Towards multilingual competence: examining beliefs and agency in first year university students’ language learner biographies

ABSTRACT

As working life across the world is increasingly multilingual, multicultural and multidisciplinary, higher education language teaching is faced with a challenge of how to prepare students for it. Many universities have recently developed multilingual pedagogies but central to their success is learners’ perceptions of these practices. To fill this gap, this article explores first year university students’ language learner biographies to gain insight into how learners construct their linguistic realities. The biographies were studied with discourse analytical methods to examine the participants’ beliefs about language learning and their sense of agency in it. The results reveal that participants tend not to portray themselves as agentive learners but rather position agency as an outside factor affecting their language learning. In addition, the participants’ discourses differed in terms of languages, which adds to previous studies noting the complexity of learner beliefs. Understanding these dynamics can help teachers address students’ beliefs in the classroom and accordingly support learners to create meaning in their language learning. The findings further suggest that beliefs should be studied from multilingual perspectives also in the future.

KEYWORDS:

• Beliefs
• agency
• higher education
• language learning
• discursive

1. Introduction

Globalisation and new technologies have led to a situation where foreign languages and communication have an increasingly important role in the labour market and hence language learners across the world operate in increasingly transnational and superdiverse contexts (Douglas Fir Group 2016; Gunnarsson 2014; OECD 2018). Yet, research in multilingualism suggests that the prevalence of L2 English in a country may negatively affect the interest to learn languages other than English (LOTEs) (Busse 2017; Henry 2010) meaning that the expectations of the evolving working life and the competencies of the labour force are in many countries incompatible. To support learners’ multilingual development and answer to the demands of the rapidly evolving knowledge economy, there is a need to develop multilingual pedagogies in higher education (HE). Policy papers and strategic documents on different societal levels mention terms such as multilingual competence (The Council of the European Union 2018), twenty-first century skills (OECD 2018) and global citizenship (UNESCO 2019), all of which are considered central for successful adaptation in changing study and work environments and which are tightly connected to SLA. This article reports on a study conducted in a HE context where multilingualism and global citizenship have recently been added in strategic documents, yet practical pedagogical applications are still under development. A careful scrutiny of pedagogical solutions is needed, as language teaching in university can paradoxically lead to increased monolingual teaching when learners’ exposure to the target language is maximised (Kelly 2015) and ‘international’ often in practise means ‘in English’ (Fabricius, Mortensen, and Haberland 2017). Furthermore, learners’ prior language learning experiences may make it challenging to adjust to a multilingual learning environment.

Policies and pedagogies do not automatically create learning if they fail to recognise the learner’s own meaning-making processes. Learner beliefs and agency have been widely documented to have a significant role in SLA (Barcelos and Kalaja 2011; Kalaja et al. 2016; Yang and Kim 2011), which is why they should be considered when developing meaningful multilingual pedagogies in HE. As previous research shows the complex nature of beliefs (Kalaja, Barcelos, and Aro 2018; Mercer 2011b; see also Busse 2017), it is also not enough to study beliefs about one additional language¹ but in addition the intricate relationship between learners’ beliefs about all of the languages in their lives. Hence, this article will explore first year university students’ discourses to examine how the students make sense of their own relationship with language learning. Its aim is to combine and build on previous discursive belief and agency studies as well as recent multilingualism research to gain insight into how multilingual learners construct their language learning reality. Subsequently, it discusses what this might mean to future HE language teaching that aims to educate global citizens with twenty-first century skills (OECD 2018). The pedagogical objective of the article is to provide new understanding of learner beliefs that could be utilised in HE SLA classes to expand students’ language awareness and reflection skills and thus help them develop their multilingual² repertoires as young professionals entering working life.

2. Literature review

2.1. Developing multilingual pedagogies in higher education

As SLA studies have taken a multilingual turn, learners are perceived as multilingual beings with complex biographies (Douglas Fir Group 2016; Meier 2017). However, while a vast amount of this research concentrates on learners from multilingual homes, learners with one home language but various FLs have received less attention. This article argues that to develop multilingual pedagogies, learners must be recognised as multilingual and their awareness of this should be supported. A Finnish university is an apt setting for exploring this notion.

