2.1Development of pragmatic awareness during study abroad

Pragmatic awareness is explicit knowledge of how pragmatic behavior is performed in social and communicative contexts (Takahashi 2012Takahashi, Satomi 2012 “Pragmatic Awareness in Second Language Learning.” In The Encyclopedia of Applied Linguistics, ed. by Carol A. Chapelle, 1–6. New Jersey: Blackwell. ). This definition first implies that pragmatic awareness includes both pragmalinguistic (form-function associations of pragmatic targets) and sociopragmatic knowledge (relationship between the pragmatic targets and the contextual and interactional rules that frame them) (Takahashi 2012Takahashi, Satomi 2012 “Pragmatic Awareness in Second Language Learning.” In The Encyclopedia of Applied Linguistics, ed. by Carol A. Chapelle, 1–6. New Jersey: Blackwell. ). Secondly, pragmatic awareness involves L2 users’ perceptions of how pragmatic behavior is performed. Such perceptions are determined by the individual’s first language (L1)-based intuition as well as by their cultural assumptions (McConachy 2019McConachy, Troy 2019 “L2 Pragmatics as ‘Intercultural Pragmatics’: Probing Sociopragmatic Aspects of Pragmatic Awareness.” Journal of Pragmatics 151: 167–176. ). Thirdly, this definition views pragmatic awareness as an umbrella term that involves different receptive skills, such as metapragmatic awareness, recognition or identification of pragmatic targets, judgement of (in)appropriateness, and comprehension of implied meaning.

Researchers have observed that pragmatic awareness contributes to the development of pragmatic production (McConachy and Spencer-Oatey 2020McConachy, Troy, and Helen Spencer-Oatey 2020 “Developing Pragmatic Awareness.” In Developmental and Clinical Pragmatics, ed. by Klaus P. Schneider, and Elly Ifantidou, 393–428. Boston: Mouton de Gruyter. ). SA research (e.g., Sánchez-Hernández and Alcón-Soler 2019Sánchez-Hernández, Ariadna, and Eva Alcón-Soler 2019 “Pragmatic Gains in the Study Abroad Context: Learners’ Experiences and Recognition of Pragmatic Routines.” Journal of Pragmatics 146: 54–71. ) has also shown that awareness of pragmatic behavior in the L2 is a first step towards the development of productive pragmatic skills in interaction. However, ILP studies have focused on the acquisition of productive skills during SA, and the development of pragmatic awareness has not received much attention. This research gap is crucial considering that during the first months abroad, participants make gains in pragmatic receptive skills, and start assigning plausible meanings to pragmatic targets recognized in the input (Alcón-Soler and Sánchez-Hernández 2017Alcón-Soler, Eva, and Ariadna Sánchez-Hernández 2017 “Learning Pragmatic Routines During Study Abroad: A Focus on Proficiency and Type of Routine.” Atlantis 39 (2): 191–210.; Almalki and Jones 2022Almalki, Ziad, and Christian Jones 2022 “ ‘Why Did You Do That?’ The Effects of Instruction on Recognition and Production of Informal Second Party Complaints.” The Language Learning Journal 50 (4): 443–459. ).

Addressing this concern, a small number of cross-sectional and longitudinal studies have explored the development of pragmatic awareness during SA. Early studies investigated pragmatic awareness by comparing SA students with their at-home counterparts, finding an advantage in favor of SA (Bardovi-Harlig and Dörnyei 1998Bardovi-Harlig, Kathleen, and Zoltán Dörnyei 1998 “Do Language Learners Recognize Pragmatic Violations? Pragmatic vs. Grammatical Awareness in Instructed L2 Learning.” TESOL Quarterly 32: 233–259. ; Schauer 2006Schauer, Gila A. 2006 “Pragmatic Awareness in ESL and EFL Contexts: Contrast and Development.” Language Learning 56: 269–318. ). An example is Bardovi-Harlig and Dörnyei (1998)Bardovi-Harlig, Kathleen, and Zoltán Dörnyei 1998 “Do Language Learners Recognize Pragmatic Violations? Pragmatic vs. Grammatical Awareness in Instructed L2 Learning.” TESOL Quarterly 32: 233–259. , where pragmatic awareness was operationalized as the ability to judge the (in)appropriateness and the severity of pragmatic infelicities in speech acts. They compared three groups: US speakers, Hungarian users of English in the US, and L2 users of English in Italy. Findings showed a higher sensitivity towards pragmatic infelicities for the Hungarian participants in the US.

Other cross-sectional studies compared the pragmatic awareness of SA students with that of L1 users (Cook and Liddicoat 2002Cook, Misty, and Anthony J. Liddicoat 2002 “The Development of Comprehension in Interlanguage Pragmatics: The Case of Request Strategies in English.” Australian Review of Applied Linguistics 25 (1): 19–39. ; Garcia 2004Garcia, Paula 2004 “Developmental Differences in Speech Act Recognition: A Pragmatic Awareness Study.” Language Awareness 13 (2): 96–115. ). For example, Garcia (2004)Garcia, Paula 2004 “Developmental Differences in Speech Act Recognition: A Pragmatic Awareness Study.” Language Awareness 13 (2): 96–115. explored the identification of indirect speech acts in academic conversations by three groups of university students in the US: twenty-one L1 users of English, sixteen high-proficiency L2 users of English, and nineteen low-proficiency L2 users of English. The findings showed that contextual cues and the type of speech act played a major role in pragmatic awareness. As expected, high-proficiency L2 English users identified speech act types with more accuracy than low-proficiency ones. However, contrary to expectations, the L1 users of English did not outperform the high-proficiency L2 users, and this was ascribed to the academic background in Applied Linguistics of the high-proficiency L2 users, which may have given these participants more knowledge of pragmatic concepts.

