Print   |   Save

CLIL in Germany: Results from Recent Research in a Contested Field of Education

Stephan Breidbach
Humboldt University Berlin

Britta Viebrock
Goethe University Frankfurt (Main)


The implementation of CLIL in Germany can be described as a grassroots development with the first (German-French) programmes dating back to the 1960s. English-speaking CLIL programmes were established on a large scale in the 1990s. CLIL-specific research would only become a full-fledged field of research after the start of the new millennium. The first empirical research projects were driven by specialists in foreign language pedagogy who initially focussed on researching language acquisition. In more recent years, CLIL research has diversified and tackled a number of questions such as subject-specific concepts, literacy, study skills, motivation, intercultural learning, stakeholder research, underachievers, the linguistic diversity of CLIL learners, modularised CLIL concepts, mental processes as well as younger learners to name but a few. In this article, we will briefly present the most important findings and discuss some of the methodology of the respective research projects.


Keywords: CLIL research, methodology, concepts and outcomes, CLIL in the German school system, the history of CLIL in Germany

1. What one should know about contexts of CLIL in Germany

This article reviews recent CLIL research in Germany – most of which is published in German – in order to make it accessible to a wider audience. Many of the research results yield insights that may be of value beyond the particular institutional context they are derived from and beyond German contexts.1 Nevertheless, the German context is briefly explained in order to make German CLIL research, its background, direction, questions and the reception of results more accessible to an international audience. We will therefore start by looking at the historical and ideological background of CLIL and at some key characteristics of the German school system.

The historical and ideological underpinnings of the introduction of what was called in the early days ‘bilingual education’2 were the result of the political endeavours of the French and German governments in the 1960s which sought after World War II to foster reconciliation between the two countries. Following the Franco-German Treaty of 1963, CLIL programmes were introduced in Germany aiming at the promotion of international understanding. This explains why the schools with the longest tradition in CLIL are mainly situated in the West of Germany along the River Rhine. These schools mainly offered CLIL in French. With the increasing importance of English as an international language, the 1990s can be considered the heyday of the implementation of English-speaking CLIL programmes. Currently, more than 700 schools offer CLIL in various forms (cf. Werner 2009).

Public CLIL discussion and also some CLIL research are closely linked with the ongoing debate about the German school system and one of its characteristic features – the diversity of school types in secondary education. While primary education for the vast majority of children takes place in comprehensive primary schools until Grade 4, in most German states, learners are streamlined into up to four different main types of schools. These types are considered to represent three different levels of intellectual demand with the Hauptschule at the bottom, Realschule at a middle-level and the Gymnasium at the top. The Gesamtschule (comprehensive schools), a remnant from the reform-oriented 1970s, comprises all three levels within one single institution.

Page 5
Page 6

Among these four school types to be found in secondary education, it is the Gymnasium (alongside comprehensive schools with upper secondary programmes) that offers learners the general university entrance qualification, the Abitur. This school type has an intake of about 45% of the annual cohort of learners. Until recently, after four years of primary education, lower and upper secondary education spanned a total of nine years (from Grade 5 until Grade 13). At present, learners take their Abitur-exams at the end of Grade 12.

Initially, CLIL programmes were predominantly offered to those perceived as the more able learners attending the German Gymnasium. Not only is the Gymnasium the typical type of school offering CLIL but the Gymnasium is also typically sought after by the well-established middle class. It is common knowledge – and has repeatedly been substantiated through findings in several rounds of the PISA study – that children from non-migrant, middle-class households are more likely to access a Gymnasium education than children from less wealthy and migrant backgrounds.

This situation has led to two peculiarities which by and large characterise the German CLIL debate. While some researchers approach the structural elitism of CLIL in tentative moves (e.g. Zydatiß 2007) and official policies tacitly acknowledge the elitist undercurrent, the general perception of CLIL as a “success story” has led school authorities to implement CLIL even in school types formerly assumed to be inappropriate for CLIL, such as primary schools. CLIL has clearly reached the political agendas of local governments as can be seen in programmes such as the North Rhine-Westphalian “CLIL for all”. In adopting policies of comprehensive CLIL, politicians just as much as some researchers downplay the fact to the point of omission that existing experiences with CLIL have mostly involved privileged learners from middle-class backgrounds (see also section 2 and Apsel in this volume).

It is true that now some schools from the middle and bottom tier (Realschule and Hauptschule) and also some vocational schools offer CLIL, however, these can certainly not be considered mainstream cases, and they often feature an experimental character. CLIL in Germany has traditionally been regarded as one way of ‘upgrading’ the top tier schools, which is why research is often concerned with issues of comparability of outcomes between mainstream and CLIL-stream learners, but also why CLIL is often considered elitist and why political initiatives such as “CLIL for all” are potential sources of debate.

The second peculiarity amounts to what must be called a split-consciousness phenomenon since CLIL learners in Germany can be divided into two groups: the first group comprises learners who receive parts of their subject-matter instruction through a modern foreign language, which is mainly English and French and, to a lesser extent, Spanish and other European languages and mostly within so-called CLIL-streams at ordinary secondary schools. Increasingly, also primary schools are offering CLIL (cf. Massler in this volume, Putsche 2011). The second group consists of speakers of German as a second or further language and who are taught almost the entire curriculum through German in mainstream schooling (leaving aside the subjects of ancient and modern languages). This group is much bigger in numbers but far less structured in terms of organised CLIL.

Against this background, it is important to be aware that CLIL in Germany has been and still is often framed within the context of foreign language learning, in particular at Gymnasium-types schools. This is true for the public discussion and the academic discourse alike (e.g. Mentz 2001). As these schools are predominantly recruiting children – and have been doing even more so in the past – from mainly monolingual, German-speaking families, speakers of other languages, who study the regular curriculum through German (which is in fact similar to unstructured, late partial immersion, have in fact never entered the picture of the official CLIL discourse. Research communities in German as a second language and CLIL in Germany almost work in isolation from each other, even though there is common ground, as is clearly argued in a special issue of Fremdsprache Deutsch (Haataja/Wicke 2009) or in the seminal work by Leisen et al. (1999) on scaffolding techniques for teaching learners of German in content-based classrooms. From the perspective of German as a second language, Pietzuch (2008) reads Breidbach’s (2007) work on reflective CLIL pedagogy as a common theoretical base for both discourse communities to build on. Empirical evidence that multilingual learners may also be good foreign language learners – and may thus to a certain extent be expected to be successful CLIL learners has been produced in a representative large-scale study of 15-year old learners from all types of schools (DESI 2008, see section 8).

