Theory and Practice of Second Language Acquisition vol. 9 (2), 2023, pp. 1/28 https://doi.org/10.31261/TAPSLA.12609 Sylwia Scheuer https://orcid.org/0000-0001-8772-1051 Sorbonne Nouvelle-Paris 3 University, Paris, France Céline Horgues https://orcid.org/0000-0001-6957-7332 Sorbonne Nouvelle-Paris 3 University, Paris, France “Three months on, I still sound like an Anglophone”: Tales of Success and Failure told by English and French Tandem Partners A b s t r a c t This paper reports on two ways in which success and failure can be operationalized and quantified in a non-institutional L2 learning context such as language tandem setting. We draw on the SITAF database, where we gathered 25 hours of video-recorded conversations held by 21 pairs, each consisting of a native speaker of English and a native speaker of French. The tandems performed collaborative tasks in both languages, thus giving each participant ample opportunity to be both the ‘expert’ and the ‘novice’ (learner) part of the dialogue. The tandem partners met regularly and autonomously outside of the recording sessions, and mak- ing progress in their L2 was one of their declared goals. Two possible measures of success in achieving this goal are: (1) the quality and quantity of learner uptake which followed the expert’s corrective feedback (CF) during the recorded conversations. Significant differences between the two L1 groups were observed: while 52% of the CF given by the native French speakers met with total uptake, over 52% of the English CF generated no uptake at all; (2) the participants’ own narratives of progress, as both the experts and the learners, obtained through questionnaires they filled out at the end of the program. Our study aims to contrib- ute to the discussion on the stakes of successful L2 informal learning (with a focus on the acquisition of L2 pronunciation) by adopting a perspective which combines learners’ spoken output data and learners’ perceptions of their own language learning activity. Keywords: corrective feedback, tandem learning, uptake, phonetic development https://creativecommons.org/licenses/by-sa/4.0/deed https://doi.org/10.31261/TAPSLA.12609 https://orcid.org/0000-0003-2451-7242 https://orcid.org/0000-0001-6957-7332 https://orcid.org/0000-0003-2451-7242 TAPSLA.12609 p. 2/28 Sylwia Scheuer, Céline Horgues Tandem Learning is a type of non-formal1 learning environment which is often used for its potential to promote L2 linguistic and cultural acquisition. More specifically, in face-to-face tandem, L2 (foreign or second language) learners from two different L1s (first languages) collaborate mainly through spoken interactions in their two languages in the hope of developing their L2 language and (inter)cultural skills (Brammerts & Calvert, 2003). The benefits of tandem learning identified by previous research include improving communi- cative competence, developing intercultural competence, increasing motivation and developing confidence in speaking a target language (see Wakisaka, 2018, for a recent summary). Tandem learning allows for reciprocity and overall symmetry in the relation between the native speaking (NS) and the non-native speaking (NNS) partner, as both participants help each other out at different points in the tandem exchange. Indeed, both participants forming a tandem pair take on two complementary roles: the position of L2 learner or novice when speaking in their L2, but also the role of the relative expert when the conversation switches to their L1. In the past few decades, various problems associated with the notion of native speaker have been raised by numerous scholars (e.g., Jenkins, 2000; Escudero & Sharwood Smith, 2001; Dewaele, 2018; Slavkov et al., 2022), with Paikeday (1985) describing it as “a convenient fiction.” In this paper, we make use of the simple labels NS and NNS precisely for the sake of convenience, since they well reflect the roles performed by the two partners, and they are the terms associated with the historical tenets of tandem partnerships (Brammerts & Calvert, 2003; O’Rourke 2005). Moreover, these are the statuses that each participant identified with when entering the tandem program (see the Methods section), declaring themselves as a NS of English and a NNS of French, or vice versa. However, we do not subscribe to the idealized views that are sometimes attached to these concepts. Importantly, we do not take native speakers to be the sole proprietors of L1 language ex- pertise or competence; we consider them to be relative expert users of their first language (and relative experts on some aspects of their L1 cultures), and we do not assume they have a full grasp of their L1. Finally, we do not equate native speakership with monolingualism, as most tandem participants in our study are indeed multilingual users. The development of language skills in the L2 is an explicit motivation for teachers to organize such tandem programs and for their L2 learners to take part in them. For example, O’Rourke (2005, p. 434) presents tandem learn- ing as “an arrangement in which two native speakers of different languages communicate regularly with one another, each with the purpose of learning 1 Schugurensky (2000) and Cedefop (2014) propose non-formal learning as a half-way house between formal learning (higher degree of explicitness in learning/teaching and in institutional integration) and informal learning (mostly implicit, disconnected from educational institution and generally unintentional from the learner’s perspective). “Three months on, I still sound like an Anglophone”… TAPSLA.12609 p. 3/28 the other’s language.” Yet, while improving one’s L2 language skills is a set objective for participating in a tandem program (and often an explicit expec- tation expressed by learners), it is not always easy to attest that it is indeed the tangible outcome of tandem learning practice. If tandem learning favors actual L2 development, it would be interesting to find indication of this de- velopment in tandem speech data. Our research analysis thus revolves around the possible evidence and measures of L2 acquisition success in the course of face-to-face tandem interactions. To that end, we will focus on the fruitful cooperation between the L2 learner and their NS interlocutor in error treatment sequences such as: NS’s Corrective Feedback > NNS’s uptake. We will there- fore investigate interactional sequences involving learner uptake following the NS’s corrective feedback, showing how the two participants’ roles go hand in hand, and putting the learner’s (re)active role in the limelight. We also want to connect this to the metacognitive dimension of representations of success or failure expressed through the learners’ introspective and declarative metadata. Since we have two different language-culture profile participant groups (the Francophones and the Anglophones), it will also be interesting to investigate if the two groups show similar uptake patterns, both in their actual productions and their representations. The research questions our study aims to address are: – Do interaction speech data and learners’ introspective data align in the pic- tures they give of L2 learning success during tandem exchanges? – Can differences be observed between the two language-culture groups in these two measures of learning success (uptake in speech and learner rep- resentations)? Literature Review Since we are interested in looking at the language features effectively learnt by NNS participants thanks to their NS partners’ input during their tandem interactions, it is necessary to first consider how previous researchers have operationalized the key concepts for our analysis such as corrective feedback (CF), learner uptake, and related terms such as modified