The context of this study is a Finnish university whose main strategic goals include fostering linguistically and culturally aware and globally responsible academic professionals. Finnish working life is increasingly multidisciplinary, multicultural and multilingual (Lehtonen 2017) and graduates are required to have good communication and language skills as well as abilities to work in international settings. Previous studies suggest that Finnish students in tertiary education generally believe that languages have an important role in working life (Jalkanen and Taalas 2013; Mutta, Lintunen, and Pelttari 2017). Meanwhile, Finnish people study fewer FLs than before (The Matriculation Examination Board 2016). Accordingly, there is a mismatch between the generally positive views on languages and the languages actually studied. Research in multilingualism suggests that the prevalence of English in Europe may negatively affect the interest to learn LOTEs (Busse 2017; Henry 2010) and in Finland this seems to be the case both with the mandatory Swedish and optional L3s. It must also be noted that although this article focuses on work and study contexts, developing multilingual pedagogies has a vital role in fostering identities, cultures and symbolic values. Multilingual competence should be seen ‘as a life-long capital and reservoir of co-ordinate experiences’ and can thus be seen as essential in democratic societies (Stratilaki-Klein 2019).

Finnish university students have some compulsory language studies in their degrees regardless of their major subject. These studies include field-specific academic literacy and communication skills in their mother tongue, the second national language and an L2, which for most students are Finnish, Swedish and English, respectively. The institution offering these courses in the present study has since 2013 developed multilingual communication and language courses where the focus is on multilingual competencies rather than individual languages. The courses force the students to reconsider their beliefs about language learning as they are faced with a new multilingual ideology they are not familiar with from school. This has posed a question of how the students adapt to the multilingual learning environment when the beginning of HE studies can be demanding as it is (Ketonen et al. 2016). Accordingly, a reflective task was given to new freshmen in this programme to gain insight into the students’ beliefs. The study was conducted by the teacher-researcher, who was a member of the teacher team.

2.2. Beliefs about language learning and agency as discursively constructed

As presented above, various measures have been taken to support learners’ linguistic competencies in the changing working life. However, a new policy or method is not enough since students may be hindered by their beliefs and they may resist a pedagogical change. Students’ relationship with language learning must therefore be examined and addressed in SLA teaching and research. This article explores this relationship through the concepts of learner beliefs and agency that are understood as discursively constructed and therefore detectable in students’ discourses. The focus is on learners’ beliefs about language learning and their own agency in it.

Drawing from contextual approaches to L2 learner beliefs (De Costa 2011; Kalaja, Barcelos, and Aro 2018), this article defines beliefs as ‘conceptions, ideas, and opinions’ about language learning that are socially and discursively constructed by the learners (Kalaja, Barcelos, and Aro 2018). As opposed to traditional etic approaches, this understanding of beliefs has moved from cognition to observing beliefs in students’ discourses arguing that beliefs are not stable entities in a person’s mind but rather constructed in discourse (Kalaja, Barcelos, and Aro 2018). Beliefs have been shown to significantly influence language learning (e.g. Barcelos and Kalaja 2011; Kalaja and Barcelos forthcoming; Kalaja et al. 2016; Yang and Kim 2011). They are complex, situated and dynamic, and they are shared in a specific context (Aragão 2011; Kalaja, Barcelos, and Aro 2018; Mercer 2011b). Previous studies also suggest that university students’ beliefs are already rather stable yet malleable (Aragão 2011; Ohata and Fukao 2014; Yang and Kim 2011). Methodologies investigating beliefs discursively include sentence completion tasks, drawings and narratives (Kalaja et al. 2019).

Beliefs are closely connected to agency, which Ahearn (2001: 112) defines as ‘the socio-culturally mediated capacity to act’. Of the various approaches to agency, this article draws from positioning theory and poststructuralist perspectives that highlight the authorship and positioning of the writer or speaker of a discourse (Davies 2000; Kayi-Aydar 2019). In this perspective, agency is a situated and dynamic discursive position that the writer or speaker may assign on themselves or others, or refuse to do so (Kayi-Aydar 2019; Warren, 2019) and thus the discursive I is a subject position available to the individual, not a personality trait (Davies 2000). Davies (2000) also discusses agency with the notion of authorship, which agentive individuals can take up in discourse to shape, resist or challenge subject positions available to them.

Previous SLA research suggests that agency shapes and is shaped by beliefs (Gao 2010; Kalaja 2016). Positioning theory can explore how learners position themselves in discourse and how these positionings affect their learner beliefs (De Costa 2011) and beliefs about agency (İnözü 2018). Since positioning theory is rather a theoretical framework than a method, previous studies have utilised various methodologies. As agency is understood here as discursive position, it is justifiable to explore with discourse analytic tools (Edwards and Potter 2005; Potter 1996; Potter and Wetherell 1987; Wetherell and Potter 1992). The two key concepts from discourse analysis in this study are interpretative repertoires, which are different meaning-making systems utilised by the learners, and subject positions, through which learners portray agency in their narratives. These can be detected in students’ discourses and they make it possible to examine students’ beliefs about language learning and agency.