More recently, a few studies have explored pragmatic awareness in SA contexts from a longitudinal perspective (Ren 2015Ren, Wei 2015 L2 Pragmatic Development in Study Abroad Contexts. Bern: Peter Lang. ; Sánchez-Hernández and Alcón-Soler 2019Sánchez-Hernández, Ariadna, and Eva Alcón-Soler 2019 “Pragmatic Gains in the Study Abroad Context: Learners’ Experiences and Recognition of Pragmatic Routines.” Journal of Pragmatics 146: 54–71. ). Findings suggest that SA aids the development of pragmatic awareness, but longer sojourns do not necessarily entail more gains. Rather than the length of residence (LoR), it is the intensity of L2 interaction that plays a role in the development of pragmatic awareness. For instance, Sánchez-Hernández and Alcón-Soler (2019)Sánchez-Hernández, Ariadna, and Eva Alcón-Soler 2019 “Pragmatic Gains in the Study Abroad Context: Learners’ Experiences and Recognition of Pragmatic Routines.” Journal of Pragmatics 146: 54–71. investigated the recognition of pragmatic routines by thirty-one Brazilian students over a semester in the US. They conducted a mixed-methods study based on a quantitative analysis of a routine recognition test and a qualitative exploration of semi-structured interviews, finding that the students’ motivation to integrate in the new environment and the amount and nature of their interactions outside of class were determinant aspects for routine recognition. Indeed, different scholars have pointed out that engagement in different socialization practices abroad provides learners with contextually appropriate input that promotes awareness of target language (TL) use (Bardovi-Harlig et al. 2022Bardovi-Harlig, Kathleen, Jesús Izquierdo, and Yunwen Su 2022 “Recognition of Conventional Expressions by EFL Learners in Mexico and China.” System 110: 102918. ).

Recent SA studies have also found that instruction and pre-departure preparation further enhances the potential of SA (see Pérez-Vidal and Shively 2019Pérez-Vidal, Carmen, and Rachel Shively 2019 “L2 Pragmatic Development in Study Abroad Settings.” In The Routledge Handbook of Second Language Acquisition and Pragmatics, ed. by Naoko Taguchi, 355–371. New York: Routledge.; Sánchez-Hernández 2022Sánchez-Hernández, Ariadna 2022 “Second Language Pragmatic Development in Study Abroad Contexts: An Introduction.” Study Abroad Research in Second Language Acquisition and International Education 7 (1): 2–23. ) by expanding the opportunities for L2 exposure and noticing. In particular, pragmatic instruction appears to especially promote metapragmatic awareness, interpreted as “the knowledge of social meanings” (van Compernolle and Kinginger 2013van Compernolle, Rémi A., and Celeste Kinginger 2013 “Promoting Metapragmatic Development Through Assessment in the Zone of Proximal Development.” Language Teaching Research 17 (3): 282–302. , 284) that can be attributed to L2 forms. For instance, Henery (2015)Henery, Ashlie 2015 “On the Development of Metapragmatic Awareness Abroad: Two Case Studies Exploring the Role of Expert-Mediation.” Language Awareness 24: 316–331. conducted a case study on the development of metapragmatic awareness of two US learners in France, one receiving concept-based pragmatic instruction, the other not receiving any. Results from language awareness interviews showed an advantage for the instructed learner.

To further explore the development of L2 pragmatic awareness, Takahashi (2012)Takahashi, Satomi 2012 “Pragmatic Awareness in Second Language Learning.” In The Encyclopedia of Applied Linguistics, ed. by Carol A. Chapelle, 1–6. New Jersey: Blackwell. called for more fine-grained studies to analyze the role of individual differences. Since then, a few studies have revealed that the acquisition of pragmatic awareness is influenced by learners’ attitudes towards the L2 community (Yang and Ren 2019Yang, He, and Wei Ren 2019 “Pragmatic Awareness and Second Language Learning Motivation.” Pragmatics and Cognition 26: 447–473. ), their proficiency (Garcia 2004Garcia, Paula 2004 “Developmental Differences in Speech Act Recognition: A Pragmatic Awareness Study.” Language Awareness 13 (2): 96–115. ), and their willingness to communicate (Lv et al. 2021Lv, Xiaoxuan, Wei Ren, and Lin Li 2021 “Pragmatic Competence and Willingness to Communicate among L2 Learners of Chinese.” Frontiers in Psychology 12: 797419. ). However, despite the emergence of these studies, L2 pragmatic awareness has still received scant academic attention.

In summary, SA studies on the development of pragmatic awareness have shown that the first months abroad are key for the development of pragmatic receptive skills. While abroad, L2 users are likely to develop their ability to understand pragmatic infelicities, identify speech acts and pragmatic routines, and understand inferred meaning. Nevertheless, to the best of our knowledge, SA students’ awareness of PM use in the L2 has yet to be explored, despite the importance of PMs for successful oral interaction, and this perspective of investigation is unique to this study.

