Acquiring the Pragmatics of a Heritage Language: A Case of Study Abroad Experience in Greece

Abstract

Throughout the English-speaking world, there are numerous Greek-speaking diaspora communities whose language is simultaneously influenced by English and local varieties of Greek. This study builds on the body of knowledge in cross-cultural and interlanguage pragmatics to explore a case of pragmatic acquisition in a study abroad context by one member of such a community. Data were collected from a third-generation young adult Greek Australian student prior to commencement of a 6-week Greek language programme in Athens, and on three other occasions. She described her experiences and responded to a set of scenarios involving Greek requests, refusals and apologies. The responses were analysed using established frameworks and subjectively evaluated for appropriateness by a matched Greek native speaker. The student showed evidence of a shift towards documented Standard Modern Greek pragmatic norms in some but not all speech acts, and change appeared to be loosely linked to opportunities for use. There was also some evidence of reversion to diaspora variants after her return. This study contributes to our understanding of the interaction between learning outcomes, individual learner variables, prior exposure, the nature of communicative events and levels of pragmatic awareness. It is argued that Greek and diaspora contexts involve subtly distinct pragmatic varieties of Greek and that learners can benefit from explicit awareness-raising regarding the nature of these differences.

1. Introduction: Study Abroad and Heritage Speakers of Greek

In comparison with other migrant communities in Australia, Greeks have shown remarkable determination to maintain their culture and language (Lo Bianco, 2021). The sense of connectedness to Greece plays an important role in the motivation for language maintenance, and the participation of young people in study abroad (SA) can assist in promoting and strengthening cultural bonds. Heritage speakers (HS) from the Australian Greek diaspora who wish to explore opportunities to enhance their proficiency in their ancestral language tend to be focused on personal and cultural exploration (Isabelli-García et al., 2018; Shively, 2016), which differs from the more instrumental motivation of non-Greek background economic migrants.

There are various programmes in which Australians of Greek heritage can undertake language study in Greece. Some of these are general-purpose courses in which heritage background students are co-taught with Greek as a foreign language (GFL) learners from a variety of backgrounds, while several Australian universities offer specially designed study tours catering more directly to their needs and interests. These immersion programmes vary in length from two weeks to a full year and are also likely to differ in the emphasis placed on the balance of explicit and implicit instruction in pragmatics. However, even without explicit teaching, opportunities for real-life interaction with local speakers can provide the basis for the development of pragmatic awareness and enhance learners’ ability to succeed in more nuanced communication. This is why SA is considered a valuable opportunity for increasing awareness and ability in pragmatics, (Barron, 2020; Bloom & Miranda, 2015).

While the pragmatic knowledge which Greek Australians bring to SA contexts can lead to the expectation of more nuanced communication compared to GFL, the idea of a heritage advantage in pragmatics is not universally accepted. Heritage speakers often blend pragmatic conventions from their dominant language (in this case English) with their heritage language (Rakhilina et al., 2016, Xiao-Desai, 2019). There is not always a clear relationship between directness and politeness (Grainger & Mills, 2016), and “folk” knowledge of cultural preferences can be misleading. As well as the differences between standard forms of English and Greek, they can be influenced by contact sociopragmatic variants used in the diaspora context. These incorporate the influence of Australian English, locally emergent varieties of Greek, and the social and regional variants used by their original migrant ancestors. Pragmatic transfer into Standard Modern Greek (SMG) may occur based on any of these factors, without its origins being readily traceable.

Although there have been some investigations of Australian Greek as a variant (Tamis, 2010; Alvanoudi, 2019), most of the evidence of Greek pragmatic hybridisation remains anecdotal and has not been systematically studied. In addition, while there is an extensive literature on SA programmes (for example, see Tajeddin and Khanlarzadeh (2024) for a recent review of factors that individuals bring to pragmatics learning), the vast majority concern the English language, and there has been little attention to Greek. Through a close examination of the experience and pragmatic development of one Greek Australian student in an SA context, this longitudinal case study aims to contribute to the understanding of Greek interlanguage pragmatics. It brings together a learner’s perspective on her own experiences and pragmatic development drawn from a set of interviews before, during and after the SA experience, and considers her responses to a set of scenarios eliciting three well-researched speech acts: refusing, requesting and apologising. Two approaches to evidence of learning are employed: using existing frameworks to map how specific language choices have aligned with GFL progress towards native speaker (NS)-like proficiency (a construct which will be critically examined later in this article) and evaluation of the responses by an age, education and gender-matched native speaker. In addition, interviews with the participant have allowed her to take an active role in reflecting on, evaluating, and offering personal insights into her own development.

2. Literature Review

2.1. Native Speakers, Heritage Speakers and Learners: Defining the Terms

The terms heritage speaker (HS) and heritage learner (HL) are used differently across subfields, reflecting distinct theoretical and pedagogical priorities. HS has been used since the 1970s to denote naturalistic bilinguals who were exposed to an ancestral language in early childhood and acquired it in a natural context through family interactions, without assuming that they will have reached the same level of proficiency displayed by monolingual speakers raised in locations where the heritage language is dominant. HS are usually most proficient in the dominant societal language, and do not necessarily have full command of all skills and registers in the heritage language (Montrul, 2015; Polinsky, 2018). The focus in this tradition is on acquisition history and linguistic outcomes, and heritage speakers are commonly contrasted with second language (L2) learners. HL is more commonly applied in linguistics and language education. It can either be contrasted with HS, denoting a person with cultural or family connections to a language they did not acquire in childhood and which they learn later in life through formal study, or used as an umbrella category including HS. The latter approach is more commonly found in the work of researchers with a pedagogical focus (Valdés, 2000; Carreira, 2004). In sociolinguistic and theoretical research on bilingualism, a distinction between the two categories is more useful, because of its relevance to the broader debate on the nature of heritage and native speakers.

In the literature of bilingualism, it has been argued that because of the nature of their acquisition, heritage speakers constitute a subset of the category of native speakers (Montrul, 2022; Rothman & Treffers-Daller, 2014; Wiese et al., 2022) and, as such, it promotes an undesirable deficit perspective to evaluate their language output using different types of native speakers as a baseline (Xiao-Desai & Wong, 2025). In addition, high levels of variation from canonical forms and variations in competence have been observed among monolingual non-heritage speakers (Wiese et al., 2022; Vulchanova et al., 2022), indicating assumptions of a native speaker standard need to be closely examined on theoretical as well as ideological grounds. Wiese explicitly advocates for a research paradigm which “does not take monolingual norms as a yardstick to identify what is missing or incorrect in a heritage speaker’s language use” (Wiese et al., 2022, p. 3).

However, as Montrul has acknowledged, “On a variety of linguistic measures, young adult heritage speakers vary significantly from prototypical native speakers, and the questions are how they vary and why” (Montrul, 2022, p. 9, emphasis mine). The same variance phenomenon has been observed with pragmatic features (Dubinina, 2021), and the same questions apply. Informed by these debates, this study focusses not only on how the pragmatic systems of diaspora Greeks differ from prototypical or un-interrupted native speakers, but how an individual may change as a result of explicit or implicit input. Thus, any investigation of changes to pragmalinguistic and sociopragmatic choices occurring during and after time spent using the heritage language in a context where it is dominant must involve some kind of comparison with local norms. The conceptualization of the adoption of closer approximation as progress in learning is not intended to devalue heritage sociopragmatic variants. On the contrary, evidence of the adoption of dominant language norms reflects progress made in broadening an individual’s repertoire of contexts of competence. It satisfies the motivation of heritage Greeks who readopt the role of learner in order to achieve a kind of pragmatic competence that they themselves would view as more native-like and would thus enhance the probability that they will be accepted and included.

2.2. Developing Pragmatic Competence in Greek

Over recent decades there have been numerous investigations of the pragmatics of Greek learners of English (Economidou-Kogetsidis, 2008, 2021; Economidou-Kogetsidis et al., 2018; Ogiermann & Bella, 2020), some of which consider pragmatic transfer. However, systematic research on Greek pragmatic interlanguage development remains extremely limited. The most significant body of work from a speech act perspective has been conducted by Spyridoula Bella, whose publications between 2009 and 2016 examined learners’ pragmatic competence in the production of invitations, refusals, requests, apologies and offers. In an article of this scope it is impossible to do full justice to the breadth, depth and value of this work. All the studies involved comparisons of different groups: for example, learners and NS1, learners at different stages of development, and SA (GFL) as opposed to economic migrants. Most also used post-task reflections to capture the perspective of the speakers. The key studies will be summarised below, and their findings and their contribution to the development of the present study outlined.

As part of the groundwork for her study of Greek interlanguage pragmatics, Bella transcribed utterances from 60 Greek NS, role-playing the speech act of making an invitation and responding to an interlocutor’s initial refusal (Bella, 2009). The study confirmed that age was a significant variable in the choice of politeness strategies. Older speakers were more likely to rely on negative politeness strategies in the invitation and to accept a refusal without moving on to an insistence move. Younger speakers were more direct in their invitations, with a higher frequency of positive politeness strategies. Although the study did not focus on interlanguage, it helped to shape the methodology of Bella’s subsequent studies. When this paper was published there were no prior studies of inviting in Greek, and to the author’s knowledge there have not been any since.

The first of the interlanguage studies (Bella, 2011) built on this work. The focus was on refusing an invitation, a dispreferred, face-threatening speech act which is usually accompanied by mitigation strategies and may occur as part of a sequence involving instances of insistence and follow up refusals. Data consisted of audio recorded roleplays. The two learner categories under comparison were SA learners of GFL who had been in Greece a short time, and economic migrants who had longer residence in Greece but fewer opportunities to have close relationships with native speakers. Both groups were compared to a group of native speaker university students, using a framework based on Brown and Levinson’s model of politeness (Brown & Levinson, 1987). The findings indicated that intensive exposure led to more nativelike behaviour than length of stay. It should be noted that there was no controversy about the use of the terms “NS” or “NS-like” in these studies because the contrast was being made between locally raised and educated Greeks and learners of Greek as a foreign language.

Bella’s second study on refusals (Bella, 2014b) was a cross-sectional comparison of NS with SA GFL learners at three different levels, and the main aim was to identify developmental patterns. There was no explicit mention of heritage speakers, but learners who had spent any amount of time in Greece were excluded from participation. The three roleplays in this study extended the research into a wider range of settings in which different power distance relationships were explored. Findings supported the existence of developmental stages for refusals, analogous to those identified in previous studies for the acquisition of requests in English (Ellis, 1992; Achiba, 2003). Unanalysed formulae were produced by the lower-level learners, with intermediate and advanced learners expanding their repertoire of strategies, demonstrating shifts towards greater indirectness (particularly in the informal situations) and increased use of mitigation. Even the advanced speakers failed to master all the subtleties and complexities that featured in the native speaker responses, struggling to mitigate their refusals in informal contexts, but, somewhat surprisingly, overusing indirect strategies in the formal situation. Where explanations for the refusal were given, patterns emerged in topic choice with NS preferring to give excuses related to family and health. Relative to all learner groups, NS were less blunt in their responses, and the solidarity strategies they employed (such as indefinite replies and promises to comply) were largely absent from the learner data. Bella identified this as due to a lack of sociocultural competence, although it could also be argued that even if the learners were aware of what native speakers would do in the situation, they may have felt it was more appropriate to perform the speech act as their own, essentially outsider, selves.