3. The study

This article reports on a study examining Finnish first year university students’ language learner beliefs and sense of agency upon entering university. The data was collected within the students’ first multilingual language and communication course at the beginning of the first semester.

3.1. Research questions

The study was guided by the following two research questions:

1. Which interpretative repertoires do the participants use to describe factors that have affected their learning of different languages (EFL, SFL³ and L3)?

2. Which subject positions are adopted in the repertoires?

3.2. Data collection

The data consisted of language learner biographies (e.g. Williams, Mercer, and Ryan 2015) which were collected as a course assignment and where the students reflected on their past language learning (Aragão 2011). Although everyone has their individual life experiences which affect their beliefs (Mercer 2011a), the participants were chosen based on their similar age and educational background to minimise variability in terms of these factors. Therefore, students with previous HE studies or gap years were ruled out from this study.

The teacher-researcher asked the students for a permission to use their course work for research purposes but it was highlighted that this would not affect their course assessment and that their work would be used anonymously. The biographies were written for a compulsory course, which could affect the responses to a certain degree. However, the students were encouraged to write honestly about their feelings and the only assessment criterion was that the text had to respond to task instructions. It is also important to think about the role of the teacher-researcher, who must critically assess their interpretations during the research process and take responsibility of how they analyse and describe the data (Johansson 2005). In order to minimise the bias of the teacher-researcher, the chosen 14 texts were anonymised.

An autobiographical account is a retrospective reflection in which the speaker constructs events by organising and interpreting their memories (Biesta and Tedder 2007). Hence, through narratives writers are able to identify and reflect on their beliefs (Aragão 2011) and agency (Biesta and Tedder 2007). This reflective tool can therefore be both a pedagogical tool and a research method: the task can foster the development of the students’ meta-cognitive skills (Williams, Mercer, and Ryan 2015) and the teachers can learn about their students’ beliefs and sense of agency. It also allows the researcher to analyse these on a discursive level.

3.3. Methods of analysis

3.3.1. Analysing beliefs and agency discursively

Drawing from positioning theory and the discursive turn in learner beliefs research, this article understands learner beliefs and agency as discursively constructed (Davies and Harré 1990; De Costa 2011; Kalaja, Barcelos, and Aro 2018). It examines them with tools originating in discursive psychology (DP) (Edwards and Potter 2005; Potter 1996; Potter and Wetherell 1987; Wetherell and Potter 1992) which have been utilised in some SLA studies (Barkhuizen 2010; De Costa 2011; Kalaja et al. 2016). DP examines how phenomena such as beliefs and agency are constructed ‘through descriptions of actions, events, objects, persons and settings’ (Edwards and Potter 2005: 242). Descriptions can be analysed in terms of how they are constructed and what is achieved by them (Potter 1996, 98). Hence, DP is apt for examining how students form their beliefs and position themselves as agentive learners in discourse. Two key concepts from DP in this article are interpretative repertoires and subject positions (Edley 2001; Potter and Wetherell 1987; Wetherell and Potter 1992).

Interpretative repertoires are relatively complete systems of meaning making which are formulated in social practices and take part in constructing social reality (Edley 2001; Jokinen, Juhila, and E Suoninen 2016). They can be identified by the common vocabulary or register of metaphors that are used to describe events or actions in discourse (Potter and Wetherell 1987). Wetherell and Potter (1992) argue that people use various interpretative repertoires to explain and justify their claims as well as disclaim other arguments. In other words, it is possible to detect different interpretative repertoires in one individual’s accounts by analysing patterns of content and form. When studying interpretative repertoires, the aim is not to find consensus but to examine the variability of language use in different situations, and the focus is solely on discourse rather than the speakers’ cognition (Potter and Wetherell 1987).

A subject position is ‘a “part” allocated to a person by the use of a story’ (Stenner 1993: 114). Detecting subject positions is a way to examine agency in the text, as agency is a discursive position assigned by the writer (Kayi-Aydar 2019). By analysing positions it is possible to examine the meaning of a narrative as opposed to merely concentrating on its content (Barkhuizen 2010). The focus is then on the effects of a story rather than whether the story is ‘true’ (Stenner 1993). In this article the interest lies particularly on learners’ ‘intentional self-positioning’ (De Costa 2011: 349), in other words, what kinds of positions the participants adopt in the discourse when it comes to agentive behaviour and taking responsibility in language learning.

