‘I feel like I have a superpower’: a qualitative study of adolescents’ experiences of multilingual identity development during an identity-based pedagogical intervention

ABSTRACT

Multilingual identity, considered here as being shaped by learners evaluations, emotions and experiences relating to languages and language learning, has gained increasing attention among both researchers and practitioners. Indeed, school-based studies have not only suggested meaningful connections between learners’ multilingual identity and attainment but have also indicated the potential for an identity-based pedagogy to enhance students’ multilingual identity. However, much research to date has tended to focus on quantitative data at the level of a class or cohort. This paper, therefore, aims to qualitatively explore how adolescents develop their understandings of multilingualism and various dimensions of their own multilingual identity during an identity-based pedagogical intervention in the languages classroom. A qualitative approach was adopted drawing on interview data from 14 Year 9 (age 13–14) learners in secondary schools in England who participated in a year-long identity-based intervention in their languages lessons. The following three profiles of development were identified which capture students’ varied experiences of the intervention: resistant multilingual identity development, emergent multilingual identity development and reflexive multilingual identity development. We reflect on each of these in turn and highlight both theoretical and pedagogical implications for developing students’ multilingual identity in the classroom.

KEYWORDS:

• Multilingual identity
• multilingualism
• pedagogical intervention
• qualitative

Introduction

While the links between language, education and identity have long been recognised (e.g. Wenger 1998), in recent years there has been increasing interest in how students experience identity development in the classroom and how teachers can positively and proactively contribute to this process. Within the field of languages education this is evidenced by a flourishing body of research into students’ linguistic identity (i.e. the way one identifies – or is identified by others – in each of the languages in one’s linguistic repertoire) and multilingual identity (i.e. an ‘umbrella’ term which encompasses, but also transcends, linguistic identities). While this will undoubtedly be influenced by a wide range of experiences such as language(s) used in the home or shifts in the sociocultural environment due to migration or study abroad, more recent research has focused on the role of the school which, for many, will include instructed language learning. Indeed, not only has multilingual identity been associated with academic attainment across the curriculum (Rutgers et al. 2021), but there is evidence to suggest that languages teachers can actively support students’ multilingual identity development through incorporating an identity-based approach (Forbes et al. 2021). However, to date, much of the research in this area has focused on quantitative data at the level of a whole class or cohort. This paper, therefore, seeks to address this gap by qualitatively exploring whether and how 14 secondary school learners in England develop their understandings of multilingualism and various dimensions of their own multilingual identity during a 1-year identity-based pedagogical intervention in the languages classroom. Such insights, in turn, have the potential to inform pedagogy by providing crucial evidence on how such interventions may be made more effective.

Literature review

Multilingual identity

Identity can be broadly defined as ‘who we are and who we think ourselves to be, [it] is not innate, but is something that is constructed as part of a dialogue with ourselves, others, and social institutions and structures’ (Chaffee 2019, 100); as such, it is both individual and social, and also subject to change. We argue that schools, as one of the key social institutions pertinent to children and young people, constitute key sites for such identity construction and (re)negotiation. Indeed, the role of dialogue mentioned here draws particular attention to the importance of language as the medium through which identity is both (re)negotiated by the individual and (re)presented to others. By extension, it is important to acknowledge that many students will have more than one language in their repertoire (whether learned in the home and/or at school) each of which may contribute in different ways to their sense of identity. The focus of this paper, therefore, is on the construct of multilingual identity which we use as an umbrella term that encompasses, but also transcends, an individual’s language-specific identities (sometimes referred to as linguistic identities) (Fisher et al. 2020; Henry 2017). As noted by Berthele (2021), there is continued debate around the way in which scholars define and conceptualise multilingualism which centres, at least partially, on complexities surrounding the countability and boundedness of languages. For the purpose of this study, we use the term multilingual in its broadest sense to include all of the languages in a learner’s repertoire, regardless of their level of exposure to or proficiency in each language. Within this, we include named languages such as Italian or Urdu, but also dialects and non-verbal forms of communication.

Drawing on Fisher et al. (2022), we further operationalise multilingual identity as being shaped by the interconnected 3Es: learners evaluations of languages and of themselves as language learners, their emotions relating to language learning and, their experiences of languages and language learning (both in and out of school). This constitutes the theoretical framework for this paper and of particular importance is the way in which learners experiences of an identity-based intervention in the classroom both shape and are shaped by their evaluations and emotions in relation to languages and language learning.