The most broadly investigated Greek speech act in Bella’s pioneering work (Bella, 2012b, 2016) was requesting. Prior to any research into the pragmatics of Greek requests, there was a very large body of work in other languages, (see Cañas Villarreal, 2025, for a recent summary) including studies involving the requesting behaviour of Greek learners of English (Economidou-Kogetsidis, 2008, 2010). Frameworks for linguistic and strategy analysis were already well established, and these could be adapted with relative ease. As per the work on refusals: one study of requesting concerned the length of residency vs. intensity of interaction, and one compared the output from lower, intermediate and advanced learners with the aim of identifying the characteristics of developmental stages. The residence vs. intensity research (Bella, 2012a) involved three groups of participants, NS, LS1 (high interaction) and LS2 (longer residence), and used three scenarios with different PD relationships. There was one major methodological difference, however, in that data were collected using DCTs, a decision taken for operational rather than theoretical reasons. Modification strategies used in each of these were analysed separately for the NS, length and interaction groups. In line with the study of refusals, learners with more frequent and diverse opportunities for interaction with native speakers (LS1) performed in a more native-like manner in modifying requests than learners with longer residence but fewer interaction opportunities (LS2). In the two informal situations, both sets of migrant learners made choices of syntactic modifications that were different from those of the GFL learners. The difference was less marked in the formal situation, because the formulaic politeness conventions readily available to the learners coincided with the NS choices. These findings supported the position that informal situations in Greek are more challenging for learners to negotiate. This study informs the current research by detailing the modification strategies preferred by young university-educated native speakers in different PD contexts.

In the cross-sectional developmental study (Bella, 2012b), 50 NS were compared with three groups of learners from a range of cultural backgrounds, potentially diluting any effect of pragmatic transfer from the dominant language. Although the data collection method and the scenarios used were the same as the context comparison study (Bella, 2012a), the analysis differed. The scope was broader in that head acts (the core part of the speech act that performs the action) were included, but a smaller range of modification strategies was considered. While advanced learners and NS preferred conventionally indirect requests, non-conventionally indirect requests only appeared in the low-level learners’ responses. Low-level learners also used more frequent direct requests, which it was argued was probably due to limitations in general language proficiency rather than a lack of pragmatic competence. This ambiguity is indeed one of the major challenges in interpreting performance data, and one of the reasons why post-task interviews can be a valuable source of supplementary data (Woodfield, 2010).

Two features characterising NS head acts and rarely found in the learner output were the sub-strategies of query preparatory with present indicative, (QPI) and negative–interrogative suggestory formula, which will be discussed in more detail in the analysis of the data in the present study. These forms, used among familiar speakers of equal status, are very common in the Greek language and align with its solidarity politeness orientation (Sifianou, 1992). The nature of the scenario influenced the choice of lexical modifiers in all groups. In terms of external modification, the picture was complex. Lower-level learners produced the lowest frequencies and relied on a restricted range of strategies: using grounders in all three situations and apologies in the formal one. When all three scenarios were considered, the two lowest level groups showed a higher frequency of external mitigation strategies, while the advanced level usage was unusually low, and below that of the NS. This suggests that advanced learners were passing through a transient stage of underuse, choosing different ways to manage the face threat of the requests.

The third speech act considered in the current study is apologies. Although there had been developmental studies in English (Trosborg, 1987; Flores Salgado, 2011; Shardakova, 2005), Bella (2014a) was the first interlanguage study involving Greek. Prior to this, Symeon (2000) had compared Greek and English apologies. Her work confirmed that Greek apologies were consistent with a positive politeness orientation and showed a higher frequency of statements of regret, self-accusation and excuses. Symeon also noted that when several strategies were deployed in a single apology there were differences in sequencing, with Greek speakers putting the positive politeness strategy before the negative.

One group of learners participated in Bella’s study. All were from England, were of a high level of proficiency (corresponding roughly to C1 on the Common European Framework for Languages) and had not previously spent time in Greece. Data were gathered through an open DCT, with two informal and two formal scenarios of contrasting severity. Participants were also asked to provide their own subjective assessment of the power distance relationship of the interlocutors and the severity of the offence, and retrospective verbal reports were collected from a subset of participants. Although both groups agreed that the offences of being late for an appointment or damaging the property of a professor were events of higher severity than the same occurrences with friends, the Greek NS rated the difference as being more pronounced. Similarly, while both considered an offence against friends as less severe, the difference in closeness between friends and professors was less pronounced for the English speakers. There were also differing perceptions of the severity of the offence, with the Greek NS basing their evaluation on power distance only whereas the nature of the offence was more salient to the learners. Overall, NS strategy choice seemed to count on the solidarity existing between friends more than GFL learners, few of whom would have had the opportunity to experience conducting close friendships with Greek people in the Greek language. As Bella (2014a, p. 707) notes, the differences observed could be attributed to pragmatic transfer, and indeed this is more likely to be an amplified effect in cases where all learners had the same L1. Unlike the requests and refusals research, this study looked at only one level of learners, so did not make any claims about developmental stages.

It must be kept in mind that pragmatic conventions within a culture are not necessarily static. For example, early cross-cultural studies of Greek and English pragmatics (Hirschon, 2001; Pavlidou, 1994; Sifianou, 1992) characterised Greek culture as focussed on positive politeness with a strong emphasis on solidarity within the in-group. This view still tends to predominate, but recent studies have considered the levelling effects of globalisation and proposed a possible shift towards a strengthening negative politeness orientation (Sifianou, 2013; Bella & Ogiermann, 2019). In addition, the rise in international communication through social media reduces the viability of regional variants (Blommaert, 2010) and this may prove eventually to be the case with HS sociopragmatic varieties of Greek.

The above review has dealt in detail with the three speech acts which will be considered in this article. Subsequently, there was a shift in emphasis away from single speech act studies in the broader landscape of pragmatics and this was mirrored in Greek research, although several more studies have been conducted on single speech acts, including offers (Bella, 2016) and gratitude for compliance (Γκούμα, 2023; Γκούμα et al., 2023).

The most studied speech act in Greek continues to be requests, including a corpus study of Greek films (Mavromati, 2021), requests in academic settings (Mavromati & Vassilaki, 2024) and requesting development in child language (Vassilaki & Selimis, 2020). In a cross-cultural study focussing on transfer from L1 to English, Ogiermann and Bella (2020) examined differences in request perspective. A very recent doctoral study examined the acquisition of requests by Spanish/Catalan bilinguals learning Greek (Cañas Villarreal, 2025). In the current volume, written email requests to managers are examined by Bella (2025), extending pragmatic analysis beyond the speech act to include rapport management and issues of identity in the hierarchical workplace.

Despite the move to a discourse-oriented research paradigm situating Greek pragmatics in a broader sociolinguistic context, existing studies of Greek interlanguage development have relied strongly on cross-sectional approaches. As Barron (2020) points out, the exploration of retention with a delayed post-test remains rare. Similarly, the changes taking place in the pragmatic systems of SA learners of heritage background are areas of potential interest and remain unexplored.

2.3. Research Questions

The current longitudinal qualitative case study was designed to build on the foundation of Bella’s work, while at the same time addressing some questions which could not be examined using a cross-sectional quantitative approach.

The questions are as follows:

• What elements of Greek pragmatics feature in an Australian heritage speaker’s reported experience?
• What can current analytical frameworks reveal about development in pragmatic competence during a 6-week SA programme where pragmatics is not explicitly taught?
• What additional information can be captured by evaluation of responses by a local NS informant?
• Is adoption of local NS norms related to in-country experience?
• What happens after return—reversion, retention or continued change towards local NS norms?

It must be noted at the outset that as a single case study involving one participant and one informant, no claims about the generalizability of findings are made. Conclusions are intended to shed light on the possible nature of individual experience, learning and change, and to contribute to the methodological debate surrounding heritage language pragmatics.

3. Materials and Methods

3.1. Participants and Data Collection

After ethics approval for the project had been obtained, convenors of university level study abroad programmes were contacted to identify and invite students of Greek heritage background to participate in the study. The pool of potential participants was quite small. One female Greek Australian student, who will be referred to as Dora, met the criteria of being Australian-born, with a family connection to Greece, childhood exposure to the language and culture, current interest in improving her language proficiency (evidenced by being enrolled in study of Greek language at tertiary level), intent to undertake a study abroad program, and willingness to participate in the study. The distinction between heritage speaker and learner was less clearcut in her case, as she had been exposed to Greek in her early years by contact with her grandparents rather than parents, and had also taken part in community-directed structured learning activities. Details of Dora’s background will be provided in the findings of Research Question 1.

After informed consent was obtained, Dora participated in four conversations with the researcher, which were conducted predominantly but not exclusively in English. The first one was before commencement of the course, one was at the halfway point 3 weeks into the programme, one was after her return and the final one was eight months later. The first two meetings were conducted online and the others face-to-face. They ranged in length from 40 to 25 min. Semi-structured interview questions were used to elicit stories of experience and explore the participant’s subjective conceptualisation of Greek and Australian pragmatics (see Appendix A). They concerned prior perceptions of local Greek and diaspora pragmatic systems and SA experiences, before during and after the study. As the required data had already been collected, no additional discursive interview was conducted after 8 months, and the focus of the final conversation was only on the scenarios. In the first, third and fourth conversations, the prompts were randomised on shuffled cards and descriptions of the scenarios presented verbally in English by the researcher. Dora was encouraged to imagine the scenario and where possible relate it to her real-world experience, before responding in Greek. Responses were audio-recorded and transcribed by the researcher. Several additional scenarios were included. These will be discussed in a separate publication, as they involved contexts and speech events which were outside the scope of Bella’s seminal work, and thus could not be compared with prior findings.

In order to address the third research question, a predominantly monolingual Greek native speaker of the same gender, age and educational background was recruited through the researcher’s personal network. The native speaker was a resident of Athens and was not strictly monolingual in that she had studied English as part of her general education, but she could nevertheless be considered monolingually raised. She was not a fluent speaker of English and had never lived outside her homeland. At the time of the study, she had just arrived on her first visit to Australia, where she had extended family. Like Dora, she was in her early twenties and tertiary educated. As Montrul notes, a prototypically educated native speaker would be expected to “recognize ambiguity and/or multiple interpretations and the pragmatic implications of words and sentences” and to be “attuned to their sociolinguistic implications” in the local pragmalect (Montrul, 2022, p. 15). For this reason, it was considered she could be treated as prototypical for the purpose of the research question. In order not to imply that heritage speakers are not native, she will be referred to as a Local Native Speaker, henceforth LNS.

Transcriptions and translations were checked by the researcher’s colleague who was a lecturer in Greek language in an Australian university. The LNS informant was given a form with the three transcribed responses to each of the scenarios, before commencement, after completion, and 8 months later, and instructions (in Greek) on how they should be rated on a scale of 1–5. These were provided at a face-to-face meeting conducted in Greek with the researcher, and the informant was allowed to complete and return the form in her own time. She knew that speaker(s) were female and of her age and background but was not told anything about the circumstances under which the responses had been produced, or whether they were from a single or multiple speakers.

The interview questions and scenarios can be found in Appendix A. The instructions provided to the NS informant can be found in Appendix B.

3.2. Data Analysis

To answer Research Question 1, all references to Greek and Australian pragmatic norms were identified in the interview data. They were extracted and documented as they arose in the participant’s narrative of experience and learning. Question 2 was approached via a linguistic analysis. Based on the taxonomies developed for each of the speech acts in Greek (Bella, 2011, 2012a, 2012b, 2014a, 2014b), the prior studies were examined to determine the features which could indicate a move towards forms that Bella had identified as native-like. Although it was part of the original research plan to use these different taxonomies, it was a complex process because it had not been part of the brief of the prior studies to create a set of criteria to be used by other researchers and, reflecting the diverse foci of politeness studies, there was a great deal of variation among them. Data collection was sometimes by roleplay and sometimes written DCT, sometimes single stage and others multiple, while analytical classification systems had evolved across the studies and were not identical even for the same speech act. In addition, some features which were meaningful in studies with multiple participants and a statistical approach (for example, the use of a broader range of strategies in one group as opposed to another) were not applicable to a single case study. The fact remains, however, that this was the only existing prior data on Greek interlanguage pragmatics and therefore it constituted the most solid foundation on which to build.

In response to Question 3, the linguistic analysis of each scenario was supplemented by comments and evaluations from the matched LNS informant, who often ranked the effectiveness of the responses differently. As a non-linguist, she commented on aspects that fell outside the scientific pragmatics-focussed taxonomies and helped to provide a wider sense of what pragmatic competence might mean to a participant in a real-world interaction. In the results sections these data are presented together.