Page 6
Page 7

In conceptual terms, the CLIL approach was usually implemented as some kind of additive late partial immersion starting with one subject in Grade 7 and then continuously adding or alternating with other subjects in the following years. In addition, a modularised approach can be found (e.g. Krechel 1999), where only selected topics were taught in a foreign language over a limited period of time. Not surprisingly, much of the early research had been inspired by the institutional arrangements (including their shortcomings). In the past few years, the conceptual settings have diversified as much as the research questions. In the following chapters we will review selected CLIL research results in the following areas: foreign language acquisition, classroom discourse and subject-specific knowledge, study skills, “intercultural” learning, motivation, stakeholder research as well as more recent developments – the focus on underachievers and the linguistic diversity in the classroom.

The Franco-German reconciliation policies pursued in the 1960s also account for a strong leaning within the CLIL curriculum towards the social sciences-related subjects. There is also a rather rigid set of prototypical CLIL subjects, which usually include History and Geography, and to a lesser extent comprising Social and Political Studies. The Natural Sciences as well as Art, Music, Drama and Physical Education have only caught up much later, usually represented through Biology, and have not passed an experimental stage in many places. Unlike in other European countries, Maths has been broadly neglected – if not actively excluded – from CLIL and has only recently been considered more seriously (cf. Prüfer forthcoming, Viebrock 2009a, b). The latter certainly has to do with two deeply-rooted folk-theories, namely about the difficulty of Maths which ought not to be “increased” by teaching the subject through a foreign language, and the other one which holds that maths is a non-linguistic subject which implies that language does not play any particular role in the Maths-classroom. Last but not least, the structure of teacher education in Germany needs to be considered as an explanation for the underrepresentation of the Sciences and Maths in CLIL programmes: unlike in many European countries, German secondary teachers receive formal training in (at least) two subjects. CLIL teachers are usually recruited from teachers combining a language and a CLIL subject. One reason why schools offering Maths through an additional language are rare in Germany is that only few teachers combine languages and Maths or Science. In some states, such as Bavaria, teaching degree courses at universities do not even allow for someone to specialise in a language and a content subject such as Maths. It is therefore not unusual to find the question of the suitability of individual subjects and languages for CLIL raised in introductory textbooks on CLIL (Mentz 2010) or full- doctoral dissertations (Rymarczik 2003, Witzigmann 2011).

2. Learning and acquiring a foreign language in and through CLIL

The first large-scale research projects in CLIL-contexts were conducted by trained linguists. This may help to explain the strong focus on issues of foreign language acquisition, linguistic competence or language awareness. We emphasise this at the beginning of this section for two reasons: Firstly, CLIL-research has not been able to gain an equally strong foothold in the didactics of any one content subject3, secondly, CLIL is still more often than not presented as a method of foreign language learning, as the following quotation from a course description of an English department at a German university for the winter term 2011-12 illustrates:

"Bilingualer Sachfachunterricht" as a form of "content and language integrated learning (CLIL)" is being implemented in many secondary schools in Germany today, and a growing number of studies seem to confirm it as a valuable addition to more traditional programmes of language teaching. 
The course is intended as an introduction to this field of teaching certain school subjects (providing examples from a range of subjects, e.g. geography, history, politics and biology) through the medium of English. We are going to investigate the many facets of CLIL in the context of a modern understanding of English language teaching. (…)

The course description repeats the single most regularly quoted selling point of CLIL in Germany. Indeed, a number of studies show CLIL learners to be language proficient if compared to their non-CLIL peers. A closer look at the linguistic CLIL research, however, yields a more diversified picture. We will therefore look at a number of recent research projects, both completed and currently ongoing.

A comprehensive study on the development of foreign language competence has been carried out in the late 1990s by Bredenbröker (2000). Following a comparative research design, he examined 195 CLIL and non-CLIL (i.e. regular foreign language) learners over a period of two years. Employing a C-test, a grammar test and a test on reading comprehension at the beginning, after one year and after two years of CLIL instruction, Bredenbröker found that CLIL had a very positive influence on foreign language competence in general, which was most pronounced for reading comprehension and the ability to use elaborate strategies (such as tolerance of ambiguity, appropriate strategies of inference etc.). Differences in the grammar test were less pronounced, which could be attributed to the fact that regular foreign language teaching is often more concerned with an explicit focus on grammar, whereas CLIL is more concerned with implicit grammatical knowledge, which is acquired in the process of exchanging subject-specific information.

Page 7
Page 8

Bredenbröker’s results have to be discussed critically with respect to the results of the first battery of tests, which already showed statistically significant differences in language competence in favour of the CLIL learners. Here, the structural selectivity of CLIL appears to have a greater impact on student achievement than CLIL itself has on student achievement. In addition, it is questionable whether the language tests mentioned indeed examine foreign language competence as an underlying mental concept or whether they rather allow insights into language performance on the surface level.

This line of research is continued in projects by Rumlich (forthcoming) and Grum (2012). Rumlich examines the linguistic accuracy of pupils’ written products. With a first case study (pre-test) relying on statistical analysis, he is able to prove that CLIL learners’ accuracy in written texts is significantly higher than those of non-CLIL learners. However, Rumlich concedes that the participants examined in a comparative approach were hardly similar in terms of exposure to CLIL (subjects and teaching hours), teacher influences, learner selection etc. Attributing the positive results to CLIL alone is thus not possible. More substantial evidence can be provided by his ongoing large-scale study with a controlled population.

