free to answer and explain their responses in their own way, making the interview feel natural, like a conversation with a friend (Creswell & Creswell, 2023). This approach helps elicit a variety of DMs as participants speak. The semi-structured interview questions were carefully designed to focus on the participants’ life experiences and personal reflections. The goal was to encourage them to share detailed responses, providing opportunities to naturally use DMs as they discussed their personal lives and events in response to the interviewer’s prompts. All interviews were recorded, and the participants were informed that their interviews would be recorded for research purposes. At the beginning of the interview, the interviewer followed a previously conducted questionnaire to verify the background of the participants. Each participant was interviewed for about 20 min. After the interviews, the researchers transcribed the recordings to analyse the data. The transcription of each recording took almost 3 h. As a result, the researchers collected a corpus consisting of 25,849 words.

4.3. Data Analysis

The analysis of the data was conducted using qualitative methods, guided by Schiffrin’s (1987) and Schourup’s (1999) frameworks of DM functions. This model categorises DMs based on their roles across five planes: exchange structure, action structure, ideational structure, participation framework, and informational state. The researchers employed these planes to systematically examine how Arab nonnative English speakers (NNESs) utilised DMs in their speech. In addition, Culpeper and Haugh’s (2014) transcription conventions were applied to ensure consistency and precision in documenting spoken interactions.

The analysis process began with transcribing the recorded semi-structured interviews, resulting in a corpus of 25,849 words. Each interview was carefully reviewed, and all instances of DMs were identified and marked. The researchers employed hard copies of the transcripts for annotation, ensuring clarity and focus on the linguistic data. Each identified DM was classified according to its primary function and categorised under one or more of Schiffrin’s planes of discourse. To ensure reliability, the analysis was reviewed iteratively, with researchers cross-verifying findings to minimise bias. The patterns identified were consistent across participants, indicating commonalities in how Arab NNESs employ DMs to construct meaning and manage discourse.

5. Findings

This section presents the study’s results. It analyses the functions of DMs used by the participants and compares the use of DMs by NNESs with that of NESs, as found in Schiffrin (1987).

5.1. Functions of DMs as Used by Arab Speakers of English

This section presents the results related to the first research question, which asked about the functions of DMs used by Arab speakers of English.

5.1.1. Functions of “and”

By analysing the NNESs’ interviews, the researchers found that the ideational function of and is emphasised by the participants while other functions are not; however, continuing the speaker’s action is used more than organising and smoothing turns. As Example (1) below shows, Eman, in her talk about the funniest moment in her life, used and to connect the events she was retelling. The following excerpt illustrates how and is used to connect sequential events in narrative discourse, emphasising its ideational role.

(1)

Eman: a. So when we took off three mins later we were row by row we could see each other from a distance so just to realise that out of nowhere we couldn’t see them so we didn’t know where they were so just to realise that that their jet ski had flipped over and they were hanging on the edge of that jet ski and my other son who sixteen years old at this point we were both on the same jet ski so we had to go closer to see what’s going on so Rayan you had met Rayan she was on one end where as my youngest one Yousef who’s eleven years old he was trying to climb on top of it because the whole jet ski had flipped all over=

Interviewer: b. Yeah

Eman: c. =And my husband was freaking out he was trying to secure both of them and one of his sandals or shoes like flip flop had you know left the area and we were just tracking that flip flop rather than securing then so-

Interviewer: d. Hhh

In this case, and links successive actions and events, showing its function as a cohesive device that maintains narrative flow rather than managing turn-taking or interaction. Thus, as Eman was telling the interviewer about that moment when they were in Lakeland, where they had jet ski boats, she used and to connect events she mentioned. And here functions on the ideational plane, and it has the three main features of DMs according to Schourup (1999): connectivity as it shows the relationships between utterances, optionality as the deletion of and in her utterances does not affect the meaning and the hearer can still obtain it, and truth-conditionality as the deletion of and does not affect the truth conditions of the meaning that is implied in an utterance. In other words, it is syntactically optional. In Example (1), it is obvious that Eman prefaced each event with and to mark a new event. The following excerpt demonstrates how but functions ideationally by marking contrastive ideas in participants’ narratives.

Example (2) illustrates a higher level of idea structure when another interviewee, Hani, was asked about the reason for being proud of his school.

(2)

Hani: a. I feel proud of being a main part of my English team and the whole school.

Interviewer: b. Yeah so why do you feel proud to be part of your school? just tell me a little bit about your job so you feel proud to be part of it.

Hani: c. Ok (.) I am a teacher of English and I spend actually most of my career as a teacher of English in my school in Saudi Arabia all I learned was in the school.

d. and I see myself as an expert teacher I learned a lot also I had the experience that I got over the years from my colleagues.

Interviewer: e. yeah

Here, but signals a sharp contrast between the expected and actual event, fulfilling an ideational function by structuring the narrative around opposing propositions. In (a), Hani is answering the interviewer’s question about what he feels proud of, and his answer was the position that he is proud of being a member of his English team; then, he added another position, which is the whole school, so here and functions at a higher level to connect the two positions. In (b), the interviewer asked Hani about the reason behind his previous positions, and he, in (c), justifies his positions; therefore, and in (c) prefaced his support. Then, Hani in (d) prefaced another position by using and to connect his support points of (a) position with his position in (d).

In Example (3), Dania is asked about something she feels proud of, and she used and to conclude her answer. The next excerpt illustrates the pragmatic function of so, where it marks a claim made by the speaker.