The last two research questions, 4 and 5, concern the contexts where change occurred in Dora’s language and evidence for change and attrition. Comments and recollections from the participant herself helped to shed some light on circumstances and reasons for change. They are integrated into the analysis of each speech act in Section 4 and examined in further detail in Section 5.

4. Results

4.1. Background, Experience and Conceptualization of Pragmatics

The findings in Section 4.1.1 and Section 4.1.2 relate to Research Question 1, dealing with Dora’s experience and perception of pragmatics.

4.1.1. Dora’s Story

In the first interview, she discussed her background and motivation for undertaking formal language study. Her grandparents had migrated to Australia in the 1950s. Both her parents had shared the common second-generation experience of rejecting the Greek language in their attempts to fit into mainstream English-speaking society, and realising its value later in life. They still preferred to use English in the home, so Dora’s main opportunity to use the Greek language had been with her grandparents and older aunts and uncles, with whom she had frequent contact. Her relationship with her siblings was conducted in English. She had taken part in heritage education opportunities from kindergarten to the end of high school. From her early years, she recalled using textbooks from Greece, learning and reading stories. She appreciated this experience mostly for the opportunities it gave her to be with people from her own cultural background. Her first visit to Greece occurred when she was 10 years old, and she valued having the chance to see the place where her grandparents had lived and to meet her cousins. It was her first real experience using Greek without having English to fall back on. She went again at 21 and stayed for the whole summer.

Our first interview took place two years after that. She described herself as fluent in Greek and able to communicate without difficulties. She was proud when people said they could not believe she was from Australia, because it validated the effort she had made to improve her ability to use the language. She considered herself to be fully bilingual in spoken Greek, but she was also taking Modern Greek as part of her university study, to refine her knowledge of grammar and improve her other skills. No formal language assessment was performed, but Dora’s background experience would be likely to place her among the advanced and high-contact learners in Bella’s studies. Her future career plan was to work in education, and she expected Greek to be a part of that career. At the time of writing, she had successfully achieved that goal and was employed as a teacher of Greek. When asked about motivation for undertaking SA, Dora described her main goal as to develop a more “authentic” Greek accent, but fluency and improvement in the accuracy of her grammar were also among her priorities.

4.1.2. Subjective Understanding of Pragmatics

• The first interview (pre-commencement)

Dora’s language study had not included any pragmatics, so she talked about it in folk terms. She was aware, for example, that Greek and English speakers had different communication styles. She described non-Greek Australians as “calm”, and more likely to say “please” and “thank you”, which was the main way she conceptualised politeness. She described Greeks as “more direct”.

In Australia we use more the “please” and the “thank you”—“excuse me” and all that, whereas in Greece it’s a bit rough—rough and tumble… like “Oh move out of the way” sort of thing in Greek … a Greek would say, “Go get me that!” …

…When I came last year that was one of the things that shocked me the most. Because when l like I’d go to a café and order coffees I’d go like “oh thank you” or “excuse me can I please have this “, and I was like—“do I need to say please or thank you? I think it’s rude if you don’t” and they’re like “no they‘ll know you’re a foreigner if you do that…” and I’m like, “ok….”

She was able to articulate her observations of the high value placed on hospitality and generosity, and she understood the way closeness was expressed through use of the singular second-person pronoun “εσύ” (single or informal “you”).

She was strongly attracted to Greek language and culture, commenting on the depth of meaning it was possible to convey.

Greek is a much deeper and richer language, and you get more meaning out of it … like if you say the word “φίλος” (friend), it’s much deeper than “friend” in English.

The excerpt above represents both an ideological position and an affective involvement with the Greek language which is highly motivational and often expressed by Greek learners and educators in the Australian context. Responses to International Greek Language Day are a case in point, where the language’s extensive vocabulary, rich and lengthy history, complex and nuanced grammar, and impact on English and other languages are celebrated in diaspora communities. This forms part of the context in which the attitudes and identities of heritage speakers and learners are formed.

• The second interview (3 weeks into the programme)

The second interview took place at the halfway point of the programme. Dora was living alone in a flat in Athens but had visited her Greek family in the south of the country and had been interacting with fellow students and some members of the community. She was very happy and satisfied with the experience. Originally, the SA students were intended to have an individual tailor-made course, but because of low numbers they were integrated into the existing year-long program. The curriculum was topic-based, focussed on vocabulary extension and fluency but with no explicit teaching of pragmatics. Dora was very positive about the course content and how it was helping to improve her proficiency.

Apart from her family, most of the interactions she had with local people were in transactional contexts. She mentioned taxi drivers and shopkeepers. She also communicated in Greek with other members of her class and found it particularly valuable to interact with non-English speaking students.

She related enthusiastically to the positive politeness society, and the sense of caring and involvement that she experienced. People she met “opened up straight away”. Taxi drivers shared their personal stories and gave her advice on her education and future. She found that young people had a deeper knowledge of politics and the economy than their Australian counterparts, and she was impressed by how mature they were. She had difficulty following some of the content of these conversations and understanding slang, but she was unable to recall any instances of pragmatic failure. She was sometimes identified as “foreign” but attributed this to her “accent”.

When I asked if she had changed anything about her communication style, she said:

I think I’ve adopted the Greek ways of asking for things and communication—when I’m speaking Greek it comes automatically to me…some of the things living here have forced me to change.

• The third interview (immediately after return to Australia)

I had asked her to prepare for this interview by thinking about the people she had interacted with in the Greek language while she was there. The response was only an approximation, but the purpose was to have an idea of opportunities she had had for noticing the gap between her pragmatic behaviour and that of NS Greeks.

This is her estimate.

• Non-native speakers of Greek (e.g., fellow students from other countries, migrants from other countries living in Greece), 80%.
• Greeks who (probably) do not speak English (like older people in villages, some taxi drivers, basic service industry workers), 10%.
• Greeks who have learnt English as a foreign language but have not spent time in an English-speaking country, 5%.
• Greeks who have lived outside Greece (e.g., migrants who have returned home from Australia or elsewhere, Greek Australians who are visiting on holiday, etc.), 5%.

One of the blocks to heritage language speakers and learners acquiring LNS-like pragmatic competence is lack of exposure to interlocutors who are not also English speakers. Even if they communicate in Greek, both participants in the conversation share a second set of pragmatic conventions and have English to fall back on if they need to resolve any misunderstandings. Monolingual Greek speakers are few in Australia and Dora’s opportunities to speak Greek there would have been almost exclusively with interlocutors who were also speakers of English.

After the programme, Dora showed a growing awareness of the specific differences in how Greek was used by speakers in Athens and in Sydney, and she felt that she was more able to switch between the variants, using language appropriate to the different sets of norms. For example, she mentioned the use of diminutives. In Greece people would address her in the familiar form of her name, (-ούλα suffix) even when they had not known her for a long time. This rarely happened in Australia. Secondly, there was the use of certain formulaic expressions, particularly those that were characteristic of informal contexts.

Describing how she used Greek in Australia, she said:

I feel like I speak it more formally. Because the Greeks in Australia have been brought up, I know because it was my experience … with the mentality of the 1950s and 1960s 60s, so a lot of the slang used in Greek now wouldn’t be appropriate. You know to speak with older people in society, with your parents, grandparents, that sort of thing, so that aspect has changed….

…so, things that might be considered rude now in Australia by Greek Australians would be considered everyday language, in Greece no-one would bat an eyelid….

…A lot of the Greeks today say things like “έλα ρε, τι μου κάνεις;” [a colloquial greeting formula using the untranslatable particle ρε] and that “ρε”… like if I said that in front of my grandparents or my parents, they’d be like you know—it’s so rude! But there it becomes everyday life—people of all ages are using it and you know, and males and females…

• Interviewer: You never used that here?

Among my friends I would but not with older people…

• I: In Greece did you hear people using it with older people?

Yes—they used it with everyone, and I was a bit taken aback I thought “oh my God how are you speaking like that?” And they were fine with it, you know!

She also referred to the freer use of insulting language and how it seemed to have less of a capacity to cause offence in Greece than in Australia. Overgeneralisation of these observations can be potentially hazardous, but either Dora had not had any experience of inadvertently causing offence or was unaware of it.

Greek is a T/V language in which the use of the second person plural to address an individual denotes respect or distance, while the singular indicates solidarity and informality. Dora also noted that she had become more able to use these forms appropriately. She said she would feel comfortable addressing her teacher in the singular form if invited to do so, but that had not happened. “In Australia”, she said, “we mainly use the ‘σου’ we don’t make that distinction much… unless it is someone, like much higher”. In Greece, she was aware that the requirements were different but, as the following extract shows, still felt unable to identify the optimal moment to switch between the forms.

• I: So do you think in Greece that you used “σας” more?

I definitely used “σας” more. It’s that common unspoken rule there: with people my age I would use “σου”—people who were older I would use “σας” unless they said, you know, “κόψε τον πληθυντικό” [lit: cut the plural]

• I: But did you ever switch over without them inviting you to?

Ummm… not really, no.

• I: Did you ever kind of wonder about whether you should switch over or did you not think…

Sometimes I would—like with the convenor, like I was using “σας” all the time and then it came to a point where we were very like comfortable with each other … she was acting like one of the kids! But because she had never actually stated for me to speak in the singular, I never did.

• I: So, did that feel alright doing it that way?

I was kind of like wondering, like you know, can I speak to you in singular, you know, but I wouldn’t ask, obviously, so, yeah….

• I: Did you notice any of the other students speaking in singular to the convenor?

Um, no.

She felt there were improvements in her ability to express her emotions and to convey closeness, especially because of the way it had been shown to her.

• I: And do you think you express closeness better now, better than you did before you were in Greece?

Definitely. I think because a lot of people showed me that closeness. To them, like, I showed that closeness to them as well and I think I got a better handle on how to express that as well.

After her return, Dora emphasised the importance of being accepted in Greece.

I think it was important because I feel so connected to Greece. I feel a really big emotional attachment to the place, so I think to have on the other hand the connection, to be accepted as well, it was a really big part, very important to me…

Our final meeting took place eight months after her return. I asked her to repeat the scenarios, asking whether she had any recollection of having had any comparable experiences in Greece or Australia. There was no reflective interview in the final phase, but incidental comments on her experience were recorded and formed part of the dataset.

4.2. Analysis of the Scenario Data

The following section deals with Research Questions 2 and 3. As mentioned in the discussion of methods, linguistic analysis was a somewhat complex procedure, as four separate baseline studies were involved, and evidence of learning could not be generalised across +/-P D and imposition contexts. Thus, it was not possible to establish one set of general criteria for pragmatic development based on the presence or absence of specific language features, even within a single speech act category. The prior findings (Bella, 2011, 2012a, 2012b, 2014a, 2014b) were examined to identify features that had been used to indicate a transition towards NS likeness in her GFL subjects, and Dora’s output before and after the SA was analysed for evidence of their presence.

Because heritage speakers are currently considered to be on a continuum of native-speakerness rather than qualitatively different, it is problematic to contrast them as if they are separate categories (Montrul, 2022). As discussed in Section 2.1, this problem is one of conceptualization rather than mere terminology as NS retains resonances with the structuralist construct of an “ideal speaker” and troubling connotations of superior legitimacy. The fact remains that if samples of language encoding different pragmalinguistic and sociopragmatic choices are being contrasted, the humans who produced them must be denoted in some way. Alternatives that have been suggested by a reviewer of an earlier version of this paper include monolingually vs. bilingually raised Greek speakers, prototypical, or baseline speakers. In order to avoid implicit judgements about the value of SMG as opposed to heritage variants, it was decided to adopt the term Local Native Speakers. (LNS) even though it was not used in Bella’s original studies. Where the term native-like is used, it should be understood that it refers to local, in this case young, educated Athenian, pragmatic norms. As we have seen, Dora is both a heritage speaker (bilingually raised but interrupted) and learner (exposed to some registers of her ancestral language only in educational institutions outside the home).

