Investigating former pupils’ experiences and perceptions of CLIL in Finland : a retrospective analysis

This is a self-archived version of an original article. This version may differ from the original in pagination and typographic details. Author(s): Title: Year: Version: Copyright: Rights: Rights url: Please cite the original version: CC BY-NC 4.0 https://creativecommons.org/licenses/by-nc/4.0/ Investigating former pupils’ experiences and perceptions of CLIL in Finland : a retrospective analysis © 2019 The Authors. Published by Informa UK Limited, trading as Taylor & Francis Group. Published version Roiha, Anssi Roiha, A. (2019). Investigating former pupils’ experiences and perceptions of CLIL in Finland : a retrospective analysis. Nordic Journal of Studies in Educational Policy, 5(2), 92-103. https://doi.org/10.1080/20020317.2019.1586514 2019 Full Terms & Conditions of access and use can be found at https://www.tandfonline.com/action/journalInformation?journalCode=znst20 Nordic Journal of Studies in Educational Policy ISSN: (Print) 2002-0317 (Online) Journal homepage: https://www.tandfonline.com/loi/znst20 Investigating former pupils’ experiences and perceptions of CLIL in Finland: a retrospective analysis Anssi Roiha To cite this article: Anssi Roiha (2019) Investigating former pupils’ experiences and perceptions of CLIL in Finland: a retrospective analysis, Nordic Journal of Studies in Educational Policy, 5:2, 92-103, DOI: 10.1080/20020317.2019.1586514 To link to this article: https://doi.org/10.1080/20020317.2019.1586514 © 2019 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. Published online: 28 Mar 2019. Submit your article to this journal Article views: 184 View related articles View Crossmark data Citing articles: 1 View citing articles Investigating former pupils’experiences and perceptions of CLIL in Finland: a retrospective analysis Anssi Roiha Department of Language and Communication Studies, University of Jyväskylä, Jyväskylä, Finland ABSTRACT The educational approach known as Content and Language Integrated Learning (CLIL), in which content is taught partly through a foreign language, has gained great popularity in Europe in the past few decades. In Finland, CLIL has been in use since 1991 and, despite some fluctuations in its popularity, has gained a relatively stable place in the Finnish education system. CLIL has been extensively studied, but previous CLIL research has mostly focused on pupils currently enrolled in CLIL. This study takes a novel perspective by investigating CLIL retrospectively, through the eyes of former pupils. The data used are in-depth interviews with 24 former pupils who attended a CLIL class in Finland in the 1990s. The interviews were analyzed using qualitative content analysis. The findings reveal that the participants had overwhelmingly positive memories of their CLIL programme. They felt strongly that CLIL had positively affected their English language proficiency, particularly their vocabulary and speaking skills. Most believed that CLIL had not adversely affected their content learning. Despite the overall satisfaction, a few participants suggested ways to develop CLIL, which are also discussed. The results of this case study broaden our understanding of CLIL and have implications for language education policy. ARTICLE HISTORY Received 13 September 2018 Accepted 16 February 2019 KEYWORDS CLIL; content learning; foreign language learning; perceptions; emic perspective Introduction Along with one’s socioeconomic and family background, education and school experiences have been shown to be significant in shaping one’s life course and determining how one finds one’s place in society (e.g. Dominicé, 2000; Kauppila, 2002; Vanttaja, 2000). Above all, negative school experiences and poor attainment often predict challenges in later life (e.g. Kuronen, 2010). It is therefore important to investigate different educational practices and their effect on pupils’identities and attitudes towards schooling. The educational context of the present study is Content and Language Integrated Learning (hereafter CLIL). In this article, CLIL is defined as an educational approach that encompasses various models in which content is partly taught through a foreign language (Coyle, Hood, & Marsh, 2010). CLIL can be seen as partly stemming from and being influenced by the Canadian immersion programme that was developed in the 1960s (Cummins & Swain, 1986). Consequently, there has been an on-going debate about the similarities and differences between CLIL and immersion (e.g. Lasagabaster & Sierra, 2010; Somers & Surmont, 2012). This article does not delve very deeply into that discussion but takes the position that within a myriad of CLIL variations, some may have more similarities with immersion while others are rather different from it. Both immersion and CLIL nevertheless share the same theoretical underpinning, the view that