As noted above, a key component of multilingual identity is learners’ evaluations which encompasses their views of languages, of multilingualism and of themselves as multilinguals. While most existing research around students’ evaluations of languages has tended to explore their attitudes and beliefs (e.g. Henry and Apelgren 2008; Horwitz 1999), what is of greater relevance to this paper is the work on students’ understandings of multilingualism and of themselves as multilinguals which is more limited and varies according to context. For example, in a questionnaire-based study involving 116 lower secondary school students in Norway, Haukås (2022) found that students predominantly understood multilingualism as knowing multiple languages and the majority (67%) identified themselves as multilingual. As part of the same wider study this was followed up with students through digital data visualisations (Storto 2022) where they elaborated that they considered knowledge of two languages (i.e. Norwegian and English) as sufficient for self-identification as a multilingual speaker. Yet, using a similar questionnaire tool with 422 secondary school students in England, Bailey, Parrish, and Pierce (2023) found that almost 60% of their participants did not consider themselves to be multilingual, despite learning at least one other language in school; the implication being that they felt they did not know enough of their other languages to identify as multilingual. This echoes Wu and Forbes’ (2023) findings from among 24 high school learners of Japanese in China where ‘one’s status of being a multilingual was perceived as strongly, if not exclusively, associated with [foreign language] proficiency’ (9), a belief which hindered many from fully claiming a multilingual identity. Therefore, in the minds of adolescent learners it seems that both the number of languages known and proficiency in those languages are key to students’ understandings of multilingualism. Such evaluations may also extend beyond those directly connected to the present learning environment and include those related to past experiences and associated with the construction of a possible future self, that is, ‘individuals’ ideas of what they might become’ (Ushioda 2011, 201).

A second key component of multilingual identity is that of emotions, which are ‘intricately involved in the way learners perceive their experiences and how these perceptions have an effect on their experiences of self’ (Miyahara 2015, 162). This encompasses not only students’ immediate emotions in relation to language learning (e.g. anxiety, enjoyment or pride), but also more broadly how languages influence the way they ‘feel’ about themselves. Indeed, there is evidence to suggest that this may differ across the various languages in their repertoire. For example, 65% of the 1039 bi/multilingual participants in Pavlenko’s (2006) study indicated that they sometimes felt like a different person when using their various languages. This was further reinforced by qualitative case study data collected from six bilingual English and Japanese-speaking women conducted by Hemmi (2014, 84) where one participant, Mako, commented: ‘I have two faces. One for the English speaking situation and the other for the Japanese. It is difficult to play both roles at the same time’, perhaps suggesting separate linguistic identities rather than a more holistic multilingual identity.

The third vital (and interconnected) component of our model of multilingual identity is that of experiences; indeed, Fisher et al. (2022) found that multilingual identity is more strongly connected with direct experiential contact with languages than with explicit beliefs about language learning. Such experiential contact may take the form of students’ heritage background (e.g. Leeman 2015), a shift in sociocultural context due to migration (e.g. Norton 2013) or study abroad (e.g. Barkhuizen 2017). However, crucially for this paper, we argue that instructed language learning in school is also key; after all, teaching and learning have long been recognised as ‘continuous processes of reconstruction of experiences’ (Dewey 1938, 111). Indeed, recent evidence suggests that school experiences of language learning can have an even greater effect on fostering a more open attitude to societal multilingualism in adulthood than experiences linked to heritage backgrounds (Chik and Melo-Pfeifer 2023).

Identity-based interventions

Yet, while such a link between education and identity has long been recognised, too often there is an assumption that this will occur implicitly, without the teacher drawing attention to such processes. If, as suggested above, multilingual identity can play an important role in students’ engagement with languages and their attainment more broadly, then it is crucial to more fully understand how teachers can more explicitly support their multilingual identity development. This has led to an increasing interest in recent years in identity-based education (Schachter and Rich 2011), which can be considered as an experience in itself that seeks to actively promote identity (re)negotiation through activities that encourage reflection and reflexivity in the classroom. This, in turn, is closely linked with the concept of agency i.e. developing learners’ awareness of multilingualism may provide them with the agency to see themselves as multilingual and, ultimately, to claim a multilingual identity.

There have been a number of studies in recent years which have explored the effect of identity-based interventions across a range of topics and subject areas, for example, motivation (Oyserman et al. 2021), educational commitment, values and persistence (Perez, Gregory, and Baker 2022) and mathematics (Heffernan et al. 2020). Yet, much less is known about the effect of identity-focused interventions in the languages classroom which, as we argue elsewhere (Fisher et al. 2020), is a key site for the (re)negotiation of learners’ multilingual identities. To explore this further, the authors developed, implemented and evaluated an identity-based intervention in the language lessons of 268 Year 9 (age 13–14) students across four secondary schools in England (Forbes et al. 2021). Analysis of the questionnaire data at the whole-group level revealed a positive shift in the evaluations of and emotions towards languages among the students in the identity-based intervention group, which led, on the whole, to a greater willingness to claim a multilingual identity. However, while this provided valuable insights into the more general effects of identity-based instruction, it raised questions around how students engaged with such interventions and developed (or not) their multilingual identity. The answer to such questions is crucial for further informing pedagogical decisions; as noted by Miyahara (2015, 166), ‘the link between identity and classroom practice has been largely downplayed or underrepresented in the literature on identity’. This paper, therefore, seeks to address this gap by exploring the following research question: how do adolescents develop their understandings of multilingualism and various dimensions of their own multilingual identity during an identity-based pedagogical intervention in the languages classroom?

