‘Having a decent understanding of more than one language’: exploring multilingualism with secondary school students in England

ABSTRACT

This study examined the linguistic lives, as well as beliefs about multilingualism, of 422 secondary school students (ages 11–16) in Lincolnshire, England. The data were collected using questionnaires, replicating Haukås et al.’s (Haukås, Å, A. Storto, and I. Tiurikova. 2021. “Developing and Validating a Questionnaire on Young Learners’ Multilingualism and Multilingual Identity.” The Language Learning Journal 49 (4): 404–419; Haukås, Å, A. Storto, and I. Tiurikova. 2022. “School Students’ Beliefs About the Benefits of Multilingualism.” Journal of Multilingual and Multicultural Development, doi:10.1080/01434632.2022.2075001.) work. Whilst the sample was predominantly monolingual by typical, often binary, categorisations of language use (e.g. first language), we also found many other instances of language use that contributed to their linguistic identity. The students held divergent views about multilingualism, as in Haukås et al.'s (Haukås, Å, A. Storto, and I. Tiurikova. 2022. “School Students’ Beliefs About the Benefits of Multilingualism.” Journal of Multilingual and Multicultural Development, doi:10.1080/01434632.2022.2075001.) study, yet these were less positive than in the original study. Predominantly, the students did not identify as multilingual despite studying a language at school, and felt this implied a linguistic proficiency beyond their own. Statistical analysis of group differences found that where they had exposure to languages in their personal lives (e.g. friends and family) they were more likely to feel positively about the benefits of multilingualism and its role in their future lives. The study demonstrates that even seemingly ‘monolingual’ students are exposed to a number of languages in, sometimes small, varied, ways and we propose that schools hold great, often untapped, potential as sites of exposure to language(s) and the development of multilingual identities.

KEYWORDS:

• Multilingualism
• multilingual identity
• education
• foreign languages
• secondary school students

England and multilingualism

England is a country with no official language or national language policy. It is home to speakers of a diverse range of languages, including amongst its school population. 8.9% of the country's population were recorded as having a main language other than English (or Welsh in Wales) in the 2021 census (ONS 2022) and in the 2021/22 academic year, 19.5% of school children are known or were believed to use English as an Additional Language (EAL) (DfE 2022). We can expect many more young people to use and know languages other than English than this number suggests, too, due to the broad and binary categorisation of EAL in government data (Cunningham 2019). We can also expect many young people to live and study in areas where most households will only speak English (Strand, Malmberg, and Hall 2015). Schools in such areas can operate as ‘monolingual bubbles’ (Lanvers, Hultgren, and Gayton 2016, 13), embedded within a monolingual habitus (Gogolin, 1997), where modern foreign languages are the only exposure to languages other than English students will gain. Schools in highly multilingual areas, in contrast, will have a wealth of linguistic experience within their student body and ample potential for creating language-rich environments where multilingualism is seen as a valuable resource (see Cummins 2005; Sridhar 1994; Conteh 2003). The degree to which schools recognise and foster linguistic diversity is not necessarily straightforwardly related to the number of languages in the school or community, but the degree to which bi-and multilingualism are seen as a norm in each is important to recognise. There is, of course, a middle-ground between highly multilingual and highly monolingual areas, too, and we must also acknowledge the increasingly important role of media and the internet as sites of exposure to language(s) (Leppänen and Peuronen 2012). In this paper, we explore multilingualism and multilingual identity amongst learners of modern foreign languages in schools, and the impact of living multilingual lives on the students’ perceptions of the value of languages and their role in their futures.

Following from Haukås et al.'s (2021) work, we focus on multilingual identity, specifically, as defined by Fisher et al. (2020) as a persona; self-identification as multilingual. People's linguistic repertoires are at the centre of this definition. We recognise that multilingualism covers people with a range of proficiency and knowledge of other languages, and may vary in the degree to which their skills are productive (being able to speak or write in two or more languages) and receptive (being able to comprehend two or more languages when written and spoken) (see Haukås, Storto, and Tiurikova 2022), and that people, particularly young people, may consider themselves multilingual, or not, according to their own definitions rather than a common understanding.