2.2The use of pragmatic markers by L2 users

PMs such as well and I mean have been defined as linguistic constructions that smooth interaction (D’Arcy 2017D’Arcy, Alexandra 2017 Discourse-Pragmatic Variation in Context: Eight Hundred Years of LIKE. Amsterdam: John Benjamins. ). They are phonologically reduced in conversation (Fedriani and Sansò 2017Fedriani, Chiara, and Andrea Sansò 2017 “Pragmatic Markers, Discourse Markers and Modal Particles: What We Know and Where Do We Go from Here?” In Pragmatic Markers, Discourse Markers and Modal Particles: New Perspectives, ed. by Chiara Fedriani, and Andrea Sansò, 1–33. Amsterdam: John Benjamins. ) and also syntactically optional; therefore, they can be removed from a sentence without affecting its intelligibility. A further feature of PMs is their multifunctionality; that is, they serve different functions depending on the context. PMs can perform textual functions when they contribute to discourse organization or can be used for interpersonal functions when they are used to signal the relationship between the speaker and the message or between the speaker and the hearer (Magliacane and Howard 2019Magliacane, Annarita, and Martin Howard 2019 “The Role of Learner Status in the Acquisition of Pragmatic Markers During Study Abroad: The Use of ‘Like’ in L2 English.” Journal of Pragmatics 146: 72–86. ; Sánchez-Hernández and Martínez-Flor 2022Sánchez-Hernández, Ariadna, and Alicia Martínez-Flor 2022 “Teaching the Pragmatics of English as an International Language (EIL): A Focus on Pragmatic Markers.” Language Teaching Research 26 (2): 256–278. ). Nevertheless, there is less consensus on the nomenclature used to refer to these linguistic items (Beeching 2016Beeching, Kate 2016 Pragmatic Markers in British English: Meaning in Social Interaction. Cambridge: Cambridge University Press. ; Fedriani and Sansò 2017Fedriani, Chiara, and Andrea Sansò 2017 “Pragmatic Markers, Discourse Markers and Modal Particles: What We Know and Where Do We Go from Here?” In Pragmatic Markers, Discourse Markers and Modal Particles: New Perspectives, ed. by Chiara Fedriani, and Andrea Sansò, 1–33. Amsterdam: John Benjamins. ), as different terms have been used in the literature (especially ‘discourse markers’ and ‘pragmatic markers’). In this article, following Aijmer and Simon-Vanderberger (2006)Aijmer, Karin, and Anne-Marie Simon-Vandenbergen 2006 “Introduction.” In Pragmatic Markers in Contrast, ed. by Karin Aijmer, and Anne-Marie Simon-Vandenbergen, 1–10. Amsterdam: Elsevier. , we use the term ‘pragmatic marker’ as a hypernym of ‘discourse marker’ and as a linguistic item that can be used both to express textual and interpersonal relations.

Research on PM use by L2 users is still emerging (Beeching 2015Beeching, Kate 2015 “Variability in Native and Non Native Use of Pragmatic Markers: The Example of Well in Role-Play Data.” In Researching Sociopragmatic Variation: Perspectives from Variational, Interlanguage and Contrastive Pragmatics, ed. by Kate Beeching, and Helen Woodfield, 147–197. London: Palgrave Macmillian. ; Liao 2009Liao, Silvie 2009 “Variation in the Use of Discourse Markers by Chinese Teaching Assistants in the US.” Journal of Pragmatics 41 (7): 1313–1328. ), even though a focus on their L2 use can provide interesting insights into sociopragmatic development (Diskin-Holdaway 2021Diskin-Holdaway, Chloé 2021 “ You Know and Like among Migrants in Ireland and Australia.” World Englishes 40 (4): 562–577. , 563). Indeed, despite their syntactic optionality, inappropriate use of PMs by L2 users can lead to communication breakdowns, as is the case with the misuse of other pragmatic features (Hansen 1998Hansen, Mai-Britt M. 1998 The Function of Discourse Particles: A Study with Special Reference to Spoken Standard French. Amsterdam: John Benjamins. , 199), because their inappropriate use can be less immediately noticed by interlocutors than lexical or grammatical inaccuracies. For instance, L2 users may come across as too direct or even rude in everyday conversation and interlocutors may not associate such behavior to the use of PMs. Moreover, the use of PMs in the L2 can be revealing to SA research as they can be an index of L2 exposure and of engagement in intercultural interactions. In fact, their use in the L2 is mainly ascribed to naturalistic exposure rather than to input received in instructed contexts, where they seem to be absent or not focused upon (Liao 2009Liao, Silvie 2009 “Variation in the Use of Discourse Markers by Chinese Teaching Assistants in the US.” Journal of Pragmatics 41 (7): 1313–1328. ).

