Language Learning beyond English in the Netherlands: A fragile future?

3.1 French, German and English – a historical triptychon of Dutch MFL education

When teachers in the Netherlands talk about MFLs they typically speak of “French, German and English”, which reflects the order in which the languages historically appeared in the Dutch curriculum (Kwakernaak 2009). Already in the 16th century, children whose parents could afford sending them to school, went to the French school, where they would study French next to Latin and Greek. From the 18th century onwards, German and later English, were added. These three MFLs were also standard in the so-called Latin Schools where the grammar-translation method was dominant be it to teach modern or classic (Latin/Greek) languages (Kwakernaak 2011).

In 1863 the Dutch government implemented the Hogere Burgerschool (HBS), which is the basis of today’s secondary school system. Teachers became public servants and were obliged to get a national teaching degree: the middelbaar onderwijs (mo) akte. This first attempt to harmonize the Dutch education system induced several societal changes: Factory work for children under 12 years was prohibited in 1874, and from 1900 onwards primary education was compulsory for children aged 5 to 12 years old. Within the HBS, French, German, and English became mandatory and subject to examination. In 1911, the first national teachers association for MFLs was established, and from 1921 onwards students could major in an MFL at university. Today, the Vereniging van Leraren in Levende Talen (the Association of Teachers in MFLs) is still an active association with more than 3500 members, hosting annual conferences and publishing a scientific journal and a practically oriented magazine for language teachers in the Netherlands (www.levendetalen.nl ).

3.2 The Mammoetwet of 1968

For almost a century, French, German and English formed a stable part of Dutch general education. However, in 1968 a large educational reform (Wet op het voortgezet onderwijs – Law for secondary education) was implemented, called the Mammoetwet (mammoth law) reflecting its enormous size. It aimed to streamline the landscape of compulsory secondary education in the Netherlands, that, despite the efforts of the HBS, was still composed of a myriad of different school types. The Mammoetwet also aimed at adapting the curriculum to the needs of modern society in the early 1970 s. Due to the importance of English as the language of the UK and the USA, both thriving cultural, economic and political nations in those days, the position of other MFLs was heavily debated, and French almost lost its position as a mandatory school subject (Kwakernaak 2011). In the end, all three MFLs remained compulsory subjects in the first three years of secondary school (called onderbouw, i. e., junior secondary school). For the bovenbouw (senior secondary school), a minimum of one (in some tracks two) MFL was set for the final exam. Virtually all students opted for English. Many policy changes followed in the 1970 s, resulting in decreasing numbers of contact hours for MFLs (Voogel 2016).

The Mammoetwet also introduced the so-called “vakkenpakket” (subject bundles): students were no longer obliged to take exams in all subjects, but they could choose a bundle of subjects that were deemed relevant for further education. The government appointed a new testing institution, CITO, that was tasked to develop a national assessment system to ensure objective, transparent, reliable and valid examination of all subjects. Today, CITO is still in charge of the national curriculum-based examinations at the end of Dutch secondary education. Students take the final CITO exam (called centraal schriftelijk eindexamen) and local exams implemented by individual schools (called schoolexamen), both of which count for 50 % towards their final grade. For MFLs, pragmatic criteria led to the decision that the central exam would include reading comprehension only, not least because this can be tested using automatically scored multiple choice questions at a high level of reliability. Unsurprisingly, the fact that reading counts for 50 % towards the final grade, teachers and students often spend substantial time on reading comprehension training in their final years (Fasoglio et al. 2015). In addition, this washback effect leads some students to think that reading is more important but also more difficult than other skills (Westhoff 2012).

3.3 Reforms since the 1990 s: communicative in English

In 1993 the Dutch government decided it was time for a substantial revision of junior secondary education, that is, for children aged 12 to 15. These reforms allowed for growing diversity in the languages on offer, including a stronger focus on a communicative approach. In the 1970 s, languages such as Spanish, Italian and Russian had been introduced as optional/elective school subjects. In 1990, also Turkish and Arabic, the languages of large immigrant populations in the Netherlands, were included to a very limited extent. Students could take exams in any of these languages, as long as they were offered in their school’s curriculum. At the same time, this reform put further pressure on MFLs other than English. In pre-vocational VMBO schools, English remained a compulsory subject, but children were now allowed to select only one other MFL. As the schools could decide what language(s) they offer, the students’ possibilities would often be limited to one other MFL (i. e., German, French OR Spanish) available at their institution.