3.3.2. Analysis process

Phase 1 of the analysis consisted of coding the 14 texts first in in-vivo style (Saldaña 2016) and then several rounds of close reading to detect interpretative repertoires in the data. The preliminary codes were read and re-read in order to find differences and consistencies in the texts (Potter and Wetherell 1987). In addition, each paragraph was coded in terms of which language it described. Based on the initial coding, three different repertoires about language learning were identified and the analysis was tested on the data. After the testing, a fourth repertoire was also detected.

Phase 2 focused on subject positions to examine ways in which the discursive I is constructed in the discourse. As Harré (1995: 134) writes:

there is fundamental difference in discursive role between those acts in which I, the speaker, take responsibility for my actions and those in which I give reasons for them. Giving reasons for an action may be a way of disclaiming responsibility (–).

Harré’s suggestion about the differences in responsibility proved to be a useful starting point for the analysis of agency in the data. The action and agent was coded in each text excerpt, which resulted in the emergence of two main subject positions.

4. Results and discussion

4.1. Overview of the results

In line with Wetherell and Potter (1992), the participants flexibly drew from different interpretative repertoires and there were evident differences and consistencies in the repertoires. Four interpretative repertoires and two subject positions emerged from the data. The repertoires were named the affordance, affection, attribute and action repertoire and will be described and discussed in Section 4.2. A numerical overview of the data is presented in Table 1 and elaborated in the subsequent sections.

Table 1. Numerical overview of the results.

Repertoire	Position	Featured language
		English	Swedish	L3	Total
Affordance	Dependent agent	24	14	15	53
Affection		15	15	8	38
Attribute		8	3	4	15
Action	Independent agent	10	5	4	19
Total		57	37	31	125

4.1.1. Subject positions

Following Harré’s (1995) notion of responsibility, two subject positions were detected in the data. The participants positioned themselves differently within the interpretative repertoires, either presenting themselves as active agents (independent agent) or placing agency on other factors than their own actions (dependent agent). The participants adopted the independent agent position in the action repertoire and the dependent position in the three others. The texts portraying a dependent agent describe an outside cause that triggers action in the learner. In the text the account starts with a cause which is followed by the triggered action. The texts portraying the independent agent explicitly state that learning was caused by the learner and hence the participants position themselves as active agents (see also Ruohotie-Lyhty 2011).

4.1.2. Languages

Learners are complex multilingual beings and learners’ perceptions of themselves as language users are language and context dependent (Busse 2017; Douglas Fir Group 2016). Accordingly, contrary to previous studies, the participants wrote about all languages in their lives. Once the data was categorised in terms of interpretative repertoires, it was possible to count the number of repertoires per language. The results are presented in Table 1. It must be noted that a quantitative analysis of the data would not be statistically valid considering the number of participants in the study. However, this numerical overview can provide insight into the usage of different discourses and agency in the language biographies and the differences in terms of languages. The next section describes the different interpretative repertoires in detail and discusses how languages feature in the stories.

4.2. Repertoires, positioning and language

4.2.1. Affordance repertoire

The affordance repertoire was the most prominent repertoire in the data. In this repertoire the participants describe their language learning in relation to affordances in their environment (Peng 2011). These affordances include both out-of-school situations (people who speak the target language, the internet) and institutional structures (teachers, timetables). The passages mention an affordance which is followed by an evaluation of its effect. By placing agency on the affordance, the learners could not be credited or blamed for their successes or failures. The passages employ different extrematisation and minimisation techniques (Potter 1996) to explain the impact of the affordance in question. Three descriptive examples from the data are presented and discussed below.

Example 1

My new Swedish teacher was the best thing that could have happened to my learning: Now I was happy to study Swedish and I even started learning it. (S1P8)⁴

Example 2

However, I had to quit studying German in first year [of upper secondary school] because of too many other courses. (S3P4)

Examples 1 and 2 describe school-related affordances that have affected the writers’ language learning. Learning is to some degree outsourced to this affordance as the writers blame or give credit to it, portraying a picture of themselves as dependent agents. The writers use their authorship (Davies 2000) to shape this subject position. Example 2 illustrates also the way the writer can resist a position (Kayi-Aydar 2019; Warren 2019), in this case that of an independent agent (active learner) by using the modal verb of necessity ‘to have to’. By doing this, the writer disclaims the agentive position available to them to take charge of their learning and continuing the German studies. The same repertoire is also used about out-of-school affordances, such as in example 3.