Most SA studies on L2 PM use have been cross-sectional and have compared how L2 users differ from L1 users in PM production (Diskin 2017Diskin, Chloé 2017 “The Use of the Discourse-Pragmatic Marker ‘Like’ by Native and Non-Native Speakers of English in Ireland.” Journal of Pragmatics 120: 144–157. ; Diskin-Holdaway 2021Diskin-Holdaway, Chloé 2021 “ You Know and Like among Migrants in Ireland and Australia.” World Englishes 40 (4): 562–577. ; Liao 2009Liao, Silvie 2009 “Variation in the Use of Discourse Markers by Chinese Teaching Assistants in the US.” Journal of Pragmatics 41 (7): 1313–1328. ). Diskin (2017)Diskin, Chloé 2017 “The Use of the Discourse-Pragmatic Marker ‘Like’ by Native and Non-Native Speakers of English in Ireland.” Journal of Pragmatics 120: 144–157. focused on the use of like by Chinese and Polish L2 users of English in Ireland and by Irish English speakers, and explored the role that linguistic aspects (position, function), contextual factors (LoR) and individual differences (proficiency, gender) played on such use. The findings pointed to an effect of gender and of LoR: female participants used like as a filler more frequently, and after three years in Ireland, participants tended to use like in a similar way to Irish English speakers. Diskin-Holdaway (2021)Diskin-Holdaway, Chloé 2021 “ You Know and Like among Migrants in Ireland and Australia.” World Englishes 40 (4): 562–577. expanded the focus of the investigation by analyzing whether predictions about frequency, position and pragmatic functions of a specific PM can be extended to other types of PMs. To do so, she analyzed the use of you know and like by L1 (Australian, Irish) speakers and L2 (Polish, Chinese) users of English. Findings suggested that the PM use by L1 and L2 users of English differed in terms of position and functions. Regarding frequency, the rate of PMs by L2 users was commensurate with that of L1 users. While Diskin (2017)Diskin, Chloé 2017 “The Use of the Discourse-Pragmatic Marker ‘Like’ by Native and Non-Native Speakers of English in Ireland.” Journal of Pragmatics 120: 144–157. and Diskin-Holdaway (2021)Diskin-Holdaway, Chloé 2021 “ You Know and Like among Migrants in Ireland and Australia.” World Englishes 40 (4): 562–577. did not find extensive differences in terms of frequencies between L1 and L2 users of English, it needs to be acknowledged that studies conducted on more temporary mobility experiences (Liao 2009Liao, Silvie 2009 “Variation in the Use of Discourse Markers by Chinese Teaching Assistants in the US.” Journal of Pragmatics 41 (7): 1313–1328. ; Magliacane and Howard 2019Magliacane, Annarita, and Martin Howard 2019 “The Role of Learner Status in the Acquisition of Pragmatic Markers During Study Abroad: The Use of ‘Like’ in L2 English.” Journal of Pragmatics 146: 72–86. ) presented different findings.

Cross-sectional research on PMs has also found a transfer effect because L2 users either tend to frequently use specific PMs that are common in their L1s or underuse PMs that have not a specific L1 equivalent. For example, Diskin-Holdaway (2021)Diskin-Holdaway, Chloé 2021 “ You Know and Like among Migrants in Ireland and Australia.” World Englishes 40 (4): 562–577. found a more frequent use of you know by her Polish participants, when compared to the L2 users from China, and ascribed such finding to the presence of a similar PM (wiesz) in Polish. Similarly, Liao (2009)Liao, Silvie 2009 “Variation in the Use of Discourse Markers by Chinese Teaching Assistants in the US.” Journal of Pragmatics 41 (7): 1313–1328. , ascribed a less frequent use of the PMs well and I mean in L2 English to the lack of a Chinese equivalent.

Regarding pragmatic functions, L2 users display a wider variability. Firstly, they tend to re-interpret the use of PMs. For instance, House (2009)House, Juliane 2009 “Subjectivity in English as Lingua Franca Discourse: The Case of You Know .” Intercultural Pragmatics 6: 171–194. observed that you know, which is mainly an interpersonal marker in L1 speech, was more frequently uttered by German users of English for textual purposes. Secondly, L2 users may create their own PMs, especially through pragmatic transfer. An example is the use in L2 English of in my point of view instead of the expressions from my point of view or in my view (Mauranen 2012Mauranen, Anna 2012 Exploring ELF: Academic English Shaped by Non-Native Speakers. Cambridge: Cambridge University Press.).

Position is an under-researched aspect of PM use in the L2. Some exceptions are Diskin (2017)Diskin, Chloé 2017 “The Use of the Discourse-Pragmatic Marker ‘Like’ by Native and Non-Native Speakers of English in Ireland.” Journal of Pragmatics 120: 144–157. , previously discussed, Nestor et al. (2012)Nestor, Niamh, Caitríona, Ní Chasaide, and Vera Regan 2012 “Discourse ‘Like’ and Social Identity — A Case Study of Poles in Ireland.” In New Perspectives on Irish English, ed. by Bettina Migge, and Máire Ní Chiosáin, 327–353. Amsterdam: John Benjamins. and Nestor and Regan (2015)Nestor, Niamh, and Vera Regan 2015 “The Significance of Age and Place of Residence in the Positional Distribution of Discourse Like in L2 Speech.” In Pragmatic Markers in Irish English, ed. by Carolina P. Amador-Moreno, Kevin McCafferty, and Elaine Vaughan, 408–432. Amsterdam: John Benjamins. . Nestor and colleagues focused on the use of like by L2 users of English residing long-term in Ireland. Although a frequent use of like in L2 English speech was present, participants predominantly used it in initial position and did not use the clause-final like, which is a salient feature in Irish English.