language is best acquired through authentic communication, and thus aim to provide pupils with both comprehensible input (Krashen, 1986) and comprehensible output (Swain, 1985). CLIL has mushroomed in Europe during the past few decades (Dalton-Puffer, 2011; Eurydice, 2006). In Finland, CLIL started in 1991, following some changes to the relevant legislation. Despite some fluctuation over the years, CLIL has gained a stable position as one of the educational approaches used in the Finnish education system (e.g. Lehti, Järvinen, & Suomela-Salmi, 2006; Nikula & Marsh, 1996; Peltoniemi, Skinnari, Mård-Miettinen, & Sjöberg, 2018). In Europe generally, the approach has been taken up as one way to increase plurilingualism and tackle the challenges of migration and globalization (Ruiz de Zarobe, 2008). Previously, the majority of CLIL studies have revolved around learning outcomes and have been conducted from etic perspectives (e.g. Dalton-Puffer, 2011; Pérez-Cañado, 2012). Recently, emic perspectives in CLIL studies have become more common as pupils’perceptions have also started to be of interest to researchers (e.g. Coyle, 2013; Pladevall-Ballester, 2015). However, most studies have been quantitative and focused on pupils CONTACT Anssi Roiha [email protected] Department of Language and Communication Studies, University of Jyväskylä, PO Box 35, FI-40014, Jyväskylä, Finland NORDIC JOURNAL OF STUDIES IN EDUCATIONAL POLICY 2019, VOL. 5, NO. 2, 92–103 https://doi.org/10.1080/20020317.2019.1586514 © 2019 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group. This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial License (http://creativecommons.org/licenses/by-nc/4.0/), which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited. currently enrolled in CLIL programmes. The aim of the present study is to provide a novel perspective by drawing attention to former pupils’experiences of CLIL in order to broaden our understanding of this multifaceted approach to learning. The results presented here are part of a larger research project that examines CLIL through the eyes of former pupils. Specifically, in this article, 24 former Finnish pupils, who attended an English-medium CLIL programme for nine years in the 1990s, retrospectively reflect on their CLIL experiences. This article relies solely on the participants’subjective perceptions and no measurements, for instance, on their language proficiency were conducted. The specific research questions for this study are: (1) How do the participants reflect on CLIL as an experience? (2) How do they perceive the effect of CLIL on a. their target language learning? b. their content learning? Literature review This section presents previous CLIL research relevant to the scope of this article. First of all, there is a brief summary of the most important research on the effect of CLIL on target language and content learning. Secondly, CLIL studies focusing on pupils’perceptions, which have become more mainstream, are reviewed. Following the conceptualization of CLIL presented in the introduction, this article focuses on studies conducted in CLIL contexts, leaving purely immersion studies aside. However, the fact that immersion has been shown to be beneficial to pupils’ target language skills without any detriment to content learning (e.g. Lazaruk, 2007) gives some support to the claims for CLIL programmes as well. Learning outcomes in CLIL A number of studies have shown that CLIL has a very positive effect on pupils’target language learning. In several studies, CLIL pupils have even outperformed their non-CLIL peers in all the measured skills. For instance, in Lasagabaster’s(2008) study, CLIL pupils (n= 113) demonstrated higher achievement in grammar, listening, speaking and writing than their nonCLIL peers (n= 28). In Lorenzo, Casal, and Moore’s (2010) study, CLIL pupils (n= 754) performed better in reading, listening, writing and speaking than the control pupils (n= 448), while in Ruiz de Zarobe’s (2008) study, CLIL pupils (n= 107) outperformed their peers (n= 54) in pronunciation, vocabulary, grammar, fluency and content. Again, Admiraal, Westhoff, and de Bot (2006) found that CLIL pupils (n= 548) demonstrated higher proficiencies in oral production and reading comprehension than their non-CLIL peers (n= 721) but the effect of CLIL on receptive word knowledge was neutral. In Finland, Järvinen (1999) found that CLIL pupils’(n= 90) foreign language learning was significantly faster than that of their non-CLIL peers (n= 47). Compared to studies on language learning, those on content learning have been less conclusive. For instance, Surmont, Struys, Van Den Noort, and Van De Craen (2016) examined the effect of French-medium CLIL on pupils’learning of mathematics and found that CLIL pupils (n= 35) outperformed their non-CLIL peers (n= 72) already after three months. The pre-test showed that there were no a priori differences