Grum analyses data from a sample of 90 learners from CLIL and non-CLIL classes in Berlin regarding the learners’ ability to respond to a complex communicative oral task. Grum finds that CLIL learners perform significantly better than their non-CLIL peers in terms of lexical scope and fluency. While Grum concludes that her results cast a shadow of doubt over the general quality of regular language teaching, she is hesitant to praise CLIL for more than its apparent potential for language acquisition. She does so probably for a good reason since her results about general language use are at odds with findings from studies researching discourse-specific language use in CLIL classrooms teacher (Dalton-Puffer 2007) and learner talk (Lose 2007), who both state that language functions that serve as tools for thought (e.g. hypothesising, explaining, etc.) are very rarely to be found.

The concept of language awareness in CLIL has been examined by Morkötter (2002) and Fehling (2008). In Morkötter’s study, CLIL was studied as a minor aspect within a general focus on learners’ attitudes towards multilingualism. Due to the qualitative set-up of her study, the results are certainly not generalisable. However, she hypothesises that English-speaking CLIL does not necessarily support the dominance of English as an international language as CLIL learners in her study claim that multilingualism (extending over proficiency in English as a foreign language only) is favourable with respect to future professional perspectives.

The comparative approach employed by Fehling again reflects the attempt to prove the added value or even superiority of CLIL, which is characteristic of early research projects. By now, the conviction has gained general acceptance that CLIL is an independent approach of teaching, which does not automatically lead to quality learning and has to face similar pedagogical questions that all teaching is concerned with. This view is reflected in the recent tendency towards CLIL-specific non-comparative research designs.

Returning to Fehling’s study, within her setting she could show that the CLIL learners she investigated possess a more highly developed language awareness compared to non-CLIL learners within her sample. Using James’ and Garret’s (1991) model of language awareness Fehling finds the advantages of the CLIL learners particularly obvious in the affective dimension, which looks at the learners’ attitudes and engagement with the foreign language in question (English in this case). The social dimension, which assesses the learners’ thoughts and feelings towards foreign language learning in general, shows more differentiated results: whereas a rise in interest in languages such as Spanish and Italian can be observed in non-CLIL learners, the CLIL learners’ motivation concerning French decreases rapidly. Reverse results could be found with regard to the language of instruction in CLIL: the high motivation of CLIL learners for learning English remains constant while the non-CLIL learners’ motivation drops over time. In the cognitive dimension of language awareness, which measures the learners’ capacity to infer unknown vocabulary, verbalise the strategies employed and deduce more general rules, CLIL learners significantly outperform non-CLIL learners. In addition, they are able to articulate more and more complex strategies. In contrast to these results, Fehling hardly found any difference between CLIL and non-CLIL learners in the political dimension, which looks at the learners’ ability to comment on different text genres. However, the general level of language awareness in both sub-samples is markedly low.

Page 8
Page 9

Similar to Bredenbröker’s study mentioned above, the learners in Fehling’s project show initial and continuing differences concerning their cognitive abilities/linguistic intelligence with the CLIL learners starting on a higher level and significantly improving their aptitude. However, a considerable variance can be observed in both groups tested by Fehling, which precipitates the question whether a more carefully selected sample would have provided insights that could have been more directly linked to CLIL-specific influences. Nevertheless, Fehling’s relatively early discovery of the relative absence of critical language awareness foreshadowed similar results by Bonnet (2004) or Lamsfuß-Schenk (2008), who both find little if any empirical evidence that CLIL learners – being in this very similar to their non-CLIL peers – develop reflective or critical competences all by themselves. In this, empirical research has certainly rocked the foundations of one of the chief arguments circulated in the German CLIL discourse in favour of CLIL, namely that the CLIL setting all by itself will cause learners to take a reflective perspective on content matter and the cultures behind the working language. Such an inbuilt mechanism does not appear to exist in CLIL settings.

3. Classroom discourse, subject-specific knowledge and cognitive “peculiarities” of CLIL learning

In the German discourse, the CLIL approach is often criticised from the perspective of non-language subject teachers, who fear substantial negative influences on the learners’ subject matter competences and their literacy development. In the early debate, a strong opposition between (foreign) language learning and subject-specific knowledge was created. Against this, advocates of CLIL – quite often with an odd-angled reference to the language-across-the-curriculum tradition – stipulate that all learning is dependent on language. In fact, both opposing positions remained insufficiently theorised until the mid-1990s.

A significant shift within this debate occurred when the notion of different forms of knowledge having corresponding realisations in language entered the discourse in the form of Cummins’ (1984) distinction between BICS and CALP (for a detailed analysis and historical account cf. Breidbach 2007). BICS (basic interpersonal communication skills) refers to the kind of language proficiency needed in context-embedded, cognitively undemanding everyday situations, while CALP (cognitive academic language proficiency) describes language proficiency needed in decontextualised, cognitively demanding academic discourse. Becoming highly influential in CLIL pedagogy, and research, this distinction served as model for the examination of subject-specific knowledge and the learners’ abilities to participate in subject-specific discourse.

Zydatiß (2007 and in this volume) carried out a large-scale study based on testing and descriptive statistics. Originally, he set out to comparatively evaluate the implementation and success of CLIL in Berlin, but his results are of a much more general nature. Employing two language tests based on the BICS/CALP distinction for proficiency as well as achievement, and a test on subject matter literacy, Zydatiß provides statistically significant correlations between the learners’ language competences (especially lexico-grammar and/or general proficiency) and their academic task performance. Relating to Cummins’ studies in immersion settings, Zydatiß suggests “that there may be a double language threshold (a lower one and an upper one) which acts as an interven­ing variable towards transferable academic discourse competencies; accounting thereby for learners' low or high scores in a test setting which tries to tap the underlying con­struct of a general academic proficiency relevant to German CLIL classrooms” (Zydatiß in this volume, cf. also for an in-depth discussion).