Methodology

Context and overarching research design

The data reported in this study derive from a larger, quasi-experimental mixed methods study which involved developing and implementing an intervention of identity-based pedagogy with 268 Year 9 (age 13–14) students and their languages teachers in four state-funded secondary schools across the East of England and London. These schools were selected to represent a range of geographical area, linguistic diversity and social deprivation (see Table 1). The students were all in their final year of compulsory foreign language education and therefore represented a wide range of attainment levels and attitudes towards language learning.

Table 1. Overview of participating schools.

School	Description
A	Semi-rural area, high level of linguistic diversity, below average levels of social deprivation.
B	Rural area, very low levels of linguistic diversity, below average levels of social deprivation.
C	Urban area, average levels of linguistic diversity and social deprivation.
D	Urban area, very high levels of linguistic diversity and social deprivation.

The quasi-experimental design of this study involved working with three intact foreign language classes in each school (French, German or Spanish). One group in each school was designated as the control group which continued with their normal lessons while the other two groups received six 1-hour intervention lessons over the course of an academic year in their timetabled language lessons taught by their regular classroom teacher. One of these intervention groups received what we refer to as a ‘partial’ version of the intervention which focused on cultivating students’ knowledge about multilingualism across a range of contexts (for example, the cognitive and social benefits of language learning, sociolinguistic knowledge such as the use of dialects, and links between culture and language). These sessions also included input (such as text, audio and video) and various tasks in the relevant target language so as not to detract from students’ language learning. The other group received the ‘full’ version of the intervention which, in addition to developing students’ knowledge, also adopted an identity-based approach by incorporating activities which encouraged them to reflect on how the knowledge presented related to themselves as users and learners of multiple languages (see Fisher et al. 2020; Forbes et al. 2021 for further details and www.wamcam.org for freely downloadable teaching resources developed from this study). The rationale for the two intervention groups was to explore the potential effect of an identity-based approach compared to more traditional, knowledge-based interventions promoting multilingualism. In light of the quantitative evidence outlined above which indicated the effectiveness of the identity-based approach in contributing to students’ multilingual identity development, we therefore focus further here on the experiences of the students in this ‘full’ intervention group.

Sampling of focal participants

While the broader study involved a range of data collection methods including questionnaires and school attainment data (collected for all students), we focus here on the qualitative data collected via interviews with 14 focal students who were part of the full identity-based intervention classes. These 14 students were purposively selected from the wider group at the beginning of the study to represent a heterogeneous range of backgrounds and perspectives in terms of their school, home language(s), language learning experiences in school and attitudes (see Table 2). These decisions were made on the basis of their responses to a pre-intervention questionnaire (which targeted each of these areas) and also in discussion with their teachers.

Table 2. Overview of participants.

School	Pseudonym	Gender	Home language(s)	School language(s)
A	Adam	Male	English	French, German
A	Eva	Female	English	French, German, Japanese
A	Helena	Female	German	French, German, Italian
A	Kenji	Male	English, Japanese	French, German, Mandarin
B	Emilia	Female	English, Afrikaans	Spanish
B	Ethan	Male	English	Spanish
B	Freya	Female	English	Spanish
C	Abigail	Female	English	French
C	Alison	Female	English	French
C	Jacob	Male	English	French
C	Jordan	Male	English	French
C	Lia	Female	Portuguese	French, Spanish
D	Camila	Female	English, Portuguese	Spanish
D	Giorgio	Male	English, Italian	Spanish

Interviews

The research question and qualitative data at the heart of this paper seek to gather insights into these students experiences of the identity-based intervention. As such, this study is underpinned by social constructionism which emphasises ‘the world of experience as it is lived, felt and undergone by people acting in social situations’ (Robson and McCartan 2016, 24), in this case, the languages classroom. Data were collected via individual, semi-structured pre- and post-intervention interviews with each of the focal students (lasting approximately 30 min each). Interviews allow the possibility of ‘understanding the lived world from the perspective of the participants involved’ (Richards 2009, 187) and, as such, are a common method in studies exploring language and identity. However, as cautioned by Duran Eppler and Codó (2016, 308), ‘talking about one’s identity is no easy task, as it goes to the heart of our most intimate beliefs and emotions […] for this reason, inquiring directly about identity is generally useless’. In developing the interview schedule, we therefore drew on the 3Es framework outlined above to ask students about their evaluations, emotions and experiences (see Table 3). While the questions were broadly the same for both interviews, the post-intervention interview included additional questions on students’ experiences of the intervention itself and the extent to which they felt their views had shifted (or not).

Table 3. Key interview questions.