Because of the nature of the analysis, data are included in the text rather than as appendices.

4.2.1. Refusals

Using the taxonomy developed for the first of Bella’s refusal studies (detailed in Bella, 2011, p. 1738), the features characterising LNS behaviour were identified. Compared to GFL participants, the LNS were more likely than the learners or migrants to refuse the invitation directly and then provide an explanation. They used a lower number of mitigation strategies, particularly apologies, and were less likely to thank the inviter. They were more likely to use expressions of solidarity and affection, and their excuses were more elaborate and detailed.

Dora’s responses are shown below.

Refusal 1 Invitation to a Party (-P-D)

Table 1. Refusing an invitation.

REF1. A friend invites you to their birthday party, but you have to refuse.
RESPONSES	Before SA	After SA	8 months later
	Συγγνώμη αλλά δεν μπορώ να έρθω, γιατί έχω κάτι άλλο σήμερα. Sorry, but I can’t come because I have something else on today.	Συγγνώμη …θα ήθελα νά’ρθω πάρα πολύ, αλλά δυστυχώς δεν μπορώ επειδή πετάω αύριο. Sorry, I would like to come very much, but unfortunately I can’t because I am flying tomorrow.	Συγγνώμη αλλά δεν μπορώ να έρθω στο πάρτι σου, επειδή έχω κανονίσει κάτι άλλο με άλλους—αλλά θα’ρθω αύριο βράδυ να πιούμε έναν καφέ και να σου δώσω το δώρο σου. Sorry, but I can’t come to your party because I have organised something else with other people but I’ll come round tomorrow evening for a coffee and I’ll give you your gift.
POLITENESS STRATEGIES AND MITIGATION DEVICES
EXTERNAL MODIFICATION
Negative ability	δεν μπορώ να έρθω but I can’t come	δυστυχώς δεν μπορώ unfortunately, I can’t	δεν μπορώ να έρθω στο πάρτι σου I can’t come to your party
INTERNAL MODIFICATION
Excuse/explanation	γιατί έχω κάτι άλλο σήμερα. because I have something else on today	επειδή πετάω αύριο because I am flying tomorrow	επειδή έχω κανονίσει κάτι άλλο με άλλους because I have organised something else with other people
Apology	Συγγνώμη Sorry	Συγγνώμη Sorry	Συγγνώμη Sorry
Alternative			αλλά θα’ρθω αύριο βράδυ να πιούμε έναν καφέ και να σου δώσω το δώρο σου but I’ll come round tomorrow evening for a coffee and I’ll give you your gift.
ADJUNCTS
Positive opinion		θα ήθελα νά’ρθω πάρα πολύ I would like to come very much
Expression of regret		δυστυχώς unfortunately

Table 1. Refusing an invitation.

REF1. A friend invites you to their birthday party, but you have to refuse.
RESPONSES	Before SA	After SA	8 months later
	Συγγνώμη αλλά δεν μπορώ να έρθω, γιατί έχω κάτι άλλο σήμερα. Sorry, but I can’t come because I have something else on today.	Συγγνώμη …θα ήθελα νά’ρθω πάρα πολύ, αλλά δυστυχώς δεν μπορώ επειδή πετάω αύριο. Sorry, I would like to come very much, but unfortunately I can’t because I am flying tomorrow.	Συγγνώμη αλλά δεν μπορώ να έρθω στο πάρτι σου, επειδή έχω κανονίσει κάτι άλλο με άλλους—αλλά θα’ρθω αύριο βράδυ να πιούμε έναν καφέ και να σου δώσω το δώρο σου. Sorry, but I can’t come to your party because I have organised something else with other people but I’ll come round tomorrow evening for a coffee and I’ll give you your gift.
POLITENESS STRATEGIES AND MITIGATION DEVICES
EXTERNAL MODIFICATION
Negative ability	δεν μπορώ να έρθω but I can’t come	δυστυχώς δεν μπορώ unfortunately, I can’t	δεν μπορώ να έρθω στο πάρτι σου I can’t come to your party
INTERNAL MODIFICATION
Excuse/explanation	γιατί έχω κάτι άλλο σήμερα. because I have something else on today	επειδή πετάω αύριο because I am flying tomorrow	επειδή έχω κανονίσει κάτι άλλο με άλλους because I have organised something else with other people
Apology	Συγγνώμη Sorry	Συγγνώμη Sorry	Συγγνώμη Sorry
Alternative			αλλά θα’ρθω αύριο βράδυ να πιούμε έναν καφέ και να σου δώσω το δώρο σου but I’ll come round tomorrow evening for a coffee and I’ll give you your gift.
ADJUNCTS
Positive opinion		θα ήθελα νά’ρθω πάρα πολύ I would like to come very much
Expression of regret		δυστυχώς unfortunately

As shown in Table 1, all three responses included an apology, a statement of negative ability and an excuse. The post-course response added a statement of positive opinion, and the delayed response included a statement of alternative. The excuses became more elaborate in the delayed example. The addition of δυστυχώς (unfortunately) in the post-course data showed regret. Thus, using the baseline criteria there were two features that were consistent with increasing LNS likeness. The LNS informant believed that development had occurred, evaluating the pre-departure response as “correct but not friendly”, and the third, more detailed response as the best, having “a better feeling”. This suggested that provision of the alternative made a stronger positive impression on her than the more formulaic additions of positive opinion and regret.

Dora’s comments revealed that she had had a similar experience during the course, when she had been invited to a party by a fellow student and had to refuse because she was leaving the next day. Interestingly, she reported that the excuse she gave in her example, “επειδή πετάω αύριο” (because I am flying tomorrow), was exactly what she recalled having said on that occasion. In her delayed response she was no longer imagining that particular event, and she generalised the excuse to “επειδή έχω κανονίσει κάτι άλλο με άλλους” (because I have organised something else with other people) and added an alternative suggestion of the kind which was not an actual option in the real-world context in Greece.

Refusal 2 Refusing to Lend Money (-P-D)

This scenario appeared in the 2014 paper, which used a slightly different taxonomy of strategies adapted from Beebe (1990) and Barron (2003) (cited in Bella, 2014b), and provided in detail with examples (Bella, 2014b, p. 59).

Table 2. Refusing a request.

REF2. A friend asks you to lend them some money, but you have to say no.
RESPONSES	Before SA	After SA	8 months later
	Συγγνώμη αλλά δεν έχω πολλά λεφτά πάνω μου, οπότες δεν μπορώ να σου δώσω κάτι. Sorry, but I don’t have much money on me, so I can’t give you anything.	Συγγνώμη αλλά δεν συνηθίζω να δανείζω λεφτά στους φίλους μου. Sorry, but I’m not used to giving money to friends.	Συγγνώμη ρε φίλε αλλά δεν έχω πολλά χρήματα αυτή τη στιγμή, και δεν μπορώ να σου δανείσω. Sorry (untranslatable particle ‘re’) friend, but I don’t have much money at the moment, and I can’t lend you anything.
REFUSAL SEMANTIC FORMULAS
DIRECT REFUSALS
Negative ability	οπότες δεν μπορώ να σου δώσω κάτι so I can’t give you anything		δεν μπορώ να σου δανείσω I can’t lend you anything
INDIRECT REFUSALS
Reason/explanation	αλλά δεν έχω πολλά λεφτά πάνω μου I don’t have much money on me	αλλά δεν συνηθίζω να δανείζω λεφτά στους φίλους μου	αλλά δεν έχω πολλά χρήματα αυτή τη στιγμή I don’t have much money at the moment
Apology	Συγγνώμη Sorry	Συγγνώμη Sorry	Συγγνώμη Sorry
LEXICAL PHRASAL DOWNGRADERS
Understaters			αυτή τη στιγμή2 at the moment
Solidarity markers			ρε φίλε (untranslatable particle ‘re’) friend,

Table 2. Refusing a request.

REF2. A friend asks you to lend them some money, but you have to say no.
RESPONSES	Before SA	After SA	8 months later
	Συγγνώμη αλλά δεν έχω πολλά λεφτά πάνω μου, οπότες δεν μπορώ να σου δώσω κάτι. Sorry, but I don’t have much money on me, so I can’t give you anything.	Συγγνώμη αλλά δεν συνηθίζω να δανείζω λεφτά στους φίλους μου. Sorry, but I’m not used to giving money to friends.	Συγγνώμη ρε φίλε αλλά δεν έχω πολλά χρήματα αυτή τη στιγμή, και δεν μπορώ να σου δανείσω. Sorry (untranslatable particle ‘re’) friend, but I don’t have much money at the moment, and I can’t lend you anything.
REFUSAL SEMANTIC FORMULAS
DIRECT REFUSALS
Negative ability	οπότες δεν μπορώ να σου δώσω κάτι so I can’t give you anything		δεν μπορώ να σου δανείσω I can’t lend you anything
INDIRECT REFUSALS
Reason/explanation	αλλά δεν έχω πολλά λεφτά πάνω μου I don’t have much money on me	αλλά δεν συνηθίζω να δανείζω λεφτά στους φίλους μου	αλλά δεν έχω πολλά χρήματα αυτή τη στιγμή I don’t have much money at the moment
Apology	Συγγνώμη Sorry	Συγγνώμη Sorry	Συγγνώμη Sorry
LEXICAL PHRASAL DOWNGRADERS
Understaters			αυτή τη στιγμή2 at the moment
Solidarity markers			ρε φίλε (untranslatable particle ‘re’) friend,

Table 2 shows two indicators of developing LNS likeness relevant to this example, which were (i) an increase in the number of lexical–phrasal downgraders, particularly solidarity markers, and (ii) a decrease in the number of direct refusal strategies. The direct refusal by negative ability was omitted immediately after the course but returned in the delayed example, where, however, two new lexical–phrasal downgraders (consistent with a more LNS-like character) also appeared.

The LNS evaluation was interesting in this case, because it showed how her conceptualisation of appropriateness was broader and captured elements that a linguistic analysis designed for a quantitative study did not. She rated the second response as the best, although she said it was “not polite”, and the third response as second best. The reason the pre-departure response was given the lowest score for appropriateness was because it contained a word that she described as “village dialect” οπότες (so or therefore), the significance of which we will discuss later in the paper. The solidarity marker in the delayed response was judged to be too informal. A reviewer of an earlier version of this paper has noted that ρε φίλε is a gendered form most commonly used by men, and its use by a female speaker may be pragmatically marked.

Dora had not had a comparable experience while in Greece, but her reflection suggested that she felt the post-return response was better and indicated that she viewed the increasing directness as an indicator of closeness. When asked if the general statement about not lending money to friends was something she would feel comfortable saying, she replied:

yeah, with some of the people I met there, because there is this real sense of closeness, you know as though I’d known them for years sort of thing… I wouldn’t have an issue saying that.

She felt that the SA experience had made her “more direct”, and her way of interacting had been shaped by the behaviour of the people she met.

4.2.2. Requests

In Bella’s (2012a) study, she compared the DCT data of 50 LNS and three groups of learners of increasing proficiency using three scenarios: the kitchen, the forgotten wallet, and the assignment extension. She analysed the effects of length of residence and intensity of contact using categories of external modification (p. 9), internal syntactic modification (p. 15), and lexical–phrasal modification (p. 19). The 2012b study uses a streamlined version of this taxonomy, detailed in Bella (2012b, p. 1921).

As the analytical framework for requests proposed by Blum Kulka and Olshtain (1984, cited in Bella, 2012b) was not fully applicable to Greek, some modifications were made, with the addition of query preparatory with present indicative (e.g., Πληρώνεις εσύ (lit), Do you pay?) and suggestory formula (e.g., Δεν καθαρίζεις λιγάκι… (lit), Don’t you clean a little bit?) to Blum Kulka’s original list of conventionally indirect head acts. Later studies (Bella & Ogiermann, 2019) have broadened the focus of request studies to include speaker/hearer orientation.