between the two groups.InFinland,Seikkula-Leino(2007)andJäppinen (2005) examined CLIL in relation to content learning. On average, in both studies, CLIL had a neutral effect on pupils’learning. However, in Seikkula-Leino’s(2007) study, which measured the learning of mathematics and mother tongue (i.e. Finnish), overachieving was much more common for non-CLIL pupils (n= 101) than for CLIL pupils (n= 217), which suggests that in CLIL classes pupils learn according to their abilities but not above that. Jäppinen (2005)lookedatpupils’cognitive development in science and mathematics and found that in some cases the CLIL pupils’(n=335) cognitive development was even faster than the non-CLIL pupils’(n= 334). Admiraal et al. (2006) also measured content learning and found that CLIL had a neutral effect on the pupils’history and geography learning. However, the data for that were very limited, yielding only tentative results. Additionally, in Dallinger et al.’s(2016) study in Germany, which took a priori differences into account, CLIL was found to have a neutral effect on pupils’history learning even though the CLIL pupils (n= 703) had received 50 per cent more history lessons than their non-CLIL peers (n= 1103). This study also investigated language learning outcomes and found that CLIL had a positive effect only on listening comprehension, but not on overall English proficiency, a finding that challenged some previous studies on target language acquisition. Among studies which have indicated that CLIL would be detrimental to content learning is one conducted by Fernández-Sanjurjo, Fernández-Costales, and Arias Blanco (2017), which investigated the science learning of primary pupils (n= 709) in Spain. Approximately half of the pupils were Englishmedium CLIL pupils and the rest were their monolingual peers. The findings showed that pupils studying in their L1 performed slightly better than the CLIL pupils, although the difference was not very substantial. Additionally, pupils’social and economic status had an effect on their performance: pupils with a more privileged background outperformed those with lower socioeconomic status. In conclusion, the studies presented here suggest that CLIL is, generally, a useful approach for enhancing pupils’foreign language proficiency. However, NORDIC JOURNAL OF STUDIES IN EDUCATIONAL POLICY 93 there is more disagreement about which areas of language skills CLIL enhances the most. This may partly be explained by the fact that CLIL is implemented in various ways, so some programmes may emphasize oral production while others focus more on writing and reading. Moreover, it seems that CLIL may not advance pupils’content learning more than monolingual teaching does, even though it has been claimed that bilingual education has a positive effect on pupils’cognitive development (e.g. Jäppinen, 2005; Lazaruk, 2007). On the other hand, CLIL does not typically seem to be detrimental to content learning. Similar conclusions were reached by Graham, Choi, Davoodi, Razmeh, and Dixon (2018) in their recent review. While the studies reviewed demonstrated mixed results on the effectiveness of CLIL on language and content outcomes, the authors concluded that overall CLIL seems to have either a neutral or a slightly positive effect on pupils’attainment in terms of both language and content. However, it is important to bear in mind that CLIL is dependent on many contextual variables. Thus, one must be cautious about making too strong generalizations on the basis of existing CLIL studies. Additionally, it is worth noting that some of the quantitative studies presented here have been somewhat small-scale (e.g. Järvinen, 1999; Lasagabaster, 2008). Previous CLIL research has also been questioned and criticized, for instance, for its different definitions of CLIL, for the lack of pre-tests with CLIL and control pupils, as well as for its neglect of socioeconomic background and pupil selection (e.g. Bruton, 2011; Cenoz, Genesee, & Gorter, 2014; Küppers & Trautmann, 2013). For instance, Rumlich (2016) found that CLIL pupils do not necessarily gain much from CLIL when a priori differences are taken into account. Thus, more research is still needed to establish the effects of CLIL on pupils’ learning outcomes. Pupils’perceptions Earlier CLIL studies have only looked at the experiences and opinions of pupils currently taking part in CLIL. Further, most studies have focused on secondary pupils, neglecting young learners’perceptions, although some researchers have examined primary pupils’perceptions, for instance Massler (2012) (n= 176), Pladevall-Ballester (2015)(n= 197) and Ramos (2007)(n= 61). In all of these studies, the majority of pupils reported that they enjoyed CLIL and regarded it as beneficial to them: the pupils in Massler’s(2012) study would have preferred to receive even more English-medium instruction, Pladevall-Ballester (2015) found that most