(3)

Dania: a. Yeah ya3ny Alhamdulillah (thank God) I’ve done a lot of things I felt like I think I feel proud when I feel like I accomplish something as part of a team,

b. and in my last job I was teaching,

c. and when I first time I teaching them I struggled you know there is a challenge to teach (.)=

Interviewer: d. Yes I know

Dania: e. =But then when I started seeing their results at the end of the year I felt very proud I didn’t think of them like my students only they are more like my younger sisters, all of them if I was a little bit older I feel like they are like my children.

f. but I definitely felt proud of how I was with them and accomplishing that they did and the improvement that happened to them(.)

g. and yeah that’s one of the things I was proud of.

In this utterance, so introduces the speaker’s evaluative claim about his father, highlighting its pragmatic role in constructing stance rather than simply marking ideational result. In the previous example, Dania was explaining her answer. In (b) and (c), she connected two events within her position (a), and then she moved to talk about their results and how her students are close to her as a teacher. Then, in (g), she concluded her answer by using and.

According to the data analysis, NNESs use and to continue the speaker’s action more than to organise and smooth the turn-taking. Example (4) illustrates the pragmatic use of and by a NNES. The following example shows how y’know is used to organise shared knowledge and ensure mutual understanding between speaker and listener.

(4)

Yasmin: a. Yeah so I remember that he bought me a swing,

Interviewer: b. Yeah

Yasmin: c. Yeah and we put it in our garden, I really remember this it was really fun

Interviewer: d. It was very fun yeah?

Yasmin: e. Yeah.

Interviewer: f. And you were the one just playing in it?

Yasmin: g. Yeah because I am the smallest one and all my brothers and sisters were bigger than me.

Interviewer: h. Also, I think they were married?

Yasmin: i. Yes.

Interviewer: j. Yeah.

Yasmin: k. And they have kids.

Here, y’know signals an assumption of shared awareness about “strong accents,” functioning to align the speaker and listener and check comprehension. In the previous example, Yasmin is asked about her childhood. She mentioned one moment when her dad bought her a swing in (a). In (c), she was telling the events when she had that swing. Then, the interviewer asked her in (f) whether she was the only one who was playing on that swing, and she answered in (g) by mentioning the reason. In (h), the interviewer requested information on whether all her siblings were married or not, and in (i), she gave compliance with falling intonation. Then, in (k), she continues her action of answering by using and to initiate the continuous action, which was an answer.

In NNESs’ data, organising and smoothing turns appeared two times. The two examples below (6 and 7) show this use. The following excerpt shows and being used pragmatically to hold the floor during hesitation.

(5)

Hashim: a. And (0.5) ye3ny (0.5) and(.) like(.) Allah is telling me that you can like(.) nothing(.) he was telling me like this happens for this reason, if this wouldn’t happen you will not like achieve whatever whatever you are in now.

Interviewer: b. Yes.

Here, and combines with pauses and hesitation markers to help the speaker maintain control of the turn. In the previous example, Hashim was explaining his answer about changing his mind about something that had recently happened in his life. In (a), Hashim, by taking pauses, showed hesitation. And marks holding floor by Hashim in (a), and also “ya3ny” (I mean) appears in this utterance to mark his hesitation. Eman also used and in the interview to give the floor or turn to the hearer. See example (6). This excerpt illustrates how and is used to signal the end of a turn and yield the floor to the interlocutor.

(6)

Eman: a. We wouldn’t really be making a big difference but I feel like our role is really important and parents are always invited as well and they are part of the decision-making process and Hhh

Interviewer: b. Ok sorry you feel that parents are satisfied about what you do with their children?

In this instance, and precedes the out-breath and marks turn-completion, facilitating smooth speaker transition. In the previous example, Eman was explaining her job as a special education teacher. In (a), she initiated her position by using but, and then she connected her position with events by using and. The last and in (a) was followed by out-breath as she finished her turn and gave the floor to the hearer. According to data analysis, and is used by all participants as a marker to connect ideas at all levels. It was used 339 times by nonnative male and female interviewees in the data as an ideational function. However, and is used in the NNESs’ group on its action structure 29 times by both males and females. However, and appears in the data two times as a marker of organising turn-taking by one male and one female.

5.1.2. Functions of “but”

Analysing the data reveals that participants use but to mark contrastive ideas on all levels more than all other functions. It is also found that they use but in its action structure more than in the exchange structure. The ideational function of but is to mark the upcoming utterance as a contrastive idea of what is preceded. Example (7) below shows when Dania was telling about the funniest moment that happened in her life, and but marks a contrastive event. The following excerpt demonstrates how but functions to introduce a contrastive event in a narrative.

(7)

Dania: a. So I go inside, I express myself to the students on the desks, I write my name, my email, the name of the course, and I start. then the door opened, a professor came, but that was not the same professor that I am supposed to help=

Interviewer: b. Hhh

Dania: c. =Then he looked at me and he like what you’re doing here? =

Interviewer: d. Hhh

Dania: d. =And I told him that I am a TA for this class, he said no this is my class, it was international business. do you want to help me?

Here, but signals the unexpected arrival of a professor, highlighting the contrast between expectation and reality. Thus, as Dania was telling the funniest moment in her life, she was connecting the events by using and, but suddenly she mentioned in (a) a contrastive event that the professor she was supposed to help was not the same as the professor who asked her to go to his class and help before his coming by using but. Example (8) illustrates more about this function. This example shows how but introduces a contrastive situation in recounting personal experience.