Referring to Halliday’s socio-semiotic approach, considerations on academic discourse functions have renewed the discussion lately (cf. Vollmer 2009, Zydatiß 2005). Discourse functions are understood as integral units that join conceptual comprehension, thought and verbalisation processes, e.g., describe, explain, evaluate and other similar terms assignable to these categories. Their (insufficient) realisation has been analysed in classroom discourse by focusing on the socio-cultural context (Dalton-Puffer 2007) and in written products by focusing on individual cognitive development (Dielmann 2007, Coetzee 2007, Vollmer 2009, Heine 2010). In his research, Vollmer comes to the conclusion that the learners’ written products do not satisfy the discipline’s discourse conventions. Analysing Austrian CLIL classrooms, Dalton-Puffer (2007) provides supporting evidence by revealing that classroom interaction does not offer enough opportunities to develop adequate thematic and rhetorical structures. The teacher as a language model often does not provide the learner with the necessary input in the first place.

In her study on problem-solving activities in Geography, Heine (2010) examines the cognitive impact of a foreign working language on subject-specific tasks and study skills. In a comparative approach with samples from CLIL learners and non-CLIL learners and relying on thinking-aloud protocols, she demonstrates a more profound semantic processing by CLIL learners, which is usually triggered by insufficient word knowledge. In the process of verbalising concepts or thoughts, CLIL learners have to put in more effort in finding adequate expressions. Often several cycles of testing alternative phrasings are necessary. This focus on language leads to more intense conceptual reflections. Linking her findings to the general discourse on the cognitive interdependency of language and thought, Heine suggests the possibility of slightly different cognitive structures if conceptual knowledge is developed in the L2. However, a more important outcome of her study is certainly a general plea for a focus on language, alternative phrasings and ways of expressions in any kind of subject-matter teaching.

Page 9
Page 10

Bonnet (2004) showed in his study on CLIL learners’ scientific literacy that CLIL learners achieved comparable competence levels to those of non-CLIL learners. In a quasi-experimental setting, Bonnet had his CLIL learners of Chemistry work on chemical problems and tasks in small groups and negotiate meaning through focused discussion. In the analysis, the centre of attention was on the way the learners employ language in the process of negotiating (scientific) meaning. While Bonnet’s data sample included numerous instances of code switching, the analysis revealed that falling back on the learners’ mother tongue (German in this case) mainly occurred due to a not yet fully developed language competence. It certainly did not help to overcome any conceptual problems: “I don’t even know what this [chemical phenomenon] is supposed to mean in German” is a typical quotation illustrating the learners’ attitude during instances of code-switching in Bonnet’s study.

Bonnet’s study also provided a fresh methodological perspective as he reconstructed learners’ language and subject-matter competences from problem-solving learner-to-learner interaction rather than measure learning outcomes through ex-post tests or evaluation of finished learner products, which are generally preferred in quantitative research designs. Bonnet thus highlights the significance of the learning process for empirical research in education. Distinguishing between an “academic task structure” (the complexity of the task) and a “social participation structure” (the involvement of the learners), Bonnet reveals that processes of social participation, such as the learner's role in a small group discussion and the established rules of distribution and partaking, exert a much stronger influence on possible learning outcomes than is usually considered in CLIL pedagogy. He stresses that these influences might even exceed the cognitive demands, which are usually understood to account for the lack of successful learning. This view may be supported by results of an empirical study from the Swiss CLIL context (Badertscher, Bieri 2009), where the authors find CLIL learners generally performing on par with with the non-CLIL control group in terms of conceptual knowledge, even though L2 performance of the CLIL learners cannot match general L1 performance of the control group learners. Stohler and Kiss (2010), working on the same set of data as Badertscher and Bieri, conclude that particular process dimensions of CLIL classrooms may account for these findings (cf. also Bonnet in this volume).

Such processes are investigated in recent studies by Wannagat (2010), Pitsch (2010) and Wegner (2010). While Wannagat contrasts teaching styles of Hong Kong EMI (English as a medium instruction) and German CLIL History teachers, focusing on the proportions of teacher and student talk or the de facto usage of English or another language in the classroom, Pitsch draws attention to a wide range of non-verbal, symbolic meaning-making going on in a CLIL History lesson. Wegner’s study links with Bonnet’s and Dalton-Puffer’s findings on the significance of substantial talk about real subject-related problems in the CLIL classroom and learners being able to engage in what Bonnet calls “intensive negotiation of meaning” as a precondition for quality learning processes. Wegner, in a micro-analysis of one Grade seven and one Grade nine CLIL class in Politics (in a German Realschule), shows that both teachers and learners construct through their subjective beliefs and diverging mindsets hardly compatible, if not altogether opposing ideas about what to expect from CLIL. In triangulating classroom-observations, retrospective interviews with teachers and learners, and group discussions with learners, Wegner summarises the two cases of CLIL Politics as a chain of missed opportunities for substantial interaction. The main problem Wegner identifies is not that learners and teachers do not have any clear-cut expectations from CLIL, but find it difficult to transform them into mutually negotiable and hence viable classroom practice.

4. “Intercultural” learning

Intercultural learning, though a highly disputable concept, has featured strongly in the international discussion as one of the central objectives of the CLIL approach (e.g. Coyle et al. 2010). One of the most influential models in Germany was Hallet’s so-called “bilingual triangle”, which particularly aimed at the development of insights into and knowledge of “foreign languages and cultures in as many domains of society as possible” (Hallet 1999: 24; our translation) and included the development of the respective cognitive academic language proficiency (CALP). The three competence areas of the triangle include “one’s own country and culture”, “other countries and cultures” as well as the “intercommunity”, i.e. global and universal phenomena independent of a particular cultural context. Hallet (ibid.: 25) mentions examples such as global warming, peace keeping missions, or laws of nature. It must be said, however, that this model has so far not been analysed for its effects on learners.