In the pre-professional (HAVO) and pre-academic (VWO) tracks of secondary schools, the trio of compulsory English plus two MFLs (typically, French and German), remained mandatory in the first three years. Yet, in these higher-level tracks, changes for senior secondary education affected MFLs. In the late 1990 s, the aforementioned subject bundles -a relic of the Mammoetwet- were replaced by four profiles (see section above on the Dutch educational system). While all four profiles allowed taking additional languages next to English, only at the VWO level an additional foreign language (which could also be Latin or ancient Greek) remained as a compulsory exam subject. Especially at HAVO level, very few students opted for an additional language. From today’s European perspective, it seems ironic that this happened just before the EU implemented its policy for education in the mother tongue plus two additional MFLs in the early 2000 s.

3.4 The CEFR: language aims to be reached since the early 2000s

With the latest reform of secondary education (basisvorming) in 1993, English was established as a core subject next to Dutch and Mathematics (Fasoglio et al. 2015). In other words, today every Dutch adolescent has to finish high-school with an exam in English, turning it into a basic skill for a globalised world (cf. Lo Bianco 2014). For all MFLs, following the establishment of the CEFR (Council of Europe 2001), the CEFR attainment levels were implemented in the official Dutch exam syllabus for VMBO (in 2009), for HAVO (in 2010), and for VWO (in 2011) https://www.slo.nl/thema/vakspecifieke-thema/mvt/erk/, see the summary in Table 1. To support teachers and learners, the Taalprofielen [language profiles] were developed, that provide examples of language acts per CEFR level (https://www.slo.nl). Given that those goals are binding only for the central reading exams designed by CITO, schools might decide to aim higher or lower for their local school exams.

In general, higher levels are expected for receptive skills, and for English and German, because these languages are both typologically related to Dutch. For example, at the end of pre-vocational VMBO, students’ proficiency needs to be at B1 for English reading and listening but only at A2 for speaking and writing. For French, A2 is the goal for all skills. Pre-academic VWO students are expected to reach C1 in reading English/German, while receptive B2 and productive B1 suffices for French/Spanish. A disturbing fact is that large groups of learners do not reach these levels, particularly for writing: only 50 % of the final year students reach the target attainment levels for writing in English, German or French (Fasoglio et al. 2015). Interestingly, the speaking proficiency levels in English reach or exceed the target attainment levels in all school types: B1 instead of A2 in upper VMBO, B2 instead of B1 in HAVO, C1 instead of B2 in VWO (Fasoglio & Tuin, 2018). So far, no such studies have been conducted for speaking in German or French.

3.5 Bilingual schools: Content and language integrated learning (CLIL)

A different story pertains to bilingual schools, in Dutch tweetalig onderwijs (tto), where subjects are taught in the foreign language. Bilingual schools, the first of which were established in the early 1990 s, are supported and monitored by the Dutch organization for internationalization in education Nuffic. By now, there are more than 130 schools offering bilingual secondary education for all three scholastic tracks (i. e., 125 VWO, 72 HAVO, and 31 VMBO, cf. Nuffic 2019). Only two schools close to the German border have chosen for German, and all other tto schools use English as medium of instruction (de Graaff 2015). This means that subjects as diverse as history and maths (with the exception of other MFLs and Dutch) are taught in English. For HAVO and VWO a minimum of 50 % of the curriculum must be taught in the MFL in the first three years, for VMBO it is a minimum of 30 % (Mearns & de Graaff 2018). After three years of tto, students are expected to reach English B2 (VWO) or B1 (HAVO and VMBO; Nuffic, n.d.), often certified by external international exams (e. g., Cambridge). Teachers in tto tracks also require a B2 starting level, aiming for C1/C2. In addition, the teachers are being trained in bilingual pedagogy and in applying a content and language integrated learning (CLIL) approach (Coyle, Hood & Marsh 2010). Irrespective of the scholastic track, at the end of secondary school tto students are required to participate in the standard final examinations designed by CITO – which are in Dutch for all subjects. As a result, several tto schools offer the bilingual education at junior-secondary school, but switch to Dutch for senior-secondary school. Schools that remain in a tto-mode, can only offer English in a limited number of subjects (Nuffic n.d.). HAVO and VWO bilingual schools can offer the International Baccalaureate (IB) program in the upper forms for English.