output and repair, and assessment of successful L2 development. As a foreword, we deem it important to stress that most of these concepts were initially posited and described for the fairly “traditional” learning context of formal language instruction and not for the semi-naturalistic, non-formal setting of tandem exchanges between L2 learners, where a teacher is absent. Classroom interactions between the teacher and the L2 students have long framed the theoretical perspective and typol- TAPSLA.12609 p. 4/28 Sylwia Scheuer, Céline Horgues ogy for describing such phenomena (e.g., Lyster & Ranta, 1997; Gass, 2003; Mackey, 1999, 2006; El Tatawy, 2002; Sheen, 2006), although some more varied learning contexts are now being explored (Nassaji & Kartchava, 2021). Some methodological adaptations will necessarily have to be made when transferring previous categories and typologies (which, as Lyster and Ranta (1997) explain, were conceptualized for the needs of a particular database) into the specific learning environment of non-informal and semi-naturalistic NS-NNS tandem interactions (see our Methods section). Corrective Feedback (CF) Corrective feedback is synonymous with the term negative evidence, that is “the type of information that is provided to learners concerning the incor- rectness of an utterance” (Gass, 2003, p. 225). One of the pivotal pieces of research on corrective feedback is that proposed in Lyster and Ranta (1997), where the authors define seven categories of CF moves at the disposal of language teachers: (i) explicit correction, (ii) recast, (iii) clarification requests, (iv) metalinguistic feedback, (v) elicitation, (vi) error repetition, (vii) multiple feedback (combination of two or more of the above). Recasts can be defined as “the teacher’s reformulation of all or part of a student’s utterance, minus the error” (Lyster & Ranta, 1997, p. 46). They are reported by many studies to be the preferred CF strategies, although their corrective power or effectiveness is often contested (Lyster & Ranta, 1997; Lyster & Saito, 2010; Sato & Loewen, 2018; Saito, 2021). This is especially the case when compared with CF moves which are more explicit (e.g., explicit correction or metalinguistic comments), or through which the learner actively generates the target form instead of rely- ing on the teacher’s provision thereof (elicitation, error repetition, clarification requests, which have been described as output-prompting). Learner Uptake, Modified Output, and Repair Lyster and Ranta (1997) define uptake as “a student’s utterance that immedi- ately follows the teacher’s feedback and that constitutes a reaction in some way to the teacher’s intention to draw attention to some aspect of the student’s initial utterance (this overall intention is clear to the student although the teacher’s linguistic focus may not be” (Lyster & Ranta, 1997, p. 49). This phenomenon can also be referred to as “Other-Initiated self-repair” by other authors (e.g., Schegloff, Jefferson, & Sacks, 1977; as cited in Lyster & Ranta, 1997). However, it does not cover cases where learners spontaneously repair their own speech without being prompted by their interlocutor (self-initiated self-repairs, also “Three months on, I still sound like an Anglophone”… TAPSLA.12609 p. 5/28 called self-corrections). They seem to put absolute (albeit momentary or local) success at the center of their distinction between two key categories: repair (= learner’s correct reformulation of the initial error after CF) and needs-repair. Lyster and Ranta define repair as referring to “the correct reformulation of an error as uttered in a single student turn and not to the sequence of turns resulting in the correct reformulation; nor does it refer to self-initiated repair” (Lyster & Ranta, 1997, p. 49). Their other main category, needs repair (non- conformity with the target form), is rather large and encompasses many dif- ferent situations ranging from: simple acknowledgement, different error, same error, hesitation, off target, partial repair (see the reproduction of the error treatment sequence in Figure 1). In Lyster and Ranta’s model, however, it is unclear in what way learner reactions such as acknowledging, repeating the same error, hesitating, produc- ing an off-target response reveal that for the L2 learner “the teacher’s overall intention is clear” as these responses could also be found in normal, conver- sational/discursive moves (backchannelling for example). Figure 1 Lyster & Ranta’s (1997, p. 44) Error Treatment Sequence TAPSLA.12609 p. 6/28 Sylwia Scheuer, Céline Horgues Egi (2010) later emphasized that not all types of learner uptake responses are equally predictive of L2 development. Sharing the concern expressed by Ellis, Basturkmen, & Loewen (2001) for investigating what may constitute “successful uptake” or not, she therefore refines Lyster and Ranta’s typology by dividing their umbrella needs repair category into three subcategories: needs-repair modified, needs-repair unmodified, needs-repair acknowledg- ment. She also stresses the importance of regrouping uptake responses around the key distinction between modified output (which she defines, after Mackey 2007, as “generally entail[ing] the learner’s modification of a problematic form that invited feedback”; Egi, 2010, p. 2) and unmodified output (absence of modification of the problematic form by learner in the case of needs-repair unmodified, needs-repair acknowledgement). Her 4a/4b distinction level is reproduced in Figure 2. Figure 2 Egi’s (2010, p. 8) Coding System Assessing Success in L2 Learning through CF-Uptake Sequences Some studies have suggested a link between (certain types of) uptake and L2 learning (Lyster & Ranta, 1997; Egi, 2010). However, the analysis of uptake responses cannot necessarily be taken to be a direct measure of L2 development. The relation between learners’ uptake responses and their learning outcomes is fairly complex and indirect. Egi (2010, p. 4) summarises this methodological issue with the following disclaimer: “Three months on, I still sound like an Anglophone”… TAPSLA.12609 p. 7/28 The use of immediate uptake as a learning measure may be called into question in light of various discoursal constraints on its occurrence, a weak (or unclear) relationship between immediate uptake and SLA, delayed re- sponses, and indications that the effects of feedback may be delayed. Still, there are some theoretical arguments and empirical evidence that suggest a potential link between learner responses and L2 development. Quality of learner uptake is therefore important to take into account when determining its potential for conducing to learning, especially since the grada- tions in uptake quality are revealing of “varying degrees of cognitive effort involved in the production of uptake, and they may be differentially related to learning outcomes” (Egi, 2010, p. 5). In line with the Noticing Hypothesis (Schmidt, 1990; Mackey, 2006), the role played by learners’ “noticing of” or “paying attention to” their interlocu- tor’s CF and to the mismatch between their erroneous output and the target-like form is posited as an essential factor ensuring L2 acquisition. Egi (2010, p. 2) summarizes this as follows: “Learners’ responses often have been viewed as a cognitive window to their mental activities; a number of SLA researchers have interpreted learners’ responses as a signal that they have noticed feedback.” Researchers like Lyster and Ranta (1997), Mackey, Gass, & McDonough (2000), Egi (2010) have used the stimulated recall technique to investigate learner post hoc interpretation of CF and motivations behind their uptake re- sponses, for example to explore their ability to identify the corrective intent, identify the correct form and