Teaching Young Learners in Bilingual Settings
A teacher cognition study in the context of CLIL at the Dortmund International Primary Schools (DIPS) network
0923
2024
978-3-3811-2802-0
978-3-3811-2801-3
Gunter Narr Verlag
Paul Berge
10.24053/9783381128020
Bilingual education, or CLIL, at primary school varies greatly across European educational contexts. Teaching Young Learners in Bilingual Settings reports on a study that explored one such CLIL context in Dortmund, Germany. Through interviews and classroom observations, the researcher and author sought not only to document some of what takes place in CLIL classrooms but to describe and understand teachers' thoughts and beliefs about their CLIL teaching practices. This research contributes to a better understanding of primary school CLIL programs and teachers and is relevant for researchers working in the fields of foreign language education, bilingual education, and language teacher cognition research. Furthermore, the insights into CLIL teachers' thinking can support CLIL teachers, administrators, and policy makers as they seek to further develop CLIL pedagogy and programs.
<?page no="0"?> Giessener Beiträge zur Fremdsprachendidaktik Paul Berge Teaching Young Learners in Bilingual Settings A teacher cognition study in the context of CLIL at the Dortmund International Primary Schools (DIPS) network <?page no="1"?> Teaching Young Learners in Bilingual Settings <?page no="2"?> G I E S S E N E R B E I T R ÄG E Z U R F R E M D S P R A C H E N D I DA K T I K Herausgegeben von Jürgen Kurtz, Michael Legutke, Hélène Martinez, Dietmar Rösler und Ivo Steininger. Begründet von Lothar Bredella, Herbert Christ und Hans-Eberhard Piepho <?page no="3"?> Paul Berge Teaching Young Learners in Bilingual Settings A teacher cognition study in the context of CLIL at the Dortmund International Primary Schools (DIPS) network <?page no="4"?> DOI: https: / / doi.org/ 10.24053/ 9783381128020 © 2024 · Narr Francke Attempto Verlag GmbH + Co. KG Dischingerweg 5 · D-72070 Tübingen Bei der vorliegenden Publikation handelt es sich um eine Dissertation aus dem Fachbereich II der Universität Trier. Das Werk einschließlich aller seiner Teile ist urheberrechtlich geschützt. Jede Verwertung außerhalb der engen Grenzen des Urheberrechtsgesetzes ist ohne Zustimmung des Verlages unzulässig und strafbar. Das gilt insbesondere für Vervielfältigungen, Übersetzungen, Mikroverfilmungen und die Einspeicherung und Verarbeitung in elektronischen Systemen. Alle Informationen in diesem Buch wurden mit großer Sorgfalt erstellt. Fehler können dennoch nicht völlig ausgeschlossen werden. Weder Verlag noch Autor: innen oder Herausgeber: innen übernehmen deshalb eine Gewährleistung für die Korrektheit des Inhaltes und haften nicht für fehlerhafte Angaben und deren Folgen. Diese Publikation enthält gegebenenfalls Links zu externen Inhalten Dritter, auf die weder Verlag noch Autor: innen oder Herausgeber: innen Einfluss haben. Für die Inhalte der verlinkten Seiten sind stets die jeweiligen Anbieter oder Betreibenden der Seiten verantwortlich. Internet: www.narr.de eMail: info@narr.de Elanders Waiblingen GmbH ISSN 0175-7776 ISBN 978-3-381-12801-3 (Print) ISBN 978-3-381-12802-0 (ePDF) ISBN 978-3-381-12803-7 (ePub) Bibliografische Information der Deutschen Nationalbibliothek Die Deutsche Nationalbibliothek verzeichnet diese Publikation in der Deutschen Nationalbibliografie; detaillierte bibliografische Daten sind im Internet über http: / / dnb.dnb.de abrufbar. <?page no="5"?> For my late uncle Richard Worthy and his family. <?page no="6"?> Acknowledgements Research projects such as this one are always the result of numerous people along the way contributing in one way or another. I would like to say thank you to some of them here. Thank you: to the teachers who participated in this project for sharing your time and thoughts with me and for opening your classrooms to me. I am indebted to you. to Prof. Dr. Henning Rossa for supervising this project and guiding me through the various twists and turns. To Prof. Dr. Anke Wegner for giving helpful feedback on an earlier draft of this dissertation. to Dr. Kieran Harrington for being interested in the project from the beginning, asking questions, and giving me thoughtful feedback. to Prof. Dr. Patricia Ronan for giving me the additional time I needed to finish this project. to Jun.-Prof. Dr. Julia Reckermann for listening to my early thoughts on the project and giving me valuable feedback. to Stella Klatte for thoughtfully reviewing and giving feedback on aspects of my data analysis process. to my parents, Richard and Angela, my brother, Eric, and my sister, Brianne, for being only vaguely aware of this project, considering it an impractical yet harmless pastime, and talking to me about other things. to my family and most of all my wife, Debora, who has tolerated this project and all that it has required of me. It is obvious that such a project would not have been possible without you. <?page no="7"?> 1 11 1.1 11 1.1.1 11 1.1.2 14 1.1.3 16 1.1.4 17 1.1.5 19 1.2 20 1.3 22 1.4 23 2 25 2.1 25 2.1.1 26 2.1.2 28 2.1.3 31 2.2 34 2.2.1 34 2.2.2 37 2.2.3 39 2.3 40 3 41 3.1 42 3.1.1 42 3.1.2 44 3.1.3 45 3.1.4 46 3.1.5 46 3.1.6 46 3.1.7 47 3.1.8 49 Contents Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Background and rationale . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Historical and theoretical foundations of CLIL . . . . . . . . . . . . . . . . . Research results on CLIL outcomes in Germany . . . . . . . . . . . . . . . The CLIL context of the present study: Dortmund International Primary Schools (DIPS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Researching teachers’ cognitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research results on CLIL teachers’ cognitions . . . . . . . . . . . . . . . . . Purpose of the study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Eliciting teachers’ cognitions related to CLIL practices . . . . . . . . . . . . . . . Outline of chapters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CLIL and its Foundations in Theories of Second Language Acquisition . . . . . . . . . . . . CLIL and Immersion: Foundations and principles . . . . . . . . . . . . . . . . . . . . Immersion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Content and Language Integrated Learning . . . . . . . . . . . . . . . . . . . CLIL goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Theories of second language acquisition . . . . . . . . . . . . . . . . . . . . . . . . . . . Usage-based theories: Input, output, and interaction . . . . . . . . . . . Sociocultural theory: Mediation, zone of proximal development, and collaborative dialogue . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Complex systems approach: Affordances, co-adaptation, and learner agency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Early Foreign Language Learning at Primary Level . . . . . . . . . . . . . . . . . . . . . . . . . . . Principles of foreign language instruction at primary school . . . . . . . . . . Authenticity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . English as the language of instruction . . . . . . . . . . . . . . . . . . . . . . . . Multisensory Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Play-based Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Visualization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Prioritizing listening and speaking . . . . . . . . . . . . . . . . . . . . . . . . . . . Differentiation and individualization . . . . . . . . . . . . . . . . . . . . . . . . Theme-oriented teaching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . <?page no="8"?> 3.2 49 3.2.1 49 3.2.2 50 3.2.3 52 3.2.4 53 3.2.5 56 3.2.6 57 3.2.7 58 3.2.8 59 3.3 59 4 61 4.1 62 4.2 63 4.3 65 4.4 68 4.4.1 69 4.4.2 70 4.4.3 73 4.5 77 5 79 5.1 79 5.2 82 5.2.1 82 5.2.2 84 5.2.3 85 5.2.4 87 5.3 89 5.3.1 89 5.3.2 95 5.3.3 96 5.3.4 104 5.4 110 5.4.1 110 5.4.2 112 5.5 118 6 121 6.1 122 6.1.1 122 CLIL at primary school in Germany . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Age and foreign language learning . . . . . . . . . . . . . . . . . . . . . . . . . . Organization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Subjects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Curricula: English and Science . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Teacher requirements and competences . . . . . . . . . . . . . . . . . . . . . . Classroom interaction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . L1 and L2 use . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Teacher Cognition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . History of the study of teacher cognition . . . . . . . . . . . . . . . . . . . . . . . . . . . Goals of teacher cognition studies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Terminology and constructs of teacher cognition . . . . . . . . . . . . . . . . . . . . CLIL teachers’ cognitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Teacher types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . CLIL teachers’ goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Teachers’ approaches . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods and Research Design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Aims and research questions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Collecting and analyzing verbal data . . . . . . . . . . . . . . . . . . . . . . . . . Semi-structured interviews . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Classroom observations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Qualitative content analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research design . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research context: Dortmund International Primary Schools . . . . . Research participants . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Data gathering procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Data analysis procedures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Validity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Reliability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research question 1: What cognitions do teachers hold about teaching and learning in primary school CLIL? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Cognitions about the role of content, L1, and L2 . . . . . . . . . . . . . . . 8 Contents <?page no="9"?> 6.1.2 128 6.1.3 133 6.1.4 136 6.1.5 140 6.1.6 142 6.2 143 6.2.1 143 6.2.2 152 6.2.3 155 6.3 156 6.3.1 157 6.3.2 164 6.3.3 167 6.3.4 176 6.3.5 179 6.3.6 181 6.3.7 182 6.3.8 183 6.3.9 183 7 187 7.1 187 7.2 191 7.3 193 7.4 198 7.5 200 8 203 8.1 207 8.2 208 8.3 209 9 213 10 231 10.1 231 10.2 232 10.3 233 10.4 234 Cognitions about young learner SLA in CLIL . . . . . . . . . . . . . . . . . . Cognitions about choosing CLIL topics and activities . . . . . . . . . . . Cognitions about teacher competences and demands . . . . . . . . . . . Cognitions about student characteristics . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research question 2: What language-related educational goals do teachers work toward? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Foreign language learning goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . General pedagogic goals that support language learning . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Research question 3: What approaches do teachers believe are best for achieving these goals? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Focus on input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Know the students . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Scaffolding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Motivate students . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Student choice of language . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Teach language explicitly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Plan for more time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Build links to other subjects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Discussion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Teachers’ cognitions about teaching and learning in primary school CLIL Teachers’ cognitions about language-related educational goals in CLIL . Teachers’ cognitions about approaches most effective for achieving CLIL goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Influence of the DIPS program design on teachers’ cognitions . . . . . . . . Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Limitations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Implications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Further research . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendices- . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix 1: Transcription conventions- . . . . . . . . . . . . . . . . . . . . . . . . . . Appendix 2: Biographical information questionnaire- . . . . . . . . . . . . . . . Appendix 3: Background interview guide- . . . . . . . . . . . . . . . . . . . . . . . . Appendix 4: Follow-up interview guide- . . . . . . . . . . . . . . . . . . . . . . . . . . Contents 9 <?page no="10"?> 10.5 235 10.6 237 245 247 Appendix 5: Tables of absolute numbers and percentage agreement for each of the four subcategories coded by two coders.- . . . . . . . . . . . . . . . Appendix 6: Codebook- . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10 Contents <?page no="11"?> 1 Introduction 1.1 Background and rationale 1.1.1 Historical and theoretical foundations of CLIL Content and language integrated learning (CLIL) is a term used to describe educational settings “in which pupils learn a subject through the medium of a foreign language” (Eurydice, 2017, p. 13). The term “CLIL” was coined in 1994 (Coyle, Hood, & Marsh, 2010) to be an all-inclusive or “umbrella term” (Eurydice, 2017, p.-13) for the variety of bilingual programs that already existed in Europe at the time (Baetens Beardsmore, 2009, p. 208). While such an open definition of CLIL is no doubt useful on a European scale, various authors in German publications have attempted to further define or more precisely describe CLIL, or Bilingualer Sachfachunterricht (BSFU) in German (Rumlich, 2016, p. 41), as it is conceptualized in the German educational landscape (see, for example, Breidbach, 2002; Zydatiß, 2002; Werlen, 2006; Diehr, 2012; Mehisto, 2013). The modern history of bilingual education, or CLIL, in Germany dates back to at least 1963 with the signing of the Franco-German Treaty (Breidbach & Viebrock, 2012) and the subsequent development of a French-German bilingual program at a Gymnasium in Baden Württemberg (Wolff, 2016). From this starting point, bilingual education in Germany con‐ tinued to develop largely at other academically-oriented Gymnasien (Breidbach &Viebrock, 2012). While the early initiation of bilingual programs in Germany was often motivated by regional language profiles and needs (for example in border regions), the expansion of bilingual education in the mid and late 1990s was sparked by increasing globalization and European integration (Wolff, 2016). Economic incentives rooted in globalization provided a metaphorical carrot for foreign language learning while a recommendation from the European Commission stating that European Union citizens should learn two European Community languages beyond their mother tongue (EC, 1995, p. 47) provided a gentle metaphorical stick. The EC further suggested that bilingual education at secondary schools may help to achieve this goal (EC, 1995, p.-47). Further motivating the early expansion of CLIL education in Europe was the notion that bilingual education is an efficient form of teaching and learning since non-language content and a foreign language are learned simultaneously. Wolff (2009, p. 546) suggests that CLIL education “saves time within the overall curriculum” and that the “length of study time both for language and content subject can thus be reduced considerably, and as a consequence, more languages can be introduced into the curriculum and more time can be devoted to the study of each language.” While such a promise of “two for the price of one” has no doubt been a factor in the initial expansion of CLIL programs, Lyster (2011, p. 612) warns that “nothing comes for free.” Instead, Lyster maintains that “a great deal of attention still needs to be drawn to the second language, which needs to be manipulated and enhanced during content teaching” (Lyster, 2011, p.-612). As an approach to foreign language learning, CLIL programs are broadly based on the idea that changing students’ linguistic environment will lead to changes in students’ linguistic competences. Because it has been argued by some that it may be possible to learn <?page no="12"?> more on the “streets” than in the classroom (see Dalton-Puffer, 2007, p. 2 for a discussion), Dalton-Puffer suggests that CLIL, as a pedagogical approach for the development of a foreign language not spoken outside of the immediate classroom, is a “clever and economical way of turning classrooms into ‘streets’” (Dalton-Puffer, 2007, p. 2). Key to such an altered linguistic environment is increased levels of L2 input and opportunities to use the L2 in various ways. Input is the “sine qua non” of any language learning context (Gass & Mackey, 2015, p. 181). In CLIL, the notion of “rich input” (Gallardo-del-Puerto & Blanco-Suárez, 2021; Mehisto, Marsh, & Frigols, 2008) suggests that input in CLIL is abundant, varied, and of high quality. Much of this input needs to come from teachers. Wode (2009) claims that with young CLIL learners, teachers are the only ones who are able to provide such rich input and therefore advocates for high amounts of L2 teacher talk in early CLIL lessons (Wode, 2009, p.-100). In addition to rich input, CLIL contexts provide opportunities for students to use the L2 for various communicative purposes (Mehisto et al., 2008; Buchholz, 2007). Foreign language development in CLIL will be “driven primarily by the discourse in which students need to engage to complete academic tasks” (Lyster, 2011, p. 619). Mehisto et al. (2008, p. 26) describe a basic CLIL language learning formula: “A CLIL environment can unleash a child’s language-learning potential by providing rich input, and opportunities for rich intake and output” (italics in original). However, they warn that “not all input becomes intake. And if there is limited intake, then there will be equally limited opportunities for output” (Mehisto et al., 2008, p. 26 italics in original). Beyond increased opportunities for input and output, other constructs from the field of second language acquisition (SLA) such as “negotiation for meaning” (Long, 1996), “collaborative dialogue” (Swain, 2000), and “affordances” (van Lier, 2000; Larsen-Freeman, 2015) also play a role in providing a theoretical basis for the acquisition of a foreign language in CLIL contexts. The theoretical basis for language learning within a CLIL context will be reviewed in chapter 2.2. Early versions of bilingual education at primary schools in Germany began in the early and middle of the 1990s. Zydatiß (2000) reports on an early trial of primary school bilingual education at a network of 14 public “European schools” in Berlin beginning in the school year 1992-93. Wode (2009) accompanied and studied a primary school bilingual program implemented at the Claus-Rixen School in Altenholz/ Kiel that began in 1996. In North Rhine Westphalia (NRW) (where this study is set), however, various forms of primary school CLIL did not begin until the early 2000s at which time English as a foreign language (EFL) instruction became compulsory at primary schools. Wode (2009) claims that foreign language learning should begin early, in preschool or at the beginning of primary school, because of the additional time that an early start affords learners on their way to achieving a high enough level of language competence. Wode argues that the time required to learn two additional languages beyond the mother tongue or, as he calls it, the “3 language formula” cannot be met if students begin learning the first foreign language in fifth or even third grade with only one or two lessons per week (Wode, 2009, p. 17). Wode goes on to argue that such early language learning should take place through an immersion approach which he claims, among other things, provides young learners with an age-appropriate and “natural” approach to language learning and is cost-neutral in that it does not require additional teachers or lessons (Wode, 2009, p. 18). 12 1 Introduction <?page no="13"?> The promise of CLIL at primary school is that it is a time and cost efficient approach to foreign language learning and, by extending the time spent learning languages at school, makes it possible for students to reach the goal of mother tongue plus two set out by the EC. CLIL programs generally target the development of competences in three areas: the L1, L2, and the non-language subject. However, the way that these and other related goals are communicated tend to differ between the broader European discourse on CLIL and the German discourse on CLIL. For example, Mehisto et al. (2008, p. 12), writing from a broader European perspective, describe five goals that CLIL programs target. They are: • grade-appropriate levels of academic achievement in subjects taught through the CLIL language; • grade-appropriate functional proficiency in listening, speaking, reading and writing in the CLIL language; • age-appropriate levels of first-language competence in listening, speaking, reading and writing; • an understanding and appreciation of the cultures associated with the CLIL language and the student’s first language; • the cognitive and social skills and habits required for success in an ever-changing world. (Mehisto et al., 2008, p.-12) In the German context, Diehr (2012, p. 29) describes two all-encompassing CLIL goals that are representative of the broader discourse on CLIL in Germany: “Ziel des BU ist es, doppelte Fachliteralität auszubilden und kulturbewusste Mehrperspektivität in fachspezi‐ fischen Zusammenhängen zu entwickeln” (The aim of CLIL is to train subject literacy in two languages and to develop culturally aware multi-perspectivity in subject-specific contexts). CLIL goals written in and for the German context are more relevant to this study which is set in the context of primary school CLIL programs in Germany. Nonetheless, briefly comparing the way CLIL goals are described in the European and German contexts can illuminate more precisely what exactly CLIL programs in Germany aim to achieve. First, the goals as communicated by Diehr (2012) have the advantage of existing in a particular national CLIL context. They reference explicitly “subject literacy” and “subject-specific contexts” indicating the guiding role that the various subject curricula in the individual German federal states play in setting the educational goals for CLIL in Germany. As stated by the Standing Conference of the Ministers of Education and Cultural Affairs (Kultusministerkonferenz), “Leitfach des bilingualen Unterrichts in den Ländern ist das Sachfach” (The guiding subject of CLIL in the federal states is the content subject) (KMK, 2013, p. 7). In Germany, the goal of CLIL is first to teach the subject as outlined in the various curriculum documents in the different federal states. Second, Mehisto et al. (2008, p. 12) describe CLIL goals relevant to the L1 as follows: The goal is to develop “age-appropriate levels of first-language competence in listening, speaking, reading and writing.” Stated this way, L1 development seems to be uncoupled from subject-specific discourses or literacies and rather a matter of age-normed development. Such a developmental orientation to the L1, one that suggests that the L1 will develop appropriately regardless of targeted L1 instruction, differs from a more instructional 1.1 Background and rationale 13 <?page no="14"?> orientation to the L1 suggested by the CLIL goal of developing subject literacies in two languages described by Diehr (2012). Such a difference in orientation to the L1 in CLIL suggests that in German CLIL contexts, the L1 should play a more active role and should be an explicit object of learning. The promise of CLIL and ultimately its efficacy for the development of increased L2 competences has been the focus of much research for more than two decades. Much of the early research on the effects of CLIL enthusiastically reported the benefits of CLIL for L2 development (Pérez Cañado, 2016). However, some of this initial research has come under scrutiny for a variety of methodological weaknesses including variable control, research design, and statistical methodology (Rumlich, 2016; Pérez Cañado, 2016). Pérez Cañado (2020) cites more recent studies (primarily from Spanish contexts) which she claims address and correct such methodological weaknesses. She concludes that “CLIL is the variable which best accounts for the differences detected in FL competence” (Pérez Cañado, 2020, p. 9). Such results are encouraging for advocates of CLIL education but nonetheless need to be met with a degree of skepticism when seeking to apply them to other educational contexts, e.g. CLIL contexts in Germany (Rumlich, 2016, p.-194). 1.1.2 Research results on CLIL outcomes in Germany In Germany, a number of studies have investigated the effects of English language CLIL instruction in secondary schools. Nold, Hartig, Hinz, & Rossa (2008) compared CLIL students’ and non-CLIL students’ English language competences in areas such as listening comprehension, reading comprehension, and grammar. The researchers analyzed a subset of data gathered in the DESI (Deutsch-English Schülerleistungen International) study. The data was gathered from 958 students from 31 academically-oriented Gymnasium classes and 7 vocationally-oriented Realschule classes. While CLIL students outperformed non-CLIL students in all competence areas when tested at the end of grade nine, the results as presented in Figure 35.1 (Nold et al., 2008, p. 455) indicate that only CLIL students’ listening comprehension improved at a rate beyond that of non-CLIL students suggesting that CLIL can be especially effective at developing receptive language competences. Dallinger, Jonkmann, Hollm, & Fiege (2016) compared the general English proficiency, listening comprehension, and History knowledge of CLIL (N=703), non-CLIL (students at schools with CLIL streams who do not participate in CLIL) (N=659), and regular students (students at schools without CLIL streams) (N=444). They tested students at the beginning and the end of grade eight. The results indicate that at the end of grade eight, CLIL students outperformed non-CLIL and regular students in measures of general English proficiency and listening comprehension. However, after prior achievement was controlled for, CLIL students’ advantage in general English proficiency was reduced to statistically insignificant levels but CLIL students’ advantage in listening comprehension remained statistically significant albeit reduced from previous levels. All groups performed roughly the same with regard to History knowledge, although CLIL students received 50 % more History instruction than non-CLIL and regular students. The authors conclude that a CLIL-program selection effect exists, that CLIL has a positive effect on students’ listening comprehension skills, and that CLIL students “need more input to achieve the same output regarding 14 1 Introduction <?page no="15"?> central content knowledge” (Dallinger et al., 2016, p. 30). This conclusion suggests that CLIL, described as an efficient form of instruction (Wolff, 2009), may not be as efficient as suggested in theory. Rumlich (2016) compared the general English proficiency, EFL self-concept, and interest in EFL classes between CLIL students (N=414), non-CLIL students (N=360), and regular students (N=179). He gathered data on background variables (e.g. verbal cognitive abilities, sex, L1 background) and tested students before the beginning of grade seven at which time CLIL instruction would begin for the future CLIL students and then again at the end of grade eight (i.e. after two years of CLIL instruction). In interpreting his statistical results for general EFL proficiency, Rumlich (2016) reports that CLIL students had an approximately 1.25 and 1.5 to two school year advantage over regular and non-CLIL students respectively. However, “the absolute mean proficiency gains of CLIL and regular students are equal” indicating that differences between the two groups at the end of grade eight were the same as they were at initial testing at the end of grade six (Rumlich, 2016, p. 424). Rumlich concludes that such results “suggest that there might be no impact of CLIL on general EFL proficiency” (Rumlich, 2016, p.-425). The partial results of the studies presented here suggest that in secondary school CLIL contexts in Germany, CLIL instruction has a positive benefit on CLIL students’ listening comprehension skills but not on their general English proficiency. Learning content, while not affected negatively by CLIL, may require more instructional time in order to achieve results similar to non-CLIL and regular students. Two studies reviewed briefly below help illuminate the effects of CLIL on content and language learning in primary school contexts in Germany. In a qualitative study of one primary school CLIL class in Germany, Botz and Diehr (2016) investigated the effects of L1 and L2 instruction on students’ conceptual and vocabulary knowledge. Students were tested and interviewed after a sequence of lessons taught exclusively through the L2 (English) and then after another sequence of lessons taught through the L2 and the L1 (German). The authors report that students were able to better verbalize their subject knowledge in the L1 after being instructed bilingually through English and the “sparsame aber gezielte Einsatz der deutschen Sprachen” (sparse yet targeted use of the German language) (Botz & Diehr, 2016, p. 251). The authors report that the effect was greatest at the lexical level and conclude that acquired subject-language competences in the L2 (English) are not automatically transferred to the L1 (German) (Botz & Diehr, 2016, p.-256). In a quantitative study, Frisch (2021) investigated the natural science competences of CLIL (N=207) and non-CLIL (N=125) fourth grade students at seven primary schools in NRW as part of the “Bilinguales Lehren und Lernen in der Grundschule - Effekte auf die naturwissenschaftliche Kompetenz” (Bilingual teaching and learning at primary school - effects on natural science competences) (BiLL-NaWi) study. The BiLL-NaWi students’ natural sciences competences were tested using a German language version of the TIMSS (Trends in International Mathematics and Science Studies) test. The results were then compared within the study group (CLIL and non-CLIL students), with a sample of regular German students (students not attending schools with a CLIL branch), and with an international sample of students. The CLIL students in the BiLL-NaWi sample outperformed the non-CLIL students in the BiLL-NaWi sample by 72 points, the German 1.1 Background and rationale 15 <?page no="16"?> sample by 9 points, and the international sample by 51 points. However, in order to determine whether or not this advantage was the result of CLIL instruction, Frisch (2021, p. 41) conducted propensity score matching in order to compare CLIL students from the BiLL-NaWi sample to students with similar background characteristics (e.g. age, sex, first language, socio-economic status) in the other samples (German and international). The results of this comparison showed no advantage for the CLIL learners in the BiLL-NaWi sample (Frisch, 2021, p. 41). Such a result suggests that while CLIL may not have a positive effect on natural science competences, it also does not have a negative effect. No negative effect on the development of natural science competences can be interpreted positively by CLIL advocates, since CLIL is motivated by the potential for increased foreign language competences and not for increased non-language subject competences. In order to ascertain whether or not the language of the test influences students’ results, Frisch (2021) also had the CLIL students in the BiLL-NaWi sample complete the TIMMS test in English. Student’s scores were drastically lower. Frisch (2021, p. 42) reports that, on average, students scored 386 fewer points on the English version of the TIMMS test than on the German version. On the German version of the test, almost 80 % of the CLIL students tested achieved scores placing them at level three or higher on a five-level scale of natural science competences. On the English version of the test, more than 90 % of the students achieved scores placing them at level one. Frisch (2021, p. 42) concludes that the CLIL students could only complete simple and routine English language tasks that are solvable with common knowledge. The results of the two studies from German primary school CLIL contexts reviewed above suggest first that content learning in the natural sciences is not negatively affected by learning through a foreign language. Second, the L2 competences of primary school CLIL students are not sufficient to master academic tasks through the L2. Frisch (2021, p. 46) therefore questions whether the “ambitious” goal of subject literacy in two languages is appropriate for primary school students. Third, the acquisition of conceptual and lexical competences in the L2 does not automatically transfer to the L1 suggesting that the targeted use of the L1 needs to be included in CLIL instruction in order to (potentially) meet the L1 aspect of the goal of developing subject literacy in two languages (Botz & Diehr, 2016). 1.1.3 The CLIL context of the present study: Dortmund International Primary Schools (DIPS) Based on the research from German CLIL contexts reviewed above, the positive effects of CLIL may not be as high as once thought or hoped. Indeed, contentious debates regarding the effects of CLIL continue (see, for example, Bruton, 2019; Pérez Cañado, 2020). However, such concerns or uncertainties have generally not prevented the expansion of CLIL offerings. For example, in addition to the various officially recognized CLIL branches at secondary schools in NRW, the Ministry of Education in NRW has initiated the “Bilingual für alle” (CLIL for everyone) program, a program which allows any secondary school in NRW to establish and offer various forms of CLIL instruction (MSB NRW, 2022, March 3). As can be assumed when implementing CLIL under programs such as “Bilingual für alle,” CLIL programs vary greatly in how they are implemented. This concern about what 16 1 Introduction <?page no="17"?> exactly CLIL is and how it is realized has been addressed in the CLIL literature. Pérez Cañado (2016) reviews such concerns. She reports that some researchers see such diversity in CLIL programs as “detrimental” to the “pedagogically coherent evolution of CLIL” (Pérez Cañado, 2016, p. 15), while others see such diversity as helping CLIL to “accommodate the linguistic diversity of the European landscape” (Pérez Cañado, 2016, p.-15). One of those realizations of CLIL is the context in which the current study is set. The Dortmund International Primary Schools (DIPS) program is a group of five primary schools in Dortmund, Germany each with a CLIL stream or branch for each of the four grade levels. The program started during the 2010-2011 school year with the aim to close the then-existing gap between CLIL programs offered at various preschools and secondary schools in Dortmund (Raunser & Steffens, 2012). By closing this gap, program organizers sought to further expand foreign language (English) learning opportunities to students in and around Dortmund (Raunser, 2012, March 5). The DIPS program is described as using an “immersion” approach to language learning in which “the students are literally immersed in the English language” (DIPS, 2020, Sept. 7, my translation). Additionally, the program is characterized by Science and Social Studies lessons that are taught by two teachers; an English-speaking as well as a German-speaking teacher. Within this team-teaching arrangement, the concept of “one person-one lan‐ guage” in which the English-speaking teacher only uses English with students and the German-speaking teacher only uses German with students is implemented (Raunser & Steffens, 2012, p.-37). Within this broad programmatic outline (to be detailed further in section 5.3.1), the teachers in this study, whether alone or in collaboration with colleagues, make decisions about learning goals and instructional approaches and put those decisions into action. Such thinking and practice is ultimately what creates the specific CLIL program in which students learn. 1.1.4 Researching teachers’ cognitions The very nature of teaching is one of thought and action. Teaching is a “reflective, thinking activity” (Calderhead 1987b, p. 1) and teachers are “active, thinking decision-makers who make instructional choices” (Borg, 2003, p. 81). Within the context of CLIL instruction, such thinking and instructional choices are further complicated by what Morton (2012, p. 12) describes as a “complex hybrid of practices [which draw] on the pedagogies of different academic subjects and those of language education.” Further complicating CLIL is the role of the L2 in CLIL as both “an intended outcome” and as “an essential prerequisite for content learning to be possible at all” (Morton, 2012, p.-12). While any CLIL context is complex, there are factors which make teaching and learning in primary school CLIL contexts even more complex. First, most primary school CLIL students are at the very beginning of the foreign language learning process and therefore have low foreign language competences. Since some degree of L2 competence is “an essential prerequisite” (Morton, 2012, p. 12) for learning content through an L2, teaching a content subject through a language that is largely not understood by students presents obvious challenges. Second, there is some evidence that primary school CLIL teachers in 1.1 Background and rationale 17 <?page no="18"?> Germany and in NRW specifically lack professional support and guidance in the form of CLIL-specific teacher training (Massler, 2012) and CLIL-specific guiding documents or handbooks from the Ministry of Education in NRW. Such a lack of professional support would seem to make planning and teaching in CLIL contexts even more difficult. In order to understand specific realizations of CLIL including those at primary school, it is important to understand the teachers who make decisions about and implement instruction. As Tedick and Cammarata (2012, p. S48) explain: Because teachers are the ultimate decision makers as to what enters their classroom and because all curricular reforms are filtered through their beliefs and perception, CBI [content-based instruction, a North-American term for CLIL-like instruction] has little chance to succeed without their support, interest, and motivation. In other words, teachers’ experience is key to CBI program implementation. Because CLIL is a complex pedagogical context and because teachers are thinkers, deci‐ sion-makers, and the ultimate arbiters of what happens in the classroom, there have been calls to investigate CLIL from a teacher cognition perspective (see Morton, 2012; Nikula, 2017, July 5). The study of teacher cognition itself or what teachers “think, know and believe” (Borg, 2003, p. 81) dates back to the 1970s and 1980s and focused primarily on teachers’ decision-making processes and knowledge (Borg, 2015) as well as “subjektive Theorien” (subjective theories) (Groeben, Wahl, Schlee, & Scheele, 1988; Koch-Priewe, 1986). Since then, teacher cognition research has established itself as a research approach to a variety of language learning contexts including CLIL (see, for example, Viebrock, 2007; Morton, 2012; Bovellan, 2014). From the beginning, teacher cognition research has been motivated by a simple premise: that “what teachers do is affected by what they think” (Clark & Yinger, 1977, p. 279). In an effort to better understand what teachers do and therefore what happens in the classroom, researchers in this newly formed field of “teacher thinking” (Clark & Peterson, 1986, p. 285), as it was known at the time, sought to describe and better understand what teachers think. Investigating, documenting, and understanding teachers’ cognitions has never been about teachers’ cognitions per se. Instead, it is about how teachers’ thoughts, knowledge, and beliefs can inform an understanding about their classroom practices. Calderhead (1987b) describes this essential value of teacher cognition research. An important role of research is to provide more realistic models of teaching that help us conceptualize the nature of this practice more clearly, enabling supportive efforts, including training and policy-making, to be more constructive. (Calderhead, 1987b, p.-4) In order for teacher cognition research to achieve this purpose, a conception of teachers’ cognitions as “highly situated and action-oriented” (Borg & Sanchez, 2020, p. 25) is necessary. However, even with an understanding of teachers’ cognitions as being based on practice, what teachers think is not synonymous nor always congruent with what they do. What teachers do in the classroom is based on a variety of factors which includes their thinking and contextual factors. Teachers’ thinking, in turn, is affected by factors which include their classroom practices (Borg, 2003, p. 82). The fact that teachers’ cognitions may be more or less aligned philosophically, theoretically, or practically with their practices 18 1 Introduction <?page no="19"?> suggests, again, that the study of teachers’ cognitions can be used to further understand and support classroom practice (Calderhead, 1987b, p. 4) but not to “predict teachers’ practices” (Borg & Sanchez, 2020, p.-17). 1.1.5 Research results on CLIL teachers’ cognitions In Germany, relatively little is known about CLIL teachers’ cognitions in the specific context of primary school. Nonetheless, teacher cognition research from other contexts such as CLIL at secondary school and in other national settings as well as primary school EFL settings in Germany can combine to offer insights into teachers’ thinking that may also be relevant to the context of primary school CLIL in Germany. There is CLIL teacher cognition research which suggests that there are various types of CLIL teachers; those who orient themselves primarily to traditional boundaries of subject disciplines and those who orient themselves to more general pedagogic principles of learning (Dirks, 2004; Bonnet & Breidbach, 2017). There is also evidence that teachers differ in the way in which they conceptualize the target language. For example, Hüttner and Dalton-Puffer (2013) found that secondary school EFL teachers in their study oriented themselves to a native-speaker conception of English while CLIL teachers oriented themselves to a conception of English as a global lingua franca. However, in Bovellan’s (2014) study of primary school CLIL teachers in Finland, teachers oriented themselves to a native-speaker ideal of English. Such various conceptions of teaching and learning in CLIL as well as various conceptions of the target language likely have implications for CLIL teachers’ goals and instructional approaches. There is also CLIL teacher cognition research focused on describing teachers’ goals. There is research from primary, secondary and tertiary contexts that suggests that CLIL teachers focus on content learning and not on language form (Skinnari & Bovellan, 2016; Bovellan, 2014; Costa, 2013; de Graaff, Koopman, Anikina, & Westhoff, 2007). Also, the goal of L2 learning in CLIL has been described by teachers (in Spain, Finland, and Austria) as a “side-effect” or “by-product” (Skinnari & Bovellan, 2016, p. 153). Such a cognition in the German context would seem to run contrary to the stated goal of developing subject literacy in two languages. When teachers do describe L2 learning goals, their descriptions tend to be limited to vocabulary knowledge and the development of oral communication competences (Bovellan, 2014; Morton, 2012; Imgrund, 2004). There is also evidence that secondary school teachers value general pedagogic goals such as the development of motivation and respect (Skinnari & Bovellan, 2016; Wegner, 2012). Finally, there is CLIL teacher cognition research that suggests that teachers use a variety of approaches in their CLIL instruction. There is evidence from the Swedish secondary school context that teachers adjust their use of the L1 and L2 based on their knowledge of individual students (Sandberg, 2019). Also, teachers in several different contexts understand language learning in CLIL to be a natural process (Hüttner & Dalton-Puffer, 2013; Costa, 2013; Skinnari & Bovellan, 2016) and therefore attempt to maximize their use of the target language (Lasagabaster, 2017). An exception to such a natural approach to L2 acquisition in CLIL seems to be the explicit instruction of subject vocabulary (Hüttner & Dalton-Puffer, 2013; Morton, 2012; Viebrock, 2007). The research on CLIL teachers’ cognitions briefly 1.1 Background and rationale 19 <?page no="20"?> summarized here comes from a variety of CLIL contexts. A more complete review of CLIL teacher cognition research will be presented in chapter 4.4. 1.2 Purpose of the study It is within the context of CLIL and teacher cognition research that this study seeks to understand the cognitions of five CLIL teachers all teaching within the Dortmund International Primary Schools (DIPS) program in Dortmund, Germany. The investigation of teachers’ cognitions is “one way of making sense of what [teachers] do” (Borg, 2016, April 25). Therefore, it is my intention in this study to explore and describe primary school CLIL teachers’ cognitions in Germany. I will do this in the specific context of the DIPS program. Achieving such a goal serves two purposes. First, it helps to fill an existing research gap in the cognitions of primary school CLIL teachers in Germany. As introduced above, teacher cognition research in CLIL contexts comes from a variety of national contexts at various school levels. None of the research introduced above comes from primary school CLIL contexts in Germany. Investigating and understanding primary school CLIL teachers’ cognitions is important not only because CLIL, regardless of school level, demands subject and language integration thus making it a “complex hybrid of practices” (Morton, 2012, p. 12), but also because CLIL teachers at primary school are faced with young learners who often have low target language competences, “an essential prerequisite” (Morton, 2012, p. 12) for CLIL learning, thus increasing the complexity and challenge of their CLIL instruction. Also, at least in some cases, primary school CLIL teachers lack support structures for their CLIL instruction (see, for example, Massler, 2012; MSB NRW, 2021, March 4; Wolff, 2020, Sept. 7). Exploring and describing primary school CLIL teachers’ cognitions can, ultimately, support the further development of CLIL pedagogy at primary school. Second, conducting the study in the specific CLIL context of the DIPS program will provide a valuable description of and insight into the DIPS program from teachers’ perspectives. As noted earlier, the ways in which CLIL programs are implemented across Europe are diverse (Pérez Cañado, 2016). An important role of CLIL research is to document and understand specific CLIL programs from a variety of research perspectives including that of teacher cognition. Woods (1996), writing from a teacher cognition perspective, argues that any research into the process of teaching and learning as well as into classroom contexts is an attempt to “evaluate the success and failure of the learning process, to determine the factors that lead to success, and render it more successful” (Woods, 1996, p. 3). While this study does not evaluate the DIPS program per se, the results of this study nonetheless have the potential to inform DIPS teachers and administrators on their unique CLIL program and pedagogy and thus, possibly, “render it more successful” (Woods, 1996, p.-3). Several key questions emerge from the literature on CLIL and CLIL teachers’ cognitions. One of them relates broadly to how teachers conceptualize various aspects of CLIL, for example, the target language itself and learning through a foreign language as well as the prioritization of major CLIL goals such as content learning and L1 and L2 development. 20 1 Introduction <?page no="21"?> CLIL teacher cognition research suggests that teachers see the process of language learning in CLIL as a natural process and prioritize content learning. However, this research does not come from primary school CLIL contexts in Germany. Considering the generally low target language competences of primary school learners and the goal in Germany of the development of subject literacies in two languages, a variety of more specific questions relevant to DIPS teachers emerge. For example, do DIPS teachers describe learning through a foreign language in CLIL as particularly challenging? Do teachers ascribe a greater role to L1 use and, if so, what is that role? Do they describe the development of L2 competences in terms of something that can and must be achieved or, like the teachers in Skinnari and Bovellan’s study, do they describe it as a “side-effect” or “by-product” (Skinnari & Bovellan, 2016, p.-153) of CLIL instruction? Another key question that emerges from the literature and research on CLIL is, What goals related to L2 learning do teachers work toward? CLIL teacher cognition research suggests that teachers focus on vocabulary knowledge and the development of oral com‐ petences. Such limited goals would seem less than the all-inclusive and “ambitious” (Frisch, 2021, p. 46) goal of developing subject literacy in two languages. Based on her research, Frisch (2021) concludes that the goal of developing subject literacy in two languages may be too ambitious and that new pedagogical conceptions for primary school CLIL in Germany need to be developed. In this context, what L2 competences do DIPS teachers focus on? Like teachers in studies from Skinnari and Bovellan (2016) and Wegner (2012), do teachers in the DIPS schools also target general pedagogic goals such as motivation and respect? The answers to such questions have potential relevance for the reconsideration of CLIL goals at primary school and the further development of pedagogical conceptions of CLIL at primary school. Finally, research on CLIL and CLIL teachers’ cognitions suggests that teachers employ or describe employing a variety of approaches including, for example, adjusting L1 and L2 use based on knowledge of students (Sandberg, 2019) and maximizing L2 input (Lasagabaster, 2017). Based on their research into language switching in primary school CLIL, Botz and Diehr (2016), found that learners who acquired subject knowledge in the L2 were not automatically able to transfer that knowledge to the L1 and therefore conclude that in order to develop subject literacies in two languages, the L1 should be used in targeted ways in CLIL instruction. The question, then, is what approaches do DIPS teachers describe using in order to meet their goals and, more specifically, how do they describe using the L1 in their CLIL instruction? Do they, for example, describe using the L1 for the purpose of L1 development or rather to support content and L2 learning? Based on the reading of the theoretical and empirical CLIL literature presented above, three research questions emerge that will guide the research project in order to achieve the two research purposes. These questions are broad and therefore allow an exploratory approach to teachers’ cognitions. 1. What cognitions do teachers hold about teaching and learning in primary school CLIL? In its breadth, this question allows for the exploration of a range of teachers’ cognitions that may go beyond what is already evident in CLIL teacher cognition research. 1.2 Purpose of the study 21 <?page no="22"?> 2. What language-related educational goals do teachers work toward? While CLIL is a dual-focused, content and language learning pedagogy, this research question focuses more specifically on language learning goals (e.g. various foreign language competences) as well as any goals that may be more tangentially related to language learning (e.g. motivation, attitudes, etc.). 3. What approaches do teachers think are most effective for achieving these goals? This question allows for the exploration of teachers’ thinking not only about what approaches, techniques, or methods they think are best suited to learning in CLIL but also to why they think such approaches work. 1.3 Eliciting teachers’ cognitions related to CLIL practices In this study, I have employed the data gathering methods of semi-structured interviews and classroom observations. By talking to teachers about their actual CLIL practices and observing those practices over time, it was my intention to elicit teachers’ “highly situated and action-oriented” (Borg & Sanchez, 2020, p. 25) cognitions through “contextualized discussion[s] of teachers’ concrete experiences in the classroom” (Borg & Sanchez, 2020, p. 25). Semi-structured interviews provided the necessary flexibility for teachers to speak about a range of topics related to their CLIL practices. Classroom observations allowed me to document classroom practices and subsequently investigate teachers’ thinking about those practices in post-observation interviews. To analyze the interview data, I have applied the method of qualitative content analysis (QCA) because it offers a systematic approach to describing the meaning of qualitative data while at the same time being flexible enough to be tailored to the needs of individual research projects (Schreier, 2012). Additionally, the coding frame, a central aspect of QCA, resulting from the data analysis offers a concrete and “quasi-statistical” (Becker, 1970, p. 81) summary of the data. Such a catalogued summary of results that includes descriptive statistics (frequencies of coded segments) provides another tool to communicate results clearly to DIPS program stakeholders and the scientific community. The study was ultimately carried out in three phases. Phase one consisted of two parts: 1) a short questionnaire (10 items) intended to gather biographical information such as age and teaching experience (see Appendix 2 for the questionnaire) and 2) an interview designed to elicit teachers’ thinking on a range of CLIL-related topics such as their CLIL goals, materials, and planning (see Appendix 3 for background interview guide). The initial interviews lasted between 45 minutes and one hour. Phase two consisted of observing CLIL lessons in all of the participating teachers’ classes. Classroom observations took place over approximately six to eight weeks and were recorded using fieldnotes. The goal of the observations was to observe the types of activities used and lesson structures as well as to record how languages are used in lessons (i.e. who uses what language when). Observations were also used as a basis on which to plan and conduct the follow-up interviews. In phase three, I interviewed teachers a second time. The focus of the follow up interviews was to 22 1 Introduction <?page no="23"?> get teachers to reflect and comment on aspects of the CLIL instruction documented during the classroom observation phase of the study (see Appendix 4 for the follow-up interview guide). The interviews were recorded using a digital audio-recorder. Classroom observations were recorded using fieldnotes. The interview recordings were transcribed using transcrip‐ tion software (f4transkript and MAXQDA) and analyzed using QCA. Because QCA can be implemented in a variety of ways and in order to avoid accusations of “name-dropping” (Kuckartz, 2019, p. 3), I will describe in detail my process of data analysis in section 5.3.4. Classroom observations were used to provide a basis for the phase three follow-up interviews and to provide additional context for the interview analysis. 1.4 Outline of chapters In chapter two of this study, I first review the foundations and principles of CLIL and immersion education and then describe key concepts and constructs from the field of second language acquisition which provide a theoretical basis for language learning in CLIL. In chapter three, I review two theories of learning important to primary school learners as well as review eight key pedagogical principles of teaching EFL at primary school which are relevant to language learning and teaching within a primary school CLIL context. I then describe the primary school CLIL context in North Rhine Westphalia (NRW) considering both organizational aspects as well as empirical evidence from a variety of sources that, taken together, illustrate a more complete picture of what primary school CLIL in NRW looks like. In chapter four I first review the history, goals, and terminology of teacher cognition research. I then review empirical literature in order to describe what CLIL teachers think about their CLIL practice with special attention to their goals and approaches. While empirical literature on primary school CLIL teachers’ cognitions is rare, section 4.4 brings together research from related fields including teacher cognition research from secondary school CLIL contexts, international contexts, and German primary school EFL contexts in order to provide an empirical basis of knowledge that can help inform an understanding of primary school CLIL teachers’ cognitions in Germany. In chapter five, I further elaborate on the aims of this study and then review theoretical considerations of the data gathering and analysis methods used. After that, I describe in detail the research context as well as the data gathering and analysis procedures. Finally, I discuss issues of research quality, specifically validity and reliability. In chapter six, I describe the results of my research. This chapter has been organized first according to research questions and then according to the highest-level subheadings in the coding frame. At the beginning of each subsection I include the corresponding part of the coding frame. I then describe and contextualize this part of the coding frame with evidence from the teacher interviews and from classroom observations. In chapter seven, I discuss key results of the study as they relate to each of the three research questions. In my discussion, I compare the results to the theoretical background of this study and to previous empirical research as well as consider what these results might 1.4 Outline of chapters 23 <?page no="24"?> mean for the DIPS program and how they contribute to the field of primary school CLIL research. Finally, in chapter eight, I summarize the study by outlining six key conclusions. I then describe limitations of the study, discuss possible implications of the results, and suggest further research. 24 1 Introduction <?page no="25"?> 2 CLIL and its Foundations in Theories of Second Language Acquisition 2.1 CLIL and Immersion: Foundations and principles Bilingual education, broadly understood, has been an established form of foreign language education in Europe since at least the late 1990s and has steadily grown in popularity since then (van Kampen, Meirink, Admiraal, & Berry, 2020; Llinares, 2017; Nikula, Dalton-Puffer, & Llinares, 2013; Bonnet, 2012). This increased popularity of bilingual education (or CLIL) can be linked to geopolitical and economic factors such as globalization and European integration (Wolff, 2016). Specifically, the European Commission’s recommendation that European Union (EU) citizens learn two European Community languages beyond their mother tongue (EC, 1995, p. 47) has motivated the spread and development of CLIL programs throughout Europe. CLIL in Europe, however, as Dalton-Puffer, Nikula, and Smit (2010b) point out, is at risk of becoming CEIL (content and English integrated learning), suggesting that CLIL programs in Europe are doing more to support English, the global lingua franca, than supporting European integration by teaching European languages other than English. If this trend continues, they warn that “[t]rue multilingualism will remain a characteristic of minority and migrant groups as well as socio-economic elites (generally involving economically more advantageous languages)” (Dalton-Puffer et al., 2010b, p. 287). This view of CLIL seems to substantiate Baetens Beardsmore’s (1993, p. 3) earlier assertion that bilingual education programs are “essentially different forms of enrichment programmes.” In Germany, the modern concept of bilingual education dates back to the mid-1960s with the signing of the Franco-German Treaty of 1963 (Breidbach & Viebrock, 2012) and the subsequent creation of a French-German bilingual program at a Gymnasium in Baden Württemberg in 1969 (Wolff, 2016). As in other European countries, bilingual education did not begin to gain popularity in Germany until the late 1990s and early 2000s (Wolff, 2016). The early development of bilingual education in Germany took place at secondary schools, primarily at the academically-oriented Gymnasien (Breidbach &Viebrock, 2012). The story of bilingual education at primary school in NRW, however, did not begin until the early 2000s. In the 2003-2004 school year, English as a foreign language (EFL) began to be taught at all primary schools in NRW starting in the third grade (MSW NRW, 2020, October 3). Students would receive two 45-minute EFL lessons per week with the overall goal of developing positive attitudes toward language learning and foreign cultures as well as acquiring initial foreign language competences with a focus on receptive competences (MSW NRW, 2008, p. 71). Shortly after this, the first attempts at implementing various bilingual programs began (Wolff, 2020, September 7). By the 2008-2009 school year, the Ministry of Education in NRW had decided that EFL lessons would begin in the second semester of first grade for all children (MSW NRW, 2020, October 3). The Ministry of Education in NRW has since reverted back to starting EFL lessons in third grade for students starting school in the 2021-22 school year (MSB NRW, 2020). <?page no="26"?> The number of bilingual programs at primary schools in NRW and in Germany more generally have, of course, increased during the less than 20-year history of German primary school bilingual education. Across Germany, the Association for Early Multilingualism in Preschools and Schools (Verein für frühe Mehrsprachigkeit an Kindertageseinrichtungen und Schulen FMKS e.V.) reports that the number of primary school CLIL programs more than tripled between 2003 and 2014 from 80 to 287 (FMKS, 2016, p. 193). In NRW, they report that there are 34 primary schools with bilingual programs (FMKS, 2016, p. 194) while a review of schools listed on the NRW Ministry of Education website estimates the number to be 33 (MSW NRW, 2015). These numbers suggest that despite an increase in primary school bilingual programs, monolingual education remains the reality for the vast majority of primary school students in NRW. The recent decision in NRW to return to teaching EFL starting in third grade (MSB NRW, 2020) makes the future of bilingual education at primary school in NRW uncertain. This brief sketch of bilingual education in Germany and NRW highlights two main features of primary school bilingual education: 1) that its history, at least in NRW, is less than 20 years old and, therefore, relatively young and 2) that despite an increase in popularity, primary school bilingual programs are rare. In Germany, primary school bilingual education is known as bilingualer Unterricht in der Grundschule (Wolff & Sudhoff, 2015) and is modelled on two main concepts within the field of bilingual education, namely immersion and CLIL. Despite these concepts having unique origins and several differentiating features, they share much in common. Their shared commonalities, though, are likely one contributing factor to the “Babylonian confusion” (Elsner & Keßler, 2013a, p. 2) created when trying to define and label various programs. I will review the historical foundations and key principles of the concepts “immersion” and “CLIL” in the following sections. 2.1.1 Immersion Immersion education is generally understood as an approach that uses a second, foreign, or indigenous language as a medium of instruction for at least 50 % of the subject curriculum (Lyster, 2018; Elsner & Keßler, 2013a). The modern concept of immersion education has its origins in St. Lambert, Quebec in the 1960s. From its beginning as an approach to language education, the concept of immersion education has evolved in several important ways. Historically, immersion was used as an approach to teach a second language, for example, French to L1 English speakers living in Quebec. In such a context, the target language is accessible to learners outside of the classroom and plays an official role in life outside of school. As the concept and term “immersion” has spread globally, foreign as well as indigenous languages have also been taught via immersion education. For example, the DIPS program, a self-identified immersion program (DIPS, 2020, Sept. 7), uses English, a foreign language for German L1 speakers, as the medium of instruction. Despite its role as a global lingua franca, English does not hold any official role in Germany. In this way, English immersion programs in Germany differ from French immersion programs in Canada. Common, however, to immersion programs is that the target language is a language other than the L1 of most of the students. 26 2 CLIL and its Foundations in Theories of Second Language Acquisition <?page no="27"?> The amount of time students spend learning in or through the target language is also important to definitions of immersion. Because immersion requires high amounts of comprehensible input to achieve its goals and generally applies a “natural approach” (Krashen & Terrell, 1983) to language acquisition, it is deemed necessary to teach at least 50 % of the curriculum in the target language (Lyster, 2018, p. 3). Total immersion programs are those in which the entire curriculum is taught in the target language while partial immersion programs are those in which only part of the curriculum is taught in the target language (Elsner & Keßler, 2013b, p. 17). According to Swain and Johnson (1997, p. 9) total immersion programs do not use students’ L1 for a year or more. Some scholars argue that because immersion programs are content-driven in which the target language is only the medium of instruction, the target language itself is not a curricular concern (Wolff & Sudhof, 2015). This understanding of immersion limits the term’s applicability to programs that do not focus on, for example, language form or language awareness. Immersion is therefore seen as a passive form of language education intended simply to increase the amount of input students are exposed to and through this develop target language competences. Wolff and Sudhof (2015, p. 37), therefore, see immersion as being limited to early childhood and primary school settings, since these contexts tend to be monolingual and focused on natural language acquisition. Definitions of immersion programs are also based on when a program begins in a child’s development. Early immersion programs typically begin in preschool or in primary school while late immersion programs begin in secondary school (Bechler, 2014, p. 42). Early immersion programs extend the time children spend exposed to and learning a foreign language and thus better prepare them for long term success learning one or more foreign languages (Wode, 2009). Students participating in late immersion programs usually have taken L2 language classes in preparation for the immersion program (Swain & Johnson, 1997). Immersion programs may also differ in the linguistic profiles of their students. In one-way immersion programs, students who are “linguistically homogenous” and who are “typically dominant in the majority language and have no or minimal immersion language (IL) proficiency on program entry” are enrolled (Tedick, Christian, & Fortune, 2011, p. 2). In two-way immersion programs, L1 speakers of both classroom languages (for example, Spanish and English in the United States) are brought together to develop their L2s, which correspond with some of their classmates’ L1s (Tedick et al., 2011, p. 2). In indigenous language immersion, a one-way or two-way immersion approach is used, depending on the student population. Indigenous language programs differentiate themselves from other immersion programs primarily through their primary goal: “to revitalize endangered indigenous cultures and languages and promote their maintenance and development” (Tedick et al. 2011, p.-2). Wolff (2016; 2020, Sept. 7) describes immersion programs in Germany as a form of elite education that are practiced at private schools with international students from society’s upper classes. According to Wolff, immersion programs are rarely found in public primary schools in Germany. Some public primary schools do, however, label their programs as immersive (DIPS, 2020, Sept. 7) despite not addressing the various criteria described above that define an immersion program. Differences between academic definitions and publicly 2.1 CLIL and Immersion: Foundations and principles 27 <?page no="28"?> used labels creates confusion when using the term “immersion.” Regardless of exactly how immersion is understood or used, it is clear that its focus on increased target language input and natural language acquisition have played a role in the development of bilingual education in Europe and in Germany as well as at primary schools. 2.1.2 Content and Language Integrated Learning The development of bilingual education in Europe is largely intertwined with Europe’s effort to integrate culturally and economically in the second half of the twentieth century. In the 1950s, a need for increased multilingualism to support integration was recognized. Not until some twenty years later, though, in 1978 did the European Commission officially recommend to member states that teaching take place through more than one language at schools (Coyle et al., 2010, p. 8). It took another 20 years, however, until bilingual education in Europe began to gain popularity. The term “content and language integrated learning” (CLIL) was first adopted in Europe in 1994 to “describe and further design good practice as achieved in different types of school environment [sic] where teaching and learning take place in an additional language” (Coyle et al., 2010, p. 3). While immersion, as a form of bilingual education, started where bilingual education had not existed, CLIL, at least as a term, was adopted as a neutral way to talk about all of the various forms of bilingual education that already existed in Europe. Various national forms of bilingual education, with their own goals and methodologies, had existed throughout Europe for some time and CLIL was seen as a term that could refer to all of these programs (Baetens Beardsmore, 2009, p.-208). Over the 10 years following its coining, CLIL became increasingly relevant in Europe. In 1995, the European Council published a white paper on education and training titled, Teaching and Learning: Towards the Learning Society (EC, 1995) in which it described the necessity of all European Union citizens to be able to “communicate in at least two Community languages in addition to their mother tongue” (EC, 1995, p. 47), sometimes referred to as the “mother-tongue plus two” policy. The paper also suggests that a bilingual education approach for secondary school students may be appropriate (EC, 1995, p. 47). Furthermore, in a resolution on the early teaching of European Union languages, the European Council called on its member states to “promote the continuous provision of teaching in several languages” (EC, 1997, p. 8). By 2005 the European Council had made recommendations that CLIL be implemented across the European Union (Coyle et al., 2010, p.-8). CLIL has been defined as an educational approach “in which pupils learn a subject through the medium of a foreign language” (Eurydice, 2017, p. 13). While this definition is very similar to the definition of immersion, there are several key conceptual differences. First, there are clear social differences between immersion, which rose out of a single national educational system and generally more homogeneous context, and CLIL, which came from a multinational and therefore more heterogeneous context (Baetens Beardsmore, 2009). Second, the goals of a Canadian-style immersion program and CLIL are different. Immersion programs tend to target native or near-native receptive competences while CLIL programs tend to target “functional competence in both receptive and productive 28 2 CLIL and its Foundations in Theories of Second Language Acquisition <?page no="29"?> skills” (Baetens Beardsmore, 2009, p. 210). Third, in addition to teaching through the target language in CLIL classes, CLIL programs usually teach the target or foreign language in dedicated language classes parallel to CLIL instruction. This is not necessarily the case in immersion programs (Baetens Beardsmore, 2009). Finally, because CLIL arose out of a supranational institution, i.e. the European Union and not out of a local bilingual context, the target language is usually a foreign language that is taught by non-native speakers (Dalton-Puffer, Nikula, & Smit, 2010a, p. 1). Because of the heterogeneous context in which CLIL has evolved, it is not used to refer to a range of “specific lingua-cultural, national, educational and disciplinary traditions (Dalton-Puffer et al., 2010a, p.-3). While the defining CLIL features described above highlight societal and curricular differences between CLIL and other forms of bilingual education such as immersion, other scholars have focused on pedagogical differences between program types. Coyle et al. (2010) focus on the integrative aspect of CLIL. They claim that CLIL differs from other forms of bilingual education due to its “planned pedagogic integration of contextualized content, cognition, communication and culture in teaching and learning practice” (Coyle et al., 2010, p. 6). This statement locates the definition of CLIL with the teacher and teaching practices. It is possible to imagine, therefore, that programs that have teachers who are not planning and teaching in the way described above - integrating the 4Cs (see Coyle et al., 2010, p. 41), for example - would not be defined as CLIL programs. Indeed, this is exactly what David Marsh claims in an interview with Cambridge University Press: The examples of failure are often where there is poor teaching through another language, be it English, be it Portuguese, whatever. And it’s often where they were teaching badly in the first language and then for some reason, political or whatever, they were told they’ve got to teach in English, and then they continue to teach badly in English. I would not call that CLIL. I would call that teaching through another language or something like that. So I can’t really say that I can identify places where you have bad CLIL. You have plenty of places where you have bad teaching through English. For sure. But bad CLIL, I can’t see it. (Cambridge University Press ELT, 2010 3: 22-4: 10) Of course, teachers and teaching matter to the implementation of any form of bilingual education. This premise is the basis of this study as well as any investigation which seeks to understand teachers, their thinking, and their practices. To deny, however, that poor quality CLIL is not CLIL would seem to be a very advantageous way for proponents of CLIL to define CLIL and one that ignores CLIL’s central feature, namely, that CLIL is an educational approach “in which pupils learn a subject through the medium of a foreign language” (Eurydice, 2017, p.-13). CLIL, then, is uniquely European with particular parameters that set it apart from other forms of bilingual education such as Canadian-style immersion. However, for any conceptualization of bilingual education to be useful in Europe, it needs to be sufficiently broad in order to include the diversity of European bilingual education programs. Among the diversity of European bilingual education are immersive programs such as those in Swedish-speaking regions of Finland or Basque-speaking regions of Spain. In order to include these types of programs in a single European term, CLIL is used as an “umbrella term [that] encompasses provision where some or all non-language subjects are taught 2.1 CLIL and Immersion: Foundations and principles 29 <?page no="30"?> through a language designated as a foreign language in the curriculum” (Eurydice, 2017, p. 13). This broad description includes immersion programs as well as programs where content courses are taught through a foreign language. Broadly speaking, descriptions of bilingual programs differentiate themselves regarding the degree to which they focus on content or language and the amount of time spent learning in the target language. Tedick and Cammarata (2012) visualize this differentiation on two axes (see Figure 1). Time is represented on the vertical axis. Content and language are represented on the horizontal axis. Programs placed above the horizontal axis are those which include 50 % or more of instructional time in the target language. These include early, mid, or late partial or full immersion programs. Programs placed below the line teach less than 50 % of the content in the target language. These include subject classes taught through a foreign language as well as language classes with a thematic approach. All of these programs can be designated as CLIL programs. While one could imagine a high time-intensive language driven program existing in an extracurricular private language school, these programs do not exist in public schools. Most primary school CLIL programs in Germany would fall in the lower left quadrant. 30 2 CLIL and its Foundations in Theories of Second Language Acquisition <?page no="31"?> High-Time-Intensive Early, mid, late total immersion CLIL CLIL CLIL CLIL CLIL Early, mid, late partial immersion CLIL Content-Driven Language-Driven CLIL CLIL CLIL CLIL CLIL CLIL Subject courses taugh t Theme-b ased language in 12 classe s CLIL CLIL CLIL Content-related early Content-base d early language learning language learning pro grams Low Time-Intensive *Figure adapted from Met (1999) and Cummins (1982). Figure 1: Range of programs that integrate content and language* (Tedick and Cammarata, 2012) 2.1.3 CLIL goals CLIL programs are fundamentally motivated by a desire to improve students’ L2 compe‐ tences while maintaining their L1 development and non-language content learning. As non-language subject courses taught through a foreign language, CLIL programs in NRW, however, have the primary goal of content learning (MSB NRW, 2022, January 29). Massler (2013) also notes that this type of CLIL (subject courses taught through a foreign language) has the primary goal of teaching the non-language subject competences. In NRW, these competences are defined and outlined for teachers in the core curriculum (MSB NRW, 2021). For the subject of Science and Social Studies (Sachunterricht), the broad goal is to support children as they explore the world and their environment (Kahlert, 2022). Aristov and Haudeck (2013, p. 45) state that the primary goal of science lessons is “to provide pupils with meaningful, memorable, science experiences upon which they will build their further 2.1 CLIL and Immersion: Foundations and principles 31 <?page no="32"?> understanding of natural phenomena.” Furthermore, the sciences core curriculum in NRW describes the goal of science lessons as promoting an attitude of curiosity and questioning that can orient students in their environment and support them to take action in their environment (MSB NRW, 2021, p. 178). Within the curriculum, these broad goals are further organized and articulated within six areas of competence (to be further described in section 3.2.4 below). CLIL and EFL instruction differ in their focus on the development of foreign language competences. While EFL lessons tend to focus on linguistic accuracy, CLIL programs focus more on the development of linguistic fluency (Baetens Beardsmore, 2009, p. 212). Specific L2 learning goals will always vary based on individual CLIL programs. With no CLIL-specific curriculum or other guiding documents for primary school CLIL in NRW, teachers are forced to seek guidance from other sources such as the English curriculum for primary school in NRW or CLIL colleagues. Mehisto (2013, p.-26) summarizes L2 goals for CLIL as “age-appropriate levels of advanced proficiency in L2 reading, writing, speaking and listening comprehension.” More contentious in the field of CLIL is the appropriate role of the L1 and therefore L1 learning goals. Attitudes toward the role of the L1 generally fall on a continuum between those that promote a monolingual L2 approach with little or no attention paid to the L1 to those that promote a bilingual approach with planned L1 and L2 learning goals. The fact that CLIL is a form of bilingual education would suggest that, as Elsner and Keßler (2013a, p. 1) point out, “two or more languages [are used] in the classroom for communicative purposes.” Diehr (2012) analyzes how various CLIL programs position L1 and L2 use and learning goals and describes three types of CLIL. In Type A, the foreign language is used almost exclusively as the medium of instruction. While the L1 can be used in this type of instruction, it is not planned for and is only used when communication and understanding has broken down. In this way, developing L1 competences is not a goal of this type of instruction. In Type B, the foreign language remains the dominant language of CLIL instruction. However, unlike Type A, the L1 is planned for and used to support input comprehension and language production. Regardless of the supportive role that the L1 may play, it is not intended to lead to the development of L1 competences. In Type C, the L1 and the L2 play complementary roles but do not necessarily receive equal amounts of time in classroom instruction. Unlike Type B, where the L1 is limited to a supportive role, in Type C L1 subject vocabulary and, importantly, L1 subject discourse competences are targeted goals. For Diehr (2012), the development of L1 subject discourse competences must be a goal of CLIL instruction. Specifically, she advocates for the development of subject literacy in both the L1 and L2 or what she terms “doppelte Fachliteralität.” The Standing Conference of the Ministers of Education and Cultural Affairs (Kultusministerkonferenz) states as much describing the goal of CLIL instruction in Germany as the “development of subject competences which includes subject discourse competence in two languages” (KMK, 2013, p. 7, my translation). The goal of developing subject literacy in two languages encompasses the three goals more traditionally cited as CLIL goals: learning the L1, L2, and content. A reformulation of these three goals into one is likely a closer representation of the relationships between languages and content. As Zydatiß (2012, p. 71) states, “Content and discourse cannot be separated from one another” (my translation). 32 2 CLIL and its Foundations in Theories of Second Language Acquisition <?page no="33"?> The language competences developed as a part of subject literacy include those labeled by Cummins (2008) as cognitive academic language proficiency (CALP). CALP describes a student’s “ability to understand and express, in both oral and written modes, concepts and ideas that are relevant to success in school” (Cummins, 2008, p. 71). While it is important that students begin to develop CALP in the L1 and L2 in primary school CLIL lessons, the development of BICS (basic interpersonal communicative skills) (Cummins, 2008) is also important. Primary school CLIL students are usually at the beginning stages of L2 development. CLIL, as a learning environment with more L2 input and opportunities for L2 output, can, therefore, simultaneously support the development of L2 BICS and CALP as well as L1 CALP. While using the L1 and L2 in the instructional register will be necessary to develop subject literacy in both languages, Dalton-Puffer (2007, p. 203), based on her investigation CLIL classroom discourse in Finland and Austria, concludes “without reservation” that the regulative register should be conducted exclusively in the L2 in order to support the development of foreign language BICS competences. The literature on CLIL also emphasizes the development of intercultural competences as a CLIL goal (see, for example, Diehr, 2012; Mehisto, 2013; Wolff, 2016). The core curriculum for primary school English in NRW emphasizes three aspects of intercultural learning: “soziokulturelles Orientierungswissen,” “interkulturelle Einstellungen und Bewusstheit,” and “interkulturelles Verstehen und Handeln” (Sociocultural knowledge, Intercultural attitudes and awareness, and Intercultural understanding and action) (MSB NRW, 2021). Students should become familiar with English-speaking cultures through a variety of media and be prepared “to communicate in authentic situations in ways that are culturally sensitive and appropriate to the situation and addressee” (MSB NRW, 2021, p.-42, my translation). A more comprehensive view of intercultural learning is described by Byram (1997). He describes the concept of intercultural communicative competence (ICC) as consisting of four principal savoirs, namely, knowledge of self and other, attitudes, skills of interpreting and relating, and skills of discovery and interaction, which, taken together, lead to critical cultural awareness (Byram, 1997). Due to young children’s cognitive development, “critical cultural awareness” is a “complex” and “unrealistic” goal for primary school children to achieve within the context of primary school EFL lessons (Brunsmeier, 2016, p. 326). Brunsmeier (2016), based on her investigation of ICC in the primary school EFL classroom, developed the concept of Reflektierten Interkulturellen Aufmerksamkeit (Reflected Intercul‐ tural Awareness) (RIA). RIA is intended to work both with Byram’s model but also to offer an alternative goal more appropriate to the development of primary school children. As Brunsmeier (2016, p.-327) describes it, “Es handelt sich bei RIA also um eine Kompetenz, die sich im Prozess entwickelt und die den Grundstein für IKK legt” (RIA is therefore a competence that develops through a process and lays the foundation for ICC). With a focus on developing subject literacy in two languages (Diehr, 2012), CLIL is uniquely positioned to support the development of Brunsmeier’s (2016) construct of RIA. One way this is accomplished is through attention to “partial equivalencies” (Lindemann & Diehr, 2016). Partial equivalencies are terms in two languages that signify many overlapping concepts but that also differ in ways related to culture and usage. Diehr (2016, p. 72) gives the example of concepts such as SCIENCE and WISSENSCHAFT. While these two concepts overlap in many ways, the terms “science” and “Wissenschaft” are used by native speakers of 2.1 CLIL and Immersion: Foundations and principles 33 <?page no="34"?> English and German to identify categories that do not always directly correspond. Within the CLIL classroom, it is the teacher’s responsibility to be sufficiently knowledgeable of how language is used in the target culture and then to identify examples of partial equivalency related to the topic of study (Lindemann & Diehr, 2016). Partial equivalencies, therefore, provide a window through which to explore and compare cultures. To summarize, CLIL and immersion are two different but related approaches to second or foreign language learning each of which grew out of a unique cultural, political, and linguistic landscape. In Europe, the term CLIL was coined to described the variety of bilingual education programs that already existed. As such, it is a term that is used broadly to include immersion-style programs seen in places such as Swedish-speaking areas of Finland as well as much more limited approaches to bilingual education such as CLIL modules. In Germany and more specifically in NRW, the main goal of CLIL programs is non-language content learning. However, this goal is intricately tied to goals of L1 and L2 development. Based on the development of subject literacy in two languages (Diehr, 2012) CLIL programs have the potential to further develop students’ intercultural awareness through, for example, attention to partial equivalencies (Lindemann & Diehr, 2016). 2.2 Theories of second language acquisition Three broad categories of second language acquisition (SLA) theories can be used to describe how a foreign language is learned or acquired within a primary school CLIL context. Key constructs in usage-based theories, sociocultural theory, and complexity theory will be described below. 2.2.1 Usage-based theories: Input, output, and interaction Linguistic input is the “sine qua non” of language acquisition (Gass & Mackey, 2015, p. 181). CLIL, in its most rudimentary form, is education with more foreign language input than non-CLIL teaching forms. With increased exposure to linguistic input, it is assumed that students will develop, at least to some degree, foreign language competences (Vollmer, 2010; Wolff, 2016). Marsh (2000, p. 3) describes CLIL as offering students a “natural situation for language development” and that “[i]t is this naturalness which appears to be one of the major platforms for CLIL’s importance and success” (italics in original). Linguistic input and the notion of natural language learning are combined in what Dalton-Puffer et al. (2010a, p.