Page 10
Page 11

At the time of its publication, Hallet’s model was widely adopted since it appeared to fill a large conceptual void. Yet, this model was criticised for its highly problematic notion of culture, which rather statically equates culture and nation, consolidates the dichotomy between “one’s own” and “the foreign” and perpetuates an essentialist concept. The adequacy of such a notion of culture has been questioned for quite some time (e.g. Hu 1996). As a reaction, in the German discourse on CLIL pedagogy a stronger focus on the plurality of perspectives and the necessity to compare, contrast and coordinate perspectives can be found. As a result following from the findings of her study, Lamsfuß-Schenk (2008) stresses that intercultural learning in CLIL does not mean the comparison of two whole entities, but rather the creation of a more or less “complete” and sufficiently differentiated picture from various perspectives. According to Lamsfuß-Schenk, who conducted a one-year action research study on her self-taught CLIL History class in French, such a picture has to be carefully established in a reflected pedagogic design of CLIL. Teaching the standard History curriculum through French alone will not suffice. Kollenrott (2008) supports these results with her study of the potential of CLIL History lessons for intercultural learning. She comes to the conclusion that theoretical concepts hardly resonate in practical classroom activities at all.

Regarding the notion of intercultural learning, an entirely new twist to the discussion has been suggested by Bonnet (2000) who introduced a culturalistic view of the discipline: i.e. chemistry as culture or political science as culture. Instead of distinguishing between national approaches, i.e. a German, a French and a British chemistry, being unfamiliar with the “culture of chemistry” (often termed the “nature of chemistry” in the Anglo-Saxon discourse), is placed at the centre of attention (cf. ibid.: 153). The “culture and/or nature of chemistry” can be understood as the discipline's naturally limited but specifically focused perspective on the world and its particular ways of creating, documenting and communicating knowledge, all of which are conventional. Intercultural learning in this sense deals much more with the question of how novices can be initiated into the language and the culture of a discipline.

Building on this, an integrated model of cognitive and social cultural difference has been suggested by Breidbach (2007). Central to this model is the notion of the identity of learners, which pedagogic designs for teaching content and culture in CLIL need to take into account for learners to be able to engage in negotiations of meaning. Breidbach suggests that learners may position themselves vis-à-vis the content either as cognitive novices (see above) or as social “in”- or “outsiders” who position themselves in a social field either within or without a social group. Thus, CLIL learners can be new to a discipline like Chemistry or Economics and its wondrous ways of conceptualising the world, or they can be new to a collective when being confronted with e.g. national views on History. Breidbach argues that both dimensions theoretically afford ample opportunities for reflection on the nature either of disciplinary knowledge or of collective meaning-making and both may eventually lead to critical reflective awareness in learners. Judging not only from failures in empirical research to find reflective learning in situ but also from the viewpoint of a philosophy of education, reflective learning must be part of teachers’ pedagogic agendas and of their teaching methods in order to actually take place. If reflective learning is not actively fostered, it is most likely that the structural conservatism of the educational systems and its established non-dialogic interactional structures will make any coincidental occurrence of reflective learning highly unlikely.

In conclusion, the notion of intercultural learning has been drawn on extensively in the process of legimatising the CLIL approach in Germany. In empirical research it has not been examined satisfactorily, which may well be due to simplified or even simplistic concepts that are often utilised in this field (cf. Weber 1993, Göbel and Hesse 2008).

5. Learner Motivation

Motivation is among the marginally examined concepts in German-speaking CLIL research. Dagmar Abendroth-Timmer’s study (2007) is an exception, but due to its specific set-up the results are predominantly of value for modularised CLIL programmes, i.e. the temporary teaching of topical units in the foreign language. In her study, Abendroth-Timmer categorises the learners into language smart vs. content smart types (based on their own judgment) and analyses their motivation development in modularised (Spanish) CLIL programmes. It is particularly interesting that the language of interaction in these modules was not spoken fluently by all learners. One of the underlying objectives of this study was to show the effects of CLIL modules on the general motivation to learn languages among the learners investigated. This study is typical of a larger set of (international) research studies, which look at the potential of CLIL for the promotion of supposedly unpopular subjects and/or languages (cf. the contributions by Hansen-Pauly, Lochtmann, and Prüfer in Breidbach and Viebrock [forthcoming 2012]).

The results of Abendroth-Timmers study present a differentiated and rather hopeful picture. Even learners who do not consider themselves to be language smart support the modularised approach and clearly seem to benefit from it. All learners corroborate their ability to decode technical or academic texts even with limited language competences, be it because they do not actually speak the language of instruction (but a related language) or because they are beginning foreign language learners. A modularised CLIL approach can certainly be considered to be functional on a receptive level.

Page 11
Page 12

As a conclusion, Abendroth-Timmer advocates the extension of a modularised CLIL approach and places it within a concept of multilingual (foreign language) teaching. This also allows for the integration of the learners’ heritage languages and hence explicit work on the learners’ linguistic as well as cultural identity. Through the dominance of the English language in international communication, which is also reflected in the massive expansion of English-speaking CLIL programmes in Germany in the 1990s, this aspect is often neglected in research focusing on English-speaking CLIL programmes.

6. Stakeholder research

The last example mentioned in the previous section already indicates that stakeholder research has been of much interest in the German-speaking world over the last decade with the focus mainly on classroom agents (teachers and learners). The majority of the projects, however, followed qualitative research designs (cf. Dirks 2004, Massler in this volume, Meyer 2003, Viebrock 2007, forthcoming, Wegner in this volume). In a way, this relative accumulation can be interpreted as a (albeit rarely explicit) critical statement against the standardisation processes that have taken place in the German educational system recently, which are entirely output-oriented and do not pay sufficient attention to how these outcomes are achieved and how this affects the central agents involved.

Whereas Meyer’s research project is comparable to Abendroth-Timmer’s study mentioned above inasmuch as it looks at the acceptance of the CLIL approach from the learners’ point of view, Dirks and Viebrock explicitly look at the teacher’s perspective. In her study, Dirks employs a biographical approach and manages to identify two types of teachers concerning their constitution and modes of operation, which she classifies as “guardians of tradition/culture” and “traveller between worlds” (Dirks 2004: 132 ff.; our translation). They display characteristic features pertaining to their motivation for CLIL teaching, typical structures of classroom interaction and other aspects. Whereas the first type is predominantly concerned with the transgenerational transmission of knowledge structured, established and authorised by academic disciplines, the latter utilises the CLIL approach for the transgression of borders between disciplines, which is reflected in methodological and topical diversity in the classroom including many process-oriented procedures and a highly communicative setting.