the mismatch with their erroneous output. For instance, Egi (2010)’s participants watched the video clips of their interactions with the L2 teacher and gave introspective interpretations for their responses to CF. This technique led these authors to reveal the relative ineffectiveness of recasts for prompting successful uptake, in particular because recasts are too implicit and learners therefore often miss their corrective purpose, as evidenced in their post-test declarations. Another traditional technique for gauging L2 development is to have learn- ers perform language tests before and after a controlled intervention or treat- ment carried out by a teacher, often in comparison to a control group. This technique is particularly adapted to laboratory experimentations or a classroom setting but it is less suited to semi-naturalistic settings such as NS-NNS in- formal conversations. TAPSLA.12609 p. 8/28 Sylwia Scheuer, Céline Horgues Methods The SITAF2 Corpus A detailed presentation of the SITAF project’s experimental design (tasks, participants, instructions, questionnaires) is offered in Horgues and Scheuer (2015). In total, our corpus consists of around 25 hours of video-recorded, face-to-face interactions held by 21 pairs of native-French speaking and native- English speaking tandem participants. Speakers The participants were all students enrolled at Sorbonne Nouvelle University (France), aged between 17 and 22. None were balanced English-French bi- linguals.3 The 21 native French-speaking students (coded F01 to F21) were undergraduate English language specialists for the most part. The 21 English- speaking students (coded A01 to A21) were international mobility students who came from various English-speaking countries (USA, Canada, UK, Ireland, and Australia). Each participant declared either English or French as their L1 in the registration questionnaire. For convenience’ sake, we refer to their partner’s L1 (French or English, as appropriate) as their L2, even though it could actually be their L3 or L4.4 Four Anglophones declared being early simultaneous bilinguals in a language other than French (one English/Spanish, one English/Lebanese, one English/Catalan, one English/Irish). Only one Francophone declared being an early bilingual (French/French creole). Most participants also reported hav- ing learned another additional language (e.g., Chinese, Spanish, Italian, German, Hebrew), so these tandem participants were highly multilingual users overall. Sessions and Tasks The speakers were recorded on two occasions: in February (Session 1) and again in May 2013 (Session 2). In between these two recording sessions, the tandem pairs met autonomously for weekly tandem conversations (12 meetings on average; not recorded). In the two recording sessions, they performed the same three tasks (in the two languages randomly sequenced). Two of these tasks were communication activities (semi-spontaneous speech): Game 1 (Liar-Liar, a sto- rytelling activity where the L2 learner tells a story containing three lies that the 2 Spécificités des Interactions verbales en Tandems Anglais-Français [Characteristics of English/ French spoken tandem interactions]. 3 None of them had a parent who was a native speaker of the L2, or had been to an L2 immer- sion school in their childhood. 4 That, however, was rare: the language was an L3 for two Anglophones and one Francophone, and an L4 for one Anglophone. “Three months on, I still sound like an Anglophone”… TAPSLA.12609 p. 9/28 NS interlocutor has to identify by asking questions) and Game 2 (Like Minds, a debating activity where both participants share their opinion on a potentially controversial subject before determining the degree of like-mindedness between them). The third activity was a reading task (text: The North Wind and the Sun read in L2 English and in L2 French). The reading by the NNS was explicitly monitored in Session 1 (i.e., the NS partner was invited to intervene to help the NNS improve their reading; this collaborative sequence was followed by a sec- ond reading of the passage by the NNS); and in Session 2 the NNS performed a simple (unmonitored) final reading of the same passage. Analysing Relative Uptake Success in the SITAF Corpus For the sake of this particular study, we analyzed CF-uptake sequences (uptake moves by the NNS learner following each corrective feedback instance performed by their NS interlocutor) in the semi-spontaneous tasks of the SITAF corpus (Game 1 and Game 2 in the two languages, in the two recording ses- sions). This represented about 15 hours of audio-visual speech analyzed for this study. We also analyzed CF in the reading task (two sessions; focus almost exclusively on pronunciation) for comparison. Our analysis of CF in this corpus (quantity, focus, strategies) was presented in previous publications (Horgues & Scheuer, 2014; Horgues & Scheuer, 2018; Scheuer & Horgues, 2020). In these analyses, we defined CF as the production of verbal negative evi- dence by NS participants in reaction to a non-target-like form (or absence of form) produced by their NNS partner, thus making it more target-like. We also draw on Lyster and Ranta (1997)’s typology for CF moves, which we adapted, as some strategies were never used by the NNS in our corpus (certainly be- cause they seem to be the prerogative of professional language teachers and would create a sense of hierarchy, something that tandem participants gener- ally want to avoid). For example, we discarded categories which are irrelevant to our specific interactional context as they are absent, that is, metalinguistic comments without provision of the correct form, elicitation, error repetition. We therefore simplified Lyster and Ranta’s categorization around four main CF types: (i) recast, (ii) clarification request, (iii) explicit correction and/or metalinguistic comment, (iv) mix (equivalent to what they termed “mix”). We also relied on Lyster and Ranta (1997)’s approach to uptake as the NNS’s verbal response following corrective feedback provided not by the teacher but by the NNS tandem partner. However, unlike these authors, we do not consider that any possible type of uptake response is informative for exploring L2 de- velopmental potential. We believe this is particularly important when looking at the acquisition of L2 phonetics and phonology, where practice at pronounc- ing the target sound form plays an essential role in learners’ internalization TAPSLA.12609 p. 10/28 Sylwia Scheuer, Céline Horgues of the L2 auditory and articulatory gestures. More precisely, we consider that categories such as “acknowledgement,” “repetition of same error,” “new error,” “hesitation,” and “off-target” are responses which do not attest clearly enough that the learner is reacting to their interlocutor’s corrective intent. We therefore labelled “repetition of the same error” or “new error” or “off target” as failed uptake (Egi’s unmodified needs-repair). In the absence of explicit verbalizing of a modified output attempt, we grouped cases of simple “acknowledgment,” “hesitation,” “no response” under the umbrella category no uptake5 since we consider there is no tangible (i.e., verbal) evidence that the learner has grasped the corrective function their interlocutor’s feedback had. Our perspective is therefore more in line with Egi’s (2010) key distinction between modified out- put and unmodified output. Indeed, we consider that the learner’s attempt at modifying their output signals that they have somewhat grasped the didactic function of the NS’s corrective feedback (as opposed to a simple conversational function) and that their attention has somewhat been drawn to the gap between their initial (erroneous) output and the target form. Thus, we will concentrate on