pupils considered CLIL to be both improving their English skills and facilitating their content learning, and most pupils in Ramos’(2007) study felt that CLIL would help them get a better job in the future and provide them with the tools to communicate with others. Despite the overall satisfaction, all the studies included a number of pupils who were less satisfied with CLIL. For instance, Massler (2012) found that 10 per cent of pupils did not want to have CLIL in other subjects, and 22 per cent had experienced difficulties in content learning in CLIL lessons. Similarly, in Pladevall-Ballester’s(2015) study, some pupils felt that the language in CLIL lessons was too challenging: more than a third of the pupils reported experiencing difficulties in comprehension, and almost a half reported difficulties in oral production in their CLIL classes. Ramos (2007) found that more than a third of the pupils were unsure whether learning in two languages would help them perform better at school, and almost half of them were unsure whether it would enhance their cognitive skills. Secondary pupils’perceptions have been examined, for instance, by Coyle (2013)(n= 670) and Hunt (2011)(n= 283) in the UK. Overall, the pupils regarded CLIL as a highly positive experience. In Coyle’s(2013) study, 85 per cent of the pupils reported that they hoped CLIL would continue at their school. Many of them considered CLIL beneficial from the point of view of their language proficiency and thought that in particular CLIL developed their speaking and communication skills. In Hunt’s (2011) study, approximately two-thirds of the pupils reported enjoying the CLIL lessons and nearly twothirds claimed that they looked forward to learning through a foreign language in the future. Additionally, many pupils compared CLIL to mainstream classes, declaring that it was ‘different’and ‘better’. As with the studies on primary pupils’perceptions, Coyle (2013) and Hunt (2011)also encountered critical voices. That is, Coyle (2013) found that 15 per cent of pupils regarded CLIL as too difficult, boring or useless. The more critical pupils felt that CLIL lessons included too much teacher talk, translation and writing, or that the topic was already familiar to them. Hunt (2011) found that 7 per cent of pupils did not enjoy CLIL and 12 per cent did not want to continue CLIL in the future. In a very recent study, Somers and Llinares (2018) looked at Spanish secondary CLIL pupils’(n= 157) motivation towards the target language as well as content learning in highand low-intensity CLIL programmes. The results showed that, in general, pupils in both groups seemed to enjoy CLIL and perceived it as useful for their future. However, the pupils in the high-intensity group were more motivated and regarded CLIL as benefiting their future studies and professional careers more than the pupils in the low-intensity groups. Despite the overall satisfaction with CLIL, only about half of pupils reported 94 A. ROIHA being at ease in CLIL lessons. The pupils’answers to open questions revealed that anxiety among the pupils in the high-intensity group mostly related to the demanding content. In contrast, anxiety among the pupils in the low-intensity group was related to what they saw as both the demanding content and their own perceived low level of proficiency in the target CLIL language. In Finland, pupils’(n= 209) attitudes towards CLIL have been examined, for instance, by Pihko (2010). The results of her mixed methods study show that the general attitude of most secondary pupils towards CLIL was very positive. A clear majority considered CLIL easy and only 6 per cent reported it as difficult. However, one fifth of the participants disagreed with the statement that studying through a foreign language is pleasant and 15 per cent would have preferred to study all the subjects in their L1. In Pihko’s(2010) study, those pupils who reported negative attitudes to CLIL in the survey were also interviewed. Many of them considered that their own language skills were insufficient to cope with CLIL, and some school subjects, such as history, mathematics, physics and chemistry, were specifically mentioned as difficult subjects to study in a foreign language. Pupils would also have liked more guidance and support in their learning as well as more opportunities to use the language in practice. To conclude, the majority of pupils themselves appear to enjoy CLIL and regard it as a positive and useful educational approach. The studies reported here, however, show that most CLIL classes also include learners who do not see the benefits of CLIL and struggle with the approach. This is vital information also for practitioners. However, previous studies on language learning, content learning and pupils’perceptions of CLIL have focused exclusively on participants who are currently enrolled in CLIL. Thus, there seems to be an important research