Literature review

Schools’ roles in students’ multilingual lives

We start from the position that knowledge of languages beyond your first is important. As well as benefits to the economy of having a multilingual workforce (Foreman-Peck and Wang 2014) and the role of languages in trade (Ayres-Bennett et al. 2022), it is recognised that multilingualism is of benefit to individuals in terms of cognitive development (Bialystok 2009; Monnier et al. 2022), intercultural understanding (see Deutscher 2011; Liddicoat 2013); language awareness (Hawkins 1984; Sierens et al. 2018) and critical multilingual language awareness (García 2017; Hedman and Fisher 2022). This awareness of language is particularly important in terms of supporting minority language speakers to maintain their first language(s) (Cummins 2000; Hélot et al. 2018). Indeed, globally, multilingualism is ‘increasingly perceived as a normal phenomenon and as a positive resource to individuals and societies’ (Haukås, Storto, and Tiurikova 2022, 1). Nevertheless, there exists what Preece (2019) calls an ‘elite bilingualism binary’ between a prestigious form of multilingualism, as practised by ‘highly educated individuals with two or more high status languages learned formally’ (Preece 2019, 406) and providing ‘social and/or material capital, a sense of belonging, prestige, excellence, privilege’ (Barakos and Selleck 2019, 362), and the non-elite forms of multilingualism found in ‘the linguistic repertoires of urban migrant communities using heritage language(s) learned in natural, rather than schooled settings’ (Preece 2019, 406).

In a largely monolingual environment, the main sites of (initial) language learning for many people are schools. This affords schools a critical role in developing students who see themselves as multilingual. Indeed, school-learnt language may also be retained long into a person's life (Schmid 2022), giving it an impact further reaching than the school years. The role of schools is also particularly pertinent in anglophone nations. Language learning in such contexts is conducted with a backdrop of the widening role of English globally, and particularly in spaces such as popular culture, the internet and education (Lo Bianco 2014). This backdrop can contribute to an ‘English is enough’ mindset (Lanvers 2017a) as well as a reduction in exposure to a range of languages in young people's lives.

In England, where this study took place, the teaching of languages other than English is compulsory between the ages of 7 and 14 at the level of national policy. This age range spans primary and secondary school, and most young people are exposed to one or two languages in this time, commonly French or Spanish, with German taught in substantially fewer schools (see Parrish 2023). School-level language teaching is also not without its problems. As has been widely reported (see for example Bowler 2020; Hagger-Vaughan 2016; Parrish and Lanvers 2019), language learning in schools in England faces a range of challenges relating to Brexit and associated political upheaval (Collen 2020; Lanvers et al. 2018; Tinsley 2019), the perceived feminisation of the subject (Parrish 2023), as well as curriculum policy (Coleman, Galaczi, and Astruc 2007; Hagger-Vaughan 2016; Lanvers 2017a) and concerns over the content of the curriculum (Scally, Parrish, and Montgomery 2021; Wingate 2018). Differences in levels of language learning between students of different socio-economic status also exist (Coffey 2018; Lanvers 2017b; 2018; Tinsley and Board 2017b) and students have been shown to struggle to find personal relevance leading to poor take-up of language qualifications amongst students in England (Taylor and Marsden 2014; Parrish and Lanvers 2019).

Understanding more about students who study within the context of the above challenges, their linguistic lives (encompassing all languages they come into contact with, including English), their beliefs about multilingualism and their anticipated use of languages in the future, can allow us to better understand those challenges. Students themselves are, after all, at the centre of the learning experience and it is important we understand the student as linguist. The concept of multilingual identity and its implications for language learning and teaching are under-researched (Haukås, Storto, and Tiurikova 2021), but it is an important relationship for a country where an increasing number of languages are being spoken within its population, while fewer students are learning them at school (Collen 2022). Problems surrounding decreased uptake and motivation in the subject in England ultimately lead to fewer linguistic opportunities for students and less opportunity to develop students’ linguistic repertoires and multilingual identities. We cannot control students’ linguistic experiences outside the classroom; these will vary depending on families, friends and communities. We can, however, use school language lessons to (a) attempt to burst the ‘monolingual bubbles’ of some students (b) challenge an ‘English is enough’ narrative (see Lanvers, 2016) (c) try to interrogate the ‘elite bilingualism binary’ and (d) compliment wider school initiatives to embed linguistic diversity within school life and promote the value of multilingualism more generally (Bailey and Marsden 2017). This has important consequences for monolingual and multilingual children alike as both are afforded the chance to develop their multilingual identities. This led to the following research questions:

1. What do the students’ linguistic lives look like?

2. What are the students’ beliefs about multilingualism: its benefits and its place in their futures?

3. What factors affect the students’ beliefs about multilingualism?

4. What factors affect the students’ views of their future multilingual selves?

Method

In order to investigate students’ multilingual identities and their language learning preferences, we developed an electronic survey based on the Ungspråk questionnaire from Haukås et al.'s (2021) study conducted in Norway. The downward trend of language qualification take-up in England, which is indicative of a broader reluctance to engage with language learning amongst young people (Lanvers, 2016), provides an interesting backdrop to the replication of this questionnaire here. The replication of the Ungspråk questionnaire was embedded within a larger survey examining the secondary school student as linguist.

Questionnaire design

This paper centres on the two main scales from Haukås et al.'s work (2021; 2022): students’ beliefs about multilingualism (BAM scale) and their future multilingual selves (FMS scale). The BAM scale examines students’ beliefs regarding the benefits of multilingualism, for example, for better cognitive functioning (Haukås, Storto, and Tiurikova 2022). The FMS scale examines the students’ future-orientated self-conception as a speaker or user of multiple languages (Haukås, Storto, and Tiurikova 2021). These were both descriptively analysed as well as used to examine group differences by statistically comparing mean scores by group membership including year group, identification as multilingual or not and language learnt at school. Our analysis followed that of the original study (see Haukås, Storto, and Tiurikova 2021).

We also used the questions from the original study relating to participants’ own linguistic experiences which attempts to map their linguistic lives in terms of the languages spoken (including English) and the context in which they are spoken (e.g. a first language, a language used with friends). In order to confirm the suitability of the scales for use in our context, factor analysis with Varimax rotation was conducted on the two scales. Scree plots suggested that they constituted single factors as they did in the original study, and reliability tests did not suggest that deleting any items would increase the Cronbach's alphas (BAM scale = .789, FMS scale = - .828). All items were retained for analysis.

Participants and data collection

The survey was administered online to maximise our access to as many schools and students as possible which was particularly important when collecting data in the period following national lockdowns. The schools were contacted through convenience sampling using networks from the authors’ institutions. 16 schools within one geographical area of England took part in the survey. Participating schools were sent a link to the online questionnaire using their electronic communication systems. Participation was voluntary, anonymous and not actively encouraged or monitored by schools. Students were asked to seek the e-consent of a parent or guardian but undertake the questionnaire alone, in line with ethical guidelines for young people (BERA 2018) and as approved by the authors’ departmental and institutional ethics committees.

In order to focus on the link between multilingual identity and language learning, only students studying a foreign language at school were eligible to take part. This was an important pre-requisite of participation for the study as a whole. 422 secondary-school pupils participated across the 16 schools. We did not ask students to identify their schools in order to ensure the students’ anonymity and increase take-up from the schools, and so it is not possible to say how many came from each. The participants were between 11 and 16 years old. 206 (49%) identified as female, 184 (44%) male and 7 (2%) as non-binary. We collected their school year rather than their age as this is more informative context in relation to their language(s) education. Table 1 shows how many participants were from each year group. The response rate was higher from younger pupils; around half the participants were in the first two-year groups of secondary school (ages 11–13). This has been shown to be a crucial time in determining students lasting attitudes towards the study of language (Taylor and Marsden 2014; Graham et al., 2016).

Table 1. The number and proportion of participants in each school year group.