Other cross-sectional studies have explored the role of variables such as proficiency, intensity of interaction and L1 on PM use, finding significant effects. Neary-Sundquist (2014)Neary-Sundquist, Colleen 2014 “The Use of Pragmatic Markers Across Proficiency Levels in Second Language Speech.” Studies in Second Language Learning and Teaching 4 (4): 637–663. observed a positive effect of proficiency on the frequency and variety of PMs, with proficient L2 speakers using you know more frequently. Hellermann and Vergun (2007)Hellermann, John, and Andrea Vergun 2007 “Language Which Is Not Taught: The Discourse Marker Use of Beginning Adult Learners of English.” Journal of Pragmatics 39 (1): 157–179. found that L2 users with more interaction with L1 speakers increased their use of like, you know, and well.

A few SA studies have explored PM development with a longitudinal design (Magliacane 2020Magliacane, Annarita 2020 “Erasmus Students in an Irish Study Abroad Context: A Longitudinal Analysis on the Use of ‘Well’ and ‘Like’.” Study Abroad Research in Second Language Acquisition and International Education 5: 89–117. ; Magliacane and Howard 2019Magliacane, Annarita, and Martin Howard 2019 “The Role of Learner Status in the Acquisition of Pragmatic Markers During Study Abroad: The Use of ‘Like’ in L2 English.” Journal of Pragmatics 146: 72–86. ). Magliacane and Howard (2019)Magliacane, Annarita, and Martin Howard 2019 “The Role of Learner Status in the Acquisition of Pragmatic Markers During Study Abroad: The Use of ‘Like’ in L2 English.” Journal of Pragmatics 146: 72–86. and Magliacane (2020)Magliacane, Annarita 2020 “Erasmus Students in an Irish Study Abroad Context: A Longitudinal Analysis on the Use of ‘Well’ and ‘Like’.” Study Abroad Research in Second Language Acquisition and International Education 5: 89–117. observed that SA in Ireland resulted in an increase in PM production by Italian users of English. However, participants used PMs predominantly for textual functions (e.g., buy thinking time), while interpersonal PMs (e.g., reaction what the speaker has said) presented differences with the Irish English group. Both studies posited that a semester abroad aided PM production but, unlike studies focusing on a longer-term mobility (Diskin 2017Diskin, Chloé 2017 “The Use of the Discourse-Pragmatic Marker ‘Like’ by Native and Non-Native Speakers of English in Ireland.” Journal of Pragmatics 120: 144–157. ), it did not suffice to approach PM frequencies of Irish English speakers. More specifically, by relying on a comparative analysis on the use of like by SA students and au pairs, Magliacane and Howard (2019)Magliacane, Annarita, and Martin Howard 2019 “The Role of Learner Status in the Acquisition of Pragmatic Markers During Study Abroad: The Use of ‘Like’ in L2 English.” Journal of Pragmatics 146: 72–86. found that it was not only the LoR or the amount of input that affected PM development, but rather the quality of such input and the type of interactions while abroad. The SA students interacted predominantly with co-patriots and other international students and lamented contact with members of the TL community. The au pairs were living with Irish families for whom they were carrying out childcare duties and, consequently, were in an environment which could offer more potential opportunities to converse with locals. However, the host parents were often taken by work commitments and participants spent most of the time with young children or by themselves with little possibility of engaging in long conversations with adults or people of their age during their stay abroad.

Altogether, SA studies on PMs exclusively focused on production and a focus on PM awareness is lacking. An exception is Davydova et al.’s (2017)Davydova, Julia, Agnieszka E. Tytus, and Erik Schleef 2017 “Acquisition of Sociolinguistic Awareness by German Learners of English: A Study in Perceptions of Quotative Be Like .” Linguistics 55 (4): 783–812. study on L2 learners’ perceptions of the quotatives be like and say, which are often used in conversation to introduce reported or inner speech. Participants evaluated the social attributes associated with their use and identified pragmatic functions in context. The results showed that L2 users’ evaluations were similar to L1 users’ ones in that they considered the use of be like as more fashionable but also a symptom of less educated social strata. Nonetheless, it needs to be acknowledged that there is no consensus regarding the inclusion of be like under the PM category as it is not syntactically optional. At the same time, this quotative seems to be predominantly used as a face-saving act to mitigate inner thoughts (Beeching 2016Beeching, Kate 2016 Pragmatic Markers in British English: Meaning in Social Interaction. Cambridge: Cambridge University Press. , 131), i.e., to convey a relationship between the speaker and the intended message, rather than for actual reporting. Therefore, although its inclusion in the PM category is still in dispute, the study by Davydova et al. (2017)Davydova, Julia, Agnieszka E. Tytus, and Erik Schleef 2017 “Acquisition of Sociolinguistic Awareness by German Learners of English: A Study in Perceptions of Quotative Be Like .” Linguistics 55 (4): 783–812. is still a key contribution to ILP awareness as it provided an understanding of how L2 users attribute pragmatic meanings to linguistic items in context.

Against this background, this study explores the development of PM awareness during SA by addressing the following research questions (RQs):

3.1Participants

Eighteen Chinese university students in their first semester in the UK (n = 12) and Ireland (n = 6) participated in the study. The sample included four males and fourteen females, with an average age of 25.3. Some participants (n = 6) were second-year undergraduate students majoring in different (Communication, Architecture, Economics, International Business, English) degrees. The other participants (n = 12) were postgraduate students primarily enrolled on language-related programs (n = 10), except for two, who were attending a master’s in nursing.