gap, an absence of studies using retrospective reflections on CLIL, which the present study aims to fill. Methodology The CLIL context and the participants This study explores the insights of pupils from one former CLIL class in Finland. The primary purpose of this research was to describe and understand the participants’experience of CLIL, and it therefore gives voice to pupils themselves constructing their narratives and memories of their CLIL times. Altogether, 29 pupils (including the researcher) studied in the class at some stage. They were all contacted via Facebook and 24 agreed to take part in the research. All the participants signed a letter of consent which outlined the aim of the research and how the data would be used. Additionally, it was made clear to the participants that they had the right to withdraw from the study at any point. To preserve their anonymity, the participants were given pseudonyms (see Table 1). Most of the participants attended English-medium CLIL for nine years during their compulsory schooling (years 1–9, pupils’age 7–15), starting in 1992. Unlike the majority of CLIL programmes in Europe, the target programme did not require applicants to take a pre-test. However, preference was given to pupils who had some prior experience of English, and this was the case for 5 pupils in the class. The number of applicants outnumbered the places in the target year, so the remaining places had to be filled by drawing lots. In primary Table 1. The participants in the study and the interview details. Pseudonym Time spent in the CLIL class Date of the interview Duration of the interview Venue of the interview Anna Pre-school–9 th Year (10 years) 8.9.2016 58:04 Researcher’s home Annika 1 st Year–6 th Year (6 years) 19.9.2016 41:37 Researcher’s home Arttu 7 th Year–9 th Year (3 years) 1.10.2016 39:18 Hotel lobby Eemeli Pre-school–9 th Year (10 years) 11.9.2016 45:44 Researcher’s home Emmi Pre-school–9 th Year (10 years) 17.9.2016 40:22 University library Hanna Pre-school–6 th Year (7 years) 15.11.2016 30:17 Researcher’s home Jere Pre-school–9 th Year (10 years) 24.1.2017 38:35 Via Skype Jonne 1 st Year–9 th Year (9 years) 30.9.2016 45:55 Participant’s home Juho 1 st Year–9 th Year (9 years) 1.10.2016 69:03 Participant’s home Jukka Pre-school–9 th Year (10 years) 6.9.2016 34:54 Researcher’s home Kaapo 1 st Year–6 th Year (6 years) 26.1.2017 36:10 Via Skype Kalle Pre-school–9 th Year (10 years) 1.10.2016 37:15 Participant’s home Kimmo 1 st Year–9 th Year (9 years) 10.12.2016 63:18 Participant’s home Lotta Pre-school–4 th Year (5 years) 21.11.2016 34:07 Researcher’s home Maria 1 st Year–6 th Year (6 years) 2.10.2016 61:08 Hotel lobby Marko 1 st –4 th Year, 6 th –9 th Year (8 years) 3.12.2016 78:31 Researcher’s home Niko Pre-school–9 th Year (10 years) 19.1.2017 39:30 Via Skype Olli 1 st Year–9 th Year (9 years) 4.9.2016 38:24 Researcher’s home Pasi Pre-school–9 th Year (10 years) 25.8.2016 39:18 Researcher’s home Riikka 7 th Year–9 th Year (3 years) 5.11.2016 45:24 Researcher’s home Roni 1 st Year–9 th Year (9 years) 22.12.2016 41:22 Participant’s home Samu 3 rd Year–9 th Year (7 years) 12.11.2016 62:14 Participant’s home Sanna 2 nd Year–7 th Year (6 years) 16.1.2017 79:56 Via Skype Tuukka Pre-school–9 th Year (10 years) 13.10.2016 63:02 Researcher’s home NORDIC JOURNAL OF STUDIES IN EDUCATIONAL POLICY 95 school, CLIL was implemented in most subjects and approximately a quarter of the overall teaching was conducted in English. In secondary school, the amount of CLIL decreased and it was less systematic than at primary level. The declared goal of the CLIL programme was to make pupils confident and competent language users and to provide them with the skills needed to function in an increasingly international society. At the time of the interviews, the participants were 30–31 years old. After the CLIL comprehensive school, 19 of them had completed upper secondary school, two had graduated from vocational school and three had obtained a dual diploma (i.e. from both). 12 participants had amaster’sdegreeorequivalent,sevenhadaBachelor’s degree or equivalent (three of them were currently finishing their Master’s studies) and four had a vocational degree. In addition, two participants were currently completing their doctoral degrees. Four participants had completed their studies entirely in English and eleven participants’degrees had included some Englishmedium courses. The main working language of three participants was English, ten used English at work regularly, and nine used it sporadically. After their CLIL schooling, seven participants had lived abroad and four of them were still currently doing so. Each participant’s post-CLIL education and English use is outlined in more detail in Appendix 1. Data collection and analysis The data of this study are 24 in-depth interviews with the participants. The interviews were very open in nature as the