In contrast to identifying teacher types, Viebrock’s research is concerned with the individual teacher’s mindset and its relevance for classroom behaviour. Teachers’ mindsets can be understood as experiential knowledge that is highly subjective, but at the same time connected through cause-and-effect relations. The teachers’ accumulations of experiential knowledge thus qualify as theories that are structurally similar to scientific theories. When it comes to their functions, the teachers’ mindsets are also comparable to scientific theories: teachers interpret and explain their perception of classroom events on the basis of their individual mindsets. In CLIL, these mindsets include several interrelated aspects such as school education and subject-specific teaching in general, foreign language teaching, the perception of language in different subjects, learner and teacher roles, learning processes and learning activities, teaching objectives as well assessment and evaluation.

The teachers’ mindsets also serve as guides for planning classroom activities and predicting classroom events. In addition, they are referred to in the process of justifying one’s decisions retrospectively. Because of their origin in the teachers’ biography, i.e. their personal biography, their learning biography and their professional biography, they are highly stable. They are so persistent that teachers have been seen to defend them against classroom (or other) experiences pointing at their lack of validity and contradicting scientific theories.

Teachers’ mindsets containing assumptions about subject matter teaching or language teaching as well as CLIL-specific elements need to be considered in an integrated manner since these elements can and do influence and reinforce each other. For that reason, CLIL lessons can have a highly innovative potential for schools: CLIL can break the cycle of fossilised routines, behaviour and topics and lead to new perspectives and positions. However, schools and teachers need to make active use of this innovative potential as innovation does not happen automatically (cf. Viebrock forthcoming for a more elaborated account).

Page 12
Page 13

7. New trends in CLIL research: underachievers and linguistic diversity in the classroom

New trends in CLIL research in the German-speaking world can be detected in a shift towards learner groups that were usually neglected in the highly selective implementation of CLIL – underachievers as well as multilingual speakers. The most recent studies also display a more critical attitude towards the pre-supposed “added value” of CLIL by definition and try to shed light on some critical and complex or even negative aspects.

Apsel’s ongoing research project is concerned with underachievers and drop-outs (cf. in this volume). While Zydatiß (2007) has already found evidence that CLIL learners turn their back on CLIL classes for strategic reasons, i.e. better grades when participating in non-CLIL classes, Apsel attempts to provide a statistical overview of all cases of a selected region as well as in-depth studies of individual cases. Following Bohnsack’s documentary method as the overarching theoretical framework, Aspel tries to determine the factors responsible for failing CLIL programmes and possibly create certain learner profiles (types). It is his intention to link the findings to existing theories concerned with weak performances of learners in comparable circumstances (such as Cummins’ above mentioned threshold hypothesis in immersion settings).

Rauschelbach’s (in progress) project is concerned with the individual learner’s multilingual background and its relationship to successful learning in CLIL settings. Within the theoretical discourse in German on CLIL, the fact has usually been neglected that CLIL does not only involve the foreign language and the official school language, but also needs to be related to the linguistic diversity of the learners and the numerous heritage languages that might be represented in contemporary classrooms. For many learners CLIL is not a second language learning activity, but it rather involves three or four languages of various competence levels. This aspect becomes increasingly important in the context of the latest developments in educational policy such as the implementation of CLIL programmes for learners of various abilities. Taking into consideration the allegedly “elitist” nature of the CLIL-approach mentioned above, it is understandable that the theoretical discourse was biased by assumptions of more or less homogeneous classrooms with little linguistic diversity. This conjecture probably resembles an over-simplistic model from the start since there is always variation in learners’ motivation, cognitive abilities, and aptitude among other factors. Apart from this, such a model is certainly no longer feasible in view of the linguistic and cultural heterogeneity of today’s learners.

National large-scale education studies such as DESI (DESI Konsortium 2006), which looked at learners’ achievements in the EFL classroom and German, have shown that students with a multilingual background, i.e. learners who grew up with two languages or learned German in early childhood in addition to another mother tongue are particularly successful in the language areas of reading comprehension, grammar and listening comprehension. The study also provides evidence for an increased socio-pragmatic competence of multilingual learners, i.e. the ability to accurately identify and classify speaking intentions and use the English language appropriately with regard to contact, situation and addressee. Against this background, Rauschelbach poses the question whether CLIL environments offer favourable circumstances, which enable the learners to employ their multilingualism as a valuable resource. In contrast to the achievement studies, Rauschelbach is more interested in questions such as the learners’ attitudes and their self-concept as a multilingual speaker and beneficial teaching methods to foster the multilingual learners’ assumed potential. An in-depth study of individual cases is located in the realm of qualitative interview and classroom observation studies and will probably not yield broadly generalisable results. But it will most likely provide valuable insights from the learners’ points of view, which will be able to serve as additional explanations for the outcomes of the large-scale achievement studies. The topic as such is certainly of great importance particularly against the background that much of what is considered to be “monolingual” content learning (thus making use of the official school language for instruction) already is some kind of a CLIL setting for many multilingual speakers whereas the focus on an additional foreign language in CLIL adds another dimension of complexity that needs to be analysed and reflected in the professional discussion.

Page 13
Page 14

8. Conclusion

This overview shows that research in the German-speaking world has been carried out in a number of fields. Even though the context may be rather specific, the results are certainly of a more general value. As a general tendency it can be stated that the early CLIL projects were driven by a rather positivist attitude, essentially assuming an “added value” by definition, and when this was confirmed by research little attention was given to the specifics of a selected learner population. The perceived positive effect of CLIL has been generalised to a more diverse student profile by policy makers who have fostered the implementation of additional CLIL programmes to additional Grade levels and types of schools. This spread of CLIL needs to be viewed with caution if teachers do not receive the requisite training. After all, CLIL teaching is first and foremost concerned with good teaching: it has to face similar pedagogical challenges as those faced in mainstream programmes. Many CLIL issues are by no means CLIL-specific.