Dimension of multilingual identity	Key questions
Evaluation	What do you think it means to be multilingual? How important do you think it is to know other languages?
	To what extent do you think you are multilingual? Why? (Post-intervention interview: Do you think this has changed over the course of the year? Why (not)?)
	How well do you think you’re doing in your language lessons? How do you know?
Emotion	Do you enjoy learning languages? Why (not)?
	Can you give me an example of a time when you felt proud of yourself in the language classroom?
Experience	Can you tell us about your experience of learning languages in school? For example, did you choose this particular language? If you didn’t have to study a language would you still choose to?
	Post-intervention interview: This year you have been doing some different activities in your language lessons – what can you remember about these? Did you learn anything about languages or multilingualism that surprised you?

As part of the wider study the students also completed a questionnaire before each of the interviews. While we do not focus here on the questionnaire data as such, in the interviews we drew on their responses to one particular item as a stimulus to further prompt their qualitative reflections: the multilingual visual analogue scale (mVAS). This was a 100 mm straight line with the labels of ‘monolingual’ and ‘multilingual’ at each end, and students were asked to put a cross to indicate where they would position themselves on this continuum (see Rutgers et al. 2021). In the interviews we asked students to reflect on this as a means of accessing their evaluations of themselves as multilingual (or not) and to think about how this may have shifted over the course of the intervention. All interviews were conducted in-person by members of the research team in the students’ schools and were audio recorded and transcribed verbatim. We fully acknowledge that it is not possible to generalise from such small-scale data and instead follow Lew et al.’s (2018) recommendations to ensure that the data are credible through being transparent about the process of data collection and analysis and providing background information about the context and participants to enable transferability of insights to other settings.

Ethical considerations

As cautioned by Hultgren, Erline, and Chowdhury (2016, 268) ‘ethics is something that is part of each phase and aspect of language and identity projects and should inform every decision made’. From the perspective of procedural ethics, we gained institutional approval and informed consent from the participants themselves along with key gatekeepers in their respective schools (e.g. Headteachers). We also ensured ongoing assent from students throughout the project by providing age-appropriate explanations of what was involved and reminding them of their right to stop the interview or withdraw at any point. Given the potentially sensitive nature of some of the questions around identity we were mindful about how questions were framed and were careful not to suggest or impose any identity positions on students both in the interviews and the intervention itself.

Analysis

In the initial stage of analysis we adopted a deductive orientation to thematic analysis which is defined as ‘researcher- or theory-driven’ (Braun and Clarke 2022, 57). The initial coding process was informed by the 3Es framework outlined above and involved looking individually at understandings of and development of multilingual identity for each participant. Within each of these, several sub-themes were subsequently identified from the data (see Table 4).

Table 4. Initial coding framework.

Theme	Sub-theme(s)	Description
Evaluation	Of language	Students’ views on languages e.g. whether they are important and useful.
	Of being multilingual	Students’ understandings of multilingualism e.g. in terms of number of or proficiency in languages.
	Of themselves as multilingual	Students’ views on whether and how they consider themselves as multilingual (or not).
Emotion	Enjoyment	References to enjoyment (or not) of language learning.
	Pride	References to feelings of pride associated with learning or using languages.
	(Dis)association	References to feeling connected to or disassociated from language learning.
Experience	Of intervention	Students’ reflections on the intervention sessions.

The next step involved looking for broader patterns (both similarities and differences) across the themes with a particular focus on changes between the pre- and post-intervention interviews. As a result, it became evident that some of the students shared similar characteristics and three distinct profiles or trajectories of development were identified (see Table 5). While we acknowledge that not all students fitted perfectly into one of these profiles, analysis indicated that they capture the broad patterns within the group and therefore provide a useful indication of how students developed (or not) their multilingual identity over the course of the intervention. The following section explores each of these profiles in turn, drawing on data from the representative students.

Table 5. Overview of student profiles.

Profile	Number of students	General characteristics
Resistant multilingual identity development	3	• Limited understanding of multilingualism throughout. • Resistance to identifying as multilingual throughout. • Limited engagement with the intervention activities.
Emergent multilingual identity development	3	• Evidence of a growing understanding of multilingualism. • Indications of the development of some aspects of multilingual identity following the intervention. • Some engagement with the intervention sessions, in particular, the knowledge dimension.
Reflexive multilingual identity development	8	• Development of a broad understanding of multilingualism. • Stronger identification as multilingual following the intervention. • Reflective and reflexive engagement with the intervention sessions.

Results

Resistant multilingual identity development

The profile of resistant multilingual identity development is characterised by a limited understanding of multilingualism and a resistance to identifying as multilingual which remained relatively consistent throughout the intervention. There were three students who demonstrated these characteristics who all came from English-speaking backgrounds (Abigail, Ethan and Freya).

Evaluation

Of languages

In terms of their evaluation of languages, these students at times expressed an appreciation of their instrumental value, for example, ‘I think it helps like with travelling’ (Abigail). However, overall, they considered them throughout as ‘difficult to learn’ (Freya) and less important than other subjects: ‘I’d benefit more from doing extra maths, science and English […] especially if I want to work in this country’ (Ethan).