emphasis was on the participants’personal perceptions and experiences. The researcher had personal experience as a pupil in the target class, and the resulting experience and position may have influenced the analysis and interpretations. The earlier relationship between the author and the participants also added a unique character to the interviews. Garton and Copland (2010) labelled these types of interviews acquaintance interviews,and they suggest that due to the shared experience, they may offer access to resources that are not always available in traditional interview settings, although acquaintance interviews are not necessarily a more valid method of data collection than other interview types. The broad interview themes were sent to the participants in advance so that they could retrieve aspects of their CLIL classes before the interviews, especially as the participants were looking back on memories and experiences from more than 20 years before. The interviews were conducted in Finnish, by the author, between August 2016 and January 2017. The duration of the interviews ranged from 30 min to 78 min, the average being 48 min (see Table 1 for more details). The interviews were recorded and transcribed verbatim by the author. The data were analysed following theory-oriented content analysis methods. That is, the relevant umbrella concepts (i.e. perceptions, language learning, content learning) and previous studies related to them were acknowledged but the analysis still relied extensively on the data (e.g. Tuomi & Sarajärvi, 2009). Results The results of the study will be presented according to the research questions. First, the participants’general attitudes towards CLIL will be discussed. Then the following part will examine the perceived effects of CLIL on the participants’target language and content learning. The results section contains quotations from the interviews, which have been translated into English by the author. CLIL as an experience The participants were generally very satisfied with their CLIL experience. Their satisfaction was clear, as all the participants said that they would choose to take part in similar CLIL again and that they would also like their own children to have that experience. Many participants emphasized how CLIL had motivated them and made learning more interesting: I think it was also quite interesting to do things in English.. that it’sanywaymorevariedthen..anditprobably says quite a bit about the teaching that you don’t remember that we’ve gone through things in English because it was quite natural already early on. (Olli) I think it was divided so that sometimes we had Finnish-medium lessons and sometimes we had more in English then.. but maybe they’re also getting a bit mixed up in my head so that I don’t really recall in which language a certain lesson was taught.. so maybe already very early on English didn’t feel like anything so disruptive. (Eemeli) The fact that the participants did not always notice whether the language of instruction was Finnish or English exemplifies how taken-for-granted and natural CLIL was for them as an approach. Both Olli and Eemeli referred to the starting age of CLIL and regarded it as a significant factor in their having a positive attitude towards English. This should for its part encourage schools to introduce CLIL more widely already at primary level: the target CLIL programme started in Year 1 (when the pupils were 7 years old). In addition, many of the participants had already had some less serious English-medium teaching in pre-school (see Table 1). Most participants expressed the view that using English had always been effortless and that CLIL had contributed to their highly positive English language self-concept (see also Roiha & Mäntylä, in press). Thus, the main objective of the programme (i.e. making pupils confident language users) seemed to have been 96 A. ROIHA fulfilled. Issues of privilege and superiority were also frequently raised in the interviews. That is, many of the interviewees had already started to realize the advantages CLIL brought to their lives while they were still at school. Anna, for instance, said: Well I do feel that I had this kind of class identity that we were a bit better than the others. (Anna) Like Anna, a few other participants also referred to their class identity and associated it with CLIL. They felt that this identity, which intersected with a sense of privilege, was partly constructed by the teachers and even by the parents. The participants made explicit references to the novelty of the CLIL approach and to the fact that it was open to only a very limited number of pupils. This, as well as the fact that the participants were chosen from among many applicants, seemed to have prompted a sense of uniqueness, and this had lasted throughout their time at school. For instance, Marko talked about the pupils in the CLIL class as follows: Were the