In Bella’s study, some characteristics of LNS-like behaviour were consistent across the three scenarios. For example, while learners were more prone to use direct strategies, gradually reducing them across the proficiency levels, and indirect strategies or hints were found only in the data from low-level learners, the LNS in Bella’s study preferred conventionally indirect strategies in all three situations. In the low power distance situations, REQ1 and 2, the NS made frequent use of the query preparatory with present indicative (money and kitchen) and the suggestory formula (kitchen only). None of the three learner groups used these. Some of the other features of LNS-like behaviour were specific to the individual scenarios and will be discussed with the analyses that follow.

Request 1 The Kitchen (-P-D)

Table 3. Requesting cleaning of the kitchen.

REQ1. You want your flatmate to clean up the kitchen because you have friends coming over.
RESPONSES	Before SA	After SA	8 months later
	Συγγνώμη αλλά έχω φίλους που έρχονται σήμερα, μπορείς να καθαρίσεις λιγάκι εδώ μέσα; Sorry, but I have friends who are coming today, can you clean up a little bit in here?	Ερχόνται κάτι φίλοι μου… μήπως μπορείς να καθαρίσεις την κουζίνα; Some friends of mine are coming… maybe you can clean the kitchen?	Συγγνώμη Τασο, αλλά έχω κατι φίλες που Θα΄ρθούν απόψε—γίνεται να καθαρίσεις την κουζίνα λιγάκι;   Sorry Tasso, but I have some friends who will come this evening, is it possible to clean the kitchen a little bit?
HEAD ACT
Conventionally indirect (CI)	Query prep ability (S0) μπορείς να καθαρίσεις λιγάκι εδώ μέσα; Can you clean up a little bit in here?	Query prep ability (SO) μήπως μπορείς να καθαρίσεις την κουζίνα; Maybe you can clean the kitchen?	Query prep ability (Impersonal) γίνεται να καθαρίσεις την κουζίνα λιγάκι; Is it possible to clean the kitchen a little bit?
MODIFICATION STRATEGIES
EXTERNAL MODIFICATION	Apology Συγγνώμη Sorry	-	Apology Συγγνώμη Sorry
	Grounder έχω φίλους που έρχονται σήμερα, I have friends who are coming today	Grounder 3Ερχόνται κάτι φίλοι μου… Some friends of mine are coming	Grounder αλλά έχω κατι φίλες που Θα΄ρθούν απόψε but I’ve got some friends who are coming this evening
INTERNAL MODIFICATION
Lexical/phrasal	understater λιγάκι a little bit		understater λιγάκι a little bit
		downtoner μήπως perhaps

Table 3. Requesting cleaning of the kitchen.

REQ1. You want your flatmate to clean up the kitchen because you have friends coming over.
RESPONSES	Before SA	After SA	8 months later
	Συγγνώμη αλλά έχω φίλους που έρχονται σήμερα, μπορείς να καθαρίσεις λιγάκι εδώ μέσα; Sorry, but I have friends who are coming today, can you clean up a little bit in here?	Ερχόνται κάτι φίλοι μου… μήπως μπορείς να καθαρίσεις την κουζίνα; Some friends of mine are coming… maybe you can clean the kitchen?	Συγγνώμη Τασο, αλλά έχω κατι φίλες που Θα΄ρθούν απόψε—γίνεται να καθαρίσεις την κουζίνα λιγάκι;   Sorry Tasso, but I have some friends who will come this evening, is it possible to clean the kitchen a little bit?
HEAD ACT
Conventionally indirect (CI)	Query prep ability (S0) μπορείς να καθαρίσεις λιγάκι εδώ μέσα; Can you clean up a little bit in here?	Query prep ability (SO) μήπως μπορείς να καθαρίσεις την κουζίνα; Maybe you can clean the kitchen?	Query prep ability (Impersonal) γίνεται να καθαρίσεις την κουζίνα λιγάκι; Is it possible to clean the kitchen a little bit?
MODIFICATION STRATEGIES
EXTERNAL MODIFICATION	Apology Συγγνώμη Sorry	-	Apology Συγγνώμη Sorry
	Grounder έχω φίλους που έρχονται σήμερα, I have friends who are coming today	Grounder 3Ερχόνται κάτι φίλοι μου… Some friends of mine are coming	Grounder αλλά έχω κατι φίλες που Θα΄ρθούν απόψε but I’ve got some friends who are coming this evening
INTERNAL MODIFICATION
Lexical/phrasal	understater λιγάκι a little bit		understater λιγάκι a little bit
		downtoner μήπως perhaps

Bella’s data shows that LNS prefer a conventionally indirect head act with query preparatory and this is less evident in learner data. As shown in Table 3, Dora already preferred a QP ability before departure, but in the delayed response there was a minor change, from the μπορείς να (can you?) hearer-oriented strategy to γίνεται να (is it possible?) which is impersonal, and thus potentially more protective of the hearer’s negative face. She did not, however, choose QP with the present indicative, which was preferred by the majority of LNS. The grounder modification strategy was constant throughout, but the apology was absent immediately after the course, returning after 8 months of re-immersion in the Australian context. One additional modifier was used in each example, varying from understater to downtowner, and back again in the delayed example. This also contained a form of address (name), but this would not be classified as a solidarity marker in the existing system, as there was no diminutive form. Applying the criteria for the absence and presence of elements, no development was evident. Nevertheless, the LNS informant rated both post-course responses higher than the pre-course one.

Dora did recall an experience similar to the kitchen scene, which occurred during her stay in Athens.

…. I had one of the other girls who was doing the course staying with me for a couple of nights, and she wasn’t the cleanest. So… I did have… it did come up!

Asked to compare her responses, Dora commented that her post-SA response matched her recollection of what she had said in the real-life situation, and was an improvement on the previous one, commenting

It’s more natural but to me it sounds … a little bit… ruder. (laughter)

• I: But you’re comfortable with saying it, right?

Yeah. If it’s someone who I’m close to, yeah I wouldn’t have any problem.

Asked if she could recall anything about her friend’s reaction, she said:

Yeah it was fine, yeah…she was fine with it, yeah….

However, as the friend was also Greek Australian, this feedback was of course less informative than if she had been a native speaker of a homeland variety.

Request 2 The Forgotten Wallet (-P-D)

Table 4. Requesting a friend to pay for lunch.

REQ2. You are having lunch with a friend and find you have forgotten your wallet, so you have to ask them to pay for you.
RESPONSES	Before SA	After SA	8 months later
	Χίλια συγγνώμη αλλά ξέχασα το πορτοφόλι μου στο σπίτι—μπορείς να πληρώσεις για μένα; A thousand apologies but I forgot my wallet at home, can you pay for me?	Χίλια συγγνώμη αλλά άφησα το πορτοφόλι μου στο σπίτι—μπορείς να μου κεράσεις το γεύμα αυτό; A thousand apologies but I left my wallet at home, can you “shout” me this lunch?	Συγγνώμη ρε φίλε, ξέχασα το πορτοφόλι μου—γίνεται να μου αγοράσεις το φαγητό σήμερα και μετά θα σε πληρώσω; Sorry (untranslatable particle ‘re’) friend, I forgot my wallet—is it possible for you to buy my food today and afterwards I will pay you?
HEAD ACT
Conventionally indirect (CI)	Query prep ability (HO) μπορείς να πληρώσεις για μένα; Can you pay for me?	Query prep ability (HO)μπορείς να μου κεράσεις το γεύμα αυτό; can you “shout” me this lunch?	Query prep ability (impersonal) γίνεται να μου αγοράσεις το φαγητό σήμερα– is it possible Is it possible for you to buy my food today?
MODIFICATION STRATEGIES
EXTERNAL MODIFICATION
	Apology Χίλια συγγνώμη A thousand apologies	Apology Χίλια συγγνώμη A thousand apologies	Apology Συγγνώμη Sorry
	Grounder ξέχασα το πορτοφόλι μου στο σπίτι I forgot my wallet at home	Grounder άφησα το πορτοφόλι μου στο σπίτι I left my wallet at home	Grounder ξέχασα το πορτοφόλι μου I forgot my wallet
			Imposition minimiser και μετά θα σε πληρώσω Afterwards I will pay you
INTERNAL MODIFICATION
			Solidarity marker ρε φίλε, (untranslatable particle ‘re’) Friend

Table 4. Requesting a friend to pay for lunch.

REQ2. You are having lunch with a friend and find you have forgotten your wallet, so you have to ask them to pay for you.
RESPONSES	Before SA	After SA	8 months later
	Χίλια συγγνώμη αλλά ξέχασα το πορτοφόλι μου στο σπίτι—μπορείς να πληρώσεις για μένα; A thousand apologies but I forgot my wallet at home, can you pay for me?	Χίλια συγγνώμη αλλά άφησα το πορτοφόλι μου στο σπίτι—μπορείς να μου κεράσεις το γεύμα αυτό; A thousand apologies but I left my wallet at home, can you “shout” me this lunch?	Συγγνώμη ρε φίλε, ξέχασα το πορτοφόλι μου—γίνεται να μου αγοράσεις το φαγητό σήμερα και μετά θα σε πληρώσω; Sorry (untranslatable particle ‘re’) friend, I forgot my wallet—is it possible for you to buy my food today and afterwards I will pay you?
HEAD ACT
Conventionally indirect (CI)	Query prep ability (HO) μπορείς να πληρώσεις για μένα; Can you pay for me?	Query prep ability (HO)μπορείς να μου κεράσεις το γεύμα αυτό; can you “shout” me this lunch?	Query prep ability (impersonal) γίνεται να μου αγοράσεις το φαγητό σήμερα– is it possible Is it possible for you to buy my food today?
MODIFICATION STRATEGIES
EXTERNAL MODIFICATION
	Apology Χίλια συγγνώμη A thousand apologies	Apology Χίλια συγγνώμη A thousand apologies	Apology Συγγνώμη Sorry
	Grounder ξέχασα το πορτοφόλι μου στο σπίτι I forgot my wallet at home	Grounder άφησα το πορτοφόλι μου στο σπίτι I left my wallet at home	Grounder ξέχασα το πορτοφόλι μου I forgot my wallet
			Imposition minimiser και μετά θα σε πληρώσω Afterwards I will pay you
INTERNAL MODIFICATION
			Solidarity marker ρε φίλε, (untranslatable particle ‘re’) Friend

According to the linguistic criteria, the head act and external modification strategy changes in this example do not point to increasing pragmatic competence during or after the SA. As demonstrated in Table 4, the CI query preparatory strategy was adopted from the outset, the overall number of modification strategies increased rather than reducing to LNS levels, and the addition of the imposition modifier in the delayed response made it less rather than more standard. The apology became less intense with the removal of χίλια (thousand), although the taxonomy did not have a category to capture this. Only the addition of the solidarity marker in the delayed example suggested a more LNS-like character, but as in REF2, the selected form, ρε φίλε (untranslatable informal marker + friend), was potentially problematic because it was a form more commonly used by males.

Dora had no experience of the forgotten wallet scenario, nor any situation where she had suddenly had to ask a friend for help, either in Greece or Australia. She could not recall anything related occurring in her teaching programme. However, when asked to compare her responses, she expressed a belief that she had made progress by achieving a lower level of formality. The LNS informant reaction did not support this view, giving a higher appropriateness rating to the response Dora had produced before the SA. She found the addition of a redundant object in the second example (να μου κεράσεις το γεύμα αυτό; to “shout” me for this lunch) made the request confusing, describing it (in direct contrast to Dora’s assessment) as “more polite but less natural”. Without identifying the gender issue, she found the “ρε φίλε” in the delayed response highly inappropriate. Without any experience of the scenario, it seems that Dora was applying a macro-strategy of reducing the focus on negative politeness and maximising solidarity, but in a way that needed more refinement in order to become effective.

The final request scenario to be discussed involved a different PD relationship, and thus there were some differences in the features characterising an LNS-like response.