From our point of view, considerably more research needs to be carried out to substantiate the CLIL approach for all learner types. In more recent projects, a greater awareness of critical aspects of CLIL teaching can be found as well as a further diversification of topics including subjects or learner groups that were formerly neglected. As a general tendency, there is more diversity among CLIL learners, among CLIL approaches and among CLIL subjects, which needs to be taken into consideration and reflected in future research.


Abendroth-Timmer, D.: 2007, Akzeptanz und Motivation: Empirische Ansätze zur Erforschung des Einsatzes von bilingualen und mehrsprachigen Modulen. Peter Lang, Frankfurt am Main.

Badertscher, H., Bieri, T.: 2009, Wissenserwerb im Content and Language Integrated Learning: empirische Befunde und Interpretationen. Haupt, Bern/Stuttgart/Wien.

Behrendt, S., Doff, S., Marx, N. and Ziegler, G.: 2011, Review of doctoral research in second language acquisition in Germany (2006-2009). Language Teaching 44.2, 237-261.

Bonnet, A.: 2000, Naturwissenschaften im bilingualen Sachfachunterricht: Border Crossings? In: Abendroth-Timmer, D. and Breidbach, S. (eds.), Handlungsorientierung und Mehrsprachigkeit. Fremd- und mehrsprachliches Handeln in interkulturellen Kontexten, (149-160). Peter Lang, Frankfurt am Main.

Bonnet, A.: 2004, Chemie im bilingualen Unterricht. Kompetenzerwerb durch Interaktion. Leske + Budrich, Opladen.

Breidbach, S.: 2007, Bildung, Kultur, Wissenschaft. Reflexive Didaktik für den bilingualen Sachfach­unterricht. Waxmann, Münster.

Breidbach, S. and Viebrock, B. (eds.): forthcoming 2012, CLIL: Research, Policy and Practice. Peter Lang, Frankfurt am Main.

Coetzee, D.: 2007, Assessment of subject-specific task performance of bilingual geography learners: Analysing aspects of subject-specific written discourse. [online]

Coyle, D., Hood, P. and Marsh, D.: 2010, CLIL. Content and Language Integrated Learning. Cambridge University Press, Cambridge.

Dalton-Puffer, Ch.: 2007, Discourse in Content and Language Integrated Learning (CLIL) Classrooms. John Benjamins Publishing Company, Amsterdam.

DESI Konsortium: 2006, Unterricht und Kompetenzerwerb in Deutsch und Englisch.

Dielmann, S.: 2007, Integriertes Sach-Sprachlernen im bilingualen Geschichtsunterricht: Analysen von Klausurtexten. In: Caspari, D.; Hallet, W.; Wegner, A. and Zydatiß, W. (Hrsg.), Bilingualer Unterricht macht Schule. Beiträge aus der Praxisforschung, (81-96). Peter Lang, Frankfurt am Main.

Dirks, U.: 2004, ‘Kulturhüter’ oder ‘Weltenwanderer’? Zwei ‘ideale’ Realtypen bilin­gu­a­len Sach­fach­unterrichts. In: Bonnet, A. and Breidbach, S. (Hrsg.), Didaktiken im Dialog. Konzepte des Lehrens und Wege des Lernens im bilingualen Sachfachunterricht, (129-140). Peter Lang, Frankfurt am Main.

European Commission: 2006, Content and Language Integrated Learning (CLIL) at School in Europe. Eurydice, Brussels.

Page 14
Page 15

Fehling, S.: 2008, Language Awareness und bilingualer Unterricht: Eine komparative Studie. Peter Lang, Frankfurt am Main.

Grum, U.: 2012, Mündliche Sprachkompetenzen deutschsprachiger Lerner des Englischen. Entwicklung eines Kompetenzmodells zur Leistungsheterogenität. Peter Lang, Frankfurt am Main.

Göbel, K. and Hesse, H.-G.: 2008, Vermittlung interkultureller Kompetenzen im Englischunterricht, in DESI-Konsortium (ed.), Unterricht und Kompetenzerwerb in Deutsch und Englisch, (398-410). Beltz, Weinheim und Basel.

Haataja, K. and Wicke, R.E. (ed.): 2009, Zweiklang im Einklang – Integriertes Sprachen- und Fachlernen (CLIL). Fremdsprache Deutsch. 40/2009. Hueber, München.

Heine, L.: 2010, Problem solving in a foreign language. A study in Content and Language Integrated Learning. Mouton de Gruyter, Berlin, New York.

Kollenrott, A.: 2008, Sichtweisen auf deutsch-englisch bilingualen Geschichtsunterricht: Eine empirische Studie mit Fokus auf interkulturelles Lernen. Peter Lang, Frankfurt am Main.

Krechel, L.: 1999,Sprach- und Textarbeit im Rahmen von flexiblen bilingualen Modulen. Zeitschrift für Interkulturellen Fremdsprachenunterricht.

Lamsfuß-Schenk, S.: 2008, Fremdverstehen im bilingualen Geschichtsunterricht. Eine Fallstudie. Peter Lang, Frankfurt am Main.

Lose, J.: 2007, The language of scientific discourse: Ergebnisse einer empirisch-deskriptiven Interaktionsanalyse zur Verwendung fachbezogener Diskursfunktionen im bilingualen Biologieunterricht. In: Caspari, D., Hallet, W., Wegner, A. and Zydatiß, W. (eds.), Bilingualer Unterricht macht Schule. Beiträge aus der Praxisforschung, (97-107). Peter Lang, Frankfurt am Main.

Mentz, O.: 2001, Ist ‚bilingualer Unterricht’ glaubwürdig? – Ein Plädoyer für fremdsprachiges Sachlernen im 21. Jahrhundert. Neusprachliche Mitteilungen aus Wissenschaft und Praxis. 54/2: 68-77.