uptake moves which clearly reveal some verbal attempt, by the NNS learner, at modifying their erroneous output, as signaled by corrective feedback from their NS interlocutor. Whether the corrected form is first produced by the NS (through input-providing CF such as recast or explicit correction, for example) or emerges from the NNS (student-generated repair for Lyster & Ranta, 1997) does not matter in our treatment of what counts as noteworthy uptake. Whether their attempt at modifying their output actually results in conforming with the target form completely or incompletely is not a determining factor in our general treatment of uptake cases either. In a bid to obtain a more accurate picture, however, we do distinguish between partial uptake (uptake responses which result in an imperfect correction towards the target or model form where only one feature is corrected but not others, equiva- lent to Lyster and Ranta’s partial repair, or Egi’s modified needs-repair) and total uptake (where the modified form represents a reasonably complete correc- tion in conformity with the target form; Lyster and Ranta’s and Egi’s repair). Notably, we discarded some of Lyster and Ranta’s categories ( peer-repair or topic continuation by peers) as they are irrelevant in the context of a NS-NNS dyadic interaction, which is very different from the classroom setting where the L2 learner is surrounded by fellow students. Here is the uptake typology we have used in our analysis of the NS-NNS informal interactions in the SITAF corpus: i) Total uptake: reasonable conformity to the target form given by the NS expert, for example: 5 The previously mentioned authors reserve the term no uptake for cases where there was no response, or there was a change of topic, on the part of the learner. They thus include simple acknowledgements—labelled needs repair—in the general uptake counts. “Three months on, I still sound like an Anglophone”… TAPSLA.12609 p. 11/28 (NNS) And I fall [talking about a past event] (NS) Oh, you fell! (NNS) Yeah, I fell. ii) Partial uptake: only part of the correction proposed by NS was implemented by the learner, for example: (NNS) … à Madrid, dans le centre du, de Espagne, de l’Espagne … [in Madrid, in the centre of the (M.SG), of Spain, of the (F/M.SG) Spain] (NS) au centre de l’Espagne [at the centre of the (F/M.SG) Spain] (NNS) oui au centre au centre d’Espagne [yes at the centre at the centre of Spain].6 iii) Failed uptake: NNS reacted to CF but failed to produce the model form, as in: (NNS) On meadow [*ˈmiːdəʊ], big meadow [*ˈmiːdəʊ], so when you’re… (NS) A meadow [ˈmedoʊ]? (NNS) Yeah, sort of meadow [*ˈmiːdəʊ], not mountain, but… iv) No uptake: no observable verbal reaction to the CF, as in: (NNS) J’ai dansé sur le table avec autres [I danced on the (M.SG) table with others]. (NS) Sur la table [On the (F.SG) table] (NNS) Oui, c’était, c’était fou, mais… [Yes, it was, it was crazy, but…]. The above examples also illustrate the variety of language areas that were targeted in the native speakers’ corrective interventions: morphosyntax (gram- mar), vocabulary, pronunciation, or any mix of the above. We are primarily interested in uptake emerging in the direct sequence fol- lowing feedback provision (subsequent speech turns). However, uptake success may also be recorded in terms of permanency and stability over time, which we could only analyze through the course of one conversation, or by comparing the two recording sessions, separated by the three months’ interval ((non-)perma- nency of uptake, delayed uptake effects). The analysis of the various instalments of the reading task in the corpus allows for the comparison of how, over time, the same (controlled) speech material is produced by the NNS following their NS partner’s CF provision (comparison between the first and second reading in Session 1, and the final reading in Session 2). However, our experimental protocol does not permit the application of this systematic longitudinal com- parison to the semi-spontaneous (uncontrolled) speech data (Games 1 and 2). Indeed, the re-occurrence of any problematic language issue at different points in time was only incidental and this un-systematicity prevents any quantitative analysis of the temporal aspect of uptake in this portion of the corpus. 6 The NNS reproduces the correct preposition (au, and not her initial incorrect one, dans) but omits the definite article l’ (which, in French, blurs the F-M gender distinction before a word starting with a vowel, such as Espagne). TAPSLA.12609 p. 12/28 Sylwia Scheuer, Céline Horgues Exploring Learners’ Tales of Success and Failure through their Written Questionnaires Upon finishing recording Session 2, all tandem participants completed two written questionnaires (see the English version in the Appendix). The general questionnaire aimed at eliciting introspective declarative data on the learners’ impressions about their tandem experience in general (benefits, corrective feedback practices and preferences, self-assessed progress). The second ques- tionnaire focused on their representations of phonetic issues related to their tandem practice (attitudes towards and communicative impact of foreign accent, phonetic development, etc.). In the general questionnaire, questions 15 and 16 (quoted in full in the Appendix) will be of particular interest when exploring the participants’ repre- sentations of success and failure in their L2 acquisition through tandem practice. They concern, respectively, self-reporting the perceived degree of improvement in different domains, and self-reporting the evolution in speaking confidence. In the Pronunciation and Tandem questionnaire, question 19 is particularly revealing as it provides information about the participants’ representation of success or failure in relation to their L2 pronunciation during tandem exchanges. It is worth noting that, contrary to other types of studies mentioned before (e.g., Lyster & Ranta, 1997; Egi, 2010; Mackey et al., 2000), we could not or- ganize stimulated recalls due to the unavailability of our participants (especially mobility students) after the end of the academic year. Results Quantifying Uptake Success We identified 492 corrective feedback instances in the portion of the SITAF corpus under investigation, that is, in both recording sessions, in both commu- nicative games, in both languages. However, there was a dramatic difference between the two language conditions: the English section7 accounted for only 156 (31.7%), and the French one for the remaining 336 (68.3%), tokens. This sharp contrast, highly significant ( p < .005), is visually presented in Figure 3. 