Example 3

I was first introduced to English before pre-school by computer games but it started properly when I was eight years old, when my neighbour, who was three years older, introduced me to the online video game Runescape which was of course in English. I played this game many days a week (almost daily) almost throughout primary school and my language skills became pretty good even before English teaching started in third grade. (S2P7)

As in examples 1 and 2, the writer in example 3 portrays themselves as learner dependent on the affordance. Although an active agent in the playing, the writer simply states that ‘my language skills became pretty good’, thus adopting no active position in the development of their language skills.

As explained above, this repertoire included both out-of-school and institutional affordances. For the purposes of this study it was justifiable to treat them as one repertoire since the repertoire was used to give reasons for language learning, in this case an affordance. However, to discuss how different languages feature in this repertoire, it is worth examining the content of these stories. When this repertoire was used with English, it most often described out-of-school affordances whereas with Swedish and L3 it was almost solely about institutional affordances (see e.g. Busse 2017). This is an expected finding since most television shows, music and social media sites Finnish people watch and use are in English or Finnish. The stories related to Swedish focused on teachers and their niceness or expertise; a few participants also mentioned the usefulness or uselessness of Swedish in the area where they grew up. The L3 stories adopting the affordance repertoire were mainly about reasons for why the students chose an L3 or why they dropped it later. There tended to be other subjects that were more worthwhile, and hence although affective reasons were sometimes the factors that got the students to start studying an L3, over time the L3 became just another school subject that they did not feel that strongly about.

4.2.2. Affection repertoire

This repertoire includes stories related to affective and emotional factors which are seen to influence language learning both positively and negatively. It has been well documented that affective factors have an important role in language learning (Aragão 2011; Williams, Mercer, and Ryan 2015) and it was therefore unsurprising that the affection repertoire was the second-most frequent repertoire. Affect is here a phenomenon outside of the learner and one the learner cannot influence. The passages include affectively charged verbs (‘liked’, ‘wanted’), nouns (‘favourite subject’) and expressions combining affectively charged adjectives, nouns and verbs (e.g. ‘my interest declined’, ‘bad attitude’) (Golombek and Doran 2014). Examples 4 and 5 illustrate the way the learning is tightly connected to affective factors.

Example 4

I suffered from a continuous lack of motivation —. So I decided to give up my French studies in the second year of upper secondary school. (S8P9)

Example 5

Swedish started in seventh grade. I’ve liked that from the very beginning and hence my grades were 10 or 9⁵ all the way through secondary and upper secondary school. (S14P3)

The writer in example 4 uses the lack of motivation almost as an illness (‘I suffered’) and the only thing they could do was to quit studying it. Such a dramatic word choice as suffering is a form of extrematisation (Potter 1996) and the aim is to emphasise the effect of affective reasons. In the data there are various similar illness-like states such as attitude problems and lack of interest. In example 5 the writer draws a connection between enjoyment and success in language learning. In the data, interest, motivation or enjoyment is the reason the learner succeeds. In the negative stories such as example 4, the writers disclaim an active agentive position in the discourse (Davies 2000; Kayi-Aydar 2019). Hence, the blame or the credit is shifted to an outside factor, which in this case is in fact in the learner’s own mind. It is noteworthy that these internal affective factors are positioned as something a learner cannot influence and accordingly the learner again adopts the subject position of a dependent agent.

In terms of languages, the stories using the affection repertoire were more about (lack of) confidence to use English rather than about the language itself, which is in contrast with the stories featuring Swedish. Almost half of all the stories related to Swedish used the affection repertoire, which could reflect its often perceived, emotionally charged status. As was noted in the previous section, the L3 stories often adopted the affection repertoire when the participants described the beginning of their language learning; however, lack of interest or institutional affordances many times later ended the learning.