-6) label “reception-based” theories of language development. Krashen (1985) laid out five hypotheses that made up his theory of second language acquisition. Central to these hypotheses is the “Input Hypothesis” (Krashen, 1985, p. 1) which states that we acquire language only through “comprehensible input” and that learners move from their current stage of foreign language development (i) to the next stage of development (i+1) only by comprehending input at the i+1 level (Krashen, 1985, p. 2). Input can be made more comprehensible by, for example, providing visual support or discussing topics already familiar to students (Krashen, 1985, p. 2) as well as through “simplification” and “elaboration” (Long, 1996, pp. 421-422). Simplification increases comprehensibility by using “shorter, syntactically less complex utterances or 34 2 CLIL and its Foundations in Theories of Second Language Acquisition <?page no="35"?> sentences” (Long, 1996, p. 421) while elaboration increases comprehensibility through repetition and paraphrasing (Long, 1996, p.-422). While Krashen (1985) assigns a causal role to the construct of “comprehensible input,” Smith (1991, 1993) assigns no such role to his construct of “input enhancement.” According to Smith (1993, p. 176) input enhancement “implies only that we can manipulate aspects of the input” but that any effect on the language learner cannot be assumed. Enhanced oral input may include, for example, “unelaborated” forms of enhancement such as using gestures or stressing particular language features as well as “highly elaborate” forms of enhancement that include the explanation of grammatical constructions (Smith, 1993, p.-177). In a CLIL context, natural input, or input that is not enhanced, is assumed to lead to acquisition of a foreign language. Krashen (1985) distinguishes between acquisition and learning. Acquisition is the subconscious processing of language, similar to that of learning a first language, while learning is the conscious process of language learning (Krashen, 1985, p. 1). It is only through acquired language competence that students can speak in a foreign language (Krashen, 1985, p. 1). By increasing the amount of foreign language input in a student’s school week, CLIL offers more opportunities for comprehensible input and, therefore, language acquisition. While comprehensible input is clearly necessary for language acquisition, Krashen (1985, p. 3) states that learners need to be “open” to the input. He uses the metaphor of the “affective filter” to describe the attitude that learners have toward the language learning process. If the filter is “up,” learners have generally negative emotions such as anxiousness or lack of motivation toward learning the language. If the filter is “down,” emotions are generally positive toward language learning (Krashen, 1985, p. 3). A low affective filter, therefore, is necessary for students to process comprehensible input. While higher motivation levels toward language learning has been cited as a reason for implementing CLIL programs (Piske, 2015), Rumlich (2015) concludes, based on his study of secondary school CLIL learners, that reaching a motivational threshold is likely more important than absolute motivation levels. Krashen (1985, p. 4) claims that factors other than input are only necessary for language acquisition to the extent that they contribute to comprehensible input or to a low affective filter. Swain (1995), however, argues that output, or speaking and writing, contributes to language acquisition in ways that go beyond producing input for oneself. At a basic level, output production serves a practicing function which potentially improves learner’s spoken fluency (Swain, 1995, p. 125) or “automatictiy” (Gass & Mackey, 2015, p. 185). Beyond merely practicing language, Swain (1995, 1998) describes three principal functions of output that support the development of linguistic accuracy. The underlying role of output for all three functions is the development of syntax and morphology (Swain, 1995, p. 128). The first function is the “noticing” function. When learners speak, they may notice a “gap between what they want to say and what they can say” (Swain, 1995, p. 126, italics in original) and hence take action to fix the problems. Swain and Lapkin (1994 as cited in Swain 1995) conducted research with eighth grade French immersion students using think aloud protocols. The researchers concluded that evidence of noticing exhibited by the students supports the noticing hypothesis. Second 2.2 Theories of second language acquisition 35 <?page no="36"?> is the “hypothesis testing” function. By producing output, either in spoken or written form, learners are able to test hypotheses they have about how the language works and receive feedback on their hypotheses. Testing a hypothesis requires action on the learner’s part, or, as Swain puts it, “learners need to do something” (Swain, 1995, p. 131, italics in original). Third, is the “metalinguistic function.” Swain explains that when considering the hypothesis testing function of output production, the output itself is the hypothesis. What learners think about language is deduced from what they say (Swain, 1995, p. 132). The metalinguistic function of output, on the other hand, involves “explicit hypothesizing” (Swain, 1995, p. 132) in which learners reflect on language through language. Within a primary school CLIL context, this function of output may be less relevant since most primary school learners lack the foreign language competences necessary to reflect on the foreign language. In this context, reflection on foreign language use would normally take place through the school language or mother tongue. Swain (1995, p. 126) suggests that producing output is about challenging learners and pushing them “to process language more deeply (with more mental effort)” than they do when processing input. While comprehending input is a cognitively active process, it is possible, from a student perspective, to passively take language in without actively engaging with it. Output, however, demands action from the learner. It puts learners in control and may encourage them to take more responsibility for their learning (Swain, 1995, p.-126). Long (1996) takes input and output a step further and proposes the “interaction hypoth‐ esis.” While Long describes input as necessary for L2 acquisition, he claims it is not sufficient for language acquisition (Long, 1996, p. 423). The interaction hypothesis and its core construct of negotiation for meaning “connects input, internal learner capacities, […] and output in productive ways” (Long, 1996, p. 452). Negotiation for meaning is “the process in which, in an effort to communicate, learners and competent speakers provide and interpret signals of their own and their interlocutor’s perceived comprehension, thus provoking adjustments to linguistic form, conversational structure, message content or all three, until an acceptable level of understanding is achieved” (Long, 1996, p. 418). “Interactional adjustments” (Long, 1996, p. 451) that lead to increased comprehensibility is key to this definition. Principally, negotiation for meaning involves the more competent speaker using “repetitions, extensions, reformulations, rephrasings, expansions and recasts” resulting in “denser than usual frequencies of semantically contingent speech” (Long, 1996, pp. 451-452). In addition to supporting comprehension, negotiation for meaning has the potential to focus learners’ attention on form-meaning relationships and on language production (Gass & Mackey, 2015, pp. 187-188). Long (1996), while not addressing classroom contexts specifically, describes linguistic environments as fundamentally providing positive and negative evidence of language use. Positive evidence includes grammatical and acceptable usage of language while negative evidence includes explicit or implicit feedback on ungrammatical or unacceptable usage (Long, 1996, p. 413). Negative evidence has also been described as “corrective feedback” (Gass, 1997). One role for primary school CLIL teachers, then, would be to provide an abundance of positive language evidence as well as providing corrective feedback in age and developmentally appropriate ways to their learners. Gass and Mackey (2015, p. 188) state 36 2 CLIL and its Foundations in Theories of Second Language Acquisition <?page no="37"?> as much when they describe negotiation as allowing input to be matched to an individual’s “communicative needs” and “developmental levels.” The need for negotiation in interaction is caused by miscommunication between inter‐ locutors (Gass, 1997). Miscommunication can involve misunderstanding, i.e. interpreting what the speaker has said as something other than what the speaker intended, or incomplete understanding, i.e. understanding none or only part of what has been said (Gass, 1997, p. 106). It is at this breakdown in communication where, in a classroom context, the teacher can provide corrective feedback on a learner’s utterance. Gass (1997) outlines three forms of this corrective feedback, referred to as the “three Cs” (Gass & Mackey, 2015, p. 187). First, comprehension checks function to check whether or not the interlocutor has understood the previous utterance, e.g. “Do you understand? ” Second, clarification requests function to amend incomplete understanding by explicitly asking for clarification, e.g. “What did you say? ” Third, confirmation checks function to check whether or not the hearer has correctly understood the speaker, e.g. “Did you say dog? ” While the focus on input and output for language learning sees the learner as essentially processing language internally, the interaction hypothesis places the locus of learning in the dynamic exchange between interlocutors. From this perspective, language learning takes on a more collaborative and social dimension. Following will be a further discussion of a sociocultural theory of language learning. 2.2.2 Sociocultural theory: Mediation, zone of proximal development, and collaborative dialogue While the previously described theories of SLA place the core of language learning on individual cognitive mechanisms, sociocultural theories of SLA focus on social contexts in which (language) learning takes place. A central construct in sociocultural theory is mediation. Mediation is a process that works to coordinate (or mediate) “the relationship between the individual and social-material world” (Lantolf, Thorne, & Poehner, 2015, p. 208). In the case of language learning, mediation is any process that helps to support the learner in acquiring the target language. This might include, for example, a teacher’s use of visualization to support meaning, a slowed speech rate to increase the comprehensibility of input, or corrective feedback on ungrammatical output. Regulation is a form of mediation. Lantolf et al. (2015, p. 209) describe three types of regulation which follow a “developmentally sequenced shift” from an external to internal locus of control. Object-regulation is about using material things in one’s environment to support thinking and mental action (Lantolf et al., 2015, p. 209). A language learner might use a pencil and paper to make a list of new vocabulary to memorize or use a translation app to translate a complex text passage. Other-regulation is about other people mediating language development (Lantolf et al., 2015, p. 209). This can include, for example, a teacher providing corrective feedback to a student or another language learner asking for clarification while working on a joint task. Self-regulation describes an individual who has “internalized external forms of mediation” (Lantolf et al., 2015, 209). Self-regulation does not necessarily describe fully competent users of a language but instead can be applied to learners who have internalized any aspect or feature of a language. 2.2 Theories of second language acquisition 37 <?page no="38"?> The process of regulation ending in self-regulation and internalization is central to a sociocultural view of learning. Internalization is the process “through which cultural artifacts, including language, take on a psychological function (Kozulin, 1990, as cited in Lantolf et al., 2015, p. 211). Lantolf et al. (2015) cite Vygotsky who describes internalization: Any function in the child’s cultural development appears twice, or on two planes. First it appears on the social plane, and then on the psychological plane. First it appears between people as an interpsychological category, and then within the child as an intrapsychological category. (Vygotsky, 1981, p.-163 as cited in Lantolf et al., 2015, p.-211) Within a primary school CLIL context, mediation of the foreign language takes place through the teachers, students, and objects present in the classroom and school environ‐ ment. As individuals engage with the objects and people in their environments in the process of learning, there is a metaphorical space, the zone of proximal development (ZPD), in which learners are able to move beyond their current level of learning with the assistance of a more competent individual. Vygotsky defines the ZPD as follows: [T]he distance between the actual developmental level as determined by independent problem solving and the level of potential development as determined through problem solving under adult guidance or in collaboration with more capable peers. (Vygotsky, 1987, p. 86 as cited in Lantolf et al., 2015, p.-212) The ZPD is the space in which targeted mediation can have an effect on learners’ language acquisition. Corrective feedback, for example, that is outside of the ZPD is unlikely to be internalized by the learner. Because the ZPD is defined, in part, by what learners can do with support from a more competent person, the ZPD is forward looking and provides evidence of a learner’s potential (Lantolf et al., 2015, p.-212). In order to learn language, “speaker/ hearer[s]” need to interact with others in order to build “grammatical, expressive, and cultural competence” (Ohta, 2000, p. 51). This collab‐ orative process of language learners speaking and listening to one another and thereby developing their language competences is known as collaborative dialogue. Collaborative dialogue is “dialogue in which speakers are engaged in problem solving and knowledge building. It heightens the potential for exploration of the product” (Swain, 2000, p. 102). The “product” in this case is the utterance that has been spoken aloud by interlocuters. Producing output in the context of collaborative dialogue not only forces cognitive activity, it creates a language product or utterance that can subsequently be investigated by any of the interlocutors (Swain, 2000, p. 102). For Swain (2000), dialogue, and specifically collaborative dialogue, is the context in which language mediates and is mediated. Learners produce utterances about a topic and sometimes that topic is the language of their dialogue. For example, learners may use an ungrammatical form (e.g. “I goed”), then question the use of that form (e.g. “Is goed the past tense of go? ”), and then, through dialogue and possibly use of a reference text, gain knowledge about that form. A sociocultural view of language learning emphasizes the role that the external envi‐ ronment has on the learning process. Learning language is not merely an internal process, it is a process that is mediated from the external to the internal. 38 2 CLIL and its Foundations in Theories of Second Language Acquisition <?page no="39"?> 2.2.3 Complex systems approach: Affordances, co-adaptation, and learner agency Learning a foreign language is complex and depends on many factors (Legutke, Müller-Hartmann, & Schocker-v.Ditfurth, 2014, p. 15). While discussions of input, output, and mediation do not deny the complexity of language learning, they do attempt to isolate aspects of language learning. Proponents of an “ecological perspective” (van Lier, 2000) toward language learning or complexity theory (CT) (Larsen-Freeman, 2015) reject this kind of “scientific reductionism” (van Lier, 2000). This kind of reductionism fails to take into account the role of the learner. It “dehumanizes” the learner and “metaphoriz[es] him or her as a computer” (Larsen-Freeman, 2015, p.-235). CLIL practices and programs are generally justified on the basic principle that changing students’ linguistic environment by increasing exposure to a foreign language will support the learning of the target language. In primary school CLIL, that means increasing the amount of L2 input and, when appropriate, supporting L2 output. A CT view of language learning takes a holistic view of the learning environment. Not only is there increased L2 input and output in, for example, the CLIL environment, there are objects in the environment as well as teachers and students, each with his or her own attitudes, moods, experiences, relationships, competences etc. All of these features of the learning environment interact in complex ways to support or hinder language acquisition. Language acquisition, therefore, is a highly subjective and individual process and often proceeds in nonlinear ways (Larsen-Freeman, 2015). Traditional views of language learning assign causality to constructs such as collabora‐ tive dialogue or negotiation for meaning. When learners engage in collaborative dialogue, for example, they may be forced to notice and fix gaps in their knowledge of language. Engaging in collaborative dialogue caused the learner to acquire a feature of the language. CT describes instead the construct of “affordance” (van Lier, 2000; Larsen-Freeman, 2015) but does not assign it a causal role. Van Lier (2000, p. 252) describes an affordance as follows: a particular property of the environment that is relevant - for good or ill - to an active, perceiving organism in that environment. An affordance affords further action (but does not cause or trigger it). What becomes an affordance depends on what the organism does, what it wants, and what is useful for it. An affordance, then, is an opportunity in which an active learner might be able to acquire some aspect of language. There is a relationship between the learner and the feature of the environment (van Lier, 2000; Larsen-Freeman, 2015). In their environments, learners are repeatedly exposed to linguistic features and adapt to those features over time resulting in language development. Co-adaptation is the process of adjusting one’s conversation patterns to the other interlocutor in a dialogue in order to match each other’s needs (Larsen-Freeman & Cameron, 2008). Iteration, or the repeated encountering of linguistic patterns, is one aspect of this process. By repeatedly encountering linguistic patterns, learners imitate them in ways that best suit their own needs and wants (Larsen-Freeman, 2015, p. 229). In this description of co-adaptation, there are clear echoes of constructs such as collaborative dialogue (Swain, 2000) and negotiation for meaning (Long, 1996). This is not surprising considering that CT “adopts a sociocognitive view of 2.2 Theories of second language acquisition 39 <?page no="40"?> second language development” (Larsen-Freeman, 2015, 229). CT is not describing a different language learning reality. It is describing language learning reality differently. In addition to affordances and co-adaptation, CT assigns a significant role to learners in language learning. Because an affordance is only an opportunity for learning, the learner must be active or agentic. The learner must do something. What the learner does is largely dependent on his or her own experiences and perceptions of the environment. A learner’s environmental perceptions include both awareness of positive evidence of language use as well as negative evidence, i.e. differences between his own language and that of others. A learner’s awareness can be self-initiated or other-initiated (see Larsen-Freeman, 2015, p. 239). As Larsen-Freeman (2015, p. 239) states, “Learning is motivated by an awareness of difference.” This description of “awareness” is similar to Schmidt’s (1990) construct of “noticing” which claims that nothing is learned unless it is noticed. Complexity theory describes language learning from an ecological perspective. In doing so, it uses and reformulates existing constructs from the SLA literature. In many ways, it seeks to unite concepts from a cognitive perspective on language learning such as “input” with concepts from a sociocultural perspective on language learning such as “collaborative dialogue” while at the same time always keeping the uniqueness of each learner, his environment, and their interaction at the center of an individual and often unpredictable learning process. As Larsen-Freeman states, “Teaching does not cause learning; learners make their own paths. They have agency. In this way, learning another language is not about conformity to uniformity” (The New School, 2018). 2.3 Summary In this chapter, I have described the origins of immersion and CLIL pedagogy, programmatic features of these pedagogies, and outlined CLIL goals in the German context. Furthermore, I have outlined three broad categories of theories of second language acquisition that are particularly relevant to the theoretical foundations of CLIL. CLIL and immersion, as forms of bilingual education, have unique origins and goals and yet, ultimately, are pedagogical approaches intended to support the development of second or foreign languages. CLIL, since its coining in 1994, has been used broadly in Europe to include the variety of bilingual instruction that exists across the continent. CLIL instruction in Germany seeks to develop subject literacies in two languages (Diehr, 2012; KMK, 2013), a goal which comprises non-language content learning as well as L1 and L2 development. Furthermore, CLIL seeks to develop intercultural competences in students. CLIL, as a pedagogical approach to language learning, is based on the assumption that a learning context with more L2 linguistic input and, to some extent, linguistic output will lead to increased L2 competences. The theories and their constructs described above suggest that, while important to language learning, a pedagogy based solely on increased input and output will not always be sufficient for learners’ language development. Instead, such a pedagogy must consider not only language use but the social and cultural context of the learning environment as well as the uniqueness of individual learners. 40 2 CLIL and its Foundations in Theories of Second Language Acquisition <?page no="41"?> 3 Early Foreign Language Learning at Primary Level CLIL practices at primary school are based in part on general principles and practices of teaching young children at primary school. Therefore, in the following section I will review the theories of constructivism and sociocultural theory (SCT), both of which have traditionally played an important role in the development of pedagogical practices at primary school (see Cameron, 2001; Pinter, 2017) and which remain relevant to a discussion of foreign language learning in CLIL contexts. Constructivism is based on the work of the Swiss psychologist Jean Piaget. The core idea of constructivism is that children learn by acting in the physical world. Learning is constructed through direct and concrete action. Action is internalized and transformed into thought (Cameron, 2001, p.-3) as learners progress through a series of four developmental stages which describe the logical abilities or “operations” of the child (Pinter, 2017, p. 6). Progress through the stages is sequential and determined by biologically controlled development (Pinter, 2017, p. 6). A constructivist understanding of learning views the child “as an active learner and thinker, constructing his or her own knowledge from working with objects or ideas” (Cameron, 2001, p. 4). Because primary school children span ages that may fall within the pre-operational, concrete-operational, and the beginnings of formal-operational stages of development, teachers need to understand these stages and their pedagogical implications (Pinter, 2017, p.-7). Sociocultural theory (reviewed above in the context of language learning and here, again, briefly, in the context of general child development) and “social-constructivism” (Pinter, 2017, p. 10) is based on the work of the Russian psychologist Lev Vygotsky. The core idea of SCT is that human learning is “fundamentally a mediated process” (Lantolf et al., 2015, p. 207). Interaction between the learner and the world is mediated by concrete and abstract tools. Learning is a process of moving from externally regulated mediation, such as using a dictionary to look up a new word in a book, to internally regulated mediation, such as already knowing the word when read in the book (Lantolf et al., 2015, p. 209). The metaphorical space between what the learner can do independently and what the learner can do with the support of an adult or more capable peer is known as the zone of proximal development (Lantolf et al., 2015, p.-212). Constructivism and SCT differ in two important ways. The first way relates to the role of other people and language. SCT gives more importance to language and other people in the learning process than does constructivism (Cameron, 2001, p. 5). Viewing learning through a constructivist lens, one sees the child as “an active learner alone in a world of objects” (Cameron, 2001, p. 6) whereas viewing learning through an SCT lens, one sees the child as “an active learner in a world full of other people” (Cameron, 2001, p. 6). The second difference relates to the role of formal instruction. One of Vygotsky’s central findings was that “instruction, especially formal instruction in school precedes and shapes development” (Lantolf et al., 2015, p. 212). This finding contradicts a Piagetian notion of exploratory learning by suggesting that instruction and interaction play a primary role in a child’s development. Where Piaget saw learning as occurring naturally through a child’s independent action in the world, Vygotsky saw learning as an interactive process between <?page no="42"?> the child, others, and the world. Piagetian and Vygotskian views of learning provide the broad foundation for the range of principles, methods, and approaches to learning in primary school CLIL and EFL contexts. A review of some of these key principles follows. 3.1 Principles of foreign language instruction at primary school Buchholz (2007, p. 51) insists that it is necessary for primary school CLIL programs to “es‐ tablish explicit connections with the didactic principles of the primary school curriculum.” These didactic, or pedagogical, principles relate to the variety of subjects taught at primary school as well as the development of primary school students. Furthermore, Morton (2012, p. 12) describes CLIL as a “complex hybrid of practices [which draw] on the pedagogies of different academic subjects and those of language education.” CLIL, as an interdisciplinary pedagogy (KMK, 2013, p. 7), is therefore built on existing pedagogies and principles that are relevant both to the age and development of learners as well as to specific subjects and language learning. No unified CLIL pedagogy exists (KMK, 2013, p. 7). In this study, I am focused on teachers’ language learning related cognitions about their CLIL practices. Therefore, in this section I will review eight key pedagogical principles of foreign language teaching at primary school that are relevant to CLIL pedagogy at primary school. 3.1.1 Authenticity “Authenticity” is a contested and often debated concept in foreign language teaching (see, for example, Holliday, 2013) and is based on various philosophical foundations (see van Lier, 1996; MacDonald, Badger, & Dasli, 2006). Any attempt at a simple and straightforward definition of authenticity in the context of language teaching is quickly met with questions and protestations. Sartre’s statement, “If you seek authenticity for authenticity’s sake, you are no longer authentic” (1992, p. 4 as cited in van Lier, 1996, p. 123) gives a sense of the tentativeness with which one should approach the concept. Traditional definitions of authenticity in foreign language teaching focus on authentic language, authentic material, and authentic tasks. At a basic level, all three domains of authenticity attempt to describe the extent to which the language, materials, or tasks used in the foreign language classroom represent or relate to the language and culture of a native-speaking context. Furthermore, Cooper (1983, p. 15 as cited in Macdonald et al., 2006, p. 250) describes “correspondence” and “genesis” as two ways in which authenticity is used in philosophy and which have relevance to the field of applied linguistics. A picture book such as Last Stop on Market Street (de la Peña, 2015) is authentic in the sense of “genesis” because it is written by an American author for American children. Alternatively, a classroom task such as role-playing a visit to the ice-cream parlor in London is authentic in the sense of “correspondence” because it attempts to recreate the language and discourse patterns needed and used in such a situation. Authentic language, simply defined, is “language used in a real context” (Larsen-Freeman & Anderson, 2011). Böttger (2020) places the responsibility of authentic language on the shoulders of the teacher. In an EFL context where the teacher is very likely not a native speaker of English, the principle of authentic language is met when the teacher is 42 3 Early Foreign Language Learning at Primary Level <?page no="43"?> able to pronounce language competently and error-free and has a command of idiomatic expressions (Böttger, 2020, p. 78). This is not to suggest that the spoken language needs to be free of an accent, as accents and dialects are naturally occurring aspects of language. Likewise, the extent to which the teacher’s spoken language is grammatically correct is not an indicator of authenticity. Instead, a teacher’s language use is more or less authentic to the extent that it is representative of native speaker speech as evidenced in corpora of native speakers’ spoken language (Böttger, 2020, p.-79). Authentic materials are another aspect of authenticity and stand in contrast to peda‐ gogical materials (Böttger, 2020, pp. 80-81). Authentic materials are generally considered to be those created by native speakers for native speakers in native speaking contexts. Nunan (2004, p. 49) defines authentic materials as those which have been “produced for purposes of communication, not for purposes of language teaching.” Böttger (2020, p. 77) describes authentic materials as those materials with a native speaking country of origin and that were not created for teaching purposes. An example of this kind of authentic material would be the picture book No, David! (Shannon, 1998) written by the American David Shannon, published in the United States, and with no explicit pedagogical purpose. However, materials created by native speakers for native-speaking children at school can also be considered authentic (Böttger, 2020, p. 77). An example of this kind of authentic material would be books from the Oxford Reading Tree series (Oxford Owl, 2020, October 7) designed to teach British children to read English. In addition to authentic written materials, there are authentic audio-visual materials (Böttger, 2020). Examples include songs, videos, and radio programs. Through the use of authentic audio-visual material in the classroom, teachers can increase students’ exposure to authentic language use. Finally, realia such as bus tickets, advertisements, and currency that originate in native-speaking countries are also authentic materials (Böttger, 2020). Böttger (2020) describes semi-authentic materials as those with native-speaker origins but tailored in various ways for foreign language audiences. Examples include materials created or published by native speakers specifically for, in this case, German-speaking children or, more commonly, authentic material that has been edited in some way to meet the linguistic needs of a specific group of children (Böttger, 2020, pp. 80-81). Ultimately, through careful planning and the use of authentic language and materials, children should get a view of “real-life” in the target culture (Böttger, 2020, p. 79). This supports the development of cultural awareness (Böttger, 2020, p. 79) and intercultural competences (MSB NRW, 2021). Finally, tasks (see Willis & Willis, 2007; Müller-Hartmann & Schocker-von Ditfurth, 2011) or what children do in the classroom can be described as either authentic, sometimes referred to as “real-world” (see Cameron, 1997; Willis & Willis, 2007), or pedagogic. Pedagogic tasks are those typically found only within a classroom setting, for example answering comprehension questions or writing vocabulary lists. Alternatively, authentic tasks are those that attempt to recreate situations in which students would use language authentically outside of the classroom. This might include tasks related to travel such as booking a flight (Willis & Willis, 2007, p. 138) or foreign employment such as interviewing for a job (Müller-Hartman & Schocker-von Ditfurth, 2004, p. 41). Of course, there is likely nothing authentic or real-world about these tasks for young children. Therefore, in the 3.1 Principles of foreign language instruction at primary school 43 <?page no="44"?> context of teaching English at primary school, authentic tasks or activities are those that are “meaningful to the students - authentic to their lives” (Holliday, 2013, p. 21) or have been authenticated by students through a “personal process of engagement” (van Lier, 1996, p.-128). 3.1.2 English as the language of instruction The English core curriculum for primary school in NRW describes “(funktional) einsprachig geführten Englischunterricht” ((functionally) monolingual English lessons) (MSB NRW, 2021) as the appropriate instructional approach for primary school English lessons. While this approach is generally accepted, there is a role for the L1. Böttger (2020, p. 86) summarizes the L1-L2 balance in English lessons with a rule of thumb, “So oft wie nur möglich Englisch, so wenig wie nötig Deutsch” (English as often as possible, German as little as necessary). Cameron (2001, p. 209) echoes this by advising, “[U]se as much of the target language as possible. […] [E]nsure that use of first language supports the children’s language learning.” Butzkamm and Caldwell (2009, p. 24) call students’ L1 their “strongest ally” and advocate for its targeted use in EFL instruction. In EFL contexts, the L1 can be used minimally to support students’ understanding and L2 development. In the context of immersion education, Swain and Lapkin (2013p. 122) offer three “principle uses” of the L1 and L2. First, students should be allowed to use their L1 alone or with others “in order to mediate their understanding and generation of complex ideas” (Swain & Lapkin, 2013, p. 122). Use of the L1 as a mediation tool often takes place during phases of a lesson in which students are preparing an oral or written end product in the L2. Swain and Lapkin further explain that as students’ L2 competences improve, they should be encouraged to use the L2 as a mediating tool but should also be allowed to revert to L1 use as necessary when, for example, content complexity increases. Second, Swain and Lapkin advise teachers to “set clear expectations about L1/ L2 use in order to create a secure classroom environment” (Swain & Lapkin, 2013, p. 123). Included in the creation of such expectations should be a process in which students and teachers make explicit their beliefs about “the cognitive/ emotive interface in language use and language learning,” thus encouraging a “climate of co-operation” (Swain & Lapkin, 2013, p. 123). Third, Swain and Lapkin call for teachers to always prioritize using the L2 as developing L2 competences is the goal of immersion education. Use of the L1, they say, should be “purposeful, not random” (Swain & Lapkin, 2013, p. 123). The L1 can be used for metalinguistic functions and for “mediat[ing] L2 development” (Swain & Lapkin, 2013, p. 123). For Swain and Lapkin, use of the L1 plays primarily a supportive or mediating role in the development of L2 competences and content understanding but not for the development of L1 competences. This role for the L1 differs from the role described in the German context by Diehr (2012) and Botz and Diehr (2016). Because students should be able to verbalize their subject knowledge and competences in German (L1) (Diehr, 2012, p. 28), Botz and Diehr (2016, p. 251) call for the sparse yet targeted use of the German language. Such targeted use of the L1 supports the development of L1 subject discourse competences. In the German context, the L1 may play a supportive role for the development of L2 and content competences but also must play a role for the development of subject literacy in the L1. 44 3 Early Foreign Language Learning at Primary Level <?page no="45"?> 3.1.3 Multisensory Learning Multisensory learning is learning with all of the senses (Böttger, 2020, p. 84). The senses, through which children perceive the physical world, provide the context in which children use language to communicate meaning (Richards & Rodgers, 2001, p. 117). Children need to be active or do something in order for their senses to be activated. Through action, language becomes concrete and, as Vollmuth (2004, p. 45) states, literally becomes “be-greifbar.” A multisensory approach to language teaching, therefore, can help engage young children in learning language. On the one hand, multisensory learning is holistic. It considers the whole child, the range of sensory channels available to her, and seeks to engage her in activities that activate each of these senses. From a neurophysiological perspective, activities that engage a variety of senses activate both hemispheres of the brain leading to expanded cognitive capacity (Waas & Hamm, 2004, p. 10). On the other hand a multisensory approach is a way to differentiate for different learner types, for example, more auditory learners or more visual learners. By offering a variety of activities that target different senses, a teacher can better support individual needs in a heterogeneous class. Music and drama are two popular ways of targeting linguistic competences through multisensory approaches in the primary school. Engh (2013) reviewed brain research that found that musical and linguistic structures are processed in the same location in the brain. This finding “leads to the hypothesis that syntax in language and music share a common set of processes (instantiated in front brain areas) that operate on different structural representations (in posterior brain areas)” (Patel, 2003, p. 674 as cited in Engh, 2013, p. 116). The exact pedagogical implications of this hypothesis are unclear. However, from a multisensory learning perspective, it would appear that using music to activate syntactic processing in the brain would be beneficial to language learning. Affective and motivational reasoning for using music in the primary school classroom is common (see, for example, Klippel, 2000; Pfitzer, 2006). Millington (2011) reviewed pedagogical research that suggests listening to songs supports the development of listening comprehension simply by providing practice listening to different forms of intonation and that singing develops pronunciation and vocabulary by providing a more enjoyable form of practice. Furthermore, the use of music within sensory, perceptual, and intuitive activities can potentially support the acquisition of rhythmic aspects of language (Vollmuth, 2004, p.-42). Drama, in both its structured and less structured forms, can be used to engage the head, heart, and hands (Haß, Kieweg, Kuty, Müller-Hartmann, & Weisshaar, 2006, p. 29) and is uniquely positioned to use auditory, visual, and haptic senses as well as kinesthetic, spatial, and verbal processing (Bland, 2015, pp. 221-222). For example, a hand puppet that only speaks English can whisper a forgotten word into the ear of a child (Haß et al., 2006) or a game of “Simon says” can be used to develop listening comprehension through movement (Klippel, 2000, p. 134). By providing a multisensory context, drama has the potential to open more learning channels for more children all while using language in context. In this way, Bland (2015) describes drama activities as holistic in two ways: in a whole language sense and in a whole person sense. From a multisensory perspective, drama activities in the EFL classroom are potentially “complete” (Bland, 2015, p.-222). 3.1 Principles of foreign language instruction at primary school 45 <?page no="46"?> 3.1.4 Play-based Learning Play-based learning is integral to any developmentally appropriate approach to the primary school EFL classroom. Games used in the EFL classroom can take a variety of forms and serve a variety of language learning goals. A game such as “memory,” for example, is usually known to students and can easily be restructured for language learning purposes (Waas & Hamm, 2004, p. 21). Games can be used in various social arrangements such as groups or the whole class (Waas & Hamm, 2004, p. 21), thus providing a flexible tool for meeting a variety of learning and organizational needs. Ultimately, for a game to be effective, it must be fun for children. This criteria for games is highly subjective (Klippel, 2000, p. 129). Therefore, teachers need to know their students and make decisions about the relative effectiveness of various games. If the game is fun and reaches the necessary level of “Spielspannung” (Klippel, 2000, p. 129) it can be an effective tool in supporting the unconscious acquisition of language (Klippel, 2000). It is because children focus on the play aspect of the activity that language is acquired unconsciously as a byproduct (Vollmuth, 2004, p.-42). 3.1.5 Visualization Visualizing language can take many forms including using pictures, graphics, tables, and diagrams as well as visualizing text through highlighting and underlining (Böttger, 2020). While visualizing language is an obvious component to a multisensory approach to language learning (described above), it is, at its core, a way to make language concrete and comprehensible to young children. In this way, it is an important form of input scaffolding and something teachers need to do to continually contextualize their language (Burmeister, 2006 as cited in Kersten & Rhode 2015). Additionally, pictures can motivate children, generate attention and curiosity, and help children retain foreign language better and longer (Böttger, 2020, p. 83) as well as provide important linguistic and cultural context (Wright, 1989, p. 2). Key to effectively visualizing language is to provide simple, concise, and comprehensible forms of visualization (Böttger, 2020, p.-84). Pictures, as a particular form of visualization, can be a valuable form of non-linguistic input that, when planned for effectively, can be used to develop not only receptive language competences but productive ones as well (Wright, 1989). Beyond, for example, the tradi‐ tional use of picture cards, body language such as gestures and facial expressions can be used to support comprehension in the primary school EFL classroom. Total physical response (TPR) (Asher, 1981) is a specific language learning methodology designed to teach language through actions. In the primary school, it is often used to develop listening comprehension skills in beginning learners (Haß et al., 2006; Legutke et al., 2014). Additionally, digital media such as interactive whiteboards and personal devices such as tablets increase the ease with which teachers can visualize materials and language for their students (Enever, 2016). 3.1.6 Prioritizing listening and speaking Listening comprehension and speaking are prioritized in the primary school EFL classroom. An emphasis in the primary school EFL classroom on developing listening and speaking 46 3 Early Foreign Language Learning at Primary Level <?page no="47"?> competences is based on an understanding of mother tongue development; children learn to understand and use spoken language before they learn to read and write it (Pinter, 2017). Young children at primary school do not yet have well-developed written language competences. An understanding of written language, therefore, cannot be relied on to teach young EFL learners. None of this is to suggest, however, that written language cannot or should not play a role in young learners’ foreign language development. The English core curriculum for primary school in NRW describes roles for oral and written communicative competences in English lessons (MSB NRW, 2021). To start teaching simple grapheme-lexeme relationships, Pinter (2017) suggests labeling classroom objects, making posters of commonly used classroom phrases, or playing popular games such as memory where children match pictures with written words. Beyond teaching basic word recognition, teaching students to read picture books has also been shown to be possible in the EFL classroom (see Frisch, 2013; Reckermann, 2018). Nonetheless, the primary school classroom is characterized by an abundance of oral language activities. At the end of fourth grade in NRW, students should be able to follow, with the help of visual support, EFL classroom discourse and understand essential information in listening texts such as short films (MSB NRW, 2021). In order to begin developing these listening competences, many beginning listening activities follow a “listen and do” (Pinter, 2017, p. 59) structure. TPR (described above), playing “Simon says,” or drawing a picture based on spoken instructions are all popular examples of listening comprehension activities in primary school. Classroom discourse also often follows a listen and do pattern. Teachers may ask children to sit down, take out their pencils, or come up to the board. All of these instructions are easily demonstrated by the teacher. As learners advance in their receptive language skills, it is important to use listening activities with concrete referents that are relevant to children’s lives and interests. Speaking competences in NRW are formulated in relation to two main goals: 1) partic‐ ipating in dialogues and discussions and 2) formulating and uttering complete sentences (MSB NRW, 2021). Activities designed to support the development of these competences should focus on personally relevant contexts and use patterned language and “chunks” (Pinter, 2017). Children may start by learning simple chunks to be used in the classroom or participate in short dialogues. These structures allow children to develop confidence in speaking and act to scaffold early speaking attempts. 3.1.7 Differentiation and individualization Primary schools in Germany, unlike secondary schools, are not externally-differentiated. As such, the student body at primary schools is even more heterogeneous than the students of German secondary schools. Differentiation at primary schools in Germany therefore takes the form of internal differentiation and is considered one of the foundational pedagogical principles of the primary school EFL classroom (Böttger, 2020). Despite the heterogeneity of primary schools, bilingual programs generally (primary and secondary schools) often employ a selection process (Rumlich, 2016) and may, therefore, be more homogeneous than typical classes. Regardless of any real or perceived homogeneity in bilingual classes, 3.1 Principles of foreign language instruction at primary school 47 <?page no="48"?> differentiation and individualization remain important principles of language education in primary school. Germany has been working to build a more inclusive school system since at least 2007 when it signed the Convention of the United Nations on the Rights of Persons with Disabilities (BB BMB, 2018). This has led to much discussion about the role of differentiation for students with disabilities. Differentiation and individualization in the context of language learning, however, extends to all students since language learning is a very individual process (Böttger, 2020; Thaler, 2012). Principles and concepts already reviewed such as play-based learning, multisensory learning, and mediation can go a long way to differentiating for individual language learning needs amongst learners. Some discussions of individual differences in the field of foreign language learning focus on different learning styles such as visual vs. auditory learners or inductive vs. deductive learners (see Brown, 2014, p. 113). Others go far beyond this to include not only language learning aptitude but differences such as anxiety and motivation toward language learning as well as different personalities and willingness to communicate (MacIntyre, Gregersen, & Clément, 2016). MacIntyre et al. (2016, p. 318-319) remind us of two important characteristics of these differences: 1) that these differences are not static but fluctuate and 2) that they are interdependent. This requires teachers to know their students well and to constantly evaluate learners’ progress. There are a multitude of ways in which teachers can differentiate within the classroom. Böttger (2020) sees differentiation taking place with regard to goals. While all students need to achieve some minimum standard as set by the curriculum, the level of competence to be achieved can be differentiated (Böttger, 2020, p.-88). Thaler, (2012) warns that consistently setting lower standards for particular students may fail to challenge these students and lacks a balanced approach to differentiating. Differentiation also needs to be planned with a longer view in mind, for example, through choices about group constellations, as well as in the immediate classroom interaction between teachers and students (Böttger, 2020, p. 88). MacIntyre et al. (2016, p. 318) suggest that teachers start by focusing on one area of individual difference and explore its interdependence with other factors in the classroom and with students. Thaler (2012, p. 130) outlines 14 ways teachers can differentiate in the classroom including, for example, different goals for different students, various social forms such as working alone or with a partner, or giving students choices regarding media and methods. To differentiate speaking tasks, for example, Verriere (2016) offers ideas such as adjusting topics based on the interests of students, increasing difficulty by choosing topics further from students' everyday lives, or adjusting the type of input students receive, for example, written or visual. While differentiation can be challenging for schools and teachers, some have embraced young children’s natural heterogeneity as a strength. The Bielefeld Laborschule teaches multi-grade level classes (grades 3-5) and in doing so has forced more differentiation which has led to more effective support for individual children (Thurn, 2011, p.-19). 48 3 Early Foreign Language Learning at Primary Level <?page no="49"?> 3.1.8 Theme-oriented teaching A thematic approach to language teaching allows for a variety of meaningful language learning activities in the classroom (Cameron, 2001). While theme-oriented teaching generally takes a holistic approach to language learning, it also allows for a form-focused approach when appropriate (Widdowson, 1998 as cited in Cameron, 2001, p.-184). A thematic approach to language education for young children serves two main pur‐ poses. First, it supports child-centered and developmentally appropriate teaching. Instead of organizing the curriculum around, for example, features of language, a theme-oriented curriculum is organized around topics relevant to children’s lives. Appropriate themes for the primary EFL class are ones that are concrete, meaningful, and interesting (Böttger, 2020). Breitung and Kirsch (1997 as cited in Böttger, 2020) describe eight main areas of interest for children that can be used to organize a theme-oriented curriculum. Examples include themes that allow children to experience the role of an explorer and inventor as well as themes allowing children to learn about their bodies and explore the world around them (Breitung & Kirsch, 1997 as cited in Böttger, 2020). Features of language such as subject vocabulary and grammar that are embedded in chunks can be taught and practiced within the context of the theme. Second, theme-oriented teaching supports authenticity in the classroom. As already discussed, one understanding of authenticity in language learning is learning authenticated by the learners (van Lier, 1996). At a minimum, this would require that learners find the topic interesting and meaningful (Böttger, 2020). Theme-oriented learning can also achieve authenticity through cross-curricular connections as well as connections to local or international festivals (Camerson, 2001). Learning that is connected broadly to children’s lives and interests has a greater possibility of being authentic to them. In summary, there are a number of principles intended to guide the teaching of English as a foreign language in primary school. These principles are based on understandings of how children learn as well as how languages are learned. As such, they are relevant to primary school CLIL pedagogies intended to support, in part, the development of foreign language competences in young children. 3.2 CLIL at primary school in Germany 3.2.1 Age and foreign language learning CLIL, whether at primary school or secondary school, is motivated by factors already described such as globalization and European integration. CLIL is seen as helping to prepare students for the linguistic and intercultural demands of an interconnected world. In addition to these motivations, CLIL at primary school (and early foreign language learning in general) is often motivated by the early start to foreign language learning which it provides thus preparing students to achieve higher levels of foreign language competence in the long term (Wode, 2009). The bilingual education concept for the Dortmund International Primary Schools makes this motivation explicit: Um mindestens eine Fremdsprache auf einem sehr hohen, berufstauglichen Level zu erwerben, ist frühe Immersion ein geeignetes Mittel, dieses Ziel zu erreichen. Auch beim Erwerb einer Zweitsprache 3.2 CLIL at primary school in Germany 49 <?page no="50"?> gilt die Devise je früher, desto besser. Aus diesem Grund arbeiten die DIPS Schulen sowohl mit den Kindertagesstätten als auch den weiterführenden Schulen, die ein ähnliches Rahmenprogramm anbieten zusammen. (Early immersion is a suitable means of achieving at least one foreign language at a very high, professional level. The motto “the earlier, the better” also applies to the acquisition of a second language. For this reason, DIPS schools work together with both kindergartens and secondary schools that offer a similar program.) (DIPS, 2020, Sept. 7) When considering age-related factors in foreign language learning, a popular yet contro‐ versial idea is the critical period hypothesis (CPH). The CPH suggests that there is a critical period during which an individual is “more sensitive to environmental stimulation […] than at other times during [his or her] life” (Colombo, 1982, p. 261 as cited in Lambelt & Berthele, 2015, p. 17). Applied to foreign language learning, the CPH posits that “complete mastery of an L2 is no longer possible if learning begins after the end of the putative critical period” (Piske, 2017, p. 45). In their book-length review of age-related factors in foreign language learning, Lambelet and Berthele (2015) conclude that while there is evidence that in the long term, learners who start early have some advantage over learners who start late, curricular factors, e.g. contact hours or instructional approaches and learner factors, e.g. motivation, learning difficulties, or individual bilingualism also play a significant role in students’ language learning. Piske (2017, p. 51), after reviewing research focused on age factors in foreign language learning, comes to a similar conclusion: an early start to foreign language learning “does not automatically lead to greater success.” Instead, based on his research review, Piske describes five factors unrelated to age that do have an effect on L2 learning. They are: 1) intensive exposure to the L2 over a long period of time, i.e. years, 2) frequent opportunities to speak and write the L2 in a variety of contexts, 3) a lot of “authentic, enriched and […] almost native-like input,” 4) targeted instruction for different linguistic skills, and 5) high levels of motivation. He also states that teaching competences and students’ language biographies play a role (Piske, 2017, pp. 51-52). While primary school EFL classes are likely to only be in a position to meet motivational factors (Piske, 2017, p. 52) primary school CLIL programs could address more of the factors. At a minimum, they are likely to provide more native-like input over a longer period of time as well as harnessing young children’s natural motivation (see Piske, 2017, p. 52). Because of the content focus of CLIL lessons and the less developed language competences of young learners, student output tends to be fairly minimal in primary school CLIL settings (see Schwab, 2020). As for targeted language instruction, again, a content focus in CLIL lessons in Germany makes it unlikely that this learning factor will be significantly addressed in the CLIL context. 3.2.2 Organization Primary school CLIL in Germany can be organized in a variety of forms. In its more formal realizations, CLIL is organized in branches or streams in which a selected group of students are chosen for the CLIL class and receive instruction though a foreign language for a set number of non-language subject courses. These courses are timetabled as the non-language subject and are generally taught through the foreign language for the entire school year. Variation in the way that CLIL branches are organized can include the subjects that are taught, how many subjects are taught, and the possibility of team-teaching (generally with 50 3 Early Foreign Language Learning at Primary Level <?page no="51"?> one teacher responsible for the L1 and the other for the L2) or a single teacher teaching the CLIL courses. As an alternative to CLIL branches, CLIL modules are more flexible and less time-intensive realizations of CLIL. Within a module structure, teachers alone or in collaboration with colleagues, choose particular topics within particular subjects that they feel are particularly suitable to a CLIL approach. There is a high degree of variation in how CLIL modules are organized and taught. One difference between CLIL branches and CLIL modules is the amount of time spent in CLIL instruction. Under a CLIL branch format, there may be upwards of 8-10 lessons taught through the foreign language per week. Under a CLIL module format, there may be as little as one or two CLIL subjects taught per week. This is in addition to the two (sometimes three) weekly EFL lessons. In addition to the time difference between models, there is a difference in student population. Because official CLIL branches require an application process (at minimum expressing interest in being in the CLIL class), CLIL branches tend to attract high-achieving and motivated students (Küppers and Trautmann, 2013). CLIL modules are generally undertaken in classes without any official CLIL designation and therefore have a student population that has not been selected for in any CLIL-specific way. The issue of CLIL selection remains a topic of debate within the CLIL literature (see, for example, Bruton, 2013; Hüttner & Smit, 2014). Regardless of the CLIL model followed and inherent program variability, primary school CLIL in Germany has several generalizable characteristics. First, the non-language subject curriculum is the guiding document for CLIL instruction. Teachers are required to teach the content as it is outlined in the core curriculum (KMK, 2013). While improved L2 development is a primary goal and motivating factor for CLIL, it is not a part of the curriculum and therefore is not a required learning outcome. Second, CLIL teachers are often German native-speakers who have studied and passed the required university courses to become a teacher in Germany. As such, they are 1) non-native speakers (NNS) of the foreign language (often English) and 2) share a common L1 with most of the students. Sharing a language with students is likely advantageous when needing to explain content in children’s L1 (or the school language more generally) as well as when working to develop students’ L1 competences alongside their emerging L2 competences. Third, CLIL students usually have little exposure to the L2 outside of the school and classroom context. This is especially true if the target language is English, a language which has no official role in Germany and which is not spoken as an official language in any of the countries bordering Germany. An exception to this would be CLIL programs operating in border regions and which target languages spoken in the border regions (e.g. German-French CLIL programs along the German-French border). CLIL programs that target English lie in marked contrast to traditionally defined immersion contexts (e.g. French immersion in Quebec) since English is generally not spoken in the community outside of the school. While the programmatic and contextual features described above are likely to be found across CLIL programs in Germany, programs nonetheless vary in how they are organized. Common organizational variations include the amount of time allocated to CLIL 3.2 CLIL at primary school in Germany 51 <?page no="52"?> instruction, the subjects taught through CLIL, and teaching arrangements, e.g. a single teacher or a team-teaching arrangement. 3.2.3 Subjects Primary school CLIL in NRW is not guided by a single, comprehensive, government-pub‐ lished set of guidelines. Instead, CLIL programs in Germany and Europe have often been initiated, planned, and implemented at the local level, i.e. by municipalities, schools, and individual teachers. This has led to widely varying CLIL programs and classes across NRW. One of the ways in which CLIL programs vary is the subjects taught. The primary school curriculum (Kernlehrplan) in NRW describes competences for the eight subjects taught at primary school: German, Math, English, Science and Social Studies (Sachunterricht), Art, Music, Physical Education, and Religion (MSB NRW, 2021). In principle, all of these subjects, with the exception of German, can be taught through a foreign language. The most common CLIL courses in primary school, however, are the Science and Social Studies, Art, Music, and Physical Education. Heim (2015) points out that subjects chosen for CLIL should allow opportunities to explore intercultural learning as well as include topics that can be easily visualized. Additionally, opportunities to engage authentically (Viebrock, 2013; Aristov & Haudeck, 2013) and actively (Heim, 2013) with content and language are crucial. Beyond these, topics which offer opportunities for multisensory and play-based approaches are suitable to CLIL instruction. Art, Music, and Physical Education are well-suited to CLIL education because of the various reasons described above. For example, Art and Physical Education are easily visualized either with materials or through actions. These subjects offer opportunities to engage authentically (in the sense of authenticity described by Holliday, 2013 and van Lier, 1996) and actively and likely require less linguistic output from students thus supporting the development of oral receptive skills in young English learners. The subject of Science and Social Studies cover a wide range of curricular topics. The primary goal of science lessons at primary school is “to provide pupils with meaningful, memorable, science experiences upon which they will build their further understanding of natural phenomena” (Aristov & Haudeck, 2013, p. 45). This content-related goal lays the foundation for an “authentic and action-oriented approach” conducive to CLIL classes (Aristov & Haudeck, 2013, p. 45). For example, students can engage in the scientific process by asking questions, conducting experiments, and observing results. Through experiments, Aristov and Haudeck (2013, p. 45) argue that children become emotionally attached to the activities. This emotional attachment to learning has the potential to motivate children not only to learn the content but to be motivated to use and develop the language associated with the activity. Because the sciences at primary school in NRW include social sciences, there are opportunities to engage with culture, further bolstering the argument that the subject Natural and Social Science is suitable for CLIL instruction. Mathematics is rarely taught as a CLIL subject at primary schools in NRW. Viebrock (2013) cites three possible reasons for this: 1) the perception that mathematics does not offer insight into culture, 2) an understanding of mathematics as one that limits opportunities 52 3 Early Foreign Language Learning at Primary Level <?page no="53"?> for language use, and 3) the fear that mathematical content learning will be jeopardized by learning through a foreign language. Viebrock (2013, p. 52) goes on to explain that there is not empirical evidence for these claims and that the claims are likely based on misconceptions about what mathematics is. Contrary to the above claims, Viebrock (2013) sees the importance of language use in mathematics as a primary reason for the suitability of mathematics for CLIL. Additionally, the fact that mathematics at primary school describes “real-life contexts” (Viebrock, 2013, p. 55) would seem to offer opportunities for authentic learning. Finally, international mathematical notation and the visual and concrete nature of mathematics at primary school, e.g. the study of space and shape or mathematical modelling (MSB NRW, 2021), would seem to offer a form of scaffolding for the accompanying use of language. 3.2.4 Curricula: English and Science CLIL classes in NRW are timetabled as the non-language subject course and are required to follow the non-language subject curriculum. Because this thesis is focused on the cognitions of teachers who are teaching the sciences bilingually, I will briefly review the Science curriculum here. I will also highlight aspects of the English curriculum that are most relevant to a CLIL teaching context. In addition to development of central science competences, the goals of Science and Social Studies lessons in NRW are to promote curiosity and an attitude of questioning in order to help students orient themselves in their environment as well as to participate responsibly and take action to shape their environment (MSB NRW, 2021, p.-178). These broad goals lay the foundation for the six areas of competence and the further content focus areas within each area of competence (see Table 1 below). These competence and focus areas as well as the individual competences stated in the curriculum are then further concretized by individual schools into an internal school curriculum (schulinternes Lehrplan) resulting in a degree of freedom for individual schools to choose topics, materials, and methods that they find best suited for achieving the outcomes outlined in the curriculum. This inevitably leads to a variety of topics being taught making it difficult to establish exactly what children are studying in their CLIL classes. However, from NRW’s core curriculum, it is clear that students should be developing the competence to, for example, “describe,” “explain,” “investigate,” and “evaluate” (MSB NRW, 2021) all of which involve language use. The extent to whether or not CLIL students develop such competences in English by the end of grade four is largely dependent on students and teachers. 3.2 CLIL at primary school in Germany 53 <?page no="54"?> Kompetenzbereich (Area of Competence) inhaltliche Schwerpunkte (Content Focus Area) Demokratie und Gesellschaft (Democracy and Society) • Zusammenleben in der Klasse, in der Schule und in der Gesell‐ schaft (Living together in class, at school and in society) • Leben in der Medien- und Konsumgesellschaft (Living in a media and consumer society) • Leben in Vielfalt (Living with diversity) Körper und Gesundheit (Body and Health) • Körper und gesunde Lebensführung (Body and healthy life‐ style) • Körper und Entwicklung (Body and development) Natur und Umwelt (Nature and Environment) • Tiere, Pflanzen, Lebensräume (Animals, plants, habitats) • Stoffe, ihre Umwandlung und Stoffkreisläufe (Matter, its transformation and matter cycles) • Energie und Ressourcen (Energy and resources) Raum und Mobilität (Space and Mobility) • Orientierung in Räumen (Orientation in spaces) • Räume nutzen und schützen (Using and protecting spaces) • Mobilität im Raum (Mobility in space) Technik, digitale Technologie und Arbeit (Technology, Digital Technology and Work) • Bauen und Konstruieren (Building and construction) • Technische und digitale Entwicklungen (Technical and digital developments) • Arbeit und Beruf (Work and career) Zeit und Wandel (Time and Change) • Orientierung in der Zeit (Orientation in time) • Früher, heute und morgen (Past, present and future) • Fakten und Fiktion (Facts and fiction) Table 1: Areas of Competence and Content Focus Areas in the Primary School Science Curriculum in NRW (MSB NRW, 2021) The English curriculum in NRW does not play an explicit role in guiding CLIL instruction. However, with the development of L2 competences as a primary goal of CLIL, the English curriculum can play some role as a guide to the development of those competences in CLIL instruction. The primary school English curriculum in NRW emphasizes the development of func‐ tional communicative competences as well as intercultural competences (MSB NRW, 2021). Additionally, the curriculum highlights the importance of developing motivated language learners and early language learning strategies and techniques (MSB NRW, 2021, p. 36). In order to achieve these broad goals and the development of the more detailed competences outlined in the curriculum, holistic approaches that take into account the individuality of students, including their individual language biographies, are recommended. Activities and tasks should scaffold meaning, be playful, and view students’ multilingualism as a resource for learning (MSB NRW, 2021, pp. 37-38). The current English curriculum does not mention CLIL or bilingual education. However, the strategies and techniques described in it are generally relevant to a variety of language learning contexts including CLIL. Two principles described in the English curriculum have relevance beyond the language classroom and are particularly relevant to CLIL instruction. First, primary school teachers of all subjects need to support the development of the school 54 3 Early Foreign Language Learning at Primary Level <?page no="55"?> language and, in EFL lessons, the target language as well as take into consideration the role of other languages spoken by students. Da in allen Fächern der Primarstufe fachliches und sprachliches Lernen eng miteinander verknüpft sind, ist es die gemeinsame Aufgabe und Verantwortung aller Fächer, die bildungssprachlichen Kompetenzen aller Schülerinnen und Schüler als wichtige Voraussetzung zum Lernen und für den Schulerfolg zu entwickeln und zu stärken. Mehrsprachigkeit wird dabei als Ressource für die sprachliche Bildung verstanden. (Because subject and language learning are closely linked in all subjects at primary level, it is the joint task and responsibility of all subjects to develop and strengthen students’ academic language skills which are an important prerequisite for learning and success at school. Multilingualism is seen as a resource for language education.) (MSB NRW, 2021, p.-37) Considering that a primary goal of CLIL is subject literacy in two languages (Diehr, 2012; KMK, 2013), attention to how languages and content relate to one another and can be taught simultaneously is central to planning and implementing CLIL instruction. Second, the curriculum highlights the importance in primary school of making connec‐ tions between students’ interests and inclinations and the range of requirements across subjects. Es ist Aufgabe der Primarstufe, die Fähigkeiten, Interessen und Neigungen aller Schülerinnen und Schüler aufzugreifen und sie mit den Anforderungen fachlichen und fächerübergreifenden Lernens zu verbinden. (It is the task of primary school to tap into the abilities, interests and inclinations of all pupils and integrate them with the requirements of subject-specific and interdisciplinary learning.) (MSB NRW, 2021, p.-37) Such student-centered and transdisciplinary learning is especially relevant to CLIL instruc‐ tion in which teachers must consider a variety of individual student differences including language competences and transdisciplinary strengths and interests in order to achieve content and language learning. The various areas of competence described in the English curriculum are outlined in Table 2 below. Kompetenzbereich (Area of Competence) Schwerpunkte/ Teilkompetenzen (Focus Areas/ Sub-Competences) Funktionale kommunikative Kompetenz (Functional Communicative Competence) -- • Hörverstehen/ Hör-Sehverstehen (Listening/ Listening-visual comprehension) • Leseverstehen (Reading comprehension) • Sprechen (Speaking) • an Gesprächen teilnehmen (Participate in conversations) • zusammenhängendes Sprechen (Connected speaking) • Schreiben (Writing) • Sprachmittlung (Language mediation) Verfügbarkeit von sprachlichen Mitteln (Availability of Linguistic Re‐ sources) - • Wortschatz (Vocabulary) • Grammatik (Grammar) • Aussprache und Intonation (Pronunciation and intonation) • Orthografie (Spelling) 3.2 CLIL at primary school in Germany 55 <?page no="56"?> Interkulturelle kommunikative Kompetenz (Intercultural Com‐ municative Competence) - • soziokulturelles Orientierungswissen (Sociocultural knowl‐ edge) • interkulturelle Einstellungen und Bewusstheit (Intercultural attitudes and awareness) • interkulturelles Verstehen und Handeln (Intercultural under‐ standing and action) Text- und Medienkompetenz (Text and Media Competence) -- Sprachlernkompetenz (Language Learning Competence) -- Sprachbewusstheit (Language Awareness) -- Table 2: Areas of Competence and Focus Areas in the Primary School English Curriculum in NRW (MSB NRW, 2021) 3.2.5 Teacher requirements and competences CLIL teachers at primary school are required to have studied the target language (often English) as well as the non-language subject at university (MSB NRW, 2022, January 29). Among those teachers qualified to teach EFL lessons at primary school, those with only a language teaching certification in English would not be allowed to teach CLIL. CLIL teachers are not required, however, to have any CLIL-specific pedagogical training. Massler (2012) found that out of the 12 primary school CLIL teachers who participated in the Pro-CLIL project, only two of them had had previous CLIL training. This finding suggests that many primary school CLIL teachers, trained in content subjects and EFL methodologies, are forced to develop CLIL pedagogical competences on their own through CLIL teaching experience. CLIL pedagogical competences are multiple and varied. A review of theoretical and empirical literature on CLIL teaching suggests that CLIL teacher competences can be organized into three broad categories: 1) general teaching competences, 2) target language competences, and 3) material creation competences. Wolff (2020, Sept. 7), in his summary of CLIL teacher competences outlines all three areas. Not only do primary school CLIL teachers need to be competent in all areas of primary school teaching such as subject knowledge, classroom management, task planning, and facilitating classroom interaction, they need to be able to do all of these things through the target language. Furthermore, CLIL teachers need to be able to create foreign language materials and evaluate, choose, and revise authentic materials. The European Centre for Modern Languages published a more comprehensive framework of competences for CLIL teachers (Marsh, Mehisto, Wolff, & Frigols Martin, 2011). Competence areas include, for example, personal reflection and classroom management (general teaching competences), content and language awareness (target language competences), as well as learning resources and environments (material creation competences) (Marsh et al., 2011). The framework goes on to describe areas such as knowledge of CLIL fundamentals, methodology and assessment, and research and evaluation (Marsh et al., 2011). 56 3 Early Foreign Language Learning at Primary Level <?page no="57"?> Empirical investigations into teachers and CLIL contexts suggest that teachers are most concerned with their own L2 competences as well as the ability to create materials. Deters-Philipp (2018) investigated English language competences of primary school EFL teachers in Germany. She found that teachers’ personal language biographies had a much stronger influence on how teachers were able to fulfill the linguistic and pedagogical de‐ mands in the EFL classroom than were factors such as age, formal qualifications, and career experience. She concludes by proposing four essential characteristics of primary school EFL teachers: 1) a variety of positive experiences with English in various educational settings, 2) active contact with English outside of lessons, 3) a positive attitude toward early foreign language learning, and 4) fluent and flexible English language skills (Deters-Philipp, 2018, p. 431). These conclusions point to the highly personal and affective nature of language teaching and suggest that teaching CLIL demands personal and emotional investment. Massler (2012) investigated teachers’ thinking about CLIL. Over the course of her ProCLIL project, two out of the 12 participating teachers dropped out of the project due to a perceived lack of language skills and others reported that their lack of language competences was a significant contributor to the high workload. Furthermore, Massler (2012) reported that all of the participating teachers reported spending significantly more time creating materials despite being provided published and team-created materials. Bechler (2014), in her study of primary school CLIL modules, also found that teachers reported spending more time rethinking lesson plans, looking up subject vocabulary, and creating materials. The finding that CLIL teachers need to invest more time in lesson preparation has been reported elsewhere as well (see, for example, Dallinger, 2015; Burmeister, Ewig, Frey, & Rimmele, 2013). Based on these three studies, CLIL teachers need to make significant personal, emotional, and time commitments to successfully carry out their work. 3.2.6 Classroom interaction CLIL classroom interaction, at least in secondary school contexts, has been shown to often be teacher-centered (Viebrock, 2007; Dalton-Puffer, 2007). Furthermore, there is some evidence that explicit attention to language form in CLIL instruction is rare (de Graaff et al., 2007, p. 615) even though, as Dalton-Puffer and Smit (2013, p. 555) report, some science educators have argued that focused attention on language has benefits for content learning. There is also evidence that the primary school EFL classroom is often characterized by teacher-centeredness. Limberg (2020) analyzed data from the Primary English Classroom Corpus (PECC). The PECC is a collection of transcripts from primary school EFL lessons in Germany. Limberg reports that “every transcript in the PECC contains longer stretches of teacher-led talk at different stages of a lesson in which new language is introduced, recycled, or learners are instructed for an activity” (Limberg, 2020, p. 39). Limberg explains that this is because primary school learners have low English competences and therefore need large amounts of exposure to the language. Evidence from the primary school CLIL classrooms, however, tells a slightly different story. Based on CLIL classroom observations, Schwab (2020) reports long stretches of student talk in the CLIL classroom: Obviously, teachers did not insist on using English as the medium of communication. Interactions were usually not interrupted when students provided utterances in their L1 and thus, we see long 3.2 CLIL at primary school in Germany 57 <?page no="58"?> stretches of talk-in-interaction where many students are involved even if the teacher remains in the focal position. (Schwab, 2020, p.-107) These findings seem to indicate a clear prioritization of content learning goals. Addition‐ ally, these longer, uninterrupted “talk-in-interaction” sequences suggest an embedded instructional style. Unlike the EFL classroom, where “classroom interaction is largely controlled by the teacher, who models the foreign language for the learners, selects pupils to respond mostly to display questions, initiates choral repetitions of either single words or phrasal expressions and recasts or fine-tunes pupils’ contributions” (Limberg, 2020, p. 51), teachers in Schwab’s (2020) study of CLIL instruction depended greatly on feedback within interaction and worked to avoid interruption and maintain conversational flow. Based on Schwab’s (2020) findings, opportunities for language learning in CLIL contexts would seem 1) primarily in the realm of receptive competences and 2) largely unplanned and integrated naturally into the content-focused interaction. 3.2.7 L1 and L2 use Teachers make use of the L1 in their CLIL instruction. Frisch (2016) identifies five functions of teachers switching to the L1 in CLIL. First, the L1 is used for cognitive reasons, e.g. to explain complex subject-related concepts. Second, the L1 is used for communicative reasons, e.g. to provide L1 vocabulary equivalents. Third, the L1 is used to save time, e.g. to explain homework. Fourth, the L1 is used for affective reasons, e.g. comforting children. Fifth, the L1 is used for pedagogical reasons, e.g. to discipline students (Frisch, 2016, p. 93). With the exception, perhaps, of providing L1 equivalents for subject terminology, all of these reasons function as a kind of scaffold for linguistic or cognitive demands. None of these uses of L1, again, with the exception of providing L1 equivalents, aim to use the L1 to actively teach L1 subject discourse competences. Glaser (2020) investigated how teachers give instructions in a primary school EFL context. She found that teachers sometimes use the L1 to give instructions but concluded that switching to the L1 for such a purpose was ineffective. Citing Limberg (2017) and Solmecke (1998) she states that having students repeat the instructions in the L1 is also not effective. Using the L1 for such purposes may be motivated by a teacher’s perception of task complexity or by the desire to save time. Regardless of exact reasons for L1 use when giving instructions, Glaser (2020, p. 80) claims that “if planned and sequenced carefully, instruction-giving can proceed successfully without the L1 even in the young learner classroom.” While Glaser admits that such planning takes “time, energy, and motivation” (Glaser, 2020, p. 80), it results in improved teacher control not only over the instructions but over task management generally. In a CLIL context where there is a goal of developing competences in two languages, there may be times when it is appropriate to give instructions in the L1. Students’ use of the L1 and L2 also varies in CLIL. Schwab (2020) found evidence that student output in a primary school CLIL context is primarily in the L1. Bechler (2014) investigated bilingual science modules at primary school. She found that when the lesson focus was on having students speak, there was a switch to German (the L1). These results are not surprising considering the generally low L2 competences of primary school children. 58 3 Early Foreign Language Learning at Primary Level <?page no="59"?> Regardless, student-produced L2 output is vital for L2 development (Swain, 1995) and a key trait of good foreign language instruction (Helmke, Helmke, Schrader, Wagner, Nold, & Schröder, 2008, p.-361). 3.2.8 Motivation Increased motivation toward foreign language learning is one reason often cited for using CLIL models of instruction (see, for example, Thompson & Sylvén, 2019). Generally speaking, it seems primary school students are more motivated in the EFL classroom than secondary school students (Piske, 2017, p. 52). Akkerman (2012) investigated learning in CLIL science modules as well as with forms of non-CLIL learning. Based on student responses to a questionnaire, she found that students in CLIL contexts were most motivated by working in groups or with a partner but that students in non-CLIL contexts were also motivated suggesting that motivation was more a function of activity or social forms than of CLIL instruction per se. Bechler’s (2014) investigation of CLIL science modules found that students often lacked motivation and focus during CLIL instruction. Bechler reports that students did not take the CLIL instruction seriously and instead chose to disengage. She suggests that this disengagement is the result of being overwhelmed by the foreign language (Bechler, 2014, p. 238). Taken together, these two studies suggest that motivation in primary school CLIL depends on factors other than teaching through a foreign language and that without careful consideration of L2 use in CLIL, the use of a foreign language can demotivate young students. 3.3 Summary In this chapter, I have described general theories of learning relevant to primary school aged children and key pedagogical principles of foreign language teaching in primary school. Furthermore, I have described programmatic features of CLIL in NRW primary schools and have reviewed empirical evidence describing the current state-of-the-art of EFL and CLIL at German primary schools. Principles key to EFL instruction such as multisensory learning, play-based learning, and visualization are relevant not only to primary school EFL instruction but to primary school instruction generally. Principles such as authenticity are more relevant to language learning contexts such as EFL and CLIL. The principle of using English as the language of instruction, while relevant for English language CLIL contexts, needs to be considered carefully in CLIL contexts which aim to develop subject competences in two languages. Primary school CLIL is motivated in part by the potential long term benefits of getting an early start in foreign language learning as well as by factors such as European integration and globalization that are also relevant to CLIL programs at secondary school. Primary school CLIL can be organized in a variety of ways, including in branches or modules, and can, in principle, be used as a pedagogical approach in all subjects except German (or L1 language class generally). In NRW, there are no CLIL specific guiding documents for primary school. Teachers follow the non-language subject curriculum. Trained in EFL, 3.3 Summary 59 <?page no="60"?> teachers likely look to the EFL curriculum for some general guidance to the development of L2 competences in their CLIL instruction. For primary school CLIL teachers, the challenge of CLIL lies in the L2 language competences necessary to teach CLIL as well as the additional time and effort required to create and revise CLIL materials. Assuming that Deters-Philipp’s (2018) four proposed characteristics for primary school EFL teachers are relevant to CLIL teachers, CLIL teachers need a wealth of positive experiences with language learning and an emotional commitment to CLIL in order to endure the sustained challenges of CLIL. Based on the evidence reviewed above, primary school CLIL instruction seems to be characterized by high amounts of teacher-produced L2 input and high amounts of student-produced L1 output. Teachers do, however, switch to the L1 for a variety of reasons outlined by Frisch (2016). While primary school aged students generally exhibit more motivation toward language learning (Piske, 2017), there is evidence that without careful consideration of primary school students’ L2 competences, teachers can overwhelm students with the L2 and demotivate them (Bechler, 2014). 60 3 Early Foreign Language Learning at Primary Level <?page no="61"?> 4 Teacher Cognition Teachers’ professional lives consist of their inner thoughts and beliefs as well as their outward words and actions. The term “teacher cognition,” what teachers “think, know and believe” (Borg, 2003, p. 81), is an “umbrella term for the unseen dimensions of teachers’ work” (Borg, 2019, p. 4), while teachers’ behaviors, i.e. what they say and do, are the observable aspects of teachers’ work. The distinction between the seen and unseen, the physical and the mental, does not, however, indicate isolated realms of action. The entire premise for the study of teacher cognition rests on the assumption that thoughts influence behavior and behavior influences thoughts and that by studying the cognitions of teachers or, simply, what they think, we can hope to understand and influence teachers’ behavior (Borg, 2003; Calderhead, 1987a; Clark & Yinger, 1977). Two foundational terms or constructs are used in the study of teachers’ cognitions. “Cognition,” as defined by the American Psychological Association, is “all forms of knowing and awareness, such as perceiving, conceiving, remembering, reasoning, judging, imagining, and problem solving” (APA, 2022a, June 7). A second term, “thinking,” is defined as: cognitive behavior in which ideas, images, mental representations, or other hypothetical elements of thought are experienced or manipulated. […] Thinking may be said to have two defining characteristics: (a) It is covert—that is, it is not directly observable but must be inferred from actions or self-reports; and (b) it is symbolic—that is, it seems to involve operations on mental symbols or representations, the nature of which remains obscure and controversial. (APA, 2022b, June 7) The nature of thinking, namely, that it is covert and symbolic, have ramifications for the study of teachers’ cognitions. These ramifications will be discussed in section 5.2. Borg (2019, p. 20) defines the study of teacher cognition: “Inquiry which seeks, with reference to their personal, professional, social, cultural and historical contexts, to under‐ stand teachers’ minds and emotions and the role these play in the process of becoming, being and developing as a teacher.” While “teacher cognition” is the most commonly used term for this field of study in the English-speaking world, the terms “teacher psychology” (Mercer & Kostoulas, 2018) or “teacher philosophies” (Crookes, 2015) are preferred by some. Mercer and Kostoulas (2018) do not explicitly define “teacher psychology ” but use the term in a superordinate way to teacher cognition suggesting that the study of “teacher cognition” and “teacher identities” are subordinate fields of study under teacher psychology. Crookes (2015, p. 490) is also skeptical of the “subsuming expansion” of teacher cognition to include constructs such as attitudes, identities, and emotions. He instead argues that “teacher cognitions” along with other constructs such as “sets of beliefs” and “bodies of knowledge” be subsumed under the heading “philosophy of teaching” (Crookes, 2015, p. 485). Because the goals of the present research are compatible with Borg’s definition of teacher cognition, namely what teachers “think, know and believe” (Borg, 2003, p. 81), the term teacher cognition will be used for this thesis. In the following sections, I will briefly review the history of language teacher cognition research and describe its goals. Then I will describe selected theoretical and psychological <?page no="62"?