Their onset English proficiency was at least upper-intermediate (according to the Common European Framework of Reference for Languages), as this is an entry requirement for both universities involved in the study. For all participants, this experience was the first in an English-speaking country and they all arrived in the host institutions at the beginning of the academic year 2022/2023. During their stay, they lived in university accommodation, sharing their flat with both co-nationals and other international students. Moreover, all attended modules in English at their host institution. With regard to their previous knowledge of PMs, following their responses in the PM awareness test (Magliacane and Sánchez-Hernández 2024Magliacane, Annarita, and Ariadna Sánchez-Hernández 2024 “Pragmatic Marker Awareness Test.” Questionnaire from Development of Pragmatic Awareness During Study Abroad: A Focus on Pragmatic Markers [Text/Materials]. University of York: IRIS Database. )11.The PM awareness test discussed in this article can be found in the IRIS repository. and the interviews, none of the participants had any instruction on L2 pragmatics prior or during the sojourn abroad, and only two had some basic knowledge of PMs at the beginning of the study.

3.2Data collection

Employing a pre- and post-test design, data collection took four months. After receiving ethical approval in the institutions involved in the study, a call for participants was circulated. Students interested in participating contacted the researchers, and if they met the recruitment criteria (L1 = Chinese, first time in an English-speaking country, arrival at the beginning of the academic year), they were sent the link to the PM awareness test. After signing a consent form, they were asked to complete the test and participate in an online semi-structured reflective interview within a week of completion of the test. The post-test was administered four months after, and the participants were invited to join a second interview.

3.3Data coding

Pragmatic awareness was operationalized as PM recognition and metapragmatic awareness of PMs. Recognition was explored in terms of the ability to recognize PMs in the oral excerpts and in their corresponding transcripts. As discussed in Section 3.2.1, the four excerpts contained a total of twenty-eight PMs. The responses of all participants for each PM were computed and then an average score value per PM was generated. The minimum value for such value was 0 in the event none of the participants recognized the PM and a maximum of 1, if all participants recognized it. The position of the mean recognition score along the continuum 0–1 indicated to what extent participants on average managed to recognize the specific PM. For example, the average score of 0.61 indicated that 61% of the participants provided a correct response and, consequently, recognized that PM.

Metapragmatic awareness of PM use was measured in terms of the ability to identify PMs and their function in context. Each excerpt contained four highlighted words, with a total of sixteen words to identify (fourteen PMs and two distractors). Subsequently, the participants were asked what function the PMs played in conversation. One point was awarded for each PM and function correctly identified. The responses of the participants were compared with the pragmatic functions identified by the researchers. These functions were identified by consulting previous literature (Aijmer 2013Aijmer, Karin 2013 Understanding Pragmatic Markers: A Variational Pragmatic Approach. Edinburgh: Edinburgh University Press. ; Beeching 2016Beeching, Kate 2016 Pragmatic Markers in British English: Meaning in Social Interaction. Cambridge: Cambridge University Press. ; Bolden 2009Bolden, Galina B. 2009 “Implementing Incipient Actions: The Discourse Marker ‘So’ in English Conversation.” Journal of Pragmatics 41: 974–998. ; House 2013House, Juliane 2013 “Developing Pragmatic Competence in English as a Lingua Franca: Using Discourse Markers to Express (Inter)Subjectivity and Connectivity.” Journal of Pragmatics 59: 57–67. ). For example, following Beeching (2016)Beeching, Kate 2016 Pragmatic Markers in British English: Meaning in Social Interaction. Cambridge: Cambridge University Press. , the function of ‘you know’ in the extract “[…] the house had to be repaired to certain historic standards too so that may have added to the timeline and also you know pandemic timeline it,’s more difficult to get projects done […]” (S1_pre-test) was gauged to be an example of “[a]ttention-getting/launching a new piece of information” (Beeching 2016Beeching, Kate 2016 Pragmatic Markers in British English: Meaning in Social Interaction. Cambridge: Cambridge University Press. , 101). Researchers first identified PMs and the pragmatic functions within the transcripts individually. Inter-rater agreement was then calculated, showing a coding agreement of 97.9% for the pre-test and 97.9% for the post-test responses. The few occurrences where disagreement was present were further discussed by the researchers, who also consulted L1 users of English for additional feedback, until an agreement was reached. As previously mentioned, the identified functions were used as a baseline to analyze the responses of the participants and as previously discussed, points were awarded to the correct identification of PMs and functions by the participants. Mean score values were then calculated to analyze trends within the data.

3.4Data analysis

Once the data was coded, normality was checked with Shapiro-Wilk tests. With regard to recognition scores (RQ1), data elicited with the pre-test did not present a normal distribution as evident by the p values of the audio (D(28) = 0.877, p = 0.003) and written scores (D(28) = 0.905, p = 0.015). Therefore, the non-parametric related-samples Wilcoxon signed-rank test was used to compare the recognition scores in oral and transcribed speech at the pre-test. The nature of this test was two tailed. Additionally, we measured the strength of the relationship between the two skills by calculating the effect size through Pearson’s r, a score that ranges from −1 to 1, and indicates that the larger the effect size the stronger the correlation between the two variables. More particularly, a r coefficient between 0.1 and 0.29 indicates a small effect size, a r between 0.3 and 0.49 is a moderate effect size, and a r between 0.5 and 1 corresponds with a large effect size. Finally, standard errors (SE) were calculated, showing how accurately the sample data represents the whole population.