Request 3 The Assignment Extension (+P+D)

The extension situation has been identified by Bella (2012b) as challenging for FL learners who come from more egalitarian cultural backgrounds and do not understand (or feel comfortable adapting to) the level of deference accorded to academic staff by Greek students.

Table 5. Asking a lecturer for an assignment.

REQ 3. You have to ask a university lecturer for an extension on an assignment.
RESPONSES	Before SA	After SA	8 months later
	Συγγνώμη αλλά δεν είχα αρκετό χρόνο να τελειώσω την εργασία, μπορώ να έχω λίγο χρόνο ακόμη; Sorry, but I did not have enough time to finish my assignment, can I have a little more time?	Συγγνώμη κυρία A., (μικρό όνομα), αλλά επειδή δεν είχα πάρα πολύ ώρα να τελειώσω την εργασία, μήπως μπορώ να έχω λιγάκι χρόνο ακόμα? Sorry, Mrs A. (given name) but because I did not have very much time to finish my assignment, can I maybe have a little (dim) more time?	Επειδή δεν είχα την ώρα να τελειώσω την εργασία μου, γίνεται να έχω μια προσθεσμία για να τελειώσω τη εργασία μου πιο αργα; Because I did not have time to finish my assignment, is it possible to have a due date to finish my assignment later?
HEAD ACT	Conventionally Indirect
	Query prep permission μπορώ να έχω λίγο χρόνο ακόμη; can I have a little more time?	Query prep permission μήπως μπορώ να έχω λιγάκι χρόνο ακόμα? can I maybe have a little (dim) more time?	Query prep permission γίνεται να έχω μια προθεσμία για να τελειώσω τη εργασία μου πιο αργα; is it possible to have a due date to finish my assignment later?
EXTERNAL MODIFICATION	Apology Συγγνώμη Sorry	Apology Συγγνώμη Sorry	-
	Grounder δεν είχα αρκετό χρόνο να τελειώσω την εργασία I did not have enough time to finish my assignment	Grounder επειδή δεν είχα πάρα πολύ ώρα να τελειώσω την εργασία because I did not have very much time to finish my assignment	Grounder Επειδή δεν είχα την ώρα να τελειώσω την εργασία μου because I did not have time to finish my assignment
INTERNAL MODIFICATION
Lexical	Understater λίγο a little	Understater (dim) λιγάκι a little (dim)	-
		Downtoner μήπως perhaps

Table 5. Asking a lecturer for an assignment.

REQ 3. You have to ask a university lecturer for an extension on an assignment.
RESPONSES	Before SA	After SA	8 months later
	Συγγνώμη αλλά δεν είχα αρκετό χρόνο να τελειώσω την εργασία, μπορώ να έχω λίγο χρόνο ακόμη; Sorry, but I did not have enough time to finish my assignment, can I have a little more time?	Συγγνώμη κυρία A., (μικρό όνομα), αλλά επειδή δεν είχα πάρα πολύ ώρα να τελειώσω την εργασία, μήπως μπορώ να έχω λιγάκι χρόνο ακόμα? Sorry, Mrs A. (given name) but because I did not have very much time to finish my assignment, can I maybe have a little (dim) more time?	Επειδή δεν είχα την ώρα να τελειώσω την εργασία μου, γίνεται να έχω μια προσθεσμία για να τελειώσω τη εργασία μου πιο αργα; Because I did not have time to finish my assignment, is it possible to have a due date to finish my assignment later?
HEAD ACT	Conventionally Indirect
	Query prep permission μπορώ να έχω λίγο χρόνο ακόμη; can I have a little more time?	Query prep permission μήπως μπορώ να έχω λιγάκι χρόνο ακόμα? can I maybe have a little (dim) more time?	Query prep permission γίνεται να έχω μια προθεσμία για να τελειώσω τη εργασία μου πιο αργα; is it possible to have a due date to finish my assignment later?
EXTERNAL MODIFICATION	Apology Συγγνώμη Sorry	Apology Συγγνώμη Sorry	-
	Grounder δεν είχα αρκετό χρόνο να τελειώσω την εργασία I did not have enough time to finish my assignment	Grounder επειδή δεν είχα πάρα πολύ ώρα να τελειώσω την εργασία because I did not have very much time to finish my assignment	Grounder Επειδή δεν είχα την ώρα να τελειώσω την εργασία μου because I did not have time to finish my assignment
INTERNAL MODIFICATION
Lexical	Understater λίγο a little	Understater (dim) λιγάκι a little (dim)	-
		Downtoner μήπως perhaps

Local native speakers preferred the permission strategy (QP P) in this scenario, as did advanced learners. Not surprisingly, (see Table 5), it was also Dora’s preferred option, which she used in the pre- and immediately post-course examples, before reverting to the impersonal γίνεται να in the delayed response. Becoming more like the LNS, her use of modifiers increased, with the addition of an understater, a downtoner and a term of address. Interestingly, after her return to Australia and re-immersion in a less formal academic environment, these features disappeared.

Despite the addition of mitigators in the second response, the LNS informant rated Dora’s pre-departure extension request as the best, and the other two as equally inappropriate. The third response was considered problematic because it lacked an apology and, interestingly, the elements that had been added to the second response were felt to detract from its appropriateness. The LNS informant felt that the downtoner λιγάκι (tiny little bit) was inappropriate to a formal academic context and the title + given name form of address, κυρία A. (Mrs A.), was so inappropriate as to be worse than nothing, suggesting the deferential κύριε/α καθηγητά (Mr/Madame Professor) as more appropriate. This was unquestionably part of Dora’s repertoire, as it was used in a later apology scenario, so her decision not to use it in this case was not due to ignorance of the form. If it were contextually inappropriate it would be more accurately classified as a sociopragmatic infelicity, possibly reflecting the less formal relationship she was accustomed to having with academics in Australia. This will be explored in more depth in the Section 5 of this paper along with the possible effect of lexical infelicities on the hearer. The LNS evaluation might also have been influenced by the choice of the Greek word for “time” in the responses. The first one, χρόνος, conveys the sense of extended time needed for an assignment extension, while ώρα used in the second and third responses typically suggests a short duration (minutes or hours), which may have made the request seem clumsy or strange.

While Dora reported that she had not had to ask for an assignment extension in Greek, there was another experience which she had evaluated as similar, when she had asked the convenor of the SA programme for permission to return independently from a scheduled excursion.

4.2.3. Apologies

Four scenarios are discussed in this section. In Bella’s comparison of LNS and GFL learners’ apologies, which will be used to probe for evidence of Dora’s development, the coding categories for strategies were based on Brown and Levinson’s politeness theory and the systems developed by Olshtain and Cohen (1983) and Blum-Kulka 1989 (as cited in Bella, p. 689, where details of the taxonomy are also set out). Adaptations made in naming and categorising strategies or sub-strategies were based on theoretical grounds rather than as a response to unique features of Greek. Lexical/phrasal upgraders/downgraders were not considered separately in this study, the author choosing instead to consider intensification as a strategy, as its importance for apologies is well-documented. A total of twelve strategies were considered. Unlike the refusal and request studies, explicit use of an Illocutionary Force Indicating Device (IFID) was considered as a strategy, rather than a head act subject to modification.

Apology 1 Late for Meeting with a Friend (-P-D)

In this scenario the LNS displayed a lower frequency of IFIDS (which learners tended to overuse). When present, LNS IFIDs were located at the end rather than the beginning of the apology. LNS used fewer intensifiers, and gave reasons related to their own personality rather than external factors. Their informal apologies were characterised by the “justify hearer” strategy, humour and appeasers.

Table 6. Apology for lateness: friend.

AP1. You are late for a meeting with a friend.
RESPONSES	Before SA	After SA	8 months later
	Συγγνώμη αλλά αργώ— θα είμαι εκεί σε 5 λεπτά, δεν θ’αργήσω. I’m sorry but I’m late. I’ll be there in 5 min I won’t be late.	Συγγνώμη που θ’αργήσω, αλλά θα είμαι εκεί σε λίγο. I’m sorry that I’m late, but I’ll be there soon.	Συγγνώμη που άργησα, ελπίζω να μην υπήρξε θέμα αλλά όταν φθάσω θα τα πούμε. I’m sorry I’m late, I hope there won’t be a problem about it, but when I get there we’ll talk.
IFID
Apologising	Συγγνώμη I’m sorry	Συγγνώμη I’m sorry	Συγγνώμη I’m sorry

Table 6. Apology for lateness: friend.

AP1. You are late for a meeting with a friend.
RESPONSES	Before SA	After SA	8 months later
	Συγγνώμη αλλά αργώ— θα είμαι εκεί σε 5 λεπτά, δεν θ’αργήσω. I’m sorry but I’m late. I’ll be there in 5 min I won’t be late.	Συγγνώμη που θ’αργήσω, αλλά θα είμαι εκεί σε λίγο. I’m sorry that I’m late, but I’ll be there soon.	Συγγνώμη που άργησα, ελπίζω να μην υπήρξε θέμα αλλά όταν φθάσω θα τα πούμε. I’m sorry I’m late, I hope there won’t be a problem about it, but when I get there we’ll talk.
IFID
Apologising	Συγγνώμη I’m sorry	Συγγνώμη I’m sorry	Συγγνώμη I’m sorry

The analysis of Dora’s responses in Table 6 revealed only the presence of the IFID, which did not change in its form or placement. None of the listed strategies were present but all three responses contained additional features that were not captured by the classification system, such as a statement of the offence and an undertaking of what was going to happen next. The delayed response included an expression of hope that the lateness had not caused a problem, a strategy which could be described as consideration for the hearer. The addition of the promise to talk about the situation when she arrived did not correspond to any of the listed strategies. Like the LNS, Dora did not use intensifiers, nor did she give any explanation for her lateness, and this did not change over the 8-month period.

The LNS informant was unimpressed with all of Dora’s responses, giving them an equal low rating and judging that they would cause confusion. This was clearly a valid comment regarding the pre-course response, when she apologised for lateness and then said she would not be late.

In the post-course interview, Dora said she had not experienced anything like this scenario while in Greece.

Apology 2 The Friend’s Broken DVD Player (-P-D)

The second scenario concerned damage to a friend’s property. In Bella’s study the British GFL learners had rated this as a much more significant offence than the Greek LNS, though it cannot be automatically assumed that Australian English speakers would share this view. Reflecting this, the LNS favoured similar strategies to the ones preferred in the lateness example. They used a wider range of strategies including explicit acknowledgment of responsibility and lack of intent, both also common in the learner data, humour and appeasers, which were rare.

As

Table 7. Apology for property damage: friend.

AP2. You have to tell a friend that you have broken their DVD player.
RESPONSES	Before SA	After SA	8 months later
	Κατερίνα μου, χίλια συγγνώμη… δεν υπάρχει δικαιoλογία αλλά έσπασε το DVD player. My Katerina, a thousand apologies. There is no excuse, but I broke your DVD player.	Θα με σκοτώσεις αλλά έχω κάτι κακό να σου πω…έπαιζα ένα DVD στον παίχτη και μου χάλασε. You will kill me, but I have something bad to tell you. I was playing a DVD in your player, and it broke.4	Συγγνώμη αλλά έβλεπα μια κασέτα, μου χάλασε το ‘DVD player…’ αλλά θα σου δώσω τα χρήματα ή θα πάμε μαζί να πάρουμε μια καινούρια. I’m sorry, but I was watching a DVD and the DVD player broke. But I will give you some money or we’ll go together to get a new one.
IFID
Apologising	Χίλια συγγνώμη A thousand apologies	-	Συγγνώμη Sorry
OTHER
Reasons/explanations5		έπαιζα ένα DVD στον παίχτη και μου χάλασε. I was playing a DVD in your player and it broke.	έβλεπα μια κασέτα, μου χάλασε το ‘DVD player I was watching a cassette and the DVD player broke.
Explicit acknowledgement of responsibility	δεν υπάρχει δικαιoλογία there’s no excuse
Justify hearer		Θα με σκοτώσεις You will kill me
Offer of repair			αλλά θα σου δώσω τα χρήματα ή θα πάμε μαζί να πάρουμε μια καινούρια. but I’ll give you the money, or we’ll go together to get a new one.

shows, Dora used more strategies than in the previous example, suggesting that she did indeed view the offence as more serious.