Of being multilingual

When reflecting on their understanding of the term ‘multilingual’ this group felt strongly that this entailed being ‘fluent in more than just one language’ (Ethan) and placed a lot of emphasis on proficiency. Such views didn’t shift as a result of the intervention and, in some cases, were further reinforced. This group also largely associated the development of such fluency with experiences outside of school. For example, Abigail commented that multilinguals have ‘obviously grown up learning more than one language’ and Freya felt that becoming multilingual would be a particular challenge for those who ‘don’t have a background from that country’. It seems, therefore, that the resistant group largely understood being multilingual as determined by factors outside of their control (such as family background), rather than something which individuals have the agency to develop.

Of themselves as multilinguals

In light of their above understandings of multilingualism, it is unsurprising that these students were largely resistant to identifying as multilingual (evidenced in part by their tendency to position themselves closer to ‘monolingual’ on the mVAS following the intervention). This was principally attributed to increased difficulty in their language learning course over the year of the intervention which made them feel further away from the perceived ‘required’ level of fluency: ‘I feel like more monolingual than multilingual […] I don’t speak French fluently so I am just trying to learn so I think I might be a bit lower [then before the intervention]’ (Abigail); ‘I just struggled more and more in Spanish, I was getting basically worse which is why I put the “x” further along the monolingual line’ (Ethan). This was underpinned by a lack of confidence in learning languages; for example, Ethan considered himself as ‘not doing very well’, while Freya commented that ‘as it got more difficult I just didn’t enjoy it and I just, I thought my confidence went a bit downhill’. Another key characteristic that emerged among this profile which may have also contributed to their reluctance to consider themselves as multilingual, was their weak future self in relation to languages. While they saw languages as potentially useful for some people (e.g. for travelling or getting a job, as noted above), they did not see them as having any relevance for their own future. Freya, in particular, in the final interview more explicitly expressed a resistance to being more multilingual in the future as she saw it as unattainable: ‘I don’t want to be like, fully multilingual because I don’t think I would be able to do it’.

Emotion

These students did not typically express any particular enjoyment in relation to language learning and any sense of pride was very strongly linked to specific instances of external validation by teachers, for example: ‘if I get my results of my tests and they’re what I need to be at then that makes me proud’ (Abigail). Overall, there was no evidence of any form of emotional connection to languages or language learning either before or after the intervention. In fact, what emerged was a more explicit disassociation, with Ethan, for example, not only reporting feeling out of his ‘comfort zone’ in the languages classroom, but indicating that he did not even feel like himself: ‘I felt like my voice sounded a bit different when I was pronouncing some words and it didn’t really sound like me so I thought, I just thought it sounded different, it was like a different person’.

Experience

Overall, the resistant profile is characterised by holding a rather limited understanding of multilingualism as requiring fluency in another language and as largely determined by family background or experiences outside school, rather than taught languages. This, along with their sense of languages as a difficult subject and not relevant to their own lives, both present and future, contributed to their continued resistance towards identifying as multilingual. As such, their experience of the intervention was limited; it seemed to have little or no effect on their understandings of multilingualism or on the development of their multilingual identity. Indeed, when asked to reflect on the intervention activities, they had little to say and often did not remember the sessions. What emerged within this group, therefore, was a rather fixed mindset: ‘I sort of had my mind set on how I wanted to be, like, in the scale of being multilingual, and I didn’t really change it’ (Freya).

Emergent multilingual identity development

The profile of emergent multilingual identity development is characterised by a growing understanding of multilingualism and more nuanced reflection among students on the various languages in their repertoire. While there was little evidence of an overall shift in their holistic multilingual identity as a result of the intervention, there were indications of the development of certain dimensions of identity. There were three students who demonstrated these characteristics (Emilia, Giorgio and Jacob).

Evaluation

Of languages

As with the first group, the emergent multilingual identity students also held rather instrumental views of languages at the beginning (e.g. as being useful for some for getting certain jobs or moving to that country). However, following the intervention there was much more variation in their evaluation of the different languages in their repertoire. For example, Emilia made a clear distinction between languages learned in the home and school: ‘cos, like, learning one language and being raised up with that one language is then like, completely different than when you learn another language [at school]’. She saw a clear separation between the two and didn’t feel that speaking a language other than English at home helped in any way with learning Spanish in school. Students in this group also tended to distinguish between different school languages and, rather than seeing all language learning as difficult (like students in the resistant group), they often had a clear preference for one language over another. For example, Giorgio chose Spanish as he felt it was ‘easier’ than French. In this group, therefore, the intervention seemed to have some influence on how students evaluated languages in relation to each other.