> constructs most relevant to the present study. Finally I will review research on teacher cognition most relevant to the field of primary school CLIL. 4.1 History of the study of teacher cognition Current understandings of teachers, their lives, and their thinking have philosophical (Ryle, 2009, originally published in 1949; Fenstermacher, 1978), sociological ( Jackson, 1968; Lortie, 1975), and psychological (Rokeach, 1968; Nisbett & Ross, 1980) foundations. For example, Ryle’s (2009) investigation into “knowing how” and “knowing that” has influenced the study of expertise in teaching (see Tsui, 2003) while Lortie’s (1975) description of the “apprenticeship of observation” continues to influence teacher education programs (see Borg, 2015). An explicit recognition that teachers’ thinking plays a role in their classroom actions and that their thinking should, therefore, be the object of scientific investigation dates back to the 1970s and 1980s. Initial research, at least in the Anglo-American context, into what is now commonly referred to as teacher cognition focused primarily on teachers’ decision-making processes and teachers’ knowledge (Borg, 2015). Among the various forms of knowledge investigated include “practical knowledge” (Elbaz, 1981), “personal practical knowledge” (Clandinin & Connelly, 1987), and “pedagogical content knowledge” (Shulman, 1986). In Germany, the study of teacher thinking has its origins in the 1980s with the formulation of, and investigation into, “subjektive Theorien” (Groeben et al. 1988; Koch-Priewe, 1986). The “Forschungsprogramm subjektive Theorien” (Groeben et al., 1988) continues to be relevant across the German educational landscape (see, for example, Kindermann, 2020; Wagner, 2016) including in CLIL contexts (see Viebrock, 2007). It wasn’t until the 1990s with publications such as Woods’s (1996) Teacher Cognition in Language Teaching that the field of language teacher cognition began to establish itself (Borg, 2019). Since the 1990s, the field has grown to include investigations into a variety of constructs subsumed under the heading teacher cognition. Borg (2015) lists a number of them: “beliefs” (Bastukmen et al., 2004), “epistemological beliefs” (Flores, 2001), “theoretical beliefs” ( Johnson, 1992b), “theories for practice” (Burns, 1996), “pedagogical principles” (Breen et al., 2001), “practical knowledge” (Meijer et al., 1999), “personal practical knowledge” (Golombek, 1998), and more. From this list, Borg (2015, p. 54) concludes that the terms indicate that teachers’ cognitions are personal, based on experience, and both inform and are informed by classroom practice. Recent research into teacher cognition has focused on teachers’ identities (Ruohotie-Lyhty, 2016; Li & De Costa, 2018; Li, 2020), teacher motivation (Sahakyan, Lamb, & Chambers, 2018; Barcelos, 2016), teacher resilience (Kostoulas & Lämmerer, 2018), teacher beliefs (Kalaja, 2016; Gkonou & Mercer, 2018) and teacher emotions (Oxford, 2020; King & Ng, 2018). Research perspectives on the study of teachers’ cognitions have steadily expanded from more individualist conceptualizations of cognition to conceptualizations that take into account social, historical, and contextual aspects of teachers’ cognitions (Burns, Freeman, & Edwards, 2015). An early study in the field of general education and an example of an individualist approach based in cognitive psychology was conducted by Yinger (1980) who studied teachers’ mental processes while planning. His findings are, in part, presented 62 4 Teacher Cognition <?page no="63"?> with detailed diagrams in the manner of a flow chart which describe teachers’ mental processes as a series of inputs and outputs and in doing so present teachers’ cognitions as computer-like. Such an individualist approach to teachers’ thinking is representative of other early research on teachers’ thinking which conceptualized teachers’ cognitive work as “information processing” that was linear and rational (see Borg, 2015, pp. 8-10). A more recent study that reflects a contextualized and situated understanding of teachers’ thinking, one that takes social and historical factors into account, is from Ruo‐ hotie-Lyhty (2016). She investigated the development of teachers’ professional identities over time. Through the interviews conducted over approximately 10 years, Rouhotie-Lyhty was able to elicit social, historical, and contextual factors such as becoming a parent or relationships to students that have shaped the teachers’ identities. The expansion of research perspectives in teacher cognition offers opportunities and challenges. An expansion of theoretical and methodological approaches has the potential to broaden our understanding of teachers’ cognitions. No longer are researchers and teachers only interested in the mental machinations of the teacher’s brain. They are also increasingly interested in how complex, agentic human beings (i.e. teachers) relate to, make sense of, and act in the foreign language classroom. Kubanyiova and Feryok (2015, p. 436) call this more holistic approach to the study of teachers’ cognition “emergent sense making in action.” The authors distinguish this approach from earlier “cognitivist” (Kubanyiova & Feryok, 2015, p. 436) approaches which sought to isolate psychological constructs such as beliefs and knowledge. However, in an effort to move beyond the “individualist” (Burns et al., 2015) and “cognitivist” (Kubanyiova & Feryok, 2015) beginnings of teacher cognition studies, Borg (2019) sees some evidence that theoretical and methodological approaches that target more contextualized understandings of teachers’ cognitions are viewed as “intrinsically superior” (Borg, 2019, p. 7) to earlier approaches. He warns against this. Instead, Borg (2019) argues that the goals for the field of language teacher cognition moving forward should be to “promote complementarity (by building on existing work), inclusivity (by promoting methodological pluralism), and accessibility (by communicating transparently)” (Borg, 2019, p. 20). An “overly complex” (Borg, 2019, p. 11) and “overly intellectualized” (Borg, 2019, p. 20) field of study will be counterproductive to the pursuit of complementarity, inclusivity, and accessibility. 4.2 Goals of teacher cognition studies Teacher cognition studies have, from the beginning, been motivated by the idea that “what teachers do is affected by what they think” (Clark & Yinger, 1977, p. 279). This foundational premise of teacher cognition studies views teachers as “active, thinking decision-makers” (Borg, 2003, p. 81) and not merely as curriculum implementation widgets. Teachers, therefore, have agency and base their decisions and actions on “complex, practically-oriented, personalized, and context-sensitive networks of knowledge, thought, and beliefs” (Borg, 2003, p.-81). Teacher cognition research seeks not only to “describe” what teachers think but to “understand” how their thinking influences what they do (Borg, 2019, p. 2). Understanding the link between cognition and practice would seem to be the key to improving practice; 4.2 Goals of teacher cognition studies 63 <?page no="64"?> with the “right” cognitions teachers will employ the “right” practices. The link between cognition and practice, however, is neither linear nor unidirectional (Borg, 2015). Further‐ more, teachers’ cognitions are not necessarily stable constructs but are influenced by context and experience (Li, 2020). Any conception of teacher cognitions, teacher practices, and student learning that would suggest a linear and unidirectional relationship would be naive. Instead, teachers and what they think, know, and believe are part of a complex and interconnected social, physical, and pedagogical context that includes, for example, students, other teachers, materials, and learning goals. Understanding teachers’ cognitions can lead to a better understanding of teaching and learning contexts. This is especially important in a relatively under researched and highly variable context such as primary school CLIL. Teacher learning is one use of teacher cognition research. Traditionally, the hope of identifying teachers’ cognitions was to bring about the “successful alteration of the teachers’ everyday actions” (Mandl & Huber, 1983, p. 98). More recently, Kubanyiova & Feryok (2015, p. 437) have described “meaningful teacher development” that can result from understanding teachers’ cognitions. Writing more than 30 years ago, Calderhead understood this: An important role of [teacher cognition] research is to provide more realistic models of teaching that help us conceptualize the nature of this practice more clearly, enabling supportive efforts, including training and policy-making, to be more constructive. Furthermore, teachers themselves need a language and concepts that realistically represent their classroom practice, and that acknowledge the complexity of their classroom work. (Calderhead, 1987b, p.-4) An understanding of teachers’ cognitions, then, can influence models of teaching and teacher training. Furthermore, by naming and describing the various concepts that repre‐ sent their professional lives, teachers can better reflect on and take action to improve their classroom instruction. Collectively, teachers’ cognitions may be used to inform programmatic decisions, for example, on how best to structure initial teacher training programs (see Rossa, 2017). Un‐ derstanding the idiosyncratic cognitions of teachers can also be used. For example, Li (2020) suggests that the value of understanding teachers’ cognition is to “empower” individual teachers as they participate in “collaborative dialogic reflective practice” to develop their own practices in their immediate teaching contexts (Li, 2020, p. 277). Furthermore, teacher education programs that attempt to accurately contextualize instruction are likely to be more successful than programs that teach decontextualized knowledge and skills (Li, 2020, p.-284). Developing pedagogy is another use of teacher cognition research. Pedagogical devel‐ opments based on a better understanding of what teachers actually think in a particular context are more likely to be successful. Of course, teacher cognition studies, in informing pedagogy, are not meant to drive pedagogical developments. Design research such as that conducted by Botz and Diehr (2016) is better suited to identifying concrete approaches and techniques that have the possibility of leading to desired results. Teacher cognition 64 4 Teacher Cognition <?page no="65"?> studies help us understand the teacher’s role and perspective in, for example, implementing particular approaches. In summary, teacher cognition studies provide evidence for what teachers think in local contexts. In doing so, teacher cognition research has the potential to inform the work of teacher educators, curriculum developers, and teachers themselves through teacher education initiatives and the development of pedagogy. 4.3 Terminology and constructs of teacher cognition Attempts to define the various constructs studied within the field of teacher cognition is problematic not least because, as Kubanyiova and Feryok (2015, p. 436) state, “the scientific and philosophical communities lack agreement on the nature of the mind and mental states.” Furthermore, it is recognized by some that constructs such as beliefs are not solely psychological but are context specific and “co-constructed in interaction with others” (Kalaja, Barcelos, Aro, & Ruohotie-Lyhty, 2016, p. 5). The inability to describe the human mind with precision coupled with a sociocultural as opposed to individualist understanding of human cognition makes precise definitions of various constructs such as beliefs or knowledge difficult. These constructs, however difficult to define, nonetheless continue to influence the field of teacher cognition studies. The construct of “knowledge,” for example, has received significant attention since the 1980s with various researchers describing teachers’ knowledge as, for example, “practical” (Elbaz, 1981) or “personal” (Lambert, 1985). Because the current study is not concerned with teachers’ knowledge or beliefs per se, I will use the term “cognitions” (Crooke, 2015) to describe what teachers “think, know, and believe” (Borg, 2003, p. 81). My reasons for using this broader and more inclusive yet imprecise term are threefold. First, this is a study in applied linguistics, not cognitive psychology. While the field of language teacher cognition certainly relies on a variety of areas of study such as psychology and philosophy, I will not attempt to label what teachers said in the interviews in this study as “beliefs,” “knowledge,” or any other specific construct. Second, I hope to maintain relevance. It would be my hope that primary school CLIL teachers generally and DIPS teachers specifically find the results of this study useful when reflecting on classroom practices and developing CLIL pedagogy. In this way I am trying to heed Borg’s (2019, p. 20) advice and not “overly intellectualiz[e]” the concepts and constructs of the study. Third, the results of this study are more relevant to the field of CLIL than to the study of teachers’ lives. It is not my intention to map the complexities of teachers’ professional mental lives in this study. If this were my goal, a more theoretically precise construct such as “subjective theories” (Groeben et al., 1988) would be more appropriate. Instead, it is my intention to describe and understand teachers’ perspectives, thoughts, beliefs etc. about the complex and challenging CLIL context in which they teach. The more inclusive term “cognition” will allow me to do this. Despite taking this more inclusive approach in this study, the constructs of knowledge and beliefs remain important. To discuss them separately is difficult. Commonly under‐ stood, knowledge and beliefs are opposites; knowledge is objective, provable, and describes facts; beliefs are subjective, unprovable, and describe opinions. Ernest (1989, p. 30), in de‐ 4.3 Terminology and constructs of teacher cognition 65 <?page no="66"?> scribing mathematics teachers’ thinking, distinguishes between knowledge, the “cognitive outcome” of teacher education, and beliefs, the “affective outcome.” However seemingly simplistic this “head vs. heart” approach is, it does describe a common understanding of knowledge and beliefs. Nonetheless, researchers point out that the distinction between knowledge and beliefs is not always so clear (Li, 2017; Woods, 1996). Early on, Fenstermacher (1978), writing from a philosophical perspective, distinguished between teachers’ “subjectively reasonable beliefs” and “objectively reasonable beliefs.” Fenstermacher uses these terms to describe the starting and ending point of teacher learning. By exposing teachers’ subjectively reasonable beliefs to empirical evidence, teachers develop objectively reasonable beliefs. While the two states of teacher thinking that Fenstermacher presents are couched in the language of “belief,” it is possible to understand his construct of “objectively reasonable belief ” as a kind of knowledge based on facts. Fenstermacher’s juxtaposition of “objectively” with “belief ” creates uncertainty when considering the boundary between knowledge and belief. Further confusing the distinction between knowledge and belief and providing evidence of their interconnectedness is Pajares’s (1992, p. 314) observation that the psychologists Nisbett and Ross (1980), in their exploration of human decision making, “subsume belief as a type of knowledge” while Rokeach (1968), in his description of belief systems, “subsumes knowledge as a component of belief.” Nespor (1987, p. 317), in his contribution to “a theoretically-grounded model of ‘beliefs systems’,” seems also to suggest that beliefs are a form of knowledge when he describes “features of beliefs that serve to distinguish them from other forms of knowledge” (Nespor, 1987, p. 318, italics added). Pajares (1992, p. 325) himself concludes, based on his review of literature, that “[k]nowledge and beliefs are inextricably intertwined.” Accepting that knowledge and beliefs are “intertwined,” or, as Woods (1996, p. 195) claims, that they are “points on a spectrum of meaning,” descriptions of what the terms mean are still necessary. Pajares (1992) acknowledges that beliefs are a “messy construct” and describes beliefs as “an individual’s judgment of the truth or falsity of a proposition” (Pajares, 1992, p. 316, italics added). Michaela Borg (2001, p. 186) echoes Pajares, describing a belief as something that is “evaluative in that it is accepted as true by the individual, and is therefore imbued with emotive commitment” (italics added). Likewise, Calderhead (1996, p. 719) describes beliefs as “untested assumptions.” According to these descriptions, beliefs are individual and subjective evaluations of truth. Woods (1996, p. 195) states that beliefs “refer to an acceptance of a proposition for which there is no conventional knowledge, one that is not demonstrable, and for which there is accepted disagreement.” Nespor (1987) called this aspect of belief systems “non-consensuality.” Non-consensuality refers to an understanding that “propositions, concepts, arguments, or whatever […] are recognized - by those who hold them or by outsiders - as being in dispute or as in principle disputable” (Nespor, 1987, p. 321). Nespor also notes that belief systems rely more on “affective and evaluative components” than systems of knowledge do (Nespor, 1987, p. 319). Nespor further distinguishes beliefs from knowledge with the concept of “unboundedness” by which he means that beliefs can be applied beyond their “core applications” in often “radical and unpredictable ways” (Nespor, 1987, p. 321). Summarizing the above descriptions, beliefs 66 4 Teacher Cognition <?page no="67"?> are emotionally motivated, individual evaluations of truth which are accepted as disputable and are sometimes applied unpredictably to various phenomena. Contrary to beliefs, knowledge is “based on objective fact” (Pajares, 1992, p. 313) and is “empirically grounded” (Golombek, 1998, p. 447). Woods (1996, p. 195) describes knowledge as the “things we ‘know’,” “conventionally accepted facts,” and that which “has been demonstrated or is demonstrable.” Where beliefs are “imbued with emotive commitment” (Borg, 2001, p. 186), knowledge is “analytical” (Elbaz, 1981, p. 45). Because it is empirical and demonstrable, knowledge is true beyond the individual’s judgment. Nespor (1987) also notes that knowledge systems “generally have relatively well-defined domains of application and can only be applied to other domains by way of “strict rules of argument” (Nespor, 1987, p. 321). Ryle (2009, originally published 1949) states that in addition to “knowing that,” teachers are often more dependent on their ability to “know how” and that a teacher’s “intelligent” actions have knowledge embedded in them. Knowledge, then, is factual, empirical, analytical, demonstrable, can be embedded in action, and limited to strictly defined contexts of applicability. Within these contexts, knowledge can make a claim to truth that is acceptable beyond the individual. The constructs “beliefs” and “knowledge” are at the core of many of the terms and varying concepts that are studied in the field of teacher cognition research. Teacher cognition researchers have struggled, however, to agree on a set of terms to use for the various constructs on which they study. Borg describes the problem: The field is characterized by an overwhelming array of concepts. […] The need for such diversity may be justified by the inherently complex nature of the phenomena under study; however, the confusion is also due to the fact [that] identical terms have been defined in different ways and different terms have been used to describe similar concepts. (Borg, 2015, p.-40) Examples of terminology from Borg’s (2015, pp. 41-45) collection include “case knowledge” (Shulman, 1986), “implicit theories” (Dirks & Spurgin, 1992), and “schema” (Carter & Doyle, 1987). Three terms and related concepts in particular highlight the blurriness between beliefs and knowledge within the field of teacher cognition studies. “Practical knowledge” (Elbaz, 1981), “personal practical knowledge” (Connelly & Clandinin, 1982), and personal knowledge (Lambert, 1985) all describe teachers’ cognitions as a kind of qualified knowledge. The terminology alone suggests that what teachers know is built on individual understandings of teachers’ practical teaching experience. The “personal” and the “practical” blur the belief-knowledge dichotomy in two ways. First, knowledge that is “personal” would seem more congruent with definitions of belief which highlight the importance of emotion and individual evaluations of truth. Second, knowledge that is “practical” and based on experience is not empirical which is one of the descriptors of knowledge. In addition to blurring caused by terminology, Elbaz’s (1981) detailed description of “practical knowledge” mixes elements of beliefs and knowledge. For example, she describes five ways in which teachers orient or direct their practical knowledge: situationally, theoretically, personally, socially, and experientially (Elbaz, 1981, p. 49). These orientations mix aspects related to belief - for example, the personal orientation of practical knowledge - with aspects related to knowledge - for example, the theoretical orientation. 4.3 Terminology and constructs of teacher cognition 67 <?page no="68"?> All of this leads to a situation in which a clear distinction between knowledge and beliefs “is not tenable” (Woods, 1996, p.-194). Woods and Ҫakır (2011) describe teachers’ knowledge as having two dimensions: the per‐ sonal vs. impersonal dimension and the theoretical vs. practical dimension. Morton (2012) describes teachers as having four “modes of knowing” which include public-theoretical, public-practical, personal-practical, and personal-theoretical. Similar to Woods and Ҫakır, Morton maps teachers’ knowledge on the dimensions of personal vs. public and theoretical vs. practical. For Woods and Ҫakır, it is through practical classroom experience that teachers’ impersonal and often isolated theoretical knowledge becomes “deconstructed, personalized and reinterpreted” to become personal theoretical knowledge (Woods and Ҫakır, 2011, pp. 388-389). This view of knowledge as one of constant deconstruction and reinterpretation of the theoretical and practical with the personal and impersonal provides a more holistic and perhaps more stable view of teachers' cognitions. The idea that knowledge and beliefs are “inextricably intertwined” (Pajares, 1992, p. 325) or that they lie on opposite ends of a “spectrum of meaning” (Woods, 1996, p. 195) seems to be borne out in more recent publications. For example, Bonnet and Breidbach (2017) speak of “explicit knowledge” and “implicit knowledge.” Explicit knowledge refers to teachers’ knowledge of, for example, subject content and “language and content teaching techniques” (Bonnet & Breidbach, 2017, p. 281). Implicit knowledge refers to “pedagogical beliefs” (Bonnet & Breidbach, 2017, p. 282) related strongly to one’s professional identity. In this case, the term “knowledge” is qualified to refer alternatively to aspects of objectivity and subjectivity. Skinnari and Bovellan (2016) use the term “beliefs” in their study. However, when reviewing research related to teacher beliefs, they cite researchers investigating both “knowledge” and “beliefs.” These examples show that researchers use the terminology of “knowledge” and “beliefs” in varying ways to talk about the more objective and more subjective aspects of teachers’ cognitions. To conclude, teachers’ cognitions are made up of knowledge and beliefs. Their cognitions can be emotive or analytical, empirically or experientially based, theoretical or practical, and personal or public. Stated in the most general way, teachers’ cognitions are simply what they think about something. Based on the review above, teachers’ cognitions are the knowledge and beliefs that guide teachers’ decision-making and actions related to their professional lives. As previously described, the purpose of the present study is to explore primary school CLIL teachers’ cognitions and describe a particular case of primary school CLIL (the DIPS program) from teachers perspectives. Labeling teachers’ thinking as a belief, knowledge, attitude or something else is not relevant for this study. 4.4 CLIL teachers’ cognitions Studying CLIL teachers’ cognitions is still rare despite calls for studying CLIL through a teacher cognition lens (Morton, 2012; Nikula, 2017, July 5). Studies investigating primary school CLIL teachers’ cognitions are even rarer. Because of this, in the following section I will not only review research on primary and secondary school CLIL teachers’ cognitions but also research on the cognitions of teachers working in primary school EFL contexts in Germany and general education primary school contexts. Including studies from related 68 4 Teacher Cognition <?page no="69"?> fields of research may be able to support a further understanding of primary school CLIL teachers’ cognitions. Primary school CLIL pedagogy is based, in part, on the pedagogy and principles of primary schools and EFL instruction (Buchholz, 2007; Morton, 2012). Taken together, it is my intention to establish a state-of-the-art to which the current study will contribute. 4.4.1 Teacher types Dirks (2004) interviewed CLIL teachers at Gymnasien in Germany in order to investigate their “cultural frames” (Dirk, 2004, p. 129) or the various orientations to teaching and learning that guide their thinking and classroom actions. Based on the interviews and group discussions, she concludes that two archetypes, the “guardian of tradition/ culture” and the “traveler between worlds” (translation from Breidbach & Viebrock, 2012), can be found teaching in CLIL contexts. The “guardian of tradition/ culture” teacher type wants to maintain closed disciplinary boundaries. He or she prioritizes the “pragmatic and efficient” (Dirks, 2004, p. 133) transmission of disciplinary knowledge and skills as defined by the core curriculum. Learning outcomes are prioritized, for example the development of dual language subject terminology equivalencies. The classroom tends to be characterized by a focus on disciplinary content and student-teacher relationships are professionally distant. The “traveler between worlds'' wants to open disciplinary boundaries. He or she prioritizes the skills and attitudes of discovery and curiosity, focuses rather on the process of learning, and works to develop target language competences through communicative practices and personal student-teacher relationships. While Dirks (2004, p. 137) suggests that the various orientations of these two archetypes could have relevance beyond the CLIL classroom, it is unclear the extent to which they are relevant to the primary school context. Perhaps in line with Dirks’s (2004) findings, Bonnet and Breidbach (2017, p. 283) describe two secondary school CLIL teachers that they claim may represent two opposite “social types” of CLIL teachers. The first type is a teacher who is personally invested in CLIL. This type of teacher has a general pedagogical orientation to teaching and learning, sees schools as “places of learning” but does not see learning as pre-determined by set learning outcomes. Alternatively, the second type of teacher has a strong subject orientation and is therefore assumed to not be heavily invested in CLIL. This type of teacher sees schools as selective institutions and therefore learning as determined. Similar to the Dirks (2004) study, it is unknown the extent to which these possible “social types” apply to the primary school. However, because primary schools in Germany are non-selective institutions, it is plausible that the first social type described above is more representative of primary school CLIL teachers than the second type. CLIL teachers also differ in the way in which they conceptualize the target language. Hüttner and Dalton-Puffer (2013) interviewed 28 teachers at a Höheren Technischen Lehranstalten (HTL) in Austria who taught technical subjects through CLIL. They found differences in the way that subject teachers and language teachers teaching in a CLIL context conceptualized their own target language learning. English teachers conceptualized the English language as that used by a native speaker whereas CLIL subject teachers conceptualized the English language as a global lingua franca (Hüttner & Dalton-Puffer, 4.4 CLIL teachers’ cognitions 69 <?page no="70"?> 2013, p. 136). Bovellan (2014), investigating teachers’ cognitions in a primary school CLIL context in Finland, found that teachers seemed to hold themselves to a standard of English competence in line with a native-speaker ideal which led teachers to refer to their English in “dismissive terms” (Bovellan, 2014, p. 174). Almost all of the teachers in Bovellan’s study were trained classroom teachers and all of the teachers were qualified to teach English at primary school (but had not necessarily studied English). It seems reasonable to assume that the way CLIL teachers relate to and conceptualize the target language has implications for the way language is used and taught in CLIL contexts. Finally, it should be noted in this brief discussion about CLIL teachers themselves, that the literature suggests that CLIL teachers often feel “insecure” (Bovellan, 2014, p. 11) about aspects of their professional lives. First and foremost, several studies report teachers feeling insecure about their foreign language competences and whether their language skills are sufficient to teach content (Gruber & Mercer, 2022; Oattes, Oostdam, de Graff, & Wilschut, 2018; Bovellan, 2014; Díaz & Requejo, 2008; Imgrund; 2004). Some teachers have also expressed insecurity about CLIL-specific pedagogy and methodology (Massler, 2012). 4.4.2 CLIL teachers’ goals Several studies indicate that CLIL teachers focus primarily on teaching non-language content and do not focus on, for example, explicit language learning or linguistic form (de Graaff et al., 2007; Costa, 2013; Skinnari & Bovellan, 2016, Bovellan, 2014). De Graaff et al. (2007) studied the classroom interaction of nine secondary school teachers in the Netherlands in order to validate the use of a CLIL observation instrument and to gather examples of effective CLIL pedagogy. The authors found that teachers rarely, if ever, facilitated “form-focused processing” (de Graaff et al., 2007, p. 616). Teachers justified this lack of corrective feedback by stating that they are subject teachers and therefore not responsible for “explaining forms and giving rules” (de Graaff et al., 2007, p.-616) Costa (2013) observed and interviewed six Italian lecturers all teaching the natural sciences at the tertiary level in a CLIL context. Teachers expressed a clear focus on the science content of the lessons and claimed not to pay attention to language issues. Despite teachers’ prioritization of content learning, Costa did find evidence that teachers attempted to teach language implicitly through, for example, metalinguistic reflection and corrective feedback (Costa, 2013, p.-125). Skinnari and Bovellan (2016) interviewed 12 secondary school CLIL teachers in Finland, Spain, and Austria who taught either Science or History. The authors report that overall, content learning goals were prioritized. However, while some teachers reported teaching “the same knowledge” (Skinnari & Bovellan, 2016, p. 151) in their CLIL and non-CLIL courses, others reported adjusting the content learning goals in various ways. For example, a Spanish teacher reported teaching history content through an Anglo-Saxon lens in her CLIL courses whereas an Austrian teacher questioned the need to teach certain curriculum-prescribed content that she felt was not important (Skinnari & Bovellan, 2016, pp. 151-152). Bovellan (2014) investigated the cognitions of eight primary school CLIL teachers in Finland during a three-stage study which included interviews, keeping a diary, and creating 70 4 Teacher Cognition <?page no="71"?> materials. She also found that teachers focus on content learning in their CLIL lessons. She explains that this is likely because all but one of the teachers in her study were trained class teachers and therefore less likely to “reflect the meaning of language for the construction of knowledge and learning” (Bovellan, 2014, p. 173). Additionally, she notes that this finding is in line with CLIL as a content-focused pedagogy. None of these studies are from the German context. However, considering that CLIL in Germany is explicitly and officially a subject-focused pedagogy (e.g. teachers follow the non-language subject curriculum, courses are timetabled as the non-language subject), it would be expected that many CLIL teachers in Germany would hold a content-focused view of CLIL. While content learning is the clear priority of teachers in the studies above, CLIL is usually motivated by the promise of increased foreign language competences. Viebrock (2007), in her investigation of the subjective theories of four CLIL geography teachers at secondary school in Germany, found that teachers considered the development of foreign language competences as very important. Additionally, teachers did not feel that learning through a foreign language hindered students’ learning of non-language content. However, Skinnari and Bovellan (2016) report that the secondary school CLIL teachers in their study described foreign language learning in CLIL as “a side-effect,” a “by-product,” or “a spice” (Skinnari & Bovellan, 2016, p. 153). The authors explain that these descriptions of language learning are evidence of a belief that foreign languages are “acquired naturally through FL content learning” (Skinnari & Bovellan, 2016, p.-153). With respect to language learning generally, CLIL teachers seem to focus primarily on vocabulary learning and communicative competences, specifically oral competences. Schauer (2019) used an online questionnaire to elicit the classroom practices of 27 EFL teachers teaching in Thüringen. She found that teachers prioritize speaking and listening skills. She notes that this result is not surprising considering that speaking and listening skills are prioritized in the primary school English curriculum in Thüringen and in many publications on ELT with young learners (Schauer, 2019, p.-272). Imgrund (2004) reports on an early trial of French CLIL lessons in Switzerland in grades five and six. She concludes that listening comprehension as well as “everyday” and subject-specific vocabulary should be primary language learning goals (Imgrund, 2004, p. 124). Bovellan (2014) reports that the majority of teachers in her study viewed language as “a system or a set of words” and therefore as a “tool to encode the world” whereas a minority of teachers viewed language as a communicative tool with which to teach content (Bovellan, 2014, p. 174). These two views of language suggest that teachers may prioritize either vocabulary learning or communicative learning goals. Teachers who view language as a collection of words may be inclined to prioritize vocabulary learning while teachers who view language as a communicative tool may prioritize the development of communicative competences such as speaking and listening. Morton (2012) in his investigation of secondary school CLIL teachers’ cognitions reports that teachers focused “almost exclusively” on vocabulary when describing “language-focused practices” in their CLIL classrooms (Morton, 2012, p.-217). 4.4 CLIL teachers’ cognitions 71 <?page no="72"?> These studies show that while foreign language learning can be seen as either very important or as a kind of side-effect in CLIL, teachers primarily focus on vocabulary learning and the development of oral competences in the context of their CLIL teaching. The development of intercultural competences is an established goal of CLIL (Mehisto, 2013; Wolff, 2016). However, teachers seem not to prioritize it in their thinking or their teaching. In Schauer’s (2019) study, primary school EFL teachers ranked the development of intercultural competences fifth out of seven possible goals. Viebrock (2007) also reports that the CLIL teachers in her study did not make any significant mention of intercultural development as a goal. Imgrund (2004, p. 124) states that when there are “genuine” intercultural themes, it is appropriate to include such goals. This conclusion would suggest that intercultural development is not a goal in all units of study. Bovellan (2014) reports that teachers did consider cultural aspects when creating materials for their students but also viewed materials from English-speaking countries as problematic due to their lack of correspondence to the local curriculum. Bovellan concludes that theoretical goals such as the development of intercultural competences “do not always correspond to the practices of the field” (Bovellan, 2014, p. 175). Taken together, these studies suggest that teachers are generally aware that intercultural competences should be developed but that it is not a driving force in their thinking about CLIL. Goals not directly related to content or language learning are also the objects of teachers’ thinking. Wegner (2012) investigated German (Hessen) secondary school students’ and teachers’ perspectives on CLIL. She conducted interviews with two CLIL teachers of Politics and Economics before and after observing and video-recording classroom practice (grade 7 and grade 9) to elicit their “theories on CLIL” (Wegner, 2012, p. 29). She also conducted group interviews with students from these classes after recording classroom practice. She found that teachers emphasized neither subject-specific learning nor specific language learning goals but instead valued the development of social, psychological, and affective aspects of students. For example, teachers wanted to “promot[e] students’ self-confidence” (Wegner, 2012, p. 30), “emphasise the motivation and activation of the students” (Wegner, 2012, p. 32), and create communicative situations in which students could “experience success” (Wegner, 2012, p. 32). Wegner concludes, in part, that such emphasis on general education learning goals and teachers’ perception that teaching through a foreign language is challenging led to the “lowering of expectations as regards to subject standards” (Wegner, 2012, p.-30). Skinnari and Bovellan (2016) also found that secondary school CLIL teachers work toward goals other than learning content and language. Teachers saw CLIL as a “new pedagogy and learning culture” that offered opportunities to “wide[n] perspectives,” “cul‐ tivat[e] respect and understanding,” and use new teaching methods (Skinnari & Bovellan, 2016, p. 162). They also report that some teachers thought that CLIL could “enhance student cognitive skills” by, for example, “offering them new schemes for thinking” (Skinnari & Bovellan, 2016, p. 162). These two studies suggest that teachers see learning opportunities unique to CLIL that go beyond traditional content and language learning goals. Nonetheless, such general pedagogic goals may result in lowered content learning expectations (Wegner, 2012). 72 4 Teacher Cognition <?page no="73"?> 4.4.3 Teachers’ approaches Teachers’ thinking about how best to plan for and implement instruction in order to meet their learning goals for students is based on a variety of factors. One of those factors is a knowledge and understanding of the students themselves. König, Krepf, Bremerich-Vos, & Buchholtz (2021) identify teachers’ knowledge of students’ learning dispositions as one of six factors for measuring pre-service teachers’ lesson planning competences. They state, “We expect the planning teacher to reflect on how both learning dispositions of students and learning tasks are related to each other, so that learners are guided into their ‘zone of proximal development’ (Vygotsky, 1978, p.-84)” (König et al., 2021, p.-471). There is evidence that CLIL and non-CLIL teachers do this. Sandberg (2019), in her study of eight upper secondary school Biology and History CLIL teachers in Sweden, found that the teachers in her study adapted their instruction, specifically their language use, to specific groups of students as well as to individuals. For example, based on students’ understanding of key biology concepts, teachers chose either to further explain the terminology in the L2 or to translate the terminology into the L1. Other studies from non-CLIL primary and middle school contexts (Sardo-Brown, 1988; Wing-mui So, 1997) and EFL instruction (Baily, 1996) have also found that knowledge and consideration of specific groups of students is a primary factor when planning for and implementing instruction. Knowledge of students is something that CLIL and non-CLIL teachers use in their planning and implementation of curriculum. The way that CLIL teachers understand the process of learning a foreign language also has ramifications for how teachers approach their CLIL instruction. A key belief of many CLIL teachers is that foreign language learning takes place naturally in CLIL. Hüttner and Dalton-Puffer (2013, p. 135), who studied vocational school CLIL teachers’ cognitions in Austria, report that teachers subscribed to a theory of “‘natural’ language acquisition comparable to an English-language work setting.” Costa (2013, p. 125) describes the university lecturers teaching through English in Italy as holding a belief that “implicit” teaching methods such as corrective written feedback is sufficient for language learning in the tertiary CLIL environment. Likewise, Skinnari and Bovellan (2016) report that most of the teachers in their study of secondary school CLIL teachers believed in “‘natural’ language learning” (Skinnari & Bovellan, 2016, p. 153) which they associated with “easy-going and relaxed ways of learning” (Skinnari & Bovellan, 2016, p.-154). The belief that foreign language acquisition occurs naturally is connected to the belief and the practice of maximizing target language use in the CLIL classroom. Lasagabaster (2017) found that all of the respondents in his study believed in maximizing English use in the classroom. Hüttner and Dalton-Puffer (2013) report that the teachers in their study conceptualized language learning as an active “doing”: “Die stärkste Überzeugung im Rahmen der subjektiven Theorien aller Beteiligten ist, dass Schülerlernen ein aktives „Tun“ ist. Das heißt, dass Lernen als wiederholtes Üben durch häufigen aktiven und passiven Gebrauch der Zielsprache Englisch konzeptualisiert wird” (The strongest conviction within the subjective theories of all participants is that student learning is an active “doing.” This means that learning is conceptualized as repeated practice through frequent active and passive use of the target language English) (Hüttner & Dalton-Puffer, 2013, p.-135). 4.4 CLIL teachers’ cognitions 73 <?page no="74"?> This belief in “natural” learning through language use has merit. As cited in section 2.2.1 above, Marsh (2000, p. 3) claims that CLIL provides a “natural situation for language development” and that CLIL’s “naturalness” (italics in original) is one of its “major platforms for […] success.” Also, usage-based theories of second language acquisition describe language acquisition as taking place through language use. Despite this belief in natural language learning in CLIL, teachers (at least in the secondary school context) also seem to believe that vocabulary is best taught explicitly (Hüttner & Dalton-Puffer, 2013; Morton, 2012; Viebrock, 2007). Teachers in Hüttner and Dalton-Puffer’s (2013) study, for example, regularly made use of vocabulary learning tools such as glossaries, dictionaries, and vocabulary lists (Hüttner & Dalton-Puffer, 2013, p. 137). It is unknown whether primary school CLIL teachers hold similar beliefs. How teachers think about the use of authentic materials is also relevant to the study of CLIL teachers’ cognitions. Bovellan (2014) investigated teachers’ cognitions about authentic materials. She categorized teachers’ responses into three groups. First, teachers consider authentic materials as those that are in the target language and have not been created for educational purposes. The majority of Bovellan’s respondents held this view of authentic materials. In Bovellan’s examples, authentic materials were both written (“texts”) and audio-visual (“small video clips”). Teachers seemed to recognize the potential challenge of authentic texts and at least one teacher explicitly mentioned this challenge as the reason for using authentic texts: “I’ll have to teach the kids that there will always be these difficult words (.) and you’ll have to get over with them (.) that you have to understand the text (.) as a whole” (Sirpa as cited in Bovellan, 2014, p. 141). Second, teachers understand authentic materials as educational materials from target language countries. One example provided by Bovellan is that of a science textbook, Science Now (2019). At least one teacher used this textbook in her sixth grade science class because it “covers the themes” and contains “rich language” (Bovellan, 2014, p. 141). Teachers recognized the linguistic challenge of such texts and therefore scaffolded texts by “going through” them together (Bovellan, 2014, p. 141). Third, a minority of teachers (two) felt that authentic materials are materials that are applicable to the local, i.e. Finnish, context. One teacher stated: [W]e are in Finland and we have Finnish systems and so on (.) so I don't know if we have to like if we should know things in similar systems as elsewhere (.) actually I wouldn't like that we don't want to be any others we want to be us like we only have the language we aim at developing the language. (Tarmo as cited in Bovellan, 2014, p.-144). As Bovellan concludes, this statement demonstrates a belief that in CLIL, the target lan‐ guage competence can be developed separately from the locally defined content and culture. While Bovellan accepts a degree of caution when considering the use of foreign-made materials in new or local contexts, she states that some of the respondents expressed “prejudiced” views which highlighted the high value they placed on “culture-specific education” (Bovellan, 2014, p.-145). It is questionable whether a teacher with “prejudiced” views toward the target language culture is the right teacher for CLIL. The view that authentic materials are those which are applicable to the local context is not a description of authentic materials presented in scientific literature on materials. The first two descriptions from Bovellan, however, are. 74 4 Teacher Cognition <?page no="75"?> Reviewed in section 3.1.1 above, Böttger (2020, p. 77) defines authentic materials both as those created by and for target language speakers with no educational intention in mind as well as educational materials created by and for target language speakers. The majority of teachers in Bovellan’s (2014) study expressed these views of authentic materials. Beyond the question of authenticity in materials, teachers in Bovellan’s (2014) study as well as those in a number of other studies (Siepmann, Rumlich, Matz, & Römhild, 2021; McDougald, 2015; Massler, 2012; Bechler, 2014; Viebrock, 2007) view the creation of CLIL materials as time-consuming and generally burdensome. However, Sandberg (2019), reports that the more experienced teachers whom she interviewed described the creation of English-language CLIL materials as a “positive challenge” and a “creative task” which deepened their own knowledge of the subject (Sandberg, 2019, p.-305). Teachers’ thinking about the use of the L1 in CLIL is varied. While an L2 only principle of CLIL education has traditionally dominated notions of best practice (Lasagabaster, 2017), Diehr (2012) has criticized such ideas. Leykum, Heinze, & Gropengießer (2012), in their interviews with seven German CLIL teachers at Gymnasien, indeed found that teachers themselves tend not to initiate use of the L1 in CLIL lessons but instead preferred a monolingual approach. There is teacher cognition research, however, which suggests that teachers value the use of the L1 for specific purposes in the CLIL classroom. Lasagabaster (2017) held three group discussions with a total of eight secondary school CLIL teachers in Spain to, in part, find out when and why teachers use the L1 and the L2. He found that teachers use the L1 for behavior issues, supporting student comprehension, and comparing the L1 and L2. He also cites a study from Lasagabaster (2013) which was conducted in Colombia. That study found that teachers use the L1 for the three reasons described above as well as to “boost debate” and to “feel comfortable in the CLIL class” (Lasagabaster, 2013 as cited in Lasagabaster, 2017, p. 264). Lasagabaster (2017) concludes that using the L1 for behavior issues, supporting student comprehension, and comparing the L1 and L2 “may be a general trend in secondary education CLIL classes irrespective of the context” (Lasagabaster, 2017, p.-264). Indeed, Sandberg (2019) found that the secondary school CLIL teachers in her study used Swedish to explain complex content as well as for translation purposes. Sandberg describes teachers adapting such L1 use to specific groups of students and to individuals. For example, some teachers described lecturing in English and taking questions in Swedish as well as responding to students’ questions in the language in which the question was asked. Sandberg also found that teachers translated subject terminology into the L1 when “they sensed that the students could not follow what they were talking about” (Sandberg, 2019, p.-306). Oattes et al. (2018) investigated the thinking of 86 secondary school CLIL Dutch teachers using a questionnaire. They found that teachers used the L1 to increase students’ English and Dutch vocabulary, to support comprehension of English and Dutch words as well as “historical or academic concepts,” and to prepare students for the transition back to “mainstream history lessons in grade 10” (Oattes et al., 2018, p. 170). The authors state that while some teachers felt they were “breaking the rules” (Oattes et al., 2018, p. 170), at least 4.4 CLIL teachers’ cognitions 75 <?page no="76"?> one teacher stated, “it is so very obvious that it should be a mix of languages, that you need Dutch to support English language learning” (Oattes et al., 2018, p.-170). In primary school CLIL contexts, teachers may switch between the L1 and the L2 for a variety of reasons. Massler (2012), in her interviews with eight primary school CLIL teachers in Germany, found that teachers used the L1 to save time and “to reduce the disadvantage to cognitively weaker students” (Massler, 2012, p. 42). Botz and Frisch (2016) used an online questionnaire to gather data on German (NRW) primary school CLIL teachers’ use of the L1 and L2 and the functions of using German (L1). Of the 16 teachers who responded to the question about L1 and L2 practices, 55 % reported sometimes using German (e.g. through a German-speaking teacher or through German phases of the lesson). Of the 12 teachers who responded to the question about the function of using German, over 80 % described using German to ensure the learning of German subject terminology and approximately 75 % used German to ensure content understanding. Only 50 % described using German to support the development of subject language skills in German (Botz & Frisch, 2016, p. 242). Taken together, these studies suggest that when teachers use the L1, they use it primarily as a scaffold for content and L2 learning. Even reasons such as saving time or reducing “disadvantage” (Massler, 2012, p. 42) are essentially about scaffolding comprehension. However, in the case of the Dutch secondary school context, there appears to be a belief that the L1 should be used to develop academic understanding in the L1 since students will need to transfer back to L1 instruction at a later date. At primary school, there is some evidence that teachers sometimes use the L1 to support the development of academic language skills in the L1. Teachers also use visualization in the creation of CLIL materials (Bovellan, 2014). Bovellan (2014) documents five reasons for this visualization. First, teachers use visuali‐ zation to make content concrete. Teachers use pictures, objects, and actions as well as digital media to be able to show and demonstrate to students content-related concepts. Bovellan concludes that teachers’ responses indicate that they use visualization primarily for content learning and not language learning (Bovellan, 2014, p. 147). Second, teachers described the need for materials to be aesthetically pleasing. According to several of the teachers, materials should be “beautiful” and “positive” and therefore motivate students by suggesting to individuals that it is “an important paper” and has been “made for her” (Bovellan, 2014, p. 151). Third, teachers used visualization to stimulate classroom conversation. Bovellan gives an example of a teacher including a picture of a fox and a wolf in her PowerPoint to introduce a new unit of study. Bovellan concludes that not only do the pictures concretize the animals in question but also that they motivate students to learn more (Bovellan, 2014, p. 152). Fourth, some teachers involved students in the creation of materials. Bovellen notes that this was rare in the data but that it indicates a learner-centered approach to CLIL learning whereas most of her data indicated a teacher-centered approach. Bovellan concludes that teachers who had students create materials held a belief that learner-centered approaches led to better student learning (Bovellan, 2014, p. 153). Fifth, teachers used digital media to make links to students’ non-school lives. Similar to using pictures to stimulate conversation, teachers used short video clips to introduce topics and motivate students and used longer educational series to teach content. By using a media format that children experience in their lives away from school, Bovellan concludes that 76 4 Teacher Cognition <?page no="77"?> teachers believe the use of such media in the classroom will positively influence students’ motivation and learning (Bovellan, 2014, p.-154). The various ways in which teachers in Bovellan’s study use visualization ultimately serve the purpose of scaffolding. Even aesthetic and motivational reasons serve the purpose of supporting student learning by drawing them into the processes and objectives of learning. Two of the main objectives of CLIL are content and foreign language learning. While Bovellan concludes that teachers use visualization primarily for the purpose of learning content, such visualization has the effect of scaffolding language as well due to the integrated nature of content and language learning in CLIL. As a scaffolding tool, visualization is fundamental for primary school CLIL teachers’ work. Finally, how primary school teachers think about and use variety has also been studied. Two studies, one from the primary school EFL context and one from the primary school CLIL context suggest that teachers use variety in different ways in their classrooms. The EFL teachers in Schauer’s (2019) study varied social forms such as group work and pair work, though whole class and pair work were the most often used by teachers. As a group, the teachers in Schauer’s study used a wide variety of children's books as well as songs. There is also evidence that individual teachers used a variety of songs in their classrooms (Schauer, 2019). Schauer’s results, however, do not include teachers’ thinking or reasoning for their use of such variety. The CLIL teachers in Bovellan’s (2014) study preferred “tasks that involve variety and diversity” because they felt that this motivated children to learn and engage with the activity (Bovellan, 2014, p. 173). Based solely on Bovellan’s (2014) results, teachers seem to use variety to motivate and engage students. 4.5 Summary In this chapter, I have described the history of language teacher cognition studies and have outlined various goals and applications of teacher cognition research. Furthermore, I have reviewed key terminology and constructs within the field of teacher cognition research as well as research on CLIL teachers’ cognitions. Teaching in CLIL contexts is complex and understanding CLIL teachers’ cognitions can help stakeholders, including teachers themselves, better understand this complexity. Because research on primary school CLIL teachers’ cognitions is rare, much of the research reviewed above comes from other, related contexts such as secondary school CLIL and primary school EFL. While this research is not a substitute for research conducted specifically in primary school CLIL contexts, it can contribute to some initial understanding of this context. The research reviewed here indicates that CLIL teachers focus primarily on non-lan‐ guage content learning goals and to a lesser extent on foreign language learning goals. Any focus on foreign language learning in CLIL is generally limited to vocabulary learning and to developing oral language competences. This is especially true in the primary school CLIL context where oral language skill development tends to be the focus of EFL lessons. There is also evidence that, at least at secondary school, teachers want to develop skills and attitudes such as self-confidence, motivation, and even general cognitive ability that lie outside of traditionally defined content and language learning goals. Notably absent from teachers’ 4.5 Summary 77 <?page no="78"?> cognitions about CLIL is an explicit focus on developing intercultural communicative competences. A number of studies presented above describe the cognition that language learning is a “natural” process and one associated with “easy-going and relaxed ways of learning” (Skinnari & Bovellan, 2016, p. 154). Teachers also report valuing the use of authentic materials. A number of studies indicate that when teachers do need to create their own CLIL materials, they find the process time-consuming and burdensome. Teachers use the L1, visualization techniques, and variety in order to scaffold meaning and motivate learners. This suggests that teachers may see CLIL learning as cognitively and emotionally challenging. Through scaffolding and motivation techniques, teachers hope to help students overcome these challenges. 78 4 Teacher Cognition <?page no="79"?> 5 Methods and Research Design 5.1 Aims and research questions With CLIL’s increasing popularity has come more research into a variety of CLIL-related topics. Primary among research topics include the effects of CLIL on L2 learning such as vocabulary acquisition (Fernández-Fontecha, 2015; Castellano-Risco, Alejo-González, & Piquer-Píriz, 2020) and listening comprehension (Nold et al., 2008; Dallinger et al., 2016; Pérez Cañado, 2018) as well as L1 development (Botz & Diehr, 2016; Zaunbauer, Bonerad, & Möller, 2005) and content learning (Frisch, 2021; Surmont, Struys, van den Noort, & van de Craen, 2016; Dallinger et al., 2016). Additional research has focused on topics such as CLIL teachers’ code-switching (Gierlinger, 2015), CLIL classroom discourse (Dalton-Puffer, 2007), student motivation (Sylven & Thompson, 2015; Rumlich, 2016), and general stakeholder (teachers, parents, students) attitudes toward and perceptions of CLIL (Denman, Tanner, & de Graaff,- 2013; Raez-Padilla, 2018). While teacher cognition studies in the CLIL context do exist (Viebrock, 2007; Morton, 2012; Bovellan, 2014), the number of such studies are comparatively few. Morton (2012, p. 12) finds this fact surprising since, as he puts it, “CLIL can be seen as a complex hybrid of practices drawing on the pedagogies of different academic subjects and those of language education.” In a presentation on subject and language integration in CLIL given at the University of Oxford, Nikula (2017, July 5) also notes that more research is needed into teachers’ “beliefs and perceptions relating to CLIL” (10: 13). Teddick and Cammarata (2012), writing from a North American CBI (content-based instruction) perspective, also argue for more research into teachers’ “actual experience” (Teddick & Cammarata, 2012, p. S48) implementing various bilingual programs. They state that “[b]ecause teachers are the ultimate decision makers as to what enters their classroom and because all curricular reforms are filtered through their beliefs and perceptions, CBI has little chance to succeed without their support, interest, and motivation” (Teddick & Cammarata, 2012, p. S48). This premise is foundational to any study of teacher cognition. It is also important that there be CLIL research in the various school forms and in various national and regional contexts. In Germany, for example, there have been important publications focused on primary school CLIL related to CLIL modules (Bechler, 2014), assessment (Massler & Stotz, 2013), subject competences (Frisch, 2021), and the use of the L1 and L2 (Botz & Diehr, 2016). Nonetheless, there is less research into primary school CLIL in Germany than into secondary school CLIL in Germany. Ohlberger and Wegner’s (2018) review of CLIL studies reveals that out of the 18 studies included that were conducted in Germany, none of them were conducted at primary schools. The 17 primary school studies reviewed were all conducted in other European countries. Because CLIL is realized differently in different contexts, it is important that researchers study these contexts. Massler (2012) recognizes this lack of research into primary school CLIL in Germany as well. She states, “This lack of research can distort the way in which CLIL implementation may be understood and formulated, since data significant to the entire process is lacking” (Massler, 2012, p.-36). <?page no="80"?> In addition to there being less research focused on various primary school CLIL contexts, there are also factors specific to primary school CLIL that complicate its implementation. First, primary school students usually have beginning and relatively low levels of target lan‐ guage competences. Target language competences are “an essential prerequisite” (Morton, 2012, p. 12) for content learning, something that is recognized in secondary schools in NRW where students receive additional EFL lessons for two school years before entering their first CLIL course. At primary school, Frisch (2021) provides evidence that CLIL students performed poorly on a test of science competences when tested through English (reviewed in section 1.1.2). Teaching curriculum-defined content through a target language that is largely undeveloped in students would seem to pose a significant and unique challenge to primary school CLIL teachers. Second, there is some evidence that primary school CLIL teachers lack CLIL-specific training and guidance. In Massler’s (2012) project (ProCLIL) that studied primary school CLIL teachers, 10 out of the 12 teachers had no previous CLIL-specific training. Such a finding raises the question of how common CLIL-specific training is for practicing CLIL teachers. While primary school CLIL teachers in NRW are trained EFL teachers, there are unique challenges to the CLIL context that teachers may be unprepared for. CLIL teacher training is vital not only for giving teachers the knowledge and skills necessary to implement CLIL but also to give them the confidence to persevere in a challenging context. As noted in the previous chapter, some teachers experience a sense of insecurity related to their perceived lack of CLIL-specific pedagogical understanding (Massler, 2012). Furthermore, whereas the Ministry of Education in NRW provides handbooks on CLIL in several subjects such as geography and biology for secondary school (MSB NRW, 2021, March 4), no such guidelines exist to support primary school CLIL teachers as they implement CLIL programs. Teachers know that they are responsible for teaching the content of the non-language subject curriculum but questions about how exactly to do this through a foreign language are left unaddressed by the written curriculum and other official documents. In, for example, the Handbook for German-English Biology at Secondary School Level 1 (Handreichung Biologie deutsch-englisch in der Sekundarstufe I) (MSW NRW, 2012), a theoretical background is provided as well as an extensive practical example of how one might plan a unit of study for CLIL Biology at secondary school. The NRW Ministry of Education website has only one article (Wolff, 2020, Sept. 7) that addresses primary school CLIL specifically. The article outlines broad theoretical as well as structural issues related to starting and implementing CLIL programs that, while important, are unlikely to help practicing CLIL teachers navigate practical pedagogical problems. As described above, the primary school CLIL context is uniquely challenging and research into primary school CLIL teachers’ cognitions is limited. Therefore, the first purpose of this study is to explore and describe primary school CLIL teachers’ cognitions. Exploring and describing what teachers think about their CLIL practice is an important first step to better understanding CLIL teachers and improving CLIL pedagogy. Research into various aspects of CLIL, especially small, qualitative studies such as this one, is always carried out in specific CLIL contexts often with unique programmatic features. Such is the case in this study which has been conducted in the context of the Dortmund International Primary Schools program (to be described in detail in section 80 5 Methods and Research Design <?page no="81"?> 5.3.1). A second purpose, then, of this study is to describe the DIPS program from teachers’ perspectives. Such descriptions have the potential to inform teachers and administrators as they work to improve their specific CLIL programs. In order to explore and describe primary school CLIL teachers’ cognitions, it is necessary to initially seek input on a variety of CLIL-related topics. Teachers’ experiences with CLIL are likely broad and varied and they may have a variety of views concerning aspects of CLIL such as perceived effects of CLIL instruction, language use and learning in CLIL, or teacher demands. There is evidence, for example, that CLIL teachers conceptualize the target language in various ways (Hüttner & Dalton-Puffer, 2013; Bovellan, 2014) and prioritize learning in the CLIL classroom differently (Skinnari & Bovellan, 2016). These and other aspects of teachers’ thinking about CLIL will be explored in this study. Beyond teachers’ cognitions about aspects of CLIL pedagogy such as students, teacher demands, learning, etc., CLIL pedagogy can be understood as a set of goals and approaches intended to lead students to those goals. CLIL goals (described in section 2.1.3 above) are described differently in different CLIL contexts. In Germany, two primary goals for CLIL have been articulated by Diehr: the development of subject literacy in two languages and the development of culturally aware multi-perspectivity in subject-specific contexts (Diehr, 2012). However, Frisch (2021) has questioned the appropriateness of developing subject literacy in two languages for primary school CLIL contexts based on her research that found primary school CLIL students achieved drastically lower scores on a test of natural science competences when tested in English rather than in German. Such a finding not only questions the appropriateness of this goal but also begs the question of what goals teachers actually set for their students in their CLIL instruction. There is a wealth of evidence from a variety of CLIL contexts that teachers prioritize non-language content learning (see, for example, Skinnari & Bovellan, 2016; Bovellan, 2014; Costa, 2013; de Graaff et al., 2007). When focusing on language learning, teachers in both primary school EFL contexts and primary and secondary school CLIL contexts tend to focus on teaching vocabulary and developing oral communication competences (Schauer, 2019; Bovellan, 2014; Morton, 2012; Imgrund, 2004) but also work toward the development of general pedagogic goals such as motivation and respect (Skinnari & Bovellan, 2016; Wegner, 2012). In exploring CLIL teachers’ thinking, it is, therefore, important to investigate what goals teachers set for students. Finally, in order to achieve the purposes of this study, it is necessary to investigate what approaches teachers employ and why they use such approaches. One aspect of teachers’ approaches in the CLIL classroom concerns their use of the L1. Because the goal of CLIL in Germany is, in part, the development of subject literacy in two languages, Botz and Diehr (2016, p. 251) have advocated for the “sparse yet targeted” use of the L1 in CLIL instruction. There is evidence from secondary and tertiary contexts that suggests teachers see foreign language learning in CLIL as a natural process (Hüttner & Dalton-Puffer, 2013; Costa, 2013; Skinnari & Bovellan, 2016) and therefore seek to maximize L2 input (Lasagabaster, 2017) but nonetheless teach subject vocabulary explicitly (Hüttner & Dalton-Puffer, 2013; Morton, 2012; Viebrock, 2007). Describing what teachers think about best to achieve their goals in CLIL is important to a broad investigation of primary school CLIL teachers’ cognitions. 5.1 Aims and research questions 81 <?page no="82"?> Based on the review of theoretical and empirical CLIL literature presented above and in previous chapters and in order to achieve the two purposes of this study, three research questions will be investigated: 1. What cognitions do teachers hold about teaching and learning in primary school CLIL? 2. What language-related educational goals do teachers work toward? 3. What approaches do teachers think are most effective for achieving these goals? These research questions are broad and therefore allow for a more expansive investigation into the thinking that guides DIPS teachers’ daily CLIL practices. 5.2 Methods In order to investigate the three research questions, data was gathered using semi-struc‐ tured interviews and classroom observations and was analyzed using qualitative content analysis. In the following sections, I will first describe the possibilities and limitations of using verbal data (interviews) for eliciting teachers’ cognitions. Then I will describe the specific data gathering methods of semi-structured interviews and classroom observations and why they have been used to meet the purposes of this study. Finally, I will describe the data analysis method of qualitative content analysis and how it fulfills the purposes of this study. 5.2.1 Collecting and analyzing verbal data Interviews or, more generally, verbal commentaries, have a long and well-established place in teacher cognition research (Borg, 2015). Two assumptions about the use of language and about teachers themselves underpin the use of interviews as a viable tool for eliciting teachers’ cognitions. First, it is assumed that language is capable of expressing thought or cognition. Second, it is assumed that teachers are explicitly aware of what they think and can, therefore, describe what they think. It is clear that while language and thought are “intertwined and internally connected” (Li, 2020, p. 52), they are not the same thing (Lund, 2003; Pinker, 2013). Their relationship, of course, is complex. One of the fundamental “design characteristics” of language is displacement or the ability to use language to talk about things removed in time and space (Hockett, 1960 as cited in Lund, 2003, p. 2). Such a characteristic is essential if interviews are to be considered a viable tool for eliciting teachers’ cognitions. Beyond this foundational characteristic of language, however, are varying hypotheses that attempt to describe the relationship between language and thought. Two seemingly opposing hypotheses are the linguistic relativity hypothesis (LRH) (alternatively known as the Sapir-Worf hypothesis or Whorfian hypothesis), which claims that language either influences (weak form) or determines (strong form) thought (Parkin, 2000; Lund, 2003) and a hypothesis put forward by Piaget (writing from a developmental psychology perspective) which claims that a child must understand a concept before being able to use language to express the concept (Lund, 2003). Moving beyond the question of which comes first, language or thought, Smith (2017, p. 95) claims that our internal “conceptual system 82 5 Methods and Research Design <?page no="83"?> contains meaning that has no expression in language.” Taken together, language either influences or determines thought, is capable of expressing thought only after the thought exists, or is simply unable to express all existing thoughts. While language may be one of the most ubiquitous tools of human communication, it limits human thought and is an imprecise tool for expressing human thought. Nonetheless, for Lakoff and Johnson (1980, p. 3) because language “is based on the same conceptual system that we use in thinking and acting, [it] is an important source of evidence for what that system is like.” Despite any imprecision, verbal data such as verbal reports and interviews have been used to study how people think (e.g. Ericsson & Simon, 1993) as well as what people think (e.g. Li & De Costa, 2018). Grotjahn (1987, p. 70), describing verbal reports of thinking processes, lists two challenges to the validity of such reports. First is the challenge of making a “valid interpretive reconstruction” of the intended meaning and second is that verbal reports always “have a representational function.” Though writing from a sociocultural as opposed to cognitive perspective, Li (2020, p. 52) echoes Grotjahn’s description of language as representational when she states, “What people say is not a channel to their mental state, rather, it is their positioning, understanding and beliefs that are manifested in the words they use and the way they communicate.” The challenges to verbal reports described by Grotjahn are, therefore, also relevant to verbal data such as data gathered with interviews. Talk is an opaque window into the content and processes of the mind. It will always be a subjective expression of thought based on internal cognitive processes as well as the external social context. It is possible to know (or think, believe, etc.) something without being conscious of it. From a philosophical perspective, this idea has been put forward by Michael Polanyi in his 1966 publication, The Tacit Dimension, in which he states, “we can know more than we can tell” (Polanyi, 1966, p. 4, italics in original). But also from the perspective of using introspection to elicit thinking processes, there is the recognition that research participants are only capable of reporting on that which they are conscious of (Cohen, 1987, p. 84). This point is obvious enough, but it warrants being mentioned here to underscore the limits of verbal data to accurately describe teachers’ cognitions. Teachers can only express cognitions which they are conscious of. Despite the shortcomings of language to express thoughts or interviews as a tool to elicit teachers’ cognitions, no viable alternative exists. People are, in part, shaped and limited by their ability to use language. However, considering the goals of this research project, not all of the limits of verbal data are necessarily problematic for this study. For example, the fact that language more likely represents teachers’ “positioning, understanding and beliefs” (Li, 2020, p. 52) as opposed to some hypothetically pure mental state is in line with the broad goal of this project to describe what teachers think about teaching and learning in their primary school CLIL context. It may not be possible to observe teachers’ thinking directly, but by listening to what they say, it is possible to observe “how they argue the course of their action and decision-making” (Li, 2020, p.-52). 5.2 Methods 83 <?page no="84"?> 5.2.2 Semi-structured interviews Semi-structured interviews are exactly what their label implies; they are interviews guided by a partially structured set of topics or questions. Dörnyei (2007, p. 136) defines them as interviews that entail a “set of pre-prepared guiding questions and prompts, the format is open-ended and the interviewee is encouraged to elaborate on the issues raised in an exploratory manner.” Richards (2009, p. 186) describes them as interviews “where the interviewer has a clear picture of the topics that need to be covered (and perhaps even a preferred order for these) but is prepared to allow the interview to develop in unexpected directions where these open up important new areas.” Semi-structured interviews generally stand on a continuum between structured interviews which are characterized by a degree of standardization (Borg, 2015, p. 223) and unstructured interviews characterized by “maximum flexibility” intended to allow the interviewee to take the interview in varied and “unpredictable directions” (Dörnyei, 2007, p.-135). As a small-scale, qualitative study designed to explore primary school CLIL teachers’ cognitions, semi-structured interviews, as opposed to, for example, questionnaires, were deemed an appropriate method of data gathering that could reveal “more detailed views of individuals” (Coyle et al., 2010, p. 143). As opposed to more or less structured interviews, semi-structured interviews were deemed a viable data gathering method because of the flexibility inherent in a method located somewhere in the middle on “a continuum ranging from more structured to less structured” (Borg, 2015, p. 221). This flexibility allows questions to be asked within the context and structure laid out by the research questions but also allows for some degree of conversational freedom (Borg, 2015, p. 236) and exploration (Dörnyei, 2007, p. 136). When seeking to explore CLIL teachers’ cognitions, researchers have sought teachers’ views on a range of CLIL topics including, for example, perceived student learning, teacher motivation, instructional techniques, and materials used (Lan‐ caster, 2018). Employing semi-structured interviews can support such a wide-ranging and exploratory approach to understanding teachers’ cognitions. As opposed to the breadth of information that researchers can gather from large numbers of teachers through the use of questionnaires, interviews allow researchers to probe deeper into teachers’ cognitions’ (Richards, 2009, p. 187). However, the extent to which researchers can gain access to teachers’ cognitions via interviews depends on several factors. First among them is simply who the interviewee and interviewer are and how their various identities relate to one another. For example, personal characteristics such as social status, ethnicity, skin color, gender, or age can influence relationships (Prior, 2018, p. 240). Likewise, power relationships, professional and cultural insider vs. outsider status, and whether or not the interviewee and interviewer share an L1 also influence the relationship (Canh & Maley, 2012, pp. 98-99). Since “interviews are interpretively active” (Holstein and Gubrium, 1995, p. 4) and “jointly constructed encounters” (Richards, 2009, p. 190 italics in original), the interviewees are “constructors of knowledge in collaboration with interviewers” (Holstein & Gubrium, 1995, p. 4). Second, far from being opportunities to freely “excavat[e]” (Holstein & Gubrium, 1995, p. 4) teachers’ cognitions, interviews require interviewing skills that allow teachers to express their thoughts. Prior (2018) notes the need to develop rapport with the interviewee. One aspect of building rapport is language use. When interviewing multilingual respondents or when the interviewee and interviewer do not share an L1, 84 5 Methods and Research Design <?page no="85"?> Pavlenko (2007) recommends allowing the interviewee to choose which language to use. This includes the choice to switch between languages. Also crucial to developing rapport is the “art of listening” (Richards, 2009, p. 189). Teachers are unlikely to be interested in sharing their thoughts if the interviewer is not interested in their responses or is unable to listen because of concerns about, for example, recording or upcoming questions (Richards, 2009). Borg (2015) highlights the awkwardness of supposed semi-structured interviews that lack interviewer listening skills. They proceed like highly structured interviews and, therefore, do not allow “social interaction” or opportunities to dig deeper into teachers’ responses (Borg, 2015, p.-243). Finally, allowing the interview to “develop naturally” (Richards, 2009, p. 186) is crucial to developing rapport and harnessing the advantages of semi-structured interviews. This naturalness and flexibility in semi-structured interviews is best achieved through preparation and practice (Borg, 2015). 5.2.3 Classroom observations Classroom observations can be employed in a variety of ways to study the language learning classroom. Used in teacher cognition research, classroom observation “provides direct information” (Dörnyei, 2007, p. 178) about the interactions and events taking place in lessons. Observations can complement the information teachers report in interviews by providing a “concrete descriptive basis” (Borg, 2015, p. 273) for the interpretation of teachers’ cognitions. The complimentary role that observations take in the study of teachers’ cognitions should not be taken to suggest their lack of importance. As Borg (2003, p. 105) states, “Ultimately, […] we are interested in understanding teachers’ professional actions, not what or how they think in isolation of what they do.” Therefore, classroom observations provide a way for researchers to validate or at least to contextualize teachers’ self-reported cognitions. The combination of classroom observations and teacher interviews has the potential to provide a more holistic view of a CLIL program from the teacher’s perspective. Based on his review of literature concerning the use of observations as a methodological tool, Borg (2015) concludes that there are eight dimensions to observations that researchers need to consider. These dimensions include, for example, “participation” which describes “[t]he extent to which the observer participates in the settings under study” as well as “authenticity” described as “[t]he extent to which the settings under observation are naturally occurring” (Borg, 2015, p. 269). The complete list is reproduced in Table 3 below. 5.2 Methods 85 <?page no="86"?> Dimension Description Options (from/ to) Participation The extent to which the observer participates in the settings under study. Participant/ Non-par‐ ticipant Awareness The extent to which those observed know they are being so and by whom Overt/ Covert Authenticity The extent to which the settings under observation are naturally occurring Real/ Contrived Disclosure The extent to which the purposes of the observation are explained to those being observed Full/ Minimal Recording How a record of the observation is made Manual/ Technological Structure The extent to which data are recorded against prede‐ termined analytical categories Closed/ Open Coding The extent to which data are coded according to existing frameworks Deductive/ Inductive Analysis The role of quantification in the analysis Quantitative/ Qualita‐ tive Scope The extent to which a range of individuals, events and times are studied Limited/ Extended Table 3: Dimensions of observational research (Borg, 2015, p.-269) The choices that researchers make regarding each of these dimensions is influenced by re‐ search aims, theoretical and methodological approaches, as well as practical considerations such as resource availability. Researchers’ choices also influence aspects of validity and reliability of their research and have ethical implications as well. Choosing to fully disclose one’s research aims may appear to be the most ethical thing to do but may also lead to less valid and unreliable research results. From the 18 teacher cognition studies Borg (2015) reviewed, 16 of them described the researcher role as non-participant, suggesting that this is the more common role for researchers to take within such studies. As a non-participant in classroom observations, researchers typically situate themselves at the back of the classroom, take notes, and do not interact with the teacher or students (Borg, 2015, p.-273). Despite this relatively unobtrusive presence in the classroom, the “observer’s paradox” or the notion that “the act of observation will change the observed person’s behavior” (Cowie, 2009, p. 177) is a feature of observational research that needs to be considered. A researcher’s presence, especially in a primary school classroom, can be made more disruptive by the use of technical recording equipment such as video-recording devices. For this reason, Hall (2007, p. 6) recommends using a “minimal equipment rig” when recording. Researchers need to work to mitigate the observer’s paradox by, for example, observing over an extended period of time during which research participants become more desensitized to the observer’s presence. Especially relevant to observing primary school classrooms, is the extent to which researchers can truly be a non-participant. Dörnyei (2007, p. 179) claims that “it is impossible 86 5 Methods and Research Design <?page no="87"?> to observe in almost every nonexperimental situation without some participation.” The primary school CLIL classroom, as an “authentic” observation context in which teachers and students are doing real work, is one such research situation. Borg (2015) highlights the ethical aspects of non-participation in classroom observation. To refuse requests from teachers or students to, for example, check the spelling of a word or answer a student question is not possible simply because one decided not to participate in the classroom. Borg (2015, pp. 274-275) advises instead to participate in such minimal ways, describe the kind and degree of participation in the research report, and not to use such interactions as data. Fieldnotes offer researchers a technologically simple and less intrusive method of recording classroom observations than, for example, video-recordings. However, they do not allow the researcher to, for example, accurately record long stretches of classroom discourse and are inherently subjective records of classroom events. As Copland (2018, p. 253) notes, fieldnotes are “always incomplete,” “always representative,” and “always evaluative.” In a study by Humphries (Humphries & Gebhard, 2012, p. 117), the researcher recorded “factual data,” “reflective remarks,” and “questions for follow-up interviews” in his field‐ notes. Copland (2018, p. 256 citing Emerson et al., 1995, p. 32) advises researchers to “describe neutrally” and use “[c]oncrete sensory details” in order to help them remember events when reviewing the notes at a later time. Cowie (2009, p. 167) advises researchers to “make detailed notes […] about the place, the people, and the interactions that occur.” Because fieldnotes are “constructed accounts over time,” they become a “corpus” (Emerson et al., 2007, p. 353 as cited in Copeland, 2018, p. 257) of observational data that researchers can analyze or use as a basis for classroom description. 5.2.4 Qualitative content analysis Schreier (2012, p. 1) defines qualitative content analysis (QCA): “QCA is a method for systematically describing the meaning of qualitative material. It is done by classifying material as instances of the categories of a coding frame.” Several aspects of this definition will be explored further below. First, QCA is a method of data analysis and not, as Kuckartz (2019, p. 11) states, a “methodology” or “epistemology.” It is not concerned with a particular philosophical or theoretical approach to the world but, as a method, is simply “a means of gaining knowledge, a planned process based on rules” (Kuckartz, 2019, p. 11). It involves applying a series of steps leading to the interpretation and presentation of results (Schreier, 2012, p.-6). Second, QCA is systematic. According to Schreier (2012, p. 9) QCA “is systematic in three ways: all relevant material is taken into account; a sequence of steps is followed during the analysis, regardless of your material; and you have to check your coding for consistency (reliability).” The “necessity of a systematic approach” (Mayring, 2015, p. 49, my translation) in QCA stems from QCA’s origins in quantitative content analysis which Berelson (1952, p. 18 as cited in Schreier, 2012, p. 13) defines as “a research technique for the objective, systematic, and quantitative description of the manifest content of communication.” The systematic approach of QCA, therefore, is a feature of quantitative content analysis that 5.2 Methods 87 <?page no="88"?> proponents of QCA have adopted in order to fend off accusations of being “impressionistic” and “arbitrary” (Mayring, 2015, p.-29) in their qualitative approach to data analysis. Third, the coding frame lies “at the heart” of QCA (Schreier, 2012, p. 58). A coding frame is a “way of structuring […] material” and “consists of main categories specifying relevant aspects and of subcategories for each main category specifying relevant meanings concerning this aspect” (Schreier, 2012, p. 61). Each category must be defined. Category definitions “are the rules for assigning data segments to categories” (Schreier, 2012, p.