RQ2 was about metapragmatic awareness, operationalized as PM identification and function identification. The results of the Shapiro-Wilk test confirmed a normal distribution of the data in the pre-test (PM identification: D(14) = 0.923, p = 0.243; function identification: D(14) = 0.922, p = 0.239), Consequently, paired-samples t-tests were used to explore the differences between awareness of PMs and awareness of their functions in the pre-test. The nature of the t-tests was two-tailed, and the confidence interval percentage was 95%. Furthermore, Cohen’s d coefficient was used to calculate the effect size between the two skills. According to Cohen (1988)Cohen, Jacob 1988 Statistical Power Analysis for the Behavioral Sciences. New York: Routledge. , a d may indicate a small (d = 0.2), medium (d = 0.5) or large (d = 0.8) effect size. In addition to this, a paired-samples correlation was conducted to explore whether the two skills are related or not.

As RQ3 aimed at assessing longitudinal changes, we firstly checked the normality of the data elicited in post-test. Post-test scores did not present a normal distribution of the data (audio recognition: D(28) = 0.902, p = 0.013; written recognition: D(28) = 0.894; p = 0.008; PM identification: D(14) = 0.770, p = 0.002), except for the pragmatic function identification scores (PM function: D(14) = 0.984, p = 0.991). Consequently, Wilcoxon signed-rank tests and their corresponding effect sizes (r) and SE were used to explore differences across post-test scores. With regard to longitudinal differences, as the condition of normality for the data was respected for the pragmatic function identification scores of both tests, paired t-tests analyzed such differences, and that was complemented with the effect size (d). For all other longitudinal analyses, the non-parametric Wilcoxon signed-rank tests were used. For all statistical tests related to RQ1, RQ2, and RQ3, the statistical significance value was established at p < 0.05.

These statistics were complemented with the thematic analysis (Braun and Clarke 2006Braun, Virginia, and Victoria Clarke 2006 “Using Thematic Analysis in Psychology.” Qualitative Research in Psychology 3 (2): 77–101. ) of the interview data. This approach firstly involved getting familiarized with the transcripts. Then, the data was qualitatively coded by producing notes and highlighting details of interest. The codes were then grouped under overarching themes. For the scope of this article, the data included under the themes ‘PM understanding’, ‘PM exposure’, and ‘social contact abroad’ were considered as they were relevant for the purpose of the study and to further understand the trends of the quantitative data. ‘PM exposure’ included excerpts dealing with ‘PM coverage in textbooks’ as well as ‘PM use by other interlocutors’.

4.1Pragmatic recognition

RQ1 analyzed the PM recognition in oral conversation and transcribed speech. Two tasks were considered for this question: (a) the audio recognition task, where the participants recognized PMs in oral extracts, and (b) the written recognition task where participants recognized PMs in the transcripts. We first calculated the average number of PMs recognized by all participants across the two modalities of the pre-test. Table 3 provides an overview of this data by presenting the descriptive statistics, including mean values (M), standard deviations (SD), the lowest (Min) and the highest scores (Max). As previously mentioned, the recognition tasks included twenty-eight PMs (N), as the four excerpts contained respectively five, four, ten and nine PMs.

As Table 3 shows, our participants seemed to outperform in written PM recognition (M = 0.370) as compared with oral PM recognition (M = 0.253). This finding was statistically confirmed through a Wilcoxon signed-rank test, which showed a significant difference between oral and written recognition, with scores on the written activity being significantly higher (z = 2.717, p = 0.007). A medium effect size was also identified (r = 0.513), indicating a moderate strength of the relationship between oral and written PM recognition, and the SE of the test was 30.737.

In sum, the quantitative analysis conducted to address RQ1 suggests that the participants recognized more PMs in writing than in oral speech. These findings were triangulated with the results from the qualitative data and, in particular, the interview data under the themes ‘PM exposure’, ‘social contact abroad’, and ‘PM understanding’ resulted revealing to understand such trends. At the beginning of the study, participants acknowledged poor classroom exposure to authentic conversation and PMs. This is also documented in previous literature (Liao 2009Liao, Silvie 2009 “Variation in the Use of Discourse Markers by Chinese Teaching Assistants in the US.” Journal of Pragmatics 41 (7): 1313–1328. ) and also illustrated by UK_P1_T1,22.Codes preserved anonymity. They consisted of three parts: the first (UK, IrE) indicated the destination of stay, the central one showed the number assigned to each participant (P) and the last part indicated whether the interview took place after the pre- (T1) or post-test (T2). who stressed that the teaching materials in their home country are generally devoid of such linguistic items:

I found most of them [excerpts] — they spoke in a normal rather than standardized listening recording — so I think they might originate from some real dialogue from the native speakers […] discourse markers place a very important role in linking the words and to — make the conversation more like a real dialogue […] but most of the recordings in my textbook in the past in China they’re consistent over standardized dialogue — there’s no pause and no informal words like you know […]

However, despite the acknowledgement of the presence of PMs in the extracts, the results of the quantitative analysis showed that the oral recognition of PMs was significantly different from the recognition in writing.