(8)

Yasmin: a. Weak he can’t even talk he can’t he is feeling pain I think

Interviewer: b. So who were in hospital when you received that he-

Yasmin: c. Yeah actually it’s a bad moment.

Interviewer: d. You were in a hospital or at home?

Yasmin: e. Actually(.) I was in the hospital, but my family told us to go to home because-

The use of but here juxtaposes the setting (being in hospital) with the family’s instruction, marking a clear opposition. Yasmin was talking about the saddest moment in her life, and she was telling that her father’s death was the worst moment ever, and then, in (b), the interviewer asked her more about the incident (request for elaboration). In (c), she interrupted the interviewer, so, in (d), the interviewer asked the question again, and then, in (e), she started with an event in which she was in the hospital, and then she mentioned another event, which is contrastive with the previous one. She used but to mark the contrastive event in (e). But also marks a contrastive position (Schiffrin, 1987). See the example below. This excerpt illustrates how but is used to mark a contrastive stance in decision-making discourse.

(9)

Interviewer: so what do u think you are going to study now?

Zaid: actually everybody is telling me to go for medicine, but I think IT is much better. because there is a lot of job opportunities especially here in UAE for it.

In this case, but highlights the divergence between others’ advice and the speaker’s own position. As the interviewer asked Zaid if he had changed his mind about something, his compliance was that he had changed his mind about studying architecture; then, the interviewer requested clarification about his decision. His compliance was seen as an event that (everybody is telling me to go for medicine); then, he added his position toward that event by using but to mark his contrastive position to the preceding event. Examples below illustrate the functions of but at action structure. The following excerpt demonstrates how but functions pragmatically to return to a requested topic.

(10)

Interviewer: a. Ok question number nine is do you go on vacations? and what are these like?

Yasmin: b. <Obviously> (0.5)-

Interviewer: c. Yeah

Yasmin: d. No Hhh never you know as I previously mentioned I was -

Interviewer: e. Afraid of maybe?

Yasmin: f. >Yes yes< I have a problem with travelling. but you know that’s all to change this week so come back and ask me this question

Interviewer: g. I ask you this question next week? Hhh

Yasmin: h. Hhh I can then answer. but I will tell you, I had a vacation, but when I was very young?

Interviewer: I. Ok tell me about it.

Here, but marks a return to the interviewer’s earlier question, showing its role in structuring interaction. In Example (10), Yasmin is asked if she goes on vacations, and then in (d) she marked background information by using y’know. Then, in (f), she said that her feeling of fear will change next week. She sarcastically told the interviewer to come back and ask her again next week the same question about vacation. Then, in (h), she turned to the question that the interviewer asked to give the compliance. She marks her return to the question with but. By doing so, Yasmin used but to mark her action of returning to the requested information to give her compliance. Another example in which but functions on the action structure by marking a reclaim. This example shows how but is used pragmatically to reclaim a misunderstood point.

(11)

Interviewer: >Yeah yeah< I am an English teacher. Ok(.) and yes I agree with u it is very difficult job. but you really change your mind about this job? so what you are going to work then?

Esraa: No↑ I did not change my mind about this job. but I have changed my mind that being a teacher is not that easy job, it is not an easy thing.

In this case, but corrects the interviewer’s misinterpretation, asserting the speaker’s intended meaning. The example previously mentioned illustrates how but functions at an action structure by marking a reclaim. Esraa was asked if she had changed her mind about something recently, and her answer was that she had changed her mind about the job of being a teacher. In the interviewer’s turn, she misunderstood the point, and she thought that she had changed her mind about teaching itself, and then the interviewer asked her what she would do for work if she changed her mind about teaching. In her turn, she started her compliance with a sharp rising intonation to reclaim her point that the interviewer misunderstood. She marks her reclaim by using but, which is followed by her corrected point. By doing so, Esraa used but pragmatically in her discourse. This function of but appeared twice in the corpus. The second example of this function is in Ashraf’s interviews when he was telling the interviewer the funniest situation that happened to him. The following excerpt demonstrates how but can mark a reclaim that clarifies meaning after misunderstanding.

(12)

Ashraf: a. I think one of the things just happened recently? I was talking to someone I calling him by his name, I didn’t see him during all the summer break, I saw him after that, then I realise that this was not his name, it was totally a different name, so I was calling him-

Interviewer: b. It happens we just go on vacations maybe three or more month.

Ashraf: c. Yes it like a memory loss.

Interviewer: d. You forgot all the names?

Ashraf: e. some not all the names. but that particular person that I was talking to.

Here, but introduces a correction, ensuring the interlocutor understands the point about memory loss accurately. In Example (12), Ashraf told the interviewer that he forgot one of his friend’s names, and he was calling him using another name, and then he realised that it was not his name. In (c), he claimed that it was a memory loss, and then the interviewer asked him if he forgot all the names in (d). In (e), he made his point clearer by using but. He used but in (e) to mark his reclaim in order to clarify his point to the interviewer due to a misunderstanding that he had not really had memory loss, but he forgot the name of the person that he was talking to. In the corpus, functional contrast appeared in the discourse. This example shows how but is used to preface a cooperative response despite partial incompliance.

(13)

Interviewer: a. So eman you are afraid of insects(.) bugs?

Eman: b. Absolutely absolutely.