4.2.3. Attribute repertoire

The third repertoire with the learner positioned as a dependent agent is the attribute repertoire. Here the participants describe language learning and competence as a trait of an individual. The participants portray their competence as something they cannot influence; they either have or do not have the skills. Language is described as a challenge (Kalaja 2016). The passages use verbs and nouns related to competence (‘was able to’, ‘challenge’), utterances related to grades (‘full points’) and phrases describing the contingency of success (‘surprise’, ‘luck’). The static nature of language competence is achieved by normalisation techniques (Potter 1996) such as by verbs with an iterative aspect. Metaphorical expressions were common in this repertoire (Potter and Wetherell 1987). Extrematisation (Potter 1996) in the passages highlights the easiness or difficultness of learning (‘quite a head for languages’, ‘a major Achilles heel’). Examples 6 and 7 illustrate this repertoire.

Example 6

— it turned out that I had quite a head for languages and studying German was as easy as pie at the beginning and in fourth and fifth class I got mostly 10s as grades from tests. (S2P9)

Example 7

Also Swedish started in upper secondary school, which turned out to be a major Achilles heel in my language skills. The problem wasn’t a lack of motivation because I wanted to learn Swedish, but somehow it just never stuck to my head. (S12P2)

Also in the attribute repertoire, the writers positioned themselves as dependent agents, disclaiming the effect of their own doings (Davies 2000). Example 6 depicts a ‘head for languages’ that was prominent in the data: the writer suggests that their success in language class is based on this inherent trait. The writer in example 7 contrasts affective factors with skills: that motivation did not help when they simply were not able to learn.

The attribute repertoire was the least used in the data, so it seems that the participants do not hold a strong belief that language competence is an unchangeable state. This repertoire was most often used with English and the stories stated either that the learner was or was not good at it, without pondering reasons behind this. Swedish and L3 competences were more strongly attributed to external affordances or affective factors, although here too the competence was ultimately a trait: for example, in Example 7 the writer acknowledges that affective factors may have a role in language learning, but that the problem lay on the writer’s own skills.

4.2.4. Action repertoire

The action repertoire differs from the three others by the writers’ positioning as active agents. Here the participants describe their own actions as the factor that has affected their learning. Such actions are language learning, inactive behaviour and practice. The passages include verbs expressing purpose or intention (‘decided to go’, ‘started’) and phrases about doing work (‘my efforts’, ‘would require work’), and the subject is the learner themselves. Extrematisation and minimisation (Potter 1996) are used to highlight the agent’s efforts. Examples 8 and 9 describe the blame and credit the students give to their own actions.

Example 8

I didn’t study English with as high motivation, and the courses in upper secondary school passed by with me just sitting in class, which annoys me now. During a possible gap year I was going to study English to make it as good as possible but instead I got straight into university. (S3P4)

Example 9

My efforts and hard work paid off in the matriculation exams. I got good grades for all languages, Swedish, English and mother tongue, and partly because of that got into university. (S13P6)

The writers in examples 8 and 9 position themselves as active, independent agents responsible for their learning (Davies 2000). In example 8 the writer implies that they should have been a more active student, stating that they were merely ‘sitting in class’ rather than doing work. In contrast, the writer in example 9 is happy with their efforts that ‘paid off’. Example 10 further analyses a learner’s own actions.

Example 10

Towards the end of upper secondary school I realised that although I’ve used English all the time [in my free time], the lack of speaking it has been slightly detrimental to my pronunciation. (S5P7)

Here the writer notes that their own actions have affected their competence. Although the affordance repertoire shows that a great deal of credit in successful learning is given to out-of-school activities, free time featured only a handful of times in the action repertoire. Even here it becomes apparent that English is learned or not learned due to informal learning situations whereas the other languages are rarely met outside of school walls. The lack of free time related stories could be an indication of a belief that language learning is either actively achieved in class or passively acquired elsewhere. Overall, the action repertoire and the independent agent’s subject position features in only 19 times in the 125 stories. Accordingly, in all the rest of the passages agency is placed on something else than the learners themselves (cf. Williams et al. 2004). The result poses the question of what makes the participants in the present study refuse the role of an independent learner.

5. Conclusions

The aim of this article has been to construct knowledge on multilingual language learner beliefs and subsequently help further develop multilingual communication and language teaching in HE. The participants used four interpretative repertoires to discuss their language learning and two subject positions to express their sense of agency. Although a relatively small-scale qualitative study, it provides an insight into the discourses of young adult learners studying multiple languages who will enter a multilingual working life. By identifying and understanding learner beliefs, the teacher can better support learners in meaningful language learning.