There were few indicators of change that could be captured by the coding system. The addition of an explanation was slightly more common in LNS data but nevertheless used by both groups, and the offer of repair occurred at a similar frequency. However, the intensifier (less commonly used by the LNS) occurred only in the pre-course example, and the inclusion of the strategy of justifying the hearer in the post-course example indicated adoption of a more LNS-like feature. This was not retained in the delayed response.

The LNS informant did not report any development, feeling the pre-course response was slightly more appropriate, possibly due to the presence of the solidarity marker. It is also possible that some inappropriate word choices affected the informant’s evaluation.

Dora had not had any comparable experience during the SA.

The final two examples repeat the offences in a +P+D context.

Apology 3 Late for an Appointment with a University Lecturer (+P+D)

Table 8. (a) Apology for lateness: university lecturer. (b) Additional features of the apology response.

AP3. You are late for an appointment with a university lecturer.
RESPONSES	Before SA	After SA	8 months later
	Συγγνώμη καθηγητή αλλά βρήκα κίνηση και δεν μπορούσα να έρθω πιο νωρίτερα αλλά είμαι στο δρόμο και έρχομαι Χίλια συγγνώμη. I’m sorry professor but I found traffic and I couldn’t come earlier but I’m on the road and I’m coming, a thousand apologies.	Χίλια συγγνώμη που έχω αργήσει αλλά υπάρχει μία καθυστέρηση, θα είμαι εκεί όσο πιο γρήγορα μπορώ A thousand apologies that I’m late but there was a delay. I’ll be there as soon as possible	Κύριε, με συγχωρείτε αλλά κάτι έπαθα και δεν μπορούσα να έρθω πιο νωρίς, και δυστυχώς άργησα, … ελπίζω να μην υπάρχει θέμα I am sorry but something happened and I couldn’t come earlier and unfortunately I’m late. I hope there won’t be a problem.
(a) Features included in existing taxonomy
IFID
Apologising	Συγγνώμη—I’m sorry (beginning) Χίλια συγγνώμη (at the end)	Χίλια συγγνώμη A thousand apologies (beginning only)	Με συγχωρείτε Forgive me
OTHER
Intensification	Χίλια συγγνώμη. A thousand apologies	Χίλια συγγνώμη A thousand apologies
Reasons/explanations for the offence	βρήκα κίνηση και δεν μπορούσα να έρθω πιο νωρίτερα I found traffic and I couldn’t come earlier	υπάρχει μία καθυστέρηση there was a delay	κάτι έπαθα και δεν μπορούσα να έρθω πιο νωρίς something happened and I couldn’t come earlier
(b) Features not included in existing taxonomy
	Before SA	After SA	8 months later
Consideration for the hearer.			ελπίζω να μην υπάρχει θέμα
Explanation of what will happen next.		θα είμαι εκεί όσο πιο γρήγορα μπορώ
Use of title (deference strategy)	Καθηγητή Professor		Κύριε, Sir
Use of modifier expressing regret.			δυστυχώς unfortunately

Table 8. (a) Apology for lateness: university lecturer. (b) Additional features of the apology response.

AP3. You are late for an appointment with a university lecturer.
RESPONSES	Before SA	After SA	8 months later
	Συγγνώμη καθηγητή αλλά βρήκα κίνηση και δεν μπορούσα να έρθω πιο νωρίτερα αλλά είμαι στο δρόμο και έρχομαι Χίλια συγγνώμη. I’m sorry professor but I found traffic and I couldn’t come earlier but I’m on the road and I’m coming, a thousand apologies.	Χίλια συγγνώμη που έχω αργήσει αλλά υπάρχει μία καθυστέρηση, θα είμαι εκεί όσο πιο γρήγορα μπορώ A thousand apologies that I’m late but there was a delay. I’ll be there as soon as possible	Κύριε, με συγχωρείτε αλλά κάτι έπαθα και δεν μπορούσα να έρθω πιο νωρίς, και δυστυχώς άργησα, … ελπίζω να μην υπάρχει θέμα I am sorry but something happened and I couldn’t come earlier and unfortunately I’m late. I hope there won’t be a problem.
(a) Features included in existing taxonomy
IFID
Apologising	Συγγνώμη—I’m sorry (beginning) Χίλια συγγνώμη (at the end)	Χίλια συγγνώμη A thousand apologies (beginning only)	Με συγχωρείτε Forgive me
OTHER
Intensification	Χίλια συγγνώμη. A thousand apologies	Χίλια συγγνώμη A thousand apologies
Reasons/explanations for the offence	βρήκα κίνηση και δεν μπορούσα να έρθω πιο νωρίτερα I found traffic and I couldn’t come earlier	υπάρχει μία καθυστέρηση there was a delay	κάτι έπαθα και δεν μπορούσα να έρθω πιο νωρίς something happened and I couldn’t come earlier
(b) Features not included in existing taxonomy
	Before SA	After SA	8 months later
Consideration for the hearer.			ελπίζω να μην υπάρχει θέμα
Explanation of what will happen next.		θα είμαι εκεί όσο πιο γρήγορα μπορώ
Use of title (deference strategy)	Καθηγητή Professor		Κύριε, Sir
Use of modifier expressing regret.			δυστυχώς unfortunately

In contrast with the apologies to friends, in this formal situation the use of IFIDs by LNS exceeded that of the learners. As we can see in Table 8, Dora’s initial response contained two IFIDs and an intensifier, but according to Bella’s criteria, the removal of these made her subsequent responses less LNS-like. Her excuse (traffic) did not resemble the family and health excuses preferred by the LNS. Interestingly, the choice of the verb form of the IFID με συγχωρείτε (forgive me) over the simple “sorry” allowed respect to be conveyed through the formal plural verb form. Some other features that were absent from the existing taxonomy but which might have influenced the success of the response are shown in section (b) of Table 8.

The LNS informant gave the pre-departure response a low ranking, and despite noting the lack of a deferential address form in the second one, she judged it to be the best. The delayed response was rated lower because of the addition of ελπίζω να μην υπάρχει θέμα (I hope there won’t be a problem) which the LNS informant viewed as inappropriate and even insulting for an academic. Dora did not recall any similar experiences when in Greece.

Apology 9 Damage to the University Lecturer’s Book (+P+D)

This was probably Dora’s least successful example, with all responses rated by the LNS informant as very poor.

Increasing LNS likeness would have been indicated by a reduction in explicit acknowledgments of responsibility and an increase in the use of reasons/explanations, self-depreciation, and expressions of regret. There was a reduction in intensifier use in the IFID, but as Table 9 shows, none of the other criteria were present.

Table 9. Apology for property damage: university lecturer.

AP4.You borrow a book from a lecturer and spill coffee on it.
RESPONSES	Before SA	After SA	8 months later
	Κύριε καθηγητή χίλια συγγνώμη αλλά διάβαζα το βιβλίο και μου έχυσε το καφέ. Μπορώ να πληρώσω για ένα καινούριο; (Mr) Professor, a thousand apologies but I was reading the book and my coffee spilt. Can I pay for a new one?	Κύριε καθηγητά, …αλλά όταν διάβαζα το βιβλίο σου έχυσα λιγό καφέ. Aν υπάρχει πρόβλημα θα σ’ αγοράσω ένα καινούριο. (Mr) Professor, …but when I was reading your (sing) book I spilt a little coffee. If there is a problem I will buy you a new one.	Συγγνώμη κύριε καθηγητά αλλά αυτό το βιβλίο που μου δανείσατε, μου έχυσε λίγος καφές, και το χάλασα. Μπορώ να σου δώσω λίγα χρήματα για να πάρετε ένα καινούριο; Sorry (Mr) Professor but that book you lent me, a little bit of coffee spilt and I destroyed it. Could I give you a little bit of money to get a new one?
IFID
Apologising	Χίλια συγγνώμη a thousand apologies	-	Συγγνώμη Sorry
OTHER
Intensification:	Χίλια συγγνώμη a thousand apologies
Reasons/explanations for the offence	διάβαζα το βιβλίο και μου έχυσε το καφέ. I was reading the book and my coffee spilt.	όταν διάβαζα το βιβλίο σου έχυσα λιγό καφέ. when I was reading your book I spilt a little coffee.	αυτό το βιβλίο που μου δανείσατε, μου έχυσε λίγος καφές, και το χάλασα. that book you lent me, a little bit of coffee spilt and I destroyed it.
Offer of repair:	Μπορώ να πληρώσω για ένα καινούριο Can I pay for a new one?	Aν υπάρχει πρόβλημα θα σ’ αγοράσω ένα καινούριο If there is a problem I will buy you (sing) a new one.	Μπορώ να σου δώσω λίγα χρήματα για να πάρετε ένα καινούριο Could I give you a little bit of money to get a new one?

Table 9. Apology for property damage: university lecturer.

AP4.You borrow a book from a lecturer and spill coffee on it.
RESPONSES	Before SA	After SA	8 months later
	Κύριε καθηγητή χίλια συγγνώμη αλλά διάβαζα το βιβλίο και μου έχυσε το καφέ. Μπορώ να πληρώσω για ένα καινούριο; (Mr) Professor, a thousand apologies but I was reading the book and my coffee spilt. Can I pay for a new one?	Κύριε καθηγητά, …αλλά όταν διάβαζα το βιβλίο σου έχυσα λιγό καφέ. Aν υπάρχει πρόβλημα θα σ’ αγοράσω ένα καινούριο. (Mr) Professor, …but when I was reading your (sing) book I spilt a little coffee. If there is a problem I will buy you a new one.	Συγγνώμη κύριε καθηγητά αλλά αυτό το βιβλίο που μου δανείσατε, μου έχυσε λίγος καφές, και το χάλασα. Μπορώ να σου δώσω λίγα χρήματα για να πάρετε ένα καινούριο; Sorry (Mr) Professor but that book you lent me, a little bit of coffee spilt and I destroyed it. Could I give you a little bit of money to get a new one?
IFID
Apologising	Χίλια συγγνώμη a thousand apologies	-	Συγγνώμη Sorry
OTHER
Intensification:	Χίλια συγγνώμη a thousand apologies
Reasons/explanations for the offence	διάβαζα το βιβλίο και μου έχυσε το καφέ. I was reading the book and my coffee spilt.	όταν διάβαζα το βιβλίο σου έχυσα λιγό καφέ. when I was reading your book I spilt a little coffee.	αυτό το βιβλίο που μου δανείσατε, μου έχυσε λίγος καφές, και το χάλασα. that book you lent me, a little bit of coffee spilt and I destroyed it.
Offer of repair:	Μπορώ να πληρώσω για ένα καινούριο Can I pay for a new one?	Aν υπάρχει πρόβλημα θα σ’ αγοράσω ένα καινούριο If there is a problem I will buy you (sing) a new one.	Μπορώ να σου δώσω λίγα χρήματα για να πάρετε ένα καινούριο Could I give you a little bit of money to get a new one?

Again, one feature which may have influenced the perception of appropriacy that was not included in the coding system was the deference strategy of use of the title, “Κύριε καθηγητά”. Reflecting the greater seriousness of the offence in comparison with the lateness apology, all of Dora’s responses contained some form of this, although her pre-course choice was grammatically incorrect. The most problematic issue, however, was the use of the informal singular second person pronoun “σου” in the post-course and delayed responses. It was not possible to compare this with the pre-course response because no “you” form was used. Interview data clearly showed that Dora was well aware of the required plural form and in the delayed response she used the plural verb δανείσατε (you (formal) lent me), suggesting the second response was a slip or mistake rather than a case of sociopragmatic failure. In addition, the use of a minimiser in the offer of repair served to make the apology not only less effective but possibly offensive.