Interviewer: c. Yeah so many situations many funny situations? =

Eman: d. Yes

Interviewer: e. =Related to this kind of insects

Eman: f. Well I don’t have one in mind but something just popped out? I remember when we went on vacation with my family, we went to Lakeland which is about four hour drive from where we live, and we had rented a jet ski which is the speed boat, kind of like a small jet ski where you really enjoy it with the beautiful sun in the-

In this case, but signals a shift from rejecting the interviewer’s assumption to offering a relevant narrative. In the previous example, Eman was telling the funniest moment in her life. She said that it was when a cockroach flew at her. In (c), the interviewer requested more funny situations that happened with her about bugs and insects, but in her compliance, first, she marked her incompliance of the requested information by using well, and she directly started talking about a funny moment that happened with her, not related to insects. She, in (f), used but to preface a cooperative response in order to give compliance with the requested information. The following excerpt demonstrates how but highlights functional contrast within a positive evaluation.

Another example that illustrates the functional contrast of but is as follows:

(14)

Interviewer: a. So how was it? this vacation was it nice?

Yasmin: b. Yes but I think it’s the weather it’s very hot there

Interviewer: c. Was very hot?

Yasmin: d. That the worst thing there.

Here, but allows the speaker to soften her earlier positive evaluation of the vacation by adding a contrasting negative detail. Yasmin was talking about her vacation to Egypt, then in (a), the interviewer asked her if her vacation was nice, her compliance was (yes), but then she added that the weather was very hot, which is the worst thing there. By doing so, Yasmin prefaced the thing that she did not like by using but. It is noticed that Yasmin did not mention from the beginning that the vacation was not nice because of the weather. She said yes, and then she used but to tell the interviewer that it was not really nice. By using but as a functional contrast, she saved the interviewer’s face because the interviewer made an inference from her prior talk that the vacation was nice.

In the corpus, but appeared only once, functioning at the exchange structure. Example (15) will illustrate but at an exchange structure by marking a floor holding. This example shows how but is used pragmatically as a floor-holding device.

(15)

Mohammed: really I didn’t get the point. But (0.7) I think ya3ny (0.6) somebody who supported <me> I think(.) I couldn’t remember somebody supported me evenly.

Interviewer: Yes.

In this case, but enables the speaker to pause and search for information without losing the turn. In Example (15), Mohammed is asked about the most important person in his life, and his compliance was with his children because he does everything for them. Then, the interviewer told him her position, which is that the most important person should support you, not the other way around. In his turn, in example (16), he said that he did not get the interviewer’s point, and then he tried to think of somebody who supported him. By using but, he holds the floor in order to search for information; then, he said “ya3ny” with a stop also to search for information. In this case, Mohammed uses but to hold the floor.

After analysing but in the corpus, NNESs mostly use the ideational function of but at the ideational structure; it appeared 46 times in the corpus. On the other hand, but appeared only 17 times at the action structure, more precisely, but is used by NNESs marking cooperative responses, turning to discourse topic, and reclaims, whereas marking disclaimers, defences, and self-repairs are not used by them. In the exchange structure, it appeared once in the corpus as a marker of the floor holder.

5.1.3. Functions of “so”

In the corpus, so as a marker of results hardly appeared. The following excerpt shows how so functions ideationally to mark a result.

(16)

Dania: but as I grow old unfortunately I have got a lot of academic reading so I don’t have much time for leisure reading, so I couldn’t read novels and stories(.) because I have other kind of reading I have to do, and then after I graduated I started working, and again it was hard to find time to read for fun ya3ny.

Interviewer: Yes so-

In this case, so connects the cause (academic reading) with its consequence (no leisure reading), highlighting resultative meaning. In Example (16), Dania used so twice to indicate the result of having a lot of academic reading and having no time, also after getting her job. However, in the corpus, so was not used as a marker of inference. This may be because of the nature of interviews, where the interviewer asks questions and the interviewee answers. Therefore, the chance of making inferences by the interviewee is very low.

So also functions at the action structure level by marking a request, a compliance, or a claim. So as a marker of actions appeared in the corpus in many places (see examples below). This example demonstrates how so introduces a claim within the discourse.

(17)

Hani: a. yeah if you ask this question to anyone? of course he will say his father or his mother=

Interviewer: b. yes

Hani: c. = but since my mother passed away? my father is my man. my father is my model role, all what my father said to me in the(.) when I was young, it became true. when I became almost this age, when he was talking to me, when I was a child=

Interviewer: d. yeah

Hani: e. =so he has all (.)he has a clear vision, he is almost right in everything because of the experience he had over the years.

Here, so marks an evaluative claim about the father’s wisdom, signalling pragmatic stance-taking. Hani was asked about the most important person in his life, and he, in (a), claimed that his father is his role model. Hani marked his claim in (e) with so.

Example (18) also illustrates how so prefaces a compliance as a speaker’s action. The following excerpt illustrates how so marks compliance after a request for clarification.

(18)

Hashim: a. the happiest moment of my life?

Interviewer: b. Yeah what is your happiest moment in your life?

Hashim: c. Well (0.5) so the happiest moment in my life when I saw my first kid actually.

In this case, so introduces the speaker’s compliance, functioning as a discourse-entry marker. As shown in the previous example, the interviewer asked Hashim about the happiest moment in his life, and he, in (a), requested clarification, and, in (c), he started his compliance after clarification using well. Hashim stopped for 5 s after well, and then he started his compliance again by using so. By doing so, Hashim started his compliance using so to mark his compliance after pausing.

The last function of so at the action structure is also used by Hashim only in the corpus, which marks a request for clarification. The following excerpt demonstrates how so pragmatically marks a request for clarification.