A major finding of the study was that the number of accounts portraying the learner as an independent agent was very low. Accordingly, it seems that the participants would rather blame or give credit to other factors than their own effort. Even affective factors were portrayed as something learners cannot influence, as if they were something external. Joy, dislike or motivation were equated with other ‘inborn’ aptitudes for learning, such a as an ear for languages. Perhaps it is easier for the learner to portray an internal factor as an external one since this gives them a valid reason for disclaiming responsibility in their learning. This is worth addressing in classroom reflection but it also merits further studies. Larger, longitudinal and possible quantitative studies on university students’ beliefs could shed more light into the matter. An important question is also whether similar results would emerge in different contexts, or whether this is unique to Finnish learners (cf. Williams et al. 2004).

This study differed from previous beliefs research by examining all languages in the students’ repertoires. The participants spoke about languages in different ways, which is related to recent multilingual studies on motivation and the self (Busse 2017; Henry 2010, 2017; Mercer 2011a). The ways in which different languages were discussed reflect well on the general mindsets, the Finnish school system and the linguistic landscape of the participants (see also Busse 2017). Learner beliefs have long been characterised as complex, dynamic and even contradictory (Kalaja, Barcelos, and Aro 2018; Mercer 2011b) and when taking a multilingual perspective, the picture is if anything more complex. Further studies on the matter could explore whether beliefs are more tied to individual languages or whether some are more general.

Pedagogically, the narrative accounts involved interesting reflection and the results could be utilised in future teaching, which was one of the aims of the study design. Apart from setting the language biography task for future students, the results of this study could be brought to discussion in class. The students could even analyse their own biographies from the point of view of these interpretative repertoires and discuss their beliefs and their sense of agency in language learning. Such a theoretical perspective can be used with university students and it could lead to meaningful, reflective conversations in class (Kalaja and Barcelos Forthcoming). Drawing from Biesta and Tedder (2007: 139), it is perhaps questionable whether understanding one’s beliefs and sense of agency is enough to trigger (a change in) action or whether ‘it is change in people’s lives that will actually lead to insight and understanding’, but in the light of previous studies it seems that reflection can be of use (Aragão 2011; Ruohotie-Lyhty 2011). Although not all teachers have resources to conduct their own research, discussing students’ stories, feelings and experiences about language learning could lead to important and even empowering insights.

It seems clear that more research on these issues is needed in order to improve SLA teaching in European HE contexts. The complexity of learner beliefs and agency together with the multilingual turn in SLA requires cooperation from language educators and researchers. Together they can foster the development of young adults’ multilingual repertoires by supporting meaningful language learning and subsequently ensure that when entering working life, these learners have the competence and study skills that is required of them.

Disclosure statement

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

Notes

1 The term ‘additional languages’ can be used when talking about both second and foreign languages.

2 Some European literature distinguishes between the terms multilingualism and plurilingualism, where plurilingual competence is that of an individual whereas multilingualism concerns wider communities and societies (e.g. Kelly 2015).

3 Swedish as a foreign language.

4 The data excerpts are translated from Finnish. The original excerpts are listed in Appendix A.2.#.

5 In Finnish schools the grading scale goes from 4 to 10, meaning that 9 and 10 are the highest grades.

Appendix

A.1.#Language learner biography instructions

A.1.1.#Translation of the task (original in Finnish below)

Draw a timeline that starts from your birth year and continues to year 2030.

Mark on the timeline all the events and time periods that have to do with languages or language competencies from your birth year to year 2017 (languages are both your mother tongue and foreign languages).

Choose those events and time periods that best describe how different languages relate to your life. They can be short moments or years-long time periods and they can have happened anywhere in your life (free time, school, etc.) or even just inside your head.

Write about a 600-word-long text in which you concisely describe the events above. What happened / what did you think or feel? How do you think they relate to your relationship with languages?

Now look at the timeline from 2017 to 2030. Mark on the timeline plans / thoughts / wishes about what you think will happen next with you and languages. What kinds of communication and language skills do you think you will learn during your studies? What kind of communication and language competence do you think you will need in the future in different parts of your life?

Continue the previous text for another 200 words where you describe your thoughts about the future.

Return the assignment in Word-format on Moodle by Thu 28.9.2017. Also take a photo of the timeline and save in to Moodle.

A.1.2.#Original instructions in Finnish

Piirrä aikajana, joka alkaa syntymävuodestasi ja jatkuu vuoteen 2030.

Sijoita aikajanalle (syntymästä vuoteen 2017) elämäsi tapahtumat ja ajanjaksot, jotka liittyvät kieliin ja kielten osaamiseen (kieliä ovat sekä äidinkieli että vieraat kielet).