Mentz, O.: 2010, Alle Sprachen eignen sich – oder doch nicht? Überlegungen zu einem bilingualen Fächerkanon. In: Doff, S. (ed.) Bilingualer Unterricht in der Sekundarstufe. Eine Einführung, (29-43). Narr, Tübingen.

Meyer, C.: 2003, Bedeutung, Wahrnehmung und Bewertung des bilingualen Geo­graphieunterrichts – Studien zum zweisprachigen Erdkundeunterricht (Englisch) in Rheinland-Pfalz. Universität Trier: Dissertation. Elektronische Veröffentlichung:

Pietzuch, J. P.: 2009, Review of Breidbach, Stephan (2007): Bildung, Kultur, Wissenschaft. Reflexive Didaktik für den bilingualen Sachfachunterricht. Münster: Waxmann. Zeitschrift für Interkulturellen Fremdsprachenunterricht 14.1, 9 pages.

Pitsch, K.: 2006, Sprache, Körper, Intermediäre Objekte: zur Multimodalität der Interaktion im bilingualen Geschichtsunterricht. Bielefeld (Germany): Bielefeld University,

Prüfer, K.: (forthcoming): English-German CLIL modules in the mathematics classroom and their effect on the students’ willingness to learn mathematics. In: Breidbach, S. and Viebrock, B. (eds.), CLIL: Research, Policy and Practice (working title). Frankfurt/Main: Peter Lang.

Putsche, J.: 2011, Spracheinstellungen von Grundschülerinnen und Grundschülern in einer Grenzregion. Qualitative Untersuchung in zwei paritätisch unterrichteten ersten Klassen mit Zielsprache Französisch. Peter Lang, Bern.

Rauschelbach, L.: in progress, Multilingual students in CLIL environments. Unpublished manuscript.

Rumlich, D.: forthcoming, CLIL and students’ linguistic accuracy in written English: Empirical findings. In: Breidbach, S. and Viebrock, B. (eds.), CLIL: Research, Policy and Practice (working title). Peter Lang, Frankfurt am Main.

Rymarczik, J.: 2003, Kunst auf Englisch. Ein Plädoyer für die Erweiterung des bilingualen Sachfachkanons. Langenscheidt-Longman, München.

Stohler, U. and Kiss, B.: 2010, The Acquisition of Knowledge in CLIL: an Empirical Study on the Role of Language in Content Learning, in: Ditze, S.-A. and Halbach, A. (eds.), Bilingualer Sachfachunterricht (CLIL) im Kontext von Sprache, Kultur und Multiliteralität (59-74). Peter Lang,Frankfurt am Main.

Viebrock, B.: 2007, Bilingualer Erdkundeunterricht. Subjektive didaktische Theorien von Lehrerinnen und Lehrern. Peter Lang, Frankfurt am Main.

Page 15
Page 16

Viebrock, B.: 2009a, M² (Multilingual x Mathematical) – Some Considerations on a Content and Language Integrated Learning Approach to Mathematics. Forum Sprache: 61-79.

Viebrock, B.: 2009b, From Teaching Numeracy to Developing Mathematical Literacy: Materials and Methods for a CLIL Approach to Mathematics. Forum Sprache: 122-130.

Viebrock, B.: forthcoming, “The situation in the CLIL classroom is quite different” – or is it? Teachers’ mindsets, methodological competencies and teaching habits. In: Marsh, D. and Meyer, O. (eds.), Quality Interfaces:Examining Evidence & Exploring Solutions in CLIL (working title). University of Eichstätt.

Vollmer, H.J.: 2009, Diskursfunktionen und fachliche Diskurskompetenz bei bilingualen und monolingualen Geografielernern. In: Ditze, S.-A. and Halbach, A. (eds.), Bilingualer Sachfachunterricht (CLIL) im Kontext von Sprache, Kultur und Multiliteralität (165-185). Peter Lang,Frankfurt am Main.

Wannagat, U.: 2010, Bilingualer Geschichtsunterricht im internationalen Fokus. Eine Vergleichsstudie zur Unterrichtspraxis in Deutschland und Hong Kong. Peter Lang, Frankfurt am Main.

Weber, R.: 1993, Bilingualer Erdkundeunterricht und internationale Erziehung. (Geo­graphiedidaktische Forschungen. Band 23). Selbstverlag des Hochschulverbandes für Geo­graphie und ihre Didaktik e.V.

Wegner, A.: 2011, Weltgesellschaft und Subjekt. Eine bildungsgangdidaktisch inspirierte Studie zum bilingualen Sachfachunterricht an Real- und Gesamtschulen. Wiesbaden: VS-Verlag.

Werner, B.: 2009, Entwicklungen und aktuelle Zahlen bilingualen Unterrichts in Deutschland und Berlin, in D. Caspari, W. Hallet, A. Wegner and W. Zydatiß (eds.), Bilingualer Unter­richt macht Schule, (19-28). Peter Lang, Frankfurt am Main.

Witzigmann, S.: 2011, Bildende Kunst in der Zielsprache Französisch als Einstieg ins bilinguale Lernen und Lehren. Explorative Studie einer fünften Realschulklasse. Peter Lang, Frankfurt am Main.

Zydatiß, W.: 2007, Deutsch-Englische Züge in Berlin (DEZIBEL). Eine Evaluation des bilingualen Sachfach­unterrichts an Gymnasien. Peter Lang, Frankfurt am Main.


1 Behrendt et al. (2011) contains a review section on doctoral CLIL research in Germany between 2006 and 2009.
return to text

2 The Eurydice report published by the European Commission on CLIL in Europe (2006) lists „Bilingualer Sachfachunterricht“ and „Bilingualer Unterricht“ as the closest terminological equivalents to CLIL in German.
return to text

3 There has been some research and a fairly vivid debate about the pedagogic legitimacy of CLIL in the field of History which we will touch upon in section 3.
return to text