(19)

Hashim: a. Alright so-

Interviewer: b. It’s ok take your time think about it.

Hashim: c. Hhh

Interviewer: d. Yeah

Hashim: e. So I think about something I am proud?

Interviewer: f. Yeah think about it its ok.

Here, so initiates a request for confirmation of the intended question, signalling interactional negotiation. In Example (19), the interviewer asked Hashim about something he feels proud of. Hashim, in (e), took the turn again to request clarification about the question that she had asked. He marked his request with so. Therefore, Hashim used so in the interview at the action structure level to ask for clarification and to give his compliance.

In the participation framework plane, it has been found that so was used as a topic transition where speakers move from the main unit of discourse to subordinate units of explanation. This example shows how so functions in the participation framework to mark a topic transition.

(20)

Dania: In general, (.) I really(.)the happiest time that I spent is in(.) I spent with my family specially after (.) when I pass from university? I was for a first time in years, we are all together in the same house, me with my brothers with my parents and so it was a very happy time for me because I haven’t been in the United states in may be seven years before? that this is the first time we all together at the same house, staying together, and having fun together, and then in weekends sometime I go out with my sisters and my brothers, and have like quality time with them.

In this case, so moves the discourse from the main point to elaborating subordinate details. In the previous example, Dania was talking about her happiest moment. As shown, Dania used so to move from the main unit to subordinate units of explanation by saying (“it was a very happy time for me because I haven’t been in the United States in maybe seven years before…”).

Example (21) illustrates how the speaker completes the adjacency pairs to mark participant transition. The following excerpt demonstrates how so completes an adjacency pair in dialogue.

(21)

Interviewer: a. Like historical places do u mean?

Ashraf: b. Yeah historical places, and landmarks, there are many many many from atomised empire.

Interviewer: c. Do u like the traditional food there?

Ashraf: d. Of course ↑ I actually(.) the traditional food was like(.) the food that we know is not something exotic or unique.

e. so yes for sure.

Here, so closes the question–answer sequence, providing explicit closure. In the previous example, the interviewer asked Ashraf about his vacation to Turkey. He, in (d), showed compliance (of course), and then he mentioned that the food is like our food. Ashraf then returned in (e) to answer the question marking with so his completion of adjacency pairs; therefore, he marks a participant transition.

Example (22) explains how so marks participant transition when the speaker uses it to direct a question to the hearer. The following excerpt illustrates how so is used to transition into addressing the hearer directly.

(22)

Yasmin: a. My niece did Hhh

Interviewer: b. And

c. ok and she answered?

Yasmin: d. Ok=

Yasmin: e. =And handed the phone to me.

Interviewer: f. oh my god° °

Yasmin: g. For sure. so guess what?

Interviewer: h. What?

Yasmin: i. I just ran around, and she was following me Hhh

Interviewer: j. Hhh

In this case, so functions to engage the interlocutor actively, inviting them into the narrative. The interviewer asked Yasmin to tell her about the funniest moment in her life. In (g), Yasmin directed a question to the hearer to guess what happened. In (e), Yasmin directed a question to the interviewer while she was telling her story, with so marking a participant transition. This way of using so to mark participant transition occurred once in the corpus. The last way by which the speakers mark participant transition is when so marks turn-initiation. The following excerpt shows how so marks the initiation of a new turn.

(23)

Dania: a. I think the most important people in my life are my parents. they have a really big influence growing up(.) especially again my mom, because obviously girls usually go more like for their moms actually most of kids.

Interviewer: b. Yeah yeah.

Dania: c. So she has usually(.) I wouldn’t say a strong connection? but she has a big influence on the children.

Here, so signals a shift in focus to a more specific aspect of the parents’ influence. In the previous example, the interviewer asked Dania about the most important person in her life. Dania, in (a), gave her compliance in which she said that her parents are the most important people. Dania, then, in (c), initiated another turn to talk about her mother marking with so. Thus, a participant transition occurred when she took the turn again.

In this study, so also functions at the exchange structure level by marking hold-the-floor and give-the-floor functions. The following excerpt demonstrates how so is used to yield the floor to the hearer.

(24)

Ashraf: a. Actually my best friend always(.) I am talking about my male best friend my colleague is Suleiman in the school, and he is my best friend because you know we do all the work all the time, we face the same challenges, we do have different way of think sometime? but we share the same goal and objectives, and we look at things from like the same big picture. So Hhh

Interviewer: b. Yeah so you share the same points of view you mean?

Ashraf: c. yes.

In this case, so marks the conclusion of a speaker’s turn and smoothly hands over to the interlocutor. In the previous example, the interviewer asked Ashraf to talk about his best friend. In his turn (a), he concluded his turn with falling intonation, and then he said so to give the hearer the floor.

By analysing so in the corpus, NNESs used so at all levels, especially at the participation level. This may be due to the nature of questions that required the interviewee to narrate stories in the answer, leading the interviewees to emphasise the use of so as a marker of topic transition in explanation and in narratives. It is also found in the analysis that they hardly used so as a marker of result and as a marker of organising the floor. Clearly, the participants in the data did not use so as a marker of inferences, and that is due to the nature of interviews in which they answer the questions that are asked, which makes the probability of making inferences very low. The participants in the corpus used so at the action structure plane to mark claim, compliance, and request as well.

5.1.4. Functions of “y’know”

Analysing the data reveals that y’know has several functions. y’know functions at the informational state. The following excerpt illustrates how y’know organises shared knowledge between interlocutors.