-94). Coding frames can be built deductively or inductively. A deductive approach involves building the coding frame based on theoretical concepts before beginning the analysis while an inductive approach involves building the coding frame based on the data (Schreier, 2012). According to Gläser-Zikuda, Hagenauer and Stephan (2020, p. 5) educational researchers often build their coding frames deductively and inductively in order to “take advantage of the strengths of both approaches.” In so doing, they often first create deductive categories based on theory followed by creating inductive categories based on the data (Gläser-Zikuda et al., 2020, p. 5). Deductive categories are also often based on the research questions as they “specif[y] certain dimensions” of the research (Schreier, 2012, p. 87). Regardless of how the coding frame is created, the frame itself represents a kind of outline or summary of the data that it describes. QCA is also a flexible data analysis method. According to Schreier (2012) this flexibility primarily concerns the coding frame which needs to be adapted to the data being analyzed. The reason for such flexibility is that a rigid and deductively constructed coding frame in the style of one used for quantitative content analysis may reduce the validity of the coding frame (Schreier, 2012, p. 7). Likewise, Mayring states that QCA is “not a standard instrument […]; it needs to be adapted to the concrete material and designed to address the specific research questions” (Mayring, 2015, p.-51, my translation). Finally, QCA reduces data. It does this first by selecting topics or themes within the data on which to focus. This is done by focusing on data relevant to the concepts entailed in the research questions (Schreier, 2012, p. 7). Second, the coding frame, as a representation of the data at a “higher level of abstraction” (Schreier, 2012, p. 7) does not include details that exist in the data itself. It instead organizes those details into broader and more inclusive categories and subcategories and in doing so reduces data. By reducing data and by organizing meaning in the data into a coding frame, QCA offers the possibility to answer the research question in a succinct, organized, and clear manner. QCA, as a qualitative method based on a quantitative tradition, has faced criticism. This criticism is “precisely because of these hybrid characteristics” (Schreier, Stamann, Janssen, Dahl, & Whittal, 2019, p. 4) of QCA. Mayring himself describes QCA as “neither a purely qualitative nor a purely quantitative method” (Mayring, 2015, p. 8 as cited in Schreier et al., 2019, p. 4). Some qualitative researchers do not recognize QCA as a qualitative research method at all (see Schreier et al., 2019, p.-4). Despite Mayring recognizing the hybridity of QCA and others criticizing QCA for this fact, Marvasti (2019, p. 4) sees QCA as “a subset of qualitative research.” He concludes “that there is not a single unifying approach that defines QCA” (Marvasti, 2019, p.-3). At least three features of QCA make it particularly conducive to the purposes of this study, namely the exploration of primary school CLIL teachers’ cognitions as well as a 88 5 Methods and Research Design <?page no="89"?> description of the DIPS program from teachers’ perspectives. First, QCA is a systematic approach to data analysis which takes all of the gathered material into account and applies a set of analytical steps to that material (Schreier, 2012). Such a structured approach to data analysis provides metaphorical guard rails to the analysis process, thus potentially increasing the validity and reliability of results and resisting “impressionistic” and “arbitrary” (Mayring, 2015, p. 29) results. Second, the ultimate result or “heart” of QCA is the coding frame (Schreier, 2012, p. 58). While the coding frame can be created deductively, inductively, or with a combination of both (Schreier, 2012), the final coding frame provides a kind of summary of the data in the form of categories and sub-categories. In the case of the present study, these categories and sub-categories can be used not only to describe teachers’ cognitions about their CLIL teaching practices, but can also be used as a tool to clearly communicate results back to program stakeholders and to the wider scientific community. While a merely descriptive result has been criticized by various authors as Schreier (2012, p. 4) summarizes, such a descriptive result can meet the purposes of this study by providing a concrete starting point from which CLIL stakeholders (e.g. teachers, researchers, and administrators) can work to further develop CLIL pedagogy and CLIL teachers’ instructional practices. Teachers’ professional development is one established use of teacher cognition research (Kubanyiova & Feryok, 2015; Calderhead, 1987b). Third, QCA employs a version of “quasi-statistics” (Becker, 1970, p. 81) that further facilitates communicating results to DIPS program stakeholders and the scientific community. In this study, the quantification of coded segments within a sub-category provides information about the internal generalizability (i.e. the strength or weakness) of a particular cognition expressed by teachers (Maxwell & Chmiel, 2013). The quantification of such an “empirical finding” (Schreier, 2012, p.-188) can be easily communicated. Given different research purposes and different research questions, any number of other analytical approaches including, for example, grounded theory (Thornberg & Charmaz, 2013), documentary method (Bohnsack, 2013), or narrative analysis (Esin, Fathi, & Squire, 2013) could be viable approaches for the investigation of teachers’ thinking. Considering the rather practical intent of this study and the need to be able to communicate results succinctly and clearly, QCA, for reasons already described, was deemed an appropriate data analysis method. 5.3 Research design 5.3.1 Research context: Dortmund International Primary Schools This study investigates teachers working within the Dortmund International Primary Schools (DIPS) program, a network of five primary schools with CLIL branches in Dort‐ mund, Germany. The DIPS network was started in the school year 2010-2011. In addition to supporting the early development of English as a foreign language, the network of primary schools was created to close a gap between bilingual preschools and bilingual secondary schools in Dortmund (Raunser & Steffens, 2012, p.-37). Each of the schools in the network has one bilingual class in each of the four grade levels at school. The various subjects that are taught through English include Science and Social Studies (Sachunterricht), Art, Music, 5.3 Research design 89 <?page no="90"?> and Physical Education. Mathematics and Religion are taught through German. English as a Foreign Language and German are taught through their respective target languages. All DIPS-school students, regardless of their participation in the CLIL class, receive three EFL lessons per week as opposed to the standard two EFL lessons at other primary schools in NRW (Raunser & Steffens, 2012, p. 37). While the total number of weekly lessons varies slightly based on grade-level, most students have between 23 and 27 lessons per week with CLIL students having roughly 10-12 of those lessons in English (EFL and CLIL subjects combined). Therefore, slightly less than 50-% of their school week is in English. The DIPS program is organized and implemented around several core features that, when combined, create the framework of its unique version of CLIL. First, all Science and Social Studies lessons in the DIPS program are taught by two cooperating teachers, a German-speaking and an English-speaking teacher. This organizational feature is the most noticeable and unique aspect of the DIPS program. Second, the DIPS program is described as implementing an immersion approach to language learning (DIPS, 2020, Sept. 7). From this immersion approach comes the third core feature of a “one person-one language” (Raunser & Steffens, 2012, p. 37) approach to teachers’ language use. The English-speaking teacher should only speak English and the German-speaking teacher should only speak German. These three features and how they were observed to be enacted will be described further below. First, while all participating DIPS schools implemented a team-teaching approach in the Science and Social Studies lessons, each of the participating schools organized teachers’ roles differently. At one school (Teacher A’s school) the German-speaking teacher was the class teacher and the English-speaking teacher came to the class to teach CLIL Science and Social Studies with the class teacher (I will refer to this teacher arrangement as Type 1). At another school (Teacher D and E’s school), the roles were reversed. The English-speaking teacher was the class teacher and the German-speaking teacher came to class during the CLIL Science and Social Studies lessons (Type 2). In the third school (Teacher B and C’s school), the German-speaking and English-speaking teachers shared the role of class teacher and therefore took joint responsibility for all of the various tasks involved with being a class teacher (Type 3). Teachers at this school made this point explicit on several occasions. Other CLIL subjects (Art, Music, and Physical Education) were taught solely by the English-speaking teacher. The way CLIL teaching was implemented in these three types of teacher arrangements was noticeably different. The cause of these differences may be related to the various teacher arrangements but may also be related to factors unrelated to teacher arrangement such as teacher personalities. Regardless of the cause, I will describe these differences in order to provide a better understanding of the research context. There were two noticeable differences between the various teacher arrangements. First, the class teacher, regardless of what language she used in the class, usually took a more active role. For example, I observed in the Type 1 classroom (the only classroom like this) that the German-speaking class teacher was very quick to interject in whole-class discussions that had been started by the English-speaking teacher. From my observation summary (dated March 13, 2019), I recorded: 90 5 Methods and Research Design <?page no="91"?> [English-speaking teacher], at 10: 14 introducing the topic of Vikings in English. She reminded students of answering the “who” questions from the previous day and then quickly moved to asking, “Where did they live? ” [The German-speaking class teacher] quickly followed up with “Wo? Wo haben sie gelebt? ” This basic pattern of the English-speaking teacher asking a question and the German-speaking teacher translating soon after was common in this class. What, if any, learning effect this had for students is unknown. The pattern did, however, seem to reduce opportunities to use English during whole-class instruction. By contrast, teacher talk in Type 2 and Type 3 classes was far more dominated by English. During whole-class instructional phases, the German-speaking teachers did contribute, sometimes by translating when a translation was deemed necessary and sometimes by posing additional questions or making their own contributions to instruction. For example, in Teacher D’s class, Teacher D posed the question, “What could that be? Waren? ” (field notes, December 5, 2019). In my notes, I recorded that in this instance, the German-speaking teacher did not offer any translation. However, on an earlier occasion, the German-speaking teacher translated the question, “Are all these countries in Europe? ” after no response was given by students (field notes, October 31, 2019). This pattern in which the English-speaking teacher is the clear leader of whole-class phases of the lesson and the German-speaking teacher plays a more supportive role was observed in all classes in Type 2 and Type 3 schools. In all three types of classes, most students responded in German. Second, because the class teacher is responsible for a variety of classroom tasks beyond instruction, more of these extra-curricular or organizational tasks were completed in English when (one of) the class teacher(s) was the English-speaking teacher (Type 2 and Type 3). For example, a class council was conducted in English during one of the lessons. Preparing a birthday present for the German-speaking teacher during her absence was led by the English-speaking teacher. A large conflict between students was dealt with almost exclusively in English by the English-speaking teacher. Based on my observations, it appeared as though there were more opportunities to use English in Type 2 and Type 3 classrooms than in Type 1 classrooms. It is beyond the scope of this study to evaluate any potential effect on (language) learning in these various teaching arrangements. For this study, it is simply important to know that despite all of the participating teachers working at schools that all share the same conceptualization of CLIL, each local (school) realization of CLIL is slightly different. This could potentially influence the way that teachers perceive or understand CLIL. The way that teachers in this study planned their Science and Social Studies lessons was also potentially affected by teaching with a German-speaking colleague. While I did not specifically investigate teachers’ planning processes in this study, I did ask teachers to speak about how they planned with their cooperating teachers. Teachers planning processes with their team teachers varied. Teacher A, for example, reported that she and her team teacher “try to meet once a week to plan the science lesson” (TA1 00: 34: 12) for about 45 minutes but also expressed some regret that more time was not available for them to plan together. In contrast, however, Teacher D reported that she and 5.3 Research design 91 <?page no="92"?> her German-speaking teaching partner do not sit down together regularly to plan anymore since they have been teaching together for some years and feel that such regular planning is unnecessary. Nonetheless, they do communicate with one another about the Science and Social Studies lessons as they feel necessary. I: Planning? Do you plan those lessons together […] ? TD: […] Nicht mehr dadurch, dass wir wirklich ein richtig ehm, durch lange Zeit eingespieltes Team sind, brauchen wir das zum Glück nicht mehr. […] Vielleicht wenn wir/ wenn wir gemerkt haben in dem Vorjahr gab es Schwierigkeiten an der ein oder anderen Stelle oder wenn ich gemerkt habe auch das könnte auch auf Englisch gehen, […] dann würde ich das Arbeitsblatt nochmal umstrukturieren und würde meinen Kollegen nur kurz mitteilen, "Ich habe das geändert" oder "Ich habe das ehm zu Auswahl gestellt […] (TD1 00: 10: 41) The majority of teachers in this study do not regularly sit down to plan with their team teacher. The exception is Teacher A. Based on interview responses, it seems that Teachers B, C, D, and E do most of the Science and Social Studies planning on their own. When they do communicate about lessons with the German-speaking teacher, the focus is often about language and vocabulary. The planning habits of all five teachers, i.e. the extent to which they regularly plan with their German-speaking team teachers, seems to mirror their classroom working relationships as observed throughout the study. The class teacher and observed leader of the lessons took more responsibility for planning the Science and Social Studies lessons. In the case of Teacher A, this meant more regular planning with the German-speaking class teacher. Because Teachers B, C, D, and E were (one of) the class teachers of their respective classes and led all of the lessons that I observed, it is possible that they simply felt more responsibility for the Science and Social Studies lessons and therefore planned them without significant input from the German-speaking teacher. Immersion, as an approach to language learning, is the second core feature of the DIPS program. In the DIPS program, immersion is conceptualized as a “Sprachbad” in which “the students are literally immersed in the English language” (DIPS, 2020, Sept. 7, my translation). Language learning through immersion is seen as similar to the process of acquiring a first language since it is learned through daily interactions and not systematically (DIPS, 2020, Sept. 7). One way that the DIPS program implements their conceptualization of immersion is through the third core feature of the DIPS program, the “one person-one language” approach to teachers’ language use. Teachers adhered to the one person-one language policy or approach most of the time. Having one teacher speak English almost exclusively to children was the primary way in which teachers in this study enacted an immersion approach to teaching and learning in DIPS. According to DIPS program documents, students should “see the English teacher as a native speaker, since they only hear the teacher speak English” (DIPS, 2020, Sept. 7, my translation). DIPS students did, however, occasionally here the English-speaking teacher use German. In the interviews, teachers describe reasons that they might use German with students (to be detailed further is section 6.3.1). In one example, Teacher 92 5 Methods and Research Design <?page no="93"?> C’s German-speaking colleague was absent. Teacher C chose to initiate a whole-class discussion of a conflict by saying, “Probably best if we talk about this in German.” She proceeded to speak German with students (field notes, November 12, 2019). Even though such German-speaking instances were rare and teachers spoke mostly English with students, it is unlikely that students thought their teachers were native English speakers. Despite identifying as an immersion program, the DIPS program falls slightly under the 50 % L2 instructional threshold to be labeled as an immersion program according to scientific definitions. Furthermore, the DIPS program’s target language is a foreign language for most of the DIPS students. Historically, immersion programs have targeted a second language, i.e. one that is spoken in the surrounding community but not by the immersion program students. The topics and methods of the lessons that I observed varied. The lessons tended, however, to be structured similarly. Typically, they opened with a whole-class phase led by the English-speaking teacher. The German-speaking teacher usually took a supportive role limited primarily to language-related issues. This phase would then be followed by students working, perhaps at different stations or on worksheets. During this phase of the lesson, both teachers would interact with students independently. Sometimes, students worked on a German language worksheet and received assistance from the German-speaking teacher. Such instances were examples of how, at times, interactions between students, materials, and teachers could be monolingual German and therefore lack any essential CLIL traits. Lessons usually concluded with another whole-class phase led by the English speaking teacher. This phase was used for a variety of purposes, including, for example, sharing findings and reviewing lesson content and language. In order to further illuminate DIPS pedagogy and how it might be implemented in a lesson, I will describe one observed lesson in some detail below. The lesson took place in Teacher D’s third grade class on November 14, 2019. The students are sitting at their desks, organized in pods of roughly four students. To the front of the classroom is a traditional chalkboard as well as a large interactive whiteboard set off to the side. The interactive whiteboard is turned off. On the right side of the classroom hangs a column of five A4-sized cards, each naming a particular phase of the unit of instruction that has taken and will take place over the next several weeks. The cards are written in German with an English translation placed under the German headings. Teacher D starts the lesson speaking English while her German-speaking team teacher stands quietly to the side. Teacher D has one student move the arrow down to the card which says, “Wir informieren uns an einer Lerntheke über Europa” (We will learn about Europe at a learning station). Teacher D then introduces students to the learning station, a table toward the front of the classroom with materials for the unit on Europe. At this point in the lesson, Teacher D only takes time to point out that there are German and English worksheets on the table as well as an atlas for students to use later when completing the worksheets. She instructs students that they will need to work on their own and try to solve any problems that they encounter on their own. After briefly introducing the learning station, Teacher D directs students’ attention to the front of the classroom. On the chalkboard there is a large map of Europe. The map is cut into 5.3 Research design 93 <?page no="94"?> puzzle pieces that are attached to the chalkboard with magnets. Each puzzle piece contains several countries. There is also a simple compass rose attached to the chalkboard. Teacher D starts by reviewing the cardinal directions in German and in English with students. For the directions in German she uses the mnemonic device “Nie ohne Seife waschen.” The students recite this. In English she use the mnemonic device “Never eat Shredded Wheat.” The students recite this. Teacher D then moves on to assembling the puzzle piece map and rehearsing with students how to describe the relative location of countries in German and in English; for example, “Deutschland liegt südlich von Dänemark; ” “Germany is south of Denmark.” Teacher D starts by having one student move a puzzle piece into the correct place. The student then needs to describe, in German or in English, the relative location of a country on the map. With help from Teacher D’s German-speaking team teacher, the student says, “Deutschland liegt westlich von Polen.” After several students have described the location of countries using German, Teacher D suggests, “I’m sure some children can do it in English.” With help from Teacher D, one student says, “Germany is south of Denmark.” In this way, students continue to describe the relative locations of countries. Both teachers continue to help students as needed. Occasionally, Teacher D will suggest to a specific student that he or she can do it in English: “I’m sure that [student’s name] can do it in English too.” While most students use German, some students use English even without direct prompting. After all students have recited at least one sentence in either German or English, Teacher D introduces the learning station in more detail. The learning station is a table placed toward the front of the classroom. On it lie a total of eight worksheets, four in German and four in English. The four English worksheets are identical to the German worksheets with the exception of language. For example, there is a worksheet titled “Die europäischen Länder.” The worksheet contains an outline map of Europe with arrows pointing from a space in which to write the name of the country to each country. The instructions include, “Setze ein: Bulgarien, Dänemark […] Turkei.” The English equivalent contains the same map with arrows and spaces to write country names. The worksheet is titled, “The European countries” and includes the instructions, “Fill in the names of the European countries: Bulgaria, Denmark […] Turkey.” The other three worksheets are titled “The Europe Bus” (Europa-Bus), “The flags of Germany’s neighbours” (Die Fahnen der Europa-Nachbarn), and “The continents” (Europa, ein Kontinent). After Teacher D has introduced the worksheets in more detail, she tells students that they can choose whether to complete the German worksheet or the English worksheet. Students do not need to complete both. She then tells students, “Think before you choose a worksheet.” Teacher D acknowledges that many students may want to do the German worksheets but challenges students to do some of the worksheets in English. Teacher D then calls one group of students at a time to come forward and select their worksheets. Students begin working on their worksheets. As they work, Teacher D and her team teacher circulate throughout the classroom independently of each other offering support to students as needed. Students move freely throughout the classroom at this time, checking the map at the back of the classroom, the puzzle map on the chalkboard at the front, the atlas at the learning station, and a bilingual list of countries and capitals hanging on a side wall. During this time, I walk through the classroom and count 12 children, approximately half of the class, completing an English version of the chosen worksheet. Students spoke German 94 5 Methods and Research Design <?page no="95"?> with one another during this phase of the lesson. On this day, students only worked for about 10 minutes before Teacher D asked students to clean up. She then ended the lesson. This lesson is typical of many of the lessons observed in several ways. First, the English-speaking teacher led the lesson while the German-speaking teacher took a more supportive role. With the exception of the two teachers in the Type 1 school (see above), all of the observed lessons proceeded in this way. In most of the lessons, English input dominated during whole-class phases of instruction. Second, there was a mix of German and English-language materials. Most observed lessons used a mix of German and English materials. German-language materials were, however, more common than English materials. Third, there were moments in the lesson in which no essential CLIL feature, namely, the use of the foreign or target language, was present. This situation occurred in many of the observed lessons when students spoke German to one another, worked on German-language materials, and were supported by the German-speaking teacher. Unique to this lesson, however, was the explicit focus on language form in German and in English during the map building portion of the lesson. In most of the lessons, high amounts of English input was provided and English vocabulary was discussed or described explicitly. However, English and German chunks or subject discourse skills in both languages were not taught or rehearsed explicitly, as they were in the lesson described above. In this way, most of the lessons observed enacted a passive or input-centered form of immersion instruction in which little English output was produced by students. 5.3.2 Research participants Five primary school CLIL teachers working at three different DIPS schools participated in this study. The schools in which data was collected range in size from between approximately 300 students with 12 classes to approximately 500 students with 19 classes. All five DIPS schools were contacted about the possibility of participating in this study. The project was initially conceived of as interviewing teaching teams, i.e. the English-speaking and the German-speaking teachers in the CLIL Science and Social Studies lessons. Only one out of approximately 10 available teaching teams agreed to participate (I delayed asking teaching teams at one school until after revising the project design). Teams that declined to participate cited time constraints or the unwillingness of one team member to participate as reasons for not participating. With this overwhelming rejection, I redesigned the project in order to get additional participants. Instead of seeking to interview both members of a teaching team, I decided to only interview the English-speaking teacher. I also removed the planning diary, the post-lesson debriefing interviews, and video-recording lessons from the original design (the final research design is described below). With this scaled-down design, I approached four CLIL teachers. All four agreed to participate. Each of the total of five participating teachers was debriefed on the project and signed a consent form. Teachers were guaranteed anonymity and given the option to rescind their consent at any time. Because Teacher A participated in the project before its redesign, her students’ parents signed consent forms for video-recording lessons. Ultimately, the video-recordings were not analyzed. Parent consent forms were not issued in the other four classes. 5.3 Research design 95 <?page no="96"?> Teachers were told that the project was being conducted as a PhD project. Additionally, they were told that the principal goals of the project were to better understand what CLIL teachers do in the classroom and the reasons for their actions. While this brief description does not include the language of “attitudes,” “cognitions,” “knowledge,” etc., it does describe the essential goals of the project. The five teachers who participated in this study form a relatively homogeneous group. All of the teachers are trained primary school teachers. All of the teachers studied English at university in addition to the other subjects required to become a primary school teacher. None of the teachers have completed an additional study program dedicated to developing CLIL-specific teaching competences. A summary of biographical descriptors is in Table 4 below. - Teacher A Teacher B Teacher C Teacher D Teacher E mother tongue German German German German German gender female female female female female age 36 37 33 36 31 years teaching 9 years 9 years 6 years 10 years 6 years years teaching CLIL 8 years 8 years 5 years 7 years 3 years teaching CLIL certificate no no no no no Table 4: Participating teachers’ biographical data 5.3.3 Data gathering procedures The research project was divided into three phases. Phase one consisted, first, of a short questionnaire (10 items) to gather biographical data (see Appendix 2). This was deemed the most efficient way to gather basic information on teachers such as age, teaching experience, and CLIL-specific teaching experience. The second part of phase one consisted of a background interview intended to elicit cognitions about CLIL teaching broadly. Topics included, for example, questions about CLIL goals, materials, and planning (see Appendix 3 for the background interview guide). Conducting a background interview before observing CLIL lessons oriented me to the teacher’s thinking about her instructional practices and therefore provided additional information which could help focus my classroom observations. For example, if a teacher described the importance of only speaking English with students, I could observe the degree to which the teacher did this and when, if ever, the teacher spoke German to students. Phase two consisted of a series of classroom observations conducted over approximately six to eight weeks. Observations were recorded using fieldnotes. The goal of the obser‐ vations was to see and record how languages are used in lessons (i.e. who uses what language when) as well as to observe the types of activities used and lesson structures. Observations also provided context in which to further interpret statements made during 96 5 Methods and Research Design <?page no="97"?> the background interviews and raised additional questions to be further explored in the follow-up interviews. Phase three consisted of follow-up interviews designed to record teachers’ reflections on the observed unit of instruction as well as to elicit teachers’ cognitions on specific aspects of the lessons observed (see Appendix 4 for the follow-up interview guide). Because the interview was semi-structured, teachers were allowed to express views tangential to questions outlined in the interview guide. The research plan is summarized below in Table 5. Phase 1 • Biographical information questionnaire (see Appendix 2) • Pre-observation semi-structured interview (45-60 minutes) (see Appendix 3) Phase 2 • Weeks 1---8 • 1 classroom observation per week Phase 3 • Follow-up interview (30-60 minutes) (see Appendix 4) Table 5: Data gathering plan The data gathering process that was implemented differs in several ways from the plan outlined above. First, Teacher A participated in a more elaborately designed pilot phase of the study. This original pilot phase consisted of interviews with the German-speaking and English-speaking teachers, video-recorded classroom observations, short (10-20 minutes) post-observation reflections, and a written planning journal in addition to the background and follow-up interviews described above. This plan proved too time-consuming for most teachers to be willing to participate. The plan was simplified to the one outlined above in order to make participation in the study feasible for more teachers. Data gathered from Teacher A (not her German-speaking team teacher) and classroom observations are included in the final analysis. Second, due to, for example, scheduling conflicts and occasionally absent teachers, there were, at times gaps longer than one week in classroom observation. I have outlined the total interview time and observation time for each participating teacher in Table 6 below. 5.3 Research design 97 <?page no="98"?> Teacher Total interview time Total classroom observation time Observation dates Additional notes Teacher A (observed Grade 3 CLIL lessons) 142 minutes 540 minutes (2x45 minutes and 5x90 mi‐ nutes) March-April 2019 80 out of the 142 minutes of inter‐ view time was conducted to‐ gether with Teacher A and her German speaking team teacher. Teacher B (observed Grade 2 CLIL lessons) 101 minutes 315 minutes (5x45 minutes and 1x90 mi‐ nutes) October-De‐ cember 2019 - Teacher C (observed Grade 4 CLIL lessons) 130 minutes 630 minutes (7x90 minutes) October 2019-- -January 2020 - Teacher D (observed Grade 3 CLIL lessons) 98 minutes 360 minutes (8x45 minutes) October-De‐ cember 2019 Teacher D de‐ cided to teach one of these lessons as an EFL lesson be‐ cause the German-speaking teacher was ab‐ sent. Teacher E (observed Grade 4 CLIL lessons) 105 minutes 315 minutes (7x45 minutes) October-De‐ cember 2019 - Table 6: Amount of data gathered for each teacher Notwithstanding the already mentioned differences in data gathering procedures for Teacher A, the quantity of data gathered for Teachers B, D, and E is very similar. The longer interview time for Teacher C is a reflection of her talking more. No additional interviews were conducted. The longer classroom observation time for Teacher C is a reflection of observing 90-minute lessons instead of 45-minute lessons. All teachers were observed on between six and eight occasions over the course of two to four months. Interviews were conducted in this study as the primary means of eliciting teachers’ cognitions. All interviews were conducted at the interviewee’s school in an empty class‐ room, office, or small library. A set of topics and possible questions within each topic were prepared ahead of time in order to guide the background interview (see Appendix 3). All background interviews were conducted before any classroom observations took place. Follow-up interviews with four out of the five teachers were conducted between one and two months after the final classroom observation. This allowed me time to review my field notes and develop an outline to guide the interview (see Appendix 4). The interviews were recorded with a small digital recorder placed on the table between the teacher and researcher. The follow-up interview with Teacher C took place three and a half months after 98 5 Methods and Research Design <?page no="99"?> the final classroom observation in April 2020 due to teacher health issues. The interview was held and recorded on Skype due to the Covid-19 pandemic lockdown. During the pilot phase of the project, post-lesson interviews were conducted with Teacher A and her German-speaking team teacher. This data was transcribed and added to the corpus of background and follow-up interview data from Teacher A. Only statements from Teacher A were analyzed. The questions that I prepared were in English. Because semi-structured interviews are “jointly constructed encounters” (Richards, 2009, p. 190 emphasis in original) that require the interviewer to listen actively and respond in flexible and creative ways (Hobbs & Kubanyiova, 2008 as cited in Canh & Maley, 2012), I deemed it necessary, based on my own evaluation of my German-language competences, to ask questions and respond in English. I gave teachers the choice of responding in English or German, but suggested to them that they may feel more comfortable responding in German. They were also told that they could switch between languages freely. Ultimately, two teachers responded almost exclusively in English, two teachers responded almost exclusively in German, and one teacher regularly switched between languages. The background interview topics and questions were broad and wide-ranging. Topics included, for example, notions of quality, success, and challenge in CLIL learning as well as goals, assessment, and materials. The purpose of this was to allow teachers to talk about CLIL from a variety of perspectives. For example, by asking questions such as, “What advice would you give a teacher new to bilingual teaching? ” or simply, “What do you want students to learn in your bilingual lessons? ” it was my hope that teachers would reveal aspects of their goals or approaches in CLIL. In structuring the questions this way, I was attempting to heed Woods’s (1996, p. 27) advice to avoid overly specific questions such as “What is your approach…? ” Furthermore, some questions were designed to elicit “personal narratives or stories about concrete events and experiences” (Woods, 1996, p. 27) such as “Tell me about a bilingual lesson or unit that you thought worked well. What did you do and why was it successful? ” By getting teachers to talk about and evaluate actual lessons taught, it was my hope that their thinking about goals and approaches would be partially revealed. The follow-up interview questions were structured differently. The purpose of the interview was to get teachers to talk more about the specific unit of study that I had observed and comment specifically on aspects of language learning. For Teacher A, these types of questions were asked during multiple interviews often with her teaching partner present. For Teachers B through E, these types of questions were asked during the final follow-up interview. In addition to asking all of the teachers to reflect broadly on the unit, I asked teachers a series of questions about the activities which I observed, the materials that were used (and sometimes created by students), and social forms such as working as a whole class or with a partner. Before teachers could answer these questions, I described or listed, for example, activities which I had observed. Teachers then commented on their reasoning for choosing such activities as well as evaluating the learning from the activities. In this way, the follow-up interviews could very loosely be seen as a kind of “stimulated recall” (Dörnyei, 2007, p. 148). It was my intention that further discussing “concrete events and experiences” (Woods, 1996, p. 27) would reveal additional cognitions about teachers’ CLIL classroom practices. In addition to discussing observed events, I asked teachers about themes that 5.3 Research design 99 <?page no="100"?> emerged from my observations such as the role of challenge, choice, errors, and explicit English teaching. Some of these themes were evident in several teachers’ classrooms and I therefore asked the relevant teachers to speak about the theme. Finally, with the exception of Teacher A, all teachers responded to questions about the role of the German-speaking teacher in their CLIL lessons. Each interview lasted approximately 30-60 minutes. It was audio-recorded digitally. Following the interview, the audio file was uploaded to multiple storage devices including computer hard drives and external hard drives. Within 24 hours of completing the background interviews, I wrote an interview summary, usually between one and three pages in length. In the interview summaries, I recorded descriptions of the interview process, reflections on interviewing techniques, initial impressions and tentative analyses of the teachers’ responses, and possible questions to take note of in the classroom observations. For example, an excerpt from a description of the interview process with Teacher B reads as follows: We met in an unused classroom at the school. The interview lasted approximately 55 minutes. There were two quiet distractions along the way. One teacher came in to gather her items. One boy came in to gather his school supplies. (Oct. 1, 2019) An excerpt from a reflection on my interviewing techniques after interviewing Teacher D reads as follows: Allow teachers to answer fully before asking for clarification. Interrupting their train of thought to ask for clarification seems to lead to them not being able to return to what they wanted to say. Ultimately, the researcher is there to get the teacher to talk. (Oct. 9, 2019) These post-interview written summaries allowed me as the researcher to take a first step toward processing and organizing the content of the interview and my own role in it. All of the interviews were transcribed using the software f4transkript or MAXQDA. “Simple transcripts” (Dressing, Pehl, & Schmieder, 2015) that excluded non-verbal aspects of communication and included, for example, punctuation, were produced (see Appendix 1 for transcription conventions). Pavlenko (2007), writing about transcribing personal oral narratives, warns against such omissions and additions. Simple transcripts were, nonetheless, produced for several reasons. First, content is the focus of the analysis. I am interested in what teachers say and not how they say it. Of course, such an approach comes with potential risks. There is meaning, for example, in teachers’ paraverbal and non-verbal communication. However, qualitative content analysis, the analytical method used in this study, is designed to reduce data (Schreier, 2012). It is also possible for teachers to say something that they know to be untrue. Dresing et al. (2015, p. 23) provide an example: “A researcher asks a manger (sic) whether there will be any job cuts the following year. The manager takes ten seconds to think, rubs her chin, looks down and softly replies “Nope”.” It is clear in this example that what was said is untrue. I did not perceive any such instances in the interviews that I conducted. Second, some of the German interviews were transcribed by German native speakers or transcribed by me and edited by German native speakers. 100 5 Methods and Research Design <?page no="101"?> Because of a lack of resources required to train transcribers in more complex transcription conventions, simple transcription conventions, ones easily explained to transcribers, were used. Third, simple transcriptions that include, for example, punctuation make the teachers’ responses more readable. Finally, simple transcriptions are commonly used in the field of teacher cognition research (see Woods, 1996; Ruohotie-Lyhty, 2016; Falout & Murphey, 2018). Classroom observations were conducted in this study to record evidence of classroom interaction. All classroom observations took place over a two to four month period. This allowed me to observe the progression of one CLIL unit of study. Teachers usually taught three to four Science and Social Studies lessons per week. With some exceptions, I was able to observe one or two CLIL lessons from each teacher every week. Gaps in observations were caused by holidays and unexpected teacher absences. Following some of the observations, I was able to speak briefly with the teacher about the lesson. These conversations were often audio recorded but not transcribed and therefore not included in the interview corpus. Also, materials and photos of materials were collected as evidence of the types of activities students completed in the lessons. As an observer, I did not actively participate in lessons. Based on Borg’s (2015) description of observer roles, I was a non-participant in the classroom. I sat in the back or at the side of the classroom and quietly took notes. However, my presence in the classroom was overt (Borg, 2015) and as such led to several kinds of interactions. First, I was introduced or was asked to introduce myself in all of the classrooms. I told them who I was and that I was interested in finding out more about how their bilingual lessons work. In some cases there were student questions (e.g. “Can you [speak] German? ”). I briefly answered the questions and then the lesson continued. Second, during independent or group work phases of the lesson, I occasionally walked through the classroom in order to see what students were doing. If a student spoke to me or asked a question, I replied and moved on. Third, students occasionally approached me as I was observing and asked for help (e.g. closing the internet browser on the computer or asking for the meaning of a word). In each case, I offered the assistance. I did not disclose the specific research questions to the participating teachers. In this sense, my disclosure about the aims of my research can be considered minimal (Borg, 2015). I told teachers that I wanted to better understand teachers’ perspectives on teaching and learning in primary school CLIL and, more specifically, that I wanted to know what they do in their CLIL lessons and why they do it that way. I did not disclose, for example, that I was particularly interested in their thinking about language learning in the CLIL classroom, as I felt that doing so might influence their classroom behavior. Also, the questions asked during the background interview were intended to be broad (e.g. What would you want an outside observer to notice about bilingual lessons? or What does good bilingual primary school teaching look like in your view? ) and therefore did not reveal more specific aims of the research. Video-recording equipment was present for four out of the seven of Teacher A’s lessons. All of Teacher A’s lessons were also recorded using field notes. All of the other lessons observed (Teachers B through E) were recorded using field notes. 5.3 Research design 101 <?page no="102"?> The field notes that I took were structured openly (Borg, 2015). I did not use an observation scheme that, for example, sampled particular classroom events or events at a set time interval (Dörnyei, 2007, p. 180). Instead, I observed initially with a degree of uncertainty (Dörnyei, 2007, p. 179) until figuring out what events were most significant in relation to my research questions. My field notes allowed me to record classroom events as well as begin to reflect on and analyze those events. My notes consist of four types of information. First, I recorded descriptions of classroom actions and events. While it is not possible to record every detail with field notes, an attempt was made to record in sufficient detail events that show which languages are used, when, and by whom. For example, while observing Teacher E’s class, I noted one of the beginning routines used: Starts with planet song. Rap like. On CD. Children sing along. [Teacher E] sings along. Other events showed language use in a classroom management context. For example, from Teacher C’s classroom: [Teacher C] with [student] - discusses behavior with him in German. Still other events were about what students were doing: Girl with the only English book copies some sentences into a notebook. “While most planets spin like tops, pale blue Uranus spins on its side like a rolling ball.” These examples highlight different occasions for the use of German and English by classroom participants. Second, I recorded short samples of teacher talk and classroom discourse. While this study does not examine how discourse functions in the primary school CLIL context, I felt that samples of language, particularly with regard to how English and German are used by teachers and students, would be necessary for understanding the CLIL classroom. For example, from Teacher D: “Where have you been on holidays? ” Student starts in English. Asks to switch to German. “Ja, klar! ” from German-speaking teacher. Or from a student in Teacher D’s class: “Darf ich auf Klo gehen? ” Teacher D: “If you use that other language you can. Student: “May I go to the toilet? ” Other examples include code-switching and metalinguistic talk. For example, when eliciting descriptive language from students about popcorn (in the context of a unit on the five senses), most students in Teacher B’s class responded in German. Teacher B then translated the word into English and recorded the word on a small poster. When a student described popcorn as feeling like papier-mâché, Teacher B said: “Pappmaché. Oh, I don’t know the English word for that. I’ll write down the German word.” Or, from Teacher C: “Probably in Druckschrift is better.” These examples provide small insights into the daily negotiation of language use between teachers and students in the CLIL classroom. Third, I recorded my impressions of and reflections on what was happening in the classroom including my presence and potential influence on classroom events. These thoughts were recorded as an initial step in understanding the CLIL classroom. For example, in Teacher A’s classroom, I wrote: A calmness to the class - from students and teachers. And later: No pressure to speak English. These comments were recorded to note my interpretation of the classroom as a peaceful and casual place where student comfort and well-being were prioritized. Reflections on my presence were 102 5 Methods and Research Design <?page no="103"?> also recorded. For example, in Teacher C’s class I recorded: Is my presence too stressful? I left to allow her to deal with [behavior] issues. Maybe I should make explicit that she can kick me out [of the classroom] if there's a situation where I make her or [the] class uncomfortable. In this situation, I did speak with the teacher about this after the lesson. She assured me she felt comfortable and even stated that from her perspective, it was good that I see the class as it is. While not all of the impressions and reflections that I recorded related specifically to language learning in CLIL, they helped me to process my thinking about the complexity of CLIL teaching and learning from a more holistic perspective. Fourth, I recorded questions, both to guide my reflections on my observations and as a potential source of questions for the follow-up interview. For example, while observing in Teacher E’s class, I noted: How important is learning subject vocabulary in English? German? While observing Teacher D’s class I noted my impression of her natural enthusiasm in the classroom and wrote the question: Does this play a language learning role? Just personality? Again, these questions were recorded primarily as a tool for me to question my own initial thinking about what I was observing in the classroom. Within 24 hours of completing each observation, I wrote an observation summary. I recorded many of the same types of notes in the observation summaries as I did in my field notes, namely, events, discourse samples, impressions, and questions. Additionally, I recorded aspects of any post-lesson conversation that I may have had with the teacher. While I was able to use my field notes as a reference for writing the summaries, I relied primarily on memory. Writing the observation summaries without the moment to moment pressure of classroom observation allowed me to do two things. First, this allowed me the time necessary to organize and summarize my memories of classroom events, checking my field notes occasionally to support my memory. Sometimes, I summarized events as a numbered list. For example, from Teacher D’s class: 1. Starts by having a child move an arrow down on the list of what they will do and then read the directions. “Wir informieren uns an einer Lerntheke über Europa.” 2. Shows the “learning counter” and worksheets to the students. 3. Introduces an atlas 4. Plays two games/ routines. a. Cardinal directions are written on the board. Children repeat NOSW (nie ohne Seife waschen) phrase in German and then NESW (Never Eat Shredded Wheat) in English to remember the cardinal directions in both languages. b. Class puts together a map of Europe that is divided into groups of countries. Then students have to say a sentence in German or in English, e.g. Norway is west of Sweden or Norwegen liegt westlich von Schweden. This activity is timed. (Nov. 14, 2019) 5.3 Research design 103 <?page no="104"?> Second, writing observation summaries allowed me to further describe my impressions and tentative interpretations more clearly. For example, after observing Teacher A’s class, I wrote the following: One interpretation of English or the role of English in this class is that English is optional or an offering within the class. It’s available for those who are interested albeit in a limited form. No one needs to speak or understand a word of English to be successful in the class. However, the offering is there. Teacher A speaks it. A minority of materials are in English. But no one is forced. English is simply offered. (Oct. 4, 2019) A table summarizing the various aspects of my observation procedures according to Borg’s (2015) categories is included in Table 7 below. Participation Non-participant Awareness Overt Authenticity Real Disclosure Minimal Recording Field notes Structure Open Coding None Analysis Qualitative (non-formal) Scope Limited Table 7: Dimensions of observational research (Borg, 2015, p.-215) 5.3.4 Data analysis procedures The teacher interviews provide the primary data source for analysis in this study. They were analyzed using qualitative content analysis (QCA) (Schreier, 2012). Classroom observations and the resulting field notes were not analyzed using QCA but were instead used as a record of classroom practices which could be further explored with teachers in the follow-up interviews. Borg (2015) argues that observations are valuable for providing direct evidence of classroom events but that researchers can only make inferences about teachers’ cognitions based on observations. Such inferences need to be verified via other data sources such as interviews (Borg, 2015, pp. 289-290). The field notes that provided the record of classroom observations (including descriptions of classroom events and samples of classroom discourse as well as impressions and questions related to classroom events) were further described and critically reflected on in the post-observation summaries. The cumulative process of observing, describing, and reflecting on classroom interaction led me, the researcher, to make “inferences about [teachers’] cognitions” (Borg, 2015, p. 289). These 104 5 Methods and Research Design <?page no="105"?> inferences were further explored in follow-up interviews with teachers and the resulting interview transcripts became part of the overall transcript corpus that was analyzed using QCA in this study. In this way, classroom observations served to further investigate and describe teachers’ cognitions about their CLIL instruction. I have applied the method of QCA to analyze interview data in a way consistent with my understanding of the method and in a way that meets the needs of this particular research project. While QCA is “based on rules” (Kuckartz, 2019, p. 11), such rules “are often modified and adapted in practice by researchers” (Marvasti, 2019, p. 5). In an effort to avoid depending on an “optimistic over-estimation of the method’s well-known rigor” (Schreier et al., 2019, p. 10) to communicate my process of analysis, it is my intention to bolster the “methodological rigor” of my analysis through “transparency and explanation” (Kuckartz, 2019, p.-16). As described above in section 5.2.4, QCA is systematic in that all relevant materials are taken into account, steps are followed sequentially, and the coding frame is checked for reliability (Schreier, 2012, p. 9). As the first criterium of systematicity, I analyzed all teacher interviews, both pre-observation and post-observation, from all five teachers in their entirety. In the remainder of this section, I will describe the series of steps I made to analyze the data and how I made decisions within those steps. In sub-chapter 5.4, I will describe how I assessed and improved the quality of this research (validity and reliability). I applied a series of iterative steps to develop the coding frame and complete the analysis. Schreier (2012, p. 6) outlines eight steps in QCA, five of which she describes as “specific to QCA.” Those five steps in order are: 1) building a coding frame, 2) dividing your material into units of coding, 3) trying out your coding frame, 4) evaluating and modifying your coding frame, and 5) main analysis (Schreier, 2012, p. 6). My steps can be summarized as: 1) build a three category coding frame deductively based on the three research questions, 2) develop the coding frame by segmenting and tentatively coding one interview transcript, 3) evaluate and modify the coding frame, 4) repeat steps two and three for all interview transcripts, 5) apply the revised and modified coding frame to all segmented units of coding, 6) evaluate and modify the coding frame further as needed. In step one, I created a three-category coding frame deductively around the three research questions. The three main categories used in this study’s coding frame are “Teacher cognitions about CLIL,” “CLIL goals,” and “CLIL approaches.” In step two, I applied these three dimensions to one teacher’s transcripts. I chose the teacher with whom to start based on my subjective sense of the quality and clarity of her interview responses. As I read through the transcripts, I looked for evidence of the teacher’s thinking in relation to the three main categories (i.e. research questions). When I came to a statement that appeared to be evidence of one of the main categories, I segmented it as a unit of coding and assigned it a tentative code that represented its core meaning. In this way, the coding frame was built inductively. This process was supported and recorded using the data management software MAXQDA. MAXQDA software was used because it offered the digital tools necessary to segment data, create and define codes, code data, organize data segments, and subsequently review data segments. After completing the initial analysis of one teacher, I had a functioning, yet very provisional coding frame. 5.3 Research design 105 <?page no="106"?> In step three, I evaluated the coding frame and modified it based on my rereading and reconsideration of the teacher’s responses. Some tentative codes were renamed or further defined. Subcategories emerged and were added to the coding frame. Coded segments were arranged to fit into the emerging structure of subcategories. In step four, I applied this evolving coding frame to the other teachers’ interview transcripts. I continued to segment and code interview data according to the existing coding frame. New codes emerged and were defined based on the data. As I read and analyzed all of the remaining interviews, I continually evaluated and, when deemed necessary, modified the codes, code definitions, and the coding frame. At the end of step four, I had a complete yet provisional coding frame based on all of my interview data. In step five, I applied this complete yet provisional coding frame to all of the coded segments in my data again. While doing this, I closely read through all of the transcripts a second time. This allowed me to make minor modifications to some of the coded segments when deemed necessary. Ultimately, I read through the complete transcripts a minimum of two times and all of the segmented units of coding a minimum of three times. Each reading was an opportunity to better understand what the teachers were saying, the relevance of their statements, if any, to my research questions, how the teachers’ statements can be best represented within the coding frame, and the extent to which the coding frame is a logical and valid representation of the data. In step six, I once again evaluated and modified the coding frame. Because the coding frame is central to the findings of studies using QCA (Schreier, 2012, p. 219), I deemed it necessary to again evaluate the coding frame in order to ensure that it accurately represents teachers’ statements (validity). As is clear from the above descriptions, steps two through six were each filled with a variety of decisions that needed to be made about coded segments, codes, and the coding frame. I will describe how I made those decisions below. First, decisions were made about how to reduce data, a key feature of QCA (Schreier, 2012). For this project, teacher statements that provided evidence of their thinking about CLIL instruction were selected and coded. Statements unrelated to thinking about CLIL instruction were left unsegmented and uncoded. The coded segments used for analysis, therefore, represent only a part of what teachers said in the interviews. Data was also reduced by creating codes that are at a “higher level of abstraction” (Schreier, 2012, p. 7) than the details contained within the data. For example, in the extract below, Teacher C describes her thinking on an aspect of good bilingual teaching. Um, good bilingual teaching gets the children. Like, motivates them to learn the language by not actually hav[ing] English classes. By doing experiments. By trying out. (TC1 00: 11: 38) This extract was selected and coded as Playful and active under the categories of Activities, Motivate students, and CLIL approaches. Teacher C does not use the terms playful and active to describe motivating activities. Instead, she describes activities that involve “doing experiments” and “trying out.” Therefore, when interpreting and coding this segment as Playful and active, the details of “doing experiments” and “trying out” are lost 106 5 Methods and Research Design <?page no="107"?> and thus data is reduced. Both types of data reduction are integral to the process of QCA and occurred throughout the analysis of data in this study. Selecting units of coding from the interviews and assigning codes also required careful consideration. As described previously, the interview questions were constructed to elicit extended responses from teachers with the intention of getting teachers to reveal aspects of their thinking about CLIL instruction. Below, I describe examples of teachers’ responses that illustrate some of the types of decisions that were made throughout the data analysis. Teachers’ responses to my questions often followed one of several patterns. The first pattern can be summarized as “I do X.” In this type of response, the teacher simply describes what action she takes. For example, Teacher A made the following statement: “When I meet these children here outside, I talk English to them” (TA1 00: 32: 34). This is a direct statement of action. Because it does not directly express a cognition, e.g. a belief, an opinion, etc., this statement and statements like it were not segmented as units of coding and were not coded. The second pattern of response can be summarized as “I think (believe, know) X.” In this type of response, the teacher directly describes a cognition, sometimes by using words such as “I think” or “I believe.” For example, Teacher C made the following statement: “I think good bilingual teaching also means using what you use in good English classes. Is, I don't know, a lot of cliparts, a lot of visual stuff to get the children” (TC1 00: 11: 38). These types of statements were interpreted as statements of cognition. Statements such as this were segmented and coded. The majority of statements, however, fall between these two fairly clear types of responses. Some of the statements follow a descriptive pattern, e.g. “This is how I do X” while others follow an explanatory pattern, e.g. “This is why I do X.” Neither of these very broad patterns is a reliable indicator that what a teacher is saying is a cognition and not a description of action. Instead, each statement had to be evaluated independently in context of the question asked and what the teacher was saying directly before and after the interview segment. For example, at the beginning of the follow-up interview, Teacher E stated: Also wir haben ja damit dann abgeschlossen ähm dass die Kinder in Gruppen Referate gehalten haben zu einem Thema. Es waren nicht immer nur Planeten. Also die meisten haben über einen bestimmten Planet (…) gesprochen und manche Kinder noch über etwas anderes. Eine Gruppe über Exoplaneten zum Beispiel/ genau und da war jetzt schon mehr Deutsch. Also der Großteil war in der Regel auf Deutsch. (TE2 00: 01.01) This is a description of what happened at the end of the unit about the solar system. In it, the teacher is simply recalling events. Although the teacher describes children using mostly German in their final presentations, she does not indicate her opinion or belief about using or prioritizing German in CLIL. Therefore, statements such as this were not segmented and not coded. 5.3 Research design 107 <?page no="108"?> However, other statements that also seemed to be merely descriptions required closer evaluation. For example, Teacher B describes her understanding of the difference between EFL and CLIL lessons. She states: Because until now they maybe presented a monster they drew or it’s just not too much but then it would be like, Where do the animals live? What do they eat? What do they look like? How many babies can they get? How old can they be? It’s more different, yeah, themes that are covered. (TB2 00: 41: 44) As a subjective description of CLIL, a statement like this could be considered a cognition in that it expresses an aspect of one particular teacher’s understanding of CLIL. Whether this understanding is based solely on professional experience or has other origins is beyond the scope of this study. Looking at the context in which this response was given provides additional information. For example, the question that initiated this response was, “How do you see students’ language competences developing within CLIL? ” Given this particular question, the response seems to indicate that this teacher sees CLIL as an approach to language learning that exposes children to language that goes beyond traditional EFL contexts. As such, this particular segment was coded as English beyond EFL under the subcategory of Foreign language learning goals and the main category of CLIL goals. Yet another example demonstrates the need for careful contextual evaluation of teachers’ statements. Below, Teacher D describes her use of picture books in CLIL lessons. Also ich mache unheimlich gerne Storytelling. Deswegen versuche ich zu jedem Thema im Sachunterricht jetzt auch eine Geschichte zu finden, die/ die das Wortmaterial aufgreift oder die/ und wenn es selbst/ wenn es eine Geschichte ist, die jetzt kein naturwissenschaftlicher Aspekt hat, sondern einfach nur das Wortmaterial beinhaltet, dann ist das einfach zum Einstieg ganz nett. (TD1 00: 25: 53) Such an extract presents multiple factors when considering whether or not to code it. For example, it seems clear that personal interest is a primary motivating factor for choosing the use of picture books in CLIL Science and Social Studies lessons. She goes on to suggest or explicitly state that subject vocabulary and science content are two criteria that she uses for choosing an appropriate story. Based on this, it seems reasonable that learning or, at minimum, being exposed to subject vocabulary is a goal for this teacher’s CLIL lessons. Going beyond this interpretation to suggest that Teacher D prioritizes learning English vocabulary over content learning because she is willing to choose a book with relevant vocabulary but void of the relevant science content would be misleading at best and likely wrong considering Teacher D’s clear statements to the contrary in the rest of the interviews. Therefore, in this case, this segment was coded as Developing linguistic means under the subcategory of Foreign language learning goals under the main category of CLIL goals. Another source of constant consideration and decision-making was the structure of the coding frame. When developing the coding frame, initial codes were assigned to units of coding and these codes were sorted into one of the three main categories. Because the coding frame was built inductively from the data, initial codes were detailed and remained close to the data. As the analysis evolved, codes were often revised to describe more abstract levels 108 5 Methods and Research Design <?page no="109"?> of meaning. For example, in an early version of the coding frame, the following extract from Teacher D was coded as Body language under the main category of CLIL approaches. Aber meistens ist es mit zeigen, hochhalten, „packen das in die Postmappe“, morgen, also ganz viel Körpersprache, ganz viel ähm, ja, ganz viel mit Hand und Fuß erklären um sich verständlich zu machen. (TD1 00: 05.25) The teacher uses the term “Körpersprache” (body language) in her response. Other teachers also spoke of using gestures, facial expressions, and other kinds of physical movement but also spoke of the use of pictures and objects to communicate meaning. It became clear that teachers were describing their belief in the use of visual scaffolding strategies. Therefore, I eventually subsumed Body language under the more abstract category Visual strategies in order to include teachers’ statements about other forms of visual scaffolding. As is evident from the above example, it became clear over time that lower level categories were describing discrete aspects of higher level categories. Visual strategies was not just one of many different approaches that teachers believed were appropriate for the CLIL classroom, they were a specific type of input scaffolding strategy. In CLIL, scaffolding is often used to describe techniques for supporting L2 input comprehension and L2 output production as well as content comprehension (Massler & Ioannou-Georgiou, 2010). With the realization that particular approaches described by teachers were aspects of broader conceptual categories, broader categories were added to the coding frame to organize codes and provide additional meaning. This resulted, for example, with the subcategory of Scaffolding input having five subcategories of its own including Translation, Visual strategies, Language strategies in L2, Material strategies, and Interactive strategies. Each of these five subcategories describes a type or dimension of input scaffolding. In the example of scaffolding, it is fairly clear that the use of visualization serves the purpose of supporting either content or language understanding. This was also expressed by the teachers. Other categories, however, were less clear. Especially in the main category of CLIL approaches, it was tempting to interpret a lower level code as belonging to a higher level subcategory even when no evidence for such a connection existed. For example, one of the subcategories directly under the main category of CLIL approaches is Motivate students. Motivate students has three of its own subcategories, each describing one way that teachers believe they can motivate students in the CLIL classroom. One of those subcategories is Social support. Without looking at what teachers actually said, it is possible to imagine Social support as a dimension of Scaffolding. Indeed, under the category of Scaffolding and Scaffolding input is the category of Interactive strategies which might easily be conflated with Social support. The difference between these two categories is, of course, the reasoning that teachers give for interaction or working with others. In the extract below, Teacher C describes student motivation as her reason for having students work together. I like lessons that are designed like that where we all work on different things and then we put one thing together. And if that also means they can present something, that's even better. So having as a group, that is my class, having something that we all worked on I think is ehm motivating for 5.3 Research design 109 <?page no="110"?> the kids and this is something we've done before and I just like to choose that way of learning something. (TC2 00: 19: 35) In the next extract, Teacher B seems to indicate that having students work together serves the role of helping them understand and complete tasks successfully. Of course they could do the listening alone. But then sometimes it also helps when you have a partner just to talk about it. "I think it's that. No it's that." So. (TB2 00: 16: 58) For this reason, two teachers’ similar practices may be assigned different codes and sorted into different subcategories based on their reasoning for their practices. 5.4 Research quality In an effort to be transparent or provide an “audit trail” (Lincoln & Guba, 1982, p. 248) on my data analysis procedures, I have described in the above section how I made decisions related to selecting and coding material and have given examples of some of those decisions. In the following section, I will describe more formal criteria for evaluating the quality of qualitative research and describe what I have done to meet these criteria. As a qualitative method of data analysis with its roots in quantitative data analysis (Schreier, 2012) QCA suffers in at least two ways from a certain kind of schizophrenia among data analysis methods. First, Roller (2019) states that the “participant-researcher relationship” and the “researcher-as-instrument” attributes of qualitative research differ in QCA. In QCA, the researcher creates data in the form of codes based on some previously gathered text (e.g. an interview transcript). In this way, the researcher is a step removed from research participants (e.g. teachers) and is therefore an “instrument” for creating codes, not for gathering interview data. From a research quality perspective, Roller (2019) concludes that this removal from the process of gathering a text (e.g. interviewing and then transcribing the interview) means that any researcher-participant bias is embedded in the content to be coded and is therefore not relevant to qualitative content analysis itself. Second, quantitative measures of quality such as internal and external validity, reliability, and objectivity have been criticized by some qualitative researchers while constructs such as credibility, transferability, dependability, and confirmability have been suggested to replace them (see, for example, Lincoln & Guba, 1985; Huttunen & Kakkori, 2020). However, as Barbour (2014, p. 498) notes, “In practice, […] it has proved harder to pin down and operationalise such standards and definition remains essentially in the eye of the beholder.” The subjectivity of such constructivist terms as well as QCA’s roots in quantitative content analysis has led Schreier (2012) to maintain that positivist constructs such as reliability and validity are more appropriate as measures of quality in QCA. Therefore, the constructs of validity and reliability will be used to describe the quality of research in this study. 5.4.1 Validity Schreier (2012) states that for data-driven coding frames such as the one produced in this study, it is “face validity” that is most relevant. Face validity is “the extent to which [the] 110 5 Methods and Research Design <?page no="111"?> instrument gives the impression of measuring what it is supposed to measure” (Schreier, 2012, p. 185). In the case of a data-driven coding frame, face validity describes the extent to which the main categories and multiple levels of subcategories describe aspects investigated through the research questions. In the case of this study, face validity is the extent to which the coding frame describes dimensions of teachers’ cognitions in relation to the three research questions. For example, aspects of this coding frame are more or less valid based on the extent to which subcategories such as Speaking or Develop linguistic means relate to or answer the research question, What language-related educational goals do teachers work toward? While face validity is, perhaps, an obvious concept, Schreier (2012) offers several signs that, if present, may suggest low face validity. First, if a residual subcategory, one in which units of coding are placed that do not fit in one of the other subcategories, is used excessively, this is very likely a sign of low face validity. No residual subcategories are used in this coding frame. Second, if one subcategory is used significantly more than other subcategories, this can be a sign that the subcategory is not “sufficiently differentiated” (Schreier, 2012, p. 187). However, as Schreier notes, this alone does not make certain that the subcategory is of low validity. Instead, it could be a significant finding of the research indicating preference for one category over another. In the case of this study, there are clear instances where one subcategory has significantly more coded segments than other subcategories. For example, under the subcategory of Foreign language learning goals, the subcategory of Develop linguistic means has 25 units of coding assigned to it while the subcategory of Comprehension strategies has only five coded segments assigned to it. This difference, however, simply reflects that teachers more often spoke of goals related to learning vocabulary than they did about developing comprehension strategies and therefore, as a group, appear to value the development of linguistic means (such as vocabulary) more than the development of comprehension strategies. This is a significant finding of this research and not an indication of low face validity. Third, Schreier (2012) warns against categories that are too abstract and therefore do not describe the content of the data in sufficient detail. In the context of QCA, this is a question of balance. As already described, one way that QCA reduces data is through subcategories that are at a higher level of abstraction than what is said in the data itself. This allows researchers to organize slightly different statements in, for example, interview data into meaningful categories. Taken too far, however, this creates categories that are meaningless. For example, the subcategory of Scaffolding in this coding frame has been further divided into Scaffolding input and Scaffolding output. Each of these subcategories has been further subdivided and some of those subcategories have also been subdivided. All of these subcategories serve the purpose of describing each superseding category in further detail. By describing the data in further detail, these subcategories increase the face validity of the coding frame. 5.4 Research quality 111 <?page no="112"?> 5.4.2 Reliability Reliability is the second measure of quality used in this study. As a test of research quality, reliability tends to be described and measured differently depending on the type of research (qualitative or quantitative) and the specific research projects (see Dörnyei, 2007; Brown & Rodgers, 2002). Schreier (2012, p. 166) states that within QCA, a coding frame is deemed reliable “to the extent that it yields data that is free of error.” In QCA, reliability can be assessed in two ways: across coders and across time (Schreier, 2012). By assessing the coding frame in one or both of these ways, the researcher is establishing the degree of internal reliability or the extent to which the coding frame consistently produces the same results for the data that has been created (Schreier, 2012). Reliability in this study has been evaluated across time and, to a more limited extent, across coders. Because this study has been carried out by a single researcher working alone, there was not access to other researchers with direct knowledge of the project. Therefore, the primary method of evaluating reliability was for a single coder (the researcher) to code the data multiple times over the course of months. Coding data multiple times over the course of months served as a kind of “time triangulation” (Canh and Maley, 2012, p. 98) applied to the data analysis process. As already described above, analyzing the data and building the coding frame was done simultaneously through a process of iteration. This process began formally at the end of January, 2020 and was largely complete by the end of August 2020 with still remaining revisions to the coding frame taking place as late as June, 2021. This roughly 18-month period allowed ample time to segment data, to code data and revise those codes, and develop the coding frame. In addition to developing the coding frame and assessing reliability over time, four subcategories of the coding frame were independently coded by a second coder to further assess reliability. Those subcategories were 1) Cognitions about the role of content, L1, L2 under the main category Teacher cognitions about CLIL, 2) General pedagogic goals that support language learning under the main category CLIL goals, 3) Scaffolding under the main category CLIL approaches, and 4) Motivate students under the main category CLIL approaches. These subcategories were chosen because of the difficulty establishing discrepancy between various categories within the subcategory. For example, when coding segments of data under the subcategory Cognitions about the role of content, L1, L2, it was sometimes difficult to decide whether a teacher statement emphasized Content learning as priority or if the focus was on another category such as German subject language learning as necessary. It is possible, of course, that such difficulty deciding what category a particular statement belongs to is an indication that the categories are not sufficiently defined and discrete or that the phenomenon being described in a teacher statement is not accurately represented by the code or category. In this particular case, it is also possible that macro-level priorities or goals such as content learning and language learning are always intertwined or “cannot be separated from one another” (Zydatiß, 2012, p. 71 my translation) and that teachers’ thinking reflects this. In this study, the second coder was first given a set of interview segments that had already been determined to belong to the subcategory. The coder did not know how they had been categorized within that subcategory. Her task was to read the segments and determine to which code each segment belongs. If she felt that none of the codes accurately described 112 5 Methods and Research Design <?page no="113"?> the segment, she was to code it as miscellaneous. To be clear, the coder did not segment the data herself, and the coder did not need to read and analyze interview segments from the entire data set. She assigned a limited number of preselected codes to a limited number of preselected interview segments. The second coder was given a section of the coding frame and a document containing all of the interview segments in a MAXQDA file. She coded the data independently. We then reviewed her coding together and discussed cases of disagreement. The discussions were not an attempt to persuade the other. Instead, they were an opportunity to describe reasoning and, as Barbour (2014, p. 501) states in a discussion on inter-rater reliability, to engage in a “continuous process of re-evaluation and revision through the duration of a research project.” I took notes on our discussions in MAXQDA. The discussions always led to a reexamination of teachers’ statements and sometimes to a change of coding. Two categories were also renamed based on these discussions and the reexamination of teachers’ statements. The following three examples give some insight into the discussions between the second coder and me as well as to the process of revising the coding frame and/ or revising my interpretation of teachers’ statements. In the first example, Teacher B made the following statement: And then the big question is always, What is possible in English and what has to be done in German? So that's always the thing that's/ like sometimes I wonder, Could I do more in English? Or is it too much for the children? (TB1 00: 27: 00) I coded this segment as Complimentary “languages learning” (later renamed). In a memo about this segment written later in the analysis, I wrote the following: Many of the teachers ask or present this question as central to their thinking about their planning for the classroom. I think they look at it in two ways. One is through a German curriculum mindset - What do the kids need to know in German? And therefore what do I have to teach in German? The other way they think about the question is through a German-as-scaffold perspective. What can the kids understand in English? With the idea that we'll do it all in English if they can handle the language demands. The second coder coded Teacher B’s statement as Content learning as priority. In a memo about this segment, she wrote: ‘Content learning’ if the teacher was thinking about what they could (not) do in English and still make sure the children get the content. But it could also be 'German as priority' [later renamed], if they meant that children just have to know certain things in German, rather than in English, even though they would get the content either way. It might also be 'English as enrichment', as there is this idea of 'they have to get this, but can I do more English.' From a methodological and research quality perspective, it should be noted that there was no agreement between the proposed codes from the two coders. I proposed one code and coder two proposed three. This lack of agreement, however, was less significant than the process of rethinking or reinterpreting exactly what teachers might be saying in various statements. 5.4 Research quality 113 <?page no="114"?> Based on the written memos and the discussion between the coders, a number of interpretations of this statement are possible. The statement from Teacher B could be interpreted as follows: Learning English is this teacher’s first priority. The teacher asks, “What is possible in English and what has to be done in German? ” This teacher sees German as only being used to make up for a lack of English when she states, “what has to be done in German? ” Further evidence for this interpretation is that the teacher asks, “Could I do more in English? ”, implying that if possible, she would maximize English use until all aspects of the CLIL lessons were in English. Another, opposite, interpretation is that learning German is this teacher’s first priority. She states, “what has to be done in German? ” This suggests that using and learning German are non-negotiable in certain contexts and that regardless of English learning goals or students’ English competences, German must be used in these contexts. Yet another interpretation is that content learning is prioritized above all else. Evidence for this interpretation also lies in the sentence, “What is possible in English and what has to be done in German? ” It suggests that regardless of language, content must be learned. Some of the content “is possible in English” because it is easy enough to teach and learn in English. Other content “has to be done in German” because it is too difficult to teach and learn in English. A final interpretation is that English and German are used for different curricular purposes. It is understood that learning content, learning English, and learning German are all, often related, CLIL goals. Instead of asking which of these goals is prioritized, the more salient questions are: What is English for? and What is German for? The teacher asks, “What is possible in English and what has to be done in German? ” This suggests that students’ English competences may be a limiting factor to its use but also that content difficulty as well as the content topic or theme may play a role. Indeed, at other moments in the interviews with Teacher B, she describes the use of German as being more appropriate to teaching certain topics such as sexual education due to the personal nature of the topic. I eventually settled on this interpretation of Teacher B’s statement and coded the statement as German and English use for different purposes. The above example demonstrates not only the various possible interpretations of a particular teacher statement but also the need to contextualize any one statement with what was said at other moments in the interviews as well as what was observed during lessons. While the second coder did not have access to the full context of teachers’ statements, she was able, at different times, to confirm my interpretations or to provide alternative interpretations which required me to reconsider my initial interpretation and reasoning. In the second example, Teacher A made the following statement: Und dieses deutsch-englisch ist zumindest für Sachunterrichts Themen sehr wichtig, da die Kinder nicht nach Hause gehen sollen und alles auf Englisch über Wikinger wissen. Es ist genauso wichtig, dass die das auch auf Deutsch wissen und deswegen ist das gerade im Sachunterricht finde ich es gut, dass es auf Deutsch Englisch läuft. (TA1 00: 14: 15) I originally coded this segment as German (language) learning as priority. From my perspective, such explicit focus on using and learning German in a CLIL context was akin 114 5 Methods and Research Design <?page no="115"?