Year group (age)	Frequency	%
7 (11–12)	106	25.1
8 (12–13)	103	24.4
9 (13–14)	91	21.6
10 (14–15)	75	17.8
11 (15–16)	35	8.3
Prefer not to say	12	2.8
Total	422	100.0

Research area

The geographical area chosen was the county of Lincolnshire. It is the second largest county in England and encompasses both substantial highly monolingual areas, as well as areas with a high concentration of migrants (predominantly from the EU) which have created more multilingual or bilingual communities (Chakelian 2016). Some areas of the county have been identified as failing to adapt to these societal changes with Boston (a small town) being referred to as ‘the most divided place in England’ in national media coverage (Gallagher 2016). It is the juxtaposition of highly monolingual areas and relatively newly multilingual ones which we believe makes the development of young people's multilingual identities especially important to study here. It makes their linguistic lives difficult to predict and, we argue, more representative of a national picture.

Results

The participants’ linguistic lives

School

In school, all students in the study were learning at least one language, as shown in Table 2. The languages they were studying broadly mimicked the pattern seen nationally (Collen 2022), with French the most commonly studied language followed by Spanish. Considerably fewer students were learning German and a small number of students were taking a language outside this ‘big three’ (nationally, figures in 2022 were 41.1%, 35.8%, 11.5% and 11.6% respectively). There were a number of students (40, 9.5%), who were taking two or more languages, which is likely to represent one or two schools whose curriculum allowed for this, as dual linguists are increasingly rare in the school system (Tinsley and Board 2017a).

Table 2. The number and proportion of students studying each language.

Language studied	Frequency	%
French	189	44.8
Spanish	144	34.1
German	27	6.4
Other	8	1.9
Multiple	40	9.5
Prefer not to say	14	3.3
Total	422	100.0

Home

62 participants (15%) told us about a language they used outside of school. For 29 of the 62 (6.9% of the total sample), this language was a first language. 13 different languages were listed as first languages and Polish was the most common within these (n = 13). Arabic, French, Gujarati, Hindi, Irish, Latvian, Lithuanian, Portuguese, Spanish, Tagalog, Thai and Turkish were all listed by one or two participants only. 25 participants told us about a language which was not their first language, but that they use with their family, and 12 used a language (other than their first) with friends. Most commonly (n = 5), this language was Spanish, but languages including German, Tamil and Japanese were also listed. In terms of context-specific language use, the languages listed were most commonly used on holiday (n = 56), rather than on the internet (n = 34) or heard through television or music (n = 46). The participants selected ‘I think I know this language well’ on 50 occasions, for 17 different languages, and most commonly listed Polish (14), followed by Spanish (5).

We also asked about students’ parents and home lives as another key source of language input aside from school. 39% of students did not have friends who knew other languages and 39% did. 22% were not sure. 47% of the students reported that their parent(s) did not speak another language, compared to 36% who reported they did and 16% who were not sure. Of these, 17% reported that their parent(s) spoke more than one language. We asked which languages their parents knew and of the 263 answers given (42 different languages), the most common were French (66), German (48) and Spanish (37), further showing the impact of the ‘big three’. We did not ask students to report which parent spoke each language in an effort to reduce response time for the questionnaire. We also did not make the distinction between a language learnt in the home, or acquired, and a language learnt at school. The gaps in information, such as this, we were left with when trying to map students’ linguistic lives highlight a limitation to our study, but also demonstrate the complexity of linguistic exposure and language learning. Future research in this area should look to develop more nuanced ways of mapping linguistic lives to afford us with a more accurate and detailed picture, and particularly of students’ home lives.

In popular culture

The questionnaire also asked students about public figures they knew who were multilingual. 70 students referenced 41 public figures. Tom Hiddleston was the most commonly referenced person (20 references), followed by Emma Watson (8 references) and Johnny Depp (4 references). Actors and singers (e.g. Camila Cabello and Shakira) were most commonly referenced. The data showed that alongside 83% of participants not naming a public figure, many of the responses were vague (e.g. ‘English footballers who have moved abroad’) or, to the best of our knowledge, incorrect. We also asked the students whether the public figure inspired them. 34 said yes (e.g. ‘a little bit as I would love to be an actor in a marvel movie’), only four of these citing their multilingualism (e.g. ‘Yeah I think learning languages is very hard and it's inspiring’).