3.6 Strengthening the trend: English at primary schools

At primary school level, foreign language teaching was made compulsory in 1986, as of when early English has been offered in the last two years of primary school (children aged 10 to 12). Again, the Dutch system does not provide extensive guidance on how to implement this, nor are there standard coursebooks, methods or tests that schools have to use. For many years, there was governmental advice to provide 80 hours of early English during the two compulsory years at the end of primary school. Accordingly, most schools implemented English for one hour a week, to help children with their entry into secondary school English classes. The teaching put emphasis on basic oral interaction with other speakers of English and understanding simple written information (Fasoglio et al. 2015). By now, more than 1000 out of 7000 schools have chosen to adopt an earlier start (for children aged 6–10), and some even start in the kindergarten years of primary education (age 4). From 2016 onwards, the government has allowed a maximum of 15 % of English in primary education, be it English lessons or other subjects taught in English.

Similar to secondary schools, bilingual education at primary schools holds different standards. In 2014, a pilot project started with 17 bilingual primary schools, where 30–50 % of the curriculum is taught in English. Results of accompanying research suggests that the English level of bilingual primary school (tpo) students is consistently higher than that of low-intensity early (age 4 or 6 start) or regular (age 10 start) students, although large individual differences exist, also between the teachers’ English level (Jenniskens et al. 2020).

Nowadays, some governmental guidance is available for each of the three approaches to early English (i. e., starting at age 4, 6 or 10, respectively): SLO has formulated core aims and gives examples of how those could be reached (tule.slo.nl/Engels/F-KDEngels.html). Similarly, CITO offers a (non-compulsory) assessment at the end of primary school that focuses on receptive skills at the A1 to A2 level. Every six years, the government arranges a study into the level of English at the end of primary education (see Inspectie van het onderwijs 2019, for results from 2012–2018).

3.7 Interim historical conclusion

To recap, we can see how building on a long tradition of French, German and English in compulsory secondary education, policy reforms starting in the 1970 s have put languages other than English under pressure. While all three languages are taught in junior secondary schools of HAVO/VWO, a second MFL next to English is a compulsory exam subject at the end of secondary school only for VWO students. This steady decline in status by policy (see Lanvers 2017 for a critical review highlighting similar trends for the UK) that threatens French and German as MFLs goes hand in hand with a growing dominance of English, both in education and society. This phenomenon is well reflected in the recent establishment of almost 150 bilingual primary/secondary schools – all but two offering English as the language of instruction. Against this background, we will discuss the challenges and opportunities Dutch MFL education is currently facing.

4.1 Primary education

After 30 years, English is well-established in Dutch primary schools. At this point, we would like to foreground three challenges: (1) drip-feed classes vs. out-of-school exposure; (2) growing socio-economic differences in times of superdiversity; (3) teacher language proficiency.

Compulsory English education starting in grade 5 or even earlier, builds on the idea of ‘the earlier the better’. Yet, apart from bilingual schools, current practice of early English delivers drip-feed education: a small amount of classroom time, typically one 45 minute lesson a week, provides a fairly limited amount of English exposure to children. Extensive research by Muñoz (2008, 2011) suggests that for children in an instructed setting this might not be the most effective way to learn language, because young learners draw much more on implicit processes, which need substantial amounts of exposure. Muñoz concluded that adolescents benefited more from the limited input in school context than younger children as they were cognitively more developed and could draw on more explicit processes for language learning. In the Dutch context, Unsworth et al. (2015) investigated vocabulary knowledge of young children (aged 4 and 5) and found small advantages for those receiving more than 60 minutes a week over those with only 45 minutes. Similarly, de Graaff (2015) and de Graaff and Costache (2020) report that children starting with early English in grade 1 reach higher levels at the end of primary school than those starting in grade 5. They highlight however, that the small differences might be overruled by school population and teaching quality differences, as well as individual differences in attitude and out-of-school exposure.

In the Netherlands, out-of-school exposure to English via different forms of (social) media is often substantial. Recent Flemish research, drawing on a comparable society in this respect, has shown that 10-to-12-year-olds that have not received any formal in-school instruction of English reach on average the A2 level (de Wilde et al. 2020). The authors hypothesize that the lack of in-class exposure might be partially compensated for by out-of-school English media consumption. Yet, an investigation commissioned by the Dutch government into the state of English at the end of primary school does not confirm this hypothesis (Inspectie van het Onderwijs 2019). Data of 2088 children showed that out-of-school use of English did only play a minor role in predicting students’ English levels and even more surprisingly, the primary-school foreign language curriculum (early English versus regular English) did not make a major difference when corrected for other factors. In contrast, individual differences were an important predictor with the school advice for the follow-up level of secondary education, a measure of scholastic aptitude, showing strongest effects.