> to prioritizing it. My interpretation of the above statement as prioritizing learning German was possibly influenced by contextual and personal factors. First, the German-speaking teacher in this class often led the CLIL lessons which resulted in very little English being used during some of the lessons. This context suggested an overall prioritization of using and learning German. Second, factors unique to me such as being a native English speaker and having teaching experience internationally and in Germany as well as with children using English as a foreign language led me to make judgments about the teaching and learning that I was observing, namely, that more English could be used and more English should be used. Of course, it is not my place to judge teachers’ thinking and practices but to try to understand them and describe them. The second coder (a German native speaker) coded it as Complementary “languages learning.” In our discussion, the second coder explained that she understood the emphasis not to be on learning or prioritizing German above learning English but instead to emphasize a bilingual priority or that learning English alone was not enough. After reflecting on my own biases and reexamining Teacher A’s statement, I re-coded the segment as Complimentary “languages learning.” The previous two examples come from the subcategory Cognitions about the role of content, L1, L2. This subcategory proved to be the most challenging from the perspective of inter-coder reliability. Based on discussions and a reexamination of teachers’ statements, some of the subcategories within this subcategory were revised. For example, German (language) learning as priority became German subject language learning as necessary to reflect the idea that learning German was not necessarily a priority for teachers but was deemed necessary for students. Also, the code Complimentary “languages learning” was renamed German and English use for different purposes to describe more precisely my interpretation of what teachers were actually saying. In the third example, Teacher B made the following statement: And having motivated children that don't like/ once in a while, I haven't really, well, I don't know if there's really a kid I ever had that said, "No, it's English, I'm not paying attention." No, that never really happened. It's not that they like shut off as soon as I come to the classroom because they know it's English and are not understanding anything. So, motivated children. (TB1 00: 26: 05) I coded this statement as Motivated language learners under General pedagogic goals that support language learning. The second coder coded it as Miscellaneous. She reasoned that because the teacher describes “having motivated children” or children that are already motivated when they come to her class, this cannot be a goal. This case exemplifies the need for further context when interpreting some teacher statements. In this example, the teacher was responding to a question asking her what she hoped an outsider visiting her classroom would notice. Interpreting the statement “having motivated children” very narrowly and suggesting that “motivated children” is not a goal is likely not the teacher’s intention. Instead, “motivated children” is likely a goal that Teacher B works toward since she would want others to notice that. I therefore maintained my original interpretation and coded the statement as Motivated language learners. 5.4 Research quality 115 <?page no="116"?> Similar discussions and considerations took place when there was disagreement between the two coders. After the discussions, I reexamined teachers’ statements and made revisions to the coding frame when deemed appropriate. I then calculated a basic percentage of agreement. The percentage agreement was calculated from my (the original coding) perspective. If the second coder agreed with all of the original codes but assigned additional segments to the code, this was still calculated at 100 % agreement. These additional coded segments were not counted and not included in the table. Therefore, the numbers in the Coder 2 column are always equal to or less than the numbers in the Coder 1 column. The additional coded segments would be taken into account as disagreement in other codes. The final absolute numbers and percentages of agreement for the subcategories coded by two coders are presented in Appendix 5. The percentages of agreement for individual subcategories and combined subcategories are generally high. However, their significance as an indicator of reliability for this study is limited. First, only about half (34 out of 70) of the individual categories were coded by the second coder. Second, the second coder was not segmenting raw text and developing codes for these segments. Instead, she was assigning existing and pre-selected codes to pre-selected text segments. This preselection process limited the second coder's choices for assigning meaning to teachers’ statements. The most challenging subcategory to code and hence the one that benefited most from the examination of a second coder was Cognitions about the role of content, L1, L2. Of the six subcategories, the first four listed in Table 8 had some disagreement and were discussed. Table 8 below summarizes the original coding and the revised coding of those four subcategories. Cognitions about the role of con‐ tent, L1, L2 - Coder 1 (researcher) Coder 2 Percentage agreement Content learning as priority original 29 29 100% final 33 33 100% German learning as priority (Revised label) German subject lan‐ guage learning as necessary original 16 12 75% final 10 10 100% English learning as curricular enrich‐ ment and choice original 15 12 80% final 15 15 100% Complimentary “languages learning” (Revised label) German and English use for different purposes original 16 4 25% final 16 14 88% Table 8: Intercoder agreement for the subcategory Cognitions about the role of content, L1, L2 The lower percentages of agreement in the original coding, especially for the subcategory German and English use for different purposes reflects some of the inherent tension that exists between these four subcategories. Key to this tension is the extent to which the goals of content learning and language learning (either L1 or L2) can be differentiated in teachers’ cognitions and teachers’ practices. Teacher B’s statement below exemplifies this. 116 5 Methods and Research Design <?page no="117"?> And then how could/ well think about before you start really planning the lessons, which content could be taught in English and which has to be taught in German. (TB1 00: 29: 07) I (Coder 1) originally coded this segment as Complimentary “languages learning” inter‐ preting it to mean that German and English would be used to teach different aspects of the curriculum based on what the teacher thought “could be taught in English” and what “has to be taught in German.” Coder 2 originally coded this as Content learning as priority. In a memo attached to this coded segment, she describes alternative ways to code the segment based on how one interprets it. Alternatively: Complementary 'languages learning': Some things have to be done in German, others MAY be done in English (Might also be English learning as enrichment, because of 'could': After the necessary facts have been learnt in German, English is an add on. (Coder 2) Coder 2’s memo highlights the reality that these subcategories are interdependent. To prioritize content learning in a CLIL context in which the school language is German suggests that learning English takes a secondary role of enrichment. The alternative codes that the second coder suggests are legitimate interpretations of Teacher B’s statement. As noted earlier regarding another statement of Teacher B, it is not clear from the statement itself exactly why particular content “has to be taught in German.” What can be concluded from the statement is that this teacher feels there is a unique curricular role for English and for German. Teacher C’s statement below also describes the use of English and German for different curricular purposes. Well when we sat down and decided, one major factor is what we think is doable and what we think is too difficult, to be honest with you. It's not that we decide ehm, "Ok, the kids have to know everything in German and on top, like the icing on the cake we're going to make sure that they know some of the things in English. (TC2 00: 31: 23) Again, I coded this statement as Complimentary “languages learning” interpreting the emphasis to be on the various roles of English and German. Coder 2 coded this originally as German learning as priority but through our discussion also came to agree with coding it as Complimentary “languages learning.” While Teacher C’s statement also suggests prioritizing content learning, she clearly differentiates herself from teachers who described English learning as a curricular enrichment. For Teacher C, the use of English and German are complimentary or can be used for different purposes in the ultimate goal of learning content. These teacher statements and the process of interpreting their meanings highlight the ways in which the categories are at times interdependent and therefore lack a degree of discrepancy. Subcategories that are not “mutually exclusive” (Schreier, 2012, p. 92) are problematic for a coding frame. In this case, however, it is the nature of the phenomenon under investigation (content learning and language learning) that itself defies clear boundaries. Through the process of two coders coding this subcategory and the resulting discussions, three actions were taken. First, Complimentary “languages learning” was changed to 5.4 Research quality 117 <?page no="118"?> German and English use for different purposes. This more explicitly describes teachers’ statements while not describing exactly what those purposes are. Second, German learning as priority was changed to German subject language learning as necessary. This revision focuses on whether or not teachers think learning German subject language is necessary without necessarily assigning it as a priority. Third, with additional clarification of these subcategories, some, but not all, segments were re-coded either by Coder 1 or by Coder 2. 5.5 Summary In this chapter, I have described my research process in detail. I have described the aims of the research as well as the theoretical foundations of the data gathering methods of interviews and classroom observations and the data analysis method of QCA. With this background established, I then described the research design and setting as well as the specific steps I took to gather, record, and analyze data. Then I addressed issues of research quality, namely validity and reliability and described steps taken to improve the reliability of the analysis. This study has been conducted in order to better describe and understand CLIL teachers’ cognitions across three broad research questions: 1. What cognitions do teachers hold about teaching and learning in primary school CLIL? 2. What language-related educational goals do teachers work toward? 3. What approaches do teachers think are most effective for achieving these goals? In order to answer these questions, teachers’ cognitions were investigated using semi-struc‐ tured interviews. I interviewed teachers before and after a period of classroom observations. While questions were prepared in English, teachers responded in German, English, or both. The background interviews covered questions broad in scope. I wrote background interview summaries after completing each interview in which I described how the interview pro‐ ceeded as well as initial impressions and further questions that I had. Follow-up interview questions were based on what I had observed in the classrooms. All interviews were transcribed using simple transcription conventions. German native-speakers transcribed or edited the German transcriptions. Classroom observations were conducted in order to contextualize teachers’ stated cognitions and to better understand their CLIL practices. These observations were recorded using field notes and further described with observation summaries that were written after the observations took place. Additionally, classroom materials were collected and brief conversations were occasionally held with teachers after observing their lessons. The teacher interviews were analyzed using the methodology of qualitative content analysis (Schreier, 2012). The primary result of the analysis is the coding frame built inductively around the three research questions. I have addressed quality in this research project using the constructs of validity and reliability and have described my reasoning for this. Additionally, I have described my data analysis procedures in an attempt to leave an “audit trail” (Lincoln & Guba, 1982, p. 248) which readers can follow to determine for themselves the quality of this investigation. It is my intention that by describing what I did and why I did it that way, I have “demonstrate[d] 118 5 Methods and Research Design <?page no="119"?> an educated awareness of the consequences of particular methodological decisions” (Seale, 1999, p.-33) that I have taken in this study. 5.5 Summary 119 <?page no="121"?> 6 Results The coding frame, as the central result or outcome of this study, is presented in a series of tables below, each corresponding to one of the subcategories in the complete coding frame. Each labeled subcategory (e.g. Content learning as priority) summarizes a dimension of the superseding category (e.g. Cognitions about the role of content, L1, and L2). These categories were developed inductively based on the interview data. The tables summarize the data in four columns. In the first column, the code name is given. The name of the code is intended to reflect the central idea of the code. In the second column, an exemplary quote or coded segment is given. The segment is intended to represent a common way in which teachers spoke about the idea represented by the code. In the final two columns, the total number of coded segments and the total number of teachers who made at least one statement coded for that particular subcategory is listed. These two numbers are a form of “quasi-statistics” (Becker, 1970, p. 81) and are intended to provide a sense of the strength or internal generalizability of each subcategory (Maxwell and Chmiel, 2013, p. 545). Definitions for each code are included in the code book (Appendix 6). While the quasi-statistics reported in the coding frame provide some indication of the internal generalizability of each subcategory (i.e. the strength of a category as indicated by the number of teachers and number of coded segments that express aspects of the category), they should be interpreted with some caution. The categories presented in the coding frame emerged from the interviews and were directly influenced by specific questions. A question such as What do you want your students to learn in your CLIL lessons? allows for a variety of responses that may inform several of the categories listed. Alternatively, a question such as What kinds of students benefit from CLIL education? would generally lead to responses that inform our understanding of teachers' cognitions about student characteristics. Therefore, the generalizability of the category as indicated by the total number of coded segments (listed in the column second from the right) is a reflection, in part, of interviewer and interviewee interest. This is one way in which the interviews have been “jointly constructed” (Richards, 2009, p.-190 italics in original) by interview participants. Following the data presented in the tables, I describe and give examples of how teachers described their thinking in the context of the various subcategories. In my descriptions, I have summarized teachers’ thinking as described in the interviews and have described some degree of nuance to the ways in which teachers talked about their thinking. These descriptions are rooted in and add context to the coding frame. Additionally, I include excerpts from my field notes and short descriptions of my classroom observations in order to further contextualize teachers’ thinking. <?page no="122"?> 6.1 Research question 1: What cognitions do teachers hold about teaching and learning in primary school CLIL? The answers to RQ1 can be divided into five categories which are used to outline section 6.1. These categories reflect teachers’ cognitions about the roles of content, L1, and L2 learning in CLIL, second language acquisition in CLIL, choosing CLIL topics and activities, teacher competences and demands in CLIL, and CLIL student characteristics. The findings for each section are discussed in detail below. One of the initial motivations for conducting this study was to better understand how CLIL teachers think about the various and often competing goals of CLIL. At a minimum, CLIL is a pedagogical approach that aims to improve students’ foreign language competences while maintaining content and L1 learning that is commensurate with non-CLIL or regular students. Understanding how teachers prioritize these learning goals and what they perceive the roles of languages to be in the CLIL classroom is central to this study. 6.1.1 Cognitions about the role of content, L1, and L2 Code Exemplary quote Coded segments Teachers Content learning as priority Since the focus in Sachunterricht of course is not the language but the content, I wouldn't stop them to say it in German because, yeah. (TB1 00: 42: 18) 33 5 German sub‐ ject language learning as necessary Also ähm noch eine deutschsprechende Lehrkraft und das ist in Sachunterricht auch wirklich wichtig und gut, weil da ja nochmal ähm, ja zum einen ganz viele Fachbegriffe vorkommt, die teilweise wirklich schwierig sind, die die Kinder auch auf Deutsch können müssen. (TE1 00: 03: 58) 11 5 German and English use for different purposes Also, in meinem Mittelpunkt, was kann die Englisch/ was kann auch den englischen Sprache geleistet werden und was muss von der deutschen Sprache aufgefangen werden. […] Und so geht man ja auch immer die Planung, dass man von der englischen Sprache ein bisschen ausgeht und das Deutsche so ein bisschen drum herum baut. Genau. (TA5 00: 01: 36) 16 4 English learning as curricular en‐ richment and choice Und das, was ich mache ist eigentlich on top und ähm, wer damit gut umgehen kann, der nimmt das auf und kann das vielleicht sogar auch selbst wiedergeben. Und ähm, wer damit noch nicht so gut umgehen kann, der konzentriert sich halt auf das, was da auf Deutsch auf dem Arbeitsblatt geschrieben steht. (TD1 00: 34: 56) 15 4 122 6 Results <?page no="123"?> English use as barrier to re‐ lationships So, sometimes it's more difficult to build up a relationship to them by jokes or something because they wouldn't understand it in the beginning. So that's sometimes like (inaudible). They/ it's not that they would talk to the German class/ German teacher more than to me, but like building a relationship with them is more difficult sometimes in the English language. (TB1 00: 10: 45) 3 2 English use as a classroom lingua franca I also learned that when you have a boy or a girl from a foreign country, not necessarily US, Canada, wherever, but, I don't know. I have this boy from the Philippines and I learned that English sometimes is the way to I don't know, connect↑ Um, I was aware of that before obviously, but it's definitely nice to see that when you speak English with somebody it definitely is the key. (TC1 00: 16: 12) 2 2 Table 9: Cognitions about the role of content, L1, L2 Teachers in this study expressed quite clearly their belief that content learning is a top priority. Some of them expressed this understanding explicitly. Teacher B describes students learning Science and Social Studies content as an indicator of successful teaching. Um, well, if I think about just the Science and Social Studies classes, I can see on the one hand that they get the meaning of the topic↑ They have learned what they are supposed to be learning. (TB1 00: 29: 13) It is clear that for Teacher B the Science and Social Studies content is “what they are supposed to be learning.” But teachers also expressed their prioritization of content learning in other, sometimes implicit, ways. One of the more common ways in which they expressed this belief is through statements related to creating and choosing materials, specifically worksheets. Most of the worksheets that teachers had readily available to them were written in German. English-language materials, whether self-created, professionally pub‐ lished, or translations of German-language materials, were rarer. While German-language materials were easier to come by, teachers made statements indicating that they would use German-language materials in situations where they feel that the content is challenging or of particular importance. Likewise, if teachers feel that the use of English obscures the content too much, they are more likely to use German-language materials. Below, Teacher D describes how materials, specifically worksheets, need to serve the purpose of learning content. Natürlich sollen sie erstmal den fachlichen Inhalt lernen. Also das, was auch die nicht bilingualen Kinder lernen, steht an erster Stelle. […] Also die Kinder sollen eigentlich was über die Säugetiere hinterher mitnehmen und brauchen 6.1 What cognitions do teachers hold about primary school CLIL? 123 <?page no="124"?> jetzt nicht zwingend das spezifische englische Vokabular zu ihrem/ zu ihrem Säugetier immer parat haben. (TD1 00: 34: 56) Worksheets were a primary way that teachers described ensuring content learning. But teachers also talked about how they choose materials generally. In doing so, teachers made clear that their choice of materials was often driven by content learning goals. Teachers first want to make sure that the materials would support content learning and then decide which language to use. Teacher E’s statement highlights the importance of content learning in her thinking. Also bei der Materialauswahl (laughs) versuche ich darauf zu achten, dass ähm unterschiedliche Sachen dabei sind, dann für unterschiedliche Vorlieben, unterschiedliche Lerntypen und so, dass die Inhalte noch einmal auf unterschiedliche Arten gefestigt und erlernt werden können. Ja. (TE2 00: 17: 24) It is less clear how teachers think about languages and the roles that the school language and target language play in their instruction. Even though the teachers in this study were responsible for the English-language portion of CLIL instruction, they all expressed the belief that learning subject-specific vocabulary in German is necessary for students. Some of the statements highlighted the belief that learning German-language subject vocabulary is especially important for students who do not speak German as a first language. Weil es gibt ja auch ein paar Kinder mit Migrationshintergrund, die dann vielleicht sowieso noch eine andere Muttersprache haben und gerade für die ist ja auch wichtig, dass Deutsch dann nicht zu kurz kommt, nicht vernachlässigt wird, weil das natürlich auch ganz grundlegend ist. (TE1 00: 11: 56) But even for German native speakers, teachers feel learning certain vocabulary in German is important. For Teacher D, learning a concept and content word such as “goods” in English before learning the equivalent German word “Waren” would only lead students to be confused. For this reason, Teacher D sought to reduce the amount of English vocabulary taught. When concepts and accompanying subject terminology were generally known by the students in German, for example the cardinal directions, Teacher D taught and used the English equivalents alongside the German words. Teachers also recognize that students will need certain German vocabulary in future school contexts and therefore make sure that students were taught the relevant vocabulary. Teacher C expresses this sentiment below. I don't know, when we talk about fruits for example, or ehm healthy/ healthy food and stuff like that, we had to make sure they knew everything in German and then add the English because when they go to other schools, yes, they have to know the different terms in German. (TC2 00: 31: 23) These statements suggest that for teachers in this study, learning German subject vocabulary is necessary first for students’ conceptual and linguistic development. Young German native speakers and students learning German as a second language need to first learn the German word when learning a new concept such as “goods.” Second, teachers feel students will need German subject vocabulary when going to other schools. 124 6 Results <?page no="125"?> There were, however, some differences in the way teachers prioritized vocabulary learning. Despite the previous statement from Teacher C expressing the necessity of learning German vocabulary, elsewhere in the interviews, she indicates that there are times when learning only the English word for something is acceptable. Well when we sat down and decided, one major factor is what we think is doable and what we think is too difficult, to be honest with you. It's not that we decide ehm, "Ok, the kids have to know everything in German and on top, like the icing on the cake we're going to make sure that they know some of the things in English. (TC2 00: 31: 23) For Teacher C, then, it may be permissible for students to have certain gaps in their knowledge of German vocabulary. Based on her statements, however, these gaps may only be permissible in certain topics. With content learning as the primary learning goal for all teachers in this study and learning German subject vocabulary viewed as necessary by all teachers, a majority of teachers in this study described learning English as a curricular enrichment or choice that students could make. The basic sentiment expressed was that because learning content through a foreign language is above and beyond the standards set out by the primary school Science and Social Studies curriculum of NRW, students are free to engage with the English language to the extent that they are able and feel comfortable. Teacher D expressed this sentiment most clearly and most consistently. For example, Teacher D sees learning a foreign language in CLIL as an opportunity to offer students “more.” […] Ich finde, es ist eine große Chance und Möglichkeit, ähm den Kindern einfach mehr anzubieten. Ähm und ihn ja so die Möglichkeit zu geben, über sich hinauszuwachsen. (TD2 00: 04: 25) Teacher D focuses on the learning potential inherent in offering instruction in a foreign language. However, Teacher A sees the “on top” nature of using English as a potential burden for some students. In the excerpt below, she describes the additional burden of learning through English for one particular student in her class. And I think then bilingual is too much. It is on top and sometimes I think it's hard for him. (TA1 00: 22: 42) Because teachers view teaching and learning through English as an addition to the curricular requirement set out in the state curriculum, teachers also feel that using English and therefore learning English is a choice that students can make on their own. For example, Teacher D describes students exercising this choice when deciding whether to complete an English worksheet or an equivalent German worksheet. Und das, was ich mache ist eigentlich on top und ähm, wer damit gut umgehen kann, der nimmt das auf und kann das vielleicht sogar auch selbst wiedergeben. Und ähm, wer damit noch nicht so gut umgehen kann, der konzentriert sich halt auf das was da auf Deutsch auf dem Arbeitsblatt geschrieben steht. (TD1 00: 34: 56) 6.1 What cognitions do teachers hold about primary school CLIL? 125 <?page no="126"?> Teachers’ descriptions of students choosing whether or not to use English could be interpreted as a form of differentiation in the CLIL classroom. However, none of the teachers in this study spoke explicitly of student choice in this way. The belief that students can choose the extent to which they use English in CLIL is a clear finding in this study, yet not a cognition that was explicitly expressed by all of the teachers. Nonetheless, whether or not a teacher explicitly expressed this cognition or not did not have a noticeable influence on classroom instruction. All of the teachers interviewed and observed in this study spoke English exclusively or almost exclusively with students. Students were therefore required to deal with English-language oral input whether they wanted to or not. The cognition that students can choose the extent to which they engage with English was primarily limited to speaking competences. However, on several rarer occasions, students could choose between a German-language worksheet and an equivalent English-language worksheet (Teacher D’s class) or whether or not they wanted to complete English-language tasks to be included in a lapbook (Teacher A’s class). With all teachers stating that learning German subject vocabulary is necessary and most teachers describing the use and learning of English as a curricular enrichment and a choice that students can make, most teachers also described different instructional roles for English and German in CLIL. The teachers that described various roles for English and German generally stated that English can be used for easier or less demanding content (either from a linguistic or content perspective) and that German can be used for more difficult content that requires more language use. For Teacher A, “the English things are basic” (TA2 00: 15: 23) and German is planned and used for more difficult content. And then we think about, “Ok it's better you do this in German” and the worksheets are in German because of course the English things are basic. Because their German is better than English of course. (TA2 00: 15: 23) Other teachers framed the balance of English and German use as a question of possibility. Teacher B asks herself, “What is possible in English and what has to be done in German? ” (TB1 00: 27: 00). In this instance, Teacher B is likely using the term “possible” to refer to what is linguistically manageable for children and not, for example, to what is allowed by curricular documents since she continues in her statement to ask, “like sometimes I wonder, Could I do more in English? Or is it too much for the children? ” (TB1 00: 27: 00). And then the big question is always, What is possible in English and what has to be done in German? So that's always the thing that's/ like sometimes I wonder, Could I do more in English? Or is it too much for the children? (TB1 00: 27: 00) Teacher B goes on to further describe how English and German might be used in tandem within a unit of instruction. For her, English can be used to introduce topics while German can be used to go more in depth into a topic. Um, if it's really bilingual, let's say it's really German and English because the topic needs both languages, English is mostly the language that you use to introduce a new topic, to 126 6 Results <?page no="127"?> talk about the topic and then German might be the language that you use to get like deeper into the knowledge. (TB1 00: 28: 29) Beyond the question of how difficult or demanding a particular topic is for students, there is also the question of whether certain topics need to be taught in German due to curricular guidelines or requirements. Teacher C expresses uncertainty about this. She wonders whether students need to “know everything” (TC2 00: 31: 23) in English and in German or if only English would sometimes suffice. But when it comes to bilingual learning you always have to decide, like do you want them to know everything in German and in English? Or are there certain/ like do we have the freedom I would call it, do we have the freedom to decide, "Ok, this is ok if they just know it in English." Or are there certain learning goals where we can take responsibility for that? And that's always a struggle to decide. (TC2 00: 31: 23) Teachers in this study described using English and German for different purposes. English is to be used for less demanding and easier content that does not require complex language skills or for introducing topics. German is to be used for more complex content that requires more complex language skills or for digging deeper into topics. Two teachers stated that English restricts their ability to build relationships with students and two teachers stated that English is sometimes used as a lingua franca in the classroom between German-speaking students and students who are new to Germany. These statements suggest that teachers experience English as both a barrier and a bridge to communication and building relationships in the CLIL classroom. The number of coded segments for both findings is small. However, for at least one teacher, the barrier to building relationships that the use of English causes is frustrating and a source of stress. In the excerpt below, Teacher D describes the difficulty of choosing words and phrases carefully in order to maintain conversations with children while on fieldtrips. While not representative of the data as a whole, this statement suggests that constant communication with students through a foreign language is a potential source of emotional challenge for CLIL teachers. This challenge is in addition to the fundamental challenge of teaching content through a foreign language. Ich muss immer darauf achten. Meine Antwort muss so konstruiert sein, dass sie verständlich ist, möglichst kurz. Ähm, mit bekanntem Wortmaterial, was jetzt nicht noch mehr Fragen aufwirft, weil dieses Kind mir vielleicht beim nächsten Mal dann nicht mehr die Geschichte erzählt, weil es denkt, "Ah, da kommt wieder sowas kompliziertes Englisches und dann/ dann lasse ich es lieber." Das will ich natürlich vermeiden. Deswegen muss ich wirklich alles geben, um mich verständlich auszudrücken. Aber ich bin selber manchmal, ähm, nicht zufrieden damit, weil ich/ weil ich es auf Deutsch vielleicht anders machen würde. Weil ich da andere Fragen stellen wurde. Weil ich da ganz anders mit umgehen könnte. (TD1 00: 42: 41) 6.1 What cognitions do teachers hold about primary school CLIL? 127 <?page no="128"?> Teachers in this study clearly prioritize content learning in CLIL. Teachers use language as a means to that primary end. Teachers state that the use of German is necessary when they deem the content or the language needed to learn that content as too difficult to be dealt with by students’ lower English competences. Teachers also state that students need to learn German subject language vocabulary and, depending on the context, will often prioritize teaching German subject vocabulary before teaching equivalent English vocabulary. Several teachers described learning English as “on top” or something optional if students are able to manage with English. Several teachers also described English as a barrier to teacher-student relationships as well as English as a facilitator of relationships between German-speaking students and newly arrived students with little or no German language skills. Though all of the teachers in this study described a role for the use of German in the CLIL classroom, all five teachers spoke English to their students exclusively or almost exclusively. The use of German for content learning usually took the form of the German-speaking teacher being more active in the lessons or in the form of German language written materials. 6.1.2 Cognitions about young learner SLA in CLIL Code Exemplary quote Coded segments Teachers High lan‐ guage learning Und ähm gerade im Vergleich/ also ich seh das halt wirklich immer besonders gut im Vergleich zu den nicht bilingualen Klassen, wo ich ja auch unterrichte, die einfach sehr viel mehr Probleme damit haben wirklich selbst Englisch zu sprechen selbst irgendwelche Äußerungen zu produzieren und auch beim Hörverstehen einfach mehr Probleme haben. Also ja das merkt man halt immer wieder ja. (TE2 00: 05: 27) 24 5 Natural learning I think especially for primary school or elementary school, the kids are not afraid. They just learn it even though they don't learn it. Like they just pick it up […] It just comes naturally to them. It's like how they learn their mother tongue. Just by, by being surrounded by it the whole time. (TB1 00: 11: 56) 18 4 Chal‐ lenging Ähm, zuerst einmal glaube ich, dass es für die Kinder einfach eine enorme Herausforderung ist, diese englische Sprache aufzunehmen und zu verstehen. (TD1 00: 54: 05) 11 4 Need to speak Like um, I always tell them, "It's a language. You can only learn a language when you speak it." […] Yeah because communication is key to learn a language. (TB1 00: 41: 08) 8 3 128 6 Results <?page no="129"?> Cogni‐ tive ad‐ vantages Unsere Kinder müssen wirklich ihren Fokus darauf haben, damit sie das alles verstehen. Und ähm es zeigt sich ja auch häufig, dass die bilingualen Klassen im Matheunterricht, im Deutschunterricht ähm bessere Leistung zeigen, weil sie einfach gelernt haben wie man/ wie man sich richtig auf etwas konzentriert damit man/ damit man was erfassen kann. (TD1 00: 54: 05) 5 3 The ear‐ lier the better Ich glaube, je früher man eine Sprache lernt, desto besser es ist. (TA1 00: 14: 15) 2 2 Table 10: Cognitions about young learner SLA in CLIL All of the teachers in this study perceived that their students learned a great deal of English by participating in CLIL. Teachers talked about students’ language competences both in general terms (e.g. “their English”) as well as in more skill-specific terms (e.g. listening comprehension). Teachers described seeing evidence of students’ learning primarily in terms of listening comprehension (e.g. following classroom discourse) and increased vocabulary learning. [A]lso das [always speaking English with the students] macht schon wirklich ganz viel aus, zum einen fürs Hörverständnis natürlich aber auch so für den Wortschatz und ähm auch so das produktive Wissen was die Kinder so haben. (TE2 00: 41: 32) Teachers tended to base their subjective evaluations of students’ language competences against two different standards. First, teachers compared students' L2 competences to the outcomes described in the English curriculum. Teacher D, for example, sees students’ English competences as surpassing the expectations laid out in the curriculum. Also wenn wir den Lehrplan anschauen, was gefordert wird nach Klasse zwei, nach Klasse vier, ähm da sind die/ da sind die bilingualen Klassen natürlich sehr/ sehr weit schon voraus in jedem Bereich. (TD1 00: 48: 28) Second, teachers compared their CLIL students’ competences to their non-CLIL students’ L2 competences. Below, Teacher B describes her CLIL students’ language competences as better than her non-CLIL students’ competences and explicitly links the advantage to language learning in CLIL. Ehm, well it's [CLIL] such a benefit for the children because I also teach English in non-bilingual classes. And, well my first grade is so strong in English, the bilingual first grade is extraordinary. I've never had so much English after a couple of months. But I also teach the 1A and 1C, the non-bilingual classes and then I see the big difference between their language learning. Because if like language learning in Sachunterricht wouldn't make any difference or in 6.1 What cognitions do teachers hold about primary school CLIL? 129 <?page no="130"?> sport or in PE that we teach or in art, there wouldn't be such a big difference between the classes. (TB2 00: 40: 28) Beyond their own judgments of CLIL students’ English competences, several teachers reported hearing from former students and from the students’ secondary school teachers. The message reported was always that those students’ who had participated in primary school CLIL had stronger English skills than those who did not and did not generally have problems with English at the beginning of secondary school (grade five in NRW). Teacher E’s comment below exemplifies this. Also das hören wir immer wieder jetzt auch von den (…), wenn Kindern dann in die 5. Klasse kommen, dann von den Lehrern da an den Schulen, da ist es immer so, dass die Kinder, die bei uns in der bilingualen Klasse waren, da gar keine Probleme haben und richtig gut sind. (TE1 00: 45: 21) For this group of teachers, their experience with their students and feedback from former students and those students’ secondary school teachers is enough to convince them of the foreign language learning benefits of CLIL, especially in the areas of listening comprehension and learning vocabulary. Debates about the effectiveness of CLIL or, more generally, about the effectiveness of English language teaching and learning at primary school in Germany (see, for example, Frisch & Diehr, 2018; Mindt, 2013) is either irrelevant for teachers or frustrating, as Teacher C expresses below. For Teacher C, the value of learning English in primary school is related, at least in part, to having fun. I think it's [CLIL] valuable. Um, and I think the studies that come out from time to time where they say, “Oh, English in primary school's terrible. No benefit.” Ah, I could freak out because they should come in a classroom and they/ they should just see the fun the children are having by learning something like a language. (TC1 00: 53: 30) While all of the teachers in this study spoke explicitly about the benefits of CLIL for students’ foreign language development, three teachers also stated the belief that students reaped cognitive advantages by participating in CLIL. One of the teachers makes specific reference to “research” which she says shows that students “don't only benefit like from their English knowledge but that their brains even function better than kids that are only taught in one language” (TB1 00: 17: 36). The other two teachers made reference to students’ performance in other classes such as Math and German or to students’ success at secondary school. Teacher D attributes this success to the concentration and focus required of students when forced to listen to and process content in a foreign language. Unsere Kinder müssen wirklich ihren Fokus darauf haben, damit sie das alles verstehen. Und ähm es zeigt sich ja auch häufig, dass sie/ bilingualen Klassen im Matheunterricht, im Deutschunterricht ähm bessere Leistung zeigen, weil sie einfach gelernt haben, wie man/ wie man sich richtig auf etwas konzentriert, damit man/ damit man was erfassen kann. (TD1 00: 54: 05). 130 6 Results <?page no="131"?> For teachers in this study, the L2 and cognitive advantages that they report observing in their students are the results of CLIL instruction and not of other factors such as selection effects (see, for example, Rumlich, 2016). Teachers described foreign language learning and learning through a foreign language in primary school CLIL differently. On the one hand, teachers described learning a foreign language in CLIL as a natural process. They described language learning as “easy,” some‐ thing children “just pick up,” an “unconscious” and “implicit” learning process, and different from the “constructed” learning of EFL lessons. On the other hand, teachers described learning through a foreign language in CLIL as an “enormous challenge” requiring a “big effort,” “ambition,” and “will.” Below, Teacher B describes foreign language learning in CLIL as both natural and easy. They just learn it [English] even though they don't learn it. Like they just pick it up and it's not, "Oh I have to speak English now." They just do. It just comes naturally to them. It's like how they learn their mother tongue. Just by, by being surrounded by it the whole time. (TB1 00: 11: 56) Teachers also described language learning in CLIL as a process that takes place implicitly. This implicit learning or acquisition of language is seen as a result of listening to high amounts of English input from the English-speaking teacher. [U]nd dann passiert aber auch ja ganz viel einfach ja unbewusst also implizit halt weil ähm ja weil ich ja die ganze Zeit Englisch rede. (TE2 00: 05: 27) Some teachers also described the view that learning through a foreign language in CLIL is challenging for young learners. This view was most clearly and emphatically described by Teacher D. She views the challenge of CLIL as lying in constantly processing and attempting to understand L2 input. This challenge is compounded by the already taxing school day and school week. When asked what she hoped a visitor to the class would notice, she responded: Ähm, zuerst einmal glaube ich, dass es für die Kinder einfach eine enorme Herausforderung ist, diese englische Sprache aufzunehmen und zu verstehen. Also gerade zu Beginn von Sachunterrichtsstunden oder zu Beginn von einer Reihe kommt sehr viel sprachlicher Input von mir. […] Und ähm es/ ähm ich wünsche mir dann halt, dass gesehen wird, wie bemerkenswert das ist, dass sie/ dass sie das können und dass sie daraufhin reagieren können und eine Aufgabe bewältigen können nach diesem/ nach diesem ganzen Sprachbad. (TD1 00: 54: 05) Though Teacher D views learning in CLIL as challenging, she also describes her view that students are able to overcome this challenge. The view that students are able to meet and overcome the challenge of learning in CLIL is shared by Teacher E. Also, ja, ich würde dann hoffen, dass die Eltern sehen, dass es schon anspruchsvoll ist, aber dass die Kinder nicht überfordert sind, sondern dass sie gut mitkommen und dass 6.1 What cognitions do teachers hold about primary school CLIL? 131 <?page no="132"?> die, ja dass man wirklich dann sieht, wie deren Englisch Fähigkeiten so wachsen. (TE1 00: 43: 35) While some teachers emphasized the ease of language learning in CLIL and others emphasized the challenge of learning through a foreign language in CLIL, the majority of teachers described the need to speak in order to learn a foreign language. The need to speak was stated both as a general principle of language learning as well as a principle that guides CLIL instruction. For example, Teacher B describes the need to speak as a general principle of foreign language learning. Like um, I always tell them, "It's a language. You can only learn a language when you speak it." Only (interruption) Yeah because communication is key to learn a language. (TB1 00: 41: 08) Other teachers described how the belief that speaking is necessary to learn a language influences their CLIL planning and instruction. Teacher C describes using social forms such as partner and group work in order to provide opportunities for students to use English and thus test hypotheses and receive corrective feedback from peers. So I think yes, eh social form/ different social forms, with a partner, in small groups, is definitely one big/ ehm big aspects of learning a language. Because that's the time where you use the language or where they should use the language, ehm that's the time where they can try it out and that's the time where they also correct each other without me even saying something. (TC2 00: 38: 11) Despite teachers describing the need to speak in order to learn a foreign language, classroom discourse in all five of the observed classrooms was noticeably marked by low levels of student output in English. 132 6 Results <?page no="133"?> 6.1.3 Cognitions about choosing CLIL topics and activities Code Exemplary quote Coded segments Teachers Concrete and actionori‐ ented Also generell ist alles schon mal ein bisschen einfacher, was anschaulich ist, wie z. B. jetzt in der E Klasse das Thema Obst und Gemüse, da ist natürlich viel auch mit “wir gucken uns Gemüse an” und man sieht es und fühlt es und kann dann halt auch die englischen Begriffe dazu lernen und so. (TE1 00: 36: 33) 18 5 Motivating topics And it also went well because it's just a nice topic for the children. Yeah. I think if you have a topic that's really focused on reading, answering questions, it's also hard to do it in English. But if you have a topic that gets the children by itself by just being a nice topic, it's also very easy to do in English. (TC1 00: 18: 15) 8 3 Available ma‐ terials Ehm, so yes, one of the reasons is I just have to work with what I have and this is why I chose it. This is why I chose to have it in the classroom. (TC2 00: 26: 51) 8 2 Manageable vocabulary Also schwierig finde ich, sind eher die Sachen, wo so ganz spezielle Fachbegriffe vorkommen, die auch auf Deutsch schon schwer sind. […] ich würde mal sagen vielleicht zum Beispiel das Thema Säugetiere hat, das hatten wir jetzt in der Klasse nicht, aber da ist ja der Vorteil, dass sie zum einen schon was an Vorwissen haben, weil ja Tiere auch ein Thema im Englisch Unterricht sind, also kennen sie schon einiges an Wörtern. (TE1 00: 24: 58) 4 4 Multifaceted activities I: Do you think the topic of the Vikings is a good or a useful topic for a bilingual Sachunterricht? TA: Ja. Finde ich schon. I: Why? TA: Also, ja, weil das natürlich jetzt so vielfältig ist, weil es so viele unterschiedliche Themen. (TA4 00: 23: 53) 2 2 6.1 What cognitions do teachers hold about primary school CLIL? 133 <?page no="134"?> Fun for the teacher Also ich mache unheimlich gerne Storytelling. Deswegen versuche ich zu jedem Thema im Sachunterricht jetzt auch eine Geschichte zu finden, die/ die das Wortmaterial aufgreift oder die/ und wenn es selbst/ wenn es eine Geschichte ist, die jetzt kein naturwissenschaftlicher Aspekt hat, sondern einfach nur das Wortmaterial beinhaltet, dann ist das einfach zum Einstieg ganz nett. (TD1 00: 25: 53) 2 2 Table 11: Cognitions about choosing CLIL topics and activities All of the teachers in this study described good CLIL topics as those topics which are concrete and visual in nature and that allow for the use of more active learning. Teachers described activities in which students could see and touch objects or conduct experiments as particularly appropriate to CLIL lessons. Teacher E below describes a typical unit of study which allows for this kind of learning. [A]lso generell ist alles schon mal ein bisschen einfacher, was anschaulich ist, wie z. B. jetzt in der E Klasse [first and second grade combined] das Thema Obst und Gemüse, da ist natürlich viel auch mit “wir gucken uns Gemüse an” und man sieht es und fühlt es und kann dann halt auch die englischen Begriffe dazu lernen und so. (TE1 00: 36: 33) In emphasizing the benefits of concrete and action-oriented themes that include activities such as experiments, some teachers explicitly described not needing to read or write extensively. Topics appropriate to CLIL instruction are those which minimize the use of written English. Everything that doesn't have to do too much with writing. That's just like experiments and stuff. […] So everything that's not written or just verbally makes it easier. Like as soon as they get something to do it's easier in English. (TB1 00: 07: 46) In addition to topics that are concrete and action-oriented, teachers described motivating topics as being well-suited to CLIL lessons. Teachers described motivating topics in terms of topics that are “nice,” “fun,” “so big for the kids,” “motivating,” and “gets the children.” Teacher C describes her thinking below. And it also went well because it's just a nice topic for the children. Yeah. I think if you have a topic that's really focused on reading, answering questions, it's also hard to do it in English. But if you have a topic that gets the children by itself by just being a nice topic, it's also very easy to do in English. (TC1 00: 18: 15) While concrete and action-oriented topics that are motivating for children were the most significant aspects of teachers’ cognitions about good CLIL topics, teachers also described choosing topics based on the availability of materials. This was a practical factor in teachers’ 134 6 Results <?page no="135"?> thinking. As Teacher B describes, the practicality of choosing topics with available materials is related to time management. Yeah, because I can't think of new units every year. I did it last year, no I didn't. I did it two/ three years ago. I had the material. Of course I add something or take something out. But it's just basically what I've been doing the last years. Not that I do the same thing every time but it's just time management. Really. Honest answer (laughs). (TB2 00: 18: 05) Teachers also described manageable vocabulary as playing a role in choosing topics and activities for CLIL. Manageable vocabulary is vocabulary that teachers deem to be “easy for the kids to remember” (TC2 00: 12: 38) or vocabulary that is non-technical and that students already know in German (see Teacher E excerpt in Table 11). Also, manageable vocabulary is L2 vocabulary that students already know in English. Teacher B describes using students’ knowledge of vocabulary from EFL lessons to support the use of English in some of her CLIL lessons. They know the body parts so we can talk about the different body parts and what their function is. So that's all possible in English almost. There might be some children that don't know it in English so my German speaking class teacher might step in and explain it again in German. (TB1 00: 05: 21) Two teachers also described choosing activities that offered variety to learning and choosing activities that were fun for the teacher herself. While the internal school curriculum usually dictated the larger topics that teachers taught in their CLIL classes throughout the year, teachers were able to negotiate with their German-speaking team teacher particular topics and activities that would be taught primarily through English. As described above, teachers prefer topics that are concrete and action oriented and thus reduce the need for written language. Teachers also described appropriate CLIL topics as those that are motivating for students. They also based their decisions on the availability of materials and the difficulty of vocabulary as well as choosing activities that provided variety and that teachers themselves enjoyed. 6.1 What cognitions do teachers hold about primary school CLIL? 135 <?page no="136"?> 6.1.4 Cognitions about teacher competences and demands Code Exemplary quote Coded segments Teachers More effort for teachers Dann Wortkarten geschrieben, Poster beschriftet, kleine Texte geschrieben selbst und uh genau. Es ist dann viele Eigenleistung, dass es ja dann auch genau zu dem Thema und zu der Lerngruppe passt. (TA5 00: 12: 16) 16 4 Strong teacher language skills needed Good bilingual teaching also means as/ for me as a teacher to keep up your language. To improve yourself. And I notice that now that I've been out for like one year, that that goes pretty fast. If you're lacking some English stuff. So I guess I have to watch more movies and go, I don't know, to America more often or something like that. So I feel like it's definitely your job to, in order to be authentic, to yeah, at least have a certain level of language and be able to use it. (TC1 00: 11: 38) 6 4 Benefit from communica‐ tion with other CLIL teachers Ehm so knowing what other teachers do and how they do certain things and observing what their experiences are and talking about it would definitely be helpful. (TC2 01: 08: 44) 4 2 Table 12: Cognitions about teacher competences and demands The majority of teachers in this study found the work of teaching CLIL lessons to be time consuming and challenging. While teachers spoke of the time and effort needed, for example, to learn new subject terminology themselves or to coordinate planning with German-speaking team teachers, the majority of statements about the demands of CLIL teaching were related to making oneself understood in English as well as to material creation and curation. Teachers described the effort required to make oneself understood to primary school students when speaking English. One of the ways that teachers described doing this was through the careful consideration of the English words and phrases that they use. This increased awareness of language use is, for Teacher B, “exhausting.” I think it's more exhausting than speaking German the whole time. Because I need to really think about the words that I use. How do I say it. Of course, in class one in the beginning you also think about how you say it in German. But in English it's even more important to really think about how I'm going to say this so they get the meaning and um what they're supposed to be doing. (TB1 00: 18: 29) 136 6 Results <?page no="137"?> In addition to careful and conscious language use, Teacher B described the need to be “an entertainer” and that this is often tiring. You need to be sure that you'll be able to be an entertainer. Because in German and Math lessons, you explain something and then they're working on it. But in English lessons or in English taught lessons, most of the time you're the active part because they need your input. So that's also (inaudible) because it's more exhausting sometimes. (TB1 00: 19: 00) Teacher B highlights what, for her, is different about foreign language teaching or teaching through a foreign language, namely that students “need your input” and are less able to work independently on tasks than they would be in classes in which they are using German. This requires the teacher to be “the active part” and therefore makes CLIL “more exhausting sometimes.” Teacher D also highlights the physical aspect of making herself understood when using English with her students. Supporting student comprehension requires effort and ambition on the part of the teacher. Also das eh, das sind schon Herausforderungen manchmal. Da ist/ ist man tatsächlich schweißgebadet, wenn man mit Händen und Füßen versucht sich verständlich zu machen und weiß, "Uh, auf Deutsch könnte ich so und so sagen" aber trotzdem dem Ehrgeiz hat, das/ das ähm, trotzdem vermitteln zu können. (TD1 00: 15: 20) Teachers in this study understand speaking English all the time to be one of their primary roles as a CLIL teacher. This understanding was enacted in their CLIL lessons as all of the teachers spoke English exclusively or almost exclusively. Teachers also spoke English for purposes beyond the curriculum. Below, Teacher C describes the efforts and demands of expressing meaning in CLIL in such non-curricular contexts. For Teacher C, using English in such situations is challenging, can be frustrating, and requires more time. [H]ow am I supposed to solve all these problems that suddenly come up? Like Aufteilen where they have to go to different classes, homework in German and Math, teaching German and Math. Ehm solving certain problems, talking about material, talking about upcoming stuff, ehm how am I supposed to do all of this in English when I'm not talking to native speakers? Like I can do my best and I can simplify, and I can visualize and I can do all that, but at the end of the day it costs a lot of time, a lot of energy, which you don't have if you have a situation like this already. (TC2 00: 40: 06) The frustration that Teacher C expresses in the above excerpt seems to be a direct consequence of the “one person-one language” (Raunser & Steffens, 2012, p. 37) policy of the DIPS program. For teachers in this study, CLIL is time consuming and demanding during the planning and instructional phases of teaching. Teachers felt that creating, curating, and revising CLIL materials took a lot of time and required additional effort beyond what would be necessary 6.1 What cognitions do teachers hold about primary school CLIL? 137 <?page no="138"?> in non-CLIL classes. They also felt that finding ways to scaffold meaning when using a foreign language, for example through body language and using simplified language, required more effort from them and more instructional time than would be necessary in a non-CLIL context. In addition to the time and effort required to make oneself understood in English, teachers described the time and effort required to create and curate English materials. For Teacher C, reviewing and making choices about materials as well as creating materials is time consuming. It's more time consuming to look at the material, to pick out what you want to do, to decide what you want to do, to make material that works in classrooms. (TC1 00: 16: 12) However, beyond just the time involved in thinking about and creating materials, Teacher C also describes the uncertainty that, for her, comes with choosing materials and creating her own materials. For Teacher C, this uncertainty is about the materials themselves and, as she expresses below, seems also to extend to her understanding of CLIL pedagogy generally. For Teacher C, her uncertainty could be alleviated, at least in part, by using “official material.” And kind of like, I don't know how to call it, official material. Like for example when you teach English there are different things at a regular school that you can pick out and you, for example, I don't know, if you want to work with Bumblebee then you work with Bumblebee. And we're missing that. So I would really like something where I know, Ok this material has been made by somebody who actually has a clue (laughs) and I can use it and when I use it I feel like, Ok, I have a clue too. So, yeah, material is definitely a big aspect. (TC1 00: 11: 38) For Teacher C, then, part of the challenge of CLIL instruction is dealing with her own uncertainty about CLIL best practices. This uncertainty is exacerbated by the lack of “official material.” While the uncertainty about choosing CLIL materials and about CLIL pedagogy ex‐ pressed by Teacher C was not expressed by other teachers, the effort and time required to curate and create materials was. For Teacher D, the effort and time required to revise materials, create materials, and curate materials from a variety of sources requires teachers to be internally motivated. Selbst wenn wir hier sehr eng hier im Team arbeiten, gucken wir uns die Arbeitsblätter immer nochmal an und müssen sich verändern. Wir brauchen für alles oft neue Bildkarten, Wortkarten dazu, ähm, […] Also es ist schon/ schon was, was man gerne machen muss. (TD1 00: 19: 48) In addition to the time investment and extra effort that teachers felt CLIL requires, four of the teachers described feeling that strong target language skills, in this case English, are necessary for CLIL teachers. Teachers spoke of strong target language competences both in terms of past experiences such as “hav[ing] been abroad somewhere” (TB1 138 6 Results <?page no="139"?> 00: 18: 29) as well as an ongoing process “to keep up your language” (TC1 00: 11: 38). Some of the teachers associated strong language skills with either travel to na‐ tive-speaking (English) countries or to experience communicating with native speakers. Below, Teacher C emphasizes communication with native speakers and suggests that while going abroad may not be necessary, from her perspective, it’s likely valuable. I think it's important that at some point during your education you were in touch with some native speakers. Um, and it doesn't necessarily mean going abroad, although I think that's probably a good point. (TC1 00: 15: 02) For Teacher B, the value of going abroad is not only to improve one’s language skills but also to develop confidence in one’s language use. This confidence can allow the teacher to use the foreign language more authentically with students by, for example, admitting to not knowing a word and then looking it up. I think they [CLIL teachers] really need to have been abroad somewhere in an English speaking country. So they need to feel safe in/ sometimes kids would ask me for a word and I would say I don't know, I have to look it up, which is just natural. So don't be afraid if you don't know a word or if you don't know how to express it. (TB1 00: 18: 29) Other teachers also pointed out that while they feel it is important for CLIL teachers to “feel at home” in a language (dass man sich zuhause fühlt in der Sprache), to be able to use it “in an authentic way,” with “ease” (Lockerheit) and “flexibility” (Flexibilität), none of the teachers described an “ideal native speaker” (see Council of Europe, 2001, p.-14) as the goal and several explicitly talked about being ok with making language mistakes. Teacher D describes her views about the foreign language competences needed by CLIL teachers. Ich finde, ganz wichtig ist, dass man/ dass man sich zuhause fühlt in der Sprache. Dass man ähm, nicht/ nicht viel nachdenken muss über seine Wortwahl, über seine Sätze. […] Dass man darüber hinwegsehen kann, dass man so eine gewisse Lockerheit ähm mitbringt, dass man nicht perfekt ist, dass man einfach je das Beste gibt, um den Kindern das zu ermöglichen. Und ähm ja, so eine gewisse Flexibilität gehört natürlich dazu. (TD1 00: 15: 20) Two teachers also described the importance of working with other CLIL teachers, whether that be at their own schools or in the broader CLIL community. For Teacher B, communication with the other CLIL teacher at school primarily served the purpose of planning, while for Teacher C, communication with CLIL teachers outside of her own immediate teaching setting, would, if available, serve the purpose of “observing what their experiences are and talking about it” (TC2 01: 08: 44). 6.1 What cognitions do teachers hold about primary school CLIL? 139 <?page no="140"?> 6.1.5 Cognitions about student characteristics Code Exemplary quote Coded segments Teachers Robust stu‐ dents benefit I: Speaking of students, what kinds of students benefit from bilingual education? TD: Ähm, alle, die eine Sprache sicher sprechen können, würde ich sagen. Ähm, natürlich sind immer wieder Kinder in meine Klassen dabei, wo ich denke, "Oo, vielleicht wäre eine nicht bilinguale Klasse besser gewesen im Nachhinein,” weil so viele Probleme vielleicht schon bestehen, dass man jetzt nicht auch noch on top dieses bilinguale zumuten muss. (TD1 00: 32: 44) 11 5 All students can benefit I: What kind of students benefit from bilingual education? TB: All of them. As long as they're open to the language. Yeah, but I think all of them benefit (…). (TB1 00: 32: 16) 3 2 Students with multicultural backgrounds can benefit Yeah and children just that have some kind of multicultural background. Doesn't have to be I don't know, British or American, whatever, just some kind of multicultural background. (TC1 00: 30: 27) 2 1 Table 13: Cognitions about student characteristics All of the teachers in this study made statements that students who are either generally strong students as evaluated by their performance in other subjects such as German or Mathematics or students who already speak either German or English well benefit the most from CLIL. Below, Teacher A describes one of her CLIL students who, in her estimation, is not benefiting from CLIL due to his “problems” in subjects such as German and Mathematics. I think for him it is difficult to be here in the bilingual class because he has problems with German but also with writing and mathematics and there's a lot of problems. And I think then bilingual is too much. It is on top and sometimes I think it's hard for him. He doesn't understand. (TA1 00: 22: 42) Teacher C echoes the view that students who struggle with learning generally do not benefit from CLIL. Yeah, students who are having a hard time learning in general. I think/ it just makes it worse for them. It doesn't help in any way. It just makes it worse. (TC1 00: 32: 33) In addition to the view that students with low general learning performance do not benefit from CLIL, several teachers also described language competences as important for 140 6 Results <?page no="141"?> describing who benefits from CLIL. While Teacher D described students who “speak a language confidently” [eine Sprache sicher sprechen können] (TD1 00: 32: 53) as benefiting from CLIL, others spoke specifically about being able to speak German or English. For students who have been raised in Germany and speak German, teachers described their general learning ability or their performance in German lessons as predictive of whether or not they will benefit from CLIL. For students new to Germany and who do not yet speak much or any German, teachers often described CLIL as not being beneficial. An exception to this view of CLIL’s suitability is if the student already speaks English. In the following excerpt, Teacher B describes her view that students who do not speak German but do speak English would benefit from CLIL while students who speak neither German nor English would benefit more from a non-CLIL class. I had some kids before, they were fluent in English but they didn't know any German. For them it makes sense to me to be in the bilingual class because then at least in the beginning they are able to communicate until they pick up enough German. But for the children that are not really fluent or near to/ able to speak English to me, I would actually suggest to put them in a non-bilingual class because then they are more surrounded by German. (TB1 00: 13: 51) By stating that “at least in the beginning they are able to communicate until they pick up enough German,” Teacher B is emphasizing the importance of linguistic communication in CLIL. For Teacher B, students who can communicate in either English or German will benefit from CLIL while students who cannot should first focus on learning how to communicate in German. Teacher C states specifically, that “native [English] speakers” benefit from CLIL “because their language is used. It makes it easier for them to just understand certain things” (TC1 00: 29: 58). Several teachers also stated that all students benefit from CLIL or that students with multicultural backgrounds could benefit. Yeah and children just that have some kind of multicultural background. Doesn't have to be I don't know, British or American, whatever, just some kind of multicultural background. (TC1 00: 30: 27) These views, however, were an exception to the more commonly described view that strong students benefit most from CLIL instruction. Teachers’ cognitions that CLIL is most beneficial to students who can already commu‐ nicate well in English or in German and who are generally strong students highlight the previously described cognition that CLIL is challenging, something that several teachers in this study expressed explicitly. For teachers in this study, those students who bring specific (English or German) linguistic or general academic strengths are most likely to benefit from the challenge of learning in CLIL. 6.1 What cognitions do teachers hold about primary school CLIL? 141 <?page no="142"?> 6.1.6 Summary The first research question this study seeks to answer is: What cognitions do teachers hold about teaching and learning in primary school CLIL? This broad and seemingly all-encompassing question has been answered within the context of the five subcategories outlined in Tables 9-13. The CLIL teachers in this study clearly think that non-language content learning is the first priority in their CLIL lessons. Languages (English and German) are primarily a means to this end. However, in using languages to meet this ultimate goal, teachers described various roles and language-specific learning goals as well. Teachers in this study describe using German to teach and learn content that teachers deem difficult for students or content that requires more advanced language competences. Teachers also describe learning German subject vocabulary as necessary. Teachers describe using English for content that they consider easier for students to learn or that requires less or simpler language. Several teachers describe learning English as a curricular enrichment and choice that students can make. The use of English for what teachers sometimes described as “easy” content coincides with teachers’ belief that concrete and action-oriented topics, those which support understanding in English through visual and tactile means, are best for CLIL. As a group, teachers describe foreign language learning in CLIL as natural and learning through a foreign language in CLIL as challenging. The description of learning through a foreign language in CLIL as challenging coincides with another cognition that all teachers in this study expressed, namely, that the students most likely to benefit from CLIL are those with the ability to communicate in German or English and who demonstrate general academic competence in other subjects. In addition to concrete and action-oriented topics and activities, teachers prefer CLIL topics and activities that children find motivating and topics in which the English termi‐ nology needed is either already known, non-technical, or known in German. Finally, teachers find CLIL teaching time-consuming and demanding during the planning and the instructional phases. During the planning phase, teachers find the creation, curation, and revision of materials to be time-consuming. During the instructional phases, teachers find the process of scaffolding meaning for students when speaking English challenging. Beyond these demands, teachers describe the need for CLIL teachers to have strong target language skills, often developed through time abroad or contact with native speakers of the target language. For teachers in this study, the goal is not to demonstrate target language perfection, but to be able to interact in more flexible and authentic ways with students. 142 6 Results <?page no="143"?> 6.2 Research question 2: What language-related educational goals do teachers work toward? The answers to RQ2 are summarized in the parts of the coding frame presented in the subsection below. Under the main category of CLIL goals are two subcategories, Foreign language learning goals and General pedagogic goals that support language learning. Teachers’ responses to these two subcategories are described in further detail below. 6.2.1 Foreign language learning goals Code Exemplary quote Coded segments Teachers The four skills speaking But I always want them to really speak. (TB1 00: 41: 08) 18 5 listening com‐ prehension Ok, in general, I would like that the children can show that they understand a lot more in English than they would without the bilingual teaching. (TC1 00: 21: 45) 17 5 reading (sen‐ tences and short texts) Und auch dann die Recherche kann man ja auch teilweise schon sogar auch Englischsprachige Kurztexte lesen und so (…) Ich würde sagen, das eignet sich dann (…) gut. (TE1 00: 24: 58) 12 5 writing [U]nd die Vorgaben dafür waren so, dass ähm (…) alle Kinder auf jeden Fall auch ein bisschen etwas auf Englisch machen sollen. Also auf dem Plakat sollte ein bisschen etwas Englisches auch stehen. (TE2 00: 01: 01) 2 1 Develop lin‐ guistic means (especially subject lexicogrammar) [S]ie müssen die Kernbegriffe. Deswegen machen sie ja dieses Picture Dictionary. Dass sie die Kernbegriffe auf Englisch können, das sie viking, ship, I don't know, sailor, what else is on there. Keine Ahnung. Genau. Wortschatzarbeit ist das im Endeffekt, was alles zum Thema Wikinger gehört. (TA4 00: 09: 38) 25 5 English be‐ yond EFL con‐ texts I: What do you see as the main goals for these types of lessons […]? -TA: […] To get in touch with the English language in different contexts. To get the chance to listen to English, to read English, to write English and not just with the typical English topics you do in the English lessons. (TA1 00: 19: 19) 6 4 6.2 Research question 2: What language-related educational goals do teachers work toward? 143 <?page no="144"?> Comprehen‐ sion strategies What I want my students to be able to do is if you don't understand every single word in what you hear when you, I don't know, when you're on vacation and you talk to somebody from another country. There are going to be words that you don't understand. But listen to the whole context. (TC1 00: 48: 42) 5 2 Table 14: Foreign language learning goals All of the teachers in this study described speaking and listening comprehension skills as two of the primary foreign language learning goals for CLIL. Taken together, teachers reported the importance of working toward oral language skills more than any other CLIL goal. This comes as no surprise since oral language development is a primary goal of primary school EFL instruction. However, despite speaking skills and listening comprehension skills being given almost equal value as described in the coding frame (Table 14), teachers described each goal differently. Teachers’ descriptions of listening comprehension as well as classroom observations suggest that developing listening comprehension skills is of primary importance for the teachers. All of the teachers described the goal of listening comprehension in terms of being able to follow classroom discourse. Because students most often spoke German while the teacher spoke English, following or understanding classroom discourse meant being able to understand the English-speaking teacher. Teachers spoke of this explicitly. Also die Kinder dürfen Deutsch reden, aber ich rede Englisch und dadurch wird auch nochmal ja der Wortschatz gefestigt und das Hörverstehen und so weiter. (TE2 00: 14: 48) Teacher B also expresses the goal of children simply being able to understand her. [C]hildren that understand what I'm saying and learn something without any German input. That would be the ideal I guess. (TB1 00: 26: 05) Similarly, Teacher C describes the goal of students being able to understand her. But in her statement below, Teacher C seems to expand the goal of listening comprehension to a more global sense of meaning making that goes beyond understanding mere vocabulary or subject-related input. Ehm because what I/ what they also should be able to do is understand not only the topics that we talk about but also understand “What is this person that speaks English to me right now, what does she want? ” (TC2 00: 40: 06) After making this statement, Teacher C went on to describe the challenge of communi‐ cating with children through English when talking about non-curriculum related topics, e.g. social issues. Despite the challenge that Teacher C describes, her statement suggests that, at least for her, the goal of listening comprehension in CLIL goes beyond content-related discourse to a broader ability to understand or make meaning in new and unplanned classroom situations. 144 6 Results <?page no="145"?> Three out of the five teachers stated, ultimately, that the goal is for all students to understand the English instruction without recourse to the German-speaking teacher or the need for additional explanations. Teacher B states this below. As soon as I know that they/ or as I see that they don't need the German speaking teacher any more or less or less, that's a sign for a success I would say. (TB1 00: 23: 33) Teacher D takes this a step further and describes the goal of listening comprehension as being able to understand English input so well that it does not affect students’ ability to work in the lesson. [O]der mir durch ihre Übersetzung zeigen, sie haben das wirklich ganz genau verstanden, durchdrungen und ähm und können so weiterarbeiten, als wäre es eine/ eine Stunde, die sie in ihre Muttersprache gerade erleben. (TD1 00: 11: 50) For teachers in this study, the goal of developing listening comprehension means being able to understand classroom discourse used in a variety of subject and non-subject areas. Because the English part of classroom discourse in all of the classes in this study came almost exclusively from the English-speaking teachers, understanding classroom discourse meant understanding what the teacher was saying. Teachers did not differentiate listening comprehension goals into, for example, listening for gist, specific information, or detailed understanding nor did they describe listening comprehension goals related to sources of input such as multimedia sources or other students. Based on the interviews, all of the teachers in this study clearly want their students to develop speaking competences. Teachers tended to describe speaking as the ultimate language learning goal for the CLIL classroom. For example, Teachers B and D describe their thinking below. But I always want them to really speak. (TB1 00: 41: 08) Und im besten Fall natürlich auch hin und wieder selber auf Englisch antworten. Also das sind für mich die schönsten Stunden, wenn ähm, wenn ich merke, es kommt langsam auch selber was auf Englisch. (TD1 00: 11: 03) These statements and others suggest that teachers recognize that developing speaking competences is challenging but nonetheless valuable. As a group, the teachers described a variety of specific types of speaking competences that include both oral production and oral interaction (Council of Europe, 2020). For the teachers in this study, oral production goals include using English in part or all of a presentation as well as repeating formalized language chunks. I would try to have them say, "The pencil floats," or "The pencil sinks." (TB1 00: 51: 30) Oral interaction goals included using English for common classroom phrases and responding to questions during whole-class discussions. Aber (inaudible) immer, das geht dadurch, dass alle classroom phrases auf Englisch kommen sollten, wenn sie auf Toilette gehen wollen, wenn sie eine Frage haben. Das haben 6.2 Research question 2: What language-related educational goals do teachers work toward? 145 <?page no="146"?> wir jetzt zu Beginn der Klasse drei direkt zu Anfang ähm geübt. (TD1 00: 13: 15) Teacher D’s goal of having students use English for routine classroom phrases was also observed in the classroom. In my observation summary dated 31.10.19, I recorded, “At one point a student asked in German to go to the toilet. [Teacher D] told him that he could if he used “that other language.” He responded, “May I go to the toilet.” With the exception of responding to questions during whole-class discussions, all of these forms of oral output provide students with an opportunity to plan their output or have their output scaffolded. Such planned and scaffolded forms of oral output may seem more achievable to teachers. As has already been stated, most of the students' oral production was in German. However, some English oral production was evident in all of the classes. In Teacher A’s class, for example, there was a discussion about the parts of a Viking ship. I recorded the following in my classroom observation summary (20.3.19): “During this time one girl offered a response (long) in English. Another boy started with English and went back to German. There was no solicitation of English and no pressure for others to do it [speak English].” Based on my observations in Teacher B’s class, I recorded the following in my observation summary (11.11.19): “[Teacher B] quickly moved to this week’s topic and asked about the CD player she brought with. One boy said “Hören.” [Teacher B] asked, “What is our English sense for that? ” Called on a [different] boy. “I can hear with my ears.” In this way, students were able to reproduce pre-learned chunks as a form of speaking in CLIL. These two classroom examples differ in two ways. First, in Teacher A’s class, the students constructed responses spontaneously and used language creatively while the student in Teacher B’s class repeated a pre-learned chunk. Second, the students in Teacher A’s class responded in English on their own volition while the student in Teacher B’s class responded to a prompt from Teacher B to use English. The teachers in this study differed with the degree to which they supported, encouraged, or required students to use English in specific situations. This will be reviewed in the section describing the results of teachers’ cognitions about CLIL approaches. Reading in English was a goal that all of the teachers spoke of. Teachers described being able to read individual words and sentences as well as being able to read short texts in English. Most of the reading that teachers described was in the form of reading words and instructions on an English worksheet. Several teachers described offering English worksheets to students. For example, Teacher D below describes translating German worksheets into English when it makes sense to do so. Ähm da/ ist es ja so, dass wir die/ an die Arbeitsblätter auf Deutsch schon vorliegen haben. Die beiden Parallelklassen nutzen die halt nur auf Deutsch. Ich habe sie dann umgewandelt auf Englisch wo ich es für sinnvoll gehalten habe. (TD2 00: 13: 27) 146 6 Results <?page no="147"?> Several teachers made reference to reading in other contexts such as reading an entire book (Teacher A) and reading for research purposes (Teacher E). However, based on my classroom observations, only Teacher C constructed lessons around the goal of reading short texts with her students. The texts that Teacher C read with students contained noticeably challenging vocabulary such as “inhabitants,” “majority,” and “hinders.” While the goals of reading texts such as this seemed to include learning vocabulary (From my field notes dated 1.10.19, Teacher C asks “What are inhabitants again? ”) Teacher C describes goals more in line with reading for gist or for specific information. Because the most important thing for me was, Ok, they read the text, they understood what it was about, and they matched it to a right term. So what more can I ask for? (TC2 01: 00: 21) Despite all of the teachers describing reading as a goal in specific contexts (e.g. reading worksheets), reading in English was sometimes presented as a choice, for example, when students could choose the English worksheet or the German equivalent in Teacher D’s class. The goal of developing reading skills in English was limited primarily to reading and understanding educational materials in the form of worksheets and short texts. None of the teachers described reading authentic English texts (defined more narrowly as non-educational texts produced for native speaking children, for example, picture books) with students in class nor did I observe any such activities during my observations. The majority of teachers in this study did not describe writing in English as a goal. The one teacher that did describe writing as a goal, Teacher E, talked about it in the context of copying English correctly onto posters. [U]nd die Kinder üben ja doch so ein bisschen das korrekte Abschreiben, was ja auch wichtig ist. Also auch beim Ab/ Beim Schreiben ist es ja so die Kinder müssen nicht aus dem Kopf heraus wissen "wie schreibe ich diese englische Wort." Sie müssen es eigentlich nur richtig abschreiben können. (TE2 00: 33: 03) Despite only one teacher describing writing as a goal, albeit in a very limited form, students had opportunities to write in some of the classes. Teacher D, for example, reported that her CLIL students write in English weekly in the form of a journal. Also, while observing in Teacher D’s class, I recorded the following student-student interaction in my field notes (dated 14.11.19). S1: Ich hab vergessen, wie man das auf Englisch schreibt. S2: Schreibs auf Deutsch. S1: Nein. Ich schreibe auf Englisch. The exchange demonstrates not only that opportunities to write in English exist, but that occasionally, those opportunities are taken. All of the teachers in this study described learning subject vocabulary and, to a lesser extent, related sentence structures as a primary goal for their CLIL lessons. The subcategory of Develop linguistic means (especially subject lexico-grammar) was the most described foreign language learning goal. That this goal would be described so frequently by the teachers is perhaps not surprising, since words are the primary carrier of meaning in a 6.2 Research question 2: What language-related educational goals do teachers work toward? 147 <?page no="148"?> language. Teacher A below describes explicitly the need for students to learn core subject vocabulary. [S]ie müssen die Kernbegriffe. Deswegen machen sie ja dieses picture dictionary. Dass sie die Kernbegriffe auf Englisch können, dass sie viking, ship, I don't know, sailor, what else is on there. Keine Ahnung. Genau. Wortschatzarbeit ist das im Endeffekt, was alles zum Thema Wikinger gehört. (TA4 00: 09: 38) When teachers spoke about learning vocabulary and, to a lesser extent, grammar, they spoke specifically about subject vocabulary or particular content words that they deemed necessary for students to understand the content of any particular unit. These content words and sentence structures played an important role in teachers’ CLIL planning and in their curation of CLIL materials. Teacher B below describes vocabulary as being a key component to understanding content and therefore something to be learned before engaging more deeply with content learning. So really think about/ and what words would the children need to understand? [Are] there any words that they need to learn before you really get into the content? (TB1 00: 51: 41) Some teachers also used subject vocabulary as a point of collaboration with their German-speaking team teachers. Teacher C, for example, coordinated with her German-speaking counterpart to identify vocabulary that students would need. I also talked to [German-speaking team teacher] about what's/ what different terms they're going to know in German and what is helpful for them to also know in English ↑ (TC2 00: 52: 54) Teacher D, for example, used subject vocabulary and sentence structures to guide her choice of materials, including multimedia materials. [Z]u dem Thema gibt’s doch bestimmte Lied, dann gucke ich bei YouTube wo ähm, wo gibt’s ein Lied wo die Satzstruktur vorkommt oder das Wortmaterial vorkommt, was/ was wir brauchen.(TD1 00: 19: 48) A number of classroom activities that I observed were explicitly focused on learning subject vocabulary. For example, Teacher A and her students labeled parts of a Viking ship. Students in Teacher B’s class tasted popcorn and described the taste. Teacher B recorded their descriptions. Students in Teacher D’s class practiced English and German expressions of geographic orientation, for example, “Norway is west of Sweden, Norwegen liegt westlich von Schweden.” This observational data confirms what teachers described in the interviews, namely, that teaching and learning subject vocabulary and sentence structures are central foreign language learning goals in their CLIL instruction. Beyond developing the four skills and learning vocabulary and grammar, a majority of the teachers described CLIL as a context in which students could use and learn English beyond what is possible in EFL lessons. One way that teachers see this happening is through the variety of topics studied in CLIL. Studying topics such as the Vikings or electricity through English allows students not only to learn English words not normally part of the EFL curriculum but to go deeper into topics such as animals that are often a part of the 148 6 Results <?page no="149"?