Being multilingual

We also asked the students ‘Are YOU multilingual?’. Although all the participants were learning at least one language at school, 248 students told us they did not see themselves as multilingual and 88 were ‘not sure’. When asked why they gave the answer they did, 21 students told us they were not multilingual because they were learning a foreign language at school, 18 told us they were for the same reason, and 10 were ‘not sure’ because they were learning a language at school. This ambiguity is interesting in that school-based learning is not regarded in the same way by all the students, that is, it does not appear to foster a multilingual identity amongst all. In this question's data, the participants primarily made judgements based on their linguistic repertoire and, specifically, their proficiency in different languages. For instance, the most common reason given for not being multilingual was not ‘knowing’ or ‘speaking’ another language (35 students). We presume these students did not see their foreign language learning at school as constituting enough knowledge here. Indeed another 43 students told us they did not know enough of a second language to be multilingual.

What did the participants think being multilingual means?

Participants indicated some knowledge of multilingualism, with most (337 of 422 responses) indicating that it was to do with different languages when asked ‘what does multilingualism mean to you?’. Of these, 189 felt it was related to speaking ‘multiple’ or ‘many’ languages, for example ‘to be able to speak multiple languages’ or ‘speak languages other than your own’. A further 148 felt it was to do with speaking more than one language. This suggests that multilingualism and bilingualism were synonymous in many participants’ minds, although some were specific that it was about being ‘able to speak more than just two languages’ or even being ‘able to communicate in three or more languages’ or being able ‘to speak every language’.

Some participants (57) referred to the level of fluency required to be multilingual. Of these, 49 referred specifically to being fluent or speaking fluently. Other responses covered a range of skill levels, from ‘being able to slightly speak other languages’ to being able to ‘hold a competent conversation in multiple languages’ and ‘know[ing] more than one language and be[ing] able to use it correctly’. Only three respondents specifically referenced language learning within their definition: ‘to learn multiple languages’; ‘learning more than one language’; ‘using or trying to learn more than one language effectively’.

200 responses referred to, specifically, ‘speaking’ multiple languages, 96 to ‘knowing’, 24 to ‘using’, 21 to ‘understanding’, three to ‘learning’ and three to ‘writing’. Beyond a focus on proficiency, eight responses indicated that multilingualism was a personal characteristic: it meant ‘you’re really cool’; ‘you’re dedicated’; ‘to be an overachiever’. A further 18 responses related to the citizenship aspect of multilingualism: they felt that to be multilingual is ‘to be more open to the world and other countries’, for example. Taken as a whole, these responses indicate that, although students may not see themselves as multilingual, they are able to share definitions of it and its benefits, despite the latter not being explicitly asked about. However, it is important to note that there were 14 respondents who expressed not knowing what multilingualism meant and a further 42 who did not give a response.

Beliefs about multilingualism (BAM) scale

The BAM scale uses the same eight statements as in the original Ungspråk questionnaire (Haukås, Storto, and Tiurikova 2021; 2022). Each was followed by a five-point Likert scale. The results in Table 3 show that there is some variation in how the students scored the items. The mean scores are also reasonably low for all items and no mean scores are above 4 (agree). The lowest scores are for items 6 and 7 which relate to the benefits of multilingualism outside language use, namely, academic success and increased empathy.

Table 3. Descriptive statistics for the BAM scale.