Of being multilingual

Within this profile, being multilingual was generally considered in a positive light as ‘a good skill to have’ (Jacob) and there was a noticeable shift in the emergent multilingual identity students’ understanding of the term. While in the first interviews they were more likely to align with definitions provided in the first profile (i.e. requirement of a certain level of fluency), after the intervention this changed to ‘you don’t have to be fluent in it’ (Jacob) and ‘you might have to like, learn the basics’ (Emilia). Interestingly, they also considered multilingualism to be dynamic and demonstrated an awareness that individuals might become more multilingual by learning a language or moving to a different country, or equally, may lose their knowledge of a language if they do not use it. However, similar to the first profile, these students largely associated multilingualism with experiences outside of school. Nonetheless, as shown in the next section, they demonstrated an ability to relate more personally to those experiences which was not evident among the first group.

Of themselves as multilinguals

This profile of students was more likely than the first to identify themselves as multilingual at the start, although it is important to note that their overall ratings on the mVAS and associated reflections during the interviews remained relatively stable over the course of the intervention. As noted above, these students largely associated their multilingualism with their experiences outside of school and felt that the different languages in their repertoire contributed to this in different ways, perhaps suggesting the emergence of stronger linguistic identities rather than a more holistic multilingual identity. The students in this group with English as an additional language were quite matter of fact about attributing their multilingualism to their home language, for example: ‘because my family is Italian’ (Giorgio); ‘because my mum speaks Afrikaans and has brought me and my brother up with it’ (Emilia). Even by the end of the intervention they did not consider languages learned in school to contribute to their overall identification as multilingual and often made an explicit distinction between the two. Emilia, for example, further clarified that the languages she learned in school didn’t necessarily make her feel more multilingual as she only knew ‘a little bit’. The first language English students in the emergent multilingual identity profile were also influenced by experiences outside of school, albeit in a different way. Jacob, for example, referred to holidays in France. Yet overall, much like the first profile, this group was similarly characterised by a weak future self in relation to languages by the end of the intervention, with Giorgio feeling certain that his future wouldn’t ‘have anything to do with Spanish’.

Emotion

Like students in the first group, those in the emergent multilingual identity development group also largely associated any sense of pride in relation to languages with achieving ‘a high mark’ (Giorgio) in a test or getting ‘something right’ (Emilia) in class. However, within this group there were some indications of a greater sense of enjoyment and some emerging deeper connections with languages following the intervention. For example, Jacob commented in the second interview that when he is in his language lessons he feels ‘like I’m a different person because it’s something that some people can’t do’. Here, feeling ‘different’ appears to be associated with more positive experiences with language learning which contrasts with Ethan’s comment in the previous profile, where feeling like a ‘different person’ was related to feelings of disassociation and unease in relation to language learning.

Experience

Overall, the emergent multilingual identity profile is characterised by some shift in students’ understandings of multilingualism following the intervention; for example, expanding their view from the need to be fluent in other languages to be considered multilingual, to only requiring some knowledge of languages. In addition, when asked to reflect on their experience of the intervention sessions the students often referred to what they learned about. For example, Giorgio was interested in learning about how English is influenced by other languages and Emilia spoke of learning about how accents can shape how people are judged. This indicates some engagement with the intervention lessons and the evidence suggests that this experience contributed to their knowledge about languages and raised their awareness about multilingualism. However, the students in this profile did not feel that the intervention sessions had any particular influence on how they viewed themselves as multilinguals. While some showed evidence of developing a stronger linguistic identity in relation to one or more of the languages in their repertoire, this was largely attributed to factors outside of school and they did not yet demonstrate a more holistic multilingual identity.

Reflexive multilingual identity development

The profile of reflexive multilingual identity development is characterised by reflective and reflexive engagement with the intervention sessions which contributed to a positive shift in their multilingual identity development. These characteristics were evident among the remaining eight students.

Evaluation

Of languages

Over the course of the intervention the views of languages among students in the reflexive profile became increasingly positive and, crucially, extended beyond the more instrumental and descriptive positions which were characteristic of the other two categories. Rather than considering languages as simply just useful for some jobs, in the final interviews the reflexive group emphasised the role of languages in understanding other people and cultures. For example, languages were seen as important to be able to ‘communicate with more people’ (Helena), Adam commented that ‘when you learn a language it’s like discovering a world as you understand the culture’ and Camila suggested that this, in turn, makes learners ‘more respectful of others […] because you kind of already understand there just are differences’. Languages were therefore understood by the reflexive group not just as a tool for communicating, but as important for connecting with others.

Of being multilingual

Within this profile students demonstrated a much broader understanding of what it means to be multilingual. Like the emergent group, they felt that you don’t need to be fluent to be multilingual, but there was much more evidence here of a shift in understanding due to the intervention sessions: ‘we had this lesson in my German class about being multilingual and before it I thought to be multilingual you had to know a different language, like, be fluent in one, but then after the lesson my opinion changed’ (Helena). In addition, conceptions of multilingualism extended beyond knowledge of the language itself and included the need to ‘have an understanding of another culture and […] how those people act’. (Adam). As with the previous categories, the reflexive group acknowledged that multilingualism can be influenced by experiences outside of school such as ‘your family or your ancestors’ (Kenji) but, crucially, they also recognised the role of experiences in school and felt that you can become multilingual ‘if you learn [a language] in a classroom’ (Helena). Learners within this group are therefore considered as having much more agency to become multilingual rather than this being a label ascribed to them because of their background.