pupils after all selected for our class? It felt as if there were a lot of educated and academic parents and everything.. and then nearly everyone performed well at school.. so that it was maybe a special class in some way. (Marko) The participants’remarks partly reflect the time of their schooling: that is, in the early 1990s CLIL was something new and was generally regarded as more exotic than it is now. It is therefore arguable that the feedback and social comparisons had a prominent role in forming their sense of privilege, which in turn reinforced their positive attitude towards the CLIL programme. CLIL has been quite widely criticized for being elitist, for instance, because of the pupil selection and the implementation of CLIL in areas with families of high socioeconomic status (Cenoz et al., 2014). Contrary to Marko’s surmise, the target CLIL programme was in fact open to everyone, and the pupils were not subjected to a pre-test. It was, however, partly selective, as priority was given to pupils who already had some experience of English before school began. The remaining pupils were chosen randomly. Additionally, the area of the school was relatively high in terms of socioeconomic status. The CLIL group was therefore not representative of an average class in Finland at the time. The interviewees’overwhelming satisfaction with CLIL is illustrated by the fact that few of them could come up with any negative remarks about their CLIL experience: I can’t honestly think of anything [negative].. I tried to come up with some negative aspects but I honestly couldn’t come up with anything. (Arttu) A few of the participants did, however, express some negative memories and experiences. For instance, there was some criticism of CLIL in secondary school: some people felt that CLIL was poorly implemented there and should have been more systematic and more goal-oriented: I mostly remember that we had very little of it [CLIL classes].. and that it was quite poorly integrated.. and somehow it felt like it didn’t really work.. like the idea of it. (Riikka) One explanation for some participants’negative recollections might be that the nature of the CLIL programme changed quite considerably when the group entered secondary school: the amount of CLIL decreased and many participants felt that some teachers were not very committed to the CLIL programme. The contrast between primary and secondary CLIL was an issue that arose in many interviews. Although, overall, the target class performed relatively well at school, there were a few participants who said that at times they had felt overwhelmed by CLIL and that their language skills were not good enough to follow the English-medium instruction. Studying certain subjects (e.g. mathematics and chemistry) partly in a foreign language was said to pose an extra challenge and to cause occasional difficulties in their learning, which echoes Pihko’s(2010) research. The most critical voice was Hanna’s: Maybe I just didn’t keep up with the others.. well enough.. I think I would’ve just needed like more personal support.. maybe I even sometimes thought to myself that I wish I wasn’t in this class. (Hanna) In general, Hanna’s English use after CLIL had been very limited and her English language self-concept seemed to be fairly negative. It is worth remarking that even those participants who experienced difficulties later on in their education assessed CLIL as a positive and beneficial experience and would still choose to go through the same sort of schooling. A few of them speculated that without CLIL they might have experienced more severe foreign language learning difficulties and greater language anxiety. For instance, despite her negative remarks, Hanna remained positive about being selected for the CLIL class: I don’t think that I’d know how to speak even this much English if I’d only started it later.. I think it’s been useful that I was there and hung in there with the others. (Hanna) Most participants were satisfied with the amount of CLIL, which was approximately 25 per cent of the overall teaching. They justified their satisfaction on various grounds. Some looked at the issue in relation to their language proficiency and considered that there was enough CLIL to give them a good command of English. Others evaluated the amount from the perspective of content learning and believed that more CLIL instruction might have led to difficulties in mastering the different school subjects. For instance, the following quotation from Juho’s NORDIC JOURNAL OF STUDIES IN EDUCATIONAL POLICY 97 interview, in which he juxtaposes different educational models, exemplifies how, on the one hand, he appreciated CLIL but, on the other, how he considered that some matters have to be learned in one’sL1: Yes I would [put my children in CLIL].. and I think exactly this is good that.. that if the options were to send them to this kind of class, a