Item	Statement	Mean	Median	SD
1	The more languages you know, the easier it is to learn a new language	3.58	4	1.26
2	People who know many languages are usually smarter than others	2.88	2	1.38
3	People who know many languages are usually more creative than others	2.58	2	1.32
4	People who know many languages, usually make more money than others	2.67	2	1.40
5	Learning new languages helps you to better understand the languages you already know	3.43	4	1.39
6	Knowing many languages makes you better at other school subjects	2.22	2	1.22
7	Knowing many languages helps you understand other people's feelings better	2.40	2	1.41
8	Knowing many languages helps you to see things in different ways	3.38	4	1.39

Items relating to benefits including economic (item 4), intelligence (item 2) and creativity (item 3) were also scored low (between ‘disagree’ and ‘not sure’) indicating the students were not convinced of these benefits to multilingualism, even if they did not disagree outright they were true. The highest scoring items relate to increased ease of learning new languages (item 1) and better understanding of existing languages (item 5), as well as ‘ … see[ing] things in different ways’ (item 8). Item 7 was scored lower than this, which we may find surprising given that both items relate to flexible thinking. However, the same pattern was found in Haukås et al.'s (2022) original use of the questionnaire in Norway.

Factors linked with students’ positive beliefs about multilingualism

We used one-way ANOVA to establish whether a range of factors impacted on students’ beliefs about multilingualism. A significant difference in scores on the BAM scale was found according to year group (F(4, 404) = 2.630, p = .034). Post-hoc Tukey tests revealed that this difference lay between students in Year 7 (mean score = 2.85) and Year 11 (3.25, p = .045). Therefore, the youngest of our participants and the oldest of them held significantly different beliefs with the oldest exhibiting the more positive beliefs. These are the students with the most school-based language learning experience, but also may be more likely to have chosen to study a language (as this is not compulsory in their age range in all schools). No significant differences were found according to gender when three categories (male, female and non-binary) were included (F(2, 394) = .005, p = .995). As the non-binary category was considerably smaller than the other two, a t-test was performed with this category excluded. This also found no significant differences (t(388) = .090, p = .388).

In relation to students’ linguistic lives, no significant differences were found according to the language they were studying (F(4, 402) = 2.257, p = .062). However, there were significant differences according to whether students identified as multilingual or not (F(2, 405) = 11.481, p < .001). Post-hoc Tukey tests revealed that differences existed between students who identified as multilingual (2.74) and those who did not (3.18, p < .001) as well as those who were unsure (3.1, p = .005). We may expect, of course, that those who identify as multilingual are more likely to believe in its benefits. There were also significant differences according to whether students reported their parents speaking another language or not (F (2, 417) = 11.015, p < .001) as well as according to whether they identified their friends as speaking another language or not (F(2, 417) = 6.104, p = .002). Post-hoc Tukey tests revealed that for parents, differences existed between students who answered yes (mean score = 3.15) and those who did not (2.78, p < .001) as well as those who were unsure (2.87, p = .01) and similarly for friends (mean_No 3.03; mean_Yes 2.77; p < .005; mean_Unsure 3.03, p = .019), highlighting the importance of exposure to languages outside of school, specifically.

Finally, whether or not students’ parents had been to university affected their views on multilingualism (F(2, 406) = 5.894, p = .003), with differences existing between those whose parents had (3.04) and those whose had not (2.77; p = .004). Here, parents’ education is used as a proxy for socio-economic status. Again, it is important to acknowledge the unnuanced nature of the use of one question to measure this complex phenomenon. This data highlights the importance of including robust social-economic measures in future research examining multilingual identity or mapping linguistic lives.

The participants’ future multilingual lives

Future multilingual selves (FMS) scale

Table 4 shows the participants’ scores for each FMS scale item. The mean for each item is reasonably low and no item was scored above a 4 (‘agree’). The mean scores for the items relating to one's self-image as a multilingual person in the future (see Henry and Thorsen 2018) indicated the participants were ‘not sure’ about these and, in particular, ‘the person I would like to be in the future speaks many languages very well’ which scored the lowest of these four items. The lowest score overall was for the item relating to language being ‘pointless’ (item 6). It is important to highlight the potential for distortion to due selection bias within this data and perhaps impacting this scale more than other data in the study. We chose to survey students who were studying a language. While those in school years 7,8 and 9 will have had no choice in this, 87 of the 110 students in years 10 and 11 had chosen to study a language (even if they had not chosen which). Therefore, the students who had chosen to study languages, in particular, are likely to see themselves as language users in the future. We may not expect the results from this scale to be so positively orientated towards languages if we had surveyed students who did not study a language.