Of themselves as multilingual

Students in the reflexive profile were characterised by an increasing willingness to identify as multilingual, as evidenced by a rise in their overall ratings on the mVAS and qualitative changes in their reflections across the two interviews. Interestingly, they largely attributed these changes to their experiences in school. While the resistant group had described a lack of proficiency and confidence in languages as inhibiting them from identifying as multilingual, the opposite was true for the reflexive group who spoke of feeling more multilingual due to having ‘learnt more French’ (Alison) or knowing ‘a bit more about different languages’ (Jordan). In addition to developing their proficiency in their languages, the reflexive group also reflected explicitly on the role of the intervention in how they viewed themselves:

I didn’t think about myself [as multilingual] before because I’ve got no cultural background being able to speak another language fluently, but the more we spoke about it in class and I think about it I think maybe I am because I’m able to study those languages. (Eva)

Another key characteristic of this group was the development of a strong future self in relation to languages. Following the intervention they described languages as being highly relevant to them in the future, whether that was to ‘study abroad at university’ (Adam), ‘to live abroad when I’m older’ (Eva) or simply a desire to learn more languages (Camila). Unlike the resistant group, who didn’t see becoming more multilingual in the future as attainable, for the reflexive group this was certainly within reach. For example, when reflecting on how her position on the mVAS might change in the future, Helena said:

I think I’m going to be on the end, very close to multilingual because I still have a few years until I finish school and then I think in that time I will still improve a lot in more languages.

This further highlights the role of the school context in achieving this goal.

Emotion

In the reflexive profile, students described language learning as enjoyable and following the intervention this was associated not just with learning the language but using it. For example, when referring to speaking in another language Kenji commented that it ‘makes me feel really happy because I actually know something that I learnt and I can use it in real life’. Similarly, these students also felt a strong sense of pride in relation to their languages which encompassed, but extended far beyond, the sense of pride connected to test scores which were characteristic of the other two categories. Rather than relying solely on external validation from teachers to feel proud, pride largely seemed to stem from their own ability to use the language either within or beyond the classroom:

I felt proud that, like, a few times when we were on an exchange [in Germany] we had to order food and a few of my classmates didn’t know how to do it so I stepped in and helped them […] I was really proud that I almost felt that I was fitting in, like I wasn’t an outsider to them. (Adam)

The sense of ‘fitting in’ or belonging in relation to languages also came through strongly in this group with Camila similarly commenting: ‘when I learn Spanish, I always learn about their culture, about their traditions and like, you just feel involved with it […] you feel like you’re someone from that country’ (our emphasis). These examples indicate that these students were not simply learning about the language but were forging a stronger emotional connection with the languages and their associated cultures and peoples. By extension, they felt languages to be empowering which further enabled them to claim a multilingual identity: ‘I feel like I have a superpower because it makes me stand out from English people and my superpower is to speak German’ (Adam).

Experience

Overall, the reflexive multilingual identity profile is characterised by developing a much broader understanding of multilingualism as involving not just learning the language, but learning about the culture and, crucially, using the language. While in line with the other categories the students here recognised the role of out-of-school experiences in becoming multilingual, they also highlighted the role of their school experiences as being influential. However, what was most characteristic of this group was the reflective and reflexive manner in which they engaged with the intervention sessions. While they developed knowledge about languages and multilingualism, as did students in the emergent profile, they went a step further and reflected on the implications of this knowledge for themselves and their own lives. For example, referring to the intervention session on languages in the school, Camila said:

When I’m walking down the corridor you’re always hearing different people speaking their language or they’re speaking English. And I never really, like, stopped and thought about it […] But after this lesson I was like, it kind of opened my eyes […] Like, it made me actually think, and that’s why I really like these lessons because I could reflect also.

Similarly, Adam commented that he enjoyed ‘learning and thinking about myself as a language learner’ and Helena mentioned that the reflection tasks during the sessions ‘make you realise what you are’. The reflexive and agentive way in which these learners engaged with the intervention seems to have been crucial in contributing to the positive shift in their multilingual identity.

Discussion

The primary aim of this study was to explore how students develop their understandings of multilingualism and various dimensions of their own multilingual identity during an identity-based pedagogical intervention in the languages classroom. Through analysis of pre- and post-intervention interviews with 14 students the following three profiles of development were identified: resistant multilingual identity development (associated with a limited and rather deterministic understanding of multilingualism), emergent multilingual identity development (associated with raised awareness about multilingualism) and reflexive multilingual identity development (associated with reflexively engaging with the ways in which languages relate to their own identity and self-constructs). The key theoretical and pedagogical insights from the results are discussed below.