Dora had never experienced or witnessed an event like this during her stay abroad.

4.2.4. Comparison of Linguistic and Informant Evidence of Change and Retention

Table 10. Indicators of development, retention and reported past experience.

Scenario	Report of Similar Experience	Linguistic Evidence of Pragmatic Development	Retention (Based on Linguistic Evidence)	LNS Informant Judgement of Pragmatic Development	Retention (Based on LNS Evaluation)
REF1 birthday party	yes	yes	m	yes	d
REF2 lending money	no	yes	d	yes, post-SA only	m
REQ1 kitchen	yes	no		yes	m
REQ2 forgotten wallet	no	no		no
REQ3 extension	yes	yes, post-SA only	r	no
AP1 late (friend)	no	no		no
AP2 broken DVD player (friend)	no	yes	r	no
AP3 late (prof)	no	no		yes, post-SA only	r
AP4 damaged book (prof)	no	yes (very minor)	d	no

reports the data from Research Questions 4 and 5. It sums up the indicators of longitudinal pragmatic development and the responses of the NS informant, and records whether or not a similar experience was recalled. The column labelled “retention” reflects whether the development continued in the delayed response. The possible parameters here are: (a) attrition, in which the 8-month response was less LNS-like, (m) maintenance, in which it had the same level of LNS-likeness as the immediate post-course response, and (d) signifying continued development towards the LNS preferred form. These are presented in Table 10, and will be considered in the Section 5.

5. Discussion

This case study collected data from three sources: participant interviews, responses to scenarios, and a monolingually raised local native speaker informant’s evaluations. Longitudinal data were examined for patterns of change.

The interviews with Dora at various stages of the 8-month period revealed a deepening of her understanding of the differences between local Greek and diaspora pragmatic systems. In the pre-commencement interview, she was already aware of the different uses of politeness markers and degrees of directness in the two contexts. By the midpoint, she commented that she had adopted “more Greek ways of communicating”. She had also begun to notice the different topics that were discussed in Greek conversations, and to appreciate their greater depth and intimacy. By the time she returned home, she was able to articulate several ways in which her awareness of pragmatics and her competence in use had been enhanced. These concerned diminutives and certain formulaic expressions and her use of the second person singular and plural forms. She was also aware that she still faced limitations in achieving full local competence. All of this occurred without any explicit pragmatic instruction in the SA programme and was mostly attributable to experience. It should be noted, however, that participation in the research project could have played a role in intensifying the focus of her attention on pragmatics and inadvertently enhanced her learning.

The scenarios were examined in two different ways, firstly by comparison with existing data and secondly LNS evaluation, for evidence of change. Viewed in terms of pedagogy, especially in the GFL study context, change in the direction of a local norm would be described as “development” or “improvement”. For a heritage speaker, who can be characterised as a different type of native speaker (Montrul, 2015, 2022), it is also a positive development, but in a different way, showing an expansion of register flexibility. As such, reversion to diaspora forms on return does not necessarily signal permanent attrition.

According to the linguistic criteria discussed in Section 4.2, Dora’s responses became closer to local native speaker norms in the two refusal scenarios, the birthday party and lending money, the request for an extension on an essay, and the apology to the friend whose DVD player she had broken. The apology for the lecturer’s damaged book was borderline, with the only indicator being a reduction in intensity of the IFID, a very minor change.

The use of taxonomies from Bella’s earlier cross-sectional studies was identified at the outset as a potentially risky strategy. However, a conscious decision was made to build on a strong and thoroughly researched foundation, and it was undertaken in full knowledge that some issues of compatibility were likely to arise. Indeed, this was the case, not because of divergences in the categorisation systems (as they were only applied to the same scenarios for which they had been originally designed) but when there were elements in Dora’s responses that the systems did not capture. For example, in the existing systems the use of a name did not qualify as a solidarity marker unless it was accompanied by a possessive marker or contained a diminutive. In addition, there may have been a subtle difference between “I can’t come” and “I can’t come to your party”, in that the latter acknowledges and somehow validates the invitation. Three of the scenarios showed a movement from the hearer-oriented μπορείς (can you) QP ability strategy and speaker-oriented μπορώ να έχω (can I have) permission strategy to the more impersonal γίνεται να (is it possible), but the classification systems did not distinguish these.

The divergence between the findings of the linguistic analysis and the LNS informant was revealing. In fact, there were only two of the nine scenarios where they were the same (see Table 10). The original research design had anticipated that the informant data would supplement and enrich the linguistic data rather than contradict it, and the findings provided an opportunity to critically examine how pragmatic competence and appropriacy might be conceptualised differently in theory and responded to in actual communication. The sharpest contrast was that while the analytical approach counted any increase in the number of occurrences of an LNS-preferred strategy as a positive development, the LNS informant reacted to their actual content. Thus, in the lending money refusal and forgotten wallet request, the presence of a solidarity marker was evaluated in the first analysis as positive progress, but the LNS informant judged the inappropriately gender-marked “ρε φίλε” as considerably worse than nothing at all.

A single informant is obviously going to be constrained by her own beliefs about appropriateness and it cannot be expected that her responses represent anything beyond that. Investigating the diversity among NS would require a completely different study. Nevertheless, elements of personal preference were evident in her responses. For example, regarding the use of address terms for academic staff, the strong negative evaluation of the use of “κυρία A.” from the informant may not have aligned with the actual person’s real-world preferences. In fact, this became evident to me when the preliminary findings were presented at a conference. I was approached by a delegate who was acquainted with the real-life κυρία A. and described her as a friendly person who was comfortable with informal communication with her students.

As a reviewer of an earlier version of this paper has helpfully pointed out, the use of κυρία + first name is considered acceptable in university language classes, but not in other academic contexts, where κυρία/κύριε + surname would be used. In addition, Dora had undertaken her university education in the much less formal Australian context where academic’s first names are commonly used by students, and titles (Doctor or Professor) are reserved for contexts of high formality. It was natural that she would find it challenging to navigate these complexities. In fact, the informant’s preference for the highly formal and stylistically marked κύριε καθηγητά6 may reveal more about her own personal style than it does about broader usage, but this in no way diminishes the value of her input.

Although they fell outside the range of conventional pragmatic analysis, it was also of interest that the informant tended to give a lower rating to responses which contained subtle lexical infelicities, and lexical choices that were direct translations of English words (such as παίχτης for “DVD player”) or dialect forms. For example, she reacted negatively to the used of οπότες, a form of “so” or “therefore” that is characteristic of less educated speakers from Southern Greece, and which Dora may have heard and adopted in interactions with her Peloponnesian grandparents. Deliberate use of regional forms can be appropriate and bonding in diaspora contexts, and although variant choice tends not to be classified as a matter of pragmatics, it is nevertheless reflective of context awareness and sensitivity.

Discrepancies between the two data sources were also found in the relationship between the immediate post-course return and the 8-month delayed response, in which there were three possible states of continued development towards LNS norms: attrition, maintenance and development (see Table 10). In fact, there was not a single example in which the analytical approach to quantifying elements and the NS informant were in agreement about whether there had been progress after Dora’s return. As discussed above, this was because they attributed significance to different things.

In terms of the subjective response, Dora expressed a positive view of the development of her pragmatic competence. Directness can be a feature of positive politeness (Grainger & Mills, 2016), and Dora clearly perceived it as more LNS-like in Greek. Though she had a clear understanding of the implications of different power and distance contexts, she was not always able to apply this knowledge either in ways that suggested development on the basis of the linguistic or strategic criteria developed for earlier studies, or in ways which an age- and gender-matched LNS viewed in a positive light. While aware that Australian Greeks might be inappropriately indirect and overuse politeness markers, she still needed to tread carefully in determining the situations where the changes were appropriate.

Three of the nine scenarios could be linked to an experience Dora recalled having during her SA. In all of these cases, there was some evidence of change towards LNS norms, either from Bella’s criteria or the LNS informant evaluation. Four of the other six cases with no recalled experience also showed some indication of development, while two did not. Obviously no generalised claims can be made for such findings, and would not be appropriate in a single case study. However, Dora’s interviews did convey a range of ways in which her experience had shaped her responses, firstly in the party invitation, when she used the same excuse that she had used in Greece, and also in the kitchen scenario, where she commented that her request was the same as the one she had used in the real-life situation, and was more LNS-like than her pre-course response, albeit “ruder.” Clearly, having had the chance to experiment in a memorable context had had some lasting effects. This result is loosely consistent with prior research outcomes of the relationship between direct real-world experience and pragmatic development (Shively, 2011; Yang & Ke, 2021) and supports the value of SA in maximising opportunities for direct experience. It must be noted that Dora did not consciously recall any instances of receiving corrective feedback and if she had made any errors, she was not alerted to the fact. Native speakers tend to avoid pointing out pragmatic mistakes, preferring to tolerate non-native pragmatic forms rather than perform face-threatening corrections (Bryfonski & Sanz, 2018; Halenko, 2021), confirming the value of explicit instruction.

Dora came to the SA with a lay awareness that her pragmatic system contained hybrid elements. For example, the overuse of politeness markers is characteristic of transfer from English, and it was already embedded in the variant of Greek that Dora had been using in the Australian context. As she struggled to overcome this negative politeness orientation, she overshot acceptable norms of familiarity. She did not take on the expected level of deference for academic staff in a successful or lasting way, and tended to revert to more familiar forms when she was back home. This was to be expected as in the months after her return she was a student, interacting with Greek Australian academics with whom over-formality would not have been comfortable or welcome. Reversion to local norms may not have been permanent, and this is something that could be considered in future studies.

6. Conclusions

This study has contributed to the fields of Greek pragmatics and heritage language speaker/learner pragmatic development, and also provided a new angle on methodological possibilities, combining and synthesising established speech act approaches with the more impressionistic input of a non-specialist local native speaker. It has confirmed the value of experience in the expansion of an SA learner’s pragmatic repertoire and explored cultural reasons why some communication contexts are more challenging than others.

While the decision to focus on a single participant and a single informant allowed an intensive examination and comparison of data from multiple sources, the study has a number of limitations. It was conducted by a single researcher and although very helpful advice was provided by experts in the field, it was not possible to independently cross-check all coding decisions. In addition, it is well known that the use of DCTs as opposed to roleplays can have validity consequences (Bella, 2012b, 2014a, 2025). Some flaws were identified in the formulation of some of the scenarios, including inconsistencies regarding the provision of a grounder in the prompt, while small divergences from the wording applied in previous studies would have been better to avoid. Cultural differences in the academic context were not fully recognised in the study design, particularly the different degrees of interchangeability of the terms “lecturer” and “professor”, as well as the changes around expectations of formality in Greek academic contexts and the effect they may have on perceptions of appropriateness.

One additional question worthy of examination in future studies of interlanguage, and which this study was not able to address, is the role of phonology. Being able to produce a pragmatically appropriate string of words is not sufficient if the intonation patterns suggest a different intent and would also be a factor in how they were received. It is hoped that it may be possible to explore this in future studies.

In terms of pedagogical implications, the findings support the argument that it is more helpful for learners to see SMG and Australian Greek as subtly different varieties rather than one being “correct”, and to acknowledge that pragmatic competence is about matching a variant to geographical as well as sociocultural contextual parameters. While Dora felt a great sense of pride at being mistaken for a local person, Greek Australians are different from Greeks in the way they communicate and not everybody would necessarily want to become indistinguishable from local native speakers, even if they could. Education in SA contexts can play a role in this by acknowledging and validating local identities

The study has shed some light on the complex nature of pragmatic competence for the Greek heritage language speaker in a learning context, but there is more to be done. Consideration of heritage speakers as a special category of language learners is richly deserving of more research attention (Xiao-Desai & Wong, 2025). Many questions of how sociopragmatic variance influences the success of interpersonal communication, and how encountering and adapting to LNS norms might impact on HS and HL cultural identity, are waiting to be explored.