Table 4. Descriptive statistics for the FMS scale.

Item	Statement	Mean	Median	SD
1	I can imagine myself in the future as someone who knows more than two languages	3.15	4	1.57
2	I hope that I can use languages other than English in my future job	3.26	4	1.54
3	In my future job, I think that knowledge of English will be enough	3.60	4	1.34
4	The person I would like to be in the future speaks many languages very well	2.75	2	1.55
5	It is important to know another foreign language apart from English	3.69	4	1.36
6	Learning another language is pointless because everybody knows English	1.73	1	1.17

Factors linked to students’ future multilingual selves

One-way ANOVA was again used to identify factors which affect students’ future multilingual selves. There were no significant differences according to gender when three categories (F(2, 393) = .562, p = .571) or two (t(387) = 1.054, p = .340) were used. Similarly, there was no significant difference in scores for using languages in the future according to year group (F(4, 402) = .916, p = .454).

Language studied was found to have a significant effect, (F(4, 400) = 3.219, p = .013), and a Tukey post-hoc test revealed that the difference lay between students studying French (2.96) and those studying multiple languages (3.29; p = .027). Identifying as multilingual also had a significant effect, Welch's F(2, 174) = 28.25, p < .001). Games-Howell post-hoc analysis (due to homogeneity of variances being violated) revealed that differences existed between those who identified as multilingual (2.61) and those who did not (3.12, p = <.001) and those who identified as multilingual and those who were not sure (3.02, p = <.001). This scale, unlike others, highlights the importance of studying or knowing multiple languages, specifically. Students studying the most commonly taught language, French, did not see languages in their future to the same extent as those who studied multiple languages. We also know that many students did not see this learning of French (or school-based language learning) to denote being multilingual.

There were also significant differences between those who did and did not have parents and friends who spoke other languages for the FMS scale scores (F_Parents(2, 415) = 11.322, p = .027; F_Friends(2, 415) = 4.028, p = .019). Post-hoc Tukey tests revealed that for both, this difference was between ‘yes’ and ‘no’, rather than ‘not sure’. For parents, ‘yes’ had a mean score of 3.18 and ‘no’ for no, 2.86 (p < .001). For friends, ‘yes’ had a mean of 3.07 and ‘no’ 2.89 (p < .025). Whether parents had attended university or not had no significant effect (F(2, 405) = 2.295, p = .102) for this scale, unlike BAM, however.

Discussion

The students in this study were all linguists at school, in that they were all studying at least one modern foreign language. For most, this was one language and was French. At home, only 6.9% were using a first language other than English and this was Polish for the majority. This figure is lower than the national average of 19.5% and the region's average of 10.5% (DfE 2022) indicating that our sample was, comparatively, largely monolingual. While we may have predicted these findings, this study also shows that such formal assessments of language use in no way encapsulate students’ entire linguistic lives. Indeed, what was less predictable was the range of languages the students had had some exposure to. Across the dataset, 55 languages were referred to in some way and 17 were reported as a language that a participant ‘knew well’. Therefore, while the students’ linguistic lives may seem highly monolingual, particularly by more formal classifications (e.g. ‘first’ language), this is by no means an entirely monolingual sample with entirely monolingual lives.

Other than, for most, French in the classroom and, for some, Polish at home, the use of Spanish on holiday was most referred to in the data. Perhaps particularly in the absence of different languages in the immediate context, travel is a key source of exposure to language(s). Crucially, travel is not a source of exposure available to all, a point of significance in relation to the socio-economic divide in language learning (Lanvers 2018). Yet, other more widely accessible forms of linguistic exposure, such as the media, the internet and through well-known figures were reported far less prominently by the students in this study. The data revealed that, alongside holidays and friends, parents were a significant source of exposure to other languages, yet we struggled to accurately map parents’ language use. This showed us that to be able to accurately map linguistic lives, we need to continue to develop complex and nuanced ways to do this, particularly in terms of personal relationships. Furthermore, given how many young people may use on- and offline social networks, as well as emerging language learning arenas such as online gaming (Jabbari and Eslami 2019), we are going to need to broaden our understanding of the parameters of linguistic lives and, indeed, the student as linguist. Increasingly, and perhaps particularly in England, where formal language learning has been in decline (Collen 2022), we may see young people's linguistic competencies become increasingly complex and difficult to define. This can present both challenges and opportunities to schools.