The first key finding of note is that for the majority of students (i.e. those in the reflexive group and, to a lesser extent, the emergent group), the intervention had a positive effect on both their understandings of multilingualism and their own multilingual identity (in line with the whole-group trends outlined in Forbes et al. 2021). Crucially, this was true both for students who had exposure to languages other than English in the home and those whose only exposure to language learning was in the context of school and, across students from different schools. This adds to the growing body of evidence which highlights the potential for identity-based interventions in the classroom (e.g. Oyserman et al. 2021; Schachter and Rich 2011). Nonetheless, it is also important to acknowledge that for a small group of students in the resistant group the intervention had little or no effect. This acts as a reminder that individuals can both take up and resist identities (Makoe 2014) and, therefore, while we can increase students’ awareness of a range of identity positions in relation to languages, we cannot (and should not) impose such identities on students.

Yet, analysis of the qualitative data presented here also provides further insights into the complex interrelationships between the various dimensions of multilingual identity (i.e. the 3Es) which contributes to our theoretical understanding of this as a construct. Students’ experiences, both in and out of school, emerged as key; across all categories the learners recognised the influence of family background (e.g. speaking a language other than English at home) and past experiences (e.g. spending time in other countries) in shaping multilingualism, yet those in the reflexive (and, to some extent the emergent) categories also crucially acknowledged the influence of school-based experiences of language learning. While teachers, of course, cannot shift students’ out of school exposure to or experiences of languages, it is important therefore to consider which particular dimensions of multilingual identity were influenced as a result of the classroom-based intervention.

The analysis presented above suggests that the intervention had the greatest direct effect on students’ evaluations. Through engaging in the sessions, students demonstrated the potential to progress from viewing multilingualism as something which was determined by external factors and which required fluency in more than one language (in line with Bailey, Parrish, and Pierce 2023), to a broader, more inclusive view of multilingualism as encompassing relatively beginner level language learners and including an understanding of culture. These stronger evaluations were, in turn, closely bound up with emotions. However, it is interesting to note that shifts in emotions (as evident predominantly within the reflexive profile) did not seem to be as directly attributable to the intervention sessions per se, but rather were mediated through students feeling a stronger association with languages as a result both of their shifted evaluations and their experiences of language use. Indeed, a disassociation with languages appeared to be a key barrier to multilingual identity development among the resistant profile which is reminiscent of the pre-service teachers in Iversen’s (2022, 145) study who positioned ‘the multilingual’ ‘as someone radically different from themselves’. For the students in this study, it was only through broadening their understandings of multilingualism to include someone like them that it was possible for them to really ‘feel involved with it’, as noted by Camila. This aligns closely with Fisher et al.’s (2022) finding that multilingual identity is strongly associated with language experiences (which may include classroom-based experiences) and language self (i.e. language self-beliefs and emotions).

This, in turn, has important pedagogical implications for developing students’ understandings of multilingualism and multilingual identity. As shown in the data above, the students for whom the intervention had the greatest effect were those who engaged reflexively with the activities. Reflexivity, which involves ‘questioning our own assumptions and taken-for-granted actions, thinking about where/who we are and where/who we would like to be, challenging our conceptions of reality and exploring new possibilities’ (Cunliffe 2004, 411), therefore emerged as the critical aspect of the intervention. In line with Chaffee (2019, 100), we see reflexivity as ‘not so much a component of identity, but a process that is a driver in its formation and maintenance on a very basic level’ and our results suggest that the emphasis of such identity-based interventions should be on encouraging learners to reflect not only on what multilingualism means but, crucially, on how it relates to them personally. Such reflexive engagement with experiences in the classroom, in turn, has the potential to strengthen students’ emotional connections with languages and to empower them (if they wish) to claim a multilingual identity.

Conclusion

This paper sought to explore how adolescents develop their understandings of multilingualism and various dimensions of their own multilingual identity through an identity-based pedagogical intervention. Using a qualitative approach, three profiles of development were identified: resistant, emergent and reflexive multilingual identity development. It is important to acknowledge that this study involved only a small group of 14 students from four heterogeneous schools and focused only on data from pre- and post-intervention interviews. The data, therefore, does not necessarily capture the nuances of how students engaged with particular intervention sessions and activities and how this may have been influenced by other factors such as the teacher, wider school context, the family and community. However, the results nonetheless have valuable implications. Theoretically, they provide new insights into the way in which classroom-based experiences of language learning can shape and broaden students’ understandings of multilingualism and of themselves as multilinguals which, in turn, can facilitate a stronger emotional connection with languages. Pedagogically, the data highlight reflexivity as a crucial component in identity-based interventions.

Acknowledgements

The authors would like to thank the Arts and Humanities Research Council, UK, for its support in funding this project under the auspices of the Open World Research Initiative. We are very grateful to the teachers involved in this study for their enthusiasm in engaging with the intervention and to the students for giving up their time to speak to us and for sharing their experiences. We also thank the reviewers for their helpful comments and suggestions.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Data availability statement

Data not available due to ethical restrictions.

Additional information

Funding

This work was supported by Arts and Humanities Research Council: [Grant Number AH/N004671/1].