7 That is, the conversations held in English, where the NSs were the Anglophones, and the NNSs were the Francophones. Analogously, the French section refers to the French conversa- tions, where the NS/NNS roles were reversed. “Three months on, I still sound like an Anglophone”… TAPSLA.12609 p. 13/28 Figure 3 Amount of CF in the English vs French Conversations The two language conditions also showed significant differences in terms of immediate uptake success. In both, the two intermediate categories presented in a previous section—partial and failed uptake—jointly accounted for just over 10% of all CF instances. The remaining 90% showed either total or no uptake: in other words, in the overwhelming majority of cases the NNS tandem partners either produced the (reasonably) correct L2 form, or did not detectably react to the CF at all. These two extremities of the spectrum, however, were distributed very differently in the English as opposed to the French conversations. While 52.6% of the corrective feedback found during the English conversations met with no uptake on the part of the French participants (NNSs) at all, total up- take occurred in just 36.5% of instances. In the French tasks the figures were almost identical, except in reverse order: it is total uptake that accompanied the NS’s corrective endeavors in 52.4% of cases, with 36.9% of CF tokens going unnoticed or—not necessarily deliberately—ignored (no uptake on the part of the NNSs, i.e., native English speakers). These divergent patterns are illustrated in Figure 4. Figure 4 Relative Share of Total vs. Partial + Failed vs. No Uptake in the Two Language Conditions (English Conversations and French Conversations) TAPSLA.12609 p. 14/28 Sylwia Scheuer, Céline Horgues Needless to say, the boundaries between the respective categories can be rather fluid. Since our use of the term uptake relates to any observable attempt by the NNS at modifying their original utterance in reaction to CF, the most salient line of demarcation on the uptake continuum is that between no uptake and the remaining three types. If viewed in this way and considered jointly for both language conditions, the results can be interpreted optimistically: uptake of some sort occurred in the case of as many as 58.1% of corrective instances, that is, more often than it did not. It has to be borne in mind, however, that the respective figures were of a significantly different order for the English and the French conversations. Success and Failure through Participants’ Tales: Questionnaire Answers As previously mentioned, the tandem participants’ answers to three ques- tionnaire items will be considered, in a bid to provide another measure of their perceived L2 learning success. Question 15 in the general questionnaire Working in tandem with your partner allowed you to improve your knowl- edge and skills in which of the following domains? The perceived improvement was shown on a 0–5 scale, separately for each of the following five domains: culture, pronunciation, grammar, vocabulary, general ease of expression. The results (group averages) are given in Table 1. Table 1 Perceived (Self-declared) Improvement Following the Tandem Program English NSs French NSs (a) Culture 3.62 3.29 (b) Pronunciation 3.36 3.05 (c) Grammar 2.90 2.57 (d) Vocabulary 3.95 3.67 (e) Ease of expression 4.24 3.95 Once again, the data obtained from the Anglophone and the Francophone participants show both striking similarities and noteworthy differences. The hierarchy formed by the five dimensions is identical for the two groups: (e) > (d) > (a) > (b) > (c). That means that both cohorts noted, on average, the biggest improvement in the area of ease of expression and the smallest in that “Three months on, I still sound like an Anglophone”… TAPSLA.12609 p. 15/28 of grammar. Out of the three specific language domains targeted by this ques- tion—pronunciation, grammar and vocabulary—it is vocabulary that emerged victorious. Interestingly, it was also by far the most frequently targeted area during the corrective episodes discussed earlier in the paper, accounting for 52.5% of all CF tokens in the English, and 49% in the French, part of the corpus (Scheuer & Horgues, 2020). However, there is a dissimilarity between the two language groups when it comes to the extent of their perceived improvement. On all five dimensions, the native English speakers tended to feel they had made better progress than their French counterparts, although the differences fail to reach the threshold of significance. Question 16 in the General Questionnaire Evaluate your confidence in speaking [L2] before beginning your tandem meetings and after, on a 0–10 scale, separately for each point in time. There was a very highly significant perceived improvement in the case of both language groups: on average, by 2.9 points for the Anglophones and by 2.24 points for the Francophones. Yet again, it is the former cohort that reported a greater benefit, even though the difference is, again, non-significant ( p < .057). Question 19 in the Pronunciation Questionnaire Speaking with your partner and receiving advice from them have helped you to improve your pronunciation. Out of the 5 options to be picked from, the moderately optimistic one—some aspects of my pronunciation—was the overall winner. It garnered 12 hits from the French NSs and 8 from the native English speakers. The latter score (8) was matched by that obtained by the most enthusiastic option (yes, most aspects of my pronunciation), which was chosen 5 times by the native French participants. Reassuringly, in neither language group was the answer definitely not circled at all. Discussion Two possible measures of success in tandem L2 learning are proposed in this paper: (1) the amount of learner uptake immediately following corrective feedback offered by the NS tandem partner, and (2) participants’ self-reports of success or failure, in the form of their answers to the questionnaires admin- istered at the end of the tandem program. Both sets of analyses have yielded results which point to the prevalence of success over failure, even though these conclusions need to be qualified in various ways. TAPSLA.12609 p. 16/28 Sylwia Scheuer, Céline Horgues The typology and definition we adopted for uptake—total, partial, failed or no uptake—shows a graduated success-failure continuum. Uptake of some degree accompanied a total of 58.1% of all CF tokens. However, there was a sharp difference between the relative uptake success shown by the L2 French and the L2 English learners: while 63.1% of CF met with some observable reaction on the part of the learner8 during the French conversations, this fig- ure drops to 47.4% during the English exchanges (Figure 4). This tendency for the native English speakers to attain greater uptake-related success in L2 than their French counterparts was matched by their more optimistic assess- ment of learning progress made during the program, as per the post-recording questionnaires. While both language-culture groups reported improvement in all the areas under investigation (vocabulary, grammar, etc.), this impression was stronger in the case of the Anglophones. It is not difficult to trace a con- nection between the two measures, which may explain why it was the English NSs who seemed to enjoy an advantage on both counts. They received signifi- cantly more corrective feedback than their French partners, which must have promoted the feeling that they actually learned something. If CF is frequent, its saliency may be enhanced and it is also probably easier to develop the habit of reacting to it. If, on the other hand, it occurs only rarely, its recipients may be more prone to miss its corrective function since they are not primed for it. This abundance of feedback in the French conversations did not, however, undermine the Anglophones’ self-assessed confidence in speaking their L2. A substantial enhancement of that confidence was reported by both language groups, although—yet again—the French NSs tended to be somewhat less en- thusiastic, with the difference between the two cohorts almost reaching a sta- tistically significant level. Interestingly, the confidence boost was not universal, since one of the native English participants actually reported regression in that domain, by 2 points on a 0–10 scale. It has to be borne in mind, however, that questionnaires of this type only tend to convey the participants’ subjective impressions and representations, which may not be reliable reflections on the actual learning progress and should therefore be treated with caution, especially if comparisons are made between groups characterized by divergent cultural and scholarly traditions. While the first of the proposed measures of success—uptake amount—may be considered less subjective of the two, it is naturally not without its prob- lems, either. One