> EFL curriculum. When asked to describe her primary goals for CLIL lessons, Teacher A described broad goals related to engaging with the English language through different modes and in different contexts. She states that one of the main goals of CLIL for her is to learn words that “are not typical.” To learn English (laughs). To get in touch with the English language in different contexts. To get the chance to listen to English, to read English, to write English and not just with the typical English topics you do in the English lessons. So, talking about the Vikings and you heard some words already in English about the boat or they have to do the picture dictionary. They have a lot of um/ these are not typical words you learn in English lessons in primary schools. (TA1 00: 19: 41) Teacher B also highlights the different topics in CLIL lessons compared with EFL lessons. For Teacher B, these different topics offer opportunities to ask different questions such as “Where do the animals live? What do they eat? What do they look like? How many babies can they get? How old can they be? ” (TB2 00: 09: 28) For Teacher C, CLIL’s unique topics include experiences that go “beyond just teaching,” for example unique cultural events or holidays. I would like them [a hypothetical visitor to the class] to notice that apart from this whole doing math, German, science, English, that we're also doing things that might not be done in a regular class like Halloween parties, sandwich parties, um Thanksgiving, Christmas in a different way. So stuff like that that's yeah, goes beyond just teaching and offering kids to experience that. (TC1 00: 27: 45) For these teachers, CLIL is a context which allows teachers and students to go beyond the standard curriculum and learning experiences normally available in non-CLIL Science and Social Studies lessons as well as English lessons. Teachers work toward goals that are both measurable (e.g. learning vocabulary) and goals that are not easily definable or measurable (e.g. experiencing new events or holidays). In addition to learning language due to expanded curricular opportunities, at least one teacher described the uniqueness of CLIL students themselves as providing lan‐ guage learning opportunities. Teacher B feels that having students with “different backgrounds” and “interesting backgrounds” (TB1 00: 12: 46) first provides an authentic context in which to talk about the various places children are from. Because some of the students not from Germany did not (yet) speak German but did have some English skills, Teacher B saw this as an opportunity for meaningful English language communication between students. Teacher B also described students in her CLIL class as being “more open minded towards other cultures” (TB1 00: 12: 46) because of their interaction with students of varied backgrounds. I think they are more open minded towards other cultures. Because the/ yeah because they know that, especially in the 6.2 Research question 2: What language-related educational goals do teachers work toward? 149 <?page no="150"?> bilingual classes, you have a lot of kids from very different backgrounds, and interesting backgrounds, so that you can really talk about where you're from and why is it important to speak English so that they get a meaning of it. We have three kids in first grade now, they don't know any German. They know some English so that's the way for them to communicate. So it has a meaning for them. (TB1 00: 12: 46) Teacher B does not describe being open minded in terms of Byram’s (1997) model of intercultural communicative competence. But her statement does suggest that she sees CLIL as a context in which students can develop attitudes in line with Byram’s model. Teacher B’s statement about students being more open minded is one of the few statements by teachers in this study that could be interpreted as teachers recognizing the development of some kind of intercultural competence as a goal of CLIL. The only explicit reference to the term “intercultural” came from Teacher D in a broad closing statement during our first interview. In her statement, she lamented the then current consideration in NRW of removing English lessons from primary school altogether or delaying English instruction until the beginning of third grade (since our interview, the Ministry of Education for NRW has delayed the beginning of English as a foreign language instruction to the beginning of third grade for children starting school in the 2021-22 school year) (MSB NRW, 2020). In this context of describing the benefits of English instruction generally, she states that learning English in primary school supports students in building a foundation for further learning in “our intercultural lives” (Was man ihn für eine Basis schaffen kann für alles weitere lernen in unserem interkulturellen Leben. TD1 00: 58: 01). It would likely be false to claim that teachers in this study did not work toward goals typically described in terms of developing intercultural competence. They did not, however, describe their goals in those terms. Their descriptions of CLIL goals and, more broadly, benefits were couched in language that described their specific classroom contexts. Teachers B and C, for example, both spoke of students using English in their classrooms with students who spoke little to no German. One could interpret such statements as, for example, working toward the development of “skills of discovery and interaction” (Byram, 1997, p. 33). However, based on their statements, the teachers’ focus seems to be simply on the use of English in meaningful contexts. For teachers in this study, those meaningful contexts are often classroom-based and therefore dependent on the unique cultural and linguistic backgrounds of the students in the class. In addition to foreign language learning goals such as learning vocabulary and devel‐ oping listening comprehension skills, teachers in this study wanted students to develop comprehension strategies. Two of the teachers described wanting their students to develop the ability to glean meaning from context and not just individual words. As Teacher D describes, students need to be confronted with the question of how to build meaning when they have not understood all of the words. "Wie kann ich mir das jetzt erschließen? Ich habe nicht alles verstanden, aber ein paar Wörter habe ich verstanden und wie ähm/ wie baue ich mir daraus den Rest des Gerüsts dann 150 6 Results <?page no="151"?> mit da ich/ dann mit da ich das ganz verstehen kann? " (TD1 00: 38: 01) This goal of working toward understanding even when being confronted with unknown words was a central goal of Teacher C’s CLIL instruction. In one of her lessons, students read short texts that included noticeably challenging words and sentences such as “possessions,” “vantage point,” “inhabitants,” and the sentence “A castle was also used as a base from which knights rode out to fight and plunder.” I asked her to comment on her use of what I described as “more advanced foreign language” with her students. Her extended response suggests two primary reasons for working toward the goal of developing comprehension strategies. First, learning a foreign language or learning through a foreign language is different from learning through one’s mother tongue. For Teacher C, the principal difference is that “you don’t have to understand every single word” (TC2 00: 49: 20) when learning a foreign language. Instead students need to learn to use context to understand the gist of a text. Learning this skill is made more difficult for children because of their experience with teaching practices that deemphasize words in context. It's very difficult for me to convey to them that when you learn a language you don't have to understand every single word. And this is something that I experience every day in class. When children learn a language they want to know every single word because that's how they grew up, in first grade we talk about every single word. We talk/ we kind of take the words out of the context and out of the sentences and just teach every single word[…]. (TC2 00: 49: 20) One way that Teacher C supports the development of comprehension strategies is to confront students with short but complex written texts. Teacher C sees such tasks as requiring contextualization in order for students to make meaning. Second, Teacher C sees the development of such skills as preparing students for future learning at secondary school and suggests that understanding foreign words in isolation will be insufficient for students’ later learning. [W]hen they come in grade five, six, seven, nobody will ask them, "Oh, do you know what a pet is? Can you tell me/ can you name ten pets? " They want to know sentences. They want to know everything around it. And that's kind of my thinking behind that. (TC2 00: 49: 20) It seems that for Teacher C, one of the primary goals of CLIL is to give students the opportunity to develop foreign language discourse skills and not just subject language vocabulary and in so doing develop strategies for comprehending complex texts. 6.2 Research question 2: What language-related educational goals do teachers work toward? 151 <?page no="152"?> 6.2.2 General pedagogic goals that support language learning Code Exemplary quote Coded segments Teachers Reducing fear and devel‐ oping confi‐ dence To be confident and to actually dare to share what you can even if, I don't know, you think, “Oh my god, my English is so bad.” Just show what you can and then we will go from there. (TC1 00: 35: 11) 12 5 Motivated lan‐ guage learners Also die Kinder sollten ähm/ sollten motiviert in meine Stunden gehen und auch motiviert, ähm (laughs) hinterher wieder rauskommen. (TD1 00: 38: 01) 9 3 Success through effort Ähm, ja es hat ganz viel damit zu tun was wir gerade besprochen haben, dass wenn man/ wenn man kein Motivator hat und sich immer so ein bisschen ausruht auf seinem Level oder das dann auch noch vorbereitet bekommt ist das Erfolgserlebnis natürlich nicht das selber wie wenn man ähm sich selbst erarbeitet hat oder sich getraut hat ähm, ja, etwas anzugehen was man sich vorher vielleicht nicht unbedingt getraut hätte. (TD2 00: 25: 40) 9 2 More re‐ spectful I also want them to have a level of respect that they don't laugh at each other's faults or misspellings or whatever like that. That's very important to me. (TC1 00: 35: 11) 4 1 Table 15: General pedagogic goals that support language learning In addition to CLIL goals more directly related to foreign language learning, teachers also expressed a range of general pedagogic goals that support foreign language learning. One of those goals is Reducing fear and developing confidence. It is in these two ways, the reduction of fear when confronted with a foreign language on the one hand and the development of confidence to use the language on the other, that teachers described this goal. While all of the teachers described the importance of reducing fear and helping students feel safe and secure in the context of using a foreign language, teachers differed in how they described the result of achieving this sense of security. For some teachers, the result seemed to be a sense for children that the use of English is “normal.” For these teachers, students arriving at a sense that using English is normal was a significant achievement and one that set CLIL students apart from non-CLIL students. Teacher A, for example, sees this difference when she uses English in non-CLIL classes. [I]t is normal for the children here when I talk English to them, they don't think about it. When I would go to another classroom, and this sometimes happens, and I'm in my English thing and I talk to them in English and they have a look to me 152 6 Results <?page no="153"?> and, "Hääh? I didn't understand." And yeah, here it is/ they are used to this from the first day on. (TA1 00: 19: 41) This sense of normality, however, isn’t built just in the classroom. In the schools I visited which employ a one person-one language model of immersion, teachers spoke English with students not only in the classrooms but in the hallways and on the playground as well. As Teacher E describes, it is the use of English outside of the classroom as well as inside the classroom that helps students develop this sense that using English is normal. Ich würde sagen, dass das so den Umgang mit der englischen Sprache total normalisiert, weil ja jetzt [other English-speaking teacher at school] und ich auch halt nicht nur im/ in der Klasse, sondern ja auch auf dem Flur und auf dem Schulhof, in den Pausen und so Englisch mit den Kindern sprechen, […] Und ich glaube, das ist auch was Gutes, weil die Kinder ja, also es ist dann nicht mehr so was Fremdes, also die fremde Sprache, die man nur im Unterricht ab und zu mal benutzt, sondern so was ganz Alltägliches. (TE1 00: 47: 19) Beyond reducing fear and achieving a sense that using English is normal, one teacher emphasized developing confidence in her students. For Teacher C, confidence was about taking risks, to “dare to share” (TC1 00: 35: 11) with others through a language that is “foreign to them and weird” (TC2 01: 03: 55). The confidence that Teacher C describes is a step beyond that of simply achieving a sense of normalcy with the English language. It suggests a degree of action on the part of the teacher beyond simply speaking English with students in a variety of school contexts. Teachers also want motivated students, students that “like learning English” (TC1 00: 41: 35) and feel “that English is fun” (TB1 00: 34: 58). Being motivated to learn English is not just an attitude that teachers feel students should bring to the class but an attitude that teachers feel could be cultivated in class. For several teachers, motivation and enjoying the process of learning a language were prioritized. When asked about goals for her CLIL lessons, Teacher B responded in a way that suggests that the goal is obvious. I: What do you want students to learn in your bilingual lesson? TB: That English is fun (laughs). (TB1 00: 34: 49) Teacher C explicitly describes “hav[ing] fun when they learn English” as a priority. That’s my personal first goal. I want them to have fun when they learn English because I think that’s the most important thing. (TC1 00: 41: 35) In the interviews, teachers described two ways that they try to support students’ motivation in CLIL lessons. Teacher C describes creating more personal interactions with students as one way of motivating students to use English in class. So I think personal situations and not only one on one but also in smaller groups, ehm laughing together, creating situations where you are just personal with each other and you can address more personal, those are situations where they 6.2 Research question 2: What language-related educational goals do teachers work toward? 153 <?page no="154"?> learn the language and where they’re/ where they, I don’t know, are more motivated. (TC2 01: 03: 55) By creating personal situations in which students and the teacher can laugh together, Teacher C seeks to couple language learning with students’ positive emotions. Teacher D saw motivation as something necessary to overcoming the challenge of learning through a foreign language and something that could be developed in response to the challenge of learning through a foreign language. As described earlier, Teacher D sees one of the challenges of learning through a foreign language as developing strategies to understand input when not all of the words have been understood. One of Teacher D’s goals for her students was to develop a sense of fun in facing and overcoming such challenges. Aber diese/ diesen Spaß daran, sich auszuprobieren, nicht aufzugeben, wirklich zu gucken, ähm “Wie kann ich mir das jetzt erschließen? Ich habe nicht alles verstanden, aber ein paar Wörter habe ich verstanden und wie ähm/ wie baue ich mir daraus den Rest des Gerüsts dann mit da ich/ dann mit da ich das ganz verstehen kann? ” (TD1 00: 38: 01) Related to teachers’ thinking that learning through a foreign language in CLIL is challenging, teachers described wanting their students to experience success when putting forth the effort required to overcome this challenge. In the interviews, Teacher C described “dealing with something that is foreign” (TC2 01: 08: 44) as essen‐ tial to what CLIL is and what CLIL has to offer. Learning through a foreign language in CLIL tends to make everything unclear. The benefit of learning in such a context is learning how to manage this uncertainty and working through it. Ehm, I think apart from this whole learning language thing, there’s another great thing they learn from it [CLIL] and that is dealing with something that is foreign, dealing with something you don’t understand at the first moment and still try to be motivated enough to find out and to deal with it. (TC2 01: 08: 44) Teacher C’s description of learning in CLIL as being about “dealing with something that is foreign” suggests that she sees learning in CLIL as something qualitatively different from the learning that takes place through the school language (German). Teacher C does not indicate how students should deal with the foreign language. However, to “deal with” the foreignness of a foreign language suggests that students need to manage a variety of emotional (e.g. insecurity, confusion, curiosity) and strategic (e.g. using context or teacher-provided scaffolds such as pictures) responses when confronted with the foreign language. Achieving success by putting effort into understanding a foreign language has benefits in and beyond the classroom. In the excerpt below, Teacher D describes benefits that include further language learning as well as personal development. Also sobald ich sehe, „Ah die/ die knien sich darein und ähm und wollen das durchdringen“ das ist glaube ich ein ganz, ganz wichtiger Schritt auch für alles weitere Fremdsprachen lernen. Dass man ähm ja diesen/ diesen Willen, 154 6 Results <?page no="155"?> diesen Ehrgeiz entwickelt, ähm auch mit nur wenigen Worten die man verstanden hat, sich daraus was zu konstruieren, vielleicht gezielt Nachfragen zu stellen, ähm denn so ist es ja im täglichen Leben auch! Das ist das, was man/ was man lernen und akzeptieren muss. (TD1 00: 38: 01) Only Teacher C described the goal of wanting students to develop an attitude of respect toward others in the CLIL context. At the core of Teacher C’s thinking about this goal is making sure that students show respect toward their classmates when using English. I also want them to have a level of respect that they don’t laugh at each other’s faults or misspellings or whatever like that. That’s very important to me and I think I can say that now in my class four and three, that never happens because they know I would freak out. (TC1 00: 35: 11) Elsewhere in the interviews with Teacher C, she described wanting students to learn to show “respect to other cultures, other language[s], tolerance toward not knowing everything and still trying” (TC1 00: 54: 14). These larger and more abstract goals, however, were described when she was giving a kind of closing argument about the value of CLIL from her perspective. It is possible that for Teacher C, such global and decontextualized goals of developing respect for other cultures and languages are realized in the context of students respecting one another in her CLIL lessons. 6.2.3 Summary The second research question is: What language-related educational goals do teachers work toward? Teachers’ responses were divided into Foreign language learning goals and General pedagogic goals that support language learning. Among the foreign language learning goals that teachers worked toward in their CLIL lessons, oral language competences and the development of linguistic means (e.g. vocabulary and grammar) were the most often described goals. While the development of speaking competences tended to be described as an ideal and ultimate goal of CLIL, the development of listening comprehension skills was described as a core goal by all teachers. The development of listening comprehension skills was often described in terms of understanding what the teacher is saying and following classroom discourse. Opportunities to develop listening comprehension skills through, for example, multimedia (e.g. music, videos) or student-student interaction were observed to be rare. In addition to these core language learning goals, teachers seek to develop basic written language competences, primarily reading, by reading directions on English language worksheets and reading short English texts. While opportunities for writing in English existed in the classroom, teachers did not emphasize these in the interviews. Teachers also seek to develop language skills that go beyond typical EFL lessons. Teachers described learning words not typical in EFL lessons and using English in more complex ways as two benefits of CLIL. 6.2 Research question 2: What language-related educational goals do teachers work toward? 155 <?page no="156"?> Finally, teachers want their students to develop comprehension strategies in CLIL. Specifically, they want students to learn how to make meaning from context when faced with not understanding single words in various forms of input. Teachers also work to develop general pedagogic goals that support language learning. Chief among these goals is helping students get over the fear of being confronted with a foreign language. With all teachers reporting that students generally achieved that, some teachers described working to develop confidence in their students. Teachers also want their students to be motivated language learners. They want their students to like learning English and to have fun. Several teachers prioritized this goal. Teachers also described the goal of experiencing success through the effort required to overcome the challenge of learning in CLIL. Learning in CLIL is about working through something that is foreign and figuring out a way to understand using available resources and context. The experience of overcoming this challenge has the potential to give students a foundation from which to continue learning foreign languages and achieve success in other aspects of their lives. Finally, one teacher described the goal of developing the attitude of respect toward others. Within the CLIL classroom context, this goal was expressed as showing respect to classmates when speaking English but was also described in more global terms, e.g. respecting other cultures and languages. 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? The answers to RQ3 are summarized in the tables below. Teachers' descriptions about what they believe to be effective CLIL teaching will be described in more detail in each of the following subsections. I have ordered the subcategories based on an internal logic as well as on the number of coded segments. The first three subcategories, Focus on input, Know the students, and Scaffolding, have been organized based on my interpretation of how they fit into teachers’ thinking. Focusing on input is the first and foundational step to how teachers in this study think about CLIL approaches. Their knowledge of students then acts as a filter through which all subsequent approaches and decisions are modified. Various forms of scaffolding are the most commonly described approach (most coded segments) followed by various other approaches and organized in decreasing order by the number of coded segments. 156 6 Results <?page no="157"?> 6.3.1 Focus on input Code Exemplary quote Coded segments Teachers Provide as much English input as possible Provide as much oral English input as possible Ich glaube, dass wirklich ganz viel halt durch dieses ständige Englisch Reden (…) also von mir aus (laughs), dass das schon einfach ganz viel ausmacht. (TE2 00: 41: 32) 33 5 Authentic ma‐ terials Ich habe natürlich den Tipp gegeben, das Ausland ist immer die beste Quelle also. […] Also das war beispielsweise ein Tipp ähm, ja also ich habe bei jedem/ bei jedem Urlaub immer so ein bisschen die Augen offen gehalten, ob es authentisches Material dort gibt, was ich mitnehmen kann. (TD1 00: 23: 47) 4 3 Independently used materials It's easy to do and it doesn't need a lot of explaining and a lot of long texts or even books or research at home. I guess the keyword is simplify. So everything that's easy to do and easy to read and doesn't take a lot of time explaining for me is good materials. (TC1 00: 46: 36) 3 2 Provide German input when necessary Manage com‐ plexity And as soon as it gets to a part where you look into blank faces, the German part comes in, and maybe helps with/ sometimes with translation depending on the age and the understanding of the kids. But, yes, as soon as it gets too complicated (TB1 00: 05: 21) 23 5 Pastoral care There are some situations you can't [use English] because if there is a conflict between the children or if they are sad and, I don't know, you have to talk to them emotionally, then it's easier to do this in German. This is what my experience/ I tried this in the beginning also to do this in English but this is not working quite well. (TA1 00: 28: 59) 10 3 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 157 <?page no="158"?> Teach local and personal topics And also if you talk about, I don't even know, Sexualkunde, we do mostly in German. Because it's a very sensitive topic and you don't want to stop the children from really being honest and talk about it and ask questions. And the questions shouldn't be about not understanding the English but like really what's going on in their mind. (TB1 00: 07: 27) 5 2 Save time I: Is that important to you that ehm that German is still used both from an input and from an output perspective in the bilingual classroom? TE: Ja, ich denke schon. Also zum einen glaube ich, wenn es nicht so wäre, dann wären wirklich viele Kinder überfordert, und dann könnte man auch nicht mehr gewährleisten, dass man den ganzen Stoff, den man ja irgendwie unterkriegen muss, auch schafft, weil wir dann auch sehr viel Zeit brauchen würden. (TE1 00: 11: 36) 1 1 Table 16: Focus on input Teachers in this study all described the importance of providing students with as much English language input as possible. The input that teachers describe is principally oral input from the teacher. All of the teachers described this English language input from the teacher as forming the bedrock of what they do in CLIL and how children learn English in CLIL. For teachers in this study, children can learn more English when they are exposed to more English input. Well the more English input they have, the more they can learn. (TB2 00: 35: 20) Because teachers feel that providing students with high amounts of English input provides the basis of English learning in CLIL, they also feel that speaking as much English as possible with students is one of their primary responsibilities as CLIL teachers. They feel their job is to “talk as much English as you can” (TA1 00: 28: 59), to “stick to the language” (TB1 00: 08: 59), to “just mak[e] sure that they are provided with more ehm input” (TC2 00: 26: 51), and to provide a “Sprachbad” (TD1 00: 07: 07). In this way, teachers spoke of providing English input in quantitative terms. The key to CLIL instruction is providing more English. Teachers feel that students' success or the extent to which they learn English is largely dependent on their exposure to more English input than what is possible in EFL classes. Teacher C makes this point explicitly. She ascribes her former students’ successful English learning not to any factors related to her personal qualities or professional actions, but instead to the fact that her students have had more English input. I would say it was successful because when they come back [to visit after attending secondary school] and they tell me, 158 6 Results <?page no="159"?> “Ok, my English is so good. Everyone tells me, “Oh my god, who was your teacher? ”” That's one indicator for me. That's not nec/ I'm not saying, Ok, it was because of me, but it was because of the amount of English they had. (TC1 00: 20: 14) In order to provide more English input, teachers needed to use English in contexts beyond teaching Science and Social Studies content. Teachers used English for instruction as well as for general pedagogical and classroom management purposes. Teacher A describes this use of English as a “good thing” (TA1 00: 32: 55) as well as something unique to CLIL instruction. Aber diese ganze Instruktion, "Hol jetzt deine Mappe raus. Schreib die Hausaufgaben in den…" All dieses was hier auch im Englischen läuft, ist nicht in den anderen Klassen so normal. (TA4 00: 22: 57) Several teachers also spoke about the benefit of using English outside of the lessons and for topics not related to the Science and Social Studies curriculum. Teacher C, for example, feels that using English for extracurricular topics during lessons was an opportunity to maintain her use of English and “be authentic” in her language use with students. For Teacher C, speaking German with students in non-instructional situations would be inauthentic. However, later in the interview, she adds that using German with students would be appropriate for talking about personal issues and building relationships. Ehm, in situations where we do Christmas cards or where we prepare for Vorlesetag, ehm, that's something where I try to just stay in English with them because that's what I think classroom discourse in English also means, by me being the one representing English for them. And it's just/ would not be authentic if suddenly just because we're talking about Vorlesetag, which is in German, “Ok this/ I'm just going to speak German with you.” (TC2 00: 40: 06) While the majority of teachers’ statements about providing as much English input as possible were related to teachers themselves speaking as much English as possible with their students, some teachers also described the importance of increasing English input in other ways. Teacher D describes reading aloud in English to students as a way to introduce students to new topics and vocabulary. Also eine Vorlesungsrunde zu machen, dann haben die Kinder viel/ viele Wörter schon gehört. Sind so ein bisschen mental in dem Thema drin. Ähm also das ist/ das ist eigentlich immer Nummer eins. (TD1 00: 25: 53) The only other oral English input observed in teachers’ classes not directly from teachers came in the form of a song used during a warm-up routine in Teacher E’s CLIL lessons. While providing large amounts of English language input is central to teachers' CLIL instructions, they also described particular contexts in which they feel German input, either in oral or written form, would be more appropriate. Teachers described using German input for managing linguistic or conceptual complexity, providing pastoral care (i.e. managing 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 159 <?page no="160"?> social and emotional situations), teaching local and personal topics, and saving instructional time. Teachers described using German primarily for the purposes of managing content and language complexity. This often came in the form of the German-speaking teacher explaining the content of some part of the lesson in order to clarify anything students did not understand or that they misunderstood. Teachers found some topics more challenging than others such as topics that were more text-based and that required the use of more words that are unknown to students. In these situations, German was used to secure a basic understanding of content. Teachers warned, however, that regularly translating immediately into German could lead students to stop trying to understand English and become “lazy.” I don't think everyone in the class needs the German input. Um, and sometimes the kids might be a little bit too lazy to really understand it in English. Like, "Oh the German translation is coming in a minute. I'm not listening now." (TB1 00: 31: 21) Instead of automatically translating spoken English into German, teachers looked for signs from students that they were not understanding. In this way, English would be used as much or for as long as possible and would be translated by the German-speaking teacher or another student if necessary. Like I would start in English and see how far they get it↑ And as soon as you see blank faces the other one [German-speaking teacher] would step in. Or maybe I would have somebody [another student] translate for the other children. (TB2 00: 06: 20) Teachers also described language-heavy topics and particular phases of lessons such as providing feedback to a presentation that might require the use of German. In these phases, German input was used to go deeper into the content of the message and to secure understanding. Wenn es sehr textlastige Themen sind oder welche die nicht so greifbar sind, […] wo man merkt, da geht's jetzt nicht mehr so richtig in die Tiefe weil die Sprachlichen Mittel fehlen oder das Verständnis fehlt, auch da würde man dann auf Deutsch umswitchen beziehungsweise wurde der Kollege dann da übernehmen, damit man so ein bisschen mehr noch aus den Kindern rausbekommt bei eine Reflexion oder bei einem Feedback zu einer Präsentation beispielsweise. (TD1 00: 08: 46) Most of the written input that students received was in German in the form of worksheets. English worksheets, when used, were either published originally in English or had been translated by the teacher from the German original. Teachers stated that the decision to translate a German worksheet into English is based on the complexity of the worksheet. German worksheets with simple content and language may be translated into English. More complex worksheets would likely not be translated into English, since an 160 6 Results <?page no="161"?> English translation that students are capable of understanding would reduce the content and complexity of the German original. But as soon as there's too much/ too much ehm written on it [a worksheet], or too complicated [German] sentences, I wouldn't translate them. Because then the translation doesn't make sense. And I make short sentences out of it and it's not what it's supposed to be. (TB2 00: 26: 44) While oral German input is used to clarify meaning and explain or discuss more complex content, written German input is used to maintain a degree of complexity that already exists in the professionally published materials. Teachers also described using German for purposes of pastoral care. Such situations were primarily described as emotionally charged situations such as students’ arguments with one another or personal issues but also situations where discipline and the safety of the students was a concern. When using German to manage situations involving the emotional and physical well-being of students, teachers described switching from speaking English to speaking German with students. This differs from how they described using German to manage linguistic and content complexity, in which the German-speaking teacher or another student translated or mediated the message. Teachers who described speaking German to students for the purpose of pastoral care sometimes also described either ambivalence about speaking German with their students or a change in their thinking over time about speaking German with their students. Teacher B, for example, described “really try[ing] to stick to the language which isn't always that easy” (TB1 00: 08: 59). It is not easy to always use English because of situations involving the students’ safety and well-being. [B]ecause sometimes it's more important to quickly get a message over to the kids than to / like if somebody's hurt or if there's a fight I need to be there and I need to be able to get my meaning across. (TB1 00: 09: 20) Teacher A describes an evolution in her thinking about the necessity to only speak English with students. While she used to think that speaking German with her students was a problem, she no longer thinks this. Based on her experience, students do not learn less English if she occasionally switches to German in order to manage emotional situations. I thought this is a problem [speaking German with students] at the beginning. But now when I look back I think the children don't care. Ja wenn sie emotional/ in eine emotionale Situation denn denken sie nicht darüber nach. Dann sind sie/ müssen sie aufgefangen werden und das geht besser in der Muttersprache. Und ich glaube nicht, dass sie dadurch weniger Englisch lernen. Ja. (TA1 00: 29: 33) In these moments of switching into German for the purposes of managing social and emotional situations, teachers’ priorities change. No longer is the goal to develop foreign language competences. Instead it is to “calm everyone down and to solve the situation” (TC2 00: 40: 06). Teacher C’s view regarding speaking German with 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 161 <?page no="162"?> students was unique in this study. She described speaking German with students in such situations not only as permissible but as an opportunity to develop relationships with students. I want them to see me as somebody who speaks English with them but also somebody who is capable to switch to German if ehm if the situation needs it. If they have, I don't know, private problems and they would not come to me because they would feel like, "Ok, she just speaks English and I'm not comfortable with that," that would make me sad. So they should know that in certain situations, yes, we can speak German, ehm, and in certain situations I think it's definitely the better idea. (TC2 00: 45: 35) Teaching local and personal topics was also a reason teachers gave for using more German with students. Examples of such topics include learning the 16 federal states of Germany, the bicycle safety course and exam required of primary school students, and sexual education. In the example of the bicycle safety course, teachers said it was more important that students understand the local safety rules in German. Students should also be able to use the curriculum materials that regional authorities have published for the course even if the teacher continues to speak English with the children. Like when we have in class four, we're practicing for the um, for the bike driving test and everything, all the material and the test they have to take in the end to really do the practical stuff is in German. So, most of the time we do that in German. I might talk in English about it with them. But still all the worksheets, all the material is in German. (TB1 00: 05: 21) When teaching a topic such as sexual education, what Teacher B described as a “sensitive topic” (TB2 00: 07: 30), it is more important that students can focus entirely on understanding the content and asking questions about it. And also if you talk about, I don't even know, Sexualkunde, we do mostly in German. Because it's a very sensitive topic and you don't want to stop the children from really being honest and talk about it and ask questions. (TB2 00: 07: 27) One teacher also reported using German with students in order to save instructional time. Though not a primary source of English input, written English input was used in teachers’ approaches to CLIL instruction. As already described, some of the English input came in the form of worksheets. Teachers also described using other English language materials such as picture books and games. When possible, teachers reported preferring “authentic materials” (TD1 00: 23: 47) or materials made “for English natives” (TB1 00: 47: 17). These materials were not provided to teachers but were collected by teachers independently and often from online sources or while traveling abroad. The central challenge described by teachers, however, to using authentic materials in CLIL lessons was balancing content and language complexity. Teachers felt that authentic 162 6 Results <?page no="163"?> materials that contained appropriate levels of content were often linguistically too complex for students. Simplifying the language in order to be appropriate to students’ foreign language competences would risk compromising the level of content. Well the problem with the material is, of course I can get material that's for English natives. But then it might be too complex. But then on the other hand if I break it down, well it might be too complex language-wise, but if I break it down it might be too easy concerning the content. So you always need to find a balance between the language and the content. (TB1 00: 47: 17) Teachers also described using materials that allowed students to work independently. Teachers described using such materials as part of station work or independent work time. Key to achieving independent use of materials is the use of simple language in, for example, instructions. Teacher D described evaluating materials such as English language games purchased abroad through the lens of what students can do alone. Am Anfang habe ich sehr viel gekauft und habe mehr das versucht, alles irgendwie zu besorgen. Ähm, und mittlerweile gucke ich/ gucke ich sehr viel genauer, was macht wirklich Sinn, was können die Kinder alleine? Viel Material, Spiele beispielsweise, die erst eine ganz, äh, ganz lange Gebrauchsanweisung oder eine Spielanleitung dabei haben, machen zum Beispiel für Freiarbeit nicht so viel Sinn. (TD1 00: 23: 47) For station work, where students need to work with a degree of independence, appro‐ priate materials were described as being one of the deciding factors in the success of such lesson forms. For Teacher C, appropriate materials can be in English or German. Decisions to translate materials into English are based on teachers’ assessments of what students can manage in English. We did this/ (…) experiments and the topic was electricity. That went well because that was a grade four and the material we had was in German so we sat down and translated everything we thought would work in this class in English. […] Um, I think it went well because of the material. (TC1 00: 18: 15) For teachers in this study, providing large amounts of English language input was a key aspect of how they thought of their CLIL instruction. Providing more English input would lead to students learning more English. While other sources of English language oral input were described and observed, teachers emphasized the importance of their own role in providing the majority of the oral input. Providing German input could be used primarily to manage complex content and language in the form of translation or mediation but also for social and emotional situations as well as for local and personal topics. The use of written English input was also described by teachers as playing a role in their CLIL instruction and observed in teachers’ CLIL lessons. However, its use was not central to teachers’ thinking about how best to approach CLIL instruction. 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 163 <?page no="164"?> 6.3.2 Know the students Code Exemplary quote Coded segments Teachers Know the class Also ich glaube muss ich wirklich die Gruppen angucken und so ein Gespür dafür entwickeln, was kann ich ähm/ was kann ich da schon rausholen und wo ähm/ wo überfordere ich aber so, dass es vielleicht ähm ja den spaß verdirbt beziehungsweise die Anstrengung zu/ zu groß wird. (TD1 00: 17: 48) 14 5 Know in‐ dividuals Aber ich ähm/ Ich versuche sie schon zu ermutigen. Manchmal/ Bei manchen Sachen also dann sage ich auch wirklich, wenn jetzt jemand Deutsch redet, dann sage ich auch "Kannst du das nicht auch auf Englisch sagen? " wenn ich weiß das ist etwas was das Kind auf jeden Fall auch kann. (TE2 00: 26: 54) 13 5 Table 17: Know the students Teachers repeatedly described basing their planning and other instructional decisions on their knowledge of students. They often described using this knowledge by qualifying various responses to questions. Instead of, for example, answering, “Yes, I do X” or “No, I don’t do X,” they often used phrases such as “If I know the child could[…]” (TB2 00: 15: 00), “wenn ich weiß, dass die Kinder das leisten können” (TD1 00: 13: 15), or “wenn ich weiß das ist etwas was das Kind auf jeden Fall auch kann” (TE2 00: 26: 54). Phrases like these suggest not only that teachers’ knowledge of students is an important component of how they implement their CLIL instruction but that it acts as a filter through which they make decisions about planning and instruction. Teachers described using their knowledge of students in two ways. First, they used their knowledge of the whole class when planning lessons and units of instruction. Teachers’ described using their knowledge of the class’s overall English competences as well as their knowledge of students' social competences. For example, Teacher D describes using her knowledge of the class’s English competences to determine how much and which content might be possible in English. For Teacher D, her knowledge of any particular group of students is built on her years of teaching experience and her ability to estimate students’ competences. She uses her knowledge of students when reflecting on and revising previously planned CLIL units of instruction. Also generell natürlich zum Unterrichten nach so einer langen Zeit weiß man, dass man/ dass man immer sehr genau auf die Lerngruppe gucken muss↑ Was kann man von der Lerngruppe erwarten? […] Also ich glaube muss ich wirklich die Gruppen angucken und so ein Gespür dafür entwickeln, was kann ich ähm/ was kann ich da schon rausholen und wo ähm/ wo 164 6 Results <?page no="165"?> überfordere ich aber so, dass es vielleicht ähm ja den spaß verdirbt beziehungsweise die Anstrengung zu/ zu groß wird. (TD1 00: 17: 48) Teachers also described using their knowledge of students’ social skills when planning lessons and units of instruction. For Teacher C, her knowledge of her class led her to limit the types of activities she planned for them. Despite preferring to use partner and group activities with students, she described these types of activities as not possible with her students. Ehm, I personally like having children learn with a partner or in groups. I think, if it comes to me, I would/ that would be a learning for I would provide for them every day and every lesson. But, in this learning group, it's just not possible. (TC2 00: 34: 15) For these teachers, planning for and implementing CLIL was incomplete without knowledge of students. Second, teachers described using their knowledge of individual students to individualize instruction. They described using this knowledge for pairing students to work together, for planning activities that will appeal to students’ different interests, and for making decisions about when to push or support students to speak English in class. Teachers most often described using their knowledge of individuals to make spontaneous and interactive decisions about when and with whom to ask or encourage to speak English. Teacher B’s statement below illustrates how she uses her knowledge to differentiate speaking goals for different students. She would ask students with stronger English competences to use more English than those with weaker English competences. She fears that asking students with weaker English competences to use more English would demotivate them. / / Well, it really depends on the child/ / Like I really would have to see like, [Student 1] is very strong in English and if he talks to me in German I would, depending on what he wants to say in German, of course, if I know, or if I'm pretty sure he would know it in English, I would maybe ask him to say that again in English. But there's other kids that would give up if I asked them to say it in English. They would "Ah, never mind." But it, yeah, it's really very individual from child to child. (TB2 00: 34: 01) In the context of making decisions about which students to encourage to speak English and which to allow to speak German, several teachers described being guided in part by a fear of silencing or discouraging students who are not comfortable speaking English. As mentioned above, this is evident in Teacher B’s statement when she says, “But there's other kids that would give up if I asked them to say it in English. They would “Ah, never mind”” (TB2 00: 34: 01). While one teacher described using her knowledge of individuals to push students to their limits (“an ihre Grenzen zu bringen” TA5 00: 22: 52), the fear of pushing beyond a particular student’s limits or comfort level seemed to play a significant role in guiding Teacher D’s thinking about when to push any particular student to speak English. Teacher 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 165 <?page no="166"?> D describes using not only knowledge of a student’s English competences but knowledge of a student’s personality as well when deciding whether or not to ask a student to use English in a particular moment. For shy students, for example, Teacher D prefers to find moments in which the student is more likely to be successful when using English. She hopes that such success leads students to using more English in the future. Ähm, bei manchen Kindern würde es vielleicht sogar auch dazu führen, dass sie dann verstummen und lieber gar nichts mehr sagen wenn/ wenn sie wissen, "Ich muss das aber jetzt auf Englisch können, dann melde ich mich lieber gar nicht." Also da ähm gerade bei schüchternen Kindern bin ich da so ein bisschen vorsichtig. Da erlaube ich lieber das es auf Deutsch sagen und ähm warte bis der Zeitpunkt gekommen ist oder gebe ihn eher einfachere Situationen wo sie sich nochmal ähm/ nochmal selbst testen können und selbst sehen können. "Ah, das kann ich eigentlich doch schon. Ich kann es das nächste Mal schon probieren." Also so ein bisschen individuell. (TD1 00: 13: 15) In the ways described above, actions that teachers took in the classroom such as when to encourage students to speak English and when to allow them to use German were contingent upon their evaluation or knowledge of individual students' various competences. Teachers’ knowledge of the whole class and of individuals acted as a filter through which planning and interactive decisions were made. 166 6 Results <?page no="167"?> 6.3.3 Scaffolding Code Exemplary quote Coded segments Teachers Scaffolding input Translation German teacher transla‐ tion Die Deutsche Kraft kommt eigentlich nur dann ins Spiel, wenn man merkt, es gibt irgendwo Verständnis Schwierigkeiten. Mann kommt mit der englischen Sprache jetzt gerade an der Stelle nicht weiter. Es fehlen wichtige Begriffe oder die sind nicht/ nicht besonders gut darstellbar, weil es nur in dem Text steht und man hat kein Bild dazu. Man kann das nicht vorspielen, beispielsweise. Da würde die deutsche Kraft dann einspringen. (TD1 00: 07: 07) 18 4 Student trans‐ lation Weil wir die Befürchtung haben, dass die Kinder sich dann zurücklehnen und sagen, "Oh, er sagt's ja gleich nochmal auf Deutsch und eh brauche ich jetzt noch nicht zuhören." Also, das versuchen wir zu vermeiden. Wenn übersetzt werden muss, dann durch ein Kind. (TD 1 00: 08: 24 13 4 Visual strat‐ egies Well, to really have the kids understand what you're talking about means that you need to use a lot of body language and miming, pictures, um and really try to stick to the language which isn't always that easy. (TB1 00: 08: 59) 23 5 Language strategies in L2 Simplify lan‐ guage And the language we use when we do/ because it is a written test obviously, the English we use and the sentences we use, we try to make them as simple as possible. (TC1 00: 44: 02) 12 4 Repeat and reuse language Und ähm man versucht, nicht in die Deutsche Sprache zu übersetzen, sondern man bleibt im Englischen ganz stur. Und ähm man macht sich verständlich, indem man alles ganz oft wiederholt und vormacht. Also selbst bei den einfachsten Sätze am Anfang, “Take your folder out.” Macht man es. Und macht die Aktion und sagt es dabei. Und das immer wieder, bis alle wissen, was dieser Satz bedeutet.(TA5 00: 10: 33) 8 4 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 167 <?page no="168"?> Explicit com‐ prehension checks [Ä]hm und dann kann es auch sein, wenn dann zum Beispiel ein bestimmtes Wort in diesem Satz irgendwie neu war, was jetzt erklärt wurde, dass ich das dann noch einmal kurz auf Englisch erkläre und die Kinder nachsprechen lasse und auch frage "Weiß jetzt jeder, was das ist? " und so. (TE2 00: 05: 27) 2 1 Materials strategies Use diction‐ aries […] or I want you to use a dictionary or I want you to use a computer to find out. And if it helps you to translate that word, if it helps you to get the content, then do that. (TC1 00: 48: 42) 6 2 Create struc‐ tured and dif‐ ferentiated worksheets Like I sometimes feel like I don't even know what to do and where to help when kids come to me with like a big [German] text, and like this is/ like what am I supposed to do? Ehm, I think with the English ones, it's/ since there is a certain structure to it, I believe it's easier for them. So, again, structuring worksheets, talking about it, taking the time to do that and not relying on just because they're in fourth grade now, eh relying on being ab/ ehm them being able to do it by themselves is definitely important. (TC2 00: 08: 25) 4 1 Interactive strategies Of course they could do the listening alone. But then sometimes it also helps when you have a partner just to talk about it. "I think it's that. No it's that" So. (TB2 00: 16: 58) 5 2 Scaffolding output Encourage chil‐ dren to speak Yeah. Because I think, sometimes the kids get lazy. Because ehm, I want to get it out of them. They have so much English in them and then I really/ If I know the child could form a sentence, why wouldn't I ask them for it? Because that's also, yeah, enlarging their ability. (TB2 00: 14: 59) 13 4 Patterned output Then we would say/ I would try to have them say, "The pencil floats", or, "The pencil sinks." So that's/ because they see it and they know what it means. It's not that nobody understands if they say it in English. (TB1 00: 51: 09) 5 4 168 6 Results <?page no="169"?> Known methods and vocabulary from EFL les‐ sons But depending on the topic, depending on their vocabulary and maybe sometimes it's vocabulary they learned/ they know from English lessons. So um, like if you talk about animals and they have to make a report or a poster about their favorite farm animal or pet or whatever, I know that they know a lot of words they can use from the English lesson. So I would ask them to, if they feel comfortable with it, to have a presentation in English of course. (TB1 00: 42: 56) 5 1 Student inter‐ action Ehm, it was also very important for me to/ making sure that they're not sitting all by themselves, but that they have a group or a partner to talk about it, so that they can gain just from each other's English abilities. (TC2 00: 19: 35) 4 1 Set achievable goals Da erlaube ich lieber das es auf Deutsch sagen und ähm warte bis der Zeitpunkt gekommen ist oder gebe ihn eher einfachere Situationen wo sie sich nochmal ähm/ nochmal selbst testen können und selbst sehen können. "Ah, das kann ich eigentlich doch schon. Ich kann es das nächste Mal schon probieren." Also so ein bisschen individuell. Aber natürlich versuche ich rauszukitzeln. (TD1 00: 13: 15) 3 3 In the language Most children would respond in German. But sometimes, like if they're really on an English worksheet and in the language and like lost in the language, they would answer in English. They wouldn't do that intentionally. It's just because/ they wouldn't think about, "Oh, it's [Teacher B], I have to say it in English now." (TB2 00: 32: 30) 2 1 Table 18: Scaffolding With a focus first on providing large amounts of English input, teachers sought to scaffold that input in various ways. Teachers described a range of strategies they used and felt were beneficial to helping students understand both content and language. Most often, they described using translation and visualization to support students’ understanding but also described strategies, for example, based on materials and interaction. Translation was a strategy that all five of the teachers reported using. While translation is not typically considered a scaffolding technique in that it simply replaces one language with another and therefore reduces the need for students to process understanding in 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 169 <?page no="170"?> the L2, teachers often reported using translation to ensure conceptual and instructional understanding for all students. In these teachers’ CLIL contexts, translation, or more generally the use of German, was used to support content and language learning They described the use of translation both as an option of last resort and as a technique for maintaining the flow of instruction. Because teachers felt it was important that they only spoke English with students, they reported that translation usually took place through the German-speaking teacher in the class or through a student. Teachers often described using the German-speaking teacher only after understanding had broken down. For many of the teachers, the German-speaking teacher’s role is a kind of insurance policy for understanding. The English-speaking teacher would provide as much of the instruction in English as possible and only when the teacher feels that students no longer understand either the concepts being described or the instructions for a particular task would the German-speaking teacher step in. [A]lso in meinem Fall ist das [German-speaking team teacher] die dann mit mir zusammen mit im Sachunterricht ist, und die greift eigentlich nur ein wenn wir merken die Kinder verstehen was nicht erst auf Englisch und es muss nochmal auf Deutsch erklärt werden oder, ja, wenn die Kinder jetzt spezielle Fragen haben. (TE1 00: 03: 58) This kind of intervention from the German-speaking teacher was sometimes referred to by one of the teachers as using the “German joker.” The term describes, with some humor, situations in which teachers feel they have no other options, no other way to achieve the goal of student understanding other than using German. Using the German-speaking teacher, however, did not always mean providing a direct translation of the English for the students. It was also described as an opportunity simply to switch the lesson into German momentarily and allow the German-speaking teacher to reestablish understanding among the students by, for example, posing questions that led students to understanding. In this way, it was reported that students would remain engaged in trying to understand the message and not simply wait for a German explanation. Providing a German translation or using the “German joker” too soon could have the effect of preventing students from trying to understand what the English-speaking teacher is saying. According to Teacher D, using the German-speaking teacher had the added effect of saving instructional time. Und der [German-speaking teacher] hat dann zwischendurch auch gesagt, “Wisst ihr genau was ihr machen müsst? ” (laughs). (inaudible) Also er übersetzt nicht meinen gesagten Teil, sondern stellt eigentlich nur nochmal Nachfragen, dass sie selber so ein bisschen näher an die Lösung rankommen. Denn wir wollen ja nicht, dass sie sich zurücklehnen und warten, bis er die Erklärung auf Deutsch dann liefert. (TD2 00: 42: 15) In most of the lessons observed, the German-speaking teacher took a secondary role. As described by Teachers D and E above, the German-speaking teacher intervened when there was a breakdown in communication and understanding. However, as described above in the introduction to chapter six, this was not always the case. In Teacher A’s CLIL lessons, her German-speaking team teacher, who was also the class teacher, often 170 6 Results <?page no="171"?> intervened with a translation immediately following Teacher A’s English input. Teacher A also never expressed the idea, as several other teachers did, that immediate translation could inhibit students’ efforts and opportunities to learn English. Teachers also described students as potential translators in class. As opposed to using the German-speaking teacher to repair more fundamental breaks in comprehension, teachers described using students as translators in order to check and ensure understanding while maintaining instructional flow. Students were often used to translate individual words or task instructions. Using students as translators avoided the need for the English-speaking teacher to switch into German, thus maintaining the principle of one person-one language. Teachers described being aware of the English competences of their students and using “the fit kids” (TB2 00: 14: 15) to translate for the rest of the class. Ahm, versuche ich immer wieder auch übersetzen zu lassen, "Was habe ich jetzt gerade gesagt? ", "Wer kann es nochmal sagen? " Ich bin da immer wieder erstaunt, wie gut die Kinder wirklich das eins zu eins tatsächlich wiedergeben können, was ich gerade gesagt habe. Also es sind in/ in jeder Klasse eigentlich immer so ein paar Kinder, auf den man sich verlassen kann, die man heranziehen kann zu übersetzen, wenn man in, äh, ratlose Gesichter schaut. (TD1 00: 05: 25) In addition to translation, all of the teachers described using various visualization tech‐ niques to support students’ understanding of English. Teachers most often described using pictures and body language including miming and demonstrations to help communicate meaning. While translation is used to check and repair student understanding, teachers described visualization as a tool to communicate meaning and maintain understanding throughout the lesson. Zwischendurch lässt man übersetzen. Aber meistens ist es mit zeigen, hochhalten, “packen das in die Postmappe,” “morgen,” also ganz viel Körpersprache, ganz viel, ähm, ja, ganz viel mit Hand und Fuß erklären, um sich verständlich zu machen. (TD1 00: 06: 10) Visualization techniques, as Teacher C describes below , are a part of good EFL lessons and work just as well in CLIL lessons. I think good bilingual teaching also means using what you use in good English classes. Is, I don't know, a lot of cliparts, a lot of visual stuff to get the children. (TC1 00: 11: 38) However, beyond this standard form of visualization that is used in EFL lessons, Teacher B described using experiments specific to primary school science lessons as a form of visual scaffolding. Concrete language can be used to describe what students see happening in the experiment, thus supporting students’ foreign language development in the context of learning simple scientific concepts. I was just thinking, in grade two we just did what floats and what sinks. […] We did it with the whole group but then we would say/ I would try to have them say, "The pencil floats," 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 171 <?page no="172"?> or "The pencil sinks." So that's/ because they see it and they know what it means. It's not that nobody understands if they say it in English. (TB1 00: 51: 09) Teachers also described using several L2 language strategies to scaffold understanding. One strategy they described was to simplify the input. Teachers described simplifying both oral and written input. When simplifying oral input, teachers described the need to carefully consider how to structure sentences when explaining content as well as the need to reformulate input when students do not understand. Ich merke ja dann auch wenn ich jetzt mit einem oder mehreren/ wenigen Kindern alleine spreche eigentlich sofort ob das Kind mich versteht oder nicht und kann dann auch direkt halt, wenn es mich nicht versteht, dann darauf reagieren indem ich dann noch einmal was ich gesagt habe irgendwie umformuliere, das noch einmal einfacher sage oder so und kann dann so besser auf das Kind selbst halt eingehen. (TE2 00: 25: 20) When simplifying written input, teachers described using simple words and sentence structures when creating, for example, worksheets as well as only using English for simple content and tasks. This corresponds with the previously described role of English, especially written English, as being reserved for simple content and tasks that require only simple language. An exception to this tendency, however, was Teacher C’s description of simplifying texts. Teacher C used texts to create linguistic challenge. She simplified texts to calibrate that challenge. Simplifying texts, for Teacher C, was part of an experimentation and revision process. Teacher C revises texts that she would use in the future based on the understanding that students demonstrate in previous lessons. So I thought, I'll just give it a shot. I'll just have them read these texts, and, to be honest with you, the texts I had before were definitely more complex. So what I did is I tried to simplify them as much as possible and then give it to them. And then, as it turns out, obviously there's still a few things they didn't understand right away which is ok, but for the next time I would just simplify a little bit more. (TC2 00: 19: 35) Teachers also described the importance of repeating and reusing language in order to support students’ understanding. Teachers felt it was important to repeat language over the course of multiple lessons and when doing so, use only one formulation with students. Reformulating sentences would just lead to confusion. And try to use the same words for the same action. Like if I tell them to put their worksheet in their folder they really/ I need to use the sentence all the time. I cannot rephrase it so they just/ "Oh yeah, she said that before. I know what it means." (TB1 00: 18: 29) One teacher also described explicitly checking students’ comprehension after, for example, explaining a new word. 172 6 Results <?page no="173"?> Ähm und dann kann es auch sein, wenn dann zum Beispiel ein bestimmtes Wort in diesem Satz irgendwie neu war was jetzt erklärt wurde, dass ich das dann noch einmal kurz auf Englisch erkläre und die Kinder nachsprechen lasse und auch frage "Weiß jetzt jeder was das ist" und so. (TE2 00: 05: 27) Teachers also described using strategies that involve the use of various materials to support students’ understanding. Teachers described using various forms of dictionaries to help students understand specific content words at the start of a new unit as well as a resource when teaching. For Teacher B, the use of dictionaries or even posters which display new content words and pictures was a strategy used to meet one of her central planning considerations of how to convey meaning to students. Sometimes you will have a lesson where/ or that where they would have like a little dictionary for one content or one unit. When we did electricity in class four last year we would have like the poster with the different picture and word in English so they could also understand because some of the worksheets were in English I remember. So they would understand what it means and they still have some help. (TB1 00: 52: 41) One teacher also described the importance of well-structured English worksheets to support students’ understanding of the content and task presented on the worksheet. Ehm, I think with the English ones [worksheets], it's/ since there is a certain structure to it, I believe it's easier for them. (TC2 00: 08: 25) As mentioned in the statement above and described further in the statement below, a well-structured English worksheet was most successful when it was combined with an interactive strategy of explanation. Teacher C describes a situation in which, from her perspective, German language worksheets could sometimes be successfully completed without attention to the structure of the worksheet and a careful explanation of the worksheet. Teachers could rely instead on students’ German language competences to understand and complete the task. CLIL teachers, when creating and introducing English language materials needed to consider the structure and explanation of a worksheet in order to scaffold understanding. I felt like the English ones were more structured and they knew what to do. Maybe also because we talked about them. It's not like, "Ok, here's your worksheet. Do your job." It's always/ every time we do something in English it's usually/ we talk about it. […] And then they do it. With the German stuff, ehm [the German speaking teacher] can rely usually on the kids ehm doing it on their own. Like reading the text and all that stuff. But although she can rely on that, I still felt like they needed a lot of help. (TC2 00: 02: 39) Teacher C suggests that the use of English language worksheets prohibits teachers from assuming that students will understand, thus forcing teachers to carefully consider the 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 173 <?page no="174"?> language and structure of the worksheet itself as well as the introduction of the worksheet and explanation of the task. From Teacher C’s perspective, English language worksheets that have been carefully structured and explained are potentially easier for students to complete despite being written in a foreign language. Teachers also described scaffolding students’ own English language production in var‐ ious ways. As recorded in my classroom observations, students often spoke German during the CLIL lessons. However, there were moments, often during whole-class instruction, when students did speak some English. Chief among teachers’ strategies for scaffolding students’ spoken English was encouraging children to speak English. When they knew that a student was capable of speaking English in a particular context, teachers described “try[ing] to get it out of them in English” (TB1 00: 41: 47) or motivating them to speak English. Such encouragement and motivation could come not only from the teacher but from other students as well. ähm ist es denke ich wichtig, dass jemanden hat, der so ein bisschen anspornend. Und entweder sind das andere Kinder in der Klasse und das passiert von selbst, dass man sagt, "Oh, der hat das jetzt auch gemacht. Dann schaffe ich das auch." Oder halt ich immer noch als Motivator der ähm, der versucht, das rauszuholen. (TD2 00: 23: 22) While teachers described encouraging students to speak English, they were explicit about not requiring students to speak English in the Science and Social Studies lessons. As described previously in this chapter, they did not want speaking English to interfere with learning non-language content. Furthermore, they viewed language learning, especially speaking, as an emotionally entangled process. Forcing a student to speak English when the student is unwilling or unable to do so could negatively affect the student’s relationship to the English language. Dann hätte ich Sorge, dass sie ganz verstummen und ähm auch das Englische dann ablehnen als was schwierig ist, was sie nicht können, was sie aber müssen und das soll kein Zwang sein. (TD2 00: 37: 06) In addition to encouraging certain students to speak English in certain situations, teachers described a range of lesser approaches or techniques that they felt could be used to support students’ oral and written English production. For example, several of the teachers described using patterned output or specific learned phrases and chunks to support students’ English production. Teachers wanted students to speak English when using classroom phrases such as asking to go to the bathroom but also described teaching language chunks related to content learning and asking students to use those phrases. Teacher B, for example, had students repeat certain sentence structures replacing individual words as necessary. But I didn't see any point to translate those sentences because they're pretty easy and they can remember them easily, because it's always the same sentence↑ structure. They just need to replace the verb or the organ they're using. (TB2 00: 14: 00) 174 6 Results <?page no="175"?> Another approach described by one teacher was planning more English language activities, including those involving speaking and writing, when she knew that students already knew related English vocabulary and methods from EFL lessons. In this way, the teacher could adjust her language use expectations based on students’ preexisting knowledge and experience. So um, like if you talk about animals and they have to make a report or a poster about their favorite farm animal or pet or whatever, I know that they know a lot of words they can use from the English lesson. So I would ask them to, if they feel comfortable with it, to have a presentation in English of course. (TB1 00: 42: 56) Another teacher emphasized the role of student interaction in supporting students’ English production. For Teacher C, learning language is best learned by using language and working with partners or in a small group. Such interaction is an opportunity for students to speak English and to receive feedback from other students on their spoken English. So I think yes, eh social form/ different social forms, with a partner, in small groups, is definitely one big/ ehm big aspects of learning a language. Because that's the time where you use the language or where they should use the language, ehm that's the time where they can try it out and that's the time where they also correct each other without me even saying something. (TC2 00: 38: 11) While the general pattern of students often speaking German with one another while working on tasks was also true for Teacher C’s class, there were occasions in which students spoke English to one another or responded in English in whole-class situations. It would seem at least possible that students in Teacher C’s class could interact in English and therefore develop their speaking skills as described by Teacher C above. Setting achievable speaking goals was also described by teachers. Teachers described, for example, requiring students to only use English for a portion of a poster presentation or waiting for an easier situation in which the student could speak English. Da erlaube ich lieber, dass es auf Deutsch sagen und ähm warte bis der Zeitpunkt gekommen ist oder gebe ihn eher einfacherer Situationen wo sie sich nochmal ähm/ nochmal selbst testen können und selbst sehen können. (TD1 00: 13: 15) Finally, one teacher described students being more likely to speak English if they are engrossed in an English language task such as an English worksheet. Teacher B did not describe planning such tasks for the purpose of supporting English production. But sometimes, like if they're really on an English worksheet and in the language and like lost in the language, they would answer in English. They wouldn't do that intentionally. It's just because they wouldn't think about, "Oh, it's [English-speaking teacher], I have to say it in English now." (TB2 00: 32: 30) 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 175 <?page no="176"?> 6.3.4 Motivate students Code Exemplary quote Coded segments Teachers Activities Varied Also bei der Materialauswahl (laughs) versuche ich darauf zu achten dass ähm unterschiedliche Sachen dabei sind dann für unterschiedliche Vorlieben, unterschiedliche Lerntypen und so, dass die Inhalte noch einmal auf unterschiedliche Arten gefestigt und erlernt werden können. Ja. (TE2 00: 17: 24) 10 4 Playful and active Ähm also dieses Ritual dass wir das Europa Puzzle zusammengesetzt haben, haben wir uns überlegt weil es einfach einer spielerische Aktivität ist. Kinder lieben Spiele, lieben Puzzle, lieben Rätsel, ähm und auf eine spielerische Art und Weise wird etwas immer wieder trainiert. (TD2 00: 07: 52) 10 3 Authentic situations Um, good bilingual teaching gets the children. Like, motivates them to learn the language by not actually have English classes. […] By creating authentic situations where you can use the language apart from this whole English lesson thing where you/ where everything is kind of constructed. (TC1 00: 11: 38) 1 1 Social support Peer role models Ähm ist es denke ich wichtig dass jemanden hat, der so ein bisschen anspornend. Und entweder sind das andere Kinder in der Klasse und das passiert von selbst, dass man sagt, "Oh, der hat das jetzt auch gemacht? Dann schaffe ich das auch." (TD2 00: 23: 50) 4 1 Working together I like lessons that are designed like that where we all work on different things and then we put one thing together. And if that also means they can present something, that's even better. So having as a group, that is my class, having something that we all worked on I think is ehm motivating for the kids and this is something we've done before and I just like to choose that way of learning something. (TC2 00: 19: 35) 4 1 176 6 Results <?page no="177"?> Relation‐ ship with teacher So I think personal situations and not only one on one but also in smaller groups, ehm laughing together, creating situations where you are just personal with each other and you can address more personal, those are situations where they learn the language and where they're/ where they, I don't know, are more motivated and ehm, confident enough I would say. (TC2 01: 03: 55) 2 1 Table 19: Motivate students Teachers also described the importance of motivation in their CLIL lessons. Though they often described most students as already being motivated when coming into their CLIL lessons, they also described ways they felt they could maintain and build motivation in CLIL. They described doing this primarily through varied and playful activities as well as through various forms of social interaction and support. Planning different kinds of activities as well as activities that offer variety and choice to students was one way that teachers tried to motivate students. For example, teachers described using books, songs, and drama for the purpose of offering children more types of activities and to have fun (“den Kindern dann schon so ein bisschen mehr bieten kann und die/ den auch mehr Spaß machen einfach” TD1 00: 25: 53) as well as activities such as lapbooks which offer students choice. Teachers described making their choices of activities based, at least in part, on offering variety to meet the various interests of students. Also bei der Materialauswahl (laughs) versuche ich darauf zu achten dass ähm unterschiedliche Sachen dabei sind dann für unterschiedliche Vorlieben unterschiedliche Lerntypen und so, dass die Inhalte noch einmal auf unterschiedliche Arten gefestigt und erlernt werden können. Ja. (TE2 00: 17: 24) Teachers also described playful activities as motivating for students. Playful activities included activities such as games, puzzles, and riddles as well as science-specific activities such as doing experiments or simply trying things out. Teacher D, for example, used a routine of assembling a large map of Europe during several of her lessons during the unit on Europe. Assembling the map was combined with the students making statements in German and in English about the location of various countries. For example, students said, “Finnland liegt westlich von Russland” and “Ireland is west of Great Britain.” In later lessons, the teacher timed the class’s assembly of the map. For Teacher D, these types of activities teach skills and knowledge in a playful context that is fun for students. Ehm also dieses Ritual, dass wir das Europapuzzle zusammengesetzt haben, haben wir uns überlegt weil es einfach einer spielerische Aktivität ist. Kinder lieben Spiele, lieben Puzzle, lieben Rätsel, ähm und auf eine spielerische Art und Weise wird etwas immer wieder trainiert. (TD2 00: 07: 52) 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 177 <?page no="178"?> In addition to planning a variety of activities and fun or playful activities, teachers described various forms of social interaction and support as motivating for students. One teacher described the role that more competent English-speaking students played in motivating other students to speak English in class. By speaking English and setting an example for other students, these so-called “draft horses” (Zugpferde) could motivate others to use English. Teacher D feels a sense of success when she sees students being motivated to speak English by other more competent or willing English-speakers in her class. Manchmal hat man so ein Paar, die vielleicht eine Native Hintergrund haben oder die besonders ambitioniert sind. Die versuchen, sich auf Englisch auszudrücken. Und dann ziehen andere mit. Also wenn man so Zugpferde hat und man merkt so, das geht auf die Klasse über, dass sie/ dass sie versuchen, sich auf Englisch auszudrücken. Das ist für mich eine Stunde, wo ich denke, “Ah, jetzt ähm/ jetzt habe ich es geschafft.” (TD1 00: 11: 50) Another teacher emphasized the role of students working with a partner or in a small group. For Teacher C, working together is more motivating and leads to better outcomes. However, despite preferring such social groupings, Teacher C said that working in groups was not possible for her particular group of students. Teacher C’s statement below suggests that working with partners and in groups is an idealized form of classroom interaction but one that is contingent upon student related factors and therefore is not always possible. Ehm, I personally like having children learn with a partner or in groups. I think, if it comes to me, I would/ that would be a learning for/ I would provide for them every day and every lesson. But, in this learning group, it's just not possible. […] Still, that's something I want to try out and [the German speaking teacher] also wants to try out. Because they’re/ it's just more motivating, it's/ they're gaining a lot more in my opinion ehm, the atmosphere is just way better, the outcome is way better, way bigger. (TC2 00: 34: 15) Teacher C also stated that maintaining a more personal relationship with students motivated them and gave them opportunities to develop their language skills. She described personal situations as one on one interactions as well as small group interactions. On the one hand, more personal relationships with students are developed by using German with students and are influenced by the fact that Teacher C is a class teacher and not only a subject teacher. Within these interactions, the teacher is able to talk about more personal topics with students. On the other hand, Teacher C describes these interactions as opportunities to learn English, in which it can be assumed that she is using English with students. Whether speaking English or German, Teacher C sees such personal interactions as motivating students. So I think personal situations and not only one on one but also in smaller groups, ehm laughing together, creating situations where you are just personal with each other and you can address more personal/ those are situations where they 178 6 Results <?page no="179"?> learn the language and where they're/ where they, I don't know, are more motivated and ehm, confident enough I would say. (TC2 01: 03: 55) More personal relationships with students, positive interactions between students, and varied and playful activities were all ways that teachers described motivating their students in CLIL lessons. 6.3.5 Student choice of language Code Exemplary quote Coded segments Teachers Student choice of language Und genau, das ist einfach immer so ein, “Ich kann mir aussuchen. Ich kann es auf Englisch machen. Ich kann es auf Deutsch machen.” Für viele ist es natürlich auf Deutsch schon schwer genug. Aber für manche ist es jetzt mittlerweile auch interessant es mal auf Englisch zu versuchen und sich die englischen Sätze zu nehmen. So diese möglichkeit einfach. (TA4 00: 11: 06) 15 3 Table 20: Student choice of language Another approach that teachers described using in the CLIL classroom was allowing students to choose using English or German. Of course, all students were confronted with spoken English input from the English-speaking teacher. In this sense, students did not have a choice about listening to English. However, several teachers described allowing students to, for example, choose English or German language worksheets or to respond orally in English or German. This approach toward language use in CLIL corresponds with teachers’ view of foreign language learning in CLIL as a curricular enrichment and choice (described in section 6.1.1 above). Teachers’ responses indicate two primary reasons for allowing students a choice concerning language use. First, teachers recognize differences in students’ language competences generally and foreign language competences specifically. Students who find learning content in German difficult are allowed to avoid the additional challenge of, for example, writing in English while others who find learning in English enjoyable or interesting can choose to speak or write something in English. Und genau, das ist einfach immer so ein, “Ich kann mir aussuchen. Ich kann es auf Englisch machen. Ich kann es auf Deutsch machen.” Für viele ist es natürlich auf Deutsch schon schwer genug. Aber für manche ist es jetzt mittlerweile auch interessant es mal auf Englisch zu versuchen und sich die englischen Sätze zu nehmen. (TA4 00: 11: 06) Second, as already described in Teacher A’s statement above, some students are more interested in using English than others. Teachers described a student’s interest in using English as related to his or her general sense of well-being and security when using 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 179 <?page no="180"?> English. The reason some students choose to speak English and others do not is, in addition to competences, a result of being interested in the language and experiencing positive emotions when using the language. Teachers in this study described not wanting to create situations in which students experience negative emotions in relation to using English and therefore give students a choice of when to use it. In the excerpt below, Teacher A explains that some students in her class are more willing than others to use English because they enjoy it and have fun trying out their use of English. Aber so eine Freude auch daran. Der macht das halt total Spaß. Sie will sich richtig ausprobieren und testen, was sie schon auf Englisch sagen kann. Das haben auch zwei, drei andere, die dann immer versuchen so was sie halt können, dann auf englisch mich irgendwas zu fragen oder so. (TA4 00: 39: 39) While teachers described students choosing on the basis of language competences or interest, one teacher also described guiding students’ choices in some situations. In the excerpt below, Teacher D describes her thinking about when to intervene in a student’s choice of completing either a German or English worksheet. She uses her knowledge of students to guide them toward the appropriate level of linguistic challenge. She encourages students with lower English competences to complete German language worksheets. She encourages students with higher English competences to complete English worksheets. When students choose a worksheet in a language that she deems too difficult or too easy for a particular student, she suggests that the student reconsider his or her choice and guides the student to the, in her estimation, more appropriate worksheet. Ja, ich habe ähm das zur Wahl gestellt, was man sich zutraut. Ich habe allerdings gerade beim ersten Ausfällen auch geschaut, was die/ was die Kinder sich auswählen und habe bei dem ein oder anderen nochmal interveniert und habe sowohl in ähm in Fällen wo die Kinder sich überschätzt haben, meiner Meinung nach. […] Und sehr, sehr gute Schüler und Schülerinnen so den einfachen Weg wählen wollten, ähm habe ich ja durch aus ähm da auch noch mal den Tipp gegeben, “Komm ne, du könntest das auch auf Englisch.” Ähm, so gesehen war es/ es war schon zu Auswahl, aber so ein bisschen angeleitet gerade zu beginn. (TD2 00: 20: 46) By allowing students to choose when to use English and when to use German, teachers individualize their CLIL instruction. This individualization extends to situations where teachers guide students’ choices based on their knowledge of students. 180 6 Results <?page no="181"?> 6.3.6 Teach language explicitly Code Exemplary quote Coded segments Teachers Teach/ high‐ light vocabu‐ lary I think it also went well because of the work we did beforehand. Making sure they know all the terms in English. (TC1 00: 18: 15) 9 3 Repeat sen‐ tences I know that we/ in the first lesson I wanted them to talk or say the whole sentence like "I can hear with my whatever." But that was just basically a repetition of most words they knew and that was the only time I think I really focused on the language. (TB2 00: 11: 54) 3 1 Table 21: Teach language explicitly While teachers in this study clearly described learning non-language content as the priority in their CLIL lessons, they also described teaching aspects of language such as vocabulary explicitly as something that supported their CLIL instruction. Based on lesson observations and teacher interviews, teaching language explicitly was done primarily to establish a basic lexicon for the topics that teachers taught. Teacher C accredits pre-teaching topic vocabulary as part of the reason that a previously taught unit on electricity was successful. I think it also went well because of the work we did beforehand. Making sure they know all the terms in English. (TC1 00: 18: 15) Teachers described teaching vocabulary or making students aware of vocabulary pri‐ marily through discussion. The explicit instruction that teachers described was done orally and not, for example, through the use of written vocabulary lists to be memorized. Teachers described and were observed discussing terminology often in whole-class instructional phases of lessons. For example, while teaching a unit on the five senses, Teacher B was observed saying, “It makes me aware. What does that mean? They make me aware.” (from observation on Nov. 4, 2019). In this way, focused discussion of terminology was integrated into content-focused instruction. Based on classroom observations, a key aspect of teachers’ vocabulary-focused discus‐ sions was the use of questions. Teacher C used questions such as “I have a question. What is a guard? ” or “Is “hall” something inside or outside? ” (from observation on Oct. 1, 2019). When asked to reflect on her use of questions to highlight vocabulary, she described being unaware of using such a technique. She reasoned, however, that asking questions had the effect of grabbing students’ attention and sometimes to building links to words students already know. Ehm, I think one aspect to be honest with you is that questions like that, short, usually short questions like that where there's one key term in it, gets the attention of the children. […] At the same time I want to just make sure that 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 181 <?page no="182"?> certain words like, I don't know, “page” [as in a servant] for example, I'm just trying to build a little bridge to knowledge we already have, so it's easier for them to memorize. (TC2 00: 54: 41) In addition to making students aware of subject terminology by discussing such terms during whole-class instruction, one teacher was observed having students recite short phrases consisting of words they already knew. For example, Teacher B had students repeat the phrase “I can hear with my ears” (from observation on Nov. 11, 2019) in order to have them repeat words they already know. I know that we/ in the first lesson I wanted them to talk or say the whole sentence like "I can hear with my whatever." But that was just basically a repetition of most words they knew. (TB2 00: 11: 52) Elsewhere in my follow up interview with Teacher B, she suggests that she uses recitation as a way to help students remember words and because it is a technique that they are familiar with from EFL lessons. 