Indeed, socio-economic status is likely to affect English learning at primary schools. While Dutch compulsory education is free of charge, most schools ask for a voluntary parental contribution, typically less than €100,– a year – yet, fees for elective bilingual education can be several hundreds (which at most schools may be covered by the school when parents cannot afford it). Consequently, the benefit of early immersive English may contribute to a socio-economic gap, as children attending bilingual schools are likely to be from more affluent households. Also, Jeffery and van Beuningen (2019) argue that in superdiverse Dutch society, where schools educate many children with a multilingual background, there is a need for more inclusive practices. There are successful initiatives that could serve as examples such as Euregio schools in the southern province of Limburg where German and French are taught at primary level (Claessen 2014), or the 3M-project that supports teachers in trilingual Dutch, Frisian, English primary schools to implement a translanguaging approach and value the linguistic contributions by migrant pupils (Duarte & Günther-van der Meij 2018).

Not least, teachers’ language proficiency is a concern in primary education. Dutch primary schools decide themselves who is teaching English: it might be the pupils’ regular teacher, a (native speaker) co-teacher or a trained MFL teacher. In other words, the person delivering early English has not always received training in foreign language pedagogy. To enter Dutch primary school teacher training, students need English B1 but there are no standards for the end of that education (Fasoglio et al. 2015). Therefore, it is currently unclear what the English level of primary school teachers is, including those who are tasked to deliver English lessons. HBOs offering primary school teacher training provide comprehensive programs that cover all subjects, of which English as MFL forms a small and at times only optional part. It is ensuring that more recently, some institutes offer an elective minor program for Early English teaching. Unsworth et al. (2015) found that teachers’ language proficiency explained 31 % of variation in the attainment of the young learners in their study, while de Graaff (2015) and de Graaff and Ostache (2020) argue that the didactic skills of a teacher are at least as important.

In summary, the major challenge for English at Dutch primary schools is to ensure that pupils receive high-level language education, and that irrespective of socio-economic status, all students will learn English to such an extent that they can participate in secondary schools.

4.2 Secondary education

At the level of secondary schools, there are many intricacies affecting all MFLs that are taught, be it as mandatory subjects in junior secondary education or as elective in senior secondary schools. Like in other parts of the world, the most prominent challenge is that the dominance of English takes a toll on other MFLs. In the Netherlands, it is a chicken and egg issue whether the declining status of other MFLs in secondary schools that is reflected in the educational policy changes over the past decades is a reaction to its perceived role in society or a result of utilitarian Dutch policies. The facts are that the number of students finishing high-school with an exam in English has grown from almost 95 % in the 1980 s to compulsory 100 % now. In contrast, figures for German and French show a steady decline from almost 60 % and 30 % in 1980 to 32 % and 16 % for German and French, respectively (cf. Voogel 2016). A growing group of students opting for Spanish (2 % according to Proft, Böttger & Ide 2016) does not counter the development that the English-only attitude goes hand in hand with declining plurilingualism (Lanvers et al. 2019).

We will discuss four developments that are currently debated in the Dutch context around the interrelated factors of (1) student motivation to learn other languages than English; (2) the washback effect of current practice in MFL assessment; (3) the target language as language of instruction; (4) the content of MFL classes.

Apart from its dominance, reasons that are given for the low uptake of languages other than English relate to myths such as, English is easier than French and German, and that for future careers those languages are less meaningful (cf. www.buurtaalonderwijs.nl). The former seems not to be warranted for German, which is a typologically close linguistic neighbor of Dutch (Roelands & ten Thije 2006). Indeed, students of all scholastic tracks report that they find German easy (van Dée et al. 2017). The latter myth stands in stark contrast to Dutch reality: Germany, Belgium, the UK, and France are the top-four trade partners (in this order, cf., CBS 2018). Good social relationships that are established via a trading partner’s language (i. e., French, German) serve as an important factor for healthy economic relationships, which is unlikely to be compensated for by English as a lingua franca. In addition, Lanvers (2017) suggests that a mere focus on utilitarian needs might be counterproductive to positive MFL attitudes. Yet, the vicious circle is further fed by schools deciding to dedicate fewer hours a week to other languages, particularly to French. Consequently, students reach lower proficiency levels, and feelings of low self-efficacy tend to lead to a declining motivation to take non-mandatory classes in senior secondary school (Voogel 2018). Furthermore, governmental campaigns to strengthen STEM profiles in senior secondary schools (and the fact that these profiles allow for greater choice of follow-up university programs) meant that the culture and society profile with a mandatory additional MFL lost ground at VWO level (only 9.8 % of students in 2018 opted for the culture and society profile, NPT 2019).