(25)

Zaid: a. The funny situation that happened that when I asked him about his nationality, he replayed so fast and I didn’t understand what he said, and when I didn’t understand what he said I told him that I am sorry I don’t know where is the country that you are talking about, and all the class started to laugh because who doesn’t know where is the united kingdom?

Interviewer: b. Yes

Zaid: c. But I didn’t get it(.) because you know he has a very strong accent?

Here, y’know ensures that the listener recognises the point about accents, aligning shared understanding. In Example (25), Zaid in (c) interrupted his narrative in order to make sure that the hearer shares the same knowledge with him about the British accent; then, he used you know to mark the newly shared information with her using rising intonation because he was not sure that the hearer shares this background knowledge with him. By doing so, Zaid tries to organise the shared information between him and the hearer by using because to mark a warrant, as mentioned before, and you know to mark also the newly shared information. Example (26) illustrates the uncertainty of the speaker of the shared information. This example shows how y’know signals uncertainty and seeks confirmation of shared knowledge.

(26)

Hashim: a. I started thinking like(.) like of what (0.5) like(.) I changing my mind about like (.) the way I am looking to whatever >whatever< is happening to me or like Hhh its destiny. you know what I am saying?

Interviewer: b. yeah?

Hashim: c. it is like not coming by chance, or it come for reason. everything is happened (.) it is happening for a reason.

In this case, y’know prompts the interlocutor’s acknowledgment, reinforcing mutual comprehension. In this example, y’know prefaces the uncertainty of the speaker with rising intonation in order to organise the shared knowledge between the participants. In this example, Hashim in (a) organised the shared knowledge with the hearer by asking her, with raising intonation (you know what I mean), which indicates his uncertainty of the shared knowledge with the hearer; then, the interviewer in (b) showed with raising her intonation a requested elaboration of what he said; and then, in (c), he elaborated what he did say in (a). Example (27) shows you know as a marker of background knowledge. The following excerpt demonstrates how y’know introduces background information assumed to be shared.

(27)

Hashim: I got married in January, and you know we get married in the middle of the year, and in(.) at the end of the year I had to like(.) I had to leave my school, and to find somewhere else.

Here, y’know frames information as given and mutually known, strengthening discourse coherence. The previous example shows that Hashim prefaced shared information, which is that he got married in the middle of the year, with you know. He was sure that the hearer was sharing this information with him because he had previously mentioned it in the interview.

Participants use you know in order to draw the hearer’s attention to a very important point in their stories. In Example (28), Hani was narrating his happiest moment when he had his first child; then, he mentioned why he considered this moment as the happiest moment in his life, using you know in order to let the hearer understand why this story is said. This example shows how y’know highlights a significant point in the speaker’s story.

(28)

Hani: a. but when I received my first child, and I saw her? that was the happiest moment(.) that was like the hardest. so that was the moment.

Interviewer: b. Yeah.

Hani: c. and you know that was the success that I receive at the end. so that’s why I consider it as the most (.) or the moment that is happiest in my life.

In this case, y’know directs the listener’s attention to the reason the moment was considered highly meaningful. After Hani drew a conclusion in (a) using so, he initiated another turn in (c), continuing his action of answering using and. In (c), he used you know to mark an important point to tell the interviewer why he felt that this moment was the happiest. Thus, he drew the hearer’s attention by using you know to tell her why he chose the moment of receiving his first baby as the happiest.

In the corpus, only Esraa used you know to show a character’s dissatisfaction about something in the story. The following excerpt demonstrates how y’know signals others’ dissatisfaction in a narrative.

(29)

Esraa: Yes, recently I have learned this. after obtaining my MA degree, most of my relatives and friends advised me to continue my studying, but I answered them I will search for a job as an English teacher in one of the schools, you know most of them were shocked, and they told me why it’s better to continue? and you don’t have an ambition? so after that I have learned that satisfying myself is more important that satisfying other others’ desire.

Here, y’know emphasises the reaction of relatives, preparing the ground for the lesson learned. This is an extract of Esraa’s answer when she was asked about a lesson she learned in her life. While narrating her story, she used you know to draw the hearer’s attention to the characters’ dissatisfaction about her decision, which is to have a job instead of continuing her post-graduate studies. By doing so, she drew the hearer’s attention to the characters’ dissatisfaction in order to understand what she is going to say, which is the lesson that she learned recently in her life. Without mentioning her relatives’ action, the hearer will never understand the lesson that she learned; thus, she drew the hearer’s attention to this part by using you know in order to get her point.

You know also functions at the exchange structure pragmatically (Schiffrin, 1987). It initiates a turn, holds the floor to show hesitation, and gives the floor. Example (30) shows this. The following excerpt illustrates how y’know pragmatically functions as a floor-holding device.

(30)

Hashim: Like(.) you know (.) like he was telling me(.) like if you move a lot? you will like(.) you will (.) you will feel like you know (0.5) to be worried. you will feel distractive. that’s your movements have to be like(.) has to be like not to have exaggerate your movements Hhh.

In this case, y’know fills hesitation gaps, allowing the speaker to maintain turn control. Hashim, in this example, was telling the interviewer what another teacher told him to do in the class as a new teacher. In his turn, it is very obvious that he made many pauses, but he did fill these pauses using like as a DM of filling gaps (Müller, 2005) and you know as a DM of holding the floor. This example shows how y’know functions at the exchange structure to yield the floor.

(31)

Interviewer: a. so you arrange it with your friends all the class?