A significant challenge highlighted by our study is that while schools can be a key source of exposure to languages and serve to foster students’ multilingual identities, students do not necessarily see them this way. Even the learning of languages at school was predominantly not seen by the participants to make a person multilingual and many students felt to be multilingual they would need to know more languages than they did and to a higher proficiency. Addressing these is not straightforward. The low level of fluency typically attained by school-level learners is something which has been previously elucidated within research in the English context (Blow and Myers 2022; Milton 2022; see also Tinsley 2013). Similarly, outside of formal language learning, we also know that many schools struggle to create language-rich environments which draw on students’ linguistic repertoires (Bailey and Marsden 2017; Cunningham & Little, 2022). Shifting focus from progression in one or two languages to the learning of many would also require a substantial overhaul to the current system and is something which, thus far (to the best of our knowledge) has only been formally trialled in primary schools in England (Barton, Bragg, and Serratrice 2009).

Finally, in relation to students’ beliefs about the benefits of multilingualism, these were neither overwhelmingly positive nor overwhelmingly negative. They were also slightly more divergent than those reported in Haukås et al.,'s (2022) study using the same scale with Norwegian students. They report their participants’ beliefs as being more divergent than those of Wei, Jiang, and Kong (2021) in a Chinese context, similarly. As Haukås et al.,'s (2022) comments, it is therefore important to consider the context as well as individual learner variables. We may have expected students in our study to exhibit more negative views towards the benefits of multilingualism given the predominantly monolingual sample, as well as the potential for a monolingual habitus (Gogolin 1997), particularly in contexts such as this (see Bailey 2022). Unlike in Haukås et al.,'s (2022) study, students who identified as multilingual scored items on this scale significantly higher than those who did not demonstrating the important potential for this identity marker in terms of shaping students’ beliefs and, thus, the important role of schools in shaping this identity marker.

Importantly, and linking each aspect of this study together, factors related to students’ linguistic experiences impacted positively on both outcome variables. Where students had parents or friends who they considered to speak another language, their scores on the future multilingual self scale were significantly higher, as they were for beliefs about multilingualism. This was found in Haukås et al.,'s (2022) original study as well. It also aligns with previous studies which have shown that students need to be able to see a personal relevance to language learning in order to be motivated (Parrish and Lanvers 2019) and to persist (Taylor and Marsden 2014). Each aspect of a student's linguistic life is part of a complex arrangement and while schooling may not necessarily be the most significant part in students’ eyes, it is a key site of linguistic exposure, ripe for empowering students who see themselves as multilingual and/or who value multilingualism and its place in their lives in the future. As part of this, schools should recognise the role that they can play in supporting students’ languages, drawing on personal and family connections, and foster the positive impact that can have on language learning, as well as multilingual identity and beliefs about multilingualism.

Conclusion

Formal schooling has the potential to be an essential component of a person's linguistic life, particularly in the absence of multilingualism in a community, family and/or a person's friendships and online life. Whilst students may experience languages through school (usually French, in formal, taught sessions), this exposure to language(s) does not seem to impact their beliefs and self-conception as multilingual, as much as more personal exposure to languages (i.e. friends and family) does. It is important that students recognise that the languages they learn in school ‘count’ and are part of their linguistic repertoires, although we recognise that this is likely to be at a substantially lower level of proficiency than other languages they have. At present, students do not seem to see themselves as linguists now, nor see multilingualism in their futures. We need to inspire the student-as-linguist to see the languages around them and challenge views that exclude them from developing a multilingual identity (i.e in their words, ‘fluently’ or ‘correctly’ ‘speaking many languages’).

Acknowledgements

We would like to thank all the students who took part in the study and all the staff members who helped make that possible. We know how many demands are on your time and we are so grateful for your help.

Disclosure statement

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