Valitse ne tapahtumat ja ajanjaksot, jotka mielestäsi parhaiten kertovat siitä, miten eri kielet liittyvät sinun elämääsi. Ne voivat olla lyhyitä hetkiä tai vuosia kestäneitä kausia ja ne ovat voineet tapahtua missä vain elämässäsi (vapaa-ajalla, koulussa, jne.) tai vaikka vaan pääsi sisällä.

Kirjoita noin 600 sanan teksti, jossa kuvailet tiivisti yllämainittuja tapahtumia. Mitä tapahtui / ajattelit / tunsit? Miten ne mielestäsi liittyvät siihen, millainen suhteesi on kieliin?

Katso nyt ajanjaksoa 2017–2030. Merkitse aikajanalle suunnitelmia / ajatuksia / toiveita siitä, mitä sinun ja kielten suhteen tapahtuu seuraavaksi. Millaisia viestintä- ja kielitaitoja arvelet oppivasi opintojesi aikana? Millaista viestintä- ja kieliosaamista arvelet tarvitsevasi tulevaisuudessa elämän eri osa-alueilla?

Jatka 4. kohdan tekstiä vielä noin 200 sanan verran, jossa kuvailet 5. kohdan pohdintojasi tulevaisuudesta.

Palauta tehtävä Word-muodossa Moodleen to 28.9.2017 mennessä. Ota myös kuva aikajanasta ja tallenna se Moodleen.

A.2.#Data excerpts in Finnish

Example 1

Uusi ruotsin opettajani oli oppimisen kannalta paras mahdollinen asia: Opiskelin nyt mielelläni ruotsia ja aloin jopa oppimaan. (S1P8)

Example 2

Jouduin kuitenkin lopettamaan saksan lukemisen [lukion] ensimmäisellä luokalla liian suuren kurssimäärän takia. (S3P4)

Example 3

Englannin kieleen sain ensikosketukseni jo ennen esikoulua tietokonepeleistä mutta toden teolla se alkoi --- ollessani 8-vuotias, kun kolme vuotta vanhempi naapurini tutustutti minut, tietysti englanninkieliseen, netissä pelattavaan videopeliin nimeltä Runescape. Tätä peliä pelasin useana päivänä viikossa (lähes päivittäin) lähes koko ala-asteen ajan ja kielitaitoni kasvoi melko hyväksi jo ennen englannin opetuksen alkamista kolmannen luokan alussa. (S2P7)

Example 4

Kärsin jatkuvasta motivaatiopulasta ---. Päädyinkin jättämään ranskanopintoni sikseen toisena lukiovuonna. (S8P9)

Example 5

Seitsemännellä luokalla alkoi ruotsi. Siitäkin olen pitänyt kovasti heti alusta alkaen ja arvosanani olivatkin aina 10 tai 9 koko yläasteen ja lukion ajan. (S14P3)

Example 6

--- selvisi että minulla on suhteellisen hyvä kielipää ja saksan opiskelu sujui alussa kuin vettä vaan ja sain nelos- ja vitosluokan kokeista useammin kymppejä kuin mitään muita numeroita. (S2P9)

Example 7

Yläkoulussa alkoi myös ruotsin kielen opiskelu, mikä osoittautui valtavaksi akilleenkantapääksi kielten osaamiseni kannalta. Motivaatiosta itselläni ei ollut puutetta, sillä halusin oppia ruotsia, mutta jotenkin oppi ei vain koskaan jäänyt päähän. (S12P2)

Example 8

Englantia en lukenut yhtä motivoituneesti, ja lukion kurssit menivät tunneilla istuessa, mikä jälkikäteen harmittaa. Mahdollisena välivuotena minun pitikin opiskella englanti mahdollisimman hyväksi mutta pääsinkin lukiosta suoraan yliopistoon. (S3P4)

Example 9

Ponnisteluni ja ahkerointini palkittiin kielten osalla ylioppilaskokeissa. Kirjoitin kaikista kielistä, ruotsista, englannista ja äidinkielestä kiitettävät arvosanat, ja osin niiden avulla pääsin tänne yliopistoon opiskelemaan. (S13P6)

Example 10

Lukion loppupuolella huomasin, että vaikka englantia olenkin kaiken aikaa paljon käyttänyt, niin puhumisen jääminen vähemmälle on ollut jokseenkin haitallista kielen lausumiselle. (S5P7)