6.3.7 Plan for more time Code Exemplary quote Coded segments Teachers Plan for more time Man muss dann schon auch Zeit einplanen, dass man dieser Wörter erst übt bevor es dann weitergehen kann mit dem Sachunterrichtlichen Thema. (TD1 00: 19: 48) 8 3 Table 22: Plan for more time Several teachers described needing more instructional time to teach content through a foreign language. As an approach to CLIL instruction, teachers described their awareness of the additional time required to teach content through a foreign language and planning for additional time. One factor that required additional instructional time was pre-teaching English terminology. Additional time was also necessary for additional and repeated explanations. Ähm, was wiederum durchaus meine Mehr Zeit kosten kann/ ist, wenn es Verständnisprobleme gibt. Wenn man merkt man braucht ein bisschen mehr Zeit um Sachen zu erklären, den nochmal zu wiederholen vielleicht, ähm Wortmaterial einzuschleifen. (TD1 00: 21: 48) 182 6 Results <?page no="183"?> 6.3.8 Build links to other subjects Code Exemplary quote Coded segments Teachers Build links to other subjects Also, sich in dieser Vorplanung zu überlegen hm, wo kann ich jetzt dieses Thema, was ich habe/ in welchen Unterrichtsfächern kann ich das einbauen und was ist auf Englisch möglich? (TA5 00: 03: 46) 3 2 Table 23: Build links to other subjects Teachers in this study taught lessons in subjects such as Art, Music, and Physical Education, in addition to Science and Social Studies, through English. Some teachers described working to build links from their Science and Social Studies lessons to these other subjects. Welche verschiedenen handlungsorientierten Bereiche kann man ähm abdecken? Wie macht man/ Kann man im Sportunterricht was aufgreifen von dem Fach? Kann man in Kunst was aufgreifen? Also ein bisschen dieses Fächerübergreifende. (TA5 00: 03: 26) One teacher also described planning activities within her Science and Social Studies lessons that were transdisciplinary in nature. Planning drama activities, for example, provided a kind of instructional enrichment. Wir haben jetzt bei Thema Kartoffel in Klasse drei ein kleines Theaterstück ähm, was wir aufführen, so in unserem Klassenfest, The Potato King, ähm und ja das ist auch immer mal ganz nett so als/ als Bereicherung so ein bisschen Fächerübergreifend auch schauspielerisch dann tätig zu werden. (TD1 00: 27: 28) For teachers in this study, building links to other subjects took place within the Science and Social Studies lessons (for example, by including music or drama activities) as well as with other CLIL subjects at school (e.g. Physical Education, Music, Art). 6.3.9 Summary The third research question this study seeks to answer is: What approaches do teachers believe are best for achieving these goals? The CLIL instructional approaches that teachers described in the interviews are summarized in the tables in section 6.3. As a broad approach to CLIL instruction, teachers reported focusing first on providing students with large amounts of oral English input. This English input was primarily provided by the English-speaking teacher in the classroom. Sources of oral input such as talk from other students or multimedia materials were described and observed less often. Teachers also described preferring written English input from materials that were created for native English-speaking children in English-speaking countries. Teachers, however, described the challenge of finding materials that met students’ needs from both a content and language perspective. 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 183 <?page no="184"?> Despite describing the importance of providing a large base of English input, teachers also described the use of German language input, both written and oral, as appropriate and necessary in certain contexts. The primary reason that teachers use German is to manage content and linguistic complexity. This usually takes the form of the German-speaking team teacher explaining or in some way ensuring student comprehension after a phase of English input. Teachers also feel that it is appropriate to use German in emotionally-charged situations or situations involving the safety and well-being of students. In these situations, the English-speaking teacher switches into German herself and does not, for example, seek the support of the German-speaking teacher. Teachers also described the use of German, especially in the form of written materials, for local and personal topics such as regional geography or sexual education. With a focus on input, teachers then described using their knowledge of students to guide their decisions about CLIL instruction. Teachers’ knowledge of the competences and interests of the whole class are used to guide lesson and unit planning decisions while knowledge of individual competences and interests are used to guide immediate or interactive decisions with students. Teachers then described a variety of techniques that they use to scaffold content and language input as well as L2 output. Chief among their input scaffolding strategies are translation and visualization. Translation was often described as an unpreferred strategy but nonetheless a necessary strategy to ensure understanding for all students. The teachers described using their German-speaking team teachers primarily for the purpose of ensuring content understanding including the repairing of understanding after comprehension had broken down. They use student translation to ensure understanding of words and phrases spoken by the teacher and to maintain a sense of flow in the lessons. Visualization strategies, including the use of body language, pictures and objects, are preferred by teachers and are used to support comprehension while maintaining the use of English. Teachers also described other strategies that maintain the use of the L2 such as simplifying their language and repeating or reusing particular words and phrases. When seeking to scaffold students’ oral output, teachers most often described simply encouraging students to speak. The extent to which they encouraged a student to speak English is based on their knowledge of that individual student. Scaffolding L2 input and output based on their knowledge of students formed the foun‐ dation of teachers’ approaches to CLIL instruction. Other strategies, however, were also described. One of those strategies is developing and maintaining motivation in students. Most teachers described planning for motivation in the form of planning activities that were varied, playful, and active. However, several teachers also described motivational strategies based on developing relationships in the class, for example, through student-student and student-teacher interaction. Teachers also described allowing students to choose the extent to which they engaged with the English language. This choice was limited primarily to productive competences (speaking and writing) as all students were confronted with spoken English input from the English-speaking teacher. Teachers described allowing students to choose whether or not to use English because of individual differences in language competences and interest. 184 6 Results <?page no="185"?> Though all teachers described prioritizing content learning in their CLIL lessons, several of the teachers reported teaching subject vocabulary explicitly in order to support the learning of content during CLIL lessons. Teachers teach vocabulary primarily through whole-class discussions intended to draw attention to the meaning of specific words as well as by having students repeat English words and phrases. Teachers reported that students need more time to learn content through a foreign language. Because of this, teachers plan for more time when developing their lessons and units. Teachers also reported trying to build links between the content that they teach in their CLIL Science and Social Studies lessons and other subject disciplines. This takes place by integrating subject disciplines such as the arts into their CLIL Science and Social Studies lessons as well as by looking to integrate Science and Social Studies content and language into other CLIL subjects such as Physical Education and the arts. 6.3 Research question 3: What approaches do teachers believe are best for achieving these goals? 185 <?page no="187"?> 7 Discussion In the following three sections, I will discuss the main results of this research as they relate to answering each of the three research questions. The results will be discussed and interpreted in the context of the theoretical foundations of CLIL pedagogy and empirical findings of CLIL teachers’ cognitions. In section 7.4, I will discuss the possible influence of the DIPS program design on several of the teachers’ cognitions and consider the generalizability of the teachers’ cognitions reported in this study to other CLIL contexts. 7.1 Teachers’ cognitions about teaching and learning in primary school CLIL While CLIL is generally motivated by the potential for L2 learning, research on CLIL outcomes and reviews of such research tend to focus on the development or learning of the L1, L2, and content subject (e.g. Science or Social Studies) (see, for example, Frisch, 2021; Pérez Cañado, 2020; Botz & Diehr, 2016; Piske, 2013). With these three learning goals in mind, one aspect of RQ1 was to find out how teachers prioritize these three goals in their thinking about CLIL. The teachers in this study clearly prioritize learning non-language subject content in their CLIL lessons. This result is not surprising since CLIL in Germany is a non-language subject class taught through a foreign language (KMK, 2013). This result corroborates previous research on teachers’ thinking from secondary and tertiary CLIL in several national contexts (de Graaff et al., 2007; Costa, 2013; Skinnari & Bovellan, 2016) as well as from primary school CLIL contexts (Bovellan, 2014) suggesting that despite CLIL being a “language education initiative” (Morton, 2012, p. 12), teachers in a variety of CLIL contexts remain focused on content learning. Teachers in this study also described the need for students to learn subject vocabulary in the L1 (or school language, German). Teachers’ concerns for learning L1 vocabulary stemmed from concerns about students’ general language development (e.g. learning German for non-native speakers of German and learning German terminology before English equivalents for native-speaking German students) and the need for such vocabulary in later school contexts. This thinking has also been found in secondary school contexts where teachers felt the importance of preparing students for a return to L1 teaching contexts (Oattes et al., 2018). Most teachers in this study described the development of English language competences as a curricular enrichment or choice for more capable and motivated students. This belief is congruent with statements about the DIPS program describing English as being learned “nebenbei” (Raunser, 2012, March 5, p.-17). Such a belief has also been found in other CLIL contexts in which teachers described learning the L2 as “a side-effect,” a “by-product,” or “a spice” (Skinnari & Bovellan, 2016, p.-153). CLIL programs have been described as “essentially different forms of enrichment programmes” (Baetens Beardsmore,1993, p. 3). In NRW, teaching content through a foreign language is enrichment in that it sets out learning goals not articulated in the core <?page no="188"?> curriculum for NRW (MSB NRW, 2021). Therefore, when teachers describe the development of English language competences as a curricular enrichment or choice, they are expressing a curriculum-centric view of teaching and learning and one that essentially agrees with Baetens Beardsmore’s (1993) assertion that CLIL programs are enrichment programs. The DIPS program and its conceptions of “immersion” and “one person-one language” (Raunser, 2012, March 5) mean that the English-speaking and German-speaking teachers speak (almost) exclusively their respective language to students. In considering the roles for the use of English and German in CLIL lessons, teachers described a basic calculation that they alone or in collaboration with their teaching partners made: What can students do in English and what do they need to do in German? While this calculation was sometimes described in the context of where English and German fit in the phases of a lesson or in curricular topics, it was usually expressed in the context of linguistic and conceptual demands. In other words, simple topics (linguistically or conceptually) can be managed in English; more difficult topics require more German. This formulation of the roles of languages within CLIL, however, did not necessarily mean that teachers switched their language use. It meant that there was greater attention to materials in one of the languages over the other and a stronger role for the teacher who spoke that particular language. While less significant in the data from a quantitative perspective, some teachers described the use of English with students as a barrier to teacher-student relationships as well as a bridge in the form of a lingua franca between students. These findings highlight the relational role that teachers feel language use has in CLIL. Noticeably absent in teachers’ statements about their CLIL instruction was any formula‐ tion of a goal equivalent to the development of subject literacy in two languages. Describing the development of L2 competences as a form of curricular enrichment or a choice that more motivated or capable students could make falls far short of such a goal. The failure to articulate such a goal for their CLIL instruction may have any number of reasons. One possibility is simply that they’re unaware of such a goal. If this is the case, CLIL teacher training may be one avenue to try to more closely align teachers’ cognitions with CLIL theory. If teachers are aware of the goal of developing subject literacy in two languages, another possibility for not articulating such a goal could be that teachers want to create a learning context in which failure to meet a higher standard of foreign language learning is acceptable to students and teachers or, in other words, students and teachers can experience success by meeting a lower standard of foreign language learning. Teachers in this study were clear that they found learning through a foreign language to be challenging. Such a perception of learning in CLIL seems to be corroborated by Frisch (2021) who found that when primary school CLIL students were tested on their content knowledge through English, 90 % of them scored at a rudimentary level one. When tested on content knowledge through German, 80 % scored at a level of at least three (with level five being the highest possible). Such evidence seems to strongly suggest that developing subject literacy at a grade appropriate level in the L2 is very challenging and rare. By framing the goal of L2 development as something that is “on top” (TD1 00: 34: 56) or a choice, teachers would seem to reduce the pressure to meet a more rigorous standard of language learning as well as 188 7 Discussion <?page no="189"?> alleviate a sense of failure that may come with not meeting a more rigorous standard of foreign language development. Yet another possibility is that the teachers in this study simply do not see the development of subject literacy in two languages as a curricular requirement and therefore do not work toward it. As stated above, by describing the unique and essential characteristic of CLIL, namely the use and the learning of a foreign language, as enrichment or a choice, teachers are expressing a curricular-centric view of teaching and learning that sees CLIL as a form of enrichment. The Science and Social Studies curriculum in NRW indeed does not require the development of such competences in a foreign language (MSB NRW, 2021). Therefore, if teachers are simply expressing cognitions that are based on their interpretations of curriculum documents, affecting change in teachers’ cognitions would seem to be a matter of curriculum reform and not only, for example, CLIL teacher training. Another aspect of the first research question was to find out what teachers think about the nature of second language acquisition in CLIL. All of the teachers in this study felt that CLIL had a strong, positive impact on students’ L2 development. Most of the teachers also felt that students reaped general cognitive advantages from their participation in CLIL education. Of course, these results don’t have any bearing on whether students actually benefit in these or other ways from CLIL. They do indicate, however, that the teachers in this study - those directly responsible for implementing CLIL instruction - believe in the foreign language learning and general cognitive benefits of CLIL. Such a positive perception of CLIL may play a role in maintaining teachers’ motivation in a challenging teaching and learning context. Teachers also expressed the view that language learning in CLIL is a natural process. That CLIL teachers believe language learning in CLIL is “natural” has been reported in previous studies as well (Hüttner & Dalton-Puffer, 2013; Costa, 2013; Skinnari & Bovellan, 2016) and has been associated with “easy-going and relaxed ways of learning” (Skinnari & Bovellan, 2016, p. 154). The view that language learning in CLIL is natural is also one of the theoretical arguments for CLIL education. As Marsh (2000, p. 3) states, “It is this naturalness which appears to be one of the major platforms for CLIL’s importance and success” (italics in original). However, in order for such naturalness to lead to language learning “success,” it would seem necessary that learners have basic communicative competences in the foreign language and a willingness to use them. This is not necessarily the case in CLIL contexts with young learners. Based on classroom observations in this study, students did not have the L2 competences required to communicate more fully using the L2 in CLIL lessons. Schwab (2020) and Bechler (2014) also report that primary school CLIL students often use the L1 when speaking in lessons. While CLIL, both theoretically and as observed in lessons in this study, provides a wealth of input beyond what is available to students in traditional EFL lessons, students, as observed in the CLIL lessons in this study, produced very little output. This is problematic since speaking and writing provide several language learning functions including noticing and hypothesis testing (Swain, 1995). Swain claims that producing output pushes students to “process language more deeply” than they do when processing input (Swain, 1995, p. 126). In a teaching and learning context in which students lack basic communicative competences and, therefore, do not just “naturally” speak, a belief in language learning as a natural process and one, therefore, that takes place 7.1 Teachers’ cognitions about teaching and learning in primary school CLIL 189 <?page no="190"?> with minimal direction or guidance from teachers, would seem to greatly limit students’ potential L2 development. Despite the majority of teachers in this study expressing the belief that language learning in CLIL is a natural process, a process associated with ease of learning (Skinnari & Bovellan, 2016), the majority also described learning through a foreign language in CLIL as challenging. This belief seems to manifest itself in other cognitions expressed by the teachers. For example, teachers described academically strong students as being most likely to benefit from CLIL education. Such a belief implies that learning in CLIL, i.e. learning through a foreign language is challenging. Likewise, the belief that German should be used for more challenging content and English for less challenging content also implies a belief that learning through a foreign language is challenging. Taken together, the variety of cognitions that teachers expressed suggest that CLIL-as-challenging is a central or foundational belief in their thinking about CLIL. As already described, the real or perceived challenge of CLIL may limit what teachers see as possible with regard to foreign language learning in their CLIL instruction and therefore contribute to the cognition that L2 development is a curricular enrichment or choice in CLIL. Teachers also described topics most appropriate for CLIL as concrete, action-oriented, motivating, and that involve less complex vocabulary. Concrete and action-oriented topics are amenable to visual scaffolding strategies while topics with manageable vocabulary could be considered as “pre-scaffolded,” i.e. the language is already simplified, one of the L2 scaffolding strategies described by teachers. Motivating topics are one way in which teachers can enact their belief that motivating students is a necessary approach for their CLIL instruction. According to the teachers in this study, the topics that are most appropriate to CLIL education are those which are amenable to various scaffolding techniques and which are motivating for students. Such topics allow teachers to more effectively address, as they perceive it, the central challenge of learning in CLIL. In addition to the belief that CLIL is challenging for students, teachers found CLIL instruction to be challenging for themselves. The challenge was described as relating primarily to making oneself understood to students. Considering the context of primary school CLIL described earlier, namely, that teaching through a foreign language to students who have low foreign language competences is unique to primary school CLIL, it is not surprising that teachers find communicating meaning in such a context challenging. At secondary schools in NRW, the challenge of communication in CLIL is mitigated in part by offering additional foreign language classes for two years (grades five and six) before entry into the first CLIL course (grade seven). In the DIPS program, it would seem that the additional German-speaking teacher in the Science and Social Studies lessons is intended to alleviate some of this challenge. However, as teachers described, the German-speaking teacher is often used only after communication has broken down. Whether there are additional measures that could be taken to support primary school teachers in their efforts to communicate effectively with their students through a foreign language is a question that may need to be addressed when considering the further development of the DIPS program. An additional challenge for teachers was the time and effort required to create and curate English language materials. That CLIL teachers find the creation of materials time-consuming and requiring more effort is well established in CLIL literature (Viebrock, 190 7 Discussion <?page no="191"?> 2007; Massler, 2012; Bovellan, 2014; Bechler, 2014; McDougald, 2015; Siepmann et al., 2021). Nonetheless, there is some evidence that CLIL teachers find material creation to be a “positive challenge” and “creative task” (Sandberg, 2019, p. 305). Teachers in this study did not necessarily characterize material creation as burdensome but did describe needing to be motivated in order to sustain the effort over the years needed to create and curate materials. In a context in which developing L2 competences is characterized as optional, the need for English language materials becomes less urgent. This may result in more German language materials being used. Strong foreign language skills were also something teachers in this study found to be important for CLIL teachers. The development of such skills was generally described positively in terms of personal development, e.g. traveling to English-speaking countries and engaging with English-speaking culture more broadly (e.g. watching movies in English). Teachers felt it was important to “feel safe” (TB1 00: 18: 29) in the language and that making mistakes or needing to look up words was not a sign of “bad English” (Bovellan, 2014, p. 112) but a normal part of foreign language learning. This finding stands in contrast to findings from other studies which have found that teachers often feel “insecure” (Bovellan, 2014, p. 11) about their foreign language skills or describe their foreign language skills in dismissive terms (Oattes et al., 2018; Bovellan, 2014; Díaz & Requejo, 2008; Imgrund; 2004). 7.2 Teachers’ cognitions about language-related educational goals in CLIL The second research question was concerned with finding out what language-related edu‐ cational goals teachers work toward in the context of their CLIL instruction. From a strictly cognition perspective, teachers described the goals of developing listening comprehension and speaking skills in almost equal measure. This result is to be expected in primary school language learning settings since they typically emphasize the development of oral language competences (Schauer, 2019). From a classroom practices perspective, however, students were observed speaking English far less than they listened to English spoken by the English-speaking teacher. This seeming disparity between teachers’ cognitions and their practices is likely less a sign of ambivalence and more a reflection of the reality of working with young EFL learners. Young students simply do not have the L2 productive competences to speak a lot in the classroom. However, as described in section 2.2.1 above, speaking, or producing output generally, serves a number of functions in the development of higher L2 competences (Swain, 1995). Without output production, language learning is likely to be inhibited. Teachers in this study wanted their students to speak but described it as a kind of ultimate goal for some students and not as a means of language learning for all students. It is also possible that teachers were unprepared to take the extra time and thought required to plan activities that develop speaking competences and content learning. If the CLIL principles of “natural” language learning (Marsh, 2000, p. 3) or “reception-based” theories of language development (Dalton-Puffer et al., 2010a) are to be taken seriously, then teachers should be able to teach content through a foreign language without excessive 7.2 Teachers’ cognitions about language-related educational goals in CLIL 191 <?page no="192"?> thought toward language or without planning activities focused on the development of speaking skills or producing output. But, in a program such as DIPS in which young learners spend slightly less than half of their weekly timetable learning English or through English (EFL lessons and CLIL lessons combined), such a natural or reception-based approach is unlikely to yield significant development of productive competences. If the goal is only to develop listening comprehension skills or, more broadly, to familiarize students with the language and support their development of a general sense for the language, then such reception-based approaches will likely suffice. But if the goal is truly to develop speaking competences, as all of the teachers described in the interviews, then more thought will need to go into how this goal might be accomplished. One example of such an activity was observed in Teacher D’s class. As a class, students assembled a regional map of Europe on the classroom blackboard. Once the map was assembled, individuals took turns describing a country’s location in relation to another country. They did this in German and in English. For example, they recited statements such as, “Schweden liegt östlich von Norwegen,” or “Sweden is east of Norway.” Such an activity would seem to make progress toward meeting the ultimate goal of developing subject literacy or discourse competences in two languages (Diehr, 2012; KMK, 2013). The further development of practices aimed at developing subject literacy in two languages would seem to be one way in which the DIPS program could be further developed in line with CLIL theory while also meeting teachers’ stated goal of developing L2 speaking competences. All of the teachers also described the goal of developing students’ subject-related vocabulary and grammar. Activities focused on learning subject vocabulary and grammar were observed in various lessons from all of the teachers. As a building block of language, it is not surprising that teachers emphasized learning words as a language-related goal for their CLIL lessons. Bovellan (2014), in a primary school context, also found that the majority of teachers in her study prioritized learning words and Morton (2012, p. 217), in a secondary school setting, found that teachers focused “almost exclusively” on teaching vocabulary when describing their explicit language instruction in CLIL. Focusing on the development of subject-specific vocabulary may be a feature of teachers’ cognitions and practices across a variety of CLIL contexts. Teachers also described valuing a range of general pedagogic goals and working to develop those goals in their CLIL lessons with students. Chief among these general pedagogic goals were reducing fear and developing confidence when confronted with or using a foreign language, developing motivated language learners, and providing opportunities for success through the effort required to learn through a foreign language. Teachers’ desire to develop such qualities has been documented in secondary school CLIL contexts. Wegner (2012) found that the CLIL teachers in her study sought to develop students’ self-confidence, motivation, and experiences of success. Skinnari and Bovellan (2016, p. 162) found that CLIL teachers worked to “wide[n] perspectives” and “cultivat[e] respect and understanding” in their students. The results of this study extend such findings to the primary school context. In a secondary school CLIL context, Wegner (2012) linked general pedagogic goals to the lowering of subject learning standards and treating subject learning standards in the classroom haphazardly at times. In the context of the present study, there is no evidence 192 7 Discussion <?page no="193"?> that teachers’ consideration of general pedagogic goals in any way lowered subject learning objectives. Instead, working toward general pedagogic goals seemed to allow teachers to maintain subject learning standards by, in many cases, reducing foreign language use. Working toward general pedagogic goals may have, instead, negatively affected foreign language learning outcomes by adjusting instructional strategies and expectations to meet teachers’ perceptions of students’ cognitive, linguistic, and affective capabilities. These findings also suggest that while CLIL is motivated by the potential for language learning in order to meet the social, political, and economic demands of a globalized world (Coyle et al., 2010), teachers see CLIL as a unique language learning context in which students can be developed as individuals capable of meeting and overcoming challenges. Such a personal development motivation for the value of CLIL adds to the social, political, and economic motivation described in CLIL literature. The findings that teachers work toward general pedagogic goals and that some teachers find the use of English with their students to be a barrier to relationships suggest that for the teachers in this study foreign language learning and use is a process saturated with affective meaning and consequences. Such an understanding of language learning as not only a cognitive or linguistic process but an affective one as well suggests that, for the teachers in this study, CLIL pedagogy cannot be reduced to a mere combination or integration of pedagogical practices from multiple disciplines (Morton, 2012, p. 12). Furthermore, it is possible that evaluations of CLIL based solely on curricular-defined outcomes are unsatisfactory for teachers. Based on the results of this study, CLIL pedagogy should make room for the individual experience of language learning and learning through language and evaluations of CLIL must make room for the subjective experience of CLIL instruction. 7.3 Teachers’ cognitions about approaches most effective for achieving CLIL goals The third research question is concerned with finding out what teachers think about the most effective ways to achieve their goals in the CLIL classroom. Their responses indicate four primary approaches: provide as much English input as possible, use knowledge of students to adapt instruction to their needs and competences, scaffold L2 input and output in a variety of ways, and motivate students through varied and playful activities and social support. Teachers in this study described providing large amounts of English input to students as a key component of their CLIL instruction. Teachers were also observed speaking English almost exclusively to students in their lessons as well as in broader school contexts. Such a belief and practice of functional monolingualism corresponds with and would seem to stem directly from the DIPS model of CLIL. The DIPS program is described as implementing an immersion model of CLIL and employing a one person-one language approach to teachers’ language use at school (Raunser, 2012, March 5, p. 17). Additionally, such a belief in the importance of oral input is in line with usage-based theories of second language acquisition, which consider input the “sine qua non” of language acquisition (Gass & Mackey, 2015, p. 181). Furthermore, the belief is in line with principles of CLIL pedagogy which describes 7.3 Teachers’ cognitions about approaches most effective for achieving CLIL goals 193 <?page no="194"?> language learning in CLIL as a natural process that takes place through increased exposure to the L2 (Marsh, 2000; Vollmer, 2010; Wolff, 2016). While the quantity of input is important for the development of L2 competences, the quality of input also matters. Lyster (2011, p. 612) states that the CLIL target language needs to be “manipulated and enhanced” during instruction in order for students to learn both content and the foreign language. Teachers in this study did describe simplifying their language, repeating and reusing key vocabulary items, and using visual scaffolding strat‐ egies to help convey their messages. According to Smith (1993, p.-177), such strategies are “unelaborated” forms of input enhancement. More elaborate forms of input enhancement would include, for example, explicit grammar instruction (Smith, 1993, p.-177). DIPS program documents describe language learning in CLIL as a natural process and one that proceeds in a similar way to mother tongue acquisition (DIPS, 2020, Sept. 7). It is unlikely, however, that language learning in the DIPS program can proceed in the manner of mother tongue development. For one, the time that a student spends exposed to or using the L2 (English) in the DIPS program is likely far less than the student spends using his or her mother tongue. Furthermore, slightly less than half of students’ instructional time is spent learning English or learning through English. Students would need far more time being exposed to the L2, exposure time that is not possible in the context of the DIPS program, in order for students’ L2 development to proceed in ways more similar to L1 acquisition processes. The quantity of L2 input that DIPS teachers can provide, while valuable and necessary for any L2 learning to take place, is likely not enough to achieve more advanced levels of L2 development. Input quality must also be considered. Various forms of input enhancement will need to be integrated into teachers’ instruction. Teachers did report using “unelabo‐ rated” (Smith, 1993, p. 177) forms of input enhancement. Using “highly elaborate” forms of input enhancement such as explicit grammar instruction (Smith, 1993, p. 177), however, may not be appropriate for content-focused CLIL instruction. Input scaffolding strategies, however, such as vocabulary instruction and corrective feedback on errors (Massler & Ioannou-Georgiou, 2010) are more elaborate forms of input enhancement that may support students’ language development. While providing as much English input as possible to students is clearly important to the teachers in this study, they also described a series of purposes for the use of German. While described in a slightly different way by teachers in this study (see section 6.3.1), the reasons for German use are largely congruent with those outlined by Frisch (2016) (see section 3.2.7). However, for the teachers in this primary school context, the use of German to teach local topics (e.g. German geography or a locally organized bicycle safety course) and personal topics (e.g. sexual education) would seem to include such functions as the cognitive, communicative, or affective function and yet be different in important ways. For example, the use of German to teach local topics was not described by teachers as being done in order to manage the complexity of the topics (cognitive function) or provide German vocabulary equivalents for English terminology (communicative function). Instead, teachers described using German because using English in such contexts “doesn’t make sense” (TB2 00: 26: 25). Likewise, teachers described using German for the topic of sexual education not to meet the spontaneous need for emotional comfort (affective function) or because the 194 7 Discussion <?page no="195"?> topic was inherently too complex (cognitive function), but because the nature of the topic required the reduction or elimination of a barrier to clear communication that is a foreign language. At least in this primary school context, teachers view the replacement of English with German as an appropriate use of German for some topics. German, therefore, is used to scaffold content and L2 learning, teach German language competences, and, for some topics, replace the use of English. Also, teachers’ descriptions of the use of German in their CLIL lessons do not necessarily fit neatly into one of Diehr’s (2012) three categories of language use in CLIL (see section 2.1.3). The DIPS model of CLIL, as described by teachers and observed in classrooms, would seem to match most closely with Diehr’s Type A and Type C descriptions of language use in CLIL. Teachers’ responses in the interviews describe the use of German in primarily unplanned and supportive terms. The use of German, including translation, is used to support students’ understanding of content and/ or language (English). A key element of Type A instruction is that L1 use is unplanned and rare (Diehr, 2012). A number of teachers’ responses suggest that decisions to use German were based on unplanned, interactive judgments of students’ understanding. Ehm usually it's [explaining something in German] something [the German-speaking teacher] automatically does when we have the feeling that what I'm trying to explain in English at that moment just doesn't come through to the kids for some reason. Maybe because we're looking for a special term that's essential in this whatever I'm talking about. Maybe it's because I'm just not explaining it right. So sometimes she's just ehm jumping in by just giving one key word or one translation. (TC2 01: 06: 33) Furthermore, Teacher B expresses a kind of monolingual L2 ideal. When asked what would impress a visitor to the class, Teacher B responded: That there's no German needed. If they really get everything in English. Especially like/ depending on the topic and depending on the age of the children I would say, yeah. (TB1 00: 26: 05) Despite qualifying this ideal as “depending on the topic and age of the children,” the use of the L1 is seen as something that is used sparingly and only when understanding breaks down. Using the L1 sparingly and only in spontaneous and unplanned situations are features of a Type A use of the L1 (Diehr, 2012). In addition to this more monolingual ideal, teachers also described the necessity of learning German subject vocabulary (discussed above in section 7.1). In this way, teachers’ cognitions seem to meet, at least in part, Diehr’s (2012) standard of planning for the devel‐ opment of L1 competences that she describes in her Type C model of CLIL language use. As part of those L1 competences, however, Diehr describes planning for the development of subject discourse competences. In the interviews, teachers did not describe such planning or goals. In the classroom observations, however, Teacher D’s example of assembling a map of Europe and using English and German phrases to describe countries’ relative locations (described in section 7.2 above) stands out as a planned activity that addresses discourse 7.3 Teachers’ cognitions about approaches most effective for achieving CLIL goals 195 <?page no="196"?> competences in both languages. Such a practice could be seen as representing an aspect of Diehr’s Type C model of L1 language use. Nonetheless, teachers’ cognitions and practices do not, in their entirety, conform to a Type C model of language use in CLIL. Teachers in this study also described using their knowledge of students to guide their learning goals and instructional practices. This finding suggests an understanding of language learning in CLIL that is conditional and unpredictable. The way that teachers teach and the degree to which learners learn is dependent on, in part, the multiple variables associated with the complexity of any individual learner or group of learners. Teachers seem not to plan and implement instruction in uniform ways for different students and different classes but instead make adjustments to their instruction based on specific learners and groups of learners. Teachers’ descriptions of using their knowledge of students to implement instruction can be seen as a form of differentiation or individualization. Teachers’ descriptions of calibrating instruction to meet the needs of individuals would appear to meet several of Tomlinson’s (2005) descriptors of effective differentiation practice. Such practices include, for example, creating safe and challenging learning environments and using resources and teaching strategies flexibly in order to meet a variety of student needs (Tomlinson, 2005, p. 263). By considering students holistically, teachers in this study may be working to create safe learning environments and by adjusting linguistic input and output expectations teachers seem to be using instructional strategies flexibly. From the perspective of teachers in this study, teachers cannot simply teach “in a foreign language” or even “with and through a foreign language” (Eurydice, 2006, p. 8 as cited in Coyle et al., 2010, p. 3) but must, based on knowledge of their students, calibrate their instruction to meet the cognitive, linguistic, and affective development of students. Additionally, using knowledge of students to inform instruction suggests an un‐ derstanding of language learning in CLIL that is in line with “complexity theory” (Larsen-Freeman, 2015) (described in section 2.2.3). Complexity theory (CT) describes language learning in holistic terms and rejects the “scientific reductionism” (van Lier, 2000) associated with assigning causality to constructs such as input and output in the process of language learning. Instead, CT views language learning as a process in which a language learner, each with his or her own set of attitudes, experiences, moods etc. is agentic and therefore makes use of (or not) an “affordance” (van Lier, 2000; Larsen-Freeman, 2015) in the learning environment. Teachers described their instruction as conditional and based on what they know about students. They reported using their knowledge of students to, for example, plan the types of activities and social forms most appropriate to the group and to make interactive decisions about when to encourage or push a student to speak English. Such cognitions suggest an approach to CLIL instruction that recognizes the student’s central role in his or her own learning. The teacher can work to plan and create “affordances” for student learning but cannot learn for the student. In this way, teachers seem to ascribe, at least in part, to a CT view of language learning. Teachers described a variety of scaffolding techniques which they used to support students’ understanding of content and language. Of the eight subcategories included under the main category of CLIL approaches, Scaffolding contains the most coded segments with 123. The second most quantitatively significant subcategory is Focus on input with 79 196 7 Discussion <?page no="197"?> coded segments. These numbers suggest a central feature of teachers’ thinking about their CLIL instruction: CLIL instruction is essentially scaffolded input. Teachers did not describe CLIL-specific methodological models of practice such as Coyle et al.’s (2010) 4Cs (Content, Communication, Cognition, Culture), Gierlinger’s (2017) S.A.L.T (Strategic languaging, All language practices, multi-modal Literacies, Topic-relevant language), or Meyer’s (2010) CLIL-Pyramid (Topic, Media, Language skills, Task, Higher order thinking). In describing a variety of scaffolding techniques such as visualization and simplifying language, teachers seem to be applying their knowledge of best practice from the foreign language teaching context to the CLIL context. Considering that none of the teachers reported attending or completing CLIL-specific training outside of their CLIL practice, applying practices from the EFL context to the CLIL context is unsurprising. Several specific types of scaffolding strategies described seem to corroborate other cognitions expressed by the teachers. For example, teachers described using translation both to maintain student understanding and to repair student understanding after it had broken down. Perhaps obviously, such a scaffolding technique would appear to corroborate teachers’ stated prioritization of content learning in CLIL. Content needs to be learned regardless of the language used. Also, all of the teachers described using visualization techniques to support content and language understanding. Such techniques would seem to be far more useful and powerful when used with concrete topics that require less language use. Indeed, these are the types of topics that teachers in this study described as being most conducive to CLIL instruction. The visualization strategies as well as the various forms of L2 strategies described by teachers in this study have also been reported by Massler and Ioannou-Georgiou (2010) based on their investigation of scaffolding strategies in primary school CLIL in Germany and Cyprus. There is also some evidence that teachers in this primary school context value the use of materials such as dictionaries to support language learning. Such use of vocabulary-focused materials was also found to be important in Hüttner and Dalton-Puffer’s (2013) investigation of vocational school CLIL teachers in Austria. When it comes to scaffolding output, teachers most often described simply encour‐ aging children to speak. Such a scaffolding technique is not described by Massler and Ioannou-Georgiou (2010). As already mentioned, students in the CLIL lessons spoke German most of the time. While this is likely due to their age and level of English competences, two teachers told me, in conversation and not in the context of the interviews, that they have or have had German speaking students who, for various reasons, have high or even native-like English competences but who are reluctant to speak English in the CLIL lessons. It is beyond the scope of this study to attempt to explain why that is the case. Regardless, it seems that there is some degree of German-speaking inertia in the CLIL classes and that teachers struggle to overcome this inertia. Therefore, one of the very basic output scaffolding strategies is simply to encourage children to speak English. Based on classroom observations, students produced very little English output in their CLIL lessons. The scaffolding strategy of encouraging children to speak English seems not to support students in producing much spoken English. As discussed briefly above, this is problematic from a second language acquisition perspective since producing output supports second or foreign language acquisition (Swain, 1995). Massler and Ioannou-Geor‐ 7.3 Teachers’ cognitions about approaches most effective for achieving CLIL goals 197 <?page no="198"?> giou (2010) offer a number of verbal output scaffolding techniques that they observed in CLIL lessons in Germany and Cyprus. Such techniques include, for example, providing key terminology and phrases, providing sufficient wait time for students, and offering supportive corrective feedback (Massler & Ioannou-Georgiou, 2010, p. 62). Bilingual strategies such as planning input in one language and output in the other (Baker, 2011) or using two languages alternately to describe pictures (Heimes, 2010) may also be helpful to more explicitly target and thus support the development of productive competences. Finally, teachers described motivating students through the types of activities planned and through social support. Motivation is a theme that emerged in the context of all three research questions. Teachers described good CLIL topics as topics that are motivating, they described developing motivated language learners as a goal for their CLIL instruction, and described the importance of motivating students through activities and social support in their CLIL instruction. Such an emphasis on motivation in the context of CLIL would seem to underline teachers’ description of learning through a foreign language in a CLIL context as challenging. Teachers felt that planning a variety of different activities, planning activities that are playful and active, having students work with one another, or having more competent peers model foreign language use all had the potential to keep students motivated and engaged and therefore to overcome the challenge of foreign language use in CLIL. 7.4 Influence of the DIPS program design on teachers’ cognitions Teachers’ cognitions are complex and influenced by a variety of factors (Borg, 2003). While no attempt was made in this study to establish direct links between teachers’ cognitions and programmatic features, a discussion of possible programmatic influences on specific cognitions may be helpful to readers considering the generalizability of the results of this study to their particular CLIL context. CLIL-style education is always implemented in ways specific and unique to local conceptualizations of CLIL. The DIPS program is organized around three core features (see section 5.3.1 above): 1) two teachers, a German-speaking and an English-speaking teacher, teach all Science and Social Studies lessons together; 2) immersion, as an approach to language teaching and learning, is implemented; and 3) the English-speaking teacher only uses English with students and the German-speaking teacher only uses German. Of these three features, the use of two teachers in Science and Social Studies lessons is likely the most unique aspect of the DIPS program. The fact that a German-speaking and an English-speaking teacher are working together in the CLIL context may have ramifications for how teachers think about language use and learning in CLIL. In this study, teachers described the need for the German-speaking teacher to translate in order to scaffold L2 and content learning. Such a cognition would seem to stem ipso facto from the use of two teachers in all Science and Social Studies lessons. CLIL teachers working alone, of course, would not describe such a cognition. Beyond this fairly obvious point, teachers described several other cognitions related to the use or learning of the L1 and L2. For example, teachers described the learning of German subject language as 198 7 Discussion <?page no="199"?> necessary. Such a cognition is in line with broad CLIL goals such as the “age-appropriate” development of L1 competences (Mehisto et al., 2008, p. 12) as well as one of the goals of CLIL in Germany, namely the development of subject discourse skills in two languages (Diehr, 2012; KMK, 2013). Nonetheless, the DIPS program specifically describes the use of two teachers in the Science and Social Studies lessons as serving, in part, the role of supporting the development of oral and written language skills in two languages: Das Fach Sachunterricht mit ca. drei Wochenstunden nimmt hierbei eine zentrale Rolle ein. Es wird gemeinsam von der deutschsprachigen und der englischsprachigen Lehrkraft unterrichtet und geplant. Daher können die Schüler*innen besonders in diesem Fach ihren mündlichen und später auch schriftlichen Wortschatz in beiden Sprachen parallel erweitern. (Science and Social Studies lessons play a central role here with approximately three lessons per week. These lessons are taught and planned jointly by the German-speaking and English-speaking teachers. This allows pupils to expand their oral and later written vocabulary in both languages at the same time, particularly in this subject.) (DIPS, 2020, Sept. 7) In this case, the cognition that L1 subject language skills are necessary likely stems from both facts; the continued development of L1 competences is a CLIL goal and the DIPS program calls for the development of language skills not only in the L2 but in the L1 as well. In addition to the thinking that learning subject language in the L1 is necessary, teachers described the use of German as serving the purpose of managing linguistic and content complexity. In this case, it is unlikely that such a cognition stems directly from any DIPS-specific feature. Indeed, Frisch (2016, p. 93) reports that teachers use the L1 in their CLIL instruction for translation and to manage complex content. That teachers think the use of the L1 is beneficial for managing complexity would likely be a cognition found in other CLIL contexts as well. The DIPS program is also built upon the concepts of immersion and one person-one-lan‐ guage. The concept of immersion is certainly not unique to the DIPS program. The implementation of the one person-one language approach stems from the immersion aspect of the DIPS program as well as from the use of two teachers in the Science and Social Studies lessons. Again, several cognitions reported in this study may stem directly from these DIPS features. For example, teachers described the process of second language acquisition in CLIL as a natural process and one that proceeds similarly to the process of L1 acquisition. As Teacher B states, “It just comes naturally to them. It’s like how they learn their mother tongue” (TB1 00: 11: 56). All of the teachers also described the need to provide as much oral English input as possible in order to support the development of the L2. Both of these cognitions are strongly linked to an immersion approach to learning. As Wolff and Sudhoff (2015) argue, immersion-style education is limited to young learner settings in which large amounts of input provide the foundation for natural language acquisition. Furthermore, the DIPS program describes immersion as the approach to language learning that most closely mirrors the L1 language learning process. Anders als im herkömmlichen Fremdsprachenunterricht erlernen die Kinder die Sprache nicht systematisch, sondern durch alltägliche Aktivitäten. Es ist die Methode, die dem muttersprachlichen 7.4 Influence of the DIPS program design on teachers’ cognitions 199 <?page no="200"?> Spracherwerb am Nächsten kommt. (Unlike in standard foreign language lessons, the children do not learn the language systematically, but through everyday activities. It is the method that comes closest to native language acquisition.) (DIPS, 2020, Sept. 7) It is plausible that these two cognitions, that L2 learning in CLIL is a natural process and that teachers should provide as much L2 input as possible, stem, in part, from DIPS program statements about immersion. But, because immersion is not unique to the DIPS program, it is unlikely that these cognitions would be unique to DIPS teachers. Teachers in this study also described two cognitions related to choice: 1) that learning English in CLIL is curricular enrichment and a choice that students can make; and 2) that students can choose which language to use in CLIL lessons. The DIPS program describes CLIL lessons as a context in which students can be encouraged to use the L2 in different situations (DIPS, 2020, Sept. 7), suggesting that students should not be required to speak English. It is doubtful that the use of the word “encourage” (anregen in the original) as opposed to, for example, “require” would alone lead the majority of teachers to describe learning and using English as a choice. It is more likely that these cognitions are linked to the fact that, in Germany, CLIL is guided by the non-language subject curriculum (KMK, 2013, p. 7) and that teachers, as reported in this study, prioritize content learning. As already noted, these cognitions are not unique to teachers in this study. Teachers in previous studies have also reported prioritizing content learning (Skinnari & Bovellan, 2016; Bovellan, 2014; Costa, 2013; de Graaff et al., 2007) and viewing the acquisition of the L2 as a welcome “by-product” to content learning in CLIL (Skinnari & Bovellan, 2016, p.-153). To summarize, the most unique feature of the DIPS program is likely its use of two teachers for Science and Social Studies lessons. The cognition that stems most directly from this feature is the use of the German-speaking teacher to translate or mediate in the lessons. Such a cognition is likely to be unique to the DIPS program considering that most other public school CLIL programs use one teacher for CLIL Science and Social Studies lessons. Other cognitions, for example, that the acquisition of a foreign language in CLIL is a natural process, may be reasonably linked to DIPS program features, for example, immersion, but will only be unique to the extent that the linked program feature (e.g. immersion) is unique to DIPS. 7.5 Summary In the above discussion, two points seem particularly relevant to the discussion of further developing primary school CLIL pedagogy. First, teachers’ positioning of L2 development in the context of CLIL learning as curricular enrichment or a choice is far away from and far more modest than the stated goal of CLIL in Germany, namely the development of subject literacies or discourse competences in two languages (Diehr, 2012; KMK, 2013). Frisch (2021), based on the results of her investigation into the language and subject competences of primary school CLIL students, questions whether the development of subject literacies in two languages is appropriate for primary school CLIL learners. However, there is currently no alternative goal that has been articulated in the German context. Any such goal will need to consider the theoretical foundations and principles of CLIL while also taking into 200 7 Discussion <?page no="201"?> account the nature of (language) teaching and learning with young children. Additionally, stakeholder perspectives, i.e. those of teachers, students, parents, and administrators, may also need to be considered when formulating a goal that teachers are willing and able to work toward. Based on the results of this study, teachers’ foreign language learning goals are largely divorced from the current theoretical discussion of such goals. Second, the lack of student L2 output as observed during CLIL instruction in this study presents a major challenge to the development of more advanced L2 competences. Producing L2 output is a central trait of good foreign language instruction (Helmke et al., 2008, p. 361) and serves several functions in the development of L2 competences including, for example, noticing gaps in productive competences and testing hypotheses about how the language works (Swain, 1995). Furthermore, it may push students to be more active and take more responsibility for their L2 development (Swain, 1995, p. 126). Without opportunities to produce, in spoken or written form, L2 output, students’ L2 development will likely be hindered. The majority of teachers in this study described wanting students to speak and understanding that speaking is key to foreign language learning. As Teacher B states, “I always tell them, “It's a language. You can only learn a language when you speak it”” (TB1 00: 41: 08). The lack of speaking and writing competences in the L2 is likely characteristic of many primary school CLIL learners and thus of primary school CLIL interaction generally. Indeed, Schwab (2020), based on her investigation into primary school CLIL classroom interaction, reports long stretches of student talk in the L1 since “[o]bviously, teachers did not insist on using English as the medium of communication” (Schwab, 2020, p.-107). While teachers in this study generally wanted their students to speak English in their CLIL lessons, the observed classroom interaction seems to be congruent with other observed primary school CLIL interaction in which student L1 output dominates. A central question then, not only for DIPS teachers but for primary school CLIL teachers generally, is how to increase student L2 output during CLIL instruction. In addition to encouraging students to speak English, DIPS teachers in this study described using structured output or chunks as well as known vocabulary from EFL lessons to scaffold learners’ L2 output. Pre-teaching vocabulary and scaffolding output through the use of chunks as well as allowing sufficient wait time and using supportive error correction are all ways of effectively scaffolding L2 output in CLIL (Massler & Ioannou-Georgiou, 2010). Such output scaffolding strategies need to be a key component of CLIL teacher in-service and pre-service training. In the case of DIPS teachers, a natural approach to language acquisition will only be effective if students naturally use the L2. Based on classroom observations, this was rarely the case. Therefore, DIPS teachers will need to find ways to target and scaffold opportunities for students to speak and write in the L2. The goal of developing subject literacy in two languages with students whose speaking competences are too low in order to be able to communicate through the L2 in the CLIL classroom is challenging and may need to be reformulated to be more appropriate for the primary school CLIL context. Regardless of whether and how this goal is reformulated, additional strategies to support L2 output in the classroom are necessary for the further development of CLIL pedagogy at primary school. 7.5 Summary 201 <?page no="202"?> Finally, the core features of the DIPS program need to be taken into account when considering the generalizability of the results of this study. Program features such as an immersion approach to language learning and the use of two teachers in the Science and Social Studies lessons may directly influence several of the teachers’ reported cognitions. 202 7 Discussion <?page no="203"?> 8 Conclusion The purpose of this study is twofold. First, the study has been conducted in order to explore and describe primary school CLIL teachers’ cognitions. Research in this field is limited but nonetheless important due to the unique and challenging teaching and learning context that primary school CLIL is. Second, because the study has been conducted in the context of the DIPS program, the results of this study serve to describe the DIPS program from teachers’ perspectives. These results can be used by DIPS teachers and administrators to further develop their unique CLIL program. In order to achieve these two primary goals, three research questions have been investigated: 1. What cognitions do teachers hold about teaching and learning in primary school CLIL? 2. What language-related educational goals do teachers work toward? 3. What approaches do teachers think are most effective for achieving these goals? These three research questions have been investigated by conducting semi-structured interviews with five teachers in the DIPS program and by observing a series of CLIL lessons in each of the teachers’ classes. Transcripts of the interview data were analyzed using qualitative content analysis through which interview data was segmented, coded, and organized into a coding frame. Classroom observations provided evidence of classroom practices and were used as a foundation on which to base the follow-up interviews. The results have been described in chapter six and discussed in chapter seven. The results, as presented in these two chapters, provide detailed answers to the three research questions and fulfill the aims of the study, namely to explore primary school CLIL teachers’ cognitions and to describe the DIPS program from teachers’ perspectives. The results of this study describe what teachers teaching in the DIPS program think about their CLIL pedagogy. The cognitions described are based on teachers’ experiences with CLIL and are therefore “complex, practically-oriented, personalized, and context-sensitive networks of knowledge, thought, and beliefs” (Borg, 2003, p.-81). Some of the results of this study may not be a direct reflection of the DIPS program itself. For example, that all of the teachers described CLIL instruction as leading to high levels of language learning does not indicate in any scientifically verifiable way how much students’ L2 competences develop due to CLIL instruction. Nonetheless, it is important to know teachers’ perceptions in relation to a variety of factors related to CLIL including, for example, learning outcomes such as L1 and L2 development and content learning. Gaps between teachers’ perceptions and empirical evidence of learning outcomes can be opportunities to reflect on and further develop teachers’ CLIL practices. Below, I will synthesize and summarize six conclusions from this study. These six conclusions describe how the teachers in this study perceive and understand their own CLIL practices and student learning in the context of the DIPS program. Each conclusion will be discussed critically in the context of literature on CLIL learning at primary school and observed classroom interaction. <?page no="204"?> 1. DIPS teachers prioritize content learning and consider learning English a curricular enrichment. Teachers in this study clearly prioritized content learning and described learning English as curricular enrichment. CLIL programs have been described as “enrichment programmes” (Baetens Beardsmore, 1993, p. 3) and have faced criticism on grounds that they are elitist (Bruton, 2013). However, far from being only a form of curricular enrichment or education for the advantaged, García (2009, p. 5) sees bilingual education as “the only way to educate children in the twenty-first century” (emphasis in original). These differing perspectives may be based on the different language contexts in which the authors are working. While Baetens Beardsmore and Bruton are writing from an English as a foreign language context in Europe, García is writing from an English as a second language context in bilingual regions of the United States. Furthermore, García (2009, p. 5) is attempting to articulate an “integrated plural vision for bilingual education,” one that attempts to redefine the boundaries of “named languages” (García & Otheguy, 2019, p. 2). Whether a bilingual program such as CLIL or immersion is deemed necessary or simply a matter of enrichment would seem to be dependent on local, regional, and national language politics as well as the language profiles of learners. In this study, teachers see their CLIL instruction as a form of language learning enrichment. Individual students can choose the extent to which they engage with (and learn) the English language. The learning priority remains on content learning. As stated previously, this is likely a reflection of the political reality in NRW and Germany, namely, that CLIL is subject learning through a foreign language. Such learning is guided by the non-language subject curriculum (KMK, 2013, p. 7). Furthermore, the use of English plays no official role in the local community nor was observed to be a significant language in the language profiles of DIPS learners. Instead, the value of English seems to remain in abstract notions of European integration and globalization (Wolff, 2016). DIPS teachers, therefore, described learning English as curricular enrichment. 2. Learning German subject vocabulary is necessary. However, German is used primarily to support understanding and ensure content learning. Most teachers in this study stated that students needed to know subject terminology in German and, especially when the term and associated concept are new, that learning the German term needs to precede learning the English term. The stated goal of learning subject terminology in the L1 falls short of developing subject literacy or subject competences in the L1. Developing subject literacy in the L1 and L2 is the stated goal of CLIL education in Germany (Diehr, 2012; KMK, 2013). German was rarely used in a way that would support the development of L1 subject literacy. Instead it was used to maintain and repair understanding and to ensure content learning. By using the L1 to maintain and repair understanding in unplanned classroom interaction, teachers’ practices, and perhaps their cognitions as well, are most congruent with Diehr’s (2012) description of Type A CLIL instruction (see section 2.1.3). 204 8 Conclusion <?page no="205"?> 3. Learning through a foreign language in CLIL is challenging. Such learning requires motivation and support and is an opportunity to develop students’ self-confidence and experience success. Teachers described learning through a foreign language as challenging. This is not particularly surprising since primary school students’ L2 competences are generally low and therefore learning through such low language competences is bound to be challenging. Teachers described such a context as one that requires them to motivate and support students and as one that presents an opportunity for students to develop self-confidence and experience success. Researchers in secondary school contexts have also found evidence that CLIL teachers work toward general pedagogic goals such as motivation (Wegner, 2012; Skinnari & Bovellan, 2016). It is possible that such goals have only been described by teachers in the absence of more readily achievable L2 learning goals. Frisch (2021) provides evidence that primary school CLIL learners largely fail to demonstrate content knowledge through the L2. Developing, for example, self-confidence by achieving initial levels of foreign language competences may only function as a kind of consolation prize for students who do not achieve high levels of L2 learning. Nonetheless, there is no logical reason that CLIL, as an environment that teachers described as challenging due to the use of learning through a foreign language, would not also be a conducive environment for the development of general pedagogic goals. 4. Foreign language learning in CLIL, in addition to being a cognitive and linguistic process, is an affective process. Students’ emotions must be considered when planning and implementing instruction. Teachers describe using their knowledge of students to calibrate instruction to individual and group needs. That affect plays a significant role in foreign language learning has long been recognized (Bown & White, 2010). Krashen’s (1985) affective filter hypothesis is an often-cited explanation of how affect affects language learning. Bown and White (2010, p. 332) equate affect with emotion and argue “for the importance of understanding the individual and social antecedents of emotions and the relationship between emotion and cognition in SLA.” Similarly, Rossa (2020, p. 176) argues that teachers should take a holistic approach, one which takes into account students’ emotions, when, for example, planning lessons and assessments. The results of this study suggest that the teachers in the DIPS program recognize the importance of such affective factors in language learning and, based on their knowledge of students, work to adapt instruction and expectations in order to meet students’ affective needs and the “individual and social antecedents” (Bown & White, 2010, p. 332) of those needs. The way, however, that teachers spoke of using their knowledge of students suggests that some students are encouraged to, for example, speak English more than others. But um, if I know that the children in the Science, in the Sachunterricht lessons, if I know that they're able to say it in English, I would try to get it out of them in English. […] But then if I know/ not all the children, I wouldn't force all of them, but some I would really force like, "Can you say that in English? You can do that." (TB1 00: 41: 47) 8 Conclusion 205 <?page no="206"?> An approach such as that expressed by Teacher B above likely has many merits and would seem in line with a differentiated approach (Tomlinson, 2005) to teaching and learning. Nonetheless, it raises the question of whether such attention to an individual’s affective response to language learning may, in some cases, inhibit an individual’s opportunities to produce L2 output and thus develop L2 competences. Such differentiation based solely on a teacher’s knowledge or assessment of students can lead to limiting a student’s learning op‐ portunities. To ensure a variety of learning opportunities for different learners, Rossa (2020, p. 177) recommends that teachers and students discuss potentially emotionally-problematic situations, e.g. producing L2 output, and how best to manage them. By addressing the affective component of language learning directly and openly with students, teachers may be able to maintain more language learning opportunities for more students. 5. Regarding L2 development, oral skills and subject vocabulary should be prioritized. Teachers in this study reported targeting primarily L2 oral skills and subject vocabulary. Such goals are typically described by primary school EFL teachers (Schauer, 2019) and CLIL teachers (Imgrund, 2004; Morton, 2012; Bovellan, 2014). However, such goals, again, fall short of developing subject literacy in two languages (Diehr, 2012; KMK, 2013). While developing oral language competences and subject vocabulary in the L2 may fall short of the goal of developing subject literacy in two languages, such goals may also be more realistic for primary school learners. Frisch (2021), based on evidence that primary school CLIL learners largely fail to demonstrate subject competences in the L2, has questioned whether the development of subject literacy in two languages is appropriate for primary school learners. 6. CLIL instruction should include high amounts of L2 input that is scaffolded in various ways. Teachers’ CLIL instruction was observed to be high on L2 input and low on L2 output. Teachers in this study overwhelmingly described the importance of producing high amounts of L2 input for students. In primary school CLIL contexts in which students’ L2 competences are generally low, high amounts of teacher talk or teacher-produced input is necessary in order for students to begin acquiring the target language (Wode, 2009, p. 100). In this study, teachers described scaffolding their input in a variety of ways including, for example, translation, visualization, and L2 strategies such as using simplified language and repeating and reusing language. Teachers also described, albeit to a lesser extent, the importance of student-produced output. To scaffold such output, teachers described encouraging children to speak and using structured language or chunks that students could repeat. While the majority of observed lessons included high amounts of oral input from teachers, students rarely used the L2 in the CLIL lessons. Schwab (2020) also reports low amounts of L2 output from students in a primary school CLIL context. It is perhaps obvious that students did not use the L2 because, at their age, they simply do not have the L2 competences to use the L2 productively to any significant extent. Schwab (2020, p. 107) also states that it is “obvious[s]” that teachers do not require their students to use the L2 when discussing the subject. Despite it perhaps being obvious that students do not and are not 206 8 Conclusion <?page no="207"?> required to speak the L2 much in CLIL lessons, it nonetheless presents a problem for the DIPS program and for primary school CLIL programs generally since producing output is vital to second or foreign language acquisition (Swain, 1995). With output production vital to foreign language acquisition and student output largely absent from the CLIL lessons observed in this study, the “ambitious” (Frisch, 2021, p. 46) goal of developing subject literacy in two languages would seem almost impossible for students in the DIPS program to achieve. 8.1 Limitations The results of this study need to be understood in the context of several limitations. First, the study is a small, qualitative study of five teachers teaching in one local CLIL context. The cognitions expressed by the teachers in this study may not always be congruent with the cognitions of other teachers teaching within the DIPS program or within the broader CLIL community. The results presented in this study are based only on a sample of DIPS teachers. The results can be used as an additional data point which adds to existing primary school CLIL research. As a single data point in the broader context of CLIL research, the results of this study can add to an understanding of CLIL teachers’ cognitions. Second, the interviews in which teachers expressed their perspectives on CLIL were “jointly constructed encounters” (Richards, 2009, p. 190 emphasis in original) in which meaning has been co-constructed between interlocutors (Holstein & Gubrium, 1995, p. 4). The verbal data which resulted from the interviews was created jointly by two individuals (the researcher and the teacher) interacting within a social context. The social context included, for example, personal characteristics (Prior, 2018, p. 240) and power relationships (Canh & Maley, 2012, pp. 98-99). Furthermore, the data gathered was influenced by researcher interest and questions asked. While measures were taken to establish rapport and trust (Dörnyei, 2007; Mirhosseini, 2020) in order to allow teachers to express their views openly, the data gathered remains the product of a “co-constructed social exchange” (Dönyei, 2007, p.-141). Third, the reliability of the data and its ultimate meaning must be understood within the context of the study and several embedded limitations. As described in section 5.4.2 above, two measures were taken to improve the internal reliability of the analysis. First, data was coded by one coder (the researcher) over time. In doing so, I intended to improve the reliability of the analysis through “time triangulation” (Canh & Maley, 2012, p. 98). Second, a second coder independently coded previously-segmented data for approximately half of the subcategories in the coding frame. While these techniques undoubtedly improved the reliability of the analysis from what it would have been had one coder coded the data only one time, they nonetheless fall short of an analysis conducted by multiple researchers working together. Fourth, there is the question of whether the descriptive nature of teachers’ cognitions reported in this study can truly be considered cognitions at all. The results of this study in their entirety are summarized and outlined in the form of the coding frame (presented in pieces throughout chapter six). As such, the results are descriptions and categories of thought. For example, the research question, “What language-related educational goals do 8.1 Limitations 207 <?page no="208"?> teachers work toward? ” is answered by two sub-categories which are each further divided into sub-categories which, taken together, ultimately answer the research question. In addition to the coding frame presented, I further describe and, to a limited extent, interpret what teachers told me in their interview responses. In chapter six, I presented multiple views expressed by teachers and presented them in a nuanced way. In chapter seven, I tried to further interpret and critically discuss the results in the context of CLIL theory and empirical evidence. By doing this, it was my intention to heed Borg’s (2019) advice not only to describe teachers’ cognitions but to try to understand them. I have tried to understand teachers’ cognitions by contextualizing them. However, this study has not attempted to understand the personal, professional, educational or other antecedents of teachers’ cognitions. Understanding such antecedents to teachers’ cognitions would no doubt be useful. Nonetheless, the results of this study are limited primarily to descriptions of teachers’ cognitions as they relate to their professional CLIL practice. Finally, the results presented in this study describe teachers’ cognitions and not teachers’ practices. In the analysis of the teacher interviews as carried out using the method of QCA, I have attempted to isolate what teachers think about their CLIL practice. As described in section 5.3.4, I only segmented and coded data which expressed a cognition about practice and not merely a description of practice. Evidence from classroom observations has been included in the results and discussion chapters (chapters six and seven) in order to further contextualize teachers’ cognitions and to provide evidence of teachers’ practices. This observational evidence has allowed me to discuss teachers’ cognitions in more meaningful ways. Nonetheless, it is teachers’ cognitions, and not their practices, that are the foundation of the results. 8.2 Implications CLIL teachers’ cognitions have been investigated and described in this study. Teacher cognition research can be used to promote “meaningful teacher development” (Kubanyiova & Feryok, 2015, p.-437) and to inform the development of pedagogy. If teacher cognition research can be used to promote teacher learning, then one further use of this study could be as a basis for teacher reflection. The results of the study can act as a mirror in which teachers can see their cognitions and practices and make adjustments to them as they see fit. Such a process of reflection can be ongoing and take place collaboratively with colleagues (Li, 2020, p. 277). Teacher cognition research can give teachers the language through which the reflection process can take place (Calderhead, 1987b). The results of this study can also inform pre-service and in-service teacher training. Teachers’ cognitions about teacher-produced input and student-produced output are two key findings of this study that can potentially inform CLIL teacher education. Teachers in this study tended to emphasize the quantity of L2 input over the quality of L2 input and described scaffolding students’ L2 output to a limited extent. CLIL lessons were also observed to be low on student-produced L2 output. For reasons already discussed, the quality of L2 input and scaffolding L2 output are important for L2 development in the CLIL context. Teacher training programs, whether in pre-service or in-service contexts, need 208 8 Conclusion <?page no="209"?> to address how to enhance and scaffold input and how to create more opportunities for students to produce L2 output. Without “manipulated and enhanced” input (Lyster, 2011, p. 612) and without more opportunities to produce L2 output that is scaffolded in various ways, L2 development in CLIL settings will remain limited. In addition to teacher learning, this study can also be used to further develop CLIL pedagogy in the context of the DIPS program. The results of this study indicate that teachers work toward language goals such as oral language competences and vocabulary learning in English (L2) but also see the learning of German (L1) subject vocabulary as necessary. While these goals are generally compatible with the ultimate CLIL goal in Germany of developing subject literacy in two languages (Diehr, 2012; KMK, 2013), there would seem to be room to further develop pedagogy within the DIPS program toward this ultimate goal. The development of discourse competences in two languages typically involves various forms of code-switching (Müller, 2017) or translanguaging (García, Johnson, & Seltzer, 2017) in which two or more languages are used together in various ways and for various purposes. One example of this from the classroom observations has already been described from Teacher D’s lesson focused on the map of Europe. Other examples include activities in which input is planned in one language and output in the other language (Baker, 2011) and “bilingual picture storming” (Heimes, 2010) in which pictures or photographs are discussed in two languages. It is likely, however, that any such forms of planned code-switching in primary school CLIL will need to make use of more structured and patterned language (e.g. chunks) such as those used in Teacher D’s example. Simply discussing a topic in two languages in primary school CLIL is unlikely to be successful. Finally, teachers described goals for their CLIL instruction that include the development of self-confidence and providing opportunities to experience success through the effort required to learn through a foreign language. If these are goals that teachers value and, in part, work toward, then evaluations of CLIL which take such goals into account may be something teachers find affirming. Evaluations of CLIL which describe students’ learning outcomes will remain vital to the development of CLIL pedagogy. However, evaluations which investigate and report “affective evidence” of CLIL (Coyle et al., 2010, p. 136) may be able to add to a fuller understanding of CLIL. 8.3 Further research CLIL, as a European term and construct, is always realized in local contexts. The ways that teachers and schools plan for and implement CLIL are affected by a range of factors including, for example, the language profile of students, the school and surrounding community, the target language, and curricular requirements. Therefore, the way that CLIL is realized in any particular context is bound to be different from other contexts. As Baetens Berdsmore (1993, p. 4) asserts, any variation of bilingual education is “unlikely to be adopted elsewhere.” This suggests that understanding specific CLIL programs as they are locally realized is important for understanding the broader CLIL landscape. One avenue for future research into CLIL, therefore, is to continue to document and describe specific cases of CLIL in order not only to understand those particular cases but to add to the broader knowledge and understanding of CLIL reality. This study has sought, in part, to 8.3 Further research 209 <?page no="210"?> understand one CLIL case, the DIPS program in Dortmund, Germany, through the lens of teachers’ cognitions. However, other qualitative and mixed method approaches that gather and analyze data from sources such as students and parents could also be fruitful for further understanding the variety of CLIL cases that exist. For example, researchers might investigate students’ and parents’ goals for CLIL instruction. Being that Frisch (2021) has questioned the appropriateness, for primary school CLIL programs, of the goal of developing subject literacy in two languages, understanding other stakeholder goals could inform future considerations of and investigations into CLIL goals appropriate to primary school. Once a CLIL case is more fully understood, more research as well as program intervention can take place in order to support the development of pedagogy in various cases of CLIL. Again, there would seem to be a variety of options available to researchers, teacher trainers, and teachers themselves interested in the development of CLIL pedagogy. For example, students in this study did not produce high levels of L2 output in their CLIL lessons. Developing methods or techniques that support student L2 output while maintaining content-focused instruction could be a focus of further research. The current study is based on the belief that the practicing teachers are themselves in the best position to know and describe their CLIL programs and instruction and by extension make changes to their programs and instruction. One possibility to make such change is through Action Research (AR). AR, applied to the educational setting, “aims at making a situation such as a classroom or whole school system better by responding to the continuous need for development or change” (Bradbury, Lewis, & Columbia Embury, 2019, p.-7). The DIPS program has indeed faced the need for change caused most recently by the adoption of the new primary school masterplan (MSB NRW, 2020) which delayed the beginning of EFL instruction in NRW’s primary schools from first grade to third grade. According to several DIPS school websites (for example, DIPS, 2022, March 3), the program continues to function as it did previously in grades three and four while offering one hour per week of an English language study group (Arbeitsgemeinschaft) during the school day for students in grades one and two. Despite offering an English language study group for students in grades one and two, it could be assumed that cutting EFL instruction and, by extension, CLIL instruction for these students will result in students entering CLIL in grade three with lower English competences than previously. Such a situation would seem to exacerbate the already existing challenge of teaching curriculum-defined content through a target language largely undeveloped in students. An AR approach to the further development of the DIPS program puts teachers in charge of asking the relevant questions as well as creating and executing a research plan for the ultimate purpose of improving instruction. Such an approach to research “seeks to identify and solve practical classroom problems” and “aims to improve professional practice” (Barnard, 2021, p. 35). AR, however, also comes with risks. Most notably, there is the risk that AR, conducted by teachers themselves, lacks rigor (McAteer, 2013; Rowell, 2019) and therefore does not accurately or meaningfully inform teachers’ practices (McAteer, 2013) nor add to the wider knowledge base established in the scientific community (Rowell, 2019). Therefore, if DIPS teachers felt that such a research project had the potential to develop and improve their classroom practices, they would need to be supported not only by providing 210 8 Conclusion <?page no="211"?