totally Finnishmedium class or a totally English-medium class in Finland.. then I would rather put them in this kind of class than in a totally English-medium class because.. because anyway there are a lot of things like this that you have to learn in Finnish. (Juho) A few participants agreed with the amount of CLIL they received because it did not have a stigmatizing effect; that is,theyfeltthatitwasenoughtoachievethegeneralaim of the CLIL programme (i.e. of making pupils confident and competent language users) but that more Englishmedium teaching might have been viewed negatively by their non-CLIL peers. Only four participants mentioned wanting to have even more CLIL. This was mostly in secondary school where, as has already been mentioned, the amount of CLIL decreased considerably. In conclusion, CLIL seems to have made the participants’schooling enjoyable, which has been identified in previous research on pupils currently enrolled in CLIL (e.g. Massler, 2012; Pladevall-Ballester, 2015). Enjoying school was not typical in Finland at that time, as the 1990s was in general a period of major cuts to school budgets, which was mirrored in pupils’dissatisfaction with school (e.g. Linnakylä, 1993; Nikkanen, 1999). However, it is important to bear in mind that, whatever the participants felt as they looked back, other factors besides CLIL may have affected their school satisfaction. Nevertheless, the participants’perception was that, at least to some extent, especially CLIL had increased their school satisfaction. Perceptions of the effect of CLIL Overall, the participants believed that CLIL had enhanced their language proficiency. The most recurrent theme was the positive English language selfconcept that CLIL had formed (see also Roiha & Mäntylä, in press). That is, the majority of the participants declared that they were confident language users who had not experienced any language anxiety, and they emphasized the role of CLIL in this. Again, the participants emphasized that the early start in CLIL had had an impact on their English proficiency, which speaks for the implementation of CLIL at an early, primary level: Exactly those, pronunciation and how easy it is to follow spoken language, so I believe the younger you study or in general hear or use the language a little, the easier it is to acquire it. (Emmi) The participants underlined how natural a language English had become to them as a result of CLIL. Most of them considered that their English competence was generally much better than average. Some even evaluated their English proficiency as being almost nativelike: All the grammatical structures and so on are as obvious as in speaking Finnish so I don’t need to think about them at all. (Olli) Just like in Finnish.. like you’d be discussing in Finnish.. you’re able to express yourself the same way in English. (Jukka) It is worth noting that English had not had a very prominent role in the lives of some of the participants who perceived themselves as highly competent English users. For instance, Olli had not used English very much after CLIL, and Jukka’s studies had included only a few English courses (see Appendix 1). In addition to overall English proficiency, the participants regarded CLIL as specifically benefiting certain language skills. Three-quarters of the participants thought that CLIL had mostly developed their vocabulary and speaking skills: I think that many words became familiar.. even if they didn’t go into your productive vocabulary they’ve still been like hold on I’ve heard that somewhere before.. and the context of the word has become in some ways familiar. (Kimmo) Well you could say the terminology in geography or biology.. of course you acquired a much richer vocabulary. (Sanna) Probably at least speaking.. pronunciation and using the language in everyday situations.. and how you pronounce a word properly. (Maria) In line with the participants’perceptions, many studies from different CLIL contexts have similarly shown that CLIL pupils are often ahead of their peers in vocabulary and oral production (e.g. Admiraal et al., 2006; Lasagabaster, 2008; Ruiz de Zarobe, 2008). However, arguably, which specific language skills CLIL enhances depends in part on how CLIL is implemented in practice, which varies a great deal within Europe (e.g. Eurydice, 2006). It can be inferred that in the CLIL context examined in this study, a great deal of emphasis was placed on oral output as well as input (see e.g. Krashen, 1986; Swain, 1985). Listening comprehension was mentioned by one fifth of the participants whereas grammar, reading and writing were referred to only occasionally. Two-thirds of the participants in the present study perceived that CLIL had had a neutral effect on their content learning. Many related this outcome to how natural CLIL was, and considered that the language of instruction did not play a role in the learning process: 98 A. ROIHA