Roa’a: b. YES we arranged this prank, and it was really funny at the beginning of the class, but then she you know she called the principal for a punishment. you know.

Interviewer: c. Oh my god Hhh.

Here, y’know signals closure and passes the turn smoothly to the interlocutor. In the previous example, Roa’a already told the interviewer about a funny moment in her class; then, the interviewer asked her if she agreed with the other girls to prank their teacher (a). She gave compliance in (b) and marked shared knowledge by using the first you know and gave the interviewer the floor by using you know as well. She used you know at the end of her turn with falling intonation and with a soft voice to mark the end of her turn.

You know also functions in the participation framework, marking a speaker seeking the hearer’s endorsement of a generalisation. The following excerpt demonstrates how y’know is used to seek the hearer’s endorsement of a generalisation.

(32)

Interviewer: So after your done with this grade? you feel maybe that you are more confident to go more through teaching(.) right?

Dania: YEAH definitely more confident, and I definitely change the way I think about them, you know teenagers with attitudes. I start to accept them like them.

Interviewer: yes.

In this case, y’know frames a generalisation about teenagers, inviting the interlocutor’s agreement. Dania was talking about her experience as a teacher with teenagers. The interviewer in (a) asked her if this experience gave her confidence. She, in (b), agreed that she is now more confident; then, she marked a generalisation about teenagers and how they have an attitude with you know in order to seek the hearer’s endorsement of the generalisation she made; thus, the interviewer in (c) agreed with her generalisation.

By analysing the corpus, NNESs used you know with the primary function of organising the shared information. Less used of you know in the participation framework in which you know marks endorsement, and as a device to convert opponents to the speaker’s side. It is also found that NNESs clearly use you know to hold the floor.

6. Discussion

The findings of this study shed light on how NNESs use DMs in conversation. When answering the first research question, the data revealed that NNESs primarily use DMs at the ideational structure level, meaning they rely on them for organising ideas rather than for pragmatic functions. In other words, they frequently use words like and, but, and so to connect thoughts, but they rarely use DMs in ways that shape interaction, such as managing conversation flow, structuring exchanges, or indicating shared understanding. Notably, y’know—a marker commonly used in natural discourse—was not as frequent among NNESs because of its broad pragmatic role. Interestingly, however, y’know did appear across different discourse planes, particularly in the informational state, exchange structure, and participation framework. This suggests that, while NNESs may not use many pragmatic markers, they do recognise some of their functions in discourse. Meanwhile, and and so stood out as the most frequently used DMs, aligning with findings from Algouzi (2014), who reported that NNESs favor so and y’know, likely because they have Arabic equivalents. In our data, so was used extensively, which may be influenced by its Arabic counterpart fa. Similarly, Al-Yaari (2013) found that NNESs tend to rely on and and but in their speech, which is consistent with the patterns observed in this study.

This study’s findings highlight subtle patterns in how Arab English speakers use DMs, offering insights into the relationship between SLA processes and the unique traits of nonnative speech. The prevalence of the ideational function of DMs like and, but, and so in nonnative speech supports Schiffrin’s (1987) framework, which highlights their key role in establishing semantic links within discourse. Arab speakers tend to use these markers primarily to maintain coherence and indicate logical connections between ideas. This observation aligns with global research on NNESs such as Fung and Carter (2007) and Liu (2017). Their emphasis on ideational functions underscores a fundamental focus in SLA, where learners initially prioritise mastering explicit, rule-based language components over the nuanced aspects of pragmatic usage.

The limited use of DMs for pragmatic purposes, as identified in this study, highlights the difficulties NNESs encounter in developing pragmatic competence. Pragmatic functions of DMs—such as managing turn-taking, indicating shared knowledge, and supporting interaction—depend on implicit learning gained through substantial exposure to authentic language use (Götz, 2013). For example, the sparse use of so as a marker of inferences and y’know to seek agreement points to a gap in learners’ ability to manage interpersonal dynamics and grasp conversational nuances in English.

NNESs tend to use DMs less pragmatically, favoring and, but, and so as their go-to markers. However, an interesting exception emerged: NNESs used so pragmatically more often than NESs in the interviews. This pattern could be attributed to so being a common DM in Arabic, making its function more intuitive for Arabic speakers. A similar trend was seen with y’know, which was used by NNESs in a similar way to that of NESs, as found by Schiffrin (1987). This shows that y’know serves an ideational function due to its semantic meaning, making it easier for NNESs to adopt (Schiffrin, 1987). In addition, while so is sometimes used by NESs to mark inferences, this was only observed once in the NES corpus.

The multifunctionality of DMs, particularly y’know, highlights their significance in bridging semantic and pragmatic competence. As evidenced in this study, y’know functions as a versatile marker, aiding in organising information, signalling shared knowledge, and engaging listeners. This adaptability makes y’know an invaluable resource for learners, helping them improve discourse coherence and foster interactional alignment. However, its limited use in specific pragmatic contexts—such as seeking endorsement—indicates that NNESs may focus more on its explicit functions while overlooking its nuanced, interpersonal roles.

This finding aligns with Schourup’s (1999) description of DMs as optional yet influential components of discourse. Although DMs are not syntactically essential, their pragmatic use greatly contributes to conversational fluency and listener engagement. For NNESs, developing the ability to use DMs multifunctionally is a vital step toward attaining native-like discourse competence.