Considering that SA contexts offer the potential for more TL exposure, our qualitative analysis also focused on how out-of-class exposure could affect pragmatic recognition. Following what participants reported, the difficulty of going beyond the social circle of co-nationals could have affected audio recognition. When the pre-test was conducted, participants predominantly interacted with other Chinese students as this cohort was the largest group in the student population (IrE_P15_T1: “I don’t have Irish people [friends] yeah — because my classmates — the big portion is Asian people”) or lamented little time to expand their social networks or for non-academic related conversations (UK_P12_T1: “I didn’t find too much time for chatting with native speakers except for the time I studied […] and shared my ideas in the class with my peers or academic instructors […]”). Therefore, prior to SA, participants received limited PM exposure. During their first months abroad, they did not report more out-of-the-class interactional opportunities, which may explain the significant differences between oral and written recognition scores.

In addition, in the interview after the pre-test, some participants reported difficulties in identifying PMs in daily conversation as they were too concerned about understanding the content to notice PMs: “[…] for the listener — these words are a little bit distract- distracting — distract my concentration *to the meaning of the sentence” (IrE_P14_T1). Consequently, a lack of familiarity with appreciating PMs in everyday oral exchanges may corroborate the significant difference between oral and written recognition scores.

4.2Metapragmatic awareness

Our RQ2 analyzed PM metapragmatic awareness. Two tasks of the test addressed this question. In the first task, the participants identified whether the highlighted words in the transcripts were PMs or not. The second task focused on the identification of the function of PMs in context. Both tasks involved working on four words per transcript, with a total of sixteen words per task. However, two words were distractors, so the total number of PMs were fourteen (N). Table 4 summarizes the descriptive statistics of PM metapragmatic awareness scores in the pre-test.

As indicated in Table 4, a comparison between the mean score of PM identification (M = 0.721) and that of pragmatic function identification (M = 0.286) indicates that participants outperformed in the PM identification task, while the identification of the pragmatic function in context represented a higher difficulty. A series of paired-samples t-tests confirmed that there was a significant difference between the PM identification scores and the scores of pragmatic function identification (t(13) = 5.474, p = 0.000). Moreover, a moderate effect size (d = 0.297) was identified. Hence, it can be affirmed that the participants significantly outperformed in the PM identification activity and presented less awareness about the pragmatic functions in context. In addition, we conducted a paired-samples correlation which revealed a non-significant relationship between PM identification and pragmatic function identification (r = 0.073; p = 0.804), pointing out that one participant could have been aware of what a PM was, but they were not necessarily able to discern the function in conversation.

In summary, the quantitative results related to RQ2 revealed that at the pre-test, it was easier for the participants to identify PMs rather than to determine what functions PMs play in conversation. This finding seems to corroborate previous SA studies (Alcón-Soler and Sánchez-Hernández 2017Alcón-Soler, Eva, and Ariadna Sánchez-Hernández 2017 “Learning Pragmatic Routines During Study Abroad: A Focus on Proficiency and Type of Routine.” Atlantis 39 (2): 191–210.; Almalki and Jones 2022Almalki, Ziad, and Christian Jones 2022 “ ‘Why Did You Do That?’ The Effects of Instruction on Recognition and Production of Informal Second Party Complaints.” The Language Learning Journal 50 (4): 443–459. ) which found that during the first months abroad, learners start noticing pragmatic phenomena and assigning plausible meanings. Our pre-test was conducted within the first trimester abroad, a period which, as mentioned in Section 2.1, has been considered crucial by previous SA studies on pragmatic awareness and our findings posit that participants started noticing PMs during such period, but the association of pragmatic meanings to these PMs was not extensively developed yet. These findings suggest that noticing pragmatic phenomena is a preliminary step for understanding their use in context. However, such correlation was not significant, as while some PM awareness was present for our participants at the pre-test, they did not present a full metapragmatic awareness towards their use in context.

These findings were also reflected in the interview data. Participants were aware of the role of PMs in conversation (UK_P7_T1: “these words can help us know the structure produced by others — that’s a key point for us to comprehend what they are talking about and also their feelings”) and specific pragmatic use (IrE_P1_T1: “these words give me some time to think”) but they also reported difficulties in identifying PMs in context (UK_P1_T1: “it’s a little bit hard for me to decide whether they are using the pragmatic markers”; UK_P17_T1: “[in the questionnaire] they ask you — what do you think about like so […] this kind of words — I didn’t notice them before I *take the questionnaire — so it took me a while to figure out and to express my own opinion about them”). In sum, the quantitative and qualitative analysis conducted to respond to RQ2 indicate that at the beginning of their stay abroad, participants started to identify PMs, but they did not have a full awareness about their functions in context.

4.3Development of L2 pragmatic awareness

RQ3 analyzed the development of L2 pragmatic awareness, understood as recognition and metapragmatic awareness of PMs, over four months in an English-speaking country. The description of this longitudinal analysis will be two-fold. We first examined whether participants presented longitudinal differences in audio and written recognition of PMs. Subsequently, we investigated longitudinally the results of metapragmatic awareness of PMs and their functions in context.