In 2018, teachers, teacher trainers and universities offering French and German joined forces in the organisation Visiegroep Buurtalen (Voogel 2018). Their manifesto, which was signed by, among others, the Goethe Institute and the chamber of commerce for the French industry in the Netherlands, lobbies for more attention to and a stronger position of French and German in Dutch secondary education. Given that the government is currently preparing a curriculum reform (cf. www.curriculum.nu) we might be cautiously hopeful that the future will guide more young people to aim for high proficiency in languages other than English be it traditional MFLs like German and French or upcoming Spanish and Chinese.

A recurring discussion of Dutch MFL secondary school education is the negative washback effect of the national central exams. As mentioned before, the standardized CITO exams (determining 50 % of the final grade) focus on reading comprehension tested through multiple-choice questions. Accordingly, a considerable part of MFL teaching has turned into an exam training for a multiple choice test, which scores low on student and teacher enjoyment (van Dée et al. 2017; Voogel 2018). Teachers complain that the emphasis on reading comes at the cost of the other skills, particularly, oral interaction, and leaves little room for discussing content related to language awareness, the target language society, culture and literature (van Dée et al. 2017).

Generally, productive target language use, including teachers speaking the language they teach, seems to be a problematic issue in the Netherlands (see e. g., Dönszelmann et al. 2016; West & Verspoor, 2017). In van Dée et al. (2017) only 29 % of students learning German at HAVO/VWO and 36 to 40 % at VMBO report to use the target language often in class – meaning that the majority of students does not do so. More detailed figures reveal that students use the target language during 5 to 10 % of a lesson only, that is, about 2 to 4 minutes per lesson.

Doeltaal is voertaal (freely translated as ‘using the target language’) initiatives train teachers to employ the target language consistently in class (e. g., Dönszelmann et al. 2016). It seems that teachers often keep target language use to social talk (e. g., opening/ending of class) while other instructions (e. g., grammar) are provided in Dutch (Haijma 2013) – a behaviour that is further induced by coursebooks printing instructions in Dutch.

Others try to challenge the prevailing tradition of explicit grammar instruction, which naturally induces more target language use. Communicative input-based pedagogy, such as AIM for French (Gombert et al. 2018; Rousse-Malpat et al. 2019), implicit instruction (e. g., Piggott et al. 2020), and job-specific interactive tasks at VMBO (van Batenburg et al. 2019) show first positive results: teachers and students are more confident with using the target language and productive language is authentically more fluent and more lexically diverse (accuracy measures provide a mixed picture). These continuous efforts will hopefully make classrooms more communicative so that secondary school pupils and their teachers gain first-hand experience with target language use, raising attainment levels, self-efficacy and motivation.

The aforementioned current curriculum.nu reform draws on a bottom-up approach (www.curriculum.nu/curriculum-nu-english): with advice from SLO experts, teams of teachers and school leaders have created a first draft of the main building blocks of the future Dutch curriculum, and now core objectives and attainment levels for the final exams are being formulated. Critical review sessions invite teachers, teacher trainers, testing and curriculum experts and researchers to provide feedback. After adjusting the exam program to these new objectives, piloting in some schools will take place in 2022/23. For the domain of English and other MFLs the Visiegroep Buurtalen is involved, as is the Meesterschapsteam, an inter-university group of experts in foreign language literature, second language acquisition and language pedagogy. Among others, they stress the need for more engaging and challenging content (Meesterschapsteam mvt, 2018). In particular for the pre-professional (HAVO) and pre-academic (VWO) tracks, language and culture related content taught using CLIL methodology (Mearns & Platteel 2020) might inspire more students to choose a MFL for senior secondary school.

To recap, there are manifold challenges for MFL education at secondary schools where the majority of MFL instruction takes place. Attempts to revalue languages other than English and to increase target language use through meaningful (oral) interaction are ongoing. The current major reform by curriculum.nu opens exciting avenues for the future of MFLs in order to attract more students to learn a foreign language, also at tertiary education.

4.3 Tertiary education

As shown in figure 1, tertiary education in the Netherlands is organized along the same three tracks of scholastic achievement as secondary schools, that is, vocational (MBO), applied-professional (HBO) and academic (universities). We will sketch currently debated issues around MFL teaching for each of them.