Finally, this study highlights NNESs’ preference for ideational over pragmatic DM use, even with immersion in English-speaking environments. While exposure enhances pragmatic competence, instructional focus on grammatical accuracy may limit DM multifunctionality. Encouraging NNESs to engage with pragmatic markers like y’know can help bridge the gap between structural proficiency and communicative fluency. Future research could explore how explicit instruction shapes DM usage in conversation.

Compared to similar studies, the present research makes several contributions. Whereas much previous work on Arab learners has focused either on single discourse markers (e.g., Algouzi, 2021, on so) or on limited classroom interactions (e.g., Kurdi, 2008; Al-Yaari, 2013), this study systematically analyses four markers (and, but, so, you know) across a corpus of spontaneous narrative interviews. This methodological choice provides richer evidence of how Arab learners deploy DMs in more naturalistic speech, extending beyond task-based or pedagogic contexts. In addition, while earlier studies noted Arab learners’ reliance on ideational functions, our findings reveal a more nuanced pattern: although and and but were indeed predominantly ideational, so and you know demonstrated broader pragmatic functions, such as managing turn-taking and signalling shared knowledge. This adds a new dimension to existing research, suggesting that Arab learners may develop pragmatic uses selectively, depending on the DM. Finally, by including participants from Jordan, Egypt, and Palestine, our study contributes a wider Arab perspective, which has rarely been attempted in prior research confined to a single national context.

Beyond documenting the use of discourse markers among Arab English speakers, the findings also carry broader implications for second language acquisition (SLA) research. First, they confirm a recurring theme in SLA that pragmatic and discourse competence lag behind structural competence, as learners in this study relied more on ideational than on pragmatic functions of DMs (Celce-Murcia, 2007; Gass et al., 2020). These results show how the Arab English speakers in this study relied more on ideational than pragmatic functions of DMs. While these findings cannot be generalised to all NNESs, they align with patterns reported in studies of other L2 learner groups (e.g., Romero-Trillo, 2002; Hellermann & Vergun, 2007; Lam, 2009). This suggests that the tendencies observed among Arab English speakers may reflect broader trends in L2 discourse marker use, though further research with larger and more diverse samples is needed. Second, the findings suggest that cross-linguistic influence plays a facilitative role in discourse–pragmatic development: Arab speakers’ relatively diversified use of so and you know appears linked to their Arabic equivalents (fa, yaʕni), pointing to L1 resources as scaffolds for L2 interactional competence. Finally, the study speaks to current work in instructed SLA (ISLA), where emphasis has traditionally been on grammar and lexis, by showing that discourse-level features such as DMs also warrant explicit attention. In this way, the study not only documents the patterns of DM use among Arab learners but also contributes to ongoing SLA discussions about the development of pragmatic competence, the role of transfer, and the pedagogical value of discourse-level instruction.

It should be noted that the findings of this study cannot be generalised to all Arab English speakers. The participants represented a small and heterogeneous group, drawn from different national backgrounds, and their discourse marker use may reflect both shared tendencies and local variation. Thus, the results should be interpreted as exploratory patterns rather than definitive cross-Arab generalisations. A limitation of this study concerns the diversity and recruitment of participants. The sample included speakers from different Arab countries (Jordan, Egypt, and Palestine), which provided a cross-national perspective but also introduced heterogeneity that may influence DM usage. Furthermore, participants were recruited on a voluntary basis from universities and schools, meaning that the selection was somewhat random rather than systematically stratified. These factors limit the representativeness of the findings and suggest that the results should be interpreted with caution as exploratory patterns rather than generalisable claims. Future research with larger, more systematically stratified samples would be valuable in identifying regional differences in discourse marker use among Arab English speakers.

With respect to RQ2, the present findings suggest that Arab speakers of English exhibit broadly similar patterns to those reported for other NNES groups, particularly in their reliance on ideational functions of and and but, and their more diversified pragmatic use of so and y’know. However, since this study did not include NES participants, these observations are based on comparisons with findings reported in prior NES-focused research. Thus, the conclusions regarding similarities and differences should be understood as interpretive rather than as the outcome of direct empirical comparison.

7. Conclusions

This study highlights the differences in how NNESs and NESs use DMs, emphasising NNESs’ limited pragmatic engagement and reliance on ideational functions. The findings of this study hold valuable implications for language teaching. To address the observed pragmatic gaps in DM usage, EFL curricula should include explicit instruction on the multifunctionality of DMs. Activities such as role-plays, peer discussions, and exposure to authentic spoken English can create meaningful contexts for learners to practice and internalise the pragmatic functions of DMs.

Furthermore, educators should emphasise the interpersonal and interactional aspects of DMs, extending beyond their semantic roles. For example, teaching the use of so for inferences and y’know for seeking endorsement can help learners gain a more nuanced understanding of these markers. Such pedagogical strategies can bridge the gap between ideational and pragmatic competence, fostering greater fluency and naturalness in learners’ speech.

Given the small and specific sample, the findings should be interpreted as illustrative of Arab English speakers’ discourse marker use rather than as generalisable claims about all NNESs. Nevertheless, the consistency with previous L2 studies suggests that the patterns observed here may resonate with broader tendencies in learner discourse.

The findings suggest a need for further research on the impact of proficiency on DMs usage and how advanced learners bridge the gap between ideational and pragmatic functions. Future studies could also explore DMs with pragmatic roles, such as oh, well, and I mean, and include more diverse participant groups to generalise findings. Moreover, future research could build on these findings by investigating discourse marker DM usage among NNESs from a wider range of linguistic and cultural backgrounds.