issue, already discussed in the section Assessing Success in L2 Learning through CF-Uptake Sequences, concerns the very validity of treating uptake as indicative of progress in the language acquisition process. Just because the learner correctly repeats the model form provided by the NS 8 It was the native English participants that were the learners (and therefore, the CF receivers) during the French conversations, and vice versa. “Three months on, I still sound like an Anglophone”… TAPSLA.12609 p. 17/28 does not automatically mean that the form has been, or will be, internalized, since “we cannot ignore the possibility that they parroted the recasts with no true understanding” (Egi, 2010, p. 16). Conversely, lack of any verbal reaction to the CF does not necessarily equal lack of noticing: the feedback may have a delayed effect on the shape of the learner’s interlanguage. To quote Egi (2010, p. 16) again, “[g]iven that responses to recasts are not discoursally required, learners may not always choose to respond to the recasts or may respond only subvocally.” Not being able to tap into the learners’ awareness and intentions through stimulated recalls is a methodological limitation of our study. This technique would have given us some supplementary introspective data about the participants’ post-recording interpretations of the functions of certain speech moves observed in the corpus. Also, the link between CF/uptake sequences and L2 development may be language area sensitive. While total immediate uptake of a syntactic or lexical structure may not deserve to be hailed as ac- tual progress, we think that the situation is rather different for pronunciation. Even “simple” parroting of the NS-generated form testifies to the NNS’s abil- ity to mobilize the corresponding articulatory gestures, giving them the motor practice facilitating further L2 learning. As noted by Saito (2021, p. 422), the benefits of pronunciation recasts can be further enhanced in a learning context involving “communicatively important and salient features,” which is largely the case with the SITAF conversation tasks. The other—related—issue regards uptake permanency. Even if uptake goes beyond the plain on-the-spot parroting (i.e., the learner repeats the model form meaningfully), its long-term effect cannot be taken for granted. It would be highly enlightening to be able to systematically verify how the learners repro- duced the previously “uptaken” forms later in the recording cycle. However, the fact that Games 1 and 2 featured semi-spontaneous speech—the participants were given specific conversation topics but those topics were different in the two recording sessions—makes such systematic comparisons impossible. Partial insight into this matter, meanwhile, can be gleaned from the reading task, where the same text was used in Sessions 1 and 2 (see section Sessions and tasks). Our analysis of the pronunciation uptake in the English task (Horgues & Scheuer, 2014) suggested, rather predictably, a quantitative deterioration over time. Calculated in relation to the CF provided during the first (monitored) read- ing, the combined amount of total and partial uptake dropped between Session 1 and 2. This decrease, however, was not statistically significant, and it still left the overall figure at 73.3%. In other words, nearly ¾ of the pronunciation errors corrected by the NS during the first reading were eliminated—partially or totally—by the NNS learner during the final reading, three months later. From the point of view of L2 learning success, this is a fairly encouraging result. TAPSLA.12609 p. 18/28 Sylwia Scheuer, Céline Horgues Conclusion The tales told both by the SITAF uptake results and by the participants themselves are certainly tales of success rather than failure, even if this suc- cess is naturally relative, subjective and far from universal. The two measures of success that we explored in this paper both interact with and complement one another: more corrective feedback tends to lead to—proportionally—more uptake, which tends to enhance self-confidence and a sense of achievement experienced by the learner. A boost in confidence in speaking L2 is a suc- cess in itself, even in the absence of immediately available tangible evidence of improved accuracy of L2 structures under scrutiny. Among other things, it is bound to lead to a reduction of foreign language anxiety, which in turn is beneficial for successful communication on many different levels. The analysis of our CF-uptake data could undoubtedly be refined in the future, for example by exploring possible interactions between various variables, such as CF type and uptake type, task type and uptake type, or effects of partner-generated repair vs. learner-generated repair (not discussed here) on the permanency of uptake. Another future perspective may concern a refinement of the description of the quality of uptake, through drawing a distinction between simple repetitions and incorporation repetitions of the model form. Contrary to what Lyster and Ranta (1997) seem to imply by conflating the two, we think that the latter type—where the learner incorporates the form into their own personal phrasing—is a more reliable sign of the L2 development progressing. Last but not least, we would like to investigate cases of no opportunity for uptake, where “the NS immediately continued on with the ongoing or another conversational topic without giving the learner a chance to respond to the recast” (Egi, 2010, p. 8). Following Egi, such episodes could be coded accordingly and therefore be removed from the no uptake category, thus providing a more ac- curate reflection on the learners’ reactions to corrective feedback. Language learning in tandem with a native partner is an excellent way of progressing in one’s L2 in a friendly, collaborative environment. In such a setting, focus-on-form episodes occur incidentally during communicative, focus-on-meaning activities. The SITAF tandem program bore fruit, both in terms of learner uptake success and—potentially even more importantly—self- reported progress and overall satisfaction. The latter was self-estimated to be at the level of 9.2/10 by the native English and the native French participants alike (Question 17 in the general questionnaire). It was therefore a highly posi- tive experience, despite the fact that, rather unsurprisingly, at the end of the three-month program some difficulties still remained, and Anglophones still sounded like Anglophones.9 9 This statement draws on the answer given by an English NS to the last question in the Tandem and Pronunciation Questionnaire. “Three months on, I still sound like an Anglophone”… TAPSLA.12609 p. 19/28 References Brammerts, H., & Calvert, M. (2003). Learning by communicating in tandem. In T. Lewis & L. Walker (Eds.), Autonomous language learning in tandem (pp. 45–59). Academy Electronic Publications. Cedefop (2014). Terminology of European education and training policy: A selection of 130 terms (2nd ed.). Publications Office of the European Union. https://www.cedefop.europa. eu/en/publications/4117 Dewaele, J.