In the introduction, I described my intended role. In this section I look back at how this enfolded during data collection. I was involved with the pupils and thus, inevitably, I ‘biased’ the outcomes of the study. Therefore, it is important to reflect on some epistemological considerations. This section draws mostly on the logbook with field notes I kept throughout the data collection.
5.2.1 Expectations
As a Validiv team member, I visited sixteen schools in the former mining area in Limburg to collect quantitative data for the project, before my own data collection started. In these schools the monolingual habitus (Gogolin 2002) was present more than I had expected (for example, one school had a banner saying ‘Here, we speak Dutch!’). Also, in every single school, both the teachers and the pupils I talked to, told me that talking Turkish was ‘not done’. This should be interpreted in both senses: it was not allowed, and according to pupils and teachers, it didn’t happen often. Thus, when I started my own data collection with my eight participants, I was a little concerned the pupils would not speak any Turkish, despite my intention to make very explicit to the pupils that Turkish was allowed in the translanguaging space. On the other hand, some teachers claimed that Turkish was used rather extensively and that it was very hard to weed out. So when the data collection started, I was very curious what would happen, and I had no clear-cut expectations.
5.2.2 A double role
To give the pupils the opportunity to interact in Turkish, my role as (Dutch- only) teacher was intended to be very minimal: (1) to install a welcoming climate toward their languages, by asking them to teach me some Turkish words; and (2) to give them the task instructions and then leave them be. However, as they were not used to doing group work, that proved impossible. Pupils repeatedly called for my help from the start of the intervention. As a result, I was obliged to interfere more than I had intended, and I started to take up a double role: that of researcher and teacher Thus, I was no longer a detached observer, but a participant observer.
I found myself struggling with the dual role of observer participant and participant observer, especially at the start of data collection. I made a lot of notes about when the pupils used Turkish: on the one hand I was excited when they used Turkish (the researcher would have data to analyze), on the other hand, I was frustrated because I could not understand them (as a teacher I wanted to be involved). In my field notes, a reoccurring theme is the dilemma between observing and intervening. Typically, I waited a few minutes. Regularly, a dispute or trouble source was resolved automatically. If not, I intervened. Over time, the tension between researcher and teacher decreased. The field notes describe how it became easier to let them interact,
because of several reasons. As the transcriptions came in, I noticed they spoke less Turkish then I had estimated. Moreover, I could read that they did not use Turkish to make jokes or gossip about me (something I worried about sometimes). Furthermore, in line with the literature, the seemed to talk Turkish for on-task purposes. Those facts were reassuring to me, and that helped me to trust them more during data collection.
5.2.3 The teacher role
The aim of our study was not only to uncover language practices of bilingual pupils, but also to provide policy makers and school teams with advice. Therefore, it was an advantage that I could share the experiences of teachers who do not understand mother tongues. I am pretty much like the prototypical Flemish teacher (female, aged in my thirties, Dutch-speaking, and speaking no minority language). This contributed to the ecological validity of the study as the pupils were guided by a ‘Dutch-only’
teacher. Moreover, it enabled me to share with teachers what it is like not to understand the children, which is not always easy, as I experienced first-hand.
As a teacher, I appeared to represent the monolingual habitus in some instances (contradictory to my intent). This became clear in my field notes where I stated about the blue group that ‘they were speaking Turkish most of the time’ (which, according to the transcripts, turned out untrue). Moreover, my (unintended) monolingual mindset became apparent in episodes where I interrupted them talking Turkish. The teacher in me reacted spontaneously, and interrupted them out of a concern that they would accomplish the task. My teacher alter ego just took over, and was apparently imbued with the monolingual habitus, as I didn’t give them the benefit of the doubt that they might actually be working on-task in Turkish. As a result, pupils might have chosen for Dutch more than they would have if I hadn’t interrupted them. In the transcriptions it can be noticed that pupils resorted to Dutch after my intervention, and then started using Turkish again after a few minutes.
5.2.4 The researcher role
An advantage of not having Turkish roots, was that the children were very eager to tell me about Turkish habits and culture in detail. For example, when I asked them advice about a Turkish wedding, or when we talked about their holidays in Turkish, they were very enthusiast (for example, they started singing traditional songs) and gave me very detailed information. This way, I got more insight in their lives than a Turkish-Dutch bilingual might have obtained, as pupils might not have discussed these topics so thoroughly. On the other hand, it might have been easier for the pupils to clarify some experiences using Turkish. To remedy this, the camera was presented as understanding all languages, including Turkish. Generally, when addressing the camera directly, the pupils talked Dutch, but on some occasions they used Turkish, for example, Nuran presented her ‘language portrait’ to the camera in Turkish.
An important point to make about my researcher role, concerns my mindset as a researcher. Some might say that people advocating a monolingual norm would never carry out this kind of research. This critique might be true. In the literature on multilingualism and multilingual learning, the vast majority of studies seem to be framed from a perspective which welcomes multilingual practices. I appeared to combine both standpoints. On the one hand, I sometimes happened to be an agent of the monolingual norm (which stems from my background as Flemish and monolingual-Dutch raised). On the other hand, my literature study provided me with the information that multilingual practices might be a resource for learning. I consider this ambivalence as an advantage, in the sense that I was not predisposed with strong expectations or opinions about what should happen.
5.2.5 Impact on the pupils
An important consideration is how the pupils perceived me (in both my role as a teacher and a researcher), as pupils might have shown socially desirable behavior (or might have tried to). On the one hand, they could perceive me as a representative of the dominant society, as I have a Belgian background and cannot speak Turkish. Based on those observations, they might have guessed I would prefer the monolingual norm.
This might account for the finding that they defend monolingual rules, but not for being enthusiast about the use of Turkish during data collection (in their actual behavior as well as in their reflections). On the other hand, they might have perceived me as an exception to the majority of teachers, promoting multilingual use. That would account for their positive reactions to the translanguaging space, but it contradicts their advice to stick to a ‘mostly’ monolingual policy, albeit with some more openness toward mother tongue use.
If I had shared their language and culture, a different dynamic might have occurred.
Maybe they would have been more open, sharing experiences with a companion with the same background. Also, task performances guided by a teacher of Turkish descent, might have elicited very different language choices from what I observed. Pupils might have talked more Turkish in that situation. However, as illustrated in the studies, Dutch was an important part of the linguistic repertoire used. I wouldn’t expect pupils to turn to the use of Turkish only, certainly not during task performance, as they appeared to need Dutch when discussing content of the instruction cards (given in Dutch). We considered the option to take the transcriber to the schools, but due to a lack of time and practical constraints, this plan was abandoned. It would, however, absolutely be an interesting idea for a follow up study.
Finally, because of my role as a participant observer, the ‘observers’ paradox’ (Labov 1972) must be considered: the mere presence of the researcher might have an impact on the behavior of participants. Indeed, at the start of the data collection, pupils turned directly to the camera and acted excited and playfully. However, during the course of
each task performance, they appeared to forget about my presence and the camera discussing matters not intended for my ears or the camera (both in Turkish and Dutch). For example, they were gossiping about peers and teachers, cursing, sharing secrets and so on. On some of these occasions, they suddenly became aware of being observed, for example, at one moment, Ali warned Kerem when he cursed, by pointing at the voice recorder. However, the vast majority of the time they neglected the recording devices and acted very naturally, which contributed to the ecological validity of the study.