> them time to conduct such research but, crucially, by giving them the methodological tools to gather and analyze data in a rigorous and scientific way (Barnard, 2021). The DIPS program as well as primary school CLIL and EFL contexts in NRW generally face challenges moving forward. Because of the new Masterplan Grundschule NRW (MSB NRW, 2020), EFL instruction in NRW now begins in third grade instead of first grade. For the DIPS schools, this means that bilingual instruction does not begin until third grade. Among other things, this change reintroduces a gap in foreign language learning generally and CLIL or bilingual instruction specifically for those students who attended bilingual preschools and now must wait two years before receiving further foreign language instruction at school. The DIPS program was initiated, in part, to close the gap between preschool bilingual programs and secondary school bilingual programs in Dortmund (Raunser & Steffens, 2012, p. 37). After 11 years of contiguous CLIL programming in Dortmund, a gap has been reintroduced. The decision to start teaching English in third grade instead of in first grade raises the question of the value of foreign language education at primary school. The Masterplan Grundschule as well as a position paper written in response to the decision to delay the start of EFL instruction to third grade describe learning foreign languages, especially English, as valuable and necessary not only for living in Europe but for an increasingly globalized and digital world (MSB NRW, 2020; Bartosch, Frisch, Kötter, & Reckermann, 2020). Living in such a world means using foreign languages and engaging with other cultures. CLIL as well as EFL at primary school offer young students initial experiences “dealing with something that is foreign” (TC2 01: 08: 44) and in this way can support the initial development of the attitudes and skills described by Byram (1997) that are conducive to communicating in intercultural contexts. 8.3 Further research 211 <?page no="213"?> 9 References Akkermann, Tomke (2012). Bilinguales Lernen in der Grundschule: Eine exemplarische Studie des englischsprachigen Sachunterrichts. Diplomica Verlag.- APA (2022a, June 7). APA dictionary of psychology: cognition. American Psychological Association. Retrieved from: https: / / dictionary.apa.org/ cognition APA (2022b, June 7). APA dictionary of psychology: thinking. American Psychological Association. Retrieved from: https: / / dictionary.apa.org/ thinking Aristov, Natasha & Haudeck, Helga- (2013). Natural science. In Daniela Elsner & Jörg Keßler (Eds.), Bilingual education in primary school: Aspects of immersion, CLIL, and bilingual modules (pp. 42-50). Narr Verlag. Asher, James (1981). The total physical response: Theory and practice. Native language and foreign language acquisition 379(1), 324-331. DOI: doi.org/ 10.1111/ j.1749-6632.1981.tb42019.x- Baetens Beardsmore, Hugo (1993). Introduction. In Hugo Baetens Beardsmore (Ed.), European models of bilingual education (pp. 1-5). Multilingual Matters.- Baetens Beardsmore, Hugo (2009). Language promotion by European supra-national institutions. In O. Garcia (Ed.), Bilingual education in the 21 st century: A global perspective (pp. 197-217). Wiley-Blackwell. Bailey, Kathleen (1996). The best laid plans: teachers’ in-class decisions to depart from their lesson plans. In Kathleen Bailey and David Nunan (Eds.), Voices From the Language Classroom (pp. 15-40). Cambridge University Press. Baker, Colin (2011). Foundations of bilingual education and bilingualism (5 th ed.). Multilingual Matters. Barbour, Rosaline S. (2014). Quality of data analysis. In Uwe Flick (Ed.), The Sage handbook of qualitative data analysis (pp. 496-509). Sage.-- Barcelos, Ana Maria F. (2016). Student teachers’ beliefs and motivation, and the shaping of their professional identities. In Paula Kalaja, Ana Maria F. Barcelos, Mari Aro & Maria Ruohotie-Lyhty (Eds.), Beliefs, agency and identity in foreign language learning and teaching (pp. 71-96). Palgrave Macmillan.- Barnard, Roger (2021). Action research. Modern English Teacher, 30(4), 35-37. Bartosch, Roman, Frisch, Stefanie, Kötter, Markus & Reckermann, Julia (2020). Stellungnahme von Vertreter*innen der Fremdsprachenforschung zur Rückverlegung des Englischunterrichts in Klasse 3 in NRW zum Schuljahr 2021/ 2022. Retrieved from: https: / / www.uni-muenster.de/ imperia/ md/ cont ent/ englischesseminar/ tefl/ stellungnahme_zum_masterplan_grundschule_nrw_nov2020.pdf BB BMB (2018). Die UN-Behindertenrechtskonvention Übereinkommen über die Rechte von Menschen mit Behinderungen. Beauftragter der Bundesregierung für die Belange von Menschen mit Behin‐ derungen. Retrieved from: https: / / www.behindertenbeauftragter.de/ SharedDocs/ Publikationen/ D E/ Broschuere_UNKonvention_KK.pdf ? __blob=publicationFile&; v=43 Bechler, Sabrina (2014). Bilinguale Module in der Grundschule: Integriertes Inhalts- und Sprachlernen im Fächerverbund Mensch, Natur, und Kultur. Peter Lang.- Becker, Howard S. (1970). Sociological work: Method and substance. Transaction. Bland, Janice (2015). Drama with young learners. In Janice Bland (Ed.), Teaching English to young learners: Critical issues in language teaching with 3-12 year olds (pp. 219-238). Bloomsbury. <?page no="214"?> Bohnsack, Ralf (2013). Documentary method. In Uwe Flick (Ed.), The Sage handbook of qualitative data analysis (pp. 217-233). Sage. Bonnet, Andreas & Breidbach, Stephan (2017). CLIL teachers’ professionalization: Between explicit knowledge and professional identity. In Ana Llinares & Tom Morton (Eds.), Applied linguistics perspectives on CLIL (pp. 269-285). John Benjamins. Bonnet, Andreas (2012). Towards an Evidence Base for CLIL: How to integrate qualitative and quan‐ titative as well as process, product, and participant perspectives in CLIL research. International CLIL Research Journal, 1 (4), 66-78. Retrieved from: http: / / www.icrj.eu/ 14/ article7.html. Borg, Michaela (2001). Key concepts in ELT: Teachers’ beliefs. ELT Journal, 55(2), 186-188. Borg, Simon (2003). Teacher cognition in language teaching: A review of research on what language teachers think, know, believe, and do. Language Teaching, 36(2), 81-109. DOI: 10.1017/ S0261444803001903 Borg, Simon (2015). Teacher cognition and language education: Research and practice. Bloomsbury. Borg, Simon (2016, April 25). Do teachers’ beliefs really matter? Simon Borg Blog. Retrieved from: http: / / simon-borg.co.uk/ do-teachers-beliefs-really-matter/ #: ~: text=Beliefs%2C%20of%20course%2 C%20do%20matter,be%20important%20areas%20of%20research. Borg, Simon (2019). Language teacher cognition: Perspectives and debates. In Xuesong Gao (Ed.), Second handbook of English language teaching (pp. 1-23). Springer Nature. DOI: http: / / doi.org/ 10. 1007/ 978-3-319-58542-0_59-1; Borg, Simon & Sanchez, Hugo S. (2020). Cognition and good language teachers. In Carol Griffiths & Zia Tajeddin (Eds.), Lessons from good language teachers (pp. 16-27). Cambridge University Press. Böttger, Heiner (2020). Englisch lernen in der Grundschule: Eine kindgerechte Fachdidaktik (3rd edition). Julius Klinkhardt.- Botz, Lieselotta & Diehr, Bärbel. (2016). Bilinguales Lernen in der Grundschule---Einblicke in das BiSY-Projekt (Bilingualer Sachunterricht - Young Learners). In Bärbel Diehr, Angelika Preisfeld & Lars Schmelter (Eds.), Bilingualen Unterricht weiterentwickeln und erforschen (pp. 245-260). Peter Lang. Botz, Lieselotta & Frisch, Stefanie (2016). Fachliteralität im bilingualen Sachunterricht der Grund‐ schule. In Heiner Böttger & Norbert Schlüter (Eds.), Forschritte im Frühen Fremdspachenlernen: Tagunsband zur 4. FFF-Konferenz (pp. 240-251). Bovellan, Eveliina (2014). Teachers' beliefs about learning and language as reflected in their views of teaching materials for Content and Language Integrated Learning (CLIL). [Doctoral dissertation, University of Jyväskylä] Retrieved from: https: / / jyx.jyu.fi/ handle/ 123456789/ 44277 Bradbury, Hilary, Lewis, Rolla & Columbia Embury, Dusty (2019). Education action research: With and for the next generation. In Craig A. Mertler (Ed.), The Wiley handbook of action research in education (pp. 117-138). Wiley. Breidbach, Stephan (2002). Bilingualer Sachfachunterricht als neues interdisziplinäres Forschungs‐ feld. In Stephan Breidbach, Gerhard Bach & Dieter Wolff (Eds.), Bilingualer Sachfachunterricht: Didaktik, Lehrer-/ Lernerforschung und Bildungspolitik zwischen Theorie und Empirie (pp. 11-27). Peter Lang. Breidbach, Stephan & Viebrock, Britta (2012). CLIL in Germany: Results from recent research in a contested field. International CLIL Research Journal, 1(4), 5-16. Brown, Douglas (2014). Principles of language learning and teaching (6th edition). Pearson.- 214 9 References <?page no="215"?> Brown, James Dean & Rodgers, Theodore S. (2002). Doing second language research. Oxford University Press. Bown, Jennifer & White, Cynthia (2010). A social and cognitive approach to affect in SLA. International Review of Applied Linguistics in Language Teaching,-48(4), 331-353.-DOI: https: / / doi.org/ 10.1515/ ir al.2010.014 Brunsmeier, Sonja (2016). Interkulturelle Kommunikative Kompetenz im Englischunterricht der Grund‐ schule: Grundlagen, Erfahrungen, Perspektiven. Narr. Bruton, Anthony (2013). CLIL: Some of the reasons why…and why not. System, 41, 587-597. DOI: htt p: / / dx.doi.org/ 10.1016/ j.system.2013.07.001 Bruton, Anthony (2019). Questions about CLIL which are unfortunately still not outdated: A reply to Pérez-Cañado. Applied Linguistics Review, 10(4), 591-602. Buchholz, Barbara (2007). Reframing young learners’ classroom discourse structure as a preliminary requirement for a CLIL-based ELT approach: An action research project on conversational language learning for primary students. In Christiane Dalton-Puffer & Ute Smit (Eds.), Empirical perspective on CLIL classroom discourse (pp. 51-77). Peter Lang. Burmeister, Petra, Ewig, Michael, Frey, Evelyn & Rimmele, Marisa (2013). CLIL -Teacher Training at the University Level: Bridging the Gap between Theory and Practice. In Stephan Breidbach & Britta Viebrock (Eds.), Content and Language Integrated Learning (CLIL) in Europe: Research perspectives on policy and practice (pp. 145-156). Peter Lang. Burns, Anne, Freeman, Donald, & Edwards, Emily (2015). Theorizing and studying the lan‐ guage-teaching mind: Mapping research on language teacher cognition. The Modern Language Journal, 99(3), 585-601. Butzkamm, Wolfgang & Caldwell, John A.W. (2009). The bilingual reform: A paradigm shift in foreign language teaching. Narr. Byram, Michael (1997). Teaching and assessing intercultural communicative competence. Multilingual Matters. Calderhead, James (Ed.). (1987a). Exploring teachers’ thinking. Cassell Education. Calderhead, James (1987b). Introduction. In James Calderhead (Ed.), Exploring teachers’ thinking (pp. 1-19). Cassell Education.- Calderhead, James (1996). Teachers: Beliefs and knowledge. In Handbook of educational psychology (pp. 709-725). Simon & Schuster Macmillan.- Cambridge University Press ELT. (2010, November 11). David Marsh on CLIL [Video]. Youtube. Retrieved from: https: / / www.youtube.com/ watch? v=-Czdg8-6mJA Cameron, Lynne (1997). The task as a unit for teacher development. ELT Journal, 51(4), 345-351. Cameron, Lynne (2001). Teaching languages to young learners. Cambridge University Press.- Canh, Le Van & Maley, Alan (2012). Interviews. In Roger Barnard & Anne Burns (Eds.), Researching language teacher cognition and practice: International case studies (pp. 90-108). Multilingual Matters. Carter, Kathy & Doyle, Walter (1987). Teachers’ knowledge structures and comprehension processes. In James Calderhead (Ed.), Exploring teachers’ thinking (pp. 147-160). Cassell Education. Castellano-Risco, Irene, Alejo-González, Rafael & Piquer-Píriz, Ana M. (2020). The development of receptive vocabulary in CLIL vs EFL: Is the learning context the main variable? System, 91, DOI: 10.1016/ j.system.2020.102263 Clandinin, D. Jean & Connelly, F. Michael (1987). Teachers’ personal knowledge: What counts as ‘personal’ in studies of the personal. Journal of Curriculum Studies, 19(6), 487-500. 9 References 215 <?page no="216"?> Clark, Christopher, & Yinger, Robert (1977). Research on teacher thinking. Curriculum Inquiry, 7(4), 279-304. Clark, Christopher & Peterson, Penelope (1986). Teachers’ thought processes. In Merlin C. Wittrock (Ed.), Handbook of research on teaching (3rd ed.) (pp. 255-296). Macmillan.- Cohen, Andrew D. (1987). Using verbal reports in research on language learning. In Claus Færch & Gabriele Kasper (Eds.), Introspection in second language research (pp. 82-95). Multilingual Matters. Connelly, F. Michael, & Clandinin, D. Jean (1982). Personal practical knowledge at Bay Street School. Ontario Inst. for Studies in Education, (pp. 2-35), Retrieved from: https: / / eric.ed.gov/ ? id=ED22297 8; Copland, Fiona (2018). Observation and fieldnotes. In Aek Phakiti, Peter De Costa, Luke Plonsky & Sue Starfield (Eds.), The Palgrave handbook of applied linguistics research methodology (pp. 249-268). Palgrave. DOI: https: / / doi.org/ 10.1057/ 978-1-137-59900-1_11 Costa, Francesca (2013). “Dealing with the language aspect? Personally, no.” Content lecturers’ views in an ICLHE context. In Stephan Breidbach & Britta Viebrock (Eds.), Content and Language Integrated Learning (CLIL) in Europe: Research perspectives on policy and practice (pp. 117-127). Peter Lang. Council of Europe (2001). Common European Framework of reference for languages: Learning, teaching, assessment, Cambridge University Press. Retrieved from: https: / / rm.coe.int/ 1680459f97 Council of Europe (2020). Common European Framework of Reference for Languages: Learning, teaching, assessment - Companion volume, Council of Europe Publishing. Retrieved from: www.coe.int/ lan g-cefr. Cowie, Neil (2009). Observation. In Juanita Heigham & Robert A. Croker (Eds.), Qualitative research in applied linguistics: A practical introduction (165-181). Palgrave McMillan.- Coyle, Do, Hood, Philip & Marsh, David (2010). CLIL: Content and Language Integrated Learning. Cambridge University Press. Crookes, Graham (2015). Redrawing the boundaries on theory, research, and practice concerning language teachers’ philosophies and language teacher cognition: Toward a critical perspective. The Modern Language Journal, 99(3), 485-499. DOI: https: / / doi.org/ 10.1111/ modl.12237 Cummins, Jim (2008). BICS and CALP: Empirical and theoretical status of the distinction. In Brian Street, & Nancy H. Hornberger (Eds.), Encyclopedia of language and education: Volume 2 literacy (2nd Edition) (pp. 71-83). Springer. DOI: https: / / doi.org/ 10.1007/ 978-0-387-30424-3_36 Dallinger, Sara (2015). Die Wirksamkeit bilingualen Sachfachunterrichts: Selektionseffekte, Leistung‐ sentwicklung und die Rolle der Sprachen im deutsch-englischen Geschichtsunterricht. [Doctoral dissertation, Pädagogischen Hochschule Ludwigsburg]. Retrieved from: https: / / phbl-opus.phlb.d e/ frontdoor/ deliver/ index/ docId/ 67/ file/ Dissertation_Sara+Dallinger.pdf; Dallinger, Sara, Jonkmann, Kathrin, Hollm, Jan, Fiege, Christiane (2016). The effect of content and language integrated learning on students' English and history competences - Killing two birds with one stone? Learning and Instruction, 41, 23-31. DOI: https: / / doi.org/ 10.1016/ j.learninstruc.20 15.09.003 Dalton-Puffer, Christiane (2007).- Discourse- in- content- and- language- integrated- learning (CLIL) classrooms. John Benjamins. Dalton-Puffer, Christiane & Smit, Ute (2013). Content and language integrated learning: A research agenda. Language Teaching, 46(4). 545-559. DOI: https: / / doi.org/ 10.1017/ S0261444813000256 216 9 References <?page no="217"?> Dalton-Puffer, Christiane, Nikula, Tarja & Smit, Ute (2010a). Charting policies, premises and research on content and language integrated learning. In Christiane Dalton-Puffer, Tarja Nikula & Ute Smit (Eds.), Language use and language learning in CLIL classrooms (pp. 1-19). John Benjamins.- Dalton-Puffer, Christiane, Nikula, Tarja & Smit, Ute (2010b). Language use and language learning in CLIL: Current findings and contentious issues. In Christiane Dalton-Puffer, Tarja Nikula & Ute Smit (Eds.), Language use and language learning in CLIL classrooms (pp. 279-291). John Benjamins. de Graaff, Rick, Koopman, Gerrit Jan, Anikina, Yulia & Westhoff, Gerard (2007). An Observation Tool for Effective L2 Pedagogy in Content and Language Integrated Learning (CLIL). The International Journal of Bilingual Education and Bilingualism, 10(5). 603-624. DOI: https: / / doi.org/ 10.2167/ beb4 62.0 de la Peña, Matt (2015). Last stop on Market Street. Penguin Group. Denman, Jenny, Tanner, Rosie & de Graaff, Rick- (2013).- CLIL in junior vocational secondary education: Challenges and opportunities for teaching and learning.- International Journal- of- Bi‐ lingual Education and Bilingualism, 16(3),- 285-300. - DOI: https: / / doi.org/ 10.1080/ 13670050.201 3.777386 Deters-Philipp, Ann-Cathrin (2018). Lehrersprache im Englischunterricht an deutschen Grundschulen: Eine Interviewstudie mit Lehrkräften. Waxmann. Díaz, Carmen, & Requejo, Maria D. (2008). Teacher beliefs in a CLIL education project. Porta Linguarum, 10, 151-161. DOI: 10.30827/ Digibug.31786 Diehr, Bärbel (2012). What’s in a name? Terminologische, typologische und programmatische Überlegungen zum Verhältnis der Sprachen im Bilingualen Unterricht. In Bärbel Diehr & Lars Schmelter (Eds.), Bilingualen Unterricht weiterdenken: Programme, Positionen, Perspektiven (17-36). Peter Lang. Diehr, Bärbel (2016). Doppelte Fachliteralität im bilingualen Unterricht: Theoretische Modelle für Forschung und Praxis. In Bärbel Diehr, Angelika Preisfeld & Lars Schmelter (Eds.), Bilingualen Unterricht weiterentwickeln und erforschen (pp. 57-84). Peter Lang. DIPS (Dortmund International Primary Schools) (2020, Sept. 7). Konzept zum Bilingualen Unterricht (englisch/ deutsch) in der Grundschule. Ostenberg Grundschule. Retrieved from: https: / / ostenberg -grundschule.de/ images/ internationale_Schule/ konzept.pdf DIPS (Dortmund International Primary Schools) (2022, March 3). Konzept zum Bilingualen Lernen (englisch/ deutsch) an den Dortmund International Primary Schools (DIPS) ab dem Schuljahr 2021/ 22. Kreuz Grundschule. Retrieved from: https: / / www.kreuz-grundschule.de/ index.php/ beispielseiten / dips Dirks, Una (2004). ‘Kulturhüter’ oder ‘Weltenwanderer’? Zwei ‘ideale’ Realtypen bilingualen Sach‐ fachunterrichts. In Andreas Bonnet & Stephan Breidbach (Eds.),-Didaktiken im Dialog. Konzepte des Lehrens und Wege des Lernens im bilingualen Sachfachunterricht (pp. 129-140). Peter Lang. Dörnyei, Zoltán (2007). Research methods in applied linguistics. Oxford University Press. Dressing, Thorsten, Pehl, Thorsten & Schmieder, Christian (2015). Manual (on) transcription: Tran‐ scription conventions, software guides and practical hints for qualitative researchers (3 rd English Ed.) Marburg. Retrieved from: https: / / www.audiotranskription.de/ wp-content/ uploads/ 2020/ 11/ m anual-on-transcription.pdf EC (1995). Teaching and Learning: Towards the Learning Society. EC. Retrieved from: https: / / europa.e u/ documents/ comm/ white_papers/ pdf/ com95_590_en.pdf 9 References 217 <?page no="218"?> EC (1997). Early teaching of community languages: Resolution 8. EC. Retrieved from: https: / / ec.europ a.eu/ commission/ presscorner/ detail/ en/ PRES_97_348 Elbaz, Freema (1981). The teacher’s “practical knowledge”: Report of a case study. Curriculum Inquiry, 11(1), 43-71. DOI: https: / / doi.org/ 10.2307/ 1179510 Elsner, Daniela & Keßler, Jörg (2013a). Aspects of immersion, CLIL, and bilingual modules: Bilingual education in primary school. In Daniela Elsner & Jörg Keßler (Eds.), Bilingual education in primary school: Aspects of immersion, CLIL, and bilingual modules (pp. 1-6). Narr Verlag. Elsner, Daniela & Keßler, Jörg (2013b). Bilingual approaches to foreign language education in primary school. In Daniela Elsner & Jörg Keßler (Eds.), Bilingual education in primary school: Aspects of immersion, CLIL, and bilingual modules (pp. 16-27). Narr Verlag. Enever, Janet (2016). Primary ELT: Issues and trends. In Graham Hall (Ed.), The Routledge handbook of English language teaching (353-366). Routledge.- Engh, Dwayne (2013). Why use music in English language learning? A survey of the literature. English Language Teaching, 6(2), 113-127. DOI: 10.5539/ elt.v6n2p113 Ericsson, K. Anders, & Simon, Herbert A. (1993). Protocol Analysis: Verbal reports as data. The MIT Press. DOI: https: / / doi.org/ 10.7551/ mitpress/ 5657.001.0001 Ernest, Paul (1989). The knowledge, beliefs and attitudes of the mathematics teacher: A model. Journal of Education for Teaching, 15(1), 13-33. DOI: 10.1080/ 0260747890150102 Esin, Cigdem, Fathi, Mastoureh, & Squire, Corinne (2013). Narrative analysis: The constructionist approach. In Uwe Flick (Ed.), The Sage handbook of qualitative data analysis (pp. 203-216). Sage. Eurydice (2017). Eurydice brief: Key data on teaching languages at school in Europe. Eurydice. DOI: 10.2797/ 828497 Falout, Joseph & Murphey, Tim (2018). Teachers crafting job crafting. In Achilleas Kostoulas & Sarah Mercer (Eds.), Language teacher psychology (pp. 211-230). Multilingual Matters. Fenstermacher, Gary D. (1978). A philosophical consideration of recent research on teacher effective‐ ness. Review of Research in Education, 6, 157-185. Retrieved from: https: / / www.jstor.org/ stable/ 11 67245 Fernández-Fontecha, Almudena (2015). Motivation and vocabulary breadth in CLIL and EFL contexts: Different age, same time of exposure. Complutense Journal of English Studies, 23, 79-96. DOI: -htt ps: / / doi.org/ 10.5209/ rev_CJES.2015.v23.51214 FMKS. (2016). Bilinguale Kitas und Grundschulen in Deutschland. In Heiner Böttger und Norbert Schlüter (Eds.), Fortschritte im Frühen Fremdsprachenlernen: Tagungsband zur 4. FFF-Konferenz (pp. 190-199). Westerman. Frisch, Stefanie (2013). Lesen im Englischunterricht der Grundschule: Eine Vergleichsstudie zur Wirk‐ samkeit zweier Lehrverfahren. Narr.- Frisch, Stefanie (2016). Sprachwechsel als integraler Bestandteil bilingualen Unterrichts. In Bärbel Diehr, Angelika Preisfeld & Lars Schmelter (Eds.), Bilingualen Unterricht weiterentwickeln und erforschen (pp. 85-102). Peter Lang. Frisch, Stefanie (2021). Bilinguales Lernen in der Grundschule: Einblicke in sprachliche und natur‐ wissenschaftliche Kompetenzen. Fremdsprachen Lehren und Lernen, 50(1), 31-49. Frisch, Stefanie & Diehr, Bärbel (2018). Stellungnahme zum Antrag der Fraktion der AfD „Englischun‐ terricht in der Primarstufe abschaffen - Deutsch und Mathematik dafür stärken! “ Retrieved October 8, 2021 from: https: / / www.landtag.nrw.de/ portal/ WWW/ dokumentenarchiv/ Dokument/ MMST17 -315.pdf 218 9 References <?page no="219"?> Gallardo-del-Puerto, Francisco & Blanco-Suárez, Zeltia. (2021). Foreign language motivation in primary education students: The effects of additional CLIL and gender. Journal of Immersion and Content-Based Language Education, 9(1), 58-84. DOI: https: / / doi.org/ 10.1075/ jicb.19023.gal García, Ofelia (2009). Bilingual education in the 21 st century: A global perspective. Wiley-Blackwell. García, Ofelia, Ibarra Johnson, Susana & Seltzer, Kate (2017). The translanguaging classroom: Lever‐ aging student bilingualism for learning. Carlson. García, Ofelia. & Otheguy, Ricardo (2019). Plurilingualism and translanguaging: Commonalities and divergences. International Journal of Bilingual Education and Bilingualism. 23(1), 17-35. DOI: http s: / / doi.org/ 10.1080/ 13670050.2019.1598932 Gass, Susan M. (1997). Input, interaction, and the second language learner. Lawrence Erlbaum Associates, Inc. Gass, Susan M. & Mackey, Alison (2015). Input, interaction, and output in second language acquisition. In Bill Van Patten & Jessica Williams (Eds.), Theories in second language acquisition (pp. 180-206). Routledge.- Gierlinger, Erwin (2015). ‘You can speak German, sir’: On the complexity of teachers’ L1 use in CLIL. Language and Education, 29(4), 347-368. DOI: 10.1080/ 09500782.2015.1023733 Gierlinger, Erwin (2017). Teaching CLIL? Yes, but with a pinch of SALT. Journal of Immersion and Content-Based Language Education, 5(2), 187-213. DOI: -https: / / doi.org/ 10.1075/ jicb.5.2.02gie Gkonou, Christina & Mercer, Sarah (2018). The relational beliefs and practices of highly socio-emo‐ tionally competent language teachers. In Achilleas Kostoulas & Sarah Mercer (Eds.), Language teacher psychology (pp. 158-177). Multilingual Matters. Glaser, Karen (2020). Instruction-giving in the primary English classroom: Creating or obstructing learning opportunities? In Friedrich Lenz, Maximiliane Frobenius & Revert Klattenberg (Eds.), Classroom observation: Researching interaction in English language teaching (pp. 57-83). Peter Lang. Gläser-Zikuda, Michaela, Hagenauer, Gerda & Stephan, Melanie (2020). The potential of qualitative content analysis for empirical educational research. Forum: Qualitative Social Research, 21(1), 1-21. DOI: https: / / doi.org/ 10.17169/ fqs-21.1.3443 Golombek, Paula (1998). A study of language teachers’ personal practical knowledge. TESOL Quar‐ terly, 32(3), 447-464. DOI: https: / / doi.org/ 10.2307/ 3588117 Groeben, Norbert, Wahl, Diethelm, Schlee, Jörg & Scheele, Brigitte (1988). Forschungsprogramm subjektive Theorien: eine Einführung in die Psychologie des reflexiven Subjekts. Francke Verlag.- Grotjahn, Rüdiger (1987). On the methodological basis of introspective methods. In Claus Færch & Gabriele Kasper (Eds.), Introspection in second language research (pp. 54-81). Multilingual Matters. Gruber, Marie-Theres & Mercer, Sarah (2022). The agency of CLIL primary school teachers in Austria. Journal of Immersion and Content-Based Language Education, 10(1), 90-112. DOI: https: / / doi.org/ 1 0.1075/ jicb.20032.gru Hall, Rogers (2007). Strategies for video recording: Fast, cheap, and (mostly) in control. In Sharon J. Derry (Ed.), Guidelines for video research in education. Retrieved from https: / / drdc.uchicago.edu/ w hat/ video-research-guidelines.pdf. Haß, Frank, Kieweg, Werner, Kuty, Margitta, Müller-Hartmann, Andreas & Weisshaar, Harald (2006). Fachdidaktik Englisch: Tradition, Innovation, Praxis. Klett.- Heim, Katja (2013). CLIL---teaching the art: Physical education, art, music. In Daniela Elsner & Jörg Keßler (Eds.), Bilingual education in primary school: Aspects of immersion, CLIL, and bilingual modules (pp. 61-71). Narr Verlag. 9 References 219 <?page no="220"?> Heim, Katja (2015). Bilingualer Unterricht in der Grundschule. In Bernd Rüschoff, Julian Sudhoff & Dieter Wolff (Eds.), CLIL revisited: Eine kritische Analyse zum gegenwärtigen Stand des bilingualen Sachfachunterrichts (pp. 43-64). Peter Lang.- Heimes, Alexander (2010). Bilinguale Methoden für den mehrsprachigen Sachfachunterricht. Praxis Fremdsprachenunterricht, 7(2), 7-10. Helmke, Tuyet, Helmke, Aandreas, Schrader, Friedrich-Wilhelm., Wagner, Wolfgang, Nold, Günter, & Schröder, Konrad (2008). Die Videostudie des Englischunterrichts. In DESI-Konsortium (Eds.), Unterricht und Kompetenzerwerb in Deutsch und Englisch: Ergebnisse der DESI-Studie (pp. 345-363). Beltz. DOI: 10.25656/ 01: 3521 Holliday, Adrian (2013). Authenticity, communities, and hidden potentials. In Philip Powell-Davies & Paul Gunashekar (Eds.), English language teacher education in a diverse environment (pp. 19-23). British Council.- Holstein, James A. & Gubrium, Jaber F. (1995). The active interview. Sage. Humphries, Simon, & Gebhard, Jerry (2012). Observation. In Roger Barnard & Anne Burns (Eds.), Researching language teacher cognition and practice: International case studies (pp. 109-127). Multilingual Matters. Hüttner, Julia & Dalton-Puffer, Christiane (2013). Der Einfluss subjektiver Sprachlerntheorien auf den Erfolg der Implementierung von CLIL-Programmen. In Stephan Breidbach & Britta Viebrock (Eds.), Content and Language Integrated Learning (CLIL) in Europe: Research perspectives on policy and practice (pp. 129-144). Peter Lang. Hüttner, Julia & Smit, Ute (2014). CLIL: The bigger picture. A response to: A. Bruton. 2013. CLIL: Some of the reasons why…and why not, System, 41 (2013): 587-597. System, 44, 160-167. DOI: htt p: / / dx.doi.org/ 10.1016/ j.system.2014.03.001 Huttunen, Rauno & Kakkori, Leena (2020). Heidegger’s theory of truth and its importance for the quality of qualitative research. Journal of Philosophy of Education, 54(3), 600-616. DOI: https: / / doi .org/ 10.1111/ 1467-9752.12429 Imgrund, Bettina. (2004). Bilingualer Sachfachunterricht auf der Primarstufe - Ein Erfahrungsbericht mit einem Ausbildungsmodul am Seminar Hitzkirch. Beiträge zur Lehrerinnen- und Lehrerbildung, 22(1), 117-126. DOI: https: / / doi.org/ 10.36950/ bzl.22.1.2004.10181 Jackson, Philip (1968). Life in classrooms. Holt, Rinehart and Winston, Inc.- Kahlert, Joachim (2022). Der Sachunterricht und seine Didaktik (5th edition). Klinkhardt. Kalaja, Paula (2016). Students teachers’ beliefs about L1 and L2 discursively constructed: A longitu‐ dinal study of interpretive repertoires. In Paula Kalaja, Ana Maria Barcelos, Mari Aro & Maria Ruohotie-Lyhty (Eds.), Beliefs, agency and identity in foreign language learning and teaching (pp. 97-123). Palgrave Macmillan.- Kalaja, Paula, Barcelos, Ana Maria, Aro, Mari & Ruohotie-Lyhty, Maria (2016). Beliefs, agency and identity in foreign language learning and teaching. Palgrave Macmillan.- Kersten, Kristen & Rohde, Andreas (2015). Immersion teaching in English with young learners. In Janice Bland (Ed.), Teaching English to young learners: Critical issues in language teaching with 3-12 year olds (pp. 71-89). Bloomsbury.- Kindermann, Katharina (2020). Die zusammenfassende Inhaltsanalyse als zentrale Methode bei der Rekonstruktion subjektiver Theorien mittels Struktur-Lege-Verfahren. Forum Qualitative Sozialforschung, 21(1), 1-12. DOI: https: / / doi.org/ 10.17169/ fqs-21.1.3324 220 9 References <?page no="221"?> King, Jim & Ng, Kwan-Yee Sarah (2018). Teacher emotions and the emotional labour of second language teaching. In Achilleas Kostoulas & Sarah Mercer (Eds.), Language teacher psychology (pp. 141-157). Multilingual Matters. Klippel, Friederike (2000). Englisch in der Grundschule: Handbuch für einen kindgemäßen Fremdspra‐ chenunterricht. Cornelsen. KMK. (2013). Konzepte für den bilingualen Unterricht: Erfahrungsbericht und Vorschläge zur Weiterent‐ wicklung. Kultusministerkonferenz. Retrieved from: https: / / www.kmk.org/ fileadmin/ Dateien/ ver oeffentlichungen_beschluesse/ 2013/ 201_10_17-Konzepte-bilingualer-Unterricht.pdf Koch-Priewe, Barbara (1986). Subjektive didaktische Theorien von Lehrern: Tätigkeitstheorie, bildungs‐ theoretische Didaktik und alltägliches Handeln im Unterricht. Haag und Herrchen Verlag. König, Johannes, Krepf, Matthias, Bremerich-Vos, Albert & Buchholtz, Christiane (2021). Meeting cognitive demands of lesson planning: Introducing the CODE-PLAN model to describe and analyze teachers’ planning competence, The Teacher Educator, 56(4), 466-487. DOI: 10.1080/ 08878730.2021.1938324 Kostoulas, Achilleas & Lämmerer, Anita (2018). Making the transition into teacher education: Resilience as a process of growth. In Achilleas Kostoulas & Sarah Mercer (Eds.), Language teacher psychology (pp. 247-263). Multilingual Matters. Krashen, Stephen (1985). The input hypothesis: Issues and implications. Longman. Krashen, Stephen & Terrell, Tracey (1983). The natural approach: Language acquisition in the classroom. Alemany Press.- Kubanyiova, Magdalena & Feryok, Anne (2015). Language teacher cognition in applied linguistic research: Revisiting the territory, redrawing the boundaries, reclaiming the relevance. The Modern Language Journal, 99(3), 435-449. DOI: https: / / doi.org/ 10.1111/ modl.12239 Kuckartz, Udo (2019). Qualitative content analysis: From Kracauer’s beginnings to today’s challenges. Forum: Qualitative Social Research, 20(3), 1-20. DOI: https: / / doi.org/ 10.17169/ fqs-20.3.3370 Küppers, Almut & Trautmann, Matthias (2008). It’s not CLIL that is a success - CLIL students are! Some critical remarks on the current CLIL boom. In Stephan Breidbach & Britta Viebrock (Eds.), Content and language integrated learning (CLIL) in Europe: Research perspectives on policy and practice (pp. 285-296). Peter Lang. Lakoff, George & Johnson, Mark (1980). Metaphors we live by. University of Chicago Press. Lambelet, Amelia & Berthele, Raphael (2015). Age and foreign language learning in school. Palgrave. Lambert, Magdalene (1985). How do teachers manage to teach? Perspectives on problems in practice. Harvard Educational Review, 55(2), 178-194. Lancaster, Nina Karen (2018). Innovations and challenges in CLIL program evaluation. Theory into Practice, 57, 250-257. DOI: https: / / doi.org/ 10.1080/ 00405841.2018.1484034 Lantolf, James, Thorne, Steven & Poehner, Matthew (2015). Sociocultural theory and second language development. In Bill Van Patten & Jessica Williams (Eds.), Theories in second language acquisition (pp. 207-226). Routledge. Larsen-Freeman, Diane (2015). Complexity theory. In Bill Van Patten & Jessica Williams (Eds.), Theories in second language acquisition (pp. 227-244). Routledge. Larsen-Freeman, Diane & Anderson, Marti (2011). Techniques and principles in language teaching. Oxford.- Larsen-Freeman, Diane & Cameron, Lynne (2008). Complex systems in applied linguistics. Oxford University Press. 9 References 221 <?page no="222"?> Lasagabaster, David (2017) “I always speak English in my classes”: Reflections on the use of the L1/ L2 in English medium instruction. In Ana Llinares & Tom Morton (Eds.), Applied linguistics perspectives on CLIL (pp. 251-267). John Benjamins. Legutke, Michael, Müller-Hartmann, Andreas & Schocker-v.Ditfurth, Marita (2014). Teaching English in the primary school. Klett.- Leykum, Simon, Heinze, Thomas & Gropengießer, Harald (2012). Erdwürmer und andere Gründe über bilingualen Biologieunterricht nachzudenken. In Bärbel Diehr & Lars Schmelter (Eds.), Bilingualen Unterricht weiterdenken (pp. 149-162). Peter Lang. Li, Li (2017). Social interaction and teacher cognition. Edinburgh University Press. Li, Li (2020). Language teacher cognition: A sociocultural perspective. Palgrave Macmillan.- Li, Wendy & De Costa, Peter (2018). Exploring novice EFL teachers’ identity development: A case study of two EFL teachers in China. In Achilleas Kostoulas & Sarah Mercer (Eds.), Language teacher psychology (pp. 86-104). Multilingual Matters. Limberg, Holger (2020). Classroom interactional competence. In Friedrich Lenz, Maximiliane Fro‐ benius, & Revert Klattenberg (Eds.), Classroom observation: Researching interaction in English language teaching (pp. 33-56). Peter Lang.- Lincoln, Yvonna S. & Guba, Egon (1982). Epistemological and Methodological Bases of Naturalistic Inquiry. Educational Communications & Technology Journal, 30(4), 233-252. DOI: https: / / doi.org/ 1 0.1007/ BF02765185 Lincoln, Yvonna S. & Guba, Egon (1985). Naturalistic Inquiry. Sage. Lindemann, Marlies & Diehr, Bärbel (2016). Two wor(l)ds? Interkulturelles Lernen an partieller Konzeptäquivalenz im bilingualen Unterricht. In Bärbel Diehr, Angelika Preisfeld & Lars Schmelter (Eds.), Bilingualen Unterricht weiterentwickeln und erforschen (pp. 261-274). Peter Lang. Llinares, Ana (2017). Learning how to mean in primary school CLIL classrooms. In María del Pilar García Mayo (Ed.), Learning foreign languages in primary school: Research insights (pp. 69-85). Multilingual Matters. DOI: 10.21832/ GARCIA8101 Long, Michael (1996). The role of linguistic environment in second language acquisition. In William C. Ritchie & Tej K. Bhatia (Eds.), Handbook of second language acquisition (pp. 413-468). Academic Press.- Lortie, Dan C. (1975). Schoolteacher: A sociological study. University of Chicago Press. Lund, Nick (2003). Language and thought. Routledge. Lyster, Roy (2011). Content-based second language teaching. In Eli Hinkel (Ed.), Handbook of research in second language teaching (Vol. 2) (pp. 612-630). Routledge. Lyster, Roy (2018). Content-based language teaching. Routledge.- MacDonald, Malcom, Badger, Richard & Dasli, Maria (2006). Authenticity, culture and language learning. Language and Intercultural Communication, 6(3&4), 250-261. DOI: 10.2167/ laic252.0 MacIntyre, Peter, Gregersen, Tammy & Clément, Richard (2016). Individual differences. In Graham Hall (Ed.), The Routledge handbook of English language teaching (pp. 310- 323). Routledge. Mandl, Heinz & Huber, Günter (1983). Subjektive Theorien von Lehrern. Psychologie in Erziehung und Unterricht, 30, 98-112. Marsh, David (2000, September 15). Using languages to learn and learning to use languages. TIECLIL. Retrieved from: http: / / www.tieclil.org/ html/ products/ pdf/ %201%20UK.pdf Marsh, David, Mehisto, Peeter, Wolff, Dieter & Frigols Martín, María Jesús (2011). European framework for CLIL teacher education: A framework for the professional development of CLIL teachers. European 222 9 References <?page no="223"?> Center for Modern Languages. Retrieved from: https: / / www.ecml.at/ Resources/ ECMLresources/ t abid/ 277/ ID/ 35/ language/ en-GB/ Default.aspx. Marvasti, Amir B. (2019). Qualitative content analysis: A novice's perspective. Forum: Qualitative Social Research, 20(3), 1-15. DOI: http: / / orcid.org/ 0000-0002-6026-357X Massler, Ute (2012). Primary CLIL and its stakeholders: What children, parents and teachers think of the potential merits and pitfalls of CLIL modules in primary teaching. International CLIL Research Journal, 1(4), 36-46. Retrieved from: http: / / www.icrj.eu/ 14/ article4.html Massler, Ute (2013). CLIL-Unterricht in der Grundschule. In Ute Massler & Daniel Stotz (Eds.), CLIL-Unterricht in der Primarstufe: Ein theoriebasierter Leitfaden für die Entwicklung von Aufgaben für Unterricht und Beurteilung (pp. 8-20). Wissenschaftlicher Verlag Trier.- Massler, Ute & Ioannou-Georgiou, Sophie (2010). Best practice: How CLIL works. In Ute Massler & Petra Burmeister (Eds.), CLIL und Immersion: Fremdsprachlicher Sachfachunterricht in der Grundschule (pp. 61-75). Westermann. Massler, Ute & Stotz, Daniel (Eds.). (2013). CLIL-Unterricht in der Primarstufe: Ein theoriebasierter Leitfaden für die Entwicklung von Aufgaben für Unterricht und Beurteilung. Wissenschaftlicher Verlag Trier. Maxwell, Joseph A., & Chmiel, Margaret (2013). Generalization in and from qualitative analysis. In Uwe Flick (Ed.), The Sage handbook of qualitative data analysis (pp. 540-553). Sage. Mayring, Philipp (2015). Qualitative Inhaltsanalyse: Grundlagen und Techniken (12th ed.). Beltz. McAteer, Mary (2013). Action research in education. Sage. McDougald, Jermaine (2015). Teachers’ attitudes, perceptions and experiences in CLIL: A look at content and language. Colombian Applied Linguistics Journal, 17(1), 25-41. DOI: 10.14483/ udis‐ trital.jour.calj.2015.1.a02 Mehisto, Peeter (2013). Integrating CLIL with other mainstream discourses. In Stephan Breidbach & Britta Viebrock (Eds.), Content and Language Integrated Learning (CLIL) in Europe: Research perspectives on policy and practice (pp. 25-50). Peter Lang. Mehisto, Peeter, Marsh, David & Frigols, María Jesús (2008). Uncovering CLIL: Content and language integrated learning in bilingual and multilingual education. Macmillan Education. Mercer, Sarah & Kostoulas, Achilleas (2018). Introduction to language teacher psychology. In Sarah Mercer & Achilleas Kostoulas (Eds.), Language teacher psychology (pp. 1-17). Multilingual Matters. Meyer, Oliver (2010). Towards quality-CLIL: Successful planning and teaching strategies. Pulso, 33, 11-29. DOI: https: / / doi.org/ 10.58265/ pulso.5002 Millington, Neil T. (2011). Using songs effectively to teach English to young learners. Language Education in Asia, 2(1), 134-141. Retrieved from: https: / / leia.org/ LEiA/ LEiA%20VOLUMES/ Downl oad/ LEiA_V2_I1_2011/ LEiA_V2I1A11_Millington.pdf Mindt, Dieter (2013). Grundschulenglisch abschaffen? Die Grundschule braucht endlich richtiges Englisch. Schulverwaltung. Nordrhein-Westfalen, 24(10), 270-272. Mirhosseini, Seyyed-Abdolhamid (2020). Collecting Interview Data. In Seyyed-Abdolhamid Mirhos‐ seini (Aut), Doing Qualitative Research in Language Education (pp. 85-109). Palgrave Macmillan. DOI: https: / / doi.org/ 10.1007/ 978-3-030-56492-6_5 Morton, Tom (2012). Teachers’ knowledge about language and classroom interaction in content and language integrated learning [Doctoral dissertation Universidad Autónoma de Madrid, Madrid]. Retrieved from: https: / / repositorio.uam.es/ handle/ 10486/ 660387 9 References 223 <?page no="224"?> MSB NRW. (2020). Masterplan Grundschule: Qualität stärken, Lehrkräfte unterstützen. Ministerium für Schule und Weiterbildung des Landes Nordrhein-Westfalen. Retrieved from: https: / / www.schulm inisterium.nrw.de/ system/ files/ media/ document/ file/ Masterplan%20Grundschule.pdf MSB NRW. (2021). Lehrplan für die Primarstufe in Nordrhein-Westfalen. Retrieved from: https: / / www .schulentwicklung.nrw.de/ lehrplaene/ lehrplan/ 284/ ps_lp_e_einzeldatei_2021_08_02.pdf MSB NRW. (2021, March 4). Bilingualer Unterricht. Retrieved from: https: / / www.schulentwicklung.n rw.de/ cms/ bilingualer-unterricht/ angebot-home/ bilingualer-unterricht.html MSB NRW. (2022, January 29). Bilingualer Unterricht in Nordrhein-Westfalen. Ministerium für Schule und Bildung des Landes Nordrhein-Westfalen.- Retrieved from: https: / / www.schulministerium.n rw.de/ themen/ schulsystem/ unterricht/ lernbereiche-und-unterrichtsfaecher/ bilingualer-unterrich t-nordrhein MSB NRW. (2022, March 3). Bilingualer Unterricht in Nordrhein-Westfalen. Ministerium für Schule und Bildung des Landes Nordrhein-Westfalen.-Retrieved from: https: / / www.schulministerium.nr w/ bilingualer-unterricht-nordrhein-westfalen MSW NRW. (2008). Richtlinien und Lehrpläne für die Grundschule in Nordrhein-Westfalen. Ministerium für Schule und Weiterbildung des Landes Nordrhein-Westfalen. Retrieved from: https: / / www.sch ulentwicklung.nrw.de/ lehrplaene/ upload/ klp_gs/ LP_GS_2008.pdf. MSW NRW. (2012). Bilingualer Unterricht: Biologie deutsch-englisch in der Sekundarstufe 1. Minis‐ terium für Schule und Weiterbildung des Landes Nordrhein-Westfalen.- Retrieved from: https: / / www.schulentwicklung.nrw.de/ cms/ upload/ bilingualer_Unterricht/ documents/ HR_BU_BioE_Se kI_0912.pdf MSW NRW. (2015, June 11). Bilinguale Grundschulen in NRW. Qualitäts- und Unterstützungsagentur: Landesinstitut für Schule des Landes Nordrhein-Westfalen. Retrieved from: https: / / www.schulen twicklung.nrw.de/ cms/ angebote/ egs/ bilingualer-unterricht/ bilinguale-grundschulen-in-nrw/ inde x.html MSW NRW. (2020, October 3). Englisch in der Grundschule. Qualitäts- und Unterstützungsagentur: Landesinstitut für Schule des Landes Nordrhein-Westfalen. Retrieved from: https: / / www.schulen twicklung.nrw.de/ cms/ angebote/ egs/ startseite/ englisch-in-der-grundschule.html. Müller, Natascha (2017). Code-Switching. Narr. Müller-Hartmann, Andreas & Schocker-v.Ditfurth, Marita (2004). Introduction to English language teaching. Klett.- Müller-Hartmann, Andreas & Schocker-v.Ditfurth, Marita (2011). Teaching English: Task.supported language learning. Verlag Ferdinand Schöningh. Nespor, Jan (1987). The role of beliefs in the practice of teaching. Journal of Curriculum Studies,-19(4),-317-328.-DOI: -10.1080/ 0022027870190403 Nikula, Tarja (2017, July 5). Language and disciplinary learning combined: CLIL challenging conceptions of language skills [Video]. University of Oxford. Retrieved from: https: / / podcasts.ox.ac.uk/ langua ge-and-disciplinary-learning-combined-clil-challenging-conceptions-language-skills Nikula, Tarja, Dalton-Puffer, Christiane & Llinares, Ana (2013). CLIL classroom discourse: Research from Europe. Journal of Immersion and Content-Based Language Education 1(1), 70-100. DOI: http s: / / doi.org/ 10.1075/ jicb.1.1.04nik Nisbett, Richard & Ross, Lee (1980). Human inference: Strategies and shortcomings of social judgment. Prentice Hall.- 224 9 References <?page no="225"?> Nold, Günter, Hartig, Johannes, Hinz, Silke & Rossa, Henning (2008). Klassen mit bilingualem Sachfachunterricht: Englisch als Arbeitssprache. In DESI-Konsortium (Eds.), Unterricht und Kom‐ petenzerwerb in Deutsch und Englisch: Ergebnisse der DESI-Studie (pp. 451-457). Beltz. Nunan, David (2004). Task-based language teaching. Cambridge University Press.- Oattes, Huub, Oostdam, Ron, de Graff, Rick & Wilschut, Arie (2018). The challenge of balancing content and language: Perceptions of Dutch bilingual education history teachers. Teaching and Teacher Education, 70, 165-174. DOI: http: / / doi.org/ 10.1016/ j.tate.2017.11.022 Ohlberger, Stephanie & Wegner, Claas (2018). Bilingualer Sachfachunterricht in Deutschland und Europa: Darstellung des Forschungsstands. Herausforderung Lehrer*innenbildung Zeitschrift, 1(1), 45-89. DOI: https: / / doi.org/ 10.4119/ hlz-2390 Ohta, Amy S. (2000). Rethinking interaction in SLA: Developmentally appropriate assistance in the zone of proximal development and the acquisition of L2 grammar. In James Lantolf (Ed.), Sociocultural theory and second language learning (pp. 51-78). Oxford University Press Oxford Owl. (2020, October 7). Oxford Owl: Help your children learn.- Retrieved from: https: / / home.o xfordowl.co.uk/ reading/ Oxford, Rebecca (2020). The well of language teachers’ emotional well-being. In Chistina Gkonou, Jean-Marc Dewaele & Jim King (Eds.), The emotional rollercoaster of language teaching (pp. 247-268). Multilingual Matters. Pajares, M. Frank (1992). Teachers’ beliefs and educational research: Cleaning up a messy construct. Review of Educational Research, 62(3), 307-332. DOI: https: / / doi.org/ 10.3102/ 00346543062003307 Parkin, Alan J. (2000). Essential cognitive psychology. Psychology Press.- Pavlenko, Aneta (2007). Autobiographic narratives as data in applied linguistics. Applied Linguistics, 28(2), 163-188. DOI: -https: / / doi.org/ 10.1093/ applin/ amm008 Pérez Cañado, María Luisa (2016). From the CLIL craze to the CLIL conundrum: Addressing the current CLIL controversy. Bellaterra Journal of Teaching & Learning Language & Literature, 9(1), 9-31. DOI: https: / / doi.org/ 10.5565/ rev/ jtl3.667 Pérez Cañado, María Luisa (2018). CLIL and educational level: A longitudinal study on the impact of CLIL on language outcomes. Porta Linguarum, 29, 51-70. DOI: 10.30827/ Digibug.54022 Pérez Cañado, María Luisa (2020). What’s hot and what’s not on the current CLIL research agenda: Weeding out the nonissues from the real issues. A response to Bruton (2019). Applied Linguistics Review, 1-21. DOI: https: / / doi.org/ 10.1515/ applirev-2020-0033 Pfitzer, Petra (2006). Kindergemäßes Fremdsprachenlernen: Zur Theorie und Praxis des Frühbeginns Englisch. Peter Lang.- Pinker, Steven (2013). Language, cognition, and human nature. Oxford University Press.- Pinter, Annamaria (2017). Teaching young language learners (2nd edition). Oxford University Press. Piske, Thorsten (2013). Bilingual education: Chances and challenges. In Elsner, Daniela & Keßler, Jörg (Eds.), Bilingual education in primary school (pp. 28-40). Narr. Piske, Thorsten (2015). Zum Erwerb der CLIL-Fremdsprache. In Bernd Rüschoff, Julian Sudhoff & Dieter Wolff (Eds.), CLIL revisited: Eine kritische Analyse zum gegenwärtigen Stand des bilingualen Sachfachunterrichts (pp. 101-126). Peter Lang.- Piske, Thorsten (2017). The earlier, the better? Some critical remarks on current EFL teaching to young learners and their implications for foreign language teacher education. In Eva Wilden & Raphaela Porsch (Eds.), The professional development of primary EFL teacher: National and international research (pp. 45-57). Waxmann.- 9 References 225 <?page no="226"?> Polanyi, Michael (1966). The tacit dimension. The University of Chicago Press. Prior, Matthew T. (2018). Interviews and focus groups. In Aek Phakiti, Peter De Costa, Luke Plonsky & Sue Starfield (Eds.), The Palgrave handbook of applied linguistics research methodology (pp. 225-248). Palgrave. DOI: https: / / doi.org/ 10.1057/ 978-1-137-59900-1_11 Raez-Padilla, Juan (2018). Parent perspectives on CLIL implementation: Which variables make a difference? - Porta Linguarum, (29), 181-196. DOI: 10.30827/ digibug.54033 Raunser, Christine & Steffens, Sarah-Maria (2012). Die Hälfte auf Englisch! Eine Erfolgsstory bilingualer Klassen. Grundschulmagazin Englisch, (5), 37-38.- Raunser, Christine (2012, March 5). DIPS - Konzept, Organisation und Schulalltag [slide presentation]. Retrieved from: https: / / slideum.com/ doc/ 1169242/ dips-%E2%80%93-konzept--organisation-und-s chulalltag---bilingual Reckermann, Julia (2018). Reading authentic English picture books in the primary school EFL classroom: A study of reading comprehension, reading strategies and FL development. Peter Lang.- Richards, Keith (2009). Interviews. In Juanita Heigham & Robert A. Croker (Eds.) Qualitative research in applied linguistics: A practical introduction (pp. 182-199). Palgrave McMillan. Richards, Jack & Rodgers, Theodore (2001). Approaches and methods in language teaching. Cambridge University Press. Rokeach, Milton (1968). Beliefs, attitudes, and values: A theory of organization and change. Jossey-Bass. Roller, Margaret R. (2019). A quality approach to qualitative content analysis: Similarities and differences compared to other qualitative methods. Forum: Qualitative Sozialforschung, 20(3), Art. 31. DOI: https: / / doi.org/ 10.17169/ fqs-20.3.3385 Rossa, Henning (2017). You teach what you believe in: BELT---Beliefs about effective language teaching. In Eva Wilden & Raphaela Porsch (Eds.), The professional development of primary EFL teachers (pp. 197-208). Waxmann.- Rossa, Henning (2020). Wir müssen nur wollen: Die emotionale Dimension fremdsprachlicher Lehr-Lernprozesse. In Eva Burwitz-Melzer, Claudia Riemer & Lars Schmelter (Eds.), Affektiv-emo‐ tionale Dimensionen beim Lehren und Lernen von Fremd- und Zweitsprachen: Arbeitspapiere der 40. Frühjahrskonferenz zur Erforschung des Fremdsprachenunterrichts (pp. 172-184). Narr. Rowell, Lonnie L. (2019). Rigor in educational action research and the construction of knowledge democracies. In Craig A. Mertler (Ed.), The Wiley handbook of action research in education (pp. 117-138). Wiley. Rumlich, Dominik (2015). Zur affektiv-motivationalen Entwicklung von Lernenden. In Bernd Rüschoff, Julian Sudhoff & Dieter Wolff (Eds.), CLIL revisited: Eine kritische Analyse zum gegen‐ wärtigen Stand des bilingualen Sachfachunterrichts (pp. 309-330). Peter Lang.- Rumlich, Dominik (2016). Evaluating bilingual education in Germany: CLIL students’ general English proficiency, EFL self-concept and interest. Peter Lang. Ruohotie-Lyhty, Maria (2016). Stories of change and continuity: Understanding the development of the identities of foreign language teachers. In Paula Kalaja, Ana Maria Barcelos, Mari Aro, & Maria Ruohotie-Lyhty (Eds.), Beliefs, agency and identity in foreign language learning and teaching (pp. 172-201). Palgrave Macmillan.- Ryle, Gilbert (2009). The concept of mind (first published in 1949). Routledge.- Sahakyan, Taguhi, Lamb, Martin & Chambers, Gary (2018). Language teacher motivation: From the ideal to the feasible self. In Achilleas Kostoulas & Sarah Mercer (Eds.), Language teacher psychology (pp. 53-70). Multilingual Matters. 226 9 References <?page no="227"?> Sandberg, Ylva (2019). Teaching and learning content through two languages: The biology and history teacher perspective. In Liss Kerstin Sylvén (Ed.), Investigating content and language integrated learning: Insights from Swedish high schools (pp. 298-314). Multilingual Matters. DOI: 10.21832/ 9781788922425 Sardo Brown, Deborah (1988). Twelve middle-school teachers’ planning. The Elementary School Journal, 89(1), 69-87.-DOI: https: / / doi.org/ 10.1086/ 461563 Schauer, Gila (2019). Teaching and learning English in the primary school: Interlanguage pragmatics in the EFL context. Springer. Schmidt, Richard W. (1990). The role of consciousness in second language learning. Applied Linguistics, 11, 129-158. DOI: https: / / doi.org/ 10.1093/ applin/ 11.2.129 Schreier, Margrit (2012). Qualitative content analysis in practice. Sage. Schreier, Margrit, Stamann, Christoph, Janssen, Markus, Dahl, Thomas & Whittal, Amanda (2019). Qualitative content analysis: Conceptualizations and challenges in research practice - Introduction to the FQS special issue. Forum: Qualitative Social Research, 20(3), 1-27. DOI: https: / / doi.org/ 10.17 169/ fqs-20.3.3393 Schwab, Götz (2020). Conversation analysis gets mobile: Student participation in a bilingual primary classroom in Germany. In Friedrich Lenz, Maximiliane Frobenius & Revert Klattenberg (Eds.), Classroom observation: Researching interaction in English language teaching (pp. 85-114). Peter Lang. Science Now. (2019). Teachers 4 teachers. Retrieved March 2, 2021, from: https: / / www.teachers4teach ers.com.au/ science-now/ Seale, Clive (1999). The quality of qualitative research. Sage.- Shannon, David (1998). No, David! The Blue Sky Press.- Shulman, Lee S. (1986). Those who understand: Knowledge growth in teaching. Educational Re‐ searcher, 15(2), 4-14. Retrieved from: https: / / www.jstor.org/ stable/ 1175860 Siepmann, Philipp, Rumlich, Dominik, Matz, Frauke & Römhild, Ricardo (2021). Attention to diversity in German CLIL classrooms: Multi-perspective research on students’ and teach‐ ers’ perceptions. International Journal of Bilingual Education and Bilingualism. 1-17 DOI: 10.1080/ 13670050.2021.1981821 Skinnari, Kristiina & Bovellan, Eveliina (2016). CLIL teachers’ beliefs about integration and about their professional roles: Perspectives from a European context. In Tarja Nikula, Emma Dafouz, Pat Moore & Ute Smit (Eds.), Conceptualising Integration in CLIL and Multilingual Education (pp. 145-167). Multilingual Matters. Smith, Michael Sharwood (1991). Speaking to many minds: On the relevance of different types of language information for the L2 learner. Second Language Research, 7(2), 118-132. DOI: https: / / do i.org/ 10.1177/ 026765839100700204 Smith, Michael Sharwood (1993). Input enhancement in instructed SLA: Theoretical bases. Studies in Second Language Acquisition. 15, 165-79. DOI: https: / / doi.org/ 10.1017/ S0272263100011943 Smith, Michael Sharwood (2017). Introducing language and cognition: A map of the mind. Cambridge University Press. Surmont, Jill, Struys, Esli, van den Noort, Maurits & van de Craen, Piet- (2016).- The effects of CLIL on mathematical content learning: A longitudinal study. Studies in Second Language Learning and Teaching, 6(2), 319-337.- DOI: https: / / doi.org/ 10.14746/ ssllt.2016.6.2.7 9 References 227 <?page no="228"?> Swain, Merrill (1995). Three functions of output in second language learning. In Guy Cook & Barbara Seidlhofer (Eds.), Principle and practice in applied linguistics: Studies in honour of H.G. Widdowson (pp. 125-144). Oxford University Press.- Swain, Merrill (1998). Focus on form through conscious reflection. In Catherine Doughty & Jessica Williams (Eds.), Focus on form in classroom second language acquisition (pp. 64-81). Cambridge University Press.- Swain, Merrill (2000). The output hypothesis and beyond: Mediating acquisition through collaborative dialogue. In James Lantolf (Ed.), Sociocultural theory and second language learning (pp. 97-114). Oxford University Press.- Swain, Merrill & Johnson, Robert (1997). Immersion education: A category within bilingual education. In Robert Johnson & Merrill Swain (Eds.), Immersion Education: International Perspectives (pp. 1-16). Cambridge University Press. Swain, Merrill & Lapkin, Sharon (2013). A Vygotskian sociocultural perspective on immersion education: The L1/ L2 debate. Journal of Immersion and Content-Based Education, 1(1), 101-129. DOI: https: / / doi.org/ 10.1075/ jicb.1.1.05swa Sylven, Liss Kerstin & Thompson, Amy S. (2015). Language learning motivation and CLIL: Is there a connection? Journal of Immersion and Content-Based Language Education, 3(1), 28-50. DOI: https: / / doi.org/ 10.1075/ jicb.3.1.02syl Tedick, Diane J. & Cammarata, Laurent (2012). Content and language integration in K-12 contexts: Student outcomes, teacher practices, and stakeholder perspectives. Foreign Language Annals, 45(S1), S28-S53.-DOI: https: / / doi.org/ 10.1111/ j.1944-9720.2012.01178.x Tedick, Diane J., Christian, Donna & Fortune, Tara Williams (2011). The future of immersion education: An invitation to ‘dwell in possibility.’ In Diane J. Tedick, D. Christian, & Tara Williams Fortune (Eds.), Immersion education: Practices, policies, possibilities (pp. 1-10). Multilingual Matters. DOI: https: / / doi.org/ 10.21832/ 9781847694041-004 Thaler, Engelbert (2012). Englisch unterrichten. Cornelsen. The New School. (2018, July 23). Diane Larsen-Freeman at the New School [Video]. Youtube. Retrieved from: https: / / www.youtube.com/ watch? v=KXWXJ7znovo Thompson, Amy S. & Sylvén, Liss Kerstin (2019). CLIL and motivation revisited: A longitudinal perspective. In Liss Kerstin Sylvén (Ed.), Investigating content and language integrated learning: Insights from Swedish high schools (pp. 76-95). Multilingual Matters. DOI: https: / / doi.org/ 10.21832 / 9781788922425 Thornberg, Robert & Charmaz, Kathy (2013). Grounded theory and theoretical coding. In Uwe Flick (Ed.), The Sage handbook of qualitative data analysis (pp. 153-169). Sage. Thurn, Susanne (2011). Gemeinsam im Kollegium etwas Neues wagen. In Susanne Thurn (Ed.), Individualisierung ernst genommen: Englisch lernen in jahrgangsübergreifenden Gruppen (3/ 4/ 5) (pp. 13-34). Klinkhardt.- Tomlinson, Carol Ann (2005). Grading and differentiation: Paradox or good practice? Theory into Practice, 44(3), 262-269. Tsui, Amy B.M. (2003). Understanding expertise in teaching: Case studies of ESL teachers. Cambridge University Press. van Kampen, Evelyn, Meirink, Jacobiene, Admiraal, Wilfried & Berry, Amanda (2020). Do we all share the same goals for content and language integrated learning (CLIL)? Specialist and practitioner 228 9 References <?page no="229"?> perceptions of ‘ideal’ CLIL pedagogies in the Netherlands.-International Journal of Bilingual Education and Bilingualism,-23(8),-855-871.-DOI: -https: / / doi.org/ 10.1080/ 13670050.2017.1411332 van Lier, Leo (1996). Interaction in the language curriculum: Awareness, autonomy, and authenticity. Longman.- van Lier, Leo (2000). From input to affordance: Social-interactive learning from an ecological perspective. In James Lantolf (Ed.), Sociocultural theory and second language learning (pp. 245-259). Oxford University Press.- Verriere, Katharina (2016). Wie können fremdsprachliche Sprechanlässe differenziert und individ‐ ualisiert in den Unterricht integriert werden? Erkenntnisse aus einer Befragung und unter‐ richtspraktische Ideen. In Sabine Doff (Ed.), Heterogenität im Fremdsprachenunterricht: Impulse, Rahmenbedingungen, Kernfragen, Perspektiven (pp. 107-120). Narr. Viebrock, Britta (2007). Bilingualer Erdkundeunterricht: subjektive didaktische Theorien von Lehrer‐ innen und Lehrern. Peter Lang.- Viebrock, Britta (2013). Mathematics. In Daniela Elsner & Jörg Keßler (Eds.), Bilingual education in primary school: Aspects of immersion, CLIL, and bilingual modules (pp. 51-60). Narr Verlag.- Vollmer, Helmut (2010). Förderung des Spracherwerbs im bilingualen Sachfachunterricht. In Gerhard Bach & Susanne Niemeier (Eds.), Bilingualer Unterricht: Grundlagen, Methoden, Praxis, Perspektiven (pp. 131-150). Peter Lang.- Vollmuth, Isabel (2004). English an der Grundschule: Wie Handreichungen den Frühbeginn sehen. Eine didaktisch-methodische Analyse. Universitätsverlag Winter. Waas, Ludwig & Hamm, Wolfgang (2004). Englischunterricht in der Grundschule konkret. Auer Verlag. Wagner, Rudi (2016). Unterricht aus Sicht der Lehrerinnen und Lehrer: subjektive Theorien zur Unter‐ richtsgestaltung und ihre Veränderung durch ein Training zu neuen Unterrichtsmethoden. Klinkhardt Wegner, Anke (2012). Seeing the bigger picture. What students and teachers think about CLIL. International CLIL Research Journal, 1(4), 29-35.- Werlen, Erika (2006). Kontexte und Ziele Bilingualen Lehrens und Lernens---Grundzüge einer Didaktik des Bilingualen Lehrens und Lernens. In Gérald Schlemminger (Ed.), Aspekte Bilingualen Lehrens und Lernens (pp. 199-219). Schneider Verlag Hohengehren. Willis, David & Willis, Jane (2007). Doing task-based teaching. Oxford University Press.- Wing-mui So, Winnie (1997). A study of teacher cognition in planning elementary science lessons. Research in Science Education, 27(1), 71-86. DOI: https: / / doi.org/ 10.1007/ BF02463033 Wode, Henning (2009). Frühes Fremdsprachenlernen in bilingualen Kindergärten und Grundschulen. Westermann. Wolff, Dieter (2009). Content and language integrated learning. In Karlfried Knapp & Barbara Seidlhofer (Eds.), Handbook of foreign language communication and learning (pp. 545-572). Mouton de Gruyter. Wolff, Dieter (2016). Der bilingual Sachfachunterricht historisch betrachtet: Ist er der Königsweg zu Mehrsprachigkeit und zum interkulturellen Verstehen? In Bärbel Diehr, Angelika Preisfeld & Lars Schmelter (Eds.), Bilingualen Unterricht weiterentwickeln und erforschen (pp. 19-36). Peter Lang. Wolff, Dieter (2020, September 7). Der bilinguale Ansatz. Qualitäts- und Unterstützungsagentur: Landesinstitut für Schule des Landes Nordrhein-Westfalen. Retrieved from: https: / / www.schule ntwicklung.nrw.de/ cms/ angebote/ egs/ bilingualer-unterricht/ der-bilinguale-ansatz/ der-bilinguale -ansatz.html 9 References 229 <?page no="230"?> Wolff, Dieter & Sudhoff, Julian (2015). Zur Definition des bilingualen Lehrens und Lernens. In Bernd Rüschoff, Julian Sudhoff & Dieter Wolff (Eds.), CLIL revisited: Eine kritische Analyse zum gegenwärtigen Stand des bilingualen Sachfachunterrichts (pp. 9-39). Peter Lang.- Woods, Devon (1996). Teacher cognition in language teaching: Beliefs, decision-making and classroom practice. Cambridge University Press. Woods, Devon & Ҫakir, Hamide (2011). Two dimensions of teacher knowledge: The case of commu‐ nicative language teaching. System, 39, 381-390. DOI: https: / / doi.org/ 10.1016/ j.system.2011.07.010 Wright, Andrew (1989). Pictures for language learning. Cambridge University Press. Yinger, Robert (1980). A study of teacher planning. The Elementary School Journal, 80(3), 107-127. DOI: https: / / doi.org/ 10.1086/ 461181 Zaunbauer, Anna C.M., Bonerad, Eva-Marie & Möller, Jens (2005). Muttersprachliches Lesever‐ ständnis immersiv unterrichteter Kinder. Zeitschrift für Pädagogische Psychologie, 19(3), 233-235. DOI: https: / / doi.org/ 10.1024/ 1010-0652.19.4.263 Zydatiß, Wolfgang (2000). Bilingualer Unterricht in der Grundschule: Entwurf eines Spracherwerbskon‐ zepts für zweisprachige Immersionsprogramme. Max Hueber Verlag. Zydatiß, Wolfgang (2002). Konzeptuelle Grundlagen einer eigenständigen Didaktik des bilingualen Sachfachunterrichts: Forschungsstand und Forschungsprogramm. In Stephan Breidbach, Gerhard Bach, & Dieter Wolff (Eds.), Bilingualer Sachfachunterricht: Didaktik, Lehrer-/ Lernerforschung und Bildungspolitik zwischen Theorie und Empirie (pp. 31-61). Peter Lang. Zydatiß, Wolfgang (2012). Zur doppelten Sprachschwelle im bilingualen Unterricht. In Horst Bayr‐ huber, Ute Harms, Bernhard Muszynski, Bernd Ralle, Martin Rothgangel, Lutz-Helmut Schön, Helmut Johannes Vollmer & Hans-Georg Weigand (Eds.), Formate fachdidaktischer Forschung, Empirische Projekte---historische Analysen---theoretische Grundlegungen (pp. 67-83). Waxmann. 230 9 References <?page no="231"?> 10 Appendices 10.1 Appendix 1: Transcription conventions (? ) - unsure of the word (inaudible) - recording was unclear / - false start or interrupted sentence / / overlap / / - overlapping speech emphasis - emphasized ↑ - increasing inflection “quote” - quoting direct speech (laughs) - teacher laughs (…) - long pause […] - text has been removed <?page no="232"?> 10.2 Appendix 2: Biographical information questionnaire Name: Geschlecht: männlich ⃞ weiblich ⃞ Alter: Staatsangehörigkeit: Muttersprache: Fremdsprachen: Welche Fächer haben Sie an der Universität studiert? Wie viele Jahre haben Sie als Lehrer*in gearbeitet? Wie viele Jahre haben Sie bilingual unterrichtet? Behalten Sie zusätzliche bilingualer Qualifikationen bzw. Zertifikate? 232 10 Appendices <?page no="233"?> 10.3 Appendix 3: Background interview guide Topics and sample questions 1. Background of bilingual program • Describe your school and its bilingual program. How does it work? 2. Bilingual program evaluation/ reflection • What does good primary school bilingual teaching look like in your view? What kind of topics are good for bilingual education? Why? 3. Curriculum and bilingual topics • What is the value of bilingual education in primary school? 4. Students • What kinds of students benefit from bilingual education? Why? 5. Goals • What do you want students to learn in your bilingual lessons? Why? 6. Planning • What is important to you in your planning of bilingual lessons? Why? 7. Activities • What kinds of activities are good for bilingual lessons? Why? 8. Materials • What do you look for in materials that you use in class? Why? 9. Assessment • How do you assess students in the bilingual class? Why? 10.3 Appendix 3: Background interview guide 233 <?page no="234"?> 10.4 Appendix 4: Follow-up interview guide Topics 1. Unit reflection 2. Foreign language learning in observed lessons 3. Observed lesson structure and social forms 4. Observed activities 5. Observed material use 6. Observed teaching of subject vocabulary 7. Role of extra-curricular activities (e.g. birthdays, school events, etc.) on English learning 8. Teacher-student interaction 9. Role of German speaking teacher 10. Additional comments Additional topics discussed with some of the teachers 11. Role of explicit English teaching 12. Role of choice 13. Role of challenge and encouragement 14. Role of positive emotions in foreign language learning 15. Role of mistakes and uncertainty 234 10 Appendices <?page no="235"?> 10.5 Appendix 5: Tables of absolute numbers and percentage agreement for each of the four subcategories coded by two coders. Cognitions about the role of content, L1, L2 Coder 1 (researcher) Coder 2 Percentage agreement Content learning as priority 33 33 100% German subject language as necessary 11 10 91% German and English use for different purposes 16 14 88% English learning as enrichment/ choice 15 15 100% English use as barrier to relationships 3 3 100% English use as a classroom lingua franca 2 2 100% Combined 80 77 96-% agreement General pedagogic goals Coder 1 (researcher) Coder 2 Percent agreement Reducing fear and developing confidence 13 13 100% Success through effort 8 6 75% Motivated language learners 9 8 89% More respectful 4 4 100% Combined 34 31 91-% agreement Scaffolding Coder 1 (researcher) Coder 2 Percentage agreement Scaffolding input German teacher translation 17 16 94% Student translation 13 13 100% Visualization 24 23 96% Simplify language 11 8 73% Repeat and reuse language 8 8 100% Explicit comprehension checks 1 1 100% Use dictionaries 6 5 83%% Create structured and differentiated worksheets 5 5 100% Interactive strategies 8 5 63% 10.5 Appendix 5: Tables of absolute numbers and percentage agreement 235 <?page no="236"?> Scaffolding output Encourage children to speak 12 10 83% Make connections to known methods and vocabu‐ lary from EFL 5 2 40% Patterned output 5 4 80% Student interaction 5 5 100% Set achievable goals 4 3 75% In the language 3 3 100% Combined 127 111 87-% agreement Motivate students Coder 1 (researcher) Coder 2 Percent agreement Activities Playful and active 11 10 91% Varied 9 7 78% Build knowledge 1 1 100% Authentic 1 1 100% Social support Peer role models 4 4 100% Working together 4 4 100% Building teacher student relationships 2 2 100% Combined 32 29 91-% agreement 236 10 Appendices <?page no="237"?> 10.6 Appendix 6: Codebook 1 Teacher Cognitions about CLIL 1.1 Cognitions about the role of content, L1, L2 1.1.1 Content learning as priority A teacher statement belongs to this code if it describes content learning as the first priority in a CLIL science class. Additionally, statements can refer to the role that non-language content plays in guiding decisions about what English to use, how much English to use, and whether or not it is possible to teach a lesson without the support of the German-speaking teacher. 1.1.2 English learning as curricular enrichment and choice A teacher statement belongs to this code if it refers to the belief that learning English in a CLIL class is enrichment, a choice, optional, "on top", or something to be learned after German and content curricular concerns have been learned. 1.1.3 German subject language learning as necessary A teacher statement belongs to this code if it indicates the need for students to know or learn certain subject vocabulary and phrases in German. 1.1.4 German and English use for different purposes A statement belongs to this code if it expresses the belief that some things can be taught or learned in German and others can be taught and learned in English. Statements of this type often describe teaching complex content in German and simple content in English. Statements may also describe using English for the "possible" and German for the "necessary". The key to these statements is that language use is conceptualized as being complementary; there can be German without English and English without German. Statements that express a lack of concern for the language that students learn content in would be coded as "content learning as priority." 1.1.5 English use as barrier to relationships A teacher statement belongs to this code if it describes the use of English in a CLIL setting as creating a barrier to developing teacher-student relationships. 1.1.6 English use as a classroom lingua franca A teacher statement belongs to this code if it describes using English in the class as a lingua franca or as a way of connecting with non-German speaking students. 1.2 Cognitions about young learner SLA in CLIL 1.2.1 High language learning A teacher statement belongs to this code if it refers to the belief that students learn a lot of English in CLIL or that there is a lot of potential for foreign language learning within CLIL. 1.2.2 Natural learning A teacher statement belongs to this code if it describes the belief that foreign language learning in bilingual education is a natural process, something the students just "pick up". Statements that language learning should be implicit or acquired are also included. 1.2.3 Challenging A teacher statement belongs to this code if it refers to the belief that learning through English in CLIL is very challenging for students. 1.2.4 Need to speak 10.6 Appendix 6: Codebook 237 <?page no="238"?> A teacher statement belongs to this code if it describes the belief that a foreign language is learned by speaking, by producing output. Statements referring to how the teacher supports output in the classroom would not belong here but instead to the CLIL Approaches category. 1.2.5 Cognitive advantages A teacher statement belongs to this code if it refers to the belief that the cognitive demands placed on children due to learning through a foreign language lead to higher general learning outcomes and more creative and flexible thinking. 1.2.6 The earlier the better A teacher statement belongs to this code if it expresses the belief that the earlier children start learning a foreign language, the more likely they are to learn it. 1.3 Cognitions about choosing CLIL topics and activities 1.3.1 Concrete and action-oriented A teacher statement belongs to this code if it refers to the belief that concrete, visual, and/ or action-oriented themes are more suitable to language and content learning in CLIL education than more language-based or abstract themes. 1.3.2 Motivating topics A teacher statement belongs to this code if it refers to the belief that certain content and themes are more motivating for students and that this motivation leads to increased learning in the CLIL classroom. 1.3.3 Available materials A teacher statement belongs to this code if it describes available materials as guiding what activities and learning take place in the unit. 1.3.4 Manageable vocabulary A teacher statement belongs to this code if it describes the benefits to CLIL teaching of topics that include already known English vocabulary, for example, from EFL lessons. 1.3.5 Multifaceted activities A teacher statement belongs to this code if it expresses the belief that topics which offer varied and multifaceted activities are best for CLIL. 1.3.6 Fun for the teacher A teacher statement belongs to this code if it describes using materials and activities that the teacher herself finds enjoyable and fun. 1.4 Cognitions about teacher competences and demands 1.4.1 More effort for teachers A teacher statement belongs to this code if it refers to the additional effort from the teacher needed to achieve goals within the class. The statements may also be related to achieving the same content goals as non-bilingual classes as well as the emotional commitment and desire needed to sustain the effort over time. 1.4.2 Strong teacher language skills A teacher statement belongs to this code if it describes the belief that teachers need to feel confident, fluent, "at home", or otherwise comfortable using English. 1.4.3 Benefit from communication with other CLIL teachers A teacher statement belongs to this code if it describes the belief that CLIL teachers benefit from regular communication with other CLIL teachers, whether that be at the same school or at other CLIL schools. 238 10 Appendices <?page no="239"?> 1.5 Cognitions about student characteristics 1.5.1 Robust students benefit A teacher statement belongs to this code if it reflects a general belief that the most successful bilingual learners are overall strong and robust students with good German language skills, good at other subjects, and motivated to learn language. 1.5.2 All students can benefit A teacher statement belongs to this code if it expresses the belief that all students can benefit from CLIL. 1.5.3 Students with multicultural backgrounds can benefit A teacher statement belongs to this code if it describes the belief that children from certain kinds of family backgrounds such as those that are multicultural or those families where books are present make good CLIL students. 2 CLIL Goals 2.1 Foreign language learning goals 2.1.1 The four skills This category has been organized to represent the four skills in isolation because teachers spoke of the skills in isolation. They did not, for example, describe the skills in terms of oral and written skills or receptive and productive skills. 2.1.1.1 speaking A teacher statement belongs to this code if it refers to the goal of developing speaking skills. Statements may be focused on free speaking in dialogic speech, structured speaking or repeating certain phrases, or using English in prepared monologic speech such as in a presentation. 2.1.1.2 listening A teacher statement belongs to this code if it expresses the teacher's desire for students to develop listening comprehension skills related to understanding classroom discourse and instructions. Statements may also focus on the goal of understanding more than they would without CLIL, developing strategies to manage not understanding certain words, and understanding so well that no German support is needed. 2.1.1.3 reading (sentences and short texts) A teacher statement belongs to this code if it describes the goal of reading English sentences, words, and instructions on English worksheets, or longer English texts. Statements may be focused on global comprehension goals for English texts. 2.1.1.4 writing A teacher statement belongs to this code if it express the goal of writing in English. 2.1.2 Develop linguistic means (especially subject lexico-grammar) A teacher statement belongs to this code if it describes the goal of developing linguistic means, especially those related to learning subject vocabulary and grammar or language structure but also including pronunciation and intonation. Statements may be in the form of discussing vocabulary and how to use that vocabulary including using vocabulary as a filter through which to choose materials. 2.1.3 English beyond EFL contexts A teacher statement belongs to this code if it describes the goal of learning English in new contexts that go beyond typical EFL language learning contexts. 2.1.4 Comprehension strategies 10.6 Appendix 6: Codebook 239 <?page no="240"?> A teacher statement belongs to this code if it describes the goals of developing comprehension strategies. 2.2 General pedagogic goals that support language learning 2.2.1 Reducing fear and developing confidence A teacher statement belongs to this code if it refers to the goal of reducing students' inhibitions when using English, developing a sense of security and confidence with the language, or developing a sense that English use is normal. 2.2.2 Motivated language learners A teacher statement belongs to this code if it describes the desire or goal of having and developing motivated language learners. Statements about language learning being fun or enjoyable also belong to this code. 2.2.3 Success through effort A teacher statement belongs to this code if it describes the belief that students need to learn how to manage and deal with linguistic uncertainty and still try to figure out what is being said, and that through this effort, they will develop self-confidence, experience positive emotions, and have an opportunity to experience success. Statements coded as Success through effort differ from those that belong to Motivated language learners in that they emphasize working through difficulty and not just enjoying the process. 2.2.4 More respectful A teacher statement belongs to this code if it describes the goal of developing respect in children; Respect for other cultures and for other language learners as they struggle to learn a language. 3 CLIL Approaches 3.1 Focus on input 3.1.1 Provide as much English input as possible 3.1.1.1 Provide as much oral English input as possible A teacher statement belongs to this code if it refers to providing students with as much spoken English input as possible. Statements that should be coded describe the belief that speaking a lot of or only English is beneficial or describe how this is done in class. Also, statements about oral input such as songs, reading aloud etc. belong to this code. 3.1.1.2 Authentic materials Teacher statements belong to this code if it refers to acquiring and using materials (such as picture books) that have been created for native English speaking children. 3.1.1.3 Independently used materials A teacher statement belongs to this code if it refers to using materials that can be used by students independently of the teacher. For example, students can read and follow simple English instructions on their own. 3.1.2 Provide German input when necessary 3.1.2.1 Manage complexity A teacher statement belongs to this code if is describes using German when it is evident that the children no longer understand, to support more complex subject learning objectives, to dig deeper into a topic, for more complex reading and writing activities, or to summarize teaching. 3.1.2.2 Pastoral care 240 10 Appendices <?page no="241"?> A teacher statement belongs to this code if it describes using German for social and emotional issues with children. This includes statements related to dealing with the health and safety of children and disciplinary issues. 3.1.2.3 Teach local and personal topics A teacher statement belongs to this code if it describes using German to teach topics that are rooted in local German culture such as local bicycle safety rules or German geography as well as topics that may be personal or sensitive such as sexual education. 3.1.2.4 Save time A teacher statement belongs to this code if it describes using German to save time. 3.2 Know the students 3.2.1 Know the class A teacher statement belongs to this code if it refers to the belief that teachers need to use their knowledge of the present and past classes to plan for and set appropriate language learning goals for the present class. 3.2.2 Know individuals A teacher statement belongs to this code if it refers to the belief that teachers need to use their knowledge of individual students’ abilities, personalities, and differences when making decisions about what language learning goals to work toward and how to motivate students toward those goals. These statements are often not explicit but instead indicate teacher's actions and reasons for those actions. 3.3 Scaffolding 3.3.1 Scaffolding input 3.3.1.1 Translation 3.3.1.1.1 German teacher translation A teacher statement belongs to this code if it describes when and how the German-speaking teacher is used. 3.3.1.1.2 Student translation A teacher statement belongs to this code if it describes using students to translate English input from the teacher into German for students who don’t understand. 3.3.1.2 Visual strategies A teacher statement belongs to this code if it refers to using body language and visual materials such as objects and pictures to communicate meaning. 3.3.1.3 Language strategies in L2 3.3.1.3.1 Simplify language A teacher statement belongs to this code if it refers to simplifying or reformulating oral or written input so that it can be better understood by students. Statements may describe the need to use language carefully. 3.3.1.3.2 Repeat and reuse language A teacher statement belongs to this code if it describes the use of repetition or regular exposure to the same language (for example in singing a song) in order to learn the language. 3.3.1.3.3 Explicit comprehension checks A teacher statement belongs to this code if it describes the use of explicit comprehension checks, either on the part of the teacher or on the part of students. 3.3.1.4 Materials strategies 10.6 Appendix 6: Codebook 241 <?page no="242"?> 3.3.1.4.1 Use dictionaries A teacher statement belongs to this code if it describes the need to use dictionaries, for example, self-made picture dictionaries, normal dictionaries, or online translation tools. 3.3.1.4.2 Create structured and differentiated worksheets A teacher statement belongs to this code if it describes the need for worksheets to be structured in a way that supports understanding and remembering information. 3.3.1.5 Interactive strategies A statement belongs to this code if it describes using strategies based on how students and teacher interact with one another, for example, the teacher clearly explaining a worksheet, students asking questions, or giving students the necessary time to figure out a written text on their own. 3.3.2 Scaffolding output 3.3.2.1 Encourage children to speak A teacher statement belongs to this code if it refers to the belief that the teacher should motivate and encourage students to speak English. 3.3.2.2 Patterned output A teacher statement belongs to this code if it describes using various structures (sentences, chunks) that students can repeat or use to produce output. 3.3.2.3 Known methods and vocab from EFL lessons A teacher statement belongs to this code if it describes using methods that students are familiar with from the EFL lessons (such as presentations) in CLIL lessons. Being familiar with the methods supports students output in the context of CLIL lessons. 3.3.2.4 Student interaction A teacher statement belongs to this code if it is focused on the linguistic interaction of students such as students correcting each other's language etc. These statements differ from those coded as Working together under Social support and Motivation in that in this code, there is a linguistic focus and not a social focus. 3.3.2.5 Set achievable goals A teacher statements belongs to this code if it describes asking students to produce some amount of English output, either written or spoken, that the teacher deems achievable for the student. 3.3.2.6 In the language A teacher statement belongs to this code if it describes the belief that students will naturally speak the target language when they are "in the language," i.e. when they have been working with the language in the form of input and output. 3.4 Motivate students 3.4.1 Activities 3.4.1.1 Varied A teacher statement belongs to this code if it describes the belief that using a variety of activities is motivating. Teachers may describe planning activities that go beyond, for example, using texts in order to open up different ways of approaching the topic. Teachers may describe making learning more interesting or allowing students to choose preferences. 3.4.1.2 Playful and active A teacher statement belongs to this code if it describes the belief that using activities that children find fun and playful such as games, puzzles, songs, and other play-based activities are motivating. 3.4.1.3 Authentic situations 242 10 Appendices <?page no="243"?> A teacher statement belongs to this code if it describes the belief that authentic language use situations are motivating for students. 3.4.2 Social support 3.4.2.1 Peer role models A teacher statement belongs to this code if it describes the use of more competent English speakers (students) being role models to less competent English speakers. 3.4.2.2 Working together A teacher statement belongs to this code if it describes having students share or present their findings or results to others so that they feel proud or take ownership of their work. 3.4.2.3 Relationship with teacher A teacher statement belongs to this code if it refers to building relationships on a personal level with students so that students feel comfortable with the teacher and are willing to then use the foreign language. 3.5 Student choice of language A teacher statement belongs to this code if it describes the belief that allowing students to make choices about using English, both input (for example choosing a German or English worksheet) and output (for example, choosing to speak English or German) is motivating. 3.6 Teach language explicitly 3.6.1 Teach/ highlight vocabulary A teacher statement belongs to this code if it refers to teaching subject vocabulary explicitly. 3.6.2 Repeat sentences A teacher statement belongs to this code if it refers to repeating sentences after the teacher or reciting phrases and sentences together as a class. 3.7 Plan for more time A teacher statement belongs to this code if it refers to the belief that additional time is needed for: 1) students to learn content through a foreign language and 2) to teach students the subject vocabulary in English. The focus here is on class time needed and not teacher planning time or personal time. 3.8 Build links to other subjects A teacher statement belongs to this code if it describes looking to integrate thematic language across CLIL subjects (e.g. sports, art, music, science and social studies). 10.6 Appendix 6: Codebook 243 <?page no="245"?> List of Figures Figure 1: Range of programs that integrate content and language* (Tedick and Cammarata, 2012) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31 <?page no="247"?> List of Tables Table 1: Areas of Competence and Content Focus Areas in the Primary School Science Curriculum in NRW (MSB NRW, 2021) . . . . . . . . . . . . . . . . . . . . . . 54 Table 2: Areas of Competence and Focus Areas in the Primary School English Curriculum in NRW (MSB NRW, 2021) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55 Table 3: Dimensions of observational research (Borg, 2015, p.-269) . . . . . . . . . . . . . 86 Table 4: Participating teachers’ biographical data . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 Table 5: Data gathering plan . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 97 Table 6: Amount of data gathered for each teacher . . . . . . . . . . . . . . . . . . . . . . . . . . . 98 Table 7: Dimensions of observational research (Borg, 2015, p.-215) . . . . . . . . . . . . . 104 Table 8: Intercoder agreement for the subcategory Cognitions about the role of content, L1, L2 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116 Table 9: Cognitions about the role of content, L1, L2 . . . . . . . . . . . . . . . . . . . . . . . . . 122 Table 10: Cognitions about young learner SLA in CLIL . . . . . . . . . . . . . . . . . . . . . . . . 128 Table 11: Cognitions about choosing CLIL topics and activities . . . . . . . . . . . . . . . . . 133 Table 12: Cognitions about teacher competences and demands . . . . . . . . . . . . . . . . . 136 Table 13: Cognitions about student characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140 Table 14: Foreign language learning goals . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 143 Table 15: General pedagogic goals that support language learning . . . . . . . . . . . . . . 152 Table 16: Focus on input . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157 Table 17: Know the students . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164 Table 18: Scaffolding . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167 Table 19: Motivate students . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176 Table 20: Student choice of language . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 179 Table 21: Teach language explicitly . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 181 Table 22: Plan for more time . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182 Table 23: Build links to other subjects . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183 <?page no="248"?> Giessener Beiträge zur Fremdsprachendidaktik ISBN 978-3-381-12801-3 Bilingual education, or CLIL, at primary school varies greatly across European educational contexts. Teaching Young Learners in Bilingual Settings reports on a study that explored one such CLIL context in Dortmund, Germany. Through interviews and classroom observations, the researcher and author sought not only to document some of what takes place in CLIL classrooms but to describe and understand teachers’ thoughts and beliefs about their CLIL teaching practices. This research contributes to a better understanding of primary school CLIL programs and teachers and is relevant for researchers working in the fields of foreign language education, bilingual education, and language teacher cognition research. Furthermore, the insights into CLIL teachers’ thinking can support CLIL teachers, administrators, and policy makers as they seek to further develop CLIL pedagogy and programs.