-M. (2018). Why the dichotomy ‘L1 versus LX user’ is better than ‘native versus non-native speaker’. Applied Linguistics, 39(2), 236–240. Egi, T. (2010). Uptake, modified output, and learner perceptions of recasts: Learner responses as language awareness. The Modern Language Journal, 94(1), 1–21. Ellis, R., Basturkmen, H., & Loewen, S. (2001). Learner uptake in communicative ESL lessons. Language Learning, 51, 281–318. El Tatawy, M. (2002). Corrective feedback in second language acquisition. Working Papers in TESOL & Applied Linguistics, 2, 1–19. Escudero, P., & Sharwood Smith, M. (2001). Reinventing the native speaker: or ‘What you never wanted to know about the native speaker so never dared to ask’. EUROSLA Yearbook, 1(1), 275–286. Gass, S. M. (2003). Input and interaction. In Doughty, C. J., & Long, M. H. (Eds.), The hand- book of second language acquisition (pp. 224–255). Blackwell. Horgues, C., & Scheuer, S. (2014). “I understood you, but there was this pronunciation thing…”: L2 pronunciation feedback in English/French tandem interactions. Research in Language, 12(2), 145–161. Horgues, C., & Scheuer, S. (2015). Why some things are better done in tandem? In J. A. Mompeán & J. Fouz-González (Eds.), Investigating English pronunciation: Trends and directions (pp. 47–82). Palgrave Macmillan. Horgues, C., & Scheuer, S. (2018). L2 pronunciation feedback in English-French tandem con- versations. In J. Volín & R. Skarnitzl (Eds.), Pronunciation of English by speakers from other languages (pp. 260–281). Cambridge Scholars Publishing. Jenkins, J. (2000). The phonology of English as an international language. Oxford University Press. Lyster, R., & Ranta, L. (1997). Corrective feedback and learner uptake. Studies in Second Language Acquisition, 19(1), 37–66. Lyster, R., & Saito, K. (2010). Oral feedback in classroom SLA. Studies in Second Language Acquisition, 32 (Special Issue 2), 265–302. Mackey, A. (1999). Input, interaction and second language development. Studies in Second Language Acquisition, 21, 557–587. Mackey, A. (2006). Feedback, noticing and instructed second language learning. Applied Linguistics, 27(3), 405–430. Mackey, A. (2007). Interaction as practice. In R. DeKeyser (Ed.), Practice in a second language: Perspectives from applied linguistics and cognitive psychology (pp. 85–110). Cambridge University Press. Mackey, A., Gass. S. M., & McDonough, K. (2000). How do learners perceive interactional feedback? Studies in Second Language Acquisition, 22(4), 471–497. Nassaji, H., & Kartchava, E. (Eds.). (2021). The Cambridge handbook of corrective feedback in second language learning and teaching. Cambridge University Press. https://www.cedefop.europa.eu/en/publications/4117 https://www.cedefop.europa.eu/en/publications/4117 TAPSLA.12609 p. 20/28 Sylwia Scheuer, Céline Horgues O’Rourke, B. (2005). Form-focused interaction in online tandem learning. CALICO Journal, 22(3), 433–466. Paikeday, T. (1985). The native speaker is dead! Paikeday Publishing. Saito, K. (2021). Effects of corrective feedback on second language pronunciation development. In H. Nassaji & E. Kartchava (Eds.), The Cambridge handbook of corrective feedback in second language learning and teaching (pp. 407–428). Cambridge University Press. Sato, M., & Loewen, S. (2018). Metacognitive instruction enhances the effectiveness of correc- tive feedback: Variable effects of feedback types and linguistic targets. Language Learning, 68(2), 507–545. Schegloff, E. A., Jefferson, G., & Sacks, H. (1977). The preference for self-correction in the organization of repair in conversation. Language, 53, 361–382. Scheuer, S., & Horgues, C. (2020). Corrective feedback in English/French spoken tandem interactions. In C. Tardieu & C. Horgues (Eds.), Redefining tandem language and culture learning in higher education (pp. 147–160). Routledge. Schmidt, R. (1990). The role of consciousness in second language learning. Applied Linguistics, 11(2), 129–158. Schugurensky, D. (2000). The forms of informal learning: Towards a conceptualization of the field. NALL [Research Network on New Approaches to Lifelong Learning] Working Paper, 19. Sheen, Y. (2006). Exploring the relationship between characteristics of recasts and learner uptake. Language Teaching Research, 10(4), 361–392. Slavkov, N., Melo-Pfeifer, S., & Kerschhofer-Puhalo, N. (2022). The changing face of the “na- tive speaker”: Perspectives from multilingualism and globalization. De Gruyter Mouton. Wakisaka, M. (2018). Face-to-face tandem and etandem: Differences that inf luence the mainte- nance of tandem learning activities. Revista Do GEL, 15(3), 42–57. Sylwia Scheuer, Céline Horgues „Nach drei Monaten klinge ich immer noch wie ein Anglophoner”: Erfolgs- und Misserfolgsgeschichten von englisch- und französischsprachigen Teilnehmern an einem Tandemprogramm Z u s a m m e n f a s s u n g Im vorliegenden Artikel werden zwei Möglichkeiten beschrieben, wie Erfolg oder Misserfolg in einem nicht-institutionellen Fremdsprachenlernkontext, und zwar in ei- nem Sprachtandem, operationalisiert und quantifiziert werden kann. Dabei wird auf die SITAF-Datenbank gestützt, wo 25 Stunden Videoaufnahmen mit Gesprächen von 21 Paaren („Tandems“), jeweils bestehend aus einem englischen und einem französischen Muttersprachler, gespeichert wurden. Die Teilnehmer wurden bei Gesprächen in beiden Sprachen aufgenom- men, so dass jeder von ihnen die Möglichkeit hatte, sowohl als „Experte“ als auch als „Anfänger“ (Lernender) zu agieren. Die Tandempartner trafen sich regelmäßig und autonom auch außerhalb der Aufnahmesitzungen miteinander und eines ihrer erklärten Ziele war es, Fremdsprachenkenntnisse zu erweitern. Zwei mögliche Maßstäbe für die Erreichung dieses Ziels waren: (1) Der Grad der Absorption (uptake) des korrigierenden Feedbacks (corrective feedback) vom Expertenpartner. In dieser Hinsicht wurden signifikante Unterschiede zwischen den beiden Sprachgruppen festgestellt: während 52% des von französischen Muttersprachlern “Three months on, I still sound like an Anglophone”… TAPSLA.12609 p. 21/28 gegebenen Feedbacks von ihren englischsprachigen Partnern vollständig absorbiert wurden, blieben mehr als 52% des in Englisch gegebenen Feedbacks bei ihren französischsprachigen Partnern wirkungslos. (2) Die Selbstberichte der Teilnehmer über die erzielten Fortschritte, sowohl aus der Sicht von Experten als auch von Anfängern. Die Berichte wurden durch Fragebögen erhoben, welche die Partner nach dem dreimonatigen Tandem-Lernprogramm ausfüllten. Die Analyse der Antworten auf die gestellten Fragen zeigt ein hohes Maß an Zufriedenheit unter den Teilnehmern, sowohl in Bezug auf die Fortschritte in bestimm- ten Domänen der Fremdsprache (Wortschatz, Aussprache usw.) als auch auf die allgemeine Sprachfertigkeit. Dabei war der angegebene Zufriedenheitsgrad bei den englischsprachigen Teilnehmern etwas höher als bei den französischsprachigen Teilnehmern. Schlüsselwörter: korrigierendes Feedbacks, Fremdsprachenlernen im Tandem, Absorption (uptake) A p p e n d i x General Tandem Questionnaire (English-speaking participants)10 University year: Concentration/Major: 1. Number of tandem meetings with your partner since the January 31, 2013, meeting: ……… Did you find the number of meetings to be (circle your answer)…: insuf- ficient / sufficient / too frequent? 2. How often did you meet your tandem partner? (circle your answer): twice per week / once per week / once every other week / once every 20 days/ once per month Did you find the frequency of meetings to be (circle your response)…: insufficient / sufficient / too frequent? 3. On average, how long did your tandem meetings last? 30 mins / 1hour / 1h30 / 2 hours / more than 2 hours 4. What percentage of time was spent speaking French in your tandem conversations? [ 0 / 10 / 20 / 30 / 40 / 50 / 60 / 70 / 80 / 90 / 100 ] % of the time 5. What percentage of time was spent speaking English in your tandem conversations? [ 0 / 10 / 20 / 30 / 40 / 50 / 60 / 70 / 80 / 90 / 100 ] % of the time 10 The French participants received equivalent versions in French. TAPSLA.12609 p. 22/28 Sylwia Scheuer, Céline Horgues 6. Over the semester, have you had the opportunity to converse with other French speakers: daily / several times a week / once a week / a few times over the semester / no Explain: 7. During your tandem conversations: always almost always often sometimes almost never never You begin the meeting in the same language (which language: ...............) You start speaking in one language at the beginning of the conversation and then switched to the other language You speak in one of the two languages most of the time You speak one language at one meeting and the other language at the next meeting You switch from one lan- guage to the other through- out the conversation (for example when there was a comprehension problem) You don’t mix languages, except to ask specific vocabulary or grammar questions Other (explain): 8. During a conversation in a foreign language, if you have doubts about how to express something or if you have problems expressing yourself (gram- mar, vocabulary, pronunciation): always almost always often some- times almost never never You stop and explain your problem in the foreign language You stop and explain your problem in your native language You continue to speak and wait for your partner to react “Three months on, I still sound like an Anglophone”… TAPSLA.12609 p. 23/28 9. When you make a mistake in French: always almost always often some - times almost never never Your tandem partner corrects you Your tandem partner corrects your vocabulary Your tandem partner corrects your grammar Your tandem partner corrects your pronunciation 10. When your tandem partner makes a mistake in English: always almost always often sometimes almost never never You correct them You correct their vocabulary You correct their grammar You correct their pronunciation 11. When your partner tells you something in English: always almost always often some- times almost never never You listen and try not to inter- rupt You listen and ask questions so as to help the conversation going on You interrupt your partner when you cannot understand what is said You interrupt your partner when they makes a mistake 12. Did you notice any differences in your partner’s conversational habits that would be linked to their culture? YES/NO If yes, explain: 13. Have ever helped your partner in their academic work? YES/NO TAPSLA.12609 p. 24/28 Sylwia Scheuer, Céline Horgues 14. Has your partner ever helped you in your academic work? YES/NO 15. Working in tandem with your partner allowed you to improve your knowledge and skills in which of the following domains? (Circle your answer) 0 = no improvement 5 = much improvement – culture 0 1 2 3 4 5 – phonetics/pronunciation 0 1 2 3 4 5 – grammar 0 1 2 3 4 5 – vocabulary 0 1 2 3 4 5 – general ease of expression 0 1 2 3 4 5 16. Evaluate your confidence in speaking French before beginning your tandem meetings and after: 0 = less confident 10 = more confident before: 0 1 2 3 4 5 6 7 8 9 10 after: 0 1 2 3 4 5 6 7 8 9 10 17. On a scale of 0 to 10, how was your experience with working in tandem with your partner? (Circle your answer) 0 = the most negative 10 = the most positive 0 1 2 3 4 5 6 7 8 9 10 18. What is your overall impression of your tandem conversations? (obsta- cles encountered, benefits gained) Tandem and Pronunciation Questionnaire (English-speaking participants) When speaking with your tandem partner in English: 1. Do you adapt the way you speak to them? YES / NO. If so, what changes do you make? “Three months on, I still sound like an Anglophone”… TAPSLA.12609 p. 25/28 YES NO I speak more slowly I articulate more clearly I speak louder My intonation is clearer My vocabulary is more simple I use more straightforward sentence structure I use shorter sentences 2. Now choose one of the terms below to describe your tandem partner’s accent [in English]: very strong / quite strong / moderate / quite slight / slight / very slight / no accent at all 3. What do you think about the French accent in English as a rule: 4. I correct my partner’s pronunciation (circle one of the following): systematically, whenever I hear a mistake / almost always / often / some- times / only when they ask me to / only when they ask me a specific question about a particular word / hardly ever / never 5. I prefer to correct my partner’s pronunciation (circle one of the following): on the spur of the moment / at the end of their sentence / when they’ve finished saying what they have to say / at the end of our tandem session 6. What exactly do you correct when it comes to your partner’s pronuncia- tion? YES NO Please give examples: Their intonation The rhythm of their English Their word stress The speed at which they speak Their consonants TAPSLA.12609 p. 26/28 Sylwia Scheuer, Céline Horgues Their vowels Their general fluency Other: 7. When you don’t correct your partner’s pronunciation, it’s because: YES NO Their mistakes are too small It would be impossible to correct all their mistakes You don’t want to make your partner feel uncomfortable You don’t want to interrupt the flow of their ideas You don’t think it’s polite to correct or to interrupt your partner You don’t like it when other people correct your mistakes when you’re speaking You don’t think correcting them would be helpful You can understand them despite their mistakes Your partner doesn’t take your corrections on board You can’t hear their mistakes 8. Your partner’s pronunciation in English prevents you from understanding them in English …….% of the time: 0 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Please provide specific examples from your tandem experience: 9. The advice you have given your partner has helped them to improve their pronunciation (circle): I haven’t noticed any changes / yes, most of aspects of their pronunciation / some aspects of their pronunciation / no, not really / no, definitely not Say what has improved: List any remaining difficulties: 10. Which pronunciation problems annoy you the most [in English by French speakers] even if they don’t hinder actual comprehension? When you’re speaking with your partner [in French]: 11. Would you say that your accent is: “Three months on, I still sound like an Anglophone”… TAPSLA.12609 p. 27/28 very strong / quite strong / moderate / quite slight / slight / very slight / no accent at all 12. What do you think about the English accent in French in general? 13. What aspects of French pronunciation do you think you still need to improve? 14. Your partner corrects your pronunciation (circle one of the following): systematically when they hear a mistake / almost always / often / sometimes / only when I ask them to / only when I ask them a specific question about a word / hardly ever / never 15. Your partner prefers to correct your pronunciation (circle one of the following): on the spur of the moment / at the end of my sentence / when I’ve finished saying what I have to say / during the tandem assessment session 16. What does your partner correct about your [French] pronunciation? YES NO Please give examples Your intonation The rhythm of your French Your word stress The speed at which you speak Your consonants Your vowels Your general fluency Other: 17. How often does your partner correct your pronunciation? YES NO Too often because it interrupts the flow of conversation Too often because it makes you feel uncomfortable Very often but you find it helpful From time to time and you’re okay with that TAPSLA.12609 p. 28/28 Sylwia Scheuer, Céline Horgues Rarely and you think it’s a shame Hardly ever, which is fine because you don’t like being interrupted when you’re speaking 18. Your pronunciation [in French] prevents you from making yourself un- derstood by your partner …..% of the time 0 1 2 3 4 5 6 7 8 9 1 0 Please provide specific examples: 19. Speaking with your partner and receiving advice from them have helped you to improve your pronunciation (circle): I haven’t noticed any changes / yes, most aspects of my pronunciation / some aspects of my pronunciation / no, not really / no, definitely not Say what has improved: List any remaining difficulties: