Text Competence and Academic Multiliteracy From Text Linguistics to Literacy Development Susanne Göpferich Te TT xt Competence and Academic Multiliteracy Europäische Studien zur Textlinguistik herausgegeben von Kirsten Adamzik (Genf) Martine Dalmas (Paris) Jan Engberg (Aarhus) Wolf-Dieter Krause (Potsdam) Arne Ziegler (Graz) Band 16 Susanne Göpferich Te TT xt Competence and Academic Multiliteracy From Text Linguistics to Literacy Development 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. Titelabbildung: © Minerva Studio - Fotolia.com © 2015 · Narr Francke Attempto Verlag GmbH + Co. KG Dischingerweg 5 · D-72070 Tübingen 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. Gedruckt auf säurefreiem und alterungsbeständigem Werkdruckpapier. Internet: http: / / www.narr.de E-Mail: info@narr.de Druck: Docupoint GmbH, Magdeburg Printed in Germany ISSN 1860-7373 ISBN 978-3-8233-6934-9 Contents List of Abbreviations................................................................................................. X List of Tables and Figures ...................................................................................... XI Acknowledgements ................................................................................................XV Introduction: The objective and scope of this book .....................................XVII Part I: Textlinguistic Foundations 1 Basic concepts ......................................................................................................3 1.1 Text linguistics and text..............................................................................3 1.2 Textuality......................................................................................................5 1.2.1 Cohesion ...........................................................................................6 1.2.1.1 Recurrence (repetition) ....................................................6 1.2.1.2 Substitution ........................................................................7 1.2.1.3 Pro-forms ...........................................................................8 1.2.1.4 Definite and indefinite articles ........................................9 1.2.1.5 Ellipsis .................................................................................9 1.2.1.6 Explicit (metacommunicative) text reference .............10 1.2.1.7 Sequence of tenses (consecutio temporum) and aspect.................................................................................10 1.2.1.8 Connectives (conjunctions and adverbs).....................10 1.2.2 Coherence .......................................................................................11 1.2.2.1 Isotopy ..............................................................................11 1.2.2.2 Presuppositions ...............................................................11 1.2.2.3 Frames and scripts...........................................................13 1.2.2.4 Functional sentence perspective (FSP).........................14 1.2.3 The relationship of cohesion and coherence .............................26 1.2.4 Intentionality and acceptability ...................................................26 1.2.5 Informativity ..................................................................................26 1.2.6 Situationality ..................................................................................27 1.2.7 Intertextuality.................................................................................27 1.2.8 Communicative quality of texts...................................................28 VI Contents Part II: Text Comprehension and Text Comprehensibility 2 Levels of cognitive processing .........................................................................31 2.1 The legibility of texts.................................................................................31 2.2 The readability of texts .............................................................................32 2.3 Text comprehension as a constructive process .....................................33 3 Text processing from a cognitive-science perspective ...............................35 3.1 Propositional models of text processing ................................................35 3.2 The structure of the human mind...........................................................37 3.3 The model of cyclic processing................................................................40 3.4 Network models ........................................................................................42 3.5 Semantic macro-structures ......................................................................42 3.6 Schema-theoretical approaches ...............................................................45 3.7 The theory of mental models ...................................................................48 3.8 Levels of comprehension..........................................................................49 3.9 Concluding remarks .................................................................................50 4 Text processing from the perspective of instructional psychology .........51 4.1 The Hamburg psychologists’ empirical inductive approach...............51 4.2 Groeben’s theoretical deductive approach.............................................55 4.3 The Karlsruhe comprehensibility concept.............................................58 4.3.1 Communicative function .............................................................60 4.3.2 Guiding features of text production ...........................................61 4.3.3 Comprehensibility dimensions....................................................65 4.3.4 Completeness of the Karlsruhe comprehensibility concept? ..........................................................................................75 5 Methods of comprehensibility assessment...................................................77 5.1 Target-group-focused methods of comprehensibility assessment .....77 5.1.1 Exteriorization of comprehension for the text as a whole and for its components .................................................................78 5.1.2 Selection of participants ...............................................................79 5.2 Research design .........................................................................................81 5.3 Data analysis ..............................................................................................86 5.4 Results .........................................................................................................88 5.5 Summary and conclusions .......................................................................95 5.6 Improvement of the experimental design........................................... 100 5.7 More efficient alternatives? ................................................................... 102 Contents VII Part III: Text Production 6 Writing (process) models.............................................................................. 107 6.1 Hayes & Flower’s (1980) writing process model................................ 108 6.1.1 Planning ....................................................................................... 110 6.1.2 Text generation (or translating) ............................................... 110 6.1.3 Revision (or reviewing).............................................................. 111 6.1.4 The monitor ................................................................................ 112 6.2 Hayes’ (1996) writing process model................................................... 113 6.2.1 Working memory ....................................................................... 114 6.2.2 Cognitive processes .................................................................... 115 6.2.3 Long-term memory .................................................................... 117 6.2.4 Motivation/ affect ........................................................................ 117 6.3 Cooper & Matsuhashi’s (1983) writing process model ..................... 118 6.4 Günther’s (1993) p hraseo riented p roduction s ystem (POPS) ........ 120 6.5 Bereiter & Scardamalia’s (1987) models of beginners’ and advanced writers’ composing processes .............................................. 124 6.6 An instruction-oriented writing process model ................................ 127 6.6.1 The task environment ................................................................ 127 6.6.2 The individual ............................................................................. 130 6.6.2.1 The long-term memory ............................................... 130 6.6.2.2 Cognitive processes ...................................................... 136 6.6.3 Concluding remarks................................................................... 140 7 Writing competence development models................................................ 141 7.1 Writing development stages according to Bereiter (1980) ............... 142 7.2 Kellogg’s (2008) macro-stages of writing competence development............................................................................................ 143 7.3 McCutchens’ s (1996) capacity theory of writing (development) ... 147 7.4 Writing competence development from the perspective of dynamic systems theory ........................................................................ 150 7.5 Academic writing competence development from a corpus-linguistic product-oriented perspective................................. 151 7.5.1 Steinhoff’s (2007) stages of academic writing competence... 152 7.5.2 Pohl’s (2007) stages of academic writing competence .......... 156 7.6 Alexander’s (2003) Model of Domain Learning ................................ 161 VIII Contents 7.7 The bioecological model of human development and its implications for the modelling of writing competence development............................................................................................ 163 7.8 Summary ................................................................................................. 167 Part IV: Writing Instruction 8 Best-practice approaches to writing instruction...................................... 171 8.1 A theoretical framework for literacy pedagogy: multiliteracies ....... 175 8.1.1 Situated Practice ......................................................................... 176 8.1.2 Overt Instruction ........................................................................ 177 8.1.3 Critical Framing.......................................................................... 178 8.1.4 Transformed Practice................................................................. 178 8.2 Teaching for transfer ............................................................................. 179 8.3 Approaches to writing instruction at the macro-level ...................... 185 8.4 Approaches to writing instruction at the meso-level ........................ 190 8.5 Approaches to writing instruction at the micro-level ....................... 195 8.5.1 Process-oriented approaches .................................................... 197 8.5.1.1 Learning by doing ........................................................ 200 8.5.1.2 Learning by observing ................................................. 203 8.5.1.3 Learning by cognitive apprenticeship ....................... 205 8.5.1.4 Learning by reflecting .................................................. 206 8.5.1.5 Collaborative learning ................................................. 207 8.5.2 Product-oriented approaches ................................................... 210 8.6 Writing-intensive seminars and the role of the teacher.................... 210 8.7 Assignments for writing courses and writing-intensive seminars .. 213 8.8 Giving feedback ...................................................................................... 216 Part V: L1 vs. L2 Writing 9 Writing in the L1 vs. writing in the L2 ....................................................... 221 9.1 Differences between composing in the L1 and the L2 ...................... 221 9.2 Translating into the L2 vs. composing in the L2................................ 226 9.3 Translating from the L1 as a subprocess of writing in the L2 .......... 227 9.4 An explanatory model of EFL writing ability ..................................... 234 Contents IX 9.5 Quality losses in L2 writing and the potential role of translation for writing instruction, text quality improvement and epistemic purposes................................................................................................... 235 9.5.1 Data analysis................................................................................ 237 9.5.2 Results and discussion ............................................................... 243 9.5.2.1 Text coherence errors .................................................. 247 9.5.2.2 Sense errors ................................................................... 250 9.5.2.3 Implicitness errors........................................................ 252 9.5.2.4 Errors concerning functional sentence perspective (FSP) .......................................................... 254 9.5.3 Conclusion................................................................................... 257 9.6 L2 writing pedagogy............................................................................... 258 References ............................................................................................................... 261 Index......................................................................................................................... 289 List of Abbreviations AI Artificial Intelligence BMBF Bundesministerium für Bildung und Forschung (German Federal Ministry of Education and Research) CBT computer-based training DST dynamic systems theory EFL English as a Foreign Language ESL English as a Second Language Fig. Figure FN footnote FSP functional sentence perspective GAT Gesprächsanalytisches Transkriptionssystem (discourse-analytical transcription system) JLU Justus Liebig University, Giessen/ Germany L1 mother tongue L2 first foreign language L3 second foreign language LTM long-term memory LTWM long-term working memory MDL Model of Domain Learning MICUSP Michigan Corpus of Upper-level Student Papers NSSE National Survey of Student Engagement n.y. no year POPS p hraseo riented p roduction s ystem R rheme RE Reading Ease RIP retrospective interview protocol SL source language SRSD Self-Regulated Strategy Development ST source text STM short-term memory T theme TA think-aloud Tab. Table TAP think-aloud protocol TEI Text Encoding Initiative TL target language transl. translation TT target text USTM ultra-short term memory WAC writing across the curriculum WID writing in the disciplines WM working memory XML Extensible Markup Language ZfbK Zentrum für fremdsprachliche und berufsfeldorientierte Kompetenzen (Centre for Competence Development) List of Tables and Figures List of Tables Tab. 4-1 Features for the description of text quality (cf. Langer/ Schulz von Thun/ Tausch 1993: 139)................................52 Tab. 5-1 The participants’ educational and professional background ............83 Tab. 5-2 Physical and psychological condition of the participants during the experiment ...........................................................................84 Tab. 5-3 Deficiencies of the original diabetes text with regard to comprehensibility ...................................................................................96 Tab. 5-4 The popular-science text and its optimized version ..........................99 Tab. 6-1 Defining features and examples of Beaufort’s five knowledge domains of writing expertise (Beaufort 2007: 221) ......................... 131 Tab. 6-2 Knowledge classifications according to Hayes (1996), Beaufort (2007) and Kruse & Chitez (2012)..................................... 132 Tab. 7-1 The MDL’s developmental stages and their characteristics ........... 162 Tab. 9-1 Error classification scheme ................................................................. 238 Tab. 9-2 Assessment scheme based on the Karlsruhe comprehensibility concept .................................................................................................. 241 Tab. 9-3 Number of errors found in the English and German texts and number of reflected errors in the German texts....................... 244 Tab. 9-4 Frequency of text coherence error types........................................... 250 List of Figures Fig. 1-1 Texts as woven fabrics with warp and weft ...........................................4 Fig. 1-2 The coherence-creating phenomenon of recurrence (adapted from Linke/ Nussbaumer/ Portmann 1994: 216)...................6 Fig. 1-3 The coherence-creating phenomenon of substitution (adapted from Linke/ Nussbaumer/ Portmann 1994: 217)...................7 Fig. 1-4 The creation of coherence by means of pro-forms (adapted from Linke/ Nussbaumer/ Portmann 1994: 218)...................8 Fig. 1-5 Types of presuppositions (adapted from Linke/ Nussbaumer/ Portmann 1994: 234).................12 Fig. 1-6 The interplay of bottom-up and top-down processes in text comprehension (adapted from Linke/ Nussbaumer/ Portmann 1994: 236) ..............................................................................14 XII List of Tables and Figures Fig. 1-7 Simple linear progression (Daneš 1974: 118)......................................16 Fig. 1-8 Progression with a continuous theme (Daneš 1974: 118) .................17 Fig. 1-9 Progression with derived themes (Daneš 1974: 118) .........................17 Fig. 1-10 Progression with split rheme (Daneš 1974: 119 with modifications made by myself) .............................................................18 Fig. 2-1 The Reading Ease formula by Flesch....................................................32 Fig. 3-1 Multi-stage model of the human memory (cf., e.g., Arbinger 1984).........................................................................38 Fig. 3-2 Network representation of sentence [3-3]...........................................42 Fig. 4-1 Assessment window showing values for a text that, according to the Hamburg model, is optimally comprehensible .......................53 Fig. 4-2 Framework and comprehensibility dimensions of the Karlsruhe comprehensibility concept.....................................................................59 Fig. 4-3 Simple mental denotation model of a friction clutch (adapted from Niess et al. 1984: 211) ...................................................62 Fig. 4-4 Complex mental denotation model of a friction clutch, in this case a diaphragm spring clutch (adapted from Volkswagen 1988: 5)...............................................................................62 Fig. 5-1 The popular-science text to be assessed...............................................82 Fig. 5-2 The assignment .......................................................................................85 Fig. 5-3 The Translog dictionary.........................................................................86 Fig. 6-1 Hayes & Flower’s (1980: 11) writing process model in the version optimized by Hayes (1996: 3) ............................................... 108 Fig. 6-2 Hayes’ (1996: 4) writing process model ............................................ 113 Fig. 6-3 Structure of the knowledge-telling model (Bereiter/ Scardamalia 1987: 8, Fig. 1.1.) 1 .......................................... 125 Fig. 6-4 Structure of the knowledge-transforming model (Bereiter/ Scardamalia 1987: 12, Fig. 1.2.) 2 ........................................ 126 Fig. 6-5 Instruction-oriented writing process model (cf. Göpferich 2002: 250)..................................................................... 129 Fig. 6-6 Planning processes according to Hayes & Nash (1996: 45) ........... 137 Fig. 7-1 Model of skill systems integration in writing development (Bereiter 1980: 81, Fig. 4.1) 3 ................................................................ 143 1 © 1987 Lawrence Erlbaum Associates, Inc.; republished with permission; permission conveyed through Copyright Clearance Center, Inc. 2 © 1987 Lawrence Erlbaum Associates, Inc.; republished with permission; permission conveyed through Copyright Clearance Center, Inc. 3 © 1980 Lawrence Erlbaum Associates, Inc.; republished with permission; permission conveyed through Copyright Clearance Center, Inc. List of Tables and Figures XIII Fig. 7-2 Macro-stages in the cognitive development of writing skill (Kellogg 2008: 4, Figure 1) .................................................................. 144 Fig. 7-3 The development of academic writing competence (adapted from Steinhoff 2007: 138)................................................... 155 Fig. 7-4 The ontogenesis of academic writing (adapted from Pohl 2007: 488, Fig. IV-1) ......................................... 158 Fig. 8-1 Three models of educational research into student writing in higher education .............................................................................. 171 Fig. 8-2 Rules for giving and taking feedback ................................................ 217 Fig. 9-1 Explanatory model of EFL writing ability (Sasaki/ Hirose 1996: 161) 4 .................................................................. 235 4 © 1996 Language Learning Research Club, Univ. of Michigan; republished with permission of John Wiley & Sons; permission conveyed through Copyright Clearance Center, Inc. Acknowledgements I wish to thank Lisa Beier, Mareike Ekkert, Bridgit Fastrich, Ina Machura, Imke Neumann and Carmen Neis, the team with whom I established the writing centre of Justus Liebig University (JLU), Giessen/ Germany, as well as Dagmar Knorr for valuable comments on parts of this book. I am also indebted to my students at the English Department of JLU for fruitful discussions in which I could test the comprehensibility of my models and refine my ideas. Special thanks also goes to Lisa Beier for proof-reading the entire manuscript from a native-speaker perspective. Any remaining errors are, of course, entirely my own. Furthermore, I am indebted to Gerd Bräuer, Peter Hermann Braun, Melanie Brinkschulte, Katrin Girgensohn and Dagmar Knorr for an intensive exchange on the theoretical foundations of writingcentre work and writing-centre research. Furthermore, I would like to thank my colleagues from the Educational Linguistics network at JLU for an inspiring exchange on writing-related issues in our educational linguistics lecture series and workshops. In addition, I would like to thank the publisher Stauffenburg for permission to use an updated and extended English version of sections of my monograph Textproduktion im Zeitalter der Globalisierung (Text Production in the Age of Globalization; Göpferich 2002, 3rd edition 2008) in Chapters 2 to 6 of this book. I am also grateful to the Gunter Narr publishing team, especially to Tillmann Bub, for their excellent cooperation and support. And finally, I would like to express my gratitude to Michael and my family for their patience and emotional support in phases when this book project absorbed all my attention that managing my child ZfbK left over. Introduction: The objective and scope of this book The idea for the present book arose out of the desire to make text linguistics, the study of meaningful linguistic units beyond the sentence level, fruitful for the development of text competence (“Textkompetenz”) in the sense of Portmann-Tselikas & Schmölzer-Eibinger (2008), i.e., the competence to read, write and learn from texts (cf. also Preußer/ Sennewald 2012), also termed literacy (Banzer/ Kruse 2011: 2). Thematically, this book therefore covers the wide range from text linguistic foundations (Part I) via text reception (Part II) and text production (Part III) to writing instruction (Part IV) and multiliteracy (Part V). These parts are related in the following manner: The textlinguistic foundations in Part I are a prerequisite for understanding what makes a sequence of words, sentences and nonverbal representations a meaningful whole rather than a mere agglomerate of unrelated units. These textlinguistic foundations, however, are insufficient to explain the extent of a text’s readability and comprehensibility. To sensitize readers to what makes texts readable and comprehensible, or unreadable and hard to understand, and to familiarize them with methods of comprehensibility assessment, Part II of this book is therefore devoted to text comprehension and text comprehensibility including methods of comprehensibility assessment. The ability to anticipate their prospective readers’ comprehension and interpretation of their texts and thus to write comprehensibly is a major asset that distinguishes proficient writers from weaker ones (cf. Kellogg 2008; Bereiter 1980; Sections 7.1 and 7.2). This ability and further components of writing competence as well as writing skills development and writing processes will be addressed in Part III. Familiarity with writing process models helps writers to analyze their own writing processes and to pinpoint potential weaknesses in them, which can then be focused on in writing instruction and coaching. Knowledge about writing competence development assists both writing instructors and subjectdomain teachers in the design of writing assignments and more complex writing arrangements which are adequate for the competence levels that their students have already attained and the realistic assessment of what text quality can be expected from students at certain stages of their literacy development. How writing development can be fostered, especially in tertiary education, including the design of adequate writing assignments and promoting the functions of writing for clearer thinking will be covered in Part IV of this book. Especially in tertiary education, writing fulfils an important function for sharpening one’s ideas and better reasoning. Therefore, it is of utmost importance that attention be devoted to this essential skill. Why this is so is very illustratively described by Gage (1986: 24): “Writing is thinking made tangible, thinking that can be examined because it is on the page and not in the head invisibly floating around. Writing is thinking that can XVIII Introduction be stopped and tinkered with. It is a way of holding thought still enough to examine its structure, its flaws. The road to clearer understanding of one’s thoughts is travelled in paper. It is through an attempt to find words for ourselves in which to express related ideas that we often discover what we think.” Writing assessment will not be addressed since there is an abundance of literature, including an academic journal, Assessing Writing , on this topic, which could fill a volume of its own. For a bibliography on writing assessment, see Silva & Brice (2004: 73); for second-language writing assessment, see Hamp- Lyons (1991). The last part of this book, Part V, is devoted to differences between L1 and L2 composition and their interrelationships. Multiliteracy as literacy in more than one language is of central importance in a globalized world, where literacy in English as the lingua franca of cross-cultural communication is a prerequisite for professional success in an increasing number of disciplines and professional domains. The concept of multiliteracy as it is used here is one component of the New London Group’s (1996) multiliteracies concept (see Section 8.1). They use it in the plural encompassing both a) literacy in more than one language and their cultural contexts as covered by the multiliteracy concept used here, and b) literacy in more than one medium, a second component that will not be addressed in this volume though it is not deemed less important. 1 Multiliteracy and thus multilinguality (or at least bilinguality) does not only play a role in Part V of this book. In the other parts, a contrastive English, German and in some contexts also French approach is taken, where deemed appropriate, to create an awareness of language-specific constructions and their rhetorical functions and thus to sensitize readers for translation-related issues and foster their translation competence, which is considered an important component of multiliteracy. Following Lea & Street (1998; 2006), writing instruction and, more generally, literacy development, in academic contexts can be conceptualized in three overlapping models: a) a study skills model, b) an academic socialization model and c) an academic literacies model (for details, see Section 8). The study skills model focuses on surface features of language or the lexical and grammatical correctness of texts. The academic socialization model is concerned with students’ acculturation into disciplinary and subject-related discourses and genres and focuses on aspects of writing that can only be taught and acquired in or in connection with subject-matter courses. The academic literacies model “is similar in many ways to the academic socialization model, except that it views the processes involved in acquiring appropriate and effective uses of literacy as more complex, nuanced, situated, and involving both epistemological issues and 1 On the debates surrounding the use of the singular multiliteracy or plural multiliteries , see also Lillis & Scott (2007). Introduction XIX social processes, including power relations among people, institutions, and social identities” (Lea/ Street 2006: 368 f.). In this book, literacy development will be addressed on all three levels, from the development of writing skills as a central study skill in compensatory writing courses (study skills) to writing arrangements in writing-intensive seminars (academic socialization) and progressive writing-across-the-curriculum (WAC) designs (academic literacy). Especially Chapter 8 “Best-practice approaches to writing instruction” addresses the post-Bologna Reform era with its new challenges for students’ (multi-)literacy development resulting from shorter degree programs, a larger student intake in universities and thus more heterogeneous entrance qualifications including writing skills among students. All of these factors make it necessary to foster students’ writing skills development in a more efficient and effective manner than this had been acceptable in the pre-Bologna Reform era. In the United States, a similar situation had to be mastered some 35 years ago (Lillis/ Scott 2007: 8). Against this background, it is not surprising that many developments that could be observed in the US more than three decades ago have now been taken over in Germany and other European countries (for an overview, see the articles in Björk et al. 2003b). Among these developments is the establishment of writing centres as central support structures in universities. Research on writing skills development and approaches to teaching writing are necessarily embedded in their respective national educational contexts. Unsurprisingly, this has led to national discourses which, to a certain extent, are being conducted independently. The present book is among the first to combine the US-American discourse on literacy development in the fields of composition studies, (applied) linguistics and rhetoric with the German discourse on writing skills development mainly conducted in Germanic language departments. Both have made major contributions to the literacies development discourse without always taking notice of each other to a sufficient extent. For example, I have not found invaluable German work on academic literacy development, such as the studies by Steinhoff (2007) and Pohl (2007), quoted in American publications. This is probably due to German-language publications in general hardly being read in the US, and publications related to English language not sufficiently taken account of in departments of Germanic languages in Germany. This book tries to bridge the gap by bringing both discourses together. In embarking on this endeavour, I had to decide in which language to write this book, my mother tongue German or my L2 and working language English. This was no easy decision. I am a supporter of the use of German as a language of academic discourse and I would like to reach a German-speaking as well as an English-speaking audience. To reach those who are responsible XX Introduction for writing instruction in German universities, writing this book in German might have been the best choice. By writing in German, however, I would have risked that the bulk of the German literature that I reviewed in this book and that I have not found quoted in English literature, although it makes important contributions to the discourse, would never have been read in Englishspeaking countries. Therefore, I decided to write this book in English. In this manner, I could also present research findings so far published only in German in English and thus ensure that they can be taken account of in the English-speaking world since they deserve to be acknowledged. I apologize in advance to my German-speaking colleagues for not using our language, however, I felt taking this risk was justifiable because I know that they can cope with English as the lingua franca of academic discourse, not only in English departments. I hope that this book contributes to a cross-cultural exchange between writing researchers and teachers from different countries. As the Director of the Centre for Competence Development at Justus Liebig University (JLU), Giessen/ Germany (ZfbK 2014), and the project leader of the JLUwide project “Starting with Success” (“Einstieg mit Erfolg”) for the improvement of students’ study skills and the quality of teaching at JLU funded by the German Federal Ministry of Education and Research (BMBF 2012) in the time span from 2012 to 2016, I also hope that this book will make at least a modest contribution to the adaptation of university education at JLU and other European universities to the post-Bologna Reform requirements. The book addresses different audiences: It provides bachelor, master and postgraduate students as well as writing instructors embarking on research projects investigating writing skills and their development with an overview of the discourses in these areas and introduces them to the pertinent models and theories on which this research can be based. At the same time, it assists them in developing their own writing skills. It does so by creating an awareness of defects in texts and how they affect text comprehensibility and by providing insights into the phases of writing processes and writing development. By doing so, it helps its readers to pinpoint the steps in which they can improve their own text production. Writing teachers, writing centre staff, such as writing tutors and writing fellows, as well as subject-domain teachers interested in founding their writing instruction, writing assignments and feedback on research-based findings will be provided with a thorough introduction to the related areas of research and their results. University administrators and program coordinators can inform themselves about best-practice approaches of how to establish writing instruction and support at the different levels of a university reaching from individual courses (micro-level) via entire programs (meso-level) to central support structures such as teaching and writing centres (macro-level). These latter aspects are focused on in Chapter 8. Introduction XXI I hope that academic literacy development will benefit from this book and would be grateful for any comments and suggestions for improvement, which may be sent to my e-mail address susanne.goepferich@zfbk.uni-giessen.de. Giessen, December 2014 Susanne Göpferich Part I Textlinguistic Foundations 1 Basic concepts 1.1 Text linguistics and text Text linguistics is a relatively recent field of linguistics that evolved in the 1960s and 1970s, a period that marks the so-called text-linguistic turn . Text linguistics focuses on linguistic units beyond the sentence level that are perceived as entities but can no longer be described and explained within the confines of syntax. Central questions that text linguists are interested in are: 1. What constitutes a text, i.e., what makes a mere agglomerate of sentences - and sometimes also smaller or even non-linguistic units - a text (text constitution, textuality)? 2. How are texts delimited, i.e., what marks their beginning and their end (text delimitation)? 3. How can texts be classified (text typology)? 4. How are texts understood (text comprehension and text comprehensibility)? Three other questions that are closely related to the previous ones but that are usually not covered in monographs on text linguistics are the following: 5. How are texts produced, especially in a competent manner (text production)? 6. In what respects does text production in a writer’s mother tongue (L1) differ from text production in a second or foreign language (L2)? 7. How does text production competence in the L1 and the L2 develop, how are they interrelated and how can they be taught? These are the central questions that will be tackled in the course of this book. Answers to these questions, especially to questions 1 and 2, have been provided from two perspectives: a) a text-grammatical and text-semantic perspective, i.e., a bottom-up perspective from the sentence to the text, and b) a pragmatic or functional-communicative perspective, i.e., a top-down perspective from the text to the individual sentences. Before we take a closer look at these approaches, the central concept of text linguistics, ‘text’, needs to be explained. Despite numerous attempts to define what a text is, there is no universally accepted text definition (Brinker 1973: 9; Reiß 1983: 2; Heinemann/ Viehweger 1991: 13). Trying to find such a definition would be a useless endeavour since what is considered, or should be considered, to be a text depends on the research question asked (Brinker 1973: 9). In what follows, I will therefore confine myself to outlining a few aspects that, to my mind, should be taken into account when attempting to define the concept of ‘text’. 4 Part I: Textlinguistic Foundations 1. The term text comes from the Latin word textus meaning a ‘woven fabric’ ( texere - ‘to weave’). A woven fabric is composed of longitudinal threads, the warp (German Kette ), that are intersected with traverse threads, the weft (German Schuss ; see Fig. 1-1). Fig. 1-1: Texts as woven fabrics with warp and weft In a text, the warp is composed of the sequence of linguistic signs, in most cases sentences, that the text consists of. The traverse threads in a text, the weft, are formed by additional relationships between the elements in a text, for example, by so-called isotopic features . Isotopic features are semantic components (semes) shared by various lexical items in a text. In the following short text (example [1-1]), there are two sets of isotopic features, one set found in words that have something to do with marrying (bold), the other one found in words denoting something negative (underlined). [1-1] My sister’s wedding While we were squeezing my sister into her wedding gown, her future husband realized that he had forgotten the trousers of his wedding suit at his parents’ home. My little brother had to drive off to get them for him. It took him ages to come back because there was a traffic congestion. When he finally came back, my bother-in-law could not remember where he had put the wedding rings. My sister got furious about it and threw her bridal bouquet out of the window. Unfortunately, the bouquet hit the priest on his head and he had to see a doctor. Further relationships in a text that can be compared to the warp and weft in a woven fabric will be explained in Section 1.2. 2. Texts can be considered the primary form of organisation in which language manifests itself. “Whenever people communicate by means of language, they do so (they speak and write) in texts. In other words: ‘texts’ are the units in which verbal communication is organized.” (Isenberg 1977: 144; my transl.; cf. also P. Hartmann 1971: 15). 3. The manner in which a text is verbalized is determined by its function. Without a function, there can be no text (Oomen 1972: 19). 4. An exhaustive description of a text must not be reduced to the formal unit that can be described by linguistic means alone; it must also take into account its function (“text-in-function”, S. J. Schmidt 1973: 145; my transl.) 1 Basic concepts 5 and the situation in which, or for which, it has been produced (“text-insituation”, Weinrich 1976: 16; my transl.). 5. Since nonverbal representations may be integral parts of verbal texts or may even replace them, for example, in comic strips, they should not be excluded from a text definition. The metaphor of the woven fabric is applicable also to sequences of nonverbal representations, combinations of verbal and nonverbal representations and even multi-modal units of communication. 6. The term text , as used in this book, does not only refer to written sequences of sentences (and sometimes also smaller units of communication) but also includes their spoken counterparts. 1.2 Textuality In their influential Introduction to Text Linguistics , Robert de Beaugrande and Wolfram Dressler (1981) introduced seven “standards” (or criteria) of textuality, by means of which, they believed, texts could be distinguished from nontexts: “A TEXT will be defined as a COMMUNICATIVE OCCURRENCE which meets seven standards of TEXTUALITY . If any of these standards is not considered to have been satisfied, the text will not be communicative. Hence, non-communicative texts are treated as non-texts” (de Beaugrande/ Dressler 1981: 3). In the following, these criteria will be introduced, explained and also criticized. Based on this criticism, it will become apparent that, today, the idea is gaining ground that the concept of ‘text’ is rather a prototype concept that cannot be defined by means of necessary and sufficient criteria. In other words, the criteria specified by de Beaugrande & Dressler are met by many texts; if some of them, however, are not fulfilled by a linguistic unit, this does not necessarily mean that the unit cannot be considered to be a text. In spite of this criticism, familiarizing oneself with these seven standards is useful for getting insight into what usually, or prototypically, constitutes a text. The seven standards of textuality are: 1. cohesion, 2. coherence, 3. intentionality, 4. acceptability, 5. informativity, 6. situationality and 7. intertextuality. Cohesion and coherence are “text-centered notions, designating operations directed at the text materials” (de Beaugrande/ Dressler 1981: 7), the other five 6 Part I: Textlinguistic Foundations ones are “user-centered notions which are brought to bear on the activity of textual communication at large, both by producers and by receivers” (de Beaugrande/ Dressler 1981: 7). 1.2.1 Cohesion According to de Beaugrande & Dressler (1981: 3), cohesion “concerns the ways in which the components of the SURFACE TEXT , i.e. the actual words we hear or see, are mutually connected within a sequence . The surface components depend upon each other according to grammatical forms and conventions, such that cohesion rests upon GRAMMATICAL DEPENDENCIES .” There are grammatical dependencies within sentences that can be described within the confines of traditional (sentence) grammar. What is relevant from a text-linguistic perspective are grammatical dependencies beyond sentence boundaries. In the next sections, whose structure basically follows Linke, Nussbaumer & Portmann (1994: 215 ff.), some of these cross-sentential relationships will be illustrated. Other relationships could be added. Note that grammatical dependencies in many cases also involve semantic dependencies (and thus coherence, see Section 1.2.2 below) without which they could not be interpreted sufficiently. 1.2.1.1 Recurrence (repetition) Recurrence refers to the (identical) repetition of an element that has been introduced in an earlier sentence, as in example [1-2] where the word cat reoccurs. [1-2] Yesterday, I observed a cat that had slipped down a slope and fallen into a pond. The cat was not very fond of getting wet and tried to get out of the water as quickly as possible. Since the slope was very slippery, the cat had to struggle hard until it finally managed to climb out of the water. Graphically, the recurrence of the word cat , which can be considered another form of weft in our fabric ‘text’, can be represented as follows: Fig. 1-2: The coherence-creating phenomenon of recurrence (adapted from Linke/ Nussbaumer/ Portmann 1994: 216) In Fig. 1-2, the identical squares represent the recurrent word cat that, in example [1-2] above, refers three times to the same entity, cat, in the extra- 1 Basic concepts 7 textual world. As example [1-3] below will show, identical words that reoccur in subsequent sentences need not necessarily refer to the same entities in the real world. [1-3] Her husband is very jealous. This is completely different with her sister’s husband. Husbands who do not worry too much are much more pleasant to live with. In this example, the first and second occurrences of husband refer to different persons, and the third occurrence refers to husbands in general. A comparison of examples [1-2] and [1-3] illustrates that text comprehension involves more than interpreting just what is given at the text surface. The reader knows that, in the first example, cat always refers to the same entity whereas, in the second example, the identical words husband refer to different entities in each case. To recognize this, we need world knowledge that goes beyond what is given in the text. 1.2.1.2 Substitution Substitution refers to the occurrence of text elements which, in the course of a text, replace other words or phrases which refer to the same extra-textual entities as the text elements that replace them, i.e., they have the same reference (co-reference). The substitutes may be synonyms, hyponyms, hypernyms, metaphors and lexemes belonging to the same lexical field, but also linguistic elements whose semantic relation to the elements they substitute is less obvious. In the following example [1-4], this type of co-reference can be found in the words printed in bold: [1-4] Last year, she bought a poodle. Very soon, she realized that this dog did not like being transported in a car. He always tried to creep under the front seat. The whining creature soon opposed heavily to approaching anything that has four wheels. Graphically, substitution can be represented as shown in Fig. 1-3. Fig. 1-3: The coherence-creating phenomenon of substitution (adapted from Linke/ Nussbaumer/ Portmann 1994: 217) In Fig. 1-3, thinner arrows were used for the relationships between the elements that substitute each other whereas thicker arrows were used for their reference to the object in the real world. The thinner arrows indicate that co- 8 Part I: Textlinguistic Foundations hesion is weaker in the case of substitution than in the case of recurrence; therefore, world knowledge plays an even more important role. Furthermore, the substituting elements introduce additional semantic features not necessarily contained in the elements they replace. Thus whining creature conveys a different message than, for example, barking monster . 1.2.1.3 Pro-forms A pro-form is an item in a sentence, typically a pronoun, a verb or an adverb, that substitutes for another phrase or clause. In the following short dialogue [1-5], they and so in B’s answer are pro-forms for the elements parents and will leave tomorrow respectively. [1-5] A: Will your parents leave tomorrow? B: They said so. Strictly speaking, the use of pro-forms is a special case of substitution where the substituting element does not carry - or hardly carries - any meaning of its own but rather derives its meaning from the element that it substitutes. Halliday & Hasan (1976) differentiate between the following three types of substitution by pro-forms: “nominal”, “verbal” and “clausal substitution”. In the example above, we find both nominal substitution (parents  they) and clausal substitution (that they will leave tomorrow  so). The following example [1- 6] illustrates verbal substitution (knows  does): [1-6] A: Do you think Sarah already knows? B: I think everybody does. Graphically, the use of pro-forms can be represented as shown in Fig. 1-4. Fig. 1-4: The creation of coherence by means of pro-forms (adapted from Linke/ Nussbaumer/ Portmann 1994: 218) As Fig. 1-4 shows, pro-forms themselves do not refer to extra-textual entities directly. They can rather be interpreted as instructions to search for other textual elements to which they refer and then to create a connection to extratextual entities via these other textual elements. These other textual elements may precede the pro-form. In this case, we speak of an anaphoric reference . These other elements, however, may also follow the pro-forms. In that case, we refer to cataphoric reference , as in example [1-7]. 1 Basic concepts 9 [1-7] Sarah brings a bottle of wine. She always does so. (anaphoric references of she and wine ) If she comes at all, Sarah will bring some wine. (cataphoric reference of she ) 1.2.1.4 Definite and indefinite articles Like pro-forms, definite and indefinite articles can also be used as coherencecreating devices. They, too, can be considered as instructions for the reader to search for elements in the cotext or their prior knowledge to which they refer. If the search instruction refers to the cotext, we speak of text deixis or endophoric reference ; if it refers to the reader’s prior knowledge, this pertains to knowledge deixis or exophoric reference (Halliday/ Hasan 1976: 33). The indefinite article is used to introduce new entities in a text. It signals to the reader: “There is something that you do not know yet or of which we have not spoken so far.” The definite article is used to refer to entities which have already been introduced in the text or which the reader already knows. In the first case, the instruction signalled by the definite article is: “There is something that has been mentioned in the text already. Search for it.” ( endophoric reference , more specifically anaphoric reference according to Halliday/ Hasan 1976: 33). In the latter case, the instruction signalled by the definite article is: “There is something that you know already. Search for it in your prior knowledge of the world.” (knowledge deixis or exophoric reference) . In the following example [1- 8], the menu has not been mentioned before. Nevertheless it can be used with the definite article because it forms part of our world knowledge that restaurants usually have menus. [1-8] We entered a restaurant. After sitting down, we asked for the menu. 1.2.1.5 Ellipsis The coherence-creating effect of ellipses is comparable with that of pro-forms. Both instruct the reader to search in the text for linguistic elements to which they refer. The difference between them consists in the fact that pro-forms are linguistic means that lack content which has to be searched for in the linguistic elements they refer to whereas ellipses effect such search instructions because they ‘consist’ of missing linguistic material; ellipses are omissions of something that has been mentioned earlier in the text. Halliday & Hasan (1976: 143) thus regard ellipsis as “substitution by zero”. According to them “[a]n elliptical item is one which, as it were, leaves the specific structural slots to be filled from elsewhere” (Halliday/ Hasan 1976: 143). The following three examples [1-9], [1-10] and [1-11], all quoted in Esser (2009: 42), illustrate different types of ellipses, indicated by [ ∅ ]: 10 Part I: Textlinguistic Foundations [1-9] It was quite a surprise, therefore, to find that most galaxies appeared redshifted: nearly all [ ∅ ] were moving away from us! (BNC 1 H7859) ([ ∅ ] = galaxies; nominal ellipsis) [1-10] The scramble might be a training exercise, or it might not [ ∅ ]. (BNC CDA 3066) ([ ∅ ] = be a training exercise; verbal ellipsis) [1-11] The cheque is still valid. The Bank can tell them [ ∅ ]. ([ ∅ ] = that it is still valid; clausal ellipsis) 1.2.1.6 Explicit (metacommunicative) text reference Whenever authors explicitly refer to other parts of their texts, i.e., whenever they make metacommunicative utterances, we speak of explicit text reference . Explicit text reference is necessary when authors wish to refer to parts of their texts which are not in the immediate neighbourhood of the location where the text reference is made. Examples of metacommunicative linguistic elements are: as shown in the previous section in the following as mentioned in Chapter X see above/ below 1.2.1.7 Sequence of tenses (consecutio temporum) and aspect The tenses and aspects that are used in a text can also have a coherencecreating function signalling the order in which certain actions or events took place and setting them in relation to each other. [1-12] After he had bought the rather expensive book, a felt obliged to read it. In example [1-12], the past perfect (had bought) indicates that the act of buying took place before the feeling of obligation, verbalized in past tense, occurred. [1-13] They were enjoying the lovely weather, when they heard a loud noise. In example [1-13], the past tense continuous form describes a background action or situation that is interrupted by a sudden event, verbalized in the simple past tense. 1.2.1.8 Connectives (conjunctions and adverbs) A last coherence-creating device that shall be mentioned here are connectives, such as conjunctions and adverbs. They specify the logical or semantic relationship between the sentences or clauses that they connect. Whereas coordinating conjunctions (coordinators), such as and and or , simply indicate that something is added, subordinating conjunctions (subordinators), such as 1 BNA = British National Corpus 1 Basic concepts 11 when , although and if , express temporal, concessive and conditional relationships respectively. 1.2.2 Coherence Whereas cohesion refers to relationships of units within a text that are signalled at the text surface, i.e., that are carried by signifiants , coherence refers to logical-semantic relationships between elements in texts which are not necessarily signalled at the text surface. As de Beaugrande & Dressler (1981: 6) state, coherence “is clearly not a mere feature of texts, but rather the outcome of cognitive processes among text users.” Consider the following example: [1-14] John kicked the ball. The window broke. The logical relationship between the two sentences is not stated explicitly. Nevertheless, we interpret the two sentences in such a way that the second one describes a consequence of the first one: The window did not break just by chance but was destroyed by the ball kicked by John. For this interpretation, readers have to fall back on their knowledge of the world to create a coherent textual world. This adding of one’s own knowledge to fill gaps in a text and thus make it coherent is called inferencing (de Beaugrande/ Dressler 1981: 6). A few coherence-creating mechanisms in texts will be explained in the following sections. 1.2.2.1 Isotopy We have already come across the concept of isotopy in Section 1.1. The concept was introduced into linguistics by Greimas in his Sémantique structurale (1966). Isotopy refers to the relationship between lexical items in a text that share certain semantic features ( meaning components or semes in the sense of componential analysis). By these shared semantic features, coherence is created. As we have seen in example text [1-1] in Section 1.1, there may be several isotopic strands in a text which hold it together. Isotopic strands in a text are mixtures of both coherenceand cohesion-creating devices: They are cohesion-creating devices because they may show at the text surface, e.g., when certain lexical items are repeated (e.g., bride - bridal ). In many cases, however, the semantic connection between elements in an isotopic strand can only be discovered through interpretation by text recipients. Thus isotopy can be considered to be a coherence-creating device. 1.2.2.2 Presuppositions The concept of presupposition has been introduced into text linguistics to explain the function of non-linguistic knowledge in the construction of text coherence. According to Linke, Nussbaumer & Portmann (1994: 231), basically two types of presuppositions can be distinguished: a) pragmatic presup- 12 Part I: Textlinguistic Foundations positions (“gebauchsgebundene Präsuppositionen”) and b) lexical presuppositions (“zeichengebundene Präsuppositionen”; see Fig. 1-5). Fig. 1-5: Types of presuppositions (adapted from Linke/ Nussbaumer/ Portmann 1994: 234) a) pragmatic presuppositions Let us take a look at how B’s answer to A’s request can be interpreted in example [1-15]: [1-15] A: I will not be able to attend next Wednesday’s text linguistics lecture. Can I copy your notes on Friday? B: I will have a wisdom tooth pulled next Tuesday. If we only take into consideration what is explicitly stated at the text surface, B’s statement cannot be interpreted as an answer to A’s request. A talks about a linguistics lecture that she cannot attend and asks for notes. B talks about a dental problem. Due to our knowledge of the world, however, we are able to interpret B’s statement as an answer to A’s request. We know that having a wisdom tooth pulled may cause serious pain, even the day after it was pulled. People who have had a wisdom tooth pulled will therefore probably not attend a lecture on the day after. If they do not attend the lecture, they cannot take notes and consequently cannot pass them on to somebody else who would like to copy them. All these pieces of knowledge that we need to interpret the answer correctly and that are not given explicitly in the text are called pragmatic presuppositions . They are called pragmatic , in contrast to lexical , because they are not included in the meaning of any of the lexical items used but can only be derived from the use that is made of the lexical items in a specific situation. Whenever we communicate, we make such pragmatic presuppositions. They are only verbalized in situations where we realize that our recipients have not understood. Their coherence-creating function consists in the fact that they allow us to establish a logical relationship even between sentences in a text that do not contain any syntactic or semantic elements that show their connection at the text surface. b) lexical presuppositions Lexical presuppositions can again be subclassified into two categories: existential presuppositions and semantic presuppositions. Existential presuppositions 1 Basic concepts 13 are triggered, for example, by definite articles or defining modifiers, as in the following example: [1-16] Her husband is very rich. Whoever makes this utterance in a real situation presupposes that she is married and that a man whom she married exists. In contrast to existential presuppositions, semantic presuppositions are based on semantic features of lexical items used, as in the following example: [1-17] He finally managed to get his paper finished on time. Here the verb to manage expresses that he was not only successful, but also that the task was difficult to accomplish, which is the semantic presupposition. Both types of presuppositions, existential presuppositions and semantic presuppositions, allow to us to convey additional meanings without expressing them explicitly. These additional meanings will still be conveyed when the original statements are negated; compare the examples above with their negated versions below: [1-16’] Her husband is not very rich. [1-17’] He finally did not manage to get his paper finished on time. Lexical presuppositions can be usurped to insinuate the truth of certain assumptions, for example, in advertisements, and then have a manipulating effect. Pragmatic presuppositions can be culture-specific and differ between laypersons and specialists in a domain. They may therefore cause communication problems when recipients are not able to infer the missing links between utterances. The term to infer (inference) refers to the logical reasoning that is necessary to reconstruct the presuppositions made by the speaker or writer. Are we always sure that the presuppositions we make in the texts we compose can be inferred by our readers? (Answering this question, of course, presupposes that an audience has been specified in advance.) In term papers, it is often the case that students do not state certain things explicitly just because they either lack motivation to verbalize them and hope that the reader can infer them or because they consider their professor to be the only reader, whom they assume to know anyhow (not always rightly so). 1.2.2.3 Frames and scripts The terms frames and scripts have been imported into linguistics from psychology. They refer to the way our knowledge is organized in our long-term memories. Our knowledge consists of individual concepts which are related to each other in an associative manner. In this way, knowledge clusters come about which either describe static conditions (e.g., what an apartment usually looks like) or dynamic processes (such as the actions involved when eating out in a restaurant, e.g., waiting to be seated, asking for the menu, choosing a meal, ordering, eating, paying the bill, etc.). The associative knowledge clusters that 14 Part I: Textlinguistic Foundations we have stored for static conditions are called frames , the associative knowledge clusters for recurrent dynamic processes are termed scripts . When we read a text, linguistic elements which occur in the text may invoke frames and scripts associated with them and thus invoke a picture in our minds that is much more complex than what could be conveyed by the individual linguistic elements in the text alone. These frames and scripts give rise to expectations of what has to come next in the text and thus help us to integrate the information we get from the text into a coherent framework in our minds. If these expectations are not met, i.e., if certain slots in the frames and scenes which we would like to fill with information from the text remain unfilled, we feel frustrated and text comprehension may suffer. By means of this frames-and-script theory, text linguists try to explain the interplay between information conveyed in the text (bottom-up processes) and knowledge clusters we have already stored in our long-term memories (top-down processes). This interplay between bottom-up and top-down processes is a prerequisite for text comprehension. Graphically it can be represented as illustrated in Fig. 1-6. Fig. 1-6: The interplay of bottom-up processes and top-down processes in text comprehension (adapted from Linke/ Nussbaumer/ Portmann 1994: 236) 1.2.2.4 Functional sentence perspective (FSP) When competent language users produce sentences, they do not only do so in accordance with the rules of (sentence) grammar, they also follow certain rules with regard to the distribution of known and new information over the sentence, the rules of communicative dynamism. Communicative dynamism refers to the arrangement of words and phrases in a sentence in such a manner that the information load they carry increases towards the end of the sentence. From a functional sentence perspective, sentences normally start with elements that carry a low amount of information. These elements are called the theme or topic. And they end with those elements whose amount of information is highest. They are called the rheme (from the Greek word for ‘new’), also known as focus or comment . 2 In such ‘normal’ cases, we speak of unmarked 2 For terminological distinctions that can be made between the concepts of ‘theme’ and ‘topic’ on the one hand and ‘rheme’ and ‘focus’ or ‘comment’ on the other hand, see Schnotz (2006). 1 Basic concepts 15 word order . The elements that carry a low amount of information usually refer to known information. This information may either be known because it has been mentioned earlier in the text or because it belongs to the recipient’s world knowledge. In certain situations, for example, when we want to contradict somebody, we deviate from the progression from theme to rheme and start with elements carrying a high amount of information. We then use a marked word order . These observations go back to the Prague structuralists Mathesius, Daneš and Firbas. If we compare German and English with regard to their word order, we will realize that word order in German is much more flexible than in English. English requires an SVO word order (subject-verb-object), whereas in German, we can also begin a sentence with an object, as in example [1-19]: [1-18] Der Hund jagt die Katze. (subject-verb-object  unmarked word order)  The dog is chasing the cat. [1-19] Die Katze jagt der Hund. (object-verb-subject  marked word order) ≠ The cat is chasing the dog.  The cat is being chased by the dog. (passive voice)  It’s the cat that the dog is chasing. (clefting construction) The reason for the relatively fixed word order in English is the fact that English has lost most of its inflectional morphology in the course of time, whereas modern German is still rich in inflectional morphology. In German, the inflectional endings indicate the case in which a noun is used ( der Hund as subject vs. den Hund as direct object). This is not possible in English; therefore, in English, word order had to take over the task of indicating, for example, what the subject and object is in a sentence. From a functional sentence perspective, however, it may be necessary to shift an element such as an object to the front of a sentence even in English. For this purpose, the English language needs auxiliary constructions such as the clefting construction in example [1-19] above. If I ask “What is the dog doing? ”, answer [1-18] above will be more natural than answer [1-19] because the information asked for (“is chasing the cat”) is the new information (the rheme) and therefore is expected in end position. If, however, somebody claims that the dog is chasing rabbits and I want to contradict, reaction [1-19] will be the more natural one because, with its contrast focus on the rhematic element “die Katze” and thus a marked word order, it emphasizes my objection. Competent writers intuitively follow the rules of functional sentence perspective. Inexperienced writers, however, deviate from these rules from time to time, even when writing in their mother tongue. This may have negative effects on text coherence and thus on text comprehensibility. The reason for deviating from the principles of functional sentence perspective may be that they have never been made aware of these principles. Word order from a func- 16 Part I: Textlinguistic Foundations tional perspective, in contrast to word order form a (sentence) grammatical perspective, is usually not covered in primary and secondary education curricula. It is noteworthy, however, that even inexperienced writers are sometimes able to realize that there is something awkward about a sentence which deviates from the principles of functional sentence perspective, even if they are unable to specify what exactly causes this awkwardness and how it can be remedied. Whereas the rheme in a sentence carries the new information and thus is most important from an informational perspective, the themes in sentences are free for sentence-connecting and thus coherence-creating functions. According to Daneš (1974), there are basically three types of how themes can be connected in texts, as well as an additional framework type. They are referred to as types of thematic progression and will be explained and illustrated in the following. 1. Simple linear progression In this type of thematic progression, the rheme of the preceding sentence is taken up as the theme of the following sentence, as in example [1-20], taken from Esser (2009: 33): [1-20] The first of the antibiotics was discovered by Sir Alexander Flemming in 1928. He was busy at the time investigating a certain species of germ which is responsible for boils and other troubles. Here, the rheme of the first sentence (Sir Alexander Flemming = R 1 ) is taken up as the theme of the second sentence (He = T 2 ) and connected with a new rheme (a certain species of germ = R 2 ), which then becomes the theme of the relative clause (which = T 3 ). Graphically this can be represented as follows: Fig. 1-7: Simple linear progression (Daneš 1974: 118) 2. Progression with a continuous theme In this type of thematic progression, a sequence of sentences is connected by sharing the same theme, as in example [1-21], taken from Esser (2009: 33) as well: [1-21] The Rousseauist especially feels an inner kinship with Prometheus and other Titans. He is fascinated by any form of insurgency … He must show an elementary energy in his explosion against the established order and at the same time a boundless sympathy for the victims of it … 1 Basic concepts 17 Graphically, progression with a continuous theme can be represented as follows: Fig. 1-8: Progression with a continuous theme (Daneš 1974: 118) 3. Progression with derived themes In this type of progression, there is a more general theme (hypertheme) from which the themes of the succeeding sentences are derived, as in example [1-22] quoted in Schubert (2008: 77): [1-22] Belize is a small Central American country located on the southeast of the Yucatan Peninsula. Its coastline on the Gulf of Honduras is approached through some 550 km (342 miles) of coral reefs and keys (cayo). The coastal area and north of the country are low-lying and swampy with dense forests inland. ( The Bloomsbury Pocket Encyclopedia of the World 1993: 194). Its population … Graphically, this type of thematic progression can be represented as shown in Fig. 1-9: Fig. 1-9: Progression with derived themes (Daneš 1974: 118) In texts, these types of thematic progression may occur in various combinations. Some of them can be regarded as types of progression of a higher order, i.e., as a type of framework into which other basic types can be integrated. According to Daneš (1974: 120), one of the most important of these framework types is progression with a split rheme. 4. Progression with split rheme In this type of progression, the rheme of one sentence is split up into several parts, which are taken up as themes in the sentences to follow. These sentences need not follow each other immediately; they may also be interrupted by other sentences which follow other types of thematic progression (Daneš 1974: 120). The following example, taken from Schubert (2008: 77), illustrates this type of progression: 18 Part I: Textlinguistic Foundations [1-23] The BBC radio network has five national stations: Radio One (98.8FM) is almost exclusively pop music, with a chart-biased view of the rock-world; Radio Two (89.1FM) is golden oldies and chat; […]. Graphically, this type of thematic progression can be represented as shown in Fig. 1-10: Fig. 1-10: Progression with split rheme (Daneš 1974: 119 with modifications made by myself) If we take a closer look at the examples above, we will realize that the selection of thematic elements is guided by principles of cognitive association. It is via cognitive association that sentences are connected to form a coherent whole (Blumenthal 1987: 46 ff.). These principles of association which date back to Aristotle, are: • similarity (with the special types inclusion and identity), • opposition and • contiguousness (from Latin contiguus - ‘neighbouring’). The following examples illustrate how these logical-semantic relationships are used to connect sentences: 1. Similarity (special types inclusion and identity) [1-24] There are different types of transmissions . Automatic transmissions are … [1-25] Fig. 3 shows an automatic transmission . It is composed of … In example [1-24], there is a logical relationship between the rheme of the first sentence and the theme of the following one. The theme of the second sentence takes up an aspect of the rheme of the first sentence. The relationship between them is one of inclusion and thus similarity. Example [1-25] illustrates a special type of similarity: the contents of the rheme of the first sentence is identical with that of the theme of the second sentence. 2. Opposition [1-26] Petrol engines operate at a pressure of … Diesel engines, however, need much higher operating pressures. 1 Basic concepts 19 In this example, the two sentences are connected logically by the fact that the theme of the second sentence is in opposition to the theme of the first sentence. 3. Contiguousness In sentences which are connected via a relationship of contiguousness, the theme of the second sentence has a spatial, temporal, causal or other relationship to the contents of the preceding sentence. In the following example, a pencil and a sheet of paper are in a relationship of contiguousness. [1-27] She could not find a pencil. But a sheet of paper was already lying in front of her. In all three cases of logical-semantic relationships, the beginning of the second sentence is connected to an element of the preceding sentence via cognitive association. Due to this association, it is not completely new and therefore easier to interpret. The associative relationships cause coherence (Blumenthal 1987: 46 ff.). Proceeding from what we know to what is new is important for text comprehensibility. 1.2.2.4.1 Auxiliary constructions used to achieve an unmarked word order In most cases, sentences with an unmarked word order begin with a thematic subject. In connection with certain verbs that are vague in meaning, such as to be , to appear and to exist , which express the mere existence or appearance of something, the subject may, however, become the element in a sentence that bears the highest amount of information and thus, from a communicative perspective, must be shifted to the end of the sentence (cf. Confais 1983: 255). To allow arranging a rhematic subject at the end of a sentence, some languages, especially languages with a rather fixed word order like English, use a number of auxiliary constructions. In the following, the most important auxiliary constructions will be presented illustrating their function for English, German and French. a) Grammatical subject as ‘dummy’ In these cases, the subject position at the beginning of the sentence is filled by a grammatical subject (a dummy subject) that allows moving the logical rhematic subject to the end of the sentence. Following Confais (1983: 252 ff.), who calls constructions that can take over the role of a grammatical subject “présentateurs”, these constructions could be called presenters in English. A typical presenter can be found at the beginning of fairy tales (Engl. once upon a time , German es war einmal ; French il était une fois ). Compare the following contrastive pairs: [1-28] Once upon a time there was a beautiful princess. (vs. A beautiful princess was once upon a time.) 20 Part I: Textlinguistic Foundations [1-29] Es war einmal eine schöne Prinzessin. (vs. Eine schöne Prinzessin war einmal.) [1-30] Il était une fois une jolie princesse. (vs. Un jolie princesse était und fois.) Another presenter in English is existential there ( there + a form of the verb to be and sometimes also other verbs; cf. Greenbaum/ Quirk 1999: 428). Typical French présentateurs are il y a and voilà ; a typical German presenter is es gibt . [1-31] There are two books on the shelf. (vs. Two books are on the shelf.) [1-32] Il y a une voiture. [1-33] Voilà ton argent. [1-34] In der Biotechnologie gibt es eine Reihe von Verfahren, mit denen … (vs. Eine Reihe von Verfahren, mit denen …, gibt es in der Biotechnologie.) Greenbaum & Quirk (1999: 423 ff.) call sentences which are initiated by means of existential there as a dummy subject “existential sentences”. According to them, this category also comprises sentences, such as the one in example [1- 35], in which it is not there which takes over the function of the grammatical subject but a nominal phrase followed by a form of the verb to have : [1-35] I have a brother working in Chicago. (vs. A brother of mine works in Chicago). b) Inversion Inversion refers to constructions in which the usual sequence of subject and predicate has been reversed. With regard to inversion, a distinction must be made between cases of inversion which are grammatically obligatory and cases in which it may occur for stylistic reasons. An inversion is obligatory in English and in German, for example, after fronted adverbs, as in the examples [1- 36] to [1-38]. In the French examples which follow, it is merely an option and usually restricted to written language use (Confais 1983: 251 ff.). [1-36] Herab kam eine alte Dame. (vs. Eine alte Dame kam herab.) [1-37] Here comes the bus. (vs. The bus comes here.) [1-38] Equally inexplicable was his behaviour towards his son. (vs. His behaviour towards his son was equally inexplicable.) [1-39] Au même instant arriva la police. (vs. Au même instant, la police arriva.) [1-40] Entre un page. (vs. Un page entre.) 1 Basic concepts 21 [1-41] l’interruption que provoqua la guerre … (vs. l’interruption que la guerre provoqua …) c) Passive voice The passive voice allows the roles of subject and object to be reversed. The subject of an active-voice sentence becomes a prepositional object in the corresponding passive-voice construction and is shifted after the verb or omitted, whereas the object of the active-voice sentence becomes the subject of the passive-voice construction and thus appears at the beginning of the sentence. [1-42] My sister makes these table mats. (active voice) [1-42’] These table mats are made by my sister. (passive voice) Example sentence [1-42] could be a natural answer to the question What does your sister do? , whereas sentence [1-42’] could answer the question Who made these table mats? d) Converses As example [1.42’] above showed, the passive voice may be useful in certain cases and have advantages over the active voice. Inexperienced writers, however, have an inclination to overuse it, which is detrimental to both the style and the comprehensibility of their texts. In many cases, passive-voice constructions can be avoided in a stylistically elegant manner by using converses, i.e., verbs and other formulations which carry more or less the same meaning but require a reversed order of the elements connected to them (cf. Greenbaum/ Quirk 1999: 415). The following examples taken from Greenbaum & Quirk (1999: 415) illustrate this: [1-43] An uncle, three cousins, and two brothers benefited from the will. [1-43’] The will benefited an uncle, three cousins, and two brothers. [1-44] An unidentified blue liquid was in the bottle. [1-44’] The bottle contained an unidentified blue liquid. [1-45] A red sports car was behind the bus. [1-45’] The bus was in front of a red sports car. d) Extraposition Another auxiliary construction that is used to achieve end focus is extraposition. Extraposition refers to a postponement in which the postponed element is replaced by a substitute. Greenbaum & Quirk (1999: 417) specify postponement as follows: “It [sc. postponement] operates almost exclusively on subordinate nominal clauses. The most important type of extraposition is that of a subject realized by a finite or nonfinite clause. The subject is moved to the end of the sentence, and the nominal subject position is filled by the anticipatory pronoun it . The resulting sentence thus 22 Part I: Textlinguistic Foundations contains two subjects, which we may identify as the POSTPONED SUBJECT (the one which is notionally the subject of the sentence) and the ANTICIPATORY SUBJECT (it).” Greenbaum & Quirk (1999: 41) illustrate extraposition using the following examples: [1-46] To hear him say that surprised me. [1-46’] It surprised me to hear him say that. [1-47] To teach her is a pleasure. [1-47’] It is a please to teach her. [1-48] That income tax is to be lowered was on the news. [1-48’] It was on the news that income tax is to be lowered . [1-49] What you do doesn’t matter. [1-49’] It doesn’t matter what you do. [1-50] To see others enjoying themselves makes her happy. [1-50’] It makes her happy to see others enjoying themselves. [1-51] *That she wanted to go into politics is said. [1-51’] It is said that she wanted to go into politics. [1-52] *For anyone to escape was considered impossible. [1-52’] It was considered impossible for anyone to escape. Apart from extraposition of a clausal subject, extraposition of a clausal object also occurs, as in the following examples, also taken from Greenbaum & Quirk (1999: 418): [1-53] You must find working here exciting. [1-53’] You must find it exciting working here. The examples above illustrate that extraposition serves both the principle of end focus and the principle of end weight, a principle according to which long sentence components should appear at the end of the sentence. 1.1.2.4.2 Auxiliary constructions used to achieve a marked word order In the previous section, auxiliary constructions were presented by means of which an unmarked word order can be achieved. This section will deal with auxiliary constructions used to attain a marked word order. In spoken language, we can simply emphasize a sentence component and thus make it the rheme of the sentence by placing stress on it. Compare the following examples: [1-54] He gáve her the book. (i.e., he did not just lend it to her) [1-54’] He gave her a bóok. (i.e., not a CD) These examples illustrate at the same time that meaning is determined by the context. 1 Basic concepts 23 In languages with a relatively free word order, such as German, we can simply place emphasis on certain words, and thus signal their rhematic function, by placing them at the front of the sentence, as in the following German example: [1-55] Ein Búch schenkte er ihr. In languages with a relatively fixed word order, such as English, this is not possible. In these languages, auxiliary constructions are needed in order to be able to place into front position what, in sentences with an unmarked word order, would appear after the verb. In the following, two such auxiliary constructions, clefting constructions and pseudo-clefting constructions , will be presented. a) Clefting constructions These constructions are called clefting constructions, also cleft sentences ( cleft - ‘partially split or divided’), because they result from splitting off a part from a sentence whereby the part split off is extended to form a full sentence itself without the sentence from which it was split off losing its grammatical completeness (cf. Greenbaum/ Quirk 1999: 412 ff.). [1-56] I gave him a book. [1-56’] (1) It’s a book (2) (that) I gave him. Clefting constructions are used especially in contradictions. The part split off (in example [1-56’] a book ), which in English is enclosed in it is … (who/ that) and in French in c’est … qui/ que , thereby obtains special emphasis and expresses a contradiction (contrast focus; highest objective communicative value). Due to this function, it is called mise en relief in French. Did you give him a CD? [1-57] No, it’s a book that I gave him. [1-57’] Non, c’est un live que je lui ai donné. Pierre t’a raconté cette histoire? [1-58] Non, c’est Paul qui me l’a racontée. In German, clefting constructions can also be found, especially in spoken language. Generally, however, they occur less frequently than in English. [1-59] Es war Max, dem sie eine klebte. [1-60] Es war der Taxifahrer , der den Unfall verursachte. In German, clefting constructions are not needed unless the part of the sentence to be emphasized is the subject of the sentence, as in example [1-60]. In example [1-59], Max can also be emphasized by simply placing it in front position, as in example [1-59’]: [1-59’] (Dem) Max klebte sie eine. If, however, the part of the sentence to be emphasized is the subject, as in example [1-60], clefting is necessary to indicate the contradiction. In such cases 24 Part I: Textlinguistic Foundations it is not possible to achieve a marked word order without clefting (compare example [1-60] with example [1-60’]): [1-60’] Der Taxifahrer verursachte den Unfall. b) Pseudo-clefting constructions A pseudo-cleft sentence is “essentially an SVC [subject-verb-complement] sentence with a nominal clause as subject or complement” (Greenbaum/ Quirk 1999: 414). Pseudo refers to the fact that the result of pseudo-clefting does not consist in two complete clauses each of which possessing its own subject and verb, as in the case of clefting constructions, but involves only a partial split in which the main clause uses the pseudo-split off part as its subject or object. Compare the following two variants: [1-61] It’s a book I gave him. (clefting construction). [1-61’] What I gave him is a book . (pseudo-clefting construction) In example [1-61’], the pseudo-cleft part What I gave him serves as the subject of the main clause, which would be grammatically incomplete without it. As example [1-61’] shows, pseudo-clefting constructions allow to indicate unmistakably what the rhematic element is. In contrast to clefting constructions, pseudo-clefting constructions can also be used to emphasize verb phrases, as in example [1-62]: [1-62] What Peter did was (to) call his boss a liar. In such cases, lexical means are frequently used in German, such as kannst du dir das vorstellen? , sogar , etc. Imitating the English pseudo-clefting construction in German would lead to an unidiomatic result (cf. example [1-62’]): [1-62’] Was Peter tat, war seinen Chef einen Lügner zu nennen. In German, it would be acceptable to simply place the rheme in front position. In addition, lexical means could be used, as in example [1-62’’]: [1-62’’]Einen Lügner nannte Peter seinen Chef./ …, kannst du dir das vorstellen? Emphasizing rhematic elements is not the only function that pseudo-clefting sentences can fulfil. In English, they are also used as “parsing aids” (Doherty 2000: 53), i.e., as a means to facilitate syntactical processing for the reader. This is important in cases, in which without the use of pseudo-clefting constructions, the reader might get on a ‘garden path’. Garden path sentences are sentence constructions which, when processed by readers, may induce them to a grammatical interpretation which turns out to be wrong once they proceed to the end of the sentence, and then has to be corrected. Doherty (2000) illustrates this by means of the following two examples taken from the Austrian playwright Thomas Bernhard and their translations into English. 1 Basic concepts 25 [1-63] Das größte Unglück des Menschen, daß seine Zeit immer und in jedem Fall zu kurz ist, hat die Erkenntnis immer unmöglich gemacht. (Thomas Bernhard Auslöschung : 125 f.; quoted in Doherty 2000: 52) [1-64] Wenn wir denken und nicht aufhören zu denken, was wir philosophieren nennen, kommen wir schließlich darauf, daß wir falsch gedacht haben. (Thomas Bernhard Auslöschung : 126; quoted in Doherty 2000: 51) None of the two German sentences contains a pseudo-clefting construction and none of them needs one; in their English translations, however, pseudoclefting constructions become indispensable for reasons of readability. The following three versions illustrate this: [1-63a] Man’s greatest misfortune that he never has enough time is what has always made knowledge impossible. (Thomas Bernhard Extinction : 80; quoted in Doherty 2000: 52) [1-63b] Man’s greatest misfortune that he has never enough time has always made knowledge impossible. (Doherty 2000: 52) [1-63c] Man’s greatest misfortune is that he never has enough time and that has always made knowledge impossible. (Doherty 2000: 52) In contrast to version [1-63a] and like the German original, versions [1-63b] and [1-63c] do not use pseudo-clefting. Version [1-63b], the one closest to the German original, is hard to understand (which does not apply to the German original). In version [1-63c], this was avoided to some extent by using the copula verb be to connect the original subject clause that he never has enough time to the subject of the sentence Man’s greatest misfortune . This made it necessary to connect what was originally a predicate phrase (i.e., has always made knowledge impossible ) to the demonstrative pronoun that as its subject. Following the principle of parallelism, however, this second that as a demonstrative pronoun may be misinterpreted as a conjunction upon first reading it, thus causing a garden-path effect. By using a pseudo-clefting construction, this effect can be avoided. Now, let us consider translations of example [1-64]: [1-64a] When we are thinking and don’t stop thinking, which is what we call philosophizing, we come to realize that our thinking has been wrong. (Thomas Bernhard Extinction : 80; quoted in Doherty 2000: 51) [1-64b] When we are thinking and don’t stop thinking, which we call philosophizing, we come to realize that our thinking has been wrong. (Thomas Bernhard Extinction : 80; quoted in Doherty 2000: 51) In version [1-64b], there is the danger that, upon first reading, the relative pronoun which is misinterpreted as an interrogative pronoun. This can be avoided again by using a pseudo-clefting construction as in example [1-64a]. Whether pseudo-clefting constructions are used to emphasize rhematic elements or to avoid syntactic misinterpretations, in both cases they facilitate text 26 Part I: Textlinguistic Foundations processing as “parsing aids” (Doherty 2000: 53) and thus should be used consciously as a means to obtain a high degree of text comprehensibility. 1.2.3 The relationship of cohesion and coherence Texts can be coherent without being cohesive (see example [ 1-65 ] below) but if they are cohesive without being coherent (see example [ 1-66 ] below), we will certainly not consider them to be proper texts. [ 1-65 ] This bag contains: 5 bolts 5 washers 5 nuts 10 plastic plugs [ 1-66 ] An old man was standing at the entrance of the cinema. This cinema is a nuclear submarine. The submarine studied literature. This subject is very popular with bluebells. They will stage another demonstration some time. The demonstration lasted for hours. (Lux 1981: 342; ex. 4) This shows that cohesion in contrast to coherence is not a necessary textuality criterion. 1.2.4 Intentionality and acceptability Whereas coherence and cohesion are text-centred notions, intentionality and acceptability are user-centred notions. According to de Beaugrande & Dressler (1981: 7), intentionality concerns “the text producer’s attitude that the set of occurrences should constitute a cohesive and coherent text instrumental in fulfilling the producer’s intention, e.g. to distribute knowledge or to attain a GOAL specified in a PLAN .” Acceptability is the counterpart of intentionality as seen from the receiver’s perspective. According to de Beaugrande & Dressler (1981: 7), it concerns “the text receiver’s attitude that the set of occurrences should constitute a cohesive and coherent text having some use or relevance for the receiver, e.g. to acquire knowledge or provide co-operation in a plan.” If recipients are able to understand for what purpose authors composed their texts, they are in many cases tolerant towards a potential lack in texts’ cohesion and coherence. Therefore, intentionality and acceptability seem to be more relevant for distinguishing between texts and non-texts than cohesion and coherence. 1.2.5 Informativity Informativity “concerns the extent to which the occurrences of the presented text are expected vs. unexpected or known vs. unknown/ certain” (de Beau- 1 Basic concepts 27 grande/ Dressler 1981: 8 f.). De Beaugrande & Dressler (1981: 8 f.) illustrate this standard using the following examples: [ 1-67 ] Call us before you dig. You may not be able to afterwards. (warning by the Bell Telephone Company) [ 1-67’ ] Call us before you dig. There might be an underground cable. If you break the cable, you won’t have phone service, and you may get a severe electric shock. Then you won’t be able to call us. Example [ 1-67 ] is much more informative - and motivating - than example [ 1-67’ ] although it leaves much more implicit and thus is cognitively more demanding. De Beaugrande & Dressler (1981: 9) describe the relationship between implicitness and motivation as follows: “The processing of highly informative occurrences is more demanding than otherwise, but correspondingly more interesting as well. Caution must be exercised lest the receivers’ processing become overloaded to the point of endangering communication.” 1.2.6 Situationality Situationality “concerns the factors which make a text RELEVANT to a SITUA- TION of occurrence.” (de Beaugrande/ Dressler 1981: 9). de Beaugrande/ Dressler (1981: 9) illustrate this standard of textuality using the following example, a text on a road sign: [ 1-68 ] SLOW CHILDREN AT PLAY Theoretically, this text can be interpreted in different ways. It could mean “Mentally retarded children at play.” or “Please drive slowly because children might be playing in the street.” The location in which the sign is put up helps us in interpreting the text in the intended manner; its situationality makes it interpretable. Another example are the letters “P” and “S” on pepper and salt shakers. If we read the letters “P” or “S” anywhere on a wall, we might not be able to interpret what they are standing for there; in that case, they are not “texts” for us. If they appear on a pepper or salt shaker, however, we interpret them as “This shaker contains pepper/ salt.” and thus as texts, although they only consist of one letter. Transforming them into whole sentences would not make the message clearer but less economic. 1.2.7 Intertextuality Intertextuality is defined by de Beaugrande & Dressler (1981: 10) as the standard of textuality that “concerns the factors which make the utilization of one text dependent upon knowledge of one or more previously encountered texts”. Intertextuality may be of different types. An intertextual relationship may just 28 Part I: Textlinguistic Foundations be one between a quotation in a text and the text this quotation is taken from. Another type of intertextuality is the relationship between a text and all other texts of the same genre. And a third type of intertextuality concerns the relationship between a text and other texts which are somehow derived from it, such as the fairly tale Little Red Riding Hood and all the parodies of it, or a source text and its translation(s). 1.2.8 Communicative quality of texts The examples discussed in the previous sections have shown that we might consider a semiotic unit a text even if it does not fulfil all seven standards of textuality at the same time. In contrast to what de Beaugrande & Dressler state, they are not all necessary features although they characterize what we would consider a prototypical text. In this section, we will take into consideration three properties of texts that are not constitutive of the text concept but have an effect on the quality of texts. Using terminology introduced by Searle (1969: 33 f.), de Beaugrande & Dressler (1981: 11) refer to their seven standards of textuality as “ CONSTITUTIVE PRINCIPLES […] of textual communication” and call the three properties that will be dealt with in this section as “regulative principles […] that control textual communication rather than define it”. These three non-constitutive text properties are efficiency, effectiveness and appropriateness. They are defined as follows by de Beaugrande/ Dressler (1981: 11): “The EFFICIENCY of a text depends on its use in communicating with a minimum expenditure of effort by the participants. The EFFECTIVENESS of a text depends on its leaving a strong impression and creating favourable conditions for attaining a goal. The APPROPRIATENESS of a text is the agreement between its setting and the ways in which the standards of textuality are upheld.” Choosing the wrong register, for example, leads to inappropriateness. Communicating something in a very long text that could be conveyed in a much shorter way leads to inefficiency. However, if something is not explained with the degree of explicitness that the target group needs, i.e., if too much is left implicit, this will lead to a lack of effectiveness because the readership will then not be able to understand everything and fail to react in the manner intended by the author. The examples above have illustrated that whether we consider a semiotic unit a text or not depends on whether we are able to make sense of it and interpret it. This interpretation cannot be explained within the confines of linguistics alone. It also involves insights from cognitive psychology, computer science and other related fields (cf. de Beaugrande/ Dressler 1981: 12). These insights will be the topic of Part II. Part II Text Comprehension and Text Comprehensibility 2 Levels of cognitive processing What happens on the way from the perception of a text to the mental pictures and movies it invokes in our minds? What factors have an influence on the ease or difficulty involved in making sense of a text? When answering these questions, we have to take into account at least three levels of cognitive processing. The first level is that of legibility. At this level, we must be able to decipher a text. If we have problems in doing so, for example, because of poor handwriting, our comprehension process cannot even be initiated because what needs to be interpreted cannot be ‘transported’ to the cognitive systems where further processing could take place. The next level is the level of readability. At this level, we must be able to understand the words a text is composed of and their grammatical relationships. It is only after having mastered this second level that we are able to proceed to the level of comprehensibility, at which our prior knowledge plays an important role. In practice, we do not pass through these three levels one after the other, the processes at the three levels interact. Processing steps at higher levels may allow us to overcome problems that we encounter at lower levels. If, for example, we cannot decipher a specific word, comprehension of the text around it may help us to guess what the word should be. Each of the three levels will be looked into in more detail in the sections to follow. 2.1 The legibility of texts The term legibility refers to the extent to which the typographical and graphical design of a text contributes to the perceptibility and discernibility of its components and the ease with which they can be recognized. The typographical and graphical design of printed texts and its influence on text reception are the object of legibility research. It is considered the predecessor, but also a component, of readability research (Groeben 1982: 174). Factors which have an influence on legibility are the fonts used, the font size, the typeface, the length of lines, line breaks, spacing between words, the contrast between the text and its background as well as the quality of print. Reading experiments have shown that the fonts which are used most frequently and with which readers are therefore confronted in most documents are also the ones which are most legible. Among them are Times New Roman and Helvetica (Kösler 1992: 157, 160). Italics are harder to read than non-italicized text. Capitalizing whole words (apart from brand names) should be avoided because capitalized letters look more similar and are therefore harder to discern than small letters. 32 Part II: Text Comprehension and Text Comprehensibility 2.2 The readability of texts Readability research, a field of investigation that evolved in the 1930s, focuses on the linguistic and stylistic quality of texts and their optimization (cf. Groeben/ Christmann 1989: 166 f.). It has yielded so-called readability formulas. The main factors that these formulas take into account are values for word and sentence difficulty, such as the average word length, the number of syllables per 100 words and the average number of words per sentence, but in some formulas also the average frequency with which the words used in a text occur in reference corpora. As these examples show, readability formulas only take into account quantitative features of texts. One of the best-known readability formulas is the Reading Ease formula by Flesch (1948; cf. Klare 1963: 23). RE (Reading Ease) = 206.835 - 0.846 wl - 1.015 sl wl = number of syllables per 100 words sl = average number of words per sentence Fig. 2-1: The Reading Ease formula by Flesch This formula was developed for the English language and needs adaptation if it is to be applied to other languages. The reason for this is that, on average, there are more monosyllabic words in English than, for example, in German. For English texts, the RE value ranges between 0 and 100, 0 meaning very poor readability and 100, optimal readability (cf. Klare 1963: 58, 78; Groeben 1982: 176 f.). How were these formulas developed? Text features which can be determined in an objective manner, such as the average word and sentence lengths mentioned above, were set in relation to criteria such as reading speed, experts’ ratings of text difficulty or the number of hits in cloze procedures (cf. Groeben 1982: 176). Readability formulas are popular because they can easily and quickly be applied by means of computer programs. However, they do not give us deeper insight into the comprehensibility of the texts to which they are applied because they take into account only certain lexical, syntactical, and stylistic aspects of what makes a text comprehensible or incomprehensible. They do not provide any information on whether the depth with which something is described in a text is adequate for its audience and whether the information is provided in a suitable order. They also do not consider information provided in nonverbal representations, which, however, may be decisive for text comprehensibility. It may be a necessary criterion for the quality of texts that they achieve acceptable results when a readability formula is applied to them, but it definitely is not a sufficient one. This becomes obvious already when we look 2 Levels of cognitive processing 33 at the result obtained by the application of a readability formula to a text whose sentences were rearranged in random order. It obtains the same result as the original text although the rearrangement of sentences may have made it completely incomprehensible. Another disadvantage of readability formulas is that they provide no indication of how to improve a text which has obtained a bad result. 2.3 Text comprehension as a constructive process In the early days of readability research, the text reader was more or less modelled as a passive recipient of information conveyed in a text, who only had to extract meaning from word forms given in the text as one would take things out of boxes. Thus text comprehension was considered to be a process that was controlled exclusively by data given in the text itself. Today, there is unanimity that such bottom-up processes only result in comprehension if they trigger appropriate top-down processes, i.e., associative processes in which the reader brings in his or her prior knowledge stored in their long-term memories. From this perspective, texts do not possess meaning; text meaning has to be constructed actively by the reader in an effort in which bottom-up processes and top-down processes interact. Since one and the same source text triggers different associations in each recipient because each recipient possesses a different basis of prior knowledge, the meaning of the same text constructed by different readers will never be the same. Nevertheless, people with a similar cultural background share large amounts of knowledge so that there is much overlap in their different interpretations of a text. It is this overlap via which communication comes about rather than the differences, which may give rise to misunderstandings. The more recent theories of text comprehension which focus on the active role of the reader in constructing text meaning are called cognitive constructivism (see, for example, Bransford/ McCarrell 1974; Hörmann 1976: 467 ff.; Hanstein 1993; the overview in Rickheit/ Strohner 1993: 78 f.; as well as Bartlett 1932 whose ideas originally did not receive much attention while the behaviourist paradigm was still prevailing). According to these theories, information which readers find in what is given in a text itself in bottom-up processes interacts with the knowledge stored in their long-term memories. The knowledge from their long-term memories is activated, i.e. made conscious, by what is taken up in bottom-up processes but also by their specific interests and reading strategies and is thus made available for meaning-creating processes. Groeben therefore refers to comprehension processes as “reader-text interaction”. This reader-text interaction can be illustrated using an example already referred to in Section 1.2.2.2: 34 Part II: Text Comprehension and Text Comprehensibility Two students are having the following conversation: [2-1] A: I will not be able to attend next Tuesday’s text linguistics lecture. Can I copy your notes on Friday? B: I will have a wisdom tooth pulled next Monday. If we only take into consideration what is explicitly stated at the text surface, i.e. contributed to text comprehension in bottom-up processes, B’s statement cannot be interpreted as an answer to A’s request, as explained in Section 1.2.2.2. Due to our knowledge of the world, which we bring in in top-down processes, however, we are able to interpret B’s statement as an answer to A’s request. All the pieces of knowledge that we need to interpret the answer correctly and that are not given explicitly in the text but implied somehow show the relevance of top-down processes for text comprehension. B did not mention them explicitly because he expected A to be able to infer them. We call them pragmatic presuppositions (cf. Section 1.2.2.2). Whenever we communicate, we make such pragmatic presuppositions. As pointed out in Section 1.2.2.2, they are only verbalized in situations where we realize that our recipients have not understood. Rickheit & Kock (1983) demonstrated the relevance of top-down processes for information processing in an experiment in which they had a group of students read texts which described scenes from every-day student life, e.g., getting enrolled, going to the university restaurant. After the students had read the texts the latter were taken away from them again and the students had to describe in as much detail as possible what the texts said. The reports of all students contained details that had not been mentioned in the original texts but formed part of their personal experiences. This suggests that, in the phase of reproduction, the students could no longer make a distinction between what they had taken up in bottom-up processes while reading the texts and what they had brought into text comprehension via top-down processes. Similar experiments had already been conducted by the psychologist Frederic Charles Bartlett in 1932 when he developed his psychological theory of the human mind. He was interested in finding out how the knowledge that we possess and that we can bring in into our text comprehension in top-down processes is partitioned, structured and organized in our brains and how we learn, i.e., add new knowledge to what we already know. In this connection, he coined the term schema. It refers to structured knowledge clusters in our longterm memories, which we will come back to in Section 3.6. 3 Text-processing from a cognitive-science perspective In their attempts to model text comprehension, cognitive scientists started out from very small units of text meaning and their processing, the so-called propositions . In the course of time, their focus widened to include larger processing units. With this widening of focus, the recipients’ prior knowledge was increasingly taken into account (cf. Christmann 1989: 86). The most important approaches towards explaining text comprehension in the cognitive sciences are, in chronological order, propositional models, network models, semantic macro-structures, schema-theoretical models and mental models. They will be explained in the following sections. 3.1 Propositional models of text processing The aim of the early models of text processing from the cognitive sciences was to describe the semantic structure of texts as precisely and objectively as possible and to use these descriptions to make predictions concerning the extent to which a text’s content can be recalled (cf. Christmann 1989: 61). What did these descriptions of semantic text structure look like? The first model of this category is Walter Kintsch’s (1974) proposition model. According to this model, a text can be subdivided into an ordered list of propositions called text base . Each proposition is composed of a so-called predicator and one or more so-called arguments . In proposition models, the term predicator refers to a concept that specifies another concept or the relation between other concepts. These other concepts are the arguments . The following example provides an illustration: [ 3-1 ] The mechanic repairs the broken parts. Following Kintsch et al. (1975), the text base of this sentence can be represented as follows: [ 3-1’ ] 1 (REPAIR, MECHANIC, PARTS) 2 (BROKEN PART) REPAIR and BROKEN represent predicators; MECHANIC and PARTS represent arguments of REPAIR, and PARTS the argument of BROKEN. Capitalization is used to make a distinction between the words and the concepts they stand for; the capitalized words represent concepts. As the text base in example [ 3-1’ ] shows, sentence [ 3-1 ] can be subdivided into two propositions. These are intended to represent textual meaning in an objective manner (Kintsch 1974: 11). 36 Part II: Text Comprehension and Text Comprehensibility In addition to rules according to which texts can be subdivided into individual propositions, Kintsch’s proposition model also includes rules that define how the individual propositions can be organized in a hierarchical manner in such a way that the connections between them become obvious. And the model also comprises empirically verifiable assumptions concerning their cognitive processability (cf. Christmann 1989: 50). Kintsch (1974) assumes that the individual propositions must form a connected (coherent) whole if they constitute a real text. The connection between individual propositions is either established via recurrent arguments, as in example [ 3-1’ ] , in which the argument PARTS occurs in both propositions and thus constitutes a bridging element, or via a whole proposition that is taken up in a subsequent proposition as an argument, as in example [ 3-2’ ] below. Propositions which are connected either via a recurrent argument (argument overlap) or via the integration of one proposition into another (argument embedding) are considered to be coherent. [ 3-2 ] The Greeks loved beautiful art. When the Romans conquered the Greeks, they copied them, and, thus, learned to create beautiful art. (Kintsch et al. 1975: 198) Following Kitsch et al. again, this text’s text base can be represented as follows: [ 3-2’ ] 1 (LOVE, GREEK, ART) 2 (BEAUTIFUL, ART) 3 (CONQUER, ROMAN, GREEK) 4 (COPY, ROMAN, GREEK) 5 (WHEN, 3, 4) 6 (LEARN, ROMAN, 8) 7 (CONSEQUENCE, 3, 6) 8 (CREATE, ROMAN, 2) In propositions 5, 6, 7 and 8, other propositions, indicated by their numbers, are resumed. What is important for their resumption is not that identical expressions are used but rather that the same concepts are referred to (referential or denotative identity). Indentations in the representation in example [ 3-2’ ] indicate hierarchical relationships. For the hierarchical arrangement of propositions, the following rules apply: 1. A proposition is superordinated to another proposition if it contains either an argument that also occurs in the other proposition following it or if it is embedded in one of the following propositions in its entirety. 2. If a proposition can be subordinated to several propositions preceding it, it will be subordinated to the one at the highest level in the hierarchy. 3. Propositions which contain new arguments only are placed at the highest hierarchy position. 3 Text-processing from a cognitive-science perspective 37 Taking text bases as a point of departure, cognitive psychologists asked the following questions: 1. Does propositional density, i.e., the number of propositions per a certain number of words in a natural text, have an effect on text processing? 2. Are propositions which occur at higher hierarchy levels processed and recalled better than propositions subordinated to them? 3. Does the number of connections a proposition has with other propositions (i.e. its relational density) have an influence on text processing and recall? 4. Are propositions which are connected to other propositions via argument overlap processed more profoundly than propositions which have fewer or no connections? Empirical investigations yielded results which allow all these questions to be answered with “yes”. Kintsch & Keenan (1973) showed that reading time increases with propositional density. Kintsch (1974: 138), Graesser (1978: 85) and Manelis (1980: 55) showed that propositions with a high relational density, i.e. with many connections to other propositions, are recalled better than propositions with a low relational density. Furthermore, it could be shown that propositions that are connected with other propositions via argument overlap or embedding are read faster and recalled better than propositions that are not connected. Superordinated propositions are recalled better and retained for a longer time than propositions at lower hierarchy positions (cf., for example, Kintsch et al. 1975; Meyer 1975: 107 ff.). From these results the following conclusions can be drawn for the production of comprehensible texts that can easily be recalled: 1. Propositional density must be adapted to the reader’s prior knowledge. The less prior knowledge the reader has concerning the topic of a text, the lower its propositional density should be. 2. Important information should be conveyed in positions high up in the hierarchy. These are usually paragraph initiating sentences. 3. Units of text which convey important information should be resumed frequently in a text to increase relational density. 4. Sentences should be connected semantically. 3.2 The structure of the human mind Comprehension of the model of cyclic processing, which will be explained in the next section, and understanding why long and complex sentences may be hard to understand necessitates a basic understanding of how the human 38 Part II: Text Comprehension and Text Comprehensibility memory works. In what follows, a multi-storage model of the human memory will be presented. Fig. 3-1: Multi-storage model of the human memory (cf., e.g., Arbinger 1984) In our environment, we are exposed to numerous stimuli. Only a selection of them enters our ultra-short-term memory. It can store information for only a very limited span of time which does not exceed approximately one third of a second. After that time span, most stimuli which entered our ultra-short-term memories will be deleted again. Only few of them, on which we focus our attention, are able to pass a filter, which, in Fig. 3-1, is represented by a broken line, and thereby enter short-term or working memory. The ultra-short-term memory is a sensory storage, i.e., it does not transform stimuli which we have perceived with one of our senses (i.e. our ears) into other representations, i.e., a visual one. Our short-term memories have a very limited storage capacity. According to Miller (1956), they can retain only about 7 ± 2 units of information. How complex these units of information are depends on our prior knowledge. The seven units may be letters, words or even larger units such as poems, if they have been stored as complex associative units (chunks) in our long-term 3 Text-processing from a cognitive-science perspective 39 memories. Thus, we may have difficulty to recall seven nonsense words such as aka , bali , duly , which we must compose of individual syllables, whereas we may well be able to recall seven real words because they are stored as readymade chunks in our long-term memories. The formation of complex units of meaning from smaller units and the grouping of units into categories on the basis of knowledge that we have stored in our long-term memories is called chunking (Miller 1956). When constructing sentences, we must take into account this limited storage capacity of our short-term memories. Complex sentences may exceed our audience’s short-term memory storage capacity and then cannot be processed by them in a single pass. This, however, need not necessarily be the case. If our readers possess a large amount of prior knowledge, their top-down processes may help them to fill in the gaps that occur when short-term memory overload occurs and processing would not be possible in a single pass. Writers’ extensive prior knowledge is frequently the reason why they presuppose too much in their texts. It disables them to suppress their own top-down processes when checking their texts for comprehensibility. Only stimuli which enter our short-term memories are raised to our consciousness. Even the knowledge stored in our long-term memories remains unconscious as long as it is not downloaded into our short-term memories. Imagine being permanently aware of the entire knowledge stored in your longterm memory. That would be awful. What triggers download of knowledge from our long-term memories into our short-term memories (top-down processes) is either information that we have taken up via our senses in bottom-up processes or our specific interest or focus of attention at the point of time in question. We are not only able to download knowledge from our long-term memories into our short-term memories exactly in the manner in which it was stored there. We can also ‘work’ with this knowledge in order to solve problems. For this purpose, we must be able to ‘play’ with it, to recombine knowledge and to make inferences from what we already know thereby creating new knowledge. Such knowledge transforming operations take place in our working memories, or, in models which do not make a distinction between working memory and short-term memory, in our short-term memories. Fig. 3-1 illustrates what the terms bottom-up processes and top-down processes refer to. Bottom-up processes are processes in which information is taken up from the bottom, i.e. the text surface, and then passed on level by level to the higher processing systems. Processes in which knowledge from our long-term memories is passed down step by step to lower processing levels are called top-down processes . 40 Part II: Text Comprehension and Text Comprehensibility 3.3 The model of cyclic processing The proposition model described in Section 3.1 focuses on the semantic structure that is assumed to be created during text comprehension but does not make any statements about the processes in which this semantic structure is created. This aspect is focused on in the model of cyclic processing developed by Kintsch & van Dijk (1978) on the basis of Kintsch’s earlier propositional model. The model of cyclic processing focuses on how recipients actually construct a text base. In so doing, it takes into account the recipients’ prior knowledge and the goals they pursue in reading a specific text, but also the limitations of their working memories, which have been introduced in the previous section. The model of cyclic processing starts from the “assumption that checking a text base for referential coherence and the addition of inferences wherever necessary cannot be performed on the text base as a whole because of the capacity limitations of working memory” (Kintsch/ van Dijk 1978: 368). Text processing is therefore assumed to be performed in cycles. In each cycle, a sequence of propositions is processed. The number of these propositions is assumed to vary depending on factors such as the surface characteristics of the text and the reader’s prior knowledge. Kintsch & van Dijk (1978: 368) further assume that, if “text bases are processed in cycles, it becomes necessary to make some provision in the model for connecting each chunk to the ones already processed”. Connection of chunks is modelled as follows: Part of working memory is a short-term memory buffer of limited size. When a chunk of propositions is processed, a selection of them is stored in the short-term buffer. Only these selected propositions in the short-term buffer are available in the subsequent cycle for connecting the new incoming chunk of propositions with the already processed one. If there exists some argument overlap or argument embedding between the new propositions and the ones retained in the buffer, the input is accepted as coherent with the previous text. If this is not the case, a cognitively demanding search of all previously processed propositions in long-term memory is made. This process is called reinstatement . If a proposition is found in this manner that shares an argument with at least one proposition in the input set, the set is accepted and processing continues. If not, inference processes are initiated which add propositions to the text base that connect the input set to the already processed propositions. Reinstatements and inferences, which involve long-term memory searches, are assumed to make relatively heavy demands on the recipients’ cognitive resources and thus make the text hard to understand (Kintsch/ van Dijk 1978: 368). Once coherence has been created, a cycle is completed. After each cycle, the coherent propositional structure is transferred from working memory into longterm memory with a selection of propositions retained in the short-term memory buffer for coherence creation in the next cycle. Which propositions in each cycle are selected for retention in the short-term buffer is hard to deter- 3 Text-processing from a cognitive-science perspective 41 mine empirically. Kintsch & van Dijk (1978: 369) assume, however, that propositions in high hierarchy levels are selected preferentially and that, in addition, recency also plays a role. This principle of selection is called leadingedge strategy (Kintsch/ van Dijk 1978: 379). If neither reinstatement nor inferences lead to a coherent text base, either the whole propositional structure has to be reorganized or the new chunk of propositions has to be stored in longterm memory as an incoherent chunk (Kintsch/ van Dijk 1978: 369; cf. Christmann 1989: 61 f.). This model, too, was used to make predictions of text difficulty and recall of propositions, which were tested empirically (cf. the overview in Christmann 1989: 65 ff.). Text difficulty was assumed to increase with the number of reinstatements, inferences and reorganizations necessary to create coherence. Recall of a proposition was furthermore assumed to be dependent on the number of cycles in which it was selected for transfer into the short-term memory buffer. These hypotheses could be confirmed although the amount of empirical data on which the respective conclusions were based was rather small (cf. Christmann 1989: 65; Groeben 1982: 53). Furthermore, Collins, Brown & Larkin (1980) could show empirically that propositional models fail to predict text comprehension in cases where propositions can be transformed into coherent text bases even without inferencing, if the texts give rise to misunderstandings which then make a reinterpretation necessary. Such texts give rise to comprehension problems which cannot be predicted and explained by proposition models. For text production, the following conclusions can be drawn from the model of cyclic processing: 1. Coherence gaps which the reader cannot be assumed to be able to close by inferencing should be avoided. 2. Propositions referred to must not be too distant unless they are so important that the reader can be assumed to have retained them due to their prominent hierarchy level in the text base. 3. References should be unambiguous. If the reader establishes wrong references when first reading a text, reinstatements and reorganizations may become necessary afterwards which are then detrimental to text comprehension. 4. Using a highly legible font, a transparent layout and simple sentence constructions saves the reader cognitive capacity which can then be used for more complex processing tasks. 42 Part II: Text Comprehension and Text Comprehensibility 3.4 Network models In the field of Artificial Intelligence (AI), semantic networks have been suggested to replace the lists of propositions used in propositional models (e.g., Norman/ Rumelhart 1975; Anderson 1976; Findler 1979). These are networks composed of nodes and edges, in which the nodes represent concepts and the edges, which must be conceived of as directed arrows, the relations between the concepts (cf., e.g., Norman/ Rumelhart 1975; Anderson 1976). The relations are usually specified using case roles such as agent, recipient, object. The following diagram, which is the network representation corresponding to the propositional representation in example [3-3] illustrates what semantic networks look like: [3-3] The atom donates an electron to its reaction partner. Propositional representation: [3-3’] (DONATE, ATOM, ELECTRON, REACTION PARTNER) Fig. 3-2: Network representation of sentence [3-3] What can be transformed into such network representations are not only individual propositions but also the line of argumentation and organization of complex texts. Depending on the purpose for which such representations are needed, generalizations and abstractions may be necessary as they are also formed in the reader’s mind while reading a text. Cognitive psychologists such as Norman & Rumelhart (1975) assume that our entire knowledge is organized in such networks which are interconnected in various ways (cf. Jahr 1996: 27). Transforming the linear sequences of signs and sentences in texts into network models can be used as a means of extracting the content of a text and thus of comprehending it. Network models, however, can also be used prior to text production processes as a text planning tool which assists in designing well structured texts. 3.5 Semantic macro-structures The processing of longer texts can be assumed to involve a cognitive reorganization of the information extracted, a reduction and condensation of information to what is most important for the recipient. This process of information reduction can be modelled as a process in which the ordered lists of propositions used in propositional models are reduced and condensed. The result of 3 Text-processing from a cognitive-science perspective 43 such reduction and condensation processes is called macro-structure . The original propositional lists can be considered as micro-structures accordingly. Macro-structures are assumed to be based not only on the micro-structures explicitly given in the text, but also on the prior knowledge that recipients bring into inferencing processes (van Dijk/ Kintsch 1983: 202). Macrostructures can thus be imagined as hypothetical semantic structures of texts which describe the global text content and its structure. They represent text meaning at a higher level of abstraction than the propositional microstructures departing from which they are created. Macro-structures are represented in the same manner as micro-structures. The propositions of which macro-structures are composed are the result of the application of so-called macro-rules on groups of micro-propositions. The resulting macro-propositions contain the contents of the micro-propositions on which they are based in a condensed manner. Macro-structures can then again be condensed by applying macro-rules to them another time. In this way macro-structures are obtained which represent text meaning at an ever increasing degree of abstraction and condensation (van Dijk 1980b: 41). Van Dijk (1980b) formulates four such macro-rules: 1. Deletion Deletion is a type of “semantic tree pruning” (van Dijk 1977b) in which irrelevant or unimportant propositions are removed. Example (taken from van Dijk 1977b: 11): Micro-structure: I needed some olive oil. So I went to the store and bought some. There was a sale on olive oil. We had a good Italian meal. Macro-structure: I needed some olive oil. So I went to the store and bought some. We had a good Italian meal. That there was a sale on olive oil is irrelevant for the topic of preparing Italian food and needing olive oil for it (can Dijk 1977b: 11). 2. Selection and integration According to this rule, propositions are selected from the text base which integrate the content of other propositions. The propositions selected are usually thematic or topical propositions. The propositions whose content is integrated into the one selected can be deleted. Example (adapted from van Dijk 1977b: 13): Micro-structure: John went to Paris yesterday. He took a cab to the station. He bought his tickets and then he boarded the train. Macro-structure: John went to Paris yesterday. 44 Part II: Text Comprehension and Text Comprehensibility 3. Generalization Example (adapted from van Dijk 1977b: 10): Micro-structure: Father was cleaning the kitchen. Mother was typing her new book. The children were painting the doghouse in the garden. Macro-structure: The whole family was working. 4. Construction Van Dijk (1977b: 14) describes this macro-rule as follows: Construction “is closely linked to INTEGRATION, and they might be considered variants of each other. The interesting difference is that CONSTRUCTION has no input proposition that organizes other propositions. Instead, a sequence of propositions is directly replaced by a macro-proposition, under the same conditions as in IN- TEGRATION. That is, we construct information at a more global level on the basis of micro-information in the text base. Typically, the micro-propositions represent conditions, components, and consequences of the (global) fact denoted by the macro-proposition.” Example: Micro-structure: Anne packed her suitcase, watered the flowers, closed the shutters and checked that all the lights were switched off, before she left the house. Macro-structure: Anne went on a journey. Frequently, the text surface contains indicators which can be used in the creation of a macro-structure, such as announcements (e.g., The following section will describe … ), illocutionary indicators of summarizing (e.g., This has shown that … ), relevance indicators (e.g., It is important to note that … ) and certain syntactical constructions (such as clefting constructions, which may point to a contrast focus) (van Dijk 1980a: 102 ff.). In addition to the four macro-rules, van Dijk (1980b: 202) also established converse macro-rules which are relevant during text recall. When applying these converse macro-rules, text recipients have to bring in their prior knowledge, interests and objectives as well (van Dijk 1080b: 184 ff.). Models of semantic macro-structures led to the hypothesis that macropropositions are recalled better and retained for longer periods of time than the contents of micro-propositions. Although few experiments have been conducted to check this hypothesis, the experimental evidence supports the hypothesis and thus the relevance of macro-structures for text processing (Christmann 1989: 72 ff.). 3 Text-processing from a cognitive-science perspective 45 3.6 Schema-theoretical approaches According to Graves (2004: 434), “ [ s ] chema theory is both one of the earliest and most influential constructs to emerge from the cognitive revolution”. Schema-theoretical approaches place even more emphasis on the relevance of the recipients’ prior knowledge, their expectations and objectives for text processing. In these approaches, the recipients’ prior knowledge is conceived of as being composed of schemata . These are basic knowledge clusters stored in long-term memory, which are relevant for information processing (cf. Christmann 1989: 75). Although the schema concept, which goes back to Bartlett (1932), has no uniform definition in the cognitive sciences, the different schema concepts share certain characteristics. Following Christmann (1989: 76), these can be described as follows (cf. also the summary in Hoppe-Graff 1984: 17): 1. Schemata represent knowledge about certain aspects of reality. We have schemata about “objects, situations, events, sequences of events, actions, and sequences of actions” (Rumelhart 1980: 34). They are composed of concepts and relations that exist between them (e.g., part-whole relations, temporal, causal, spatial relations). Concepts are the smallest building blocks of our knowledge. 1 We acquire them by abstracting from what we perceive. This can be illustrated by the way in which children acquire their mother tongue. Let us assume a child sees an animal on a meadow and is told by its mother that this is a cow. On another meadow, the child may see a horse and will then also call this animal a cow . The reason for this overgeneralization is that the child has formed a concept of cow that is so general that it also encompasses horses. Only after it learns that the animal on the other meadow is called a horse, will it modify its prior concept of cow in such a way that the perceivable differences between cows and horses are taken account of. 2. Schemata are organized in a hierarchical manner determined by the degree of specificity of the concepts they are composed of. 3. Schemata may be embedded in other schemata, i.e., a schema may have several sub-schemata. 4. Schemata contain variables, so-called slots , which can be filled by information gained in bottom-up processes. In this way, new information can be integrated into existing knowledge structures. 5. Schemata give rise to expectations about information that fits into the context. These expectations control the interpretation of incoming information and inference processes that are necessary to fill gaps in the coherence 1 From a neuro-physiological perspective, concepts can be conceived of as composed of bundles of activated neurons, so-called engrams . 46 Part II: Text Comprehension and Text Comprehensibility chain. In this way, schema theory can explain why we interpret the two sentences in example [3-4] below in such a way that we consider the ball the object that has broken the pane although this is not stated explicitly. [3-4] Peter kicked the ball. The window pane broke into a thousand pieces. Due to our experience, our schema of playing children includes the possibility that something is destroyed, especially if they are playing with a ball. Whenever window panes break, they usually do so because they are either knocked in or hit with a ball. When window panes are mentioned in the context of playing children, we cannot help expecting that window panes get broken. 6. Schemata are built up inductively on the basis of numerous individual experiences. 7. In information processing, information that enters our cognitive structures via bottom-up processes activates schemata stored in our long-term memories. These again activate sub-schemata and lead to the generation of hypotheses and expectations which are then used in our comprehension processes. As Graves (2004: 434) puts it, “we make sense of what we read and of our experiences more generally by a tacit process that in essence tells us, ‘Aha. This is an instance of such and such.’” 8. Schemata are culture-dependent, i.e., schemata of people from different cultures may have culture-dependent differences (cf. Reynolds et al. 1982). According to schema theory, text comprehension and recall depend on the schemata stored in the recipient’s memory. These can be activated by information which enters memory via bottom-up processes and thus can be used in text processing. If relevant schemata or sub-schemata are missing, however, recipients may encounter comprehension problems because they cannot use the prior knowledge that the author presupposed in his readership and thus are unable to make the necessary inferences. Missing schemata or subschemata may also lead to misunderstandings when recipients fall back on schemata in their cognitive structures whose use was not intended by the author. Schemata stored in our memories are relevant both for learning, i.e., the integration of new information into our knowledge structures, but also for information recall. In learning processes, they help us to interpret incoming information and provide a structure for its integration. In recall processes, they control our memory search and help us to reconstruct via inference processes information which is no longer available directly (cf. Christmann 1989: 76 f.). How can schema theory explain why different readers may rate the comprehensibility of one and the same text differently? The experiences we have in the course of our lives differ individually. The more intensively we work in a specific domain, the more complex our schemata for this domain will become. 3 Text-processing from a cognitive-science perspective 47 And the more complex the schemata become, the more knowledge will be activated indirectly whenever these schemata are activated, i.e., called up in our consciousness. The fact that, during text reception, not only those concepts are activated that are directly connected to the individual word forms that occur in a text but more comprehensive schemata is illustrated by the following example: [3-5] Peter ordered a pizza. He tipped the waitress generously. Reading this sentence does not only invoke a picture of Peter, the waitress and the tip but also a whole restaurant schema, although the restaurant is not mentioned explicitly in the text. Our experience tells us that the scene must be in a restaurant. We assume that Peter has chosen the pizza from a menu, that he has ordered it, etc. Reading the text simply invokes a whole restaurant scene which integrates the concepts explicitly mentioned in the text and against whose background they must be interpreted. This also explains why the noun waitress can be used with a definite article although she has not been mentioned before and would then normally have to be introduced using the indefinite article ( a ). Waitress can be used with the definite article here because waiters and waitresses form part of a restaurant schema and such a schema is activated when we read the text. Within this activated restaurant schema, the waitress can be considered to be ‘given’ information. Let us return to the question how schema theory can explain why different readers may rate the comprehensibility of one and the same text differently. Since we all have different experiences, the schemata stored in our minds differ as well. As a consequence, the top-down processes triggered by identical bottom-up processes vary from person to person. When writing texts for larger groups of recipients, we have to abstract from these differences to a certain degree and make sure that enough information is provided explicitly so that the common basis of comprehension is large enough for the text to achieve its objectives. Whenever technical communicators or journalists, for example, adapt academic texts for a lay audience, i.e., when they popularize texts (knowledge transfer - or better adaptation - from specialist to lay person) they have to take into account the differences in their schemata which have their origins in their different professional backgrounds. If a cognitive psychologist, for example, comes across the term schema in an academic text, the simple term will invoke a complex scene in his mind. If we wish to invoke the same complex scene in a person’s mind who is unfamiliar with cognitive psychology, much more effort is required to achieve this. The schemata and sub-schemata that the simple term schema activates in the specialist’s mind, have to be activated by mentioning their component concepts explicitly in the text. Whereas technical communicators and journalists have to take into account differences in prior knowledge that have something got to do with their read- 48 Part II: Text Comprehension and Text Comprehensibility ers’ different professional backgrounds, translators have to take account of culture-specific differences. Let us consider the British institution of the public school, for example: When Englishmen read this term, it invokes a complex scene in their minds that, depending on the context in which it occurs, may include characteristics such as ‘expensive private school’, ‘excellent education and training’, ‘upper class institution’, and places where public schools exist, such as Winchester and Eton. All this happens, because they have been confronted with all these concepts in related situations or contexts and therefore have established connections between these concepts in their long-term memories. The average German readers cannot be expected to have been confronted with these concepts in context. Therefore, they will not associate much with public school and perhaps even the wrong things if they translate it literally: öffentliche Schule , which a public school is not. For this reason, additional explanations will have to be provided in German translations of texts in which the term public school occurs, such as Public Schools, teure englische Eliteschulen . Another example of the fact that schemata are culture-dependent is provided by Reynolds et al. (1982). They requested both Afro-American and Euro- American students to read a text that described a quarrel while students were collecting their meals in a university restaurant. The description of the quarrel was so vague that it could either have been interpreted as a verbal dispute only or as a bodily fight. When the students had to interpret the texts, there were significant differences between the interpretations provided by the Afro- American students on the one hand side and the Euro-American students on the other hand side which can be explained with their different cultural backgrounds. None of the Afro-American students interpreted the scene as a bodily fight whereas 25% of the Euro-American students did. These examples illustrate how schema theory can explain how our prior knowledge influences our text comprehension. They should also have raised to our consciousness that the cognitive processes involved in free text production and in translation may be very similar. Both technical communicators and translators have to perform audience analyses in order to be able to decide how explicit they have to be. 3.7 The theory of mental models The theory of mental models provides a framework for the integration, or synthesis, of both propositional models and approaches such as schema theory, in which the recipients’ prior knowledge is emphasized (cf. Christmann 1989: 86). The theory of mental models assumes that text meaning is represented cognitively in two distinct manners. First, it is represented in the form of propositions. And second, these propositions are used to evoke pictures and 3 Text-processing from a cognitive-science perspective 49 movies in the mind, the so-called mental models , in accordance with Craik’s (1943) idea of thinking as manipulating internal representations of the world. Understanding the world is considered to be the creation of mental working models. Mental models represent specific sections of reality; they are usually incomplete and simpler than the reality they reflect (Johnson-Laird 1985: 10). According to Johnson-Laird, mental models have the function of allowing us to make inferences, to make predictions, to understand phenomena, to make decisions and control their execution and, above all, to let things happen in our minds. They allow us to use language to build representations which are comparable to those which we gain experiencing the real world (Johnson- Laird 1985: 397). Thus, in mental models, states of affair, events and actions can be simulated cognitively. They do not represent areas of reality in a symbolic or digital manner (as proposition models do) but represent them in an analogous manner (Schnotz 1985a; Christmann 1989: 87). With reference to the degree of abstraction of mental models, Christmann (1989: 88) explains that whereas propositional models are abstractions, mental models are specifications of states of affair. The specificity of mental models is comparable to that of pictures: A picture cannot depict the concept of a triangle in an abstract manner; it can only depict a specific triangle (cf. Johnson-Laird 1985: 157). In an analogous manner, propositional models represent the meaning of a text, whereas mental models represent their significance or meaningfulness (Johnson-Laird 1985: 381). Following the theory of mental models, text comprehension can be visualized as follows: Text meaning is represented in propositions. Step by step these propositions are then used to construct mental models. In doing so, we can utilize the schemata (as ready-made building blocks) that we have already stored in our long-term memories. Incoming propositions are then interpreted against the background of activated mental models of real or imagined sections of the world that have already been invoked. These mental models are then completed, checked for consistency and modified by the information from newly incoming propositions. In this manner, two degrees of comprehension can be distinguished: first, a relatively surface-oriented shallow form of comprehension at the propositional level; and second, a more profound form of comprehension that departs from the linguistic surface structure at the mental model level (Schnotz 1985a; Christmann 1989: 88). 3.8 Levels of comprehension Strictly speaking, we cannot only differentiate between a shallow form of comprehension and a more profound form, but between several levels of compre- 50 Part II: Text Comprehension and Text Comprehensibility hension. Rickheit & Strohner (1993: 79) differentiate between the following four levels of cognitive language processing: 1. At the sensomotoric level, perceptual comprehension takes place. This is the level of word recognition and the level that legibility research focuses on. 2. At the second level, syntactical comprehension takes place. Here parts of speech have to be recognized and morphological structures be interpreted. This is the level (apart from the first level) that readability research is primarily interested in. 3. At the semantic level, meaning has to be extracted. For semantic comprehension, first the concept connected to a word has to be activated (conceptual comprehension) . Then the objects referred to must be recognized (referential comprehension) . Third, the concepts referred to have to be integrated into a coherent structure. For this purpose, prior knowledge has to be integrated in top-down processes (semantic sense comprehension) . 4. At the highest level, the level of pragmatic comprehension, the situation of communication and the status of the people communicating has to be taken into account (pragmatic sense comprehension) . This level involves recognition of illocutions, for example. 3.9 Concluding remarks The theory of mental models has important implications for the construction of coherence and inferencing. According to Johnson-Laird (1985: 370), a text is coherent if it allows us to construct a mental model that is free from contradictions. Coherence is no longer dependent on argument overlap or argument embedding but rather a text characteristic that is actively created by the recipient in reader/ text interactions. It is thus dependent to a considerable extent on the recipient’s prior knowledge. Whereas, in propositional models, inferences are considered to be additional steps in text processing needed only when coherence gaps occur, they are an integral part of the comprehension process in the theory of mental models. Johnson-Laird does not believe that there are rule-based inference processes. In mental models, coherence construction is raised from the propositional level to the level of mental models. Schnotz (1985a) therefore refers to mental model approaches as “holistic” approaches, whereas propositional approaches are referred to as “additive-elementaristic”. 4 Text processing from the perspective of instructional psychology The two most important approaches towards explaining text comprehensibility from the perspective of instructional psychology are the co-called Hamburg comprehensibility concept developed by the psychologists Langer, Schulz von Thun & Tausch (1993), who were professors at the University of Hamburg when they developed it, and Groeben’s (1982) comprehensibility concept, which he developed at the same time as a psychologist at the University of Heidelberg. Both concepts are products of the 1970s (cf. Langer/ Tausch 1972; Langer et al. 1973; Schulz von Thun 1974; Groeben 1978). Based on these two models as well as on findings from cognitive psychology, I developed my own comprehensibility concept, which I called Karlsruhe comprehensibility concept because I was working at the Karlsruhe University of Applied Sciences at that time. Although the Hamburg group of psychologists on the one hand and Groeben on the other followed different approaches, an empirical inductive approach and a theoretical deductive approach respectively, they both came to the conclusion that there are characteristics in four so-called dimensions that have an effect on text comprehensibility. These four dimensions are: 1. ‘(linguistic) simplicity’ (“Einfachheit”/ “sprachliche Einfachheit”) 1 2. ‘organization - structure’/ ‘cognitive structure’ (“Gliederung - Ordnung”/ “kognitive Gliederung”) 3. ‘concision’ (“Kürze - Prägnanz”/ “semantische Kürze/ Redundanz”) 4. ‘motivation’ (“anregende Zusätze”/ “motivationale Stimulanz”) How these four dimensions were found, how they are defined and how they can be applied to the assessment of texts will be described and evaluated critically in the next sections. The results of this discussion will then be used for the development of my own communication-oriented framework of reference for the assessment of text comprehensibility, the Karlsruhe comprehensibility concept. 4.1 The Hamburg psychologists’ empirical inductive approach For the development of their comprehensibility concept, the Hamburg group of psychologists followed an empirical inductive approach. Taking account of 1 The terms in parentheses are the original designations used by the Hamburg group and Groeben respectively. 52 Part II: Text Comprehension and Text Comprehensibility the findings of stylistics, readability research and related areas of investigation, they compiled a list of features that can be used to describe the quality of texts. These features are listed in Tab. 4-1. Tab. 4-1: Features for the description of text quality (cf. Langer/ Schulz von Thun/ Tausch 1993: 139) interesting ........................................... - ........ boring too concise .......................................... - ........ too long easily comprehensible ....................... - ........ hard to understand logical................................................... - ........ confusing for pupils few hard words ................. - ........ for pupils too many hard words motivating ........................................... - ........ dull well-structured ................................... - ........ poorly structured simple sentences................................. - ........ highly complex sentences concrete ............................................... - ........ abstract with a clear layout .............................. - ........ with a confusing layout fluent .................................................... - ........ clumsy illustrative............................................ - ........ obscure simple................................................... - ........ complex easy to learn ........................................ - ........ hard to remember to the point.......................................... - ........ not to the point appropriate for children.................... - ........ not appropriate for children good distinction between important missing distinction between important and less important information - ........ and less important information diversified............................................ - ........ monotonous The Hamburg psychologists requested participants (students, teachers, etc.) to assess a total of 94 texts covering four topics (from pragmatic texts as we encounter them in our everyday lives to academic texts) on a seven-point scale from -3 to +3 (-3, -2, -1, 0, 1, 2, 3) after they had followed a specific training program. By means of a factorial analysis, they determined correlations between features. Features whose values correlated were combined to form dimensions, and in this manner the four comprehensibility dimensions came about (cf. the overview in Langer/ Schulz von Thun/ Tausch 1993: 160). The first item of criticism that can be levelled against the list of features in Tab. 4-1 is that Langer, Schulz von Thun & Tausch do not justify why they used these 18 pairs of features and not others or additional ones (cf. also Krause 1991: 396 ff.). Thus, we cannot exclude that the selection of other features or of additional ones would have led to different dimensions of comprehensibility. This raises the question whether, in addition to the four dimensions, there are still others that may be relevant for comprehensibility. I will return to this question in Section 4.2. Furthermore, the argumentation becomes circular in that the features easily comprehensible vs. hard to understand are used. To test the relevance of the four dimensions that the first step yielded, the Hamburg psychologists requested 1,200 5th to 7th graders to read the texts 4 Text-processing from the perspective of instructional psychology 53 used in the first step and to answer questions on them or to work on assignments related to them. In these experiments, texts which had achieved high values in the dimensions ‘simplicity’ and ‘structure’ and moderate values in the dimension ‘concision’ achieved the best results. ‘Motivation’ only had a positive effect on the comprehensibility of texts if they were well-structured (Langer/ Schulz von Thun/ Tausch 1993: 160). From these results, Langer, Schulz von Thun & Tausch derived an ideal value for each dimension, which they presented in a so-called assessment window (cf. the overview in Langer/ Schulz von Thun/ Tausch 1993: 28). Simplicity +2 Structure +2 0 or +1 Concision 0 or +1 Motivation Fig. 4-1: Assessment window showing the values for a text that, according to the Hamburg model, is optimally comprehensible In this window, the 7-point Likert scale used originally has been reduced to a five-point Likert scale. Langer, Schulz von Thun & Tausch (1993: 27 f.) consider the dimension ‘simplicity’ to be the most important one, followed by the dimension ‘structure’. They do not, however, provide a justification for this, neither for their ranking nor for the ideal values. It is interesting to note that Groeben (1982: 197, 207) ranks the dimensions differently. He considers the dimension ‘structure’ to be the most important one. What Groeben objects against the assumption that ‘simplicity’ is the most important dimension are findings from the psychology of language concerning the relation of grammatical structure and content. These show that grammatical structure only plays an interim role in the comprehension of utterances. As soon as we are able to extract meaning from an utterance, we forget about the precise grammatical structure in which it was conveyed and only retain the content (cf. the overview in Groeben 1982: 38 f.). To my mind, it does not make sense to rank the dimensions in a fixed order, nor does it make sense to claim fixed ideal values in each dimension since both are dependent on factors such as the genre, the addressed readership’s prior knowledge and their reading habits. ‘Simplicity’ is definitely less important for an academic readership than for typical readers of the yellow press. ‘Concision’ is more important in an abstract than in a popular-science article. ‘Motivating elements’ are important in popular-science texts but definitely not in abstracts and summaries. 54 Part II: Text Comprehension and Text Comprehensibility In a third experiment, the Hamburg psychologists had 20 original texts (from everyday life, textbooks and scientific articles) optimized in such a manner that they achieved ideal values in all four dimensions. The fact that the optimized versions were really comprehensible was confirmed by independent raters. Then the optimized versions were presented to a group of experimental participants (pupils, students, employees), whereas the original versions were presented to an equivalent control group. The members of both groups were administered comprehension and recall tests. In these experiments, the Hamburg group of psychologists found that the optimized versions, without exception, achieved better results than the originals and that the differences in comprehensibility and recall between the originals and their optimized versions increased with the discrepancies in the values in each dimension that existed between them (cf. the summary in Langer/ Schulz von Thun/ Tausch 1993: 161). Whether these findings really confirm the correctness of the Hamburg concept appears to be doubtful for the following four reasons: 1. Finding the four dimensions and validating them are based on assessments by trained persons, so-called expert raters . In many cases, these expert raters did not belong to the target group of the texts in question and, consequently, may have had a completely different background with regard to their education and training, their prior knowledge on the topics of the texts and their motivation to read them. They may have been able to imagine how the intended readership might approach these texts but, in the end, they could only verbalize opinions about how comprehensible these texts would be for their readership. Therefore, the assessment procedures employed by the Hamburg group of psychologists must be considered to be speculative (cf. also Groeben 1982: 197). 2. Speculation is compensated for to a certain extent by not only elicitating expert raters’ opinions about the comprehensibility of the texts but also having them assessed by experimental participants. These experimental participants, too, however, did not necessarily have to belong to the target groups for which the texts had been written originally. A text only has to be comprehensible for the audience for which it was written. Consequently, it does not make sense to determine a text’s comprehensibility in experiments that involve participants which do not belong to the text’s target group. Field trials would have been more useful here. 3. If we compare the optimized versions that Langer, Schulz von Thun & Tausch (1993) present for certain text passages, it becomes obvious that the optimized versions do not always address the same audience as the original versions and that they do not always belong to the same genre. This is illustrated by the following original passage and its optimized version that Langer, Schulz von Thun & Tausch (1993: 17) present: 4 Text-processing from the perspective of instructional psychology 55 Complex version: Was ist Raub? „Raub ist dasjenige Delikt, das jemand durch Entwendung eines ihm nicht gehörenden Gegenstandes unter Anwendung von Gewalt oder von Drohungen gegenüber einer anderen Person begeht, sofern die Intention der rechtswidrigen Aneignung besteht.“ Simple version: Was ist Raub? „Jemand nimmt einem anderen etwas weg. Er will es behalten. Aber es gehört ihm nicht. Beim Wegnehmen wendet er Gewalt an oder droht dem anderen, daß er ihm etwas Schlimmes antun werde. Dieses Verbrechen heißt Raub.“ The ‘optimized’ version has features of spontaneous oral communication and seems to address a child, whereas the original text does not have these features. Taking this into account, it is comprehensible that Heringer (1984: 62) feels that some passages in the optimized versions sound rather strange or even worse than the original ones. If, in the experiments conducted by the Hamburg psychologists, academic texts were ‘optimized’ in this manner for a lay audience, it is not surprising that the optimized versions achieved better comprehensibility results. It has to be doubted, however, whether they would still have been taken seriously by their scientific community. This is related to the next item of criticism. 4. The comprehensibility of the optimized versions was tested exclusively in artificial experimental situations. It should be questioned, however, whether some of the optimized versions, in contrast to their originals (see the examples above), would have been understood by the participants (or even their intended readership) in the same manner, whether they would have been interpreted in the way intended by the author and whether they would have been read at all. Next, let us have a look at how Groeben arrived at his four comprehensibility dimensions. 4.2 Groeben’s theoretical deductive approach Groeben arrives at his comprehensibility construct following a theoretical deductive approach, i.e., he starts from different theories, which will be presented in the following, and derives features from them which should have a positive effect on text comprehensibility. The theories on which he founds his dimension of simplicity are those of hermeneutic stylistics (cf. Reiners’ Stilfibel 1963) as well as those of empirical readability research, which were also used by Langer, Schulz von Thun & Tausch (1993). In addition, he also falls back on psycholinguistic models which, following Chomsky’s (1969) transformational grammar, focus on the 56 Part II: Text Comprehension and Text Comprehensibility grammatical features of sentences and texts and their impact on text comprehension and recall (Groeben 1982: 199). One of the assumptions based on these models is that recipients transform sentences during comprehension until they are available in a specific structure, their deep structure , as so-called kernel sentences . The number of transformations necessary in this process depends on the grammatical surface structure of the sentences. The more transformations are required, the more difficult to understand a sentence is assumed to be. According to this theory, affirmative passive-voice sentences and negated active-voice sentences require more transformations than affirmative active-voice sentences; negated passive-voice sentences require even more transformations than negated active-voice sentences or affirmative passive-voice sentences. There is no clear empirical evidence for this, however (cf. Heringer 1979: 263, who comments on this critically; as well as Christmann 1989: 13; and Schwarz 1992: 140 f.). That active-voice sentences are understood better and faster than passive-voice sentences could be confirmed in experiments in which participants were confronted with isolated sentences. If sentences are embedded in a context, however, passive-voice sentences may obtain better results than active-voice sentences due to reasons of thematic progression (cf. Section 1.2.2.4). Due to doubts concerning the individual theories Groeben draws on (the shortcomings of readability research have already been pointed out in Section 2.2), the complex of theories selected by Groeben to back up his dimension of simplicity is prone to criticism as well. Groeben’s dimension of concision is based on models of information theory (Groeben 1982: 199). Groeben’s dimension of structure, which he, in contrast to the Hamburg group, considers to be the most important dimension, is derived from Ausubel’s (1968) cognitive theory of learning, the so-called subsumption theory (Groeben 1982: 199 ff.). According to this theory, we understand and learn best if we are able to bring the new information to be processed and learned into a thematic connection with what we already know, i.e., if we are able to connect new information to more general knowledge that we have already stored in our long-term memories. In other words, processing and learning involves subsuming new knowledge under knowledge that we have already stored in our memories. The most important recommendations that can be derived from this theory for the structuring of texts are: 1. First of all, an audience analysis has to be performed to find out what this audience already knows and in which contexts this knowledge was acquired. These contexts determine how our knowledge is structured in our memories. Existing knowledge structures may provide text producers with hints as to the areas from which metaphors may be taken when explaining new phenomena. 4 Text-processing from the perspective of instructional psychology 57 2. Before confronting our readership with new information, we must make sure that this information ‘encounters’ knowledge components in our readers’ cognitive structures with which it is related and can be connected thematically and logically. 2a) If such knowledge clusters cannot be expected to be available in our readers’ cognitive structures already, they have to be built up first step by step. 2b) If such knowledge clusters can be expected to be available already, they must be activated in our readers’ minds first so that they can then be used as ‘hooks’ for new knowledge. This may happen by means of bottom-up processes, for example, by using advance organizers (Ausubel 1960, 1968; Groeben 1982: 235). Advance organizers are composed of meta-communicative information that is presented before the actual information to be conveyed, as in example [4-1], and that provides information with a degree of abstraction or generalization that is higher than that of the actual information to be conveyed. [4-1] In the following, we will first have a look at engine design in general and then take a closer look at the individual systems that engines are composed of. Advance organizers are assumed to activate the relevant prior knowledge learners already possess so that is can be used as a ‘hook’ to which new information can be attached. They help to build bridges between what is known already and what has to be learned (Ausubel 1968: 131 f.). Subsumption theory sets text structure in relation to the reader’s cognitive system, i.e., it takes audience characteristics into account. Furthermore, it is compatible with both schema theory and the theory of mental models and thus appears to form a sound foundation for the dimension of structure. For his dimension of motivation, Groeben uses the curiosity arousal theory by Berlyne (1960/ 1974). According to this theory, the reader’s motivation to read a text can be fostered by creating conceptual conflicts. Conceptual conflicts may be created by providing information that seems to be contradictory at first. These contradictions arouse the desire in the reader to resolve these conflicts and thus to read on in order to find the solution (Groeben 1982: 202). The statement: “Books are like tetra packs.” may create such a conceptual conflict and thus induce the reader to read on in order to find out what the tertium comparationis is. Like motivating additions, conceptual conflicts are not appropriate in all genres. From a global perspective, Groeben’s comprehensibility construct lacks a theoretical framework into which the individual theories he falls back on can be integrated, an item of criticism that can also be levelled against the Hamburg comprehensibility concept. Without such a theoretical framework, the selection of theories used remains arbitrary. Groeben himself describes the use of 58 Part II: Text Comprehension and Text Comprehensibility his comprehensibility construct as follows. It provides a theoretical framework for applied technological research: “the number and explanation of the dimensions of text comprehensibility forms an (empirically based) heuristic for the detection of further text features relevant for comprehensibility and their specification; the weighting of the individual dimensions provides an orientation as to how to weight the comprehensibility-relevant text features pertaining to them as well as related instructions for text optimization” (Groeben 1982: 216 f.; my transl.). I agree with him insofar as the four comprehensibility dimensions can be considered to be a heuristic framework that assists us in searching for further criteria (within each dimension) and in specifying them. I do not believe, however, that anything can be derived from his ranking of the dimensions as outlined in Section 4.1. In what follows, the four dimensions will not be described in more detail but rather included in the presentation of my own Karlsruhe comprehensibility concept. In this presentation, an explanation will be provided of what distinguishes my comprehensibility dimensions from Groeben’s and those of the Hamburg group of psychologists. 4.3 The Karlsruhe comprehensibility concept Graves (2004: 443) complains that students are rarely familiarized with the genres they will have to produce outside their educational contexts and formulates the demand that writing instruction “should include instruction on the functions and purposes of different sorts of texts”. According to Devitt (2008: 204 f.), antecedent genres, i.e., genres students are already familiar with, can work as primers for genres students will have to compose in the future. For writing pedagogy, he concludes from this that “selecting genres with the most potential as antecedent genres for a particular students population while teaching how to learn genres [ ... ] may be the most responsible reaction” (Devitt 2008: 205). As Rounsaville (2012: 12) 2 points out: “In this approach, transfer occurs as a consequence of the student having both the ‘right’ antecedent genres that approximate future ones and a theoretical understanding of genre as social action”. Providing students with an understanding of genres as social action with specific communicative functions can be accomplished by means of the Karlsruhe comprehensibility concept, which is not limited to specific genres and allows to specify any text to be analyzed or produced according to its communicative function. 2 The online document has no page numbers. The page numbers given refer to the pages of the printout. 4 Text-processing from the perspective of instructional psychology 59 Like the Hamburg comprehensibility concept and the comprehensibility construct developed by Groeben, my own Karlsruhe comprehensibility concept provides an expert-judgment-focused method of comprehensibility assessment (Göpferich 2001; 2002; 2009). I developed it when training technical writers at the Karlsruhe University of Applied Sciences. It represents an extended and improved version of the two comprehensibility concepts discussed in the previous two sections. These two comprehensibility concepts have attracted much criticism (cf. Heringer 1979 and 1984), but they nevertheless could not be banned from the classroom. The reason for this is that these concepts have turned out to be didactically useful in courses on text production and text optimization and that no better alternatives had been presented. A survey of the criticism levelled against these two comprehensibility concepts from educational psychology, which induced me to develop my Karlsruhe concept, has been given in the previous sections (cf. also Göpferich 2002: 136 ff.). The most serious drawback of them is that they are text-focused and lack a text-external frame of reference (cf. Biere 1989: 41 ff.). In the following, such a frame of reference will be presented. Furthermore, the four dimensions will be specified and two additional ones (‘correctness’ and ‘perceptibility’) will be introduced. Finally, it will be illustrated which text characteristics help to fulfil the requirements in each dimension. Fig. 4-2: Framework and comprehensibility dimensions of the Karlsruhe comprehensibility concept 60 Part II: Text Comprehension and Text Comprehensibility The development of my model is based on concepts and insights from the cognitive sciences (schema theory and mental models), from educational psychology (the four comprehensibility dimensions), from linguistics (stylistics, text linguistics, psycho-linguistics, LSP research/ terminology science), as well as from communication theory and semiotics (communication model and sign concept). Fig. 4-2 gives an overview of the framework, the comprehensibility dimensions and their range of application. The concepts introduced in this framework will be explained in the sections to follow. 4.3.1 Communicative function The quality of a text cannot be determined without taking into account its communicative function. This also applies to the comprehensibility of a text, which forms one component of its quality. In fact, text quality can even be defined as the degree to which a text fulfils its communicative function, the latter representing a specification to which a text can be composed. If we want to use a text’s communicative function as such a specification or as a frame of reference against which its quality can be evaluated and optimized, it has to be specified with a certain degree of precision taking into account a) the purpose of the text, b) its target group and c) its sender. a) Purpose The concept of purpose is rather vague because it can be specified with varying degrees of precision. If we want to use this concept as a frame of reference for text production, text evaluation and text optimization, the precision with which it is specified must have a certain minimum degree. A general definition of this minimum degree of precision which applies to all texts cannot be given. I must therefore confine myself to defining a few purposes with the necessary degree of precision as examples: enabling the target group to make a phone call with a mobile phone; making the target group understand why a friction clutch is inevitably subject to wear; enabling the target group to understand the invoice of their car’s clutch repair; informing people about therapies for patients infected with AIDS. b) Target group What a text must look like in order to fulfil its specified purpose also depends on its target group. It goes without saying that a text about AIDS therapies written for doctors who have specialized in this field must differ from a text about the same topic written for patients suffering from AIDS. Differences will not only occur in the terminology used but, among other things, also in the explicitness and depth in which the therapies have to be described. Examples 4 Text-processing from the perspective of instructional psychology 61 of target-group features which may have an impact on text comprehension are: age; sex; social, regional, and cultural background; education and training; hobbies; other prior knowledge and prejudices regarding the topic concerned; homogeneity of the target group, especially with regard to their prior knowledge on the topic covered (for reasons cf. Göpferich 2002: 157). c) Sender Apart from the purpose and the target group, the sender is another factor which has and must have an impact on the characteristics of a text. A leaflet informing students about modifications in their degree programs and issued by the ministry of education will differ clearly from a leaflet with the same purpose and target group which is issued by student representatives. A manufacturer of luxury cars will use corporate wording in its driver’s manuals that is different from that used by a manufacturer of cheap cars (cf. our automobiles vs. our cars ). In addition to characteristics such as age; sex; social, regional, and cultural background; education and training; and hobbies, which are relevant targetgroup features, too, the following characteristics may be relevant sender features: the situation in which the text is issued; the person or institution in whose name the text is issued (individual, association, institution, company); and the social relation between sender and recipient (for reasons cf. Göpferich 2002: 158). The purpose, the target group and the sender of a text make up its communicative function. In the conceptional phase, this communicative function first determines the guiding features of text production (arrow  in Fig. 4-2) and then also the encoding itself (arrow  in Fig. 4-2) wherever the guiding features of text production leave the author room for personal decisions. 4.3.2 Guiding features of text production The guiding features of text production comprise a) the mental model of the objects, processes, events, etc. covered in the text (mental denotation model) , b) the mental model of the genre to be used (mental convention model) , c) the medium in which the information is conveyed, and d) legal requirements and author’s guidelines, if applicable. a) Mental denotation model This is the mental picture or movie of the objects, processes, events, etc. which must be invoked during text reception if the text fulfils its communicative function. Ideally, the signs used in the text and the top-down processes they induce invoke the desired pictures and movies in the reader’s mind. In the conceptional phase, the mental denotation model represents the author’s men- 62 Part II: Text Comprehension and Text Comprehensibility tal picture or movie of the objects, processes, events, etc. to be conveyed, which s/ he encodes, i.e. transforms into signs, in the exteriorization phase. What the mental denotation model conveyed in a text must look like depends on the text’s communicative function. The following examples will illustrate this. A very simple mental denotation model of a friction clutch can be visualized as shown in Fig. 4-3. 1 clutch cover 2 pressure plate 3 friction disk 4 diaphragm spring 5 pivot rings 6 release bearing 7 clutch release lever 8 flywheel 9 clutch shaft Fig. 4-3: Simple mental denotation model of a friction clutch (adapted from Niess et al. 1984: 211) Fig. 4-4: Complex mental denotation model of a friction clutch, in this case a diaphragm-spring clutch (adapted from Volkswagen 1988: 5) For a text with the purpose of ‘explaining the functional principle of a friction clutch’ and the target group ‘laypersons in the domain of automotive engineering’ the mental denotation model in Fig. 4-3 will be sufficient: When the two disks are pressed against each other with sufficient force and the shaft on the left hand side turns, the disk on the right hand side will also be caused to turn due to the friction force at the mating faces of the two disks. This simple mental denotation model will also suffice if the purpose of the text is to explain why a friction clutch is inevitably subject to wear. Mentioning further details, 4 Text-processing from the perspective of instructional psychology 63 such as the full range of components of a clutch (pressure plate, friction disk, diaphragm spring, etc.; cf. Fig. 4-4) would be superfluous for these purposes. If the purpose of the text on clutches is to enable a car owner to understand the details in his clutch repair invoice, however, a more complex mental denotation model will be necessary, as shown in Fig. 4-4. This more complex mental denotation model must also include the components of the clutch which can be replaced and may therefore appear in the invoice. b) Mental convention model When encoding their mental denotation models during the exteriorization phase, authors are not completely free in choosing the signs to use, but have to follow the conventions of a genre appropriate for the text’s communicative function. In contrast to legal requirements and author’s guidelines, these genre conventions are not codified in the written form but have gradually developed into patterns of language use and text composition (cf. Reiß/ Vermeer 1984: 177). These have been internalized by competent language users in the form of mental (genre) convention models, at least as far as text reception is concerned. Mental convention models may comprise schemata of the structure (macrostructure) of genres and genre-specific standardized formulations. Familiarity with the convention model of a genre controls and facilitates text comprehension. If this familiarity is not only a passive one but also comprises the productive level, it also facilitates the production of texts of the respective genre. During text reception, the mental convention model takes over the function of an advance organizer , which makes it easier for the recipients to integrate the information from the text into a consistent whole. It controls the recipients’ expectations of what will come next in the text and also of the way in which this will be verbalized (style, standardized formulations, terminology, etc.). If these expectations are met, the mental convention model will facilitate comprehension (cf. Kintsch/ Green 1978). If they are disappointed, as in example [4-2] below, they may have a negative effect on text reception and comprehension. It seems plausible to assume that mental convention models remain in the subconscious and do not require any processing capacity in the reader’s working or short-term memory as long as they are not broken. If they are broken, however, the recipient’s expectations will be disappointed and this may cause parts of the mental convention model to be raised to consciousness thereby triggering top-down processes which help to interpret the discrepancies. In such cases, comprehension no longer occurs , but has to be struggled for (cf. Heringer 1984: 60). Interpretation efforts triggered by broken conventions can be illustrated by the following example. In negative replies to job applications, hedging is conven- 64 Part II: Text Comprehension and Text Comprehensibility tionally used to convey the message in a polite manner. These hedges make the text longer (cf. the following variants): [4-2a] We will not employ you. [4-2b] We have decided in favour of another applicant. [4-2c] We regret to inform you that we have offered the position you applied for to another applicant. Variants [4-2a] and [4-2b] are shorter than variant [4-2c] (cf. the requirements to be derived from of the comprehensibility dimension of concision below), but this does not make them more comprehensible in the genre ‘negative feedback on an application’. On the contrary, since variants [4-2a] and [4-2b] do not conform to the conventions of this genre, they sound unusual to the reader, which may already have a negative effect on text comprehension. The discrepancy may trigger unwanted top-down processes in the reader’s mind, such as reflections on why the author did not conform to the conventions (impoliteness? , humiliation? , etc.). Such reflections require memory capacity which will then not be available for processing the central information of the text. c) Medium The medium in which the message is conveyed may also be determined by the text’s communicative function. In addition to this, it also depends on the mental denotation model. If this comprises processes which are difficult to describe by means of words and/ or static pictures only, a medium has to be chosen which allows animated nonverbal representations, such as computer-based trainings (CBTs). The medium, in turn, may have an influence on the mental denotation model. If multi-media representations can be used, this leaves more room for the creation of mental denotation models than situations in which information can be conveyed in the written form only. The medium also has an impact on the mental convention model and vice versa. This becomes obvious when we compare the conventions in a letter with those in an e-mail (here the medium determines the conventions) and take into account that certain standard situations (such as quitting a job) require the use of a specific medium (e.g., the written form) (here the convention model determines the medium). d) Legal requirements and author’s guidelines Examples of such requirements are guidelines to be met by technical documentation. In many companies, they can be found in so-called style guides , collections of formatting, layout and writing rules to be followed to achieve more consistent and comprehensible texts which are easier to translate. Other requirements such as the ones in EU directives have to be met for legal reasons. They specify which contents must be covered in operating instructions, 4 Text-processing from the perspective of instructional psychology 65 how they have to be structured, which style has to be used in them, etc. (cf. Göpferich 1998: Chapter 11). These requirements and guidelines, which exist in the written form, have an impact on mental convention models, which only exist in the language users’ minds, and vice versa. On the one hand, existing conventions have an impact on what can be prescribed in written regulations and guidelines. Written regulations and guidelines only narrow down the range of options conforming to conventions. On the other hand, regulations and guidelines have an impact on conventions causing them to be reduced to certain options in the course of time. All the text production guiding features together determine the leeway authors have in their text production process (arrow  in Fig. 4-2). The assignment data (communicative function comprising the purpose, the target group, and the sender of a text), the text production guiding features (mental denotation model, mental convention model, medium as well as legal requirements and author’s guidelines) and the interdependences between them form a framework for text production and evaluation. The next section will deal with the six comprehensibility dimensions (cf. Fig. 4-2) and the components within this framework to which they refer. 4.3.3 Comprehensibility dimensions Following the comprehensibility concepts developed in educational psychology, the Karlsruhe comprehensibility concept differentiates between the comprehensibility dimensions of structure (derived from the dimensions ‘organization - structure’ and ‘cognitive structure’ respectively), concision, motivation, and simplicity (derived from the dimension ‘[linguistic] simplicity’). New terms are introduced here to avoid confusion of my concepts with the rather vague ones of the educational psychologists. In addition to the four dimensions mentioned above, my comprehensibility concept comprises two further dimensions: the dimension of correctness and the dimension of perceptibility. They take account of the fact that the correctness of a text at all levels (contents, layout, typography, etc.) and the ease with which it can be perceived and thus transferred to the reader’s cognitive systems for further processing are two additional important factors which determine text comprehensibility. Another major difference between the dimensions of the Karlsruhe comprehensibility concept and the ones introduced by the educational psychologists is the following: The Karlsruhe concept makes a distinction between requirements from dimensions which refer to the text (the encoded units) only and requirements from dimensions which, in addition to this, have to be fulfilled by the mental denotation model to be conveyed. The educational psycholo- 66 Part II: Text Comprehension and Text Comprehensibility gists’ comprehensibility concepts do not make a distinction between these two types of requirements. a) Concision The dimension of ‘concision’ refers both to the mental denotation model to be conveyed in the text and the textual code itself. A text has an ideal degree of concision if 1. the mental denotation model conveyed in the text has been reduced to the minimum of information that is absolutely necessary or relevant for the text to fulfil its communicative function taking into account the requirements to be derived from the guiding features of text production (mental convention model, medium, legal requirements and author’s guidelines) and the other five dimensions, and 2. the mental denotation model that fulfils these requirements has been exteriorized with the minimum of signs possible to achieve its communicative function without violating the guiding features of text production and the requirements from the other comprehensibility dimensions. Examples of maximum economy in mental denotation models, which depends on the communicative function of the texts, have been given in Figures 4-3 and 4-4. Mental denotation models must have neither gaps nor include superfluous details. Superfluous details may not automatically have a negative effect on text comprehensibility but they increase text reception effort and this is not in the reader’s interest. The second requirement specified above implies that a mental denotation model can be conveyed with varying amounts of signs so that the different presentations can be compared with regard to the amount of signs used in them. A prerequisite for such a comparison would be that the versions compared were equivalent apart from the amount of signs used in them. This equivalence requirement has been discussed extensively and for a long time in Translation Studies (cf. Neubert 1973; Jäger/ Müller 1982; Nord 1989; Koller 1992; Schreiber 1993; etc.). The results of this discussion are that various types of equivalence must be distinguished, such as denotative equivalence, formal equivalence, stylistic equivalence, etc., which cannot all be achieved at the same time, so that the translator has to establish a hierarchy (cf. Schreiber 1993: 66 ff.). The criterion for establishing this hierarchy cannot be found in the source text itself. It can only be derived from the function (skopos) of the translation (cf. Vermeer 1978; Reiß/ Vermeer 1984; Holz-Mänttäri 1984; Hönig/ Kußmaul 2 1984; Nord 1993; Hönig 1995; Schmitt 1999). This skopos - at least according to the functionalists - can be completely different from that of the source text so that it may be necessary to deviate from the source text on all levels and aim for zero-equivalence. This insight has led to the equivalence 4 Text-processing from the perspective of instructional psychology 67 requirement being rejected - at least by the functionalists - and replaced with the requirement of adequacy for a specific function, the so-called skopos (Reiß/ Vermeer 1984: 133). This skopos theory can also be applied to text optimizations as types of ‘intralingual translations’. Absolute equivalence between different text variants cannot be achieved within the same language either, nor is it desirable: The optimized version is intended to be ‘better’ than the original one, i.e., it has to be different. What can be achieved is an optimization for a specific communicative function. This communicative function has to be specified before the optimization, something which neither Langer et al. nor Groeben do to a sufficient extent. Other important text-external factors which determine the minimum amount of signs that can be used to convey a message, are the target group with their specific characteristics (e.g., their prior knowledge in the respective subject domain and the terminology they are familiar with) and the genre with its genre-specific conventions (cf. example [4-2] above). Violations against the requirements to be derived from the dimension of concision can be categorized into four groups: 1. Missing or superfluous details in the mental denotation model In the following example, the information that the pressure plate is guided by pins and that the friction lining is riveted to the friction disk are superfluous details in the mental denotation model. The text’s function is to explain to the readers why a friction clutch is subject to wear and to help them to understand their clutch repair invoice “Der Kupplungsdeckel ist mit der Schwungscheibe des Motors fest verschraubt. In ihm befindet sich die durch Bolzen beweglich geführte Druckplatte, die durch Federdruck die Kupplungsscheibe gegen die Schwungscheibe preßt. Die beidseitig mit aufgenieteten Reibbelägen versehene Kupplungsscheibe sitzt auf der längsverzahnten Hauptantriebswelle des Getriebes. Sie bildet das eigentliche Bindeglied zwischen Motor und Getriebe.” (Volkswagen 1988: 3; highlighting by S.G.) What guides the pressure plate is absolutely irrelevant for the text’s communicative function. What is important here is only that the pressure plate can be moved forwards and backwards in one direction (axially). What is of no interest either is the way in which the friction linings are fixed to the friction disk. The reader needs to know only that the friction disk has friction linings on both sides. 2. Long formulations instead of appropriate shorter ones which convey the same text-relevant information Examples of this type of violation against the dimension of concision are the formulations mit Hilfe , es ist möglich , auf andere Werte einstellen in the 68 Part II: Text Comprehension and Text Comprehensibility original version in [4-3]. They can be replaced by the shorter versions mit , Sie können and ändern . [4-3] (to be used in a software manual): Avoid: Mit Hilfe der vorhandenen Unterfunktion SETUP ist es möglich, die standardmäßig eingestellten Vorgaben auf andere Werte einzustellen. Use instead: Mit (der Unterfunktion) SETUP können Sie die Standardwerte ändern. (example adapted from Schmitt 1999: 392) [4-4] (extract from a car repair manual) Avoid: The steps below are defined in order to instruct the technician how to proceed with a diagnosis. Use instead: Follow these steps in order to make a diagnosis. (example from Means 1999) 3. Tautologies A tautology can be found in the original version in example [4-3] in the formulation standardmäßig eingestellte Vorgaben (instead of Standardwerte ). 4. Redundancies between the information given in the written text itself and the information the recipient can find in the material accompanying it (such as a software user interface) When writing instructive texts, the author has to take into account that the readers of these texts do not only get the information they need from the written text itself but also from the product (status messages and other information on the screen, flashing lights, etc.). What must be avoided in such texts are not only redundancies within text itself but also redundancies between a message conveyed both by the text itself and by the product at the same time. Instead of creating such redundancies, the readers’ attention should be directed to the product reactions so that they learn to use the information given by the product. Whereas excessive concision always leads to poorer comprehension, a lack of concision may not have this type of negative effect. Whether we use mit Hilfe der Unterfunktion SETUP or mit der Unterfunktion SETUP in example [4-3] will certainly have no effect on text comprehensibility. Nevertheless, pragmatic texts, especially instructive texts, should be concise to minimize the effort and time the reader has to invest in reading them. Apart from comprehensibility, reception economy is another characteristic of high-quality pragmatic texts. b) Correctness ‘Correctness’ is a comprehensibility dimension Langer et al. and Groeben do not take into account. They seem to take it for granted that the texts to be optimized and their optimized versions do not contain any mistakes, which is rarely the case (cf. Schmitt 1999: 59 ff.). The correctness requirement applies 4 Text-processing from the perspective of instructional psychology 69 to all components in the framework in Fig. 4-2: from wrong assumptions about the target group’s prior knowledge via an unsuitable mental denotation model, convention model or medium to linguistic mistakes in the text itself. Since text production, text evaluation, and text optimization are impossible without knowing the text’s communicative function and since misconceptions with regard to the audience or the sender will become evident either in the text itself or in the guiding features of text production in box B (especially in the mental denotation model), the requirements from the dimension of correctness primarily refer to the mental denotation model and the encoding in the text itself. Violations against conventions (violations against the mental convention model), against legal requirements and author’s guidelines as well as an unsuitable medium will become visible in the mental denotation model or the encoding in the text so that the requirement of correctness need not be formulated for these components separately. The mental denotation model and the encoding in the text, however, have to be checked for correctness separately, although the mental denotation model is reflected by the encoding in the text. The reason for this is the following: Inexperienced writers often start from a correct mental denotation model but are unable to exteriorize it without distorting it. If such writers are made aware of the distortions, they often explain that they had meant the right thing (i.e., that their own mental denotation model had been correct) but that they were unable to express it in such a way that the text conveyed the right message and induced the creation of an appropriate denotation model in the reader’s mind (this is what my technical communication students often explained to me). There are also cases, however, where the author’s mental denotation model was wrong from the beginning, and even cases in which both an unsuitable mental denotation model and distortions in the encoding phase occur. Pinpointing the origin of the defects is important from a pedagogical perspective. If the mental denotation model was wrong already, the author can only solve the problem by acquiring more subject knowledge. If the mental denotation model, however, was correct and the mistakes have occurred in the encoding phase, the problem can be tackled by conveying linguistic and semiotic knowledge. c) Motivation What is covered by the comprehensibility dimension of motivation is only the text-induced type of motivation, i.e., the motivation aroused by the text itself. Requirements that can be derived from this dimension are that a text must, first of all, attract the reader’s attention. This applies to popular-science articles, for example, but not so much to operating instructions, which readers are automatically forced to read if they do not know how to handle a product. Furthermore, the attention that a text has attracted must be kept alive. This 70 Part II: Text Comprehension and Text Comprehensibility requirement, too, applies to popular-science articles but not so much to operating instructions. Like the requirements from the dimensions of concision and correctness, the requirements derived from the dimension of motivation, too, refer both to the mental denotation model and to the encoding in the text. Motivation at the level of the mental denotation model can be aroused by exemplifying and illustrating things by means of examples from the target group’s personal experiences. This is the case in the following example from an information leaflet informing potential future students about a new program in sensor systems technology: [4-5] If your father drives his Mercedes into a car wash and does not leave it with a Fiat Panda, this is due to sensor systems. This example from everyday life makes the reader aware of the fact that we encounter sensors everywhere in our lives. It is unlikely that motivation for reading the text would be aroused by introducing the program of sensor systems technology with a lexicon definition of sensor systems such as “the object of sensor systems technology is the development of intelligent sensor/ actor systems”. In instructive texts, motivation at the encoding level can be aroused, for example, by addressing the readers directly instead of using impersonal constructions: Avoid: This chapter deals with creating tables in Word . Use instead: This chapter explains how to create tables with Word . Or: In this chapter, you will learn how to create a table with Word . Another option to arouse motivation at the encoding level is the use of comic strips, especially for young readers who like this format. In some cases, arousing motivation may lead to longer texts. In such cases, the author has to decide which of the two dimensions, ‘concision’ or ‘motivation,’ plays the more important role with regard to the communicative function of the text. Furthermore, we have to take into account that the ways in which motivation can be aroused are culture-specific and therefore may require adaptive measures during the translation process (cf. Göpferich 2002: 169 ff.). d) Structure In contrast to the educational psychologists, whose dimension ‘organization - structure’/ ‘cognitive structure’ comprises both the content structure and the graphical and typographical design of a text, the dimension ‘structure’ in the Karlsruhe concept refers to the content structure only. Like the requirements from the dimensions ‘concision’, ‘correctness’ and ‘motivation’, the requirements derived from the dimension ‘structure’, too, refer 4 Text-processing from the perspective of instructional psychology 71 both to the mental denotation model and to the encoding in the text. The mental denotation model must have been broken down into adequate components (schemata) which, in the course of the text, must be joined together in an appropriate sequence. The structure of a mental denotation model can only be accessed via the encoded text in which it becomes evident. To differentiate between the global structure of a text, which results from the features of the objects, events, processes, etc. described in it, and a local structure, which is determined by grammatical requirements (among other things), the two concepts macro-level structure and micro-level structure will be introduced here: The term macrolevel refers to the level which encompasses more than two adjacent sentences (usually the level of passages and longer units). The micro-level is confined to two adjacent sentences. It seems legitimate to associate the macro-level with the mental denotation model and the micro-level with the encoding in the text because readers can remember the exact wording and grammatical structure (the encoding) of a maximum of two sentences they have just read, whereas they forget the wording and structure of sentences which are further away; what they can remember here is only the mental denotation model conveyed by them. Within the macro-level structure a distinction can be made between the content structure, which refers to the objects, processes, events, etc. dealt with in the text ( object structure , cf. recommendation 1 below), and the metacommunicative structure , which is created by meta-communicative elements such as advance organizers and statements concerning the reception situation such as You may have known this function already. (cf. the recommendations 2 and 3 below). Strategies which may help to fulfil the requirements to be derived from the dimension of structure at the macro-level are: 1. The splitting up of complex actions into individual steps must be adapted to the target group’s prior knowledge. 2. Before describing how something is done, technical authors should describe exactly what the result will look like. This helps readers to construct a rough mental model of their goal against which the individual steps can be interpreted much better. This can be achieved by illustrations (such as screenshots) showing the result. This ensures that readers always know what they are doing and that they can find out whether they are still on the right track. System reactions should be described after every major step. 3. The less conventionalized a genre is, especially with regard to its macrostructure (cf. Göpferich 1995: 217 ff.), and the more leeway authors have in their text composition, the more important it is to introduce advance or- 72 Part II: Text Comprehension and Text Comprehensibility ganizers . They control the readers’ expectations and help them to structure the information conveyed in the text. On the micro-level, the order in which the individual concepts and schemata are conveyed in the text and their logical relation to each other (indicated by conjunctions, adverbial phrases, etc.) have to be taken into account and evaluated. Features which play an important role at this level are the functional sentence perspective (see Section 1.2.2.4) and the logical linking of sentences. Strategies which may help to fulfil the requirements to be derived from the dimension of structure on the micro-level are: 4. Information (especially instructions) should be given in the order in which the user needs them: Avoid: Loosen X after discharging Y. Use instead: Discharge Y, then loosen X. 5. For the reasons mentioned under item 4, conditional clauses should precede the main clause in instructive texts: Avoid: Press [ESC] when you want to leave the program without saving your data. Use instead: When you want to leave the program without saving your data, press [ESC]. 6. Begin sentences with thematic elements (what is already known) and put rhematic elements (the new information) in the end position unless you want to express a contradiction or make a very emotional remark. Avoid: Vielleicht haben Sie sich schon gewundert, warum das eine Auto schon nach einigen zehntausend Kilometern neue Bremsbeläge braucht, während sie bei einem anderen scheinbar ewig halten. Fahrweise und Einsatzort beinflussen stark den Verschleiß der Bremsbeläge. Use instead: ... Der Verschleiß der Bremsbeläge wird stark von der Fahrweise und dem Einsatzort beeinflusst. The topic of the first sentence is the wear of brake linings. This wear can be put in thematic position in the second sentence. In the criticized version above, however, it also appears in rhematic position in the second sentence although it is no longer new. This has been corrected in the optimized version. e) Simplicity In contrast to the dimensions of concision, correctness, motivation and structure, which refer to both the mental denotation model as an analogous representation of what is conveyed in the text, and to its encoding as its digital representation, the dimension of simplicity refers to the encoding in the text only. 4 Text-processing from the perspective of instructional psychology 73 The simplicity of the mental denotation model is covered by the dimension of concision. In the educational psychologists’ comprehensibility concepts, the dimension of simplicity or linguistic simplicity refers to the lexis and syntax only. To determine which words and sentence constructions can be regarded as simple, a frame of reference is needed, which is not provided by the educational psychologists. In my framework, this frame of reference is formed by all components which directly or indirectly (via other components) determine the text, i.e., its communicative function (box A) and the guiding features of text production (box B). Questions which have to be answered for assessing a text’s simplicity are: • Is the choice of words adequate (lexical simplicity)? - What has to be assessed here is whether the unexplained terms and abbreviations that occur in the text can be regarded as familiar to the target group and as appropriate for the genre. Furthermore, we have to check whether the specialized terminology that has been introduced for reasons of economy but cannot be regarded as familiar to the audience has been explained sufficiently. If several synonymous expressions for a concept exist, we have to make sure that the expression that the target group is most familiar with is used. In a research report: Avoid: The fish were observed to exhibit a 100% mortality response. Use instead: All the fish died. • Is the syntax adequate (grammatical simplicity)? - Here we have to answer questions such as: Can sentence complexity or the number of hypotaxes be reduced without violating the text’s communicative function, its genre conventions or the requirements from the other comprehensibility dimensions? Can nominalizations be transformed into more verbal constructions? Can passive-voice constructions be transformed into active-voice constructions? Can negative sentences be transformed into affirmative ones? Will such measures make the text more readable for its target group? Verbal constructions instead of nominalizations: Avoid: Through the introduction of measures aimed at the creation of jobs, it is the intention of the commission to facilitate the improvement of the economic and social situation. Use instead: By introducing measures to create jobs, the Commission intends to help improve the economic and social situation (adapted from European Commission Translation Service n.y.) 74 Part II: Text Comprehension and Text Comprehensibility Active voice instead of passive voice: Avoid: New guidelines have been laid down by the President in the hope that the length of documents submitted by DGs will be restricted to 20 pages. Use instead: The President has laid down new guidelines in the hope that DGs will restrict the length of documents to 20 pages. (European Commission Translation Service n.y.: 5) Affirmative sentences instead of negative ones: Avoid: It is not uncommon for applications to be rejected, so do not complain unless you are sure you have not completed yours in correctly. Use instead: It is quite common for applications to be rejected, so complain only if you are sure you have completed yours correctly . (European Commission Translation Service n.y.: 9) In addition to these questions which are also asked by the educational psychologists in connection with their dimension of (linguistic) simplicity, the dimension of simplicity in the Karlsruhe comprehensibility concept also covers the following questions: • Is the degree of directness which is achieved on the illocutionary level adequate for the genre? - Maximum directness on the illocutionary level is not always the best option. Letters with negative feedback on an application, for example, require indirectness through the use of hedging (cf. example [4-2] above); in instructive texts, however, indirect instructions must be avoided, as in the following example: Avoid: The door is opened by pressing the switch. Use instead: Open the door by pressing the switch. • Are the words and constructions used precise enough, i.e., is ambiguity avoided? - Ambiguous terms, illocutionary indicators and grammatical constructions should be avoided. • Are lexis and syntax used consistently? - Here we have to take into account that some genres, such as popular-science articles, require elegant variation for reasons of motivation, whereas it may be misleading in genres such as instructive texts. In these texts, elegant variation also has a negative effect on the efficiency with which translation memory systems can be used to translate them. f) Perceptibility This comprehensibility dimension covers those features which determine the ease with which texts can be perceived with our senses and thus made accessible to our cognitive systems for further processing as well as the features which support the reader’s recognition of content structures nonverbally (cf. the dimension of structure in Section 4.3.3 d). It refers to layout and design 4 Text-processing from the perspective of instructional psychology 75 (macro-typography) , the fonts used and other paraverbal features (microtypography) as well as nonverbal elements. Thus, it includes the features which are the object of legibility research, and additionally also includes the features which determine the ‘legibility’ and ‘readability’ of nonverbal elements (cf. the Gestalt Laws in Göpferich 1998: 55 f.) as well as those elements of the formal text structure which educational psychologists group under the dimensions of ‘organization - structure’ and ‘cognitive structure’ respectively (such as the use of bullets in enumerations). Like the dimension of simplicity, the dimension of perceptibility refers to the encoding only. 4.3.4 Completeness of the Karlsruhe comprehensibility concept? According to the Karlsruhe comprehensibility concept, ideal comprehensibility depends on six dimensions of a text: simplicity, structure, correctness, motivation, concision and perceptibility. Here the question arises whether this enumeration of dimensions is exhaustive. In the comprehensibility study which will be described in the next section and which used the target-groupcentred method of optimizing reverbalization with think-aloud (cf. also Göpferich 2006a, b, c), the participants’ criticism of the text to be assessed and the optimization maxims and strategies 3 they employed revealed what they considered to be the basic ‘ingredients’ of comprehensibility. A comparison of these ‘ingredients’ of comprehensibility with the six dimensions of the Karlsruhe comprehensibility concept revealed that there were no items of criticism that could not be attributed to one of the dimensions of the Karlsruhe concept. This suggests that the comprehensibility concept underlying the Karlsruhe model matches at least the intuitive comprehensibility concepts of the participants in the experiment. Comprehensibility assessments may lead to the wrong impression that a text which is comprehensible must also be usable (cf. Krause 1991: 396 f.). Especially for instructive texts, which tell the reader how to do something, optimal legibility, readability and comprehensibility are no guarantee for optimal usability. In addition to the three requirements mentioned above, these texts must also be complete, correct, conform to legal requirements with regard to form and contents (cf. Göpferich 1998: Chapter 11), save reading time (reception economy), provide quick and selective access to the information needed 3 The term maxim refers to goals of action the participants strive for at a specific point in the optimization process. The term strategy is used in the sense of Faerch & Kasper (1983: 36), who define strategies as “potentially conscious plans for solving what to an individual presents itself as a problem in reaching a particular communicative goal” (cf. Krings [1986: 175], who also adopts this strategy definition). Thus, maxims are targets whereas strategies are potentially conscious plans that the participants believe will lead to these targets when problems occur. 76 Part II: Text Comprehension and Text Comprehensibility by the reader (via a table of contents, an index and a user-friendly layout) and help the reader to discern descriptions from instructions. These additional requirements are also covered by the comprehensibility concept presented in this chapter: The requirement of completeness of content has been taken into account insofar as the dimension of concision does not only refer to the encoding in the text, but also to the mental denotation model. The correctness requirement has been taken up in the dimension of correctness. Conformance to legal requirements is covered by the ‘legal requirements and author’s guidelines’ in the text production guiding features. Reception economy is taken into account by requiring concision for the encoding of the text. Whether quick information access is possible is evaluated in the dimension of perceptibility, and speech-act theoretical aspects have been taken into account in the dimension of simplicity. Since the main evaluation criterion in the Karlsruhe comprehensibility concept is whether a text fulfils its communicative function (or skopos) and since text quality can be defined as the degree to which a text fulfils its communicative function, the framework presented here cannot only be used for text comprehensibility assessment but for all kinds of text quality assessment including translation quality assessment in the functionalist paradigm (for examples of the application of this framework, cf. the analyses by Göpferich 2002: Chapter 4.7.3). Although the Karlsruhe comprehensibility concept has proved a reliable instrument in pre-optimizing non-instructive texts (cf. Göpferich 2002: 154 ff.), it cannot replace target-group-centred empirical research into text comprehensibility. As Schriver (1989: 247) points out: “[E]xpert-judgement-focused methods should not be used in isolation; they need to be supplemented with other document evaluation methods, particularly those which are readerfocused.” Such methods will be discussed in the next chapter. 5 Methods of comprehensibility assessment Following Schriver (1989), methods of text comprehensibility assessment can be classified into three categories: “text-focused methods”, “expert-judgementfocused methods” and “reader-focused methods”. An example of a textfocused method is the application of readability formulas (see Section 2.2). These are popular because they can easily and quickly be applied by means of computer programs. However, they do not give us deeper insight into the comprehensibility of the texts to which they are applied because they take into account only certain lexical, syntactical and stylistic aspects of what makes a text comprehensible or incomprehensible. The use of comprehensibility concepts as presented in Chapter 4 pertains to the category of expert-judgementfocused methods. The Karlsruhe concept, for example, has proved a reliable instrument in pre-optimizing non-instructive texts (cf. Göpferich 2002: 154 ff.) but cannot replace reader-focused, or, to be more precise, targetgroup-focused methods of comprehensibility assessment (cf. Schriver 1989: 247), which yield the most reliable results. The term target-group-focused is preferred to the term reader-focused because texts need not be comprehensible for all their readers but only for those of their intended audience or target group. One such target-group-focused method, the method of optimizing reverbalization with think-aloud, keystroke logging and, additionally, screen recording and eye tracking will be focused on in this chapter, which is based on Göpferich (2006a, b, c). 5.1 Target-group-focused methods of comprehensibility assessment Target-group-focused methods undoubtedly provide the least speculative and most reliable results on text comprehensibility because this is a relative text quality, which depends on the audience, whose comprehension and comprehension problems are central for text evaluation. Schriver (1989: 241) emphasizes that an optimal text evaluation method should provide two types of information: “(1) information about whole-text or global aspects of text quality, and (2) information about how the audience may respond to the text”. A target-group-focused method of comprehensibility assessment employed for instructive texts is usability testing (cf. Rubin 1994; Byrne 2006; Nielsen/ Pernice 2009). Usability testing provides reliable results for this type of texts because they make their readers do things. These actions are exteriorizations of how the readers have comprehended the text and make their comprehension observable for others. The comprehensibility of non-instructive texts, such as popular-science texts, however, is harder to analyze. Target-group-focused 78 Part II: Text Comprehension and Text Comprehensibility methods employed to test their comprehensibility are cloze procedures (i.e., fill-in-the gap tests), questions on texts and reproductions of their contents. These methods have the disadvantage, however, that they measure either only aspects of text comprehensibility (e.g., the predictability of words and phrases that fill gaps in cloze procedures, the comprehensibility of words or passages relevant for answering the questions asked) or merely their rough overall comprehensibility. Furthermore, some of these methods lead to a confusion of the concepts of comprehensibility and retainability (cf. the research review in Schriver [1989: 244 ff.] and in Göpferich [2002: Chapter 4]). From these insights, requirements can be derived for an ‘ideal’ method of determining the comprehensibility of non-instructive texts: 1. It should be target-group-focused. 2. It should cover not only the text as a whole but also all its components. 3. It should elicit exteriorizations of the results of comprehension that members of the target group are gaining while reading the text. In the following, methodological reflections will be provided on each of these requirements. The requirements 2 and 3 will be dealt with first. 5.1.1 Exteriorization of comprehension for the text as a whole and for its components To determine whether a text is comprehended in the way intended by its author, we need insight into what reading it elicits in the audience’s short-term or working memory. For non-instructive texts, the only methods which give us this insight are introspective ones. If we do not only want to determine a reader’s global text comprehension (or a text’s overall comprehensibility) but also the reader’s comprehension (or the comprehensibility) of every component of it, the online-method of think-aloud appears to be a viable option. It elicits what is going on in the reader’s mind while reading the text (and working with it) in a direct and unfiltered way out of his or her short-term memory (“level 1 verbalizations” according to Ericsson/ Simon 1993). Since the publication of Ericsson & Simon’s book Protocol Analysis: Verbal Reports as Data in 1984 (Ericsson/ Simon 1993), think-aloud has found its way into the exploration of writing processes (cf., e.g., Schindler 2004; Göpferich/ Nelezen 2013; 2014) and (interlingual) translation processes (cf. Krings 2005; Englund Dimitrova 2005; Hansen 2006; Göpferich 2008). The application of think-aloud in comprehensibility research is not new either. In the past, however, this method was only used here either in readability research to determine text comprehension during reading processes, or it was used during revision processes (for a review of such studies cf. Schriver 1989: 249). 5 Methods of comprehensibility assessment 79 Think-aloud during revision processes has the disadvantage that this is not a reader-focused but an expert-judgement-focused method and the revisers have to speculate about the text’s comprehensibility for its audience. Think-aloud during mere reading processes has a disadvantage, too: As translators know from experience, text reception is much more profound when a text is not just read for reasons of interest but in order to translate it into another language or to paraphrase it (cf. also Shreve et al. 1993). Working with texts in this way often makes obstacles to comprehension arise which the reader would never have realized without the additional text processing task. For this reason, in text comprehensibility studies with think-aloud, reading a text should be combined with an additional processing task such as optimization (optimizing reverbalization). Such an optimization can be considered a type of intralingual translation. It takes into account every detail of the texts to be analyzed, is independent of the texts’ retainability and does not only allow the researcher to detect where the texts are incomprehensible, but also where they are hard to understand and where they give rise to misunderstandings or a demand for further information which is not given in the texts. 5.1.2 Selection of participants To determine the comprehensibility of a text by means of optimizing reverbalization, it seems to be advisable to choose a group of participants which is representative of the text’s target group. This, however, is problematic if the target group also includes people with a lower level of education. Empirical studies have shown that people with a higher level of education have a higher metacognitive competence (on the concept of metacognition, cf. Flavell 1976: 232). This includes their metacognitive reading competence. An important component of this competence is comprehension monitoring. This is composed of two sub-processes: comprehension evaluation, i.e., “evaluating the current state of one’s ongoing comprehension” (Baker 1985: 155), and comprehension regulation, which “comes into play when the reader has evaluated his or her understanding and found it inadequate” (Baker 1985: 155). For comprehension evaluation a range of different criteria is applied. In her framework for the differentiation of these criteria, Baker (1985: 156 ff.) distinguishes the following: “lexical criteria”, which refer to comprehensibility at the level of individual words; “syntactic criteria”, which refer to grammatical correctness within sentence boundaries; and “semantic criteria” (a more correct term would be text-linguistic criteria), which require “consideration of the meanings of individual sentences and the text as a whole” (Baker 1985: 156). She derives these criteria from both theoretical and empirical work on the cognitive processes involved in comprehension and comprehension monitoring. Since the last criterion is the most crucial to effective text comprehension, Baker subdivides it into five subcategories, which she describes as follows: 80 Part II: Text Comprehension and Text Comprehensibility “(1) propositional cohesiveness, checking that the ideas expressed in adjacent propositions can be successfully integrated; (2) structural cohesiveness, checking that the ideas expressed throughout the text are thematically compatible; (3) external consistency, checking that the ideas in the text are consistent with what one already knows; (4) internal consistency, checking that the ideas expressed in the text are consistent with one another; and (5) information clarity and completeness, checking that the text clearly states all of the information necessary to achieve a specific goal.” (Baker 1985: 156 f.) People with higher reading competence apply a wider range of these criteria to their comprehension evaluation (Baker 1989: 15). If we start from the assumption that people with a higher reading competence also have a higher communicative competence, their think-aloud protocols can be expected to be more illuminating as this allows them to give more detailed and precise descriptions of their comprehension problems and optimizing maxims and strategies 1 than persons with a lower level of education. This was confirmed in experiments by Rooken (1993) and Rijnders (1993), in which more educated people produced more feedback. For this reason, it may be prudent to select participants which belong to the target group of the texts to be evaluated but to give preference to people with a higher level of education. To minimize the problem of neglecting comprehension and comprehension problems of people with a lower level of education, the participants can be asked to optimize the text for the intended readership as a whole (and not only for themselves). This causes them to optimize sections they find comprehensible for themselves but consider incomprehensible for people with lower education. Their judgement on such elements of the text must be considered speculative. We must not forget, however, that people with a lower standard of education might find additional things incomprehensible that were not criticized or optimized in the experiment. Furthermore, they may not realize that they have not understood certain aspects. Optimizing texts with regard to these problems as well, however, cannot be detrimental to comprehensibility for people with lower education. Moreover, the comments by participants with a higher level of education can at least provide questions which can be used to find out how well participants with lower educational standards have understood certain passages of the text. Asking such questions is important because empirical studies have shown that participants with lower education (such as children) do not become aware of certain comprehension problems unless they are explicitly asked for them (cf. the research review by Baker 1985). In a pilot study that will be described in the following sections, five participants with a relatively high level of education and a high communicative com- 1 For a definition of the terms maxims and strategies , see FN 3 on p. 75. 5 Methods of comprehensibility assessment 81 petence (students in the degree program “Translation and Interpreting” at the Department of Translation Studies of the University of Graz, graduates from this program and a lecturer there) had to think aloud while optimizing a popular-science text on diabetes mellitus for a wide audience (Göpferich 2006a, b, c). In this pilot study, all seven criteria between which Baker distinguishes were applied by the participants. This is an indicator of both the participants’ metacognitive, metalinguistic and metacommunicative competence (cf. Baker 1989: 28, 33 f.) and the depth of text processing that reception for optimizing reverbalization requires. Experiments which require less text processing depth, such as the application of the plus/ minus-method, where participants simply have to put plus and minus signs in the margin of a text depending on whether they find a text section comprehensible (+) or hard to understand (-), provide much less feedback (cf. Pander Maat 1996: 29). Pander Maat (1996: 28) explains this as follows: “ [ A ] s long as there is no need to use a certain piece of information, there is no need to comprehend at a more than superficial level.” The optimization task forces the participants into more profound text processing; it forces them to ‘use’ the information in the text. The fact that one of Baker’s criteria was applied to every item criticized by the participants of the diabetes study and that no additional criteria were used suggests that Baker’s framework comprises all relevant criteria. 5.2 Research design Five participants, who belong to the target group of the popular-science text on diabetes reproduced in Fig. 5-1 were asked to reverbalize this text using the keystroke-logging software Translog in such a way that the result was optimally comprehensible for its target group. The target group is specified as follows on the website where the text to be optimized appears: “These contributions provide basic information on diabetes mellitus for which no prior knowledge is required.” (Deutsches Diabetes-Zentrum 2004; my transl.). During the experiment, the participants had to think aloud (“level 1 verbalizations” according to Ericsson & Simon 1993: 79). All participants were female and either students in the degree program “Translation and Interpreting” at the Department of Translation Studies of the University of Graz, had graduated from this program or were lecturers there. Their mother tongue was German; four of them were Austrians and one (YG) came from Switzerland. A short description of the participants’ educational and professional backgrounds is given in Tab. 5-1. Tab. 5-2 provides information on their physical and psychological condition during the experiment as described by themselves. 82 Part II: Text Comprehension and Text Comprehensibility Fig. 5-1: The popular-science text to be assessed 5 Methods of comprehensibility assessment 83 Tab. 5-1: The participants’ educational and professional background 8 Age (years) Academic degree/ profession Foreign languages and proficiency (selfestimation: A = excellent down to E = poor) Additional training/ experience Proficiency as a student (selfestimation) JS 25 student 2nd year English (A) Italian (A) Spanish (C) French (E) Russian (E) Dropped out of program in translation and interpreting; new start with different language combination 20 pages of translations; 40 hours of interpreting excellent - good EK 25 student 6th year French (A) Hungarian (B) English (C) 30 pages of translations; 10 days of interpreting excellent - good NL 25 student 7th year English (A) Russian (B) French (D) Latin (D) Italian (E) 4 pages of translations; 5 days of interpreting excellent - good YG 26 Mag. phil. (Master) employed in thirdparty funded research project (for 6 months) English (A) Russian (B) 8 months student assistant 20 pages of translations per year excellent - good SF 58 Mag. phil. (Master) university lecturer (for 20 years) English (A) French (B) Italian (E) 4 years secretary; 30 years of translation and interpreting no answer 84 Part II: Text Comprehension and Text Comprehensibility Tab. 5-2: Physical and psychological condition of the participants during the experiment (1 = applies very much, 2 = applies to some degree, 3 = applies not at all) Participant Fully concentrated Under time pressure Nervous Relaxed Stressed Tired, exhausted Enjoying the job Comments JS 1 3 3 1 3 1 1 — EK 2 2 3 2 3 2 1 Suffering from a headache; under time pressure during the retrospective interview NL 1 3 3 2 3 3 1 — YG 1 3 2 2 3 3 1 — SF 1 3 3 1 3 3 1 — Although all participants belonged to the target group of the text to be assessed, they were not representative of this target group. As can be seen from Tab. 5-1, their education and training was above that of the average reader. This means that whatever was incomprehensible for them can also be regarded as incomprehensible for the intended readership in general. Since the participants were asked to optimize the text for the intended readership (and not only for themselves), they may have optimized sections of the text that they have found comprehensible for themselves but considered incomprehensible for people with lower education. To reduce subjectivity, it was taken into account how many of the five participants judged elements of the text to be optimized incomprehensible or hard to understand. Each participant had to reverbalize the text in Fig. 5-1 in such a way that it would be tailored to the requirements of its intended readership. Passages that the participants considered perfect could be copied into the target version. Prior to the actual experiment a trial run with a different text was carried out 5 Methods of comprehensibility assessment 85 „Definition des Diabetes mellitus“ Source and function of the text The text you will have to work on was taken from the website www.diabetesdeutschland.de issued by the German Diabetes Research Centre (Deutsches Diabetes Forschungsinstitut - DDFI) at the University of Düsseldorf. The audience of the texts on this website, which also includes the text you will have to work on, is described as follows by the editor: „Diese Beiträge geben Ihnen grundlegende Informationen zum Diabetes mellitus, ohne dass Sie eigenes Vorwissen benötigen.“ (The following contributions provide basic information on diabetes that you will be able to understand without possessing any specific prior knowledge; my transl.) Assignment Please reverbalize the text in such as manner that, to your mind, it will optimally meet the requirements of the audience specified above. Sections that you find ideal can simply be copied into your reverbalized version. Whatever you feel could be improved, should be optimized. In doing so, please also observe spelling rules and rules for paragraph formation. For software-technical reasons, please do not use other font sizes and colours, no italics, bold type face or underlinings. In the Translog software, in which you will work, a small dictionary for the text will be provided. To access its entries, please proceed as follows: 1. Click on the dictionary icon. 2. Click on the word for which you need additional information. Translog only possesses a restricted editing functionality. The following functions are available: • arrow keys • Backspace key • Delete key • Return key • tabulator key • Cut, Copy and Paste via the menu bar • Home (Pos 1), End (Ende) (jumping to the beginning or end of a line respectively) • Control Home (Strg Pos 1), Control End (Strg Ende) (jumping to the beginning or end of the text respectively) • Page Up (Bild ↑ ), Page Down (Bild ↓ ) • mouse Please think aloud while working on the text. Fig. 5-2: The assignment 86 Part II: Text Comprehension and Text Comprehensibility to acquaint the participants with the functionality of Translog (editing functions and Translog dictionary). Only after all the questions on the software and the test setting had been answered was the actual experiment begun. The assignment was explained to the participants by the supervisor and also handed out to them in writing (see assignment in Fig. 5-2). 2 The dictionary entries provided with the source text in Translog are reproduced in Fig. 5-3. No other material could be used during the experiment. Fig. 5-3: The Translog Dictionary During the experiment, each participant was sitting in a quiet room 3 together with an experimenter. The participants wore headsets; their verbalizations were recorded with the freeware Audacity and exported in MP3 format. The recordings were transcribed according to the GAT conventions for basic transcripts (Selting et al. 1998) and then proof-read by at least one other person. The complete transcripts of all five participants can be downloaded as a PDF file from Göpferich (2005). During the experiment, the participants were not put under time pressure (cf. Tab. 5-2). They were informed that what was analyzed in the experiment was not their competence but the comprehensibility of the diabetes text. After the optimization process, the participants were asked whether they had any questions for a diabetes specialist which had cropped up during the experiment and which they could not answer using the information in the text and the Translog dictionary. These retrospective interviews were recorded and transcribed, too. They can be found at the end of each of the transcripts in Göpferich (2005). 5.3 Data analysis Apart from the information given in Tab. 5-1 and Tab. 5-2, the experiment yielded the following data: 1. the optimized versions of the diabetes text, 2 The participants, who were all native speakers of German, were given the assignment in German. The text above is a translation. 3 Any disturbances which occurred during the experiment can be found in the transcripts (cf. Göpferich 2005). 5 Methods of comprehensibility assessment 87 2. the log files, 3. the think-aloud protocols (TAPs) as well as the protocols of the retrospective interviews (RIPs; cf. Göpferich 2005). The data were analyzed as follows: 1. Each participant’s version was compared to the original text. Passages in which changes had been made were numbered and juxtaposed to their original version in a table. For each change the comprehensibility dimension in which this change occurred according to the Karlsruhe comprehensibility concept (see Section 4.3) was determined as was whether the change really improved the text, made it worse, or represented neither an improvement nor a deterioration. 2. The TAPs were analyzed for comments on why the changes had been made as well as on the maxims the participants followed and the strategies they used in optimizing the text. 3. The TAPs were analyzed for further comments on the quality of the original text which did not result in changes in their optimized versions. 4. The questions which the participants had on the text, which they could not answer using the material available to them (i.e., the text to be optimized itself and the Translog dictionary) and which they therefore would have liked to ask a specialist were collected. If we start from the assumption that an optimally comprehensible text does not give rise to questions in the reader’s mind that it does not answer, these remaining questions are additional indicators of deficiencies with regard to comprehensibility. This also applies to dictionary consultations. 5. For each element of the original text that had been subject to criticism or questions in the experiment, it was determined how many participants had criticized it or had questions on it (cf. Tab. 5-3 in Section 5.5 below). The more participants commented on it, the higher the probability that it may really lead to comprehension problems. 6. An optimized version was written in which the criticism of all participants was taken into account and which answers the questions they had. In this optimized version, only real improvements suggested by the participants were considered. If a participant formulated a maxim for a specific section of the text without providing a solution fulfilling this maxim, I tried to provide such a solution by myself. Deteriorations and ‘cosmetic’ changes were ignored. Linguistic mistakes made in the source text or by the participants were corrected in the optimized version (see Tab. 5-4 in Section 5.5 below). 7. The participants’ maxims and strategies were analyzed and classified. For each participant the repertoire of maxims she had and strategies she used were determined. From these results, conclusions for text production di- 88 Part II: Text Comprehension and Text Comprehensibility dactics can be drawn. The participants’ maxims and strategies, however, will not be dealt with here. Some of them will be mentioned; a detailed analysis of them is provided in Göpferich (2006c). 5.4 Results In the following, a survey will be given of all the elements of the original text that either one or more participants in the experiment considered hard to understand or incomprehensible, as well as of the questions the text gave rise to in the participants’ minds without providing an answer. For each element, extracts from the TAPs and/ or RIPs will be quoted which show that the participant(s) found it difficult and, if applicable, what maxims and strategies they followed to improve the corresponding passage. A distinction is made between a) completely incomprehensible elements of the source text and passages giving rise to questions that are not answered in the text and b) passages which are simply hard to understand. Incomprehensible elements and missing information 1. The text says that a distinction is made between two types of diabetes, but then three types are introduced: type 1 diabetes, type 2 diabetes and gestational diabetes. This makes four of the five participants wonder whether gestational diabetes is a variant of type 2 diabetes (cf. TAP YG 29-30) or a type of its own. Thinking aloud, YG says, „wenn i jetzt wüsst, (-) ob (.) der (.) schwangerschaftsdiabetes, (.) ob de: : s jetzt (.) typ drei isch ((tippt öfters auf die Tastatur ohne zu schreiben)) odersch, (--) typ=zwei (-) <<sich selbst beim Tippen diktierend> werden generell, im allgemeinen> (-) woteva (--) <<sich selbst beim Tippen diktierend> im allgemeinen zwei typen unterschieden>“ (TAP YG 82-88). She solves the problem by adding “im Allgemeinen” ( In general , a distinction between two types of diabetes is made) and referring to gestational diabetes as a “Spezialfall”, a special type of diabetes. The log file reveals that she first puts down “zwei Typen” (two types of diabetes), then changes this into three types, and uses “zwei Typen” again in her final version. In the retrospective interview (RIP YG 476-482), she says that she would like to ask a specialist about this because she is still not sure which one is the correct version. To my mind, the real problem in the text here is that the author informs us about what happens in the bodies of patients with type 1 diabetes and type 2 diabetes, but not about what happens in the bodies of women with gestational diabetes. If this information were given, it would be clear that gestational diabetes is neither a variant of type 1 nor of type 2. In contrast to type 1 diabetes, which occurs when the body produces too little or no in- 5 Methods of comprehensibility assessment 89 sulin, and type 2 diabetes, which occurs when the body cannot use the insulin it produces, gestational diabetes is caused when pregnancy hormones and hormones produced by the placenta lead to an increase in the blood glucose level that the pregnant woman’s pancreas can no longer compensate for by an increased insulin production. Adding this information together with the coherence increasing elements in general (im Allgemeinen) and a special type of diabetes (ein Spezialfall der Zuckerkrankheit) eliminates this incomprehensibility. 2. Some of the participants do not know the (exact) meanings of the following terms used in the text without explanations: chronisch (TAP SF 74-78; TAP EK 26-29), Insulin (TAP SF 261-289; TAP EK 224-232), Inselzellen (cf. TAP YG 159-161; TAP NL 129-134; TAP JS 84-106 and 177-185; TAP EK 66-67 and 265-268), T-Lymphozyten (cf. TAP JS 267-283 and 652-655), Gestation (cf. RIP SF 534-540; TAP YG 133-134; TAP EK 329- 331), Glukose (cf. TAP SF 303-312; TAP YG 200-206; TAP NL 232-233 and 243-246; TAP EK 135-162) and Körperzellen (cf. TAP JS 312-320). To solve this problem, these terms must either be left out or explained. They should be retained and explained if they are used in doctor-patient conversations (cf., e.g., TAP NL 98-101), but can be deleted if this is rather unlikely and they are not needed again in the text (cf. RIP JS 609-616). 3. For YG, the last sentence in the text “Dabei ist jedoch das Risiko für die spätere Entwicklung eines Typ 2 oder Typ 1 Diabetes [sic] stark erhöht.” gives rise to the question whether this refers to the mother or the child: “beim kind oder bei der mutter? ” (TAP YG 34). That it refers to the mother can easily be made more explicit here. 4. For JS (cf. TAP JS 493-516), the original text does not make clear whether type 1 diabetes is caused only by a combination of all three factors mentioned in the text (i.e., first, genetic predisposition, second, influences from outside such as certain virus infections, and third, a disorder of the immune system) or whether it may be caused by one or two of these factors alone. Although the combination of all three factors seems more plausible to her, she combines the factors by or in her optimized version, which shows that she is still not sure. Since the factors that cause the disease provide important information for the patient, the text must be made more explicit here. What makes the text hard to understand 1. Since Diabetes mellitus is a specialized term for which there is the more general and thus comprehensible expression Zuckerkrankheit in German, YG, SF and EK feel that this more comprehensible designation should be added already in the title of the article (cf. TAP SF 41-44, cf. also TAP YG 90 Part II: Text Comprehension and Text Comprehensibility 40-44). This makes sure that the reader knows what the text is about from the beginning. 2. SF feels that the semantic relation between the general designation Zuckerkrankheit and its scientific synonym Diabetes mellitus might not become clear if the latter is simply put in parentheses without additional explanations, and therefore should be made explicit: „<<sich selbst beim Tippen diktierend> krankheit, in der folge diabetes mellitus>, eigentlich das brauch ma (--) gor nix so wer ma des machen, (---) in der folge, nein das is keine gute idee (--) zuckerkrank (.) heit, des kommt a mal weg, (-) diabetes mellitus, (.) jawoll, na mach ma schon so, in der folge diabetes mellitus. (4.0) <<zustimmend> mhm> (---) nein, das müsst man auch besser erklären, weil ich weiß das ja was das is, aber jemand der nicht latein und nicht griechisch kann weiß das nicht, (.) in der folge (--) der ha lateinischen oder griechischen Bezeichnung ( ) wahrscheinlich, nein wissenschaftlichen <<sich selbst beim Tippen diktierend> wissenschaftlichen bezeichnung so, diabetes mellitus,>“ (TAP SF 55-67) She follows this maxim also when introducing other alternative designations such as juvenil , T-Lymphozyten , Alterszucker and Gestationsdiabetes (cf. TAP SF 167-169, 219-223, 381-384, 432-437). EK does not comment on this, but she too makes the relation between Diabetes mellitus and Zuckerkrankheit explicit (“Diabetes mellitus ist der Fachausdruck für Zuckerkrankheit”). 3. YG wonders about the term Definition in the title because, to her mind, the text provides more information on diabetes mellitus than just a definition (cf. TAP YG 344-349) so that the term does not fit. Furthermore, one may object that the term Definition is a hard word for many readers, which is a second reason for taking it out. (In spite of her objections, YG takes the term over in her final version.) EK deletes the term Definition in her title, but does not comment on this. 4. The participants YG and NL wonder whether “ist gekennzeichnet durch” is the correct expression in the definition of diabetes mellitus or could be replaced by a simpler verb. Both of them feel that the latter is the case (cf. TAP NL 44-82; TAP YG 568-632). 5. For YG, the first sentence does not make clear whether the participle construction “verbunden mit dem Risiko für schwere Begleit- und Folgeerkrankungen” (combined with the risk for serious other diseases which accompany or follow it) describes necessary features of the term diabetes mellitus and thus forms part of its definition, or gives just additional information. Thinking aloud, she comments on this: „<<sich selbst beim Tippen diktierend> erhöhung des blutzuckers, (.) verbunden mit dem risiko> (2.0) MO: MENT (3.0) oke=verbunden mit dem risiko des interess: iert mich hier eigentli net wirklich (5.0) verbunden mit dem risiko (3.0) aber die zuckerkrankheit is: t eigentlich nur eine chronische erhöhung des blutzuckers; 5 Methods of comprehensibility assessment 91 (---) UND wenn ma erhöhtn blutzucker: : , (.) wenn sich der erhöht (.) dann: : .hh können schwere beglEIT und FOLgeerkrankungen (--) FOLgen“ (TAP YG 45-52). In the retrospective interview, she comes back to this comprehension problem saying, „beim erstn satz hab i mi a bissl gwundert, (--) weil <<den Ausgangstext lesend> zuckerkrankheit (.) ist gekennzeichnet durch chronische erhöhung des blutzuckers verbunden mit dem RIsiko.> (--) is eigentlich (--) a: : (-) <<all> ahso> .h sog ma so- (.) i persönlich versteh des anders. für mi is zucker; zuckerkrankheit chronische erhöhung des blutzuckers (-) und danach (---) also (.) beziehungsweise (.) durch die erhöhung (--) kommen begleit- und folgeerkrankungen. aber es is net (3.0) zuckerkrankheit is net (2.5) gleich (-) begleit- (.) und folgeerkrankungen.“ (RIP YG 568-578) The information in the participle construction does not form part of the definition, which should be made clear in the optimized version. 6. Four of the five participants (SF, YG, NL, EK) are amazed that the term Zuckerspiegel (sugar level) is used in the plural; they have only heard of der Zuckerspiegel in the singular and wonder whether there are several sugar levels (cf. TAP YG 220-223; cf. also TAP SF 320-329). In fact, there is only one blood sugar level, so that the plural is wrong and must be changed into a singular. Even if there were several blood sugar levels, using the singular would be the option to be preferred in this context because a distinction between different sugar levels is not necessary in the text and wondering about the plural requires memory capacity which will then not be available for processing the central information on diabetes. If a differentiation were relevant, the plural should be introduced explicitly so that the reader need not wonder about it. YG opts for the plural because, as she says in the retrospective interview (RIP YG 498-507), the author of the text is an expert and should know what he is talking about. This is also the reason why SF uses the plural (cf. TAP SF 326-329). NL and EK prefer to use the more common singular (cf. TAP NL 264-268; TAP EK 209-210). 7. In the original version, information which belongs closely together such as a) the different designations of each type of diabetes, b) the age when type 2 diabetes occurs, c) the beginning and the end of gestational diabetes and d) the first mentioning of the destruction of the insulin producing cells and the detailed explanation of how they are destroyed and why are given at different places in the text. YG brings some of them together (a and c), however, without commenting on it in her TAP (information clustering maxim). EK brings the information under a) together (TAP EK 69-71). NL explicitly states that she wants to bring the information when type 2 diabetes occurs together (cf. TAP NL 345-354). SJ comments on d) saying, „<<sich selbst beim Tippen diktierend> beruht .hh auf einem (.) mangel an insu .hhh hh (--) infolge (-) einer zerstörung> .hhh hh (2.0) der sogenannten bet (2.0) ah das klingt glaub i komisch, wenn das da hinten so (.) .h der insulin produzier- 92 Part II: Text Comprehension and Text Comprehensibility enden zellen (2.0) .hh ja aber wodurch werden die zerstört? (2.0) <<den Ausgangstext lesend> diese zellen gehören zur bauchspeicheldrüse und sind ein bestimmter typ der sogenannten inselzellen. am höchsten ist die neuerkrankungsrate bei kindern,> (2.0) ah=so (.) do unten steht des erst (.) ha? “ (TAP JS 58-68) This comment, too, shows that information belonging closely together should be provided together. 8. Without mentioning this explicitly, YG seems to follow a parallelism maxim. This can be seen in her optimized version, where she tries to give the sections on the three types of diabetes a parallel structure (age groups and designations, disease itself, causes, effects). 4 NL follows the parallelism maxim explicitly stating twice that she wants to structure the information on type 2 diabetes in the same way as that on type 1 diabetes (cf. TAP NL 318-323 and 332-341, cf. also 381-382). Parallelism has been propagated by stylistics for a long time. For the reader, it increases the predictability of what comes next in a text and therefore is important for text comprehensibility. 9. In the original text, Beta-Zellen (beta cells) is introduced as the specialized term for Insulin produzierende Zellen (insulin producing cells). YG decides not to eliminate this alternative designation from the text because it might be used in doctor-patient conversations (cf. TAP YG 152-156), but she introduces it only once and then - in contrast to the author of the original text - goes on using the more telling expression Insulin produzierende Zellen , which I consider a good decision. 10. Several of the participants (SF, YG, NL) wonder about the noun Untergang (decline). YG laughs when she reads it (TAP YG e. g. 186-187); SF says: “sagt man wirklich den untergang? ” (TAP SF 244-245), “ untergang gefällt mir überhaupt nicht” (TAP SF 254). NL comments: „in folge davon kommt es zu einem unter <<f> untergang der insulin produzierenden zellen> (3.5) also des kann i ma a net gut vorstellen (.) untergang der (---) zellen (3.0) da hab ich gleich diese assoziation dass (.) diese zellen ausschauen wie schiffe und irgendwie attakiert werden und dann .hh sinken; und untergehen. was kann damit gemeint sein? =is da irgendwas im wörterbuch? ((schlägt im Wörterbuch nach)) natürlich nicht.“ (TAP NL 174-182) This unintentional foregrounding of linguistic elements takes away memory capacity necessary for processing the content on diabetes and therefore should be avoided. As NL’s TAP shows, she intends to replace the metaphor by a semantically more precise formulation: „also untergang des klingt für mich komisch. das führt zu einem verlust an, .h dadurch werden die insulin produzierenden zellen (2.0) geschädigt, oder beschä- 4 If certain types of information are not given in the original text, she cannot provide this information in her optimized version, of course. 5 Methods of comprehensibility assessment 93 digt, oder werden sie wirklich zerstört? .hh das würd ich jetz gern noch an fachmann fragen (-) was dann wirklich mit diesen zellen passiert. (3.5) ob einfach die anzahl reduziert wird, oder ob sie beschädigt und daher funktionsunfähig sin, (---) oder ob sie wirklich völlig zerstört werden. das müsste man noch mal fragen.“ (TAP NL 195-203) 11. The original text says that type 2 diabetes, also called Altersdiabetes (oldage diabetes), occurs after age 40 in most cases, which is the reason for its designation. In recent years, however, people have been affected by this type of diabetes at an ever earlier age. For YG, age 40 is not really old: “<<len> vierzge isch jo (.) no net (.) speziell: (-) alt.>” (TAP YG 404-405). Therefore she feels that the designation is not motivated semantically and wonders whether type 2 diabetes occurred later than age 40 in the past and the age has gone down to 40 only in recent times (cf. TAP YG 252-260). Since YG is the only of the five participants who had this problem, this is not taken into account in the optimized version. 12. The expression “nach Beendigung der Schwangerschaft” causes YG to laugh when she first reads it (cf. TAP YG 32); later on she reads it with a disgusted undertone (TAP YG 429-446). She does not comment, however, on what is wrong with it. NL comments on it explicitly: „beendigung des klingt mir aber viel zu aktiv (--) natürlich die geburt is das logische ende der schwangerschaft aber (.) einfach so beendigen? beenden kamma schwangerschaft ja, das ende der schwangerschaft. (4.0) <<sich selbst beim Tippen diktierend> in der regel verschwindet diese form des diabetes (2.5)> nach beendigung nein, (.) nach ende ((tippt)) (4.0) oder einfach nach der schwangerschaft (1.5) oder nach der geburt (10.0) nja ende der schwangerschaft kann auch a tragischeres (-) resultat sein als geburt (--) wemma des kind verliert (5.0) also kamma nicht einfach (.) also wär das vielleicht nicht ganz eindeutig (.) das was damit gemeint is wenn ich geburt schreibe (6.5) <<den optimierten Text lesend> in der regel verschwindet diese form des diabetes> (-) nach ende der schwangerschaft (-) oder nach der schwangerschaft (5.0) oder eben nach der geburt aber des is (-) nicht ganz optimal. (3.0) [...] diese form des (2.0) diabetes .h verschwi: ndet ((klickt)) in der (2.0) in der regel (.) nach ende (1.5) nja schreib ma nach ende der schwangerschaft ((tippt)) (.) <<sich selbst beim Tippen diktierend> ( )ngerschaft> schwangerschaft, (1.5) hh oder mit dem ende der schwangerschaft (--) nein, verschwindet in der regel nach ende der schwangerschaft“ (TAP NL 414-443) Her criticism is justified because the German verb beenden possesses the semantic features <controlled by one’s will>, <intentionally>, which may lead to the wrong interpretation that the author does not (also) refer to the end of a pregnancy marked by the birth of the child but only to an end caused by abortion. Both ends are meant. To make this clear, a more general formulation is needed such as nach der Schwangerschaft (after the pregnancy). 94 Part II: Text Comprehension and Text Comprehensibility 13. The author of the original text has a strong tendency to use nominal style, which the participants seem to find hard to understand. They transform several nominal formulations into verbal ones: “chronische Erhöhung des Blutzuckers” (NL), “Neuerkrankungsrate” (YG), “Untergang der insulinproduzierenden Zellen” (SF, YG, NL, EK), “das Ansprechen der Körperzellen auf Insulin” (SF, NL, EK), “das Alter beim Auftreten des Diabetes” (SF, YG) and “das Risiko für die spätere Entwicklung eines Typ 2 oder Typ 1 Diabetes [sic]” (YG, NL, EK). SF transforms “infolge einer Zerstörung der insulinproduzierenden Zellen” into verbal style saying, „<<den Ausgangstext lesend> infolge einer zerstörung der insulin produzierenden zellen> ich würde des auflösen mit einem nebensatz.“ (TAP EF 107-9) SF and NL also feel that “Neuerkrankungsrate” is a hard word for somebody who does not know it and also try to transform it into a verbal expression, but do not succeed (cf. e.g. NL TAP 140-149). SF therefore decides to split it up into “Rate von Neuerkrankungen” (cf. TAP SF 143- 153). SF and NL explicitly mention their avoid-nominal-style strategy, saying, „ein vermindertes ansprechen auf körperzellen das müsst ich sicherlich auch anders auflösen weil das is vielleicht doch ein bisschen zu schwierig zu verstehn (.) .hhh (4.0) beim typ zwei diabetes (3.0) .h würd ich nicht (.) nominalisieren sondern vielleicht (.) eher verbalisieren“ (TAP SF 345-350) „na ma könnt nominalstil noch auflösen“ (TAP NL 452-453) 14. SF and NL feel that the original text does not make clear that the sentence following the expression “Fehlsteuerung des Immunsystems” explains what is meant by it. Thinking aloud, SF says, „<<sich selbst beim Tippen diktierend> fehlsteuerung des immunsystems> (21.0) .hhh <<ff>> das passt mir nicht.> <<den optimierten Text lesend> <<pp>> als ursache des typ eins diabetes sieht heute die wissenschaft das zusammenwirken (3.0) des immunsystems> (.) doch (.) punkt. (9.0) hm immunsystems (.) <<den Ausgangstext lesend> bestimmte weiße blutkörperchen richten sich speziell gegen die betazellen> (5.0) .hh das würd ich auf jeden fall noch amal wiederholen (--) immunsystems, ich werde da hier auch einen doppelpunkt machen damit man weiß das hängt damit zusammen“ (TAP SF 209-219) Text coherence should be improved here. Both SF and NL use the strategy of inserting coherence increasing elements (e.g., a colon after “Fehlsteuerung” in the case of SF and dabei in the case of NL) in other places, too (cf. TAP SF 296-298, 323-324; TAP NL 182-189). NL also feels that the logical relation between the two sentences “Die Zuckerspiegel im Blut steigen an” (The sugar levels in the blood increase) and “der Körper muss als Energiequelle sein Fettgewebe aufzehren” (the body must fall back on its fatty tissues as a source of power) does not become clear. Thinking aloud, she says, 5 Methods of comprehensibility assessment 95 „das versteh ich jetz nicht ganz <<den Ausgangstext lesend> ohne insulin kann jedoch glukose nicht mehr aus dem blut in die körperzellen (1.5) aufgenommen (--) und verwertet werden. die blutzuckerspiegel im (2.5) die zuckerspiegel im blut steigen an (.) und der körper muss als energiequelle> (5.0) ach so hh anstatt dass die energie aus dem blut gewonnen wird, (4.0) muss der körper sein fettgewebe aufzehren.“ (TAP NL 233-240; cf. also TAP NL 269-314). She solves the problem by making explicit that the body cannot use the glucose in the increased blood sugar level and therefore has to fall back on its fatty tissue (and protein reserves, which is not mentioned in the original text) instead, which, to my mind, is an excellent solution. 15. For EK the explanation of type 2 diabetes is not clear. She wonders “was bedeutet ein vermindertes ansprechen? “ (TAP EK 307). In the retrospective interview, she wants to ask an expert about this: „unter der frage was versteht man unter dem .h typ zwei diabetes genau und wodurch wird er hervorgerufen ist jetzt als vage; also das müsst ma noch einen experten fragen und (--) .hh klären; hh was genau der typ zwei diabetes ist.“ (RIP EK 470-474) 16. NL wonders why type 1 and type 2 diabetes are not mentioned in the usual order in the last sentence: „<<sich selbst beim Tippen diktierend> dIabetes (---) typ> (--) warum is des (1.5) zuerst des typ zwei und dann des typ eins (2.0) warum steht da des (.) zuerst? (3.0) na ja wahrscheinlich weil (-) sobald eine frau im (1.5) gebärfähigen alter is, ist es wahrscheinlicher dass sie erst später an diesem altersdiabetes erkrankt (6.0) größer (4.0) ah, trotzdem find ichs irgendwie seltsam zuerst zwei und dann eins zu schreiben (4.0) ((tippt)) (4.0) ((tippt)) so. (2.0) ah (-) das risiko später an typ eins oder zwei diabetes zu erkrANken gehört da natürlich noch rein“ (TAP NL 456-465) The unusual order leads to a foregrounding of this information as can been seen from NL’s reflections. If there is a reason for the unusual order, this reason should be given explicitly. In fact, there is a reason: The risk of being affected by type 2 diabetes after a pregnancy is higher than of being affected by type 1. If no reason is given, the usual order should be used (cf. the solution in the optimized version in Tab. 5-4 in Section 5.5 below). 5.5 Summary and conclusions Tab. 5-3 gives an overview of the problematic elements in the source text, of how many participants had problems with them and of a few maxims and strategies used by the participants to solve these problems. 96 Part II: Text Comprehension and Text Comprehensibility Tab. 5-3: Deficiencies of the original diabetes text with regard to comprehensibility No. Problem in the diabetes text Participants who found the item problematic Maxims and strategies followed in optimizing the text 5 Incomprehensible elements and missing information 1 Is gestational diabetes a third type of diabetes or a variant of type 1 or type 2? SF, YG, JS, EK coherence increasing maxim insertcoherenceincreasingelement strategy 2 Unknown terms: chronisch Insulin Inselzellen T-Lymphozyten Gestation Glukose Körperzellen SF, EK SF, EK YG, NL, JS, EK JS SF, YG, (JS) 6 , EK SF, YG, NL, EK JS terminology explanation maxim 3 After gestational diabetes, is there a higher risk of developing type 1 or type 2 diabetes for the mother or the child? YG correctness, clarity and optimality maxim 4 Is type 1 diabetes caused only by a combination of all three factors mentioned or may it also be caused by one or two of these factors alone? JS correctness, clarity and optimality maxim What makes the text hard to understand? 1 Unexplained terminology (Diabetes mellitus) in the title for which there is a common designation (Zuckerkrankheit) SF, YG, (EK) terminology explanation maxim add-generaldesignation strategy 5 These maxims and strategies are only a selection of the complete repertoire used by the participants. A detailed discussion of them is provided in Göpferich (2006c). 6 (XY) means that XY’s TAP does not contain any explicit comment on this problem, but that she also replaces or adapts the element in question. 5 Methods of comprehensibility assessment 97 2 Semantic relation between designations and their synonyms in parentheses unclear SF, (EK) explainsemantic-relation maxim 3 Title does not fit (Definition) YG, (EK) correctness, clarity and optimality maxim 4 Is “is gekennzeichnet durch” in the definition of diabetes mellitus correct? YG, NL, (JS) correctness, clarity and optimality maxim 5 Imprecise definition of diabetes mellitus YG correctness, clarity and optimality maxim 6 Unusual plural form of Zuckerspiegel SF, YG, NL, EK take-over strategy (believeauthority strategy) 7 Information which belongs together given in different places YG, NL, SJ, EK information clustering maxim 8 Paragraphs on comparable phenomena do not have a parallel structure YG, NL parallelism maxim 9 Abstract designations ( Betazellen instead of Insulin produzierende Zellen ) YG replace-bygeneraldesignation strategy (after introducing the less telling synonyms if necessary) 10 Use of designations which lead to unintentional foregrounding (Untergang) SF, YG, NL correctness, clarity and optimality maxim 11 Semantic motivation of the designation Altersdiabetes (old-age diabetes) YG correctness, clarity and optimality maxim 12 Misleading designations (Beendigung) YG, NL correctness, clarity and optimality maxim 13 Tendency to use nominal style: chronische Erhöhung des Blutzuckers infolge einer Zerstörung der insulinproduzierenden Zellen NL, (JS) SF, NL avoid-nominalstyle strategy split-up-longdesignation strategy 98 Part II: Text Comprehension and Text Comprehensibility Neuerkrankungsrate Untergang Auftreten des Diabetes Ansprechen der Körperzellen auf Insulin Risiko für die Entwicklung SF, YG, NL SF, YG, NL, (EK) SF, YG, NL, (JS) SF, NL, EK YG, NL, (EK) 14 Intersentencial text coherence is not clear between: - “Fehlsteuerung des Immunsystems” and its explanation in the sentence following it - “Die Zuckerspiegel im Blut steigen an” and “und der Körper muss als Energiequelle sein Fettgewebe aufzehren” SF, NL NL coherence increasing maxim insertcoherenceincreasingelement strategy 15 What does “vermindertes Ansprechen” mean (explanation of type-2 diabetes unclear)? EK correctness, clarity and optimality maxim 16 Why is type-2 diabetes mentioned before type-1 diabetes in the last sentence? NL correctness, clarity and optimality maxim The criticism summarized in Tab. 5-3 can be used to produce a version optimized on an empirical basis. Such a version is juxtaposed to the original version in Tab. 5-4 below. In this optimized version, all elements criticized have been changed except for three items which were criticized by only one participant and in an unconvincing manner: T-Lymphozyten , Körperzellen and the semantic motivation of Alterdiabetes . Ideally, this optimized version should again become the object of optimizing reverbalization, which is an iterative method, until no further deficiencies can be recognized. If we compare the insight into text comprehensibility that the method of optimizing reverbalization described here gives us with the results obtainable with readability formulas, cloze procedures, questions on texts whose comprehensibility is to be assessed or text reproductions, it becomes obvious that the method described here provides much more differentiated and reliable results. At the same time, however, it is also much more time-consuming. The participants’ criticism of the diabetes text and their maxims and strategies, of which only a few could be mentioned here, reveal what the ‘ingredients’ of comprehensibility are to their mind. A comparison of these ‘ingredients’ of comprehensibility with the six dimensions of the Karlsruhe comprehensibility model reveals that there were no items of criticism that could not be attributed to one of the dimensions of the Karlsruhe model. This shows that the comprehensibility concept underlying the Karlsruhe model seems to match the intuitive comprehensibility concepts of the participants in the experiment. 5 Methods of comprehensibility assessment 99 The method described here may also be used to refine the Karlsruhe model. If the analysis of a larger number of texts reveals items of criticism that cannot be subsumed under one of the six dimensions of the Karlsruhe model or that complete the list of example optimizations provided for each dimension, the Karlsruhe model can be extended accordingly. Tab. 5-4: The popular-science text and its optimized version Original version Optimized version Definition des Diabetes mellitus Diabetes mellitus (Zuckerkrankheit) ist gekennzeichnet durch eine chronische Erhöhung des Blutzuckers, verbunden mit dem Risiko für schwere Begleit- und Folgeerkrankungen. Es werden zwei Typen unterschieden. Der Typ 1 Diabetes beruht auf einem Mangel an Insulin infolge einer Zerstörung der insulinproduzierenden Zellen (Beta-Zellen). Diese Zellen gehören zur Bauchspeicheldrüse und sind ein bestimmter Typ der sogenannten Inselzellen. Am höchsten ist die Neuerkrankungsrate bei Kindern zwischen 11 und 13 Jahren. Deshalb wurde der Typ 1 Diabetes früher auch als jugendlicher oder juveniler Diabetes bezeichnet. Als Ursache des Typ 1 Diabetes gilt heute das Zusammenwirken von erblicher Veranlagung und äußeren Faktoren (z.B. bestimmte Virusinfektionen) und einer Fehlsteuerung des Immunsystems. Bestimmte weiße Blutkörperchen (T-Lymphozyten) richten sich speziell gegen die Beta-Zellen. In Folge davon kommt es zum Untergang der insulinproduzierenden Zellen und zum absoluten Insulinmangel. Ohne Insulin kann jedoch Glukose nicht mehr aus dem Blut in die Körperzellen aufgenommen und verwertet werden. Die Zuckerspiegel im Blut steigen an und der Körper muss als Energiequelle sein Fettgewebe aufzehren. Der Typ 2 Diabetes beruht auf einem verminderten Ansprechen der Körper- Was ist Diabetes mellitus (Zuckerkrankheit)? Unter Diabetes mellitus (im Volksmund auch Zuckerkrankheit genannt) versteht man eine Stoffwechselkrankheit, die unbehandelt zu einer chronischen (dauerhaften) Erhöhung des Blutzuckers führt. Hiermit können schwere Begleit- und Folgeerkrankungen einhergehen. Man unterscheidet zwei Grundtypen von Diabetes: Der Typ-1-Diabetes wurde früher auch als jugendlicher (oder juveniler) Diabetes bezeichnet, weil an diesem Diabetes-Typ am häufigsten Kinder zwischen 11 und 13 Jahren neu erkranken. Er beruht auf einem Mangel an dem Blutzucker senkenden Hormon Insulin. Er entsteht, wenn die Insulin produzierenden Betazellen in der Bauchspeicheldrüse zerstört werden. Dies kann durch eine Fehlsteuerung des Immunsystems geschehen, bei der die Insulin produzierenden Zellen durch bestimmte weiße Blutkörperchen, die T-Lymphozyten, angegriffen werden, was zu deren völligen Zerstörung und damit zum absoluten Insulinmangel führen kann. Eine solche Fehlsteuerung kann erblich bedingt sein, aber auch von äußeren Faktoren (wie z. B. bestimmten Virusinfektionen) ausgelöst werden. Ohne Insulin kann Traubenzucker (Glukose) nicht mehr aus dem Blut in die Körperzellen zur Energieversorgung aufgenommen und verwertet werden. Der Zuckerspiegel im Blut steigt daher an, während der Körper als Energiequelle 100 Part II: Text Comprehension and Text Comprehensibility zellen auf Insulin. Er macht sich meist nach dem 40. Lebensjahr erstmals bemerkbar und wurde deshalb früher auch als Altersdiabetes oder Alterszucker bezeichnet. Auslösende Faktoren sind fettreiche Kost, Übergewicht und Bewegungsmangel. In den letzten Jahren hat sich das Alter beim ersten Auftreten des Diabetes zunehmend nach unten verlagert. Ein erstmals während der Schwangerschaft auftretender Diabetes wird als Schwangerschafts- oder Gestationsdiabetes bezeichnet. In der Regel verschwindet diese Form des Diabetes nach Beendigung der Schwangerschaft. Dabei ist jedoch das Risiko für die spätere Entwicklung eines Typ 2 oder Typ 1 Diabetes stark erhöht. Prof. Dr. med. Werner Scherbaum, Deutsches Diabetes-Forschungsinstitut Düsseldorf u. a. sein Fettgewebe aufzehren muss. Der Typ-2-Diabetes wurde früher auch als Alterszucker (oder Altersdiabetes) bezeichnet, weil er sich meist erst nach dem 40. Lebensjahr erstmals bemerkbar macht. In den letzten Jahren hat sich das Alter beim ersten Auftreten dieses Diabetes aber zunehmend nach unten verlagert. Bei diesem Diabetes-Typ können die Körperzellen Insulin nur noch vermindert nutzen. Auslöser können sein: fettreiche Kost, Übergewicht und Bewegungsmangel. Ein Sondertyp ist der Gestationsdiabetes (Gestation = Schwangerschaft), der erstmals während einer Schwangerschaft auftritt und danach meist wieder verschwindet. Er entsteht dadurch, dass Schwangerschaftshormone und Hormone, die der Mutterkuchen (Plazenta) bildet, Blutzucker erhöhend wirken, so dass die Bauchspeicheldrüse der Schwangeren zum Ausgleich immer größere Mengen an Insulin produzieren muss. Reicht die erhöhte Insulinproduktion zum Ausgleich nicht mehr aus, entsteht der Schwangerschaftsdiabetes. Für die betroffenen Schwangeren ist das Risiko, später an Diabetes Typ 1 (seltener) oder Typ 2 (häufiger) zu erkranken, stark erhöht. 5.6 Improvement of the experimental design In the pilot study described above, the method of think-aloud was combined with keystroke logging using the software Translog developed by Jakobsen and Schou (Jakobsen 1999). This software records (logs) all keystrokes and mouse clicks during writing processes as well as the time intervals between them without the user of this software realizing this. The log files provide useful additional information because pauses in the writing process and revisions recorded in them are indicators of how confident the participants are with regard to their comprehension results. Furthermore, the dictionary provided in Translog helps to find out what lexemes the participants do not understand. In the pilot study, the Translog dictionary did not satisfy all the participants’ needs for text-external sources of information. It contained only those terms which, according to the experimenter’s opinion, might have been problematic 5 Methods of comprehensibility assessment 101 for the participants. As the experiments showed, however, the participants also consulted the dictionary for terms which were not contained. Furthermore, the participants did not only have terminological questions, i.e., questions which can be answered on the lexical level, but also questions that involved more complex knowledge clusters where they would have liked to ask an expert. Since such an expert was not available during the experiment and since the participants were not allowed to use other sources of information such as the Internet, we must assume that the optimization depth the participants aimed at and were able to achieve in the experiment is lower than it would have been with an expert and other sources of information available. The use of other sources of information may again have led to further questions. The options the participants had for comprehension regulation (Baker 1985: 155) after recognizing a comprehension problem had clearly been limited. These methodical problems can be solved, however, by allowing the participants to use the Internet and any electronic as well as conventional resources they might deem useful for their task. Online searches can be recorded using screen-recording software such as ClearView. To determine what exactly the participants read on the screen when searching, a remote eye-tracker can be used. In its replay function, lines and bubbles, which increase with the duration of fixation, show what was actually read and fixated. Searches in conventional sources can be recorded by the experimenter. Such an optimized experimental design is particularly useful in studies which focus on comprehension regulation after comprehension problems have been recognized, i.e., in experiments concentrating on research processes and knowledge integration as they are necessary during text production and translation processes. Since even with this experimental design, questions may arise which the participants cannot answer in spite of the extended range of text-external sources of information available to them, retrospective interviews should be conducted in which the participants are asked what has remained unclear to them and what they would like to ask a specialist. If we expect an optimally comprehensible text not to contain incomprehensible components or elements that are hard to understand and, in addition to this, not to give rise to questions related to its topic and function that it does not answer, these remaining questions are additional reliable indicators of deficiencies with regard to comprehensibility. This time-consuming method certainly cannot be applied to every document produced in a commercial environment on a regular basis. However, results from empirical studies carried out by Schriver (1987) show that exposure of writers to the results of so-called protocol-aided revisions like the ones described above do not only help these writers to optimize the texts which were tested but also to improve their ability to anticipate their audience’s needs in general. As a consequence, optimizing reverbalization, analyzing the data which result from such experiments and confronting the authors who origi- 102 Part II: Text Comprehension and Text Comprehensibility nally produced the texts with the results may be an effective method for helping them to improve their text production competence. 5.7 More efficient alternatives? Since text optimization using think-aloud is an enormously time-consuming and painstaking method, the question arises whether there are more efficient alternatives of similar reliability and validity. One such option may be eyetracking during reading processes (Rayner 1994; Rayner 1998; Vitu/ O’Regan 1995). To investigate whether the positions in which the popular-science text on diabetes mellitus turned out to be hard to understand or incomprehensible in the study mentioned above can also be detected with eye-tracking, the same text was presented to five participants as a document to be read on a Tobii eyetracker screen. The participants were asked to read the text in such a way that they would be able to answer questions on it after reading it. They were allowed to read the text only once, but could proceed as slowly (or as quickly) as they liked. Re-reading passages or sentences was allowed, but not a second pass through the whole text. In this experiment, we started from the assumption that items in the text which are hard to understand may lead to longer or more fixations, shorter saccades, and/ or more regressive eye movements (cf. O’Regan 1990: 396, 400; Duchowski 2003: 140). We hoped that such phenomena would be visible in the gaze replays of the reading processes. To determine saccade lengths and regressive eye movements, the gaze replay function of the eye-tracker we had used turned out to be unsuitable because the gaze replay provides no exact quantitative data. Extremely long fixations are easy to spot in the gaze replay. They occurred on extremely long words such as insulinproduzierend and unknown terms such as T-Lymphozyten , which had already turned out to be hard to understand in the optimizing reverbalization study. However, since prolonged fixations occurred in apparently unproblematic positions too, they are no clear indicator of lexical difficulty. In fact, according to Rayner (1994: 211), there is experimental evidence that the following variables influence fixation time: “(1) word frequency; (2) contextual constraint; (3) semantic relationships between words in a sentence; (4) anaphor and co-reference; (5) lexical ambiguity; and (6) syntactic disambiguation.” Furthermore, fixation durations also depend on factors such as whether the fixations occur on function or content words and whether they occur in thematic or rhematic positions, etc. (Rayner 1978: 630 f.). The participants’ reading competence and their personal interest in the text also influence fixation durations, but also saccade lengths, and regressions (Rayner 1978: 636 f.). Additionally, we have to take into account that there is consider- 5 Methods of comprehensibility assessment 103 able interindividual and intraindividual variability in these parameters (Rayner 1994: 206), that there might be an eye-mind span, i.e., perceptual and cognitive processing in the mind may lag behind what is being fixated with the eyes (Rayner et al. 1989: 25 ff.), but also a “spill-over effect”, i.e., the processing of a word is not always completed by the time the eyes move and “spills over” onto the next word (cf. Rayner 1998: 377), and that there may be “purely motor or visuo-motor constraints that place limitations on saccade accuracy and fixation durations” (O’Regan 1990: 396). If we take all these factors into account in experiments, it will become obvious that even if eye-tracking should turn out as a method by means of which text comprehensibility can be determined, it will certainly be even less efficient and more time-consuming than optimizing reverbalization with think-aloud and provide less precise cues to what exactly makes a text difficult than the method described in this chapter. Part III Text Production 6 Writing (process) models From our own experience and from observing others while they are writing, we know that we cannot simply write down a text in a single burst. Writing phases are interrupted by pauses, in which we reflect, evaluate either what we have written so far or intend to write and in which we plan how to proceed. In both writing phases and pauses, mental processes occur in our brains that we are aware of but also mental processes that remain beneath the threshold of our awareness. What is going on in our minds while we are writing is not visible for an external observer. However, observations that we can make of writing processes, such as the duration of writing pauses in relation to the positions in texts at which they occur (see, e.g., Matsuhashi 1981; Bereiter/ Scardamalia 1987; Wrobel 1988: 202; 1995: 43 ff.; Günther 1993: 57 ff.), and verbal data from think-aloud during writing (e.g., Gould 1980; Hayes/ Flower 1980; Flower/ Hayes 1981) allow us to draw conclusions as to what must have gone on in the writer’s mind at a certain point and to use these insights to develop models of the writing process. These models can then be used to generate hypotheses about writing processes which can be tested empirically, and the results of empirical investigations can then again be used to refine the models. It is the dialectic interplay of theory and model building, on the one hand, and empirical investigations, on the other, that allows us to develop viable writing process models (cf. Günther 1993: 13). Why are writing models relevant for writing instruction? A prerequisite for improving writing instruction is that we know more about the text production problems that our target groups have and the phases of the writing process at which they occur. Writing problems may become visible at the text surface, i.e., in the final products of writing processes. They may show, for example, in poor structure, grammatical mistakes, a lack of coherence, violations of the rules of functional sentence perspective and of genre conventions. Just pointing out these ‘surface’ mistakes and weaknesses will not be the most efficient method when helping the authors overcome these writing problems. We need to learn more about the entire processes in which texts come about in order to be able to pinpoint the origins of the deficiencies that occur and to address them explicitly in writing instruction. In this manner, processes can be improved before they lead to deficient products. The first writing model in the history of writing process research was suggested by John R. Hayes & Linda S. Flower (1980). It has provided a lot of inspiration for the relatively young field of writing process research, which has its origins in the 1970s. The models that have followed this first one have all been inspired by it to some extent. Therefore this model, which has become a 108 Part III: Text Production landmark in writing process research, will be the first one to be discussed in this chapter. 6.1 Hayes & Flower’s (1980) writing process model Hayes & Flower’s (1980) writing process model was inspired by think-aloud protocols of writing processes. They explicitly state that they regard their model, which describes the writing processes of competent writers, as “tentative” and “provisional”, that “it provides a first approximate description of normal composition that can guide research and afford a valuable starting point in the search for more refined models” (Hayes/ Flower 1980: 10, 29). Hayes himself, among many others, proposed a revised and more elaborate model in 1996, which will be discussed in Section 6.2. Fig. 6-1 depicts the pioneering model in a version optimized graphically and terminologically by Hayes (1996: 3). Fig. 6-1: Hayes & Flower’s (1980: 11) writing process model in the version optimized by Hayes (1996: 3) The model is composed of three major components: 1. the task environment, 2. the writer’s long-term memory and 3. the cognitive writing processes. The cognitive writing processes are the component that the model focuses on (Hayes/ Flower 1980: 10). 6 Writing (process) models 109 The task environment According to Hayes & Flower (1980: 11), the “task environment includes everything outside the writer’s skin that influences the performance of the task”. It includes the writing assignment, i.e., the topic of the text to be produced and information about its intended audience. Additionally, it may include information that has an impact on the writer’s motivation (‘motivating cues’). Once the first words have been written down, the task environment also includes the text produced so far, which is important because writers refer to it repeatedly during the composition process (Hayes/ Flower 1980: 10). The writer’s long-term memory In their long-term memories, writers have stored knowledge about topics and audiences as well as generalized writing plans. Such writing plans may, for example, take the form of lists of questions, such as who, what, where, when and why? guiding the organization of a report (Hayes/ Flower 1980: 12). The cognitive writing processes Hayes & Flower (1980: 12 ff.) conceive of cognitive writing processes as problem-solving processes and propose that they can be subdivided into three major subprocesses: ‘planning’, ‘text generation’ (or ‘translating’ in their original 1980 terminology) and ‘revision’ (or ‘reviewing’ in their original terminology). ‘Planning’ can again be subdivided into ‘idea generation’ (or ‘generating’), ‘organizing’ and ‘goal setting’, which will be discussed in Section 6.1.1. While planning, writers take information from the task environment and from their long-term memories and use it to set goals and to develop a writing plan that guides the production of their texts in such a manner that the evolving text meets their goals. Writing plans can in part be retrieved from long-term memory or they may be developed during the planning process (on planning, see also Hayes/ Nash 1996). ‘Text generation’ (or ‘translating’) involves transforming ideas stored in the writer’s memory into language in accordance with the writing plan. ‘Revision’ (or ‘reviewing’), which consists of the subprocesses ‘reading’ and ‘editing’, has the function of improving the text produced. While revising, writers try to detect weaknesses in their texts both with regard to their content/ meaning and with regard to the linguistic means used, and check whether their texts meet their goals. In the following, planning processes will be looked at in more detail (Hayes/ Flower 1980: 12). 110 Part III: Text Production 6.1.1 Planning Idea generation (or generating) The function of idea generation is retrieving information that is relevant for the writing task from long-term memory. Hayes & Flower (1980: 12 ff.) assume that this process is initiated by information about the topic and the intended audience given in the writing task. Each item retrieved from memory may be used as a new memory probe so that bits of information can be retrieved in an associative manner. If items are retrieved which are irrelevant for the writing task, the associative chain is broken and memory search restarted “with a new memory probe derived from the task environment or from useful material already retrieved” (Hayes/ Flower 1980: 13). Relevant items retrieved from memory may induce the writer to take notes. These may consist of individual words, sentence fragments or even whole sentences (Hayes/ Flower 1980: 13 f.). Organization ‘Organization’ involves selecting the most useful ideas retrieved in the generating process and to structure them in a writing plan. Writing plans may have a temporal structure (e.g., “First, I’ll say A, then B.”), a hierarchical structure (e.g., “Under topic #1, I should discuss A, B, and C.”) or both (Hayes/ Flower 1980: 14). Notes taken while organizing are often arranged accordingly, i.e., they may be indented or numbered systematically (Hayes/ Flower 1980: 14 f.). Goal-setting Hayes & Flower (1980: 15) describe ‘goal-setting’ as follows: “Some of the materials retrieved by the GENERATING process are not topics to be written about but rather are criteria by which to judge a text. Often such criteria appear in the protocol when the writer is considering the audience or features of the text. At such times the writer may say, “Better keep it simple,” or, “I need to write a transition here.” The GOAL SETTING process identifies and stores such criteria for later use in EDITING.” 6.1.2 Text generation (or translating) In the text generation process, ideas retrieved from memory are transformed into well-formed sentences in accordance with the writing plan. Hayes & Flower (1980: 15) assume that knowledge in memory is stored in the form of propositions and not necessarily in the form of language (cf. Section 3.1). Other writing process models (see, e.g., Günter 1993) assume that knowledge is stored in the form of mental models, i.e., in an analogous manner (see Section 3.7). Furthermore, an investigation by Keseling, Wrobel & Rau (1987: 6 Writing (process) models 111 364; cf. FN 7) suggests that some knowledge components, especially genrespecific fixed formulations, can be retrieved ready-made in language form already in generating processes and thus need no longer be translated into language but only be expanded or completed. Presumably, the output of generating processes may take various forms. To illustrate text generation processes, Hayes & Flower (1980: 15) provide the following invented example of a think-aloud protocol that could have been elicited in a scenario in which a student had to write an essay on the French painter Henri Rousseau. ‘Text generation’ occurs in step 3: “1. Get next part of writing plan: ‘I’ve covered the early years, now I’ve got to say how he got into painting.’ 2. Plan next sentence: Retrieve propositions. Proposition A: [(Rousseau) (showed) (some early promise)] Proposition B: [(Rousseau) (did) (very little painting until 40)] Sentence plan: (Proposition A) (Proposition B) 3. Express next proposition part: ‘Rousseau … Rousseau, what? Rousseau displayed …. Although Rousseau displayed some early promise …, etc.’ 6.1.3 Revision (or reviewing) Revision processes have the function of improving the quality of the written text. They are composed of ‘reading’ and ‘editing’. Reading Reading is more or less self-explanatory. What has to be taken into account, however, is that reading for revision clearly differs from reading for interest in content. Whereas in the latter, errors in texts may be ignored as long as they do not impair comprehension, they have to be focused on in reading for revision. We will return to this in Section 6.2.2. Editing In editing processes, material verbalized - whether by reading, writing or speaking - is checked for linguistic correctness and precision of content as well as evaluated with respect to its conformance with the writing goals. Where discrepancies are detected, corrections are made (Hayes/ Flower 1980: 16 ff.). Hayes & Flower (1980: 18) assume that editing processes are triggered automatically whenever the writer spots something that needs correction or improvement, and that this interrupts any other ongoing process. In contrast to editing processes, revision processes do not occur spontaneously but only in phases which writers devote to the systematic examination and improvement of their texts. For this purpose, the text has to be re-read. 112 Part III: Text Production 6.1.4 The monitor The monitor controls the sequence in which the various subprocesses are executed as well as their interplay. Hayes & Flower (1980: 19) introduced the monitor to take account of the following observations: 1. Editing and generating processes may interrupt other processes and take priority over goal-setting processes. 2. Writing processes are controlled by goals, i.e., when writers are in the process of organizing material, they will persistently return to organizing even if their organizing processes are interrupted by editing and generating processes. 3. The manner in which writers set their goals depends on their writing styles. Some writers keep refining a sentence and do not move on until they consider it to be perfect. Their planning, text generation and revision is completed before they proceed to the next sentence. Other writers simply write down their thoughts as they occur to them and revise them later. Still other writers first plan their entire texts with all their details before starting to write. According to Hayes & Flower’s model, the writing process can be split up into phases but only in the sense that, in the first phase, planning dominates (first generating and then organizing), in the second phase, text generation, and in the third phase, revision. They do not want their writing model to be misunderstood as a “stage model” in the sense that one phase must be completed before the next can start. They rather conceive of the writing process as goaloriented and recursive. Based on this assumption, the whole writing process ‘planning  text generation  reviewing’ may be called up as a subprocess of editing. Since editing may interrupt any other process, any process may be embedded in any other process (Hayes/ Flower 1980: 29). Since Hayes & Flower’s writing model is based on think-aloud protocols, it can only take account of processes of which the writer is aware because subconscious or unconscious processes are not available for verbalization. As a consequence, it does not reflect any of the unconscious processes which must occur at the interface of text planning and text generation. This is criticized by Günter (1993: 20), who states that the model overemphasizes the “rationality of writing processes”. Contributions that can fill these gaps are Hayes’ revised and extended version of the writing model of 1980, the writing process model by Cooper & Matsuhashi (1983), Günther’s (1993) p hraseo riented p roduction s ystem (POPS) and Bereiter & Scardamalia’s (1987) models of beginners’ and advanced writers’ composing processes, which feed into my instruction oriented model of writing, that will be presented in Section 6.6. 6 Writing (process) models 113 6.2 Hayes’ (1996) writing process model Progress in empirical writing process research led Hayes to revise and extend his first writing process model, which he had developed together with Flower (see Section 6.1), resulting in the model depicted in Fig. 6-2. He acknowledges that this revised model, too, is a “framework […] intended to be added to and modified as more is learned” (Hayes 1996: 1). Fig. 6-2: Hayes’ (1996: 4) writing process model The new model has two major components: the ‘task environment’, which already existed in the earlier model in a slightly modified form and again “includes all those factors influencing the writing task that lie outside the writer’s skin” (Hayes 1996: 3), and the ‘individual’, which covers all other components of the 1980 model as well as the ‘working memory’ and ‘motivation/ affect’. In the 1980 model, the latter was attributed to the ‘writing assignment’ in the ‘task environment’, a modification that we will return to in Sections 6.2.4 and 6.6 (cf. Fig. 6-2). What are the main differences between the 1996 and the 1980 model? Hayes (1996) subdivides the ‘task environment’ into the ‘social environment’ and the ‘physical environment’. The social environment encompasses the audience but 114 Part III: Text Production also the social (and cultural) environment of the text producers, which determines their language and the culture-specific genre conventions they follow, as well as potential collaborators who contribute to the text production process and thus influence it (Hayes 1996: 5 f.). Hayes emphasizes that: “Writing is primarily a social activity. We write mostly to communicate with other humans. But the act of writing is not social just because of its communicative purpose. It is also social because it is a social artifact and is carried out in a social setting. What we write, how we write, and who we write to is shaped by social convention and by our history of social interaction.” (Hayes 1996: 5) A comparison of the two models shows that the social (and thus also cultural) component had not been considered explicitly in the early model. It is obvious, however, that it is extremely important, especially for text production in multicultural environments and for cross-cultural communication. In the 1996 model, the ‘text produced so far’ from the 1980 model becomes part of the ‘physical environment’. This also includes the composing medium, which, in more recent research, has been found to have a considerable influence on writing processes (cf. Hayes 1996: 6 f.; Knorr/ Jakobs 1997). The main differences between the 1980 and the 1996 models in the area of the individual are (Hayes 1996: 5, 7 ff., 26): 1. the central role of the working memory (cf. Section 6.2.1), 2. the fact that visual-spatial representations are taken into account in addition to linguistic ones (cf. also Section 6.2.2), 3. the emphasis on motivation and affect, which have been removed from the task environment and attributed to the individual as one of its four components, a modification that will be discussed in Section 6.2.4, and 4. the reorganization of the core of the 1980 writing model, which places more emphasis on text interpretation processes during writing. In the 1996 model, ‘planning’ from the 1980 model is included in the more encompassing category ‘reflection’, ‘revision’ from the 1980 model becomes ‘text interpretation’, and ‘text generation’ is incorporated into the more encompassing category ‘text production’. We will return to these processes in Section 6.2.2. The four components of the individual will be explained in the following sections. 6.2.1 Working memory In working memory, all non-automated processes, i.e., all conscious processes, take place. This is the place where the information currently needed is stored and where it is used for problem solving. 6 Writing (process) models 115 Hayes’ working memory model is based on Baddeley’s (1986) general model of working memory. Following Baddeley (1986), it is composed of a central executive, which is not mentioned explicitly in Hayes’ model but “serves such cognitive tasks as mental arithmetic, logical reasoning, and semantic verification” (Hayes 1996: 8), and two specialized memories: the phonological loop (‘phonological memory’) and the visual-spatial sketchpad. The phonological loop stores phonologically coded information. In accordance with Baddeley & Lewis (1981), it can be imagined as “an inner voice that continually repeats the information to be retained (e.g., telephone numbers or the digits in a memory span test)” (Hayes 1996: 8). The visual-spatial sketchpad stores visually or spatially coded information. With reference to the Karlsruhe comprehensibility concept described in Section 4.3, Baddeley’s visual-spatial sketchpad can be visualized as the place in which the mental denotation model - or at least those components of it to be verbalized in the next step - are being stored until they are verbalized using the phonological loop. Apart from these similarities, there are also differences between Hayes’ conception of working memory and Baddeley’s. In Baddeley’s (1986) model, the central executive also performs a number of control functions whereas in Hayes’ model it only has storage and processing functions. The additional control functions in Baddeley’s model include “retrieving information from long-term memory and managing tasks not fully automated or that require problem solving or decision making” (Hayes 1996: 8). In Hayes’ writing model, planning and decision making form part of the reflection process and are not attributed to working memory (which is not plausible to me because reflection has to take place in working memory). Furthermore, Hayes includes a ‘semantic memory’ in his working memory component, in which packages of semantic content are stored before being transformed into a linguistic representation, which is then stored in the articulatory buffer (Hayes 1996: 8, 23). The semantic memory thus seems to be the place to which the section from the mental denotation model to be verbalized next is transferred from the visual-spatial sketchpad before it is passed on to the phonological loop. In all other respects, Hayes’ working memory model and Baddeley’s are identical. 6.2.2 Cognitive processes The primary cognitive processes in Hayes’ writing model are ‘text interpretation’, ‘reflection’ and ‘text production’. ‘Text interpretation’ creates mental representations from linguistic and graphic input. Cognitive subprocesses that play a role in text interpretation are reading, listening and scanning graphical representations (Hayes 1996: 13). ‘Reading’, which in Hayes & Flower’s 1980 model was simply reading for revision, has a wide range of functions in Hayes’ 1996 model. These include: 116 Part III: Text Production 1. Reading to evaluate This type of reading is relevant in revision (Hayes 1996: 13 ff.). 2. Reading to comprehend the source text(s) This type of reading serves the collection of information. It does not only provide the reader with information on the topic, but also on the sourcetext writer’s personality and on the source text as a spatial display (Hayes 1996: 19). 3. Reading to understand the task Texts that do not fulfil the requirements of the assignment are often the result of a writing assignment that has not been read properly or that has not been fully understood (cf. Hayes 1996: 20). The mental representations produced in text interpretation processes can subsequently be used to produce other mental representations. Hayes refers to this type of cognitive activity “that operates on internal representations to produce other internal representations” (Hayes 1996: 13) as ‘reflection’ . In Hayes & Flower’s 1980 model, planning was the only reflection process that was mentioned explicitly. Reflection processes in Hayes’ 1996 model also include a) problem solving, b) decision making and c) inferencing. a) Problem-solving processes are required whenever writers know their goals but not the steps to be taken to achieve them (Hayes 1996: 20 f.; for planning also see Hayes/ Nash 1996 and Section 6.6.2.2.2). b) Decisions have to be made when writers have several options and have to choose which of them is the best one, or when texts or text sections have to be evaluated during revision processes. In revision processes, numerous evaluation criteria have to be taken into account, such as the adequacy of the diction and tone, clarity, suitability for the intended audience, etc. (Hayes 1996: 21). c) Inferencing is a process in which new information is derived from known information. This may happen in a goal-oriented or a non-goal-oriented manner, consciously or unconsciously. Inferencing occurs both during reading and writing processes: For example, if a paragraph is not clear, its readers may infer what its main point is. Writers make inferences, for example, about their audiences’ knowledge and interests (Hayes 1996: 21 f.). During ‘text production’ , mental representations are used to produce verbal (written or spoken) or nonverbal output. Hayes includes spoken language because spoken language may provide valuable input in the form of content and comments for written verbalization and many writers plan their sentences, either vocally or subvocally, in the medium of speech (Hayes 1996: 13). Hayes models the details of text production as follows: “Cues from the writing plan and from the text produced so far are used to retrieve a package of semantic content. This content is stored in working memory but not 6 Writing (process) models 117 in the articulatory buffer. […] A surface form to express this content is then constructed and stored in the articulatory buffer. […] When all the content is expressed or when the capacity limit of the articulatory buffer is reached, the sentence part is articulated either vocally or subvocally. If all the current content has been expressed, then the writer may show evidence for searching for new content. If the articulated sentence part is positively evaluated, then it is written down and the process is repeated. If it is rejected, a new sentence part is constructed and evaluated.” (Hayes 1996: 23 f.) 6.2.3 Long-term memory Long-term memory contains the writer’s entire knowledge. This includes their vocabulary, knowledge of grammar, genre knowledge, knowledge about the audience and task schemas, i.e., packages of knowledge which include how to solve a (complex) task (step by step). Knowledge about one’s intended readership is crucial for successful text production. Think-aloud protocols of writing processes in which writers produce text for an audience of strangers contain few indications that the writers reflect on their target group (Hayes 1996: 25). This observation as well as the traditional belief that experts have difficulty writing for novices led Hayes et al. (1986) to hypothesize “that writers may use themselves as their primary model for the audience” (Hayes 1996: 25), i.e., that they would only consider a text to be incomprehensible if they did not understand it themselves. To explore this hypothesis, they requested two groups of participants to read a difficult text and to underline everything that they assumed would be unclear to another reader. Prior to this, one group of the participants was given information that clarified some of the points in the text, whereas a control group was not given this information. The result was that the experimental group underlined significantly fewer items than the members of the control group, i.e., the hypothesis was confirmed. Suggestions on how this problem can be solved can be derived from an investigation by Swaney et al. (1991). They were able to show that instruction manuals could be improved by providing writers with think-aloud protocols of their users. This technique, called protocol-aided revision , helped writers to model their audience. Schriver (1987) was able to show that this technique does not only help writers to optimize the specific manual tested but creates an awareness that will also improve the writers’ text production competence for other genres and audiences (Hayes 1996: 25; cf. also Section 5.1.7). 6.2.4 Motivation/ affect In his category ‘motivation/ affect’, Hayes (1996: 8 f.) differentiates between the ‘goals’ that writers want to achieve with their texts, ‘predispositions’ as their preparedness to commit themselves in specific fields, ‘beliefs and attitudes’, 118 Part III: Text Production which may be the reason for certain predispositions, and ‘cost/ benefit estimates’. An example of the fact that beliefs and attitudes may have an impact on motivation is provided by Dweck (1986). She found that people who believed that successful performance in a specific area depended on innate and unchangeable intelligence responded very differently to failure than people who believed that successful performance depended on acquirable skills. The first group had an inclination to hide failure and to avoid situations in which failure was experienced. The second group asked for help and worked harder. There is no doubt that motivation and affect with the subcomponents that Hayes lists have an influence on cognitive writing processes and depend on the individual’s personality. Motivation and affect, however, may also be influenced from the outside, for example, through incentives (e.g., payment, prestige) or the risk to fail in an exam. Therefore, factors which may have an impact on motivation and affect should also be included in the task environment as in Hayes & Flower’s early model. The manner in which motivation can be influenced in the writing classroom will be dealt with in Chapter 8. We will return to Hayes’ model in Section 6.6, where I will introduce my own writing process model, which is based on Hayes’ and other models. 6.3 Cooper & Matsuhashi’s (1983) writing process model Cooper & Matsuhashi’s (1983) model specifies Hayes & Flower’s (1980) and Hayes’ (1996) models at the interface of sentence planning and text generation, on which it focuses. In a first step, Cooper & Matsuhashi subdivide writing processes into a) the development of global discourse plans and b) the development of sentence plans. The latter ends with the execution of the graphomotor plan, the actual writing action. The development of global discourse plans involves the definition of the purpose and the audience of the text to be written and includes the selection of the appropriate structure of the text. Once the global discourse plan has been developed as a frame of reference for sentence generation, sentence planning can begin, which, according to Cooper & Matsuhashi (1983), can be subdivided into the following steps, which make up an “idealized explanation of how a competent writer produces one sentence well into a discourse for which the writer has essentially solved - or is in control of - the global discourse decisions” (Cooper/ Matsuhashi 1983: 22): 1. Formulating a proposition In this step, the message to be conveyed in a sentence is planned. For this purpose, writers search their long-term memories for the information to be encoded in the next sentence (Cooper/ Matsuhashi 1983: 22). 6 Writing (process) models 119 2. Framing the proposition Framing the proposition involves determining a predicate and one or several arguments. Drawing on Fillmore’s case grammar (1968), Cooper & Matsuhashi (1983: 23) assume that writers determine what type of action is to be described (the predicate) and what relations (roles) have to be established between the arguments and the predicate (e.g., agentive, recipient, instrumental, locative, temporal, effected). This stage does not yet involve linguistic output (Cooper/ Matsuhashi 1983: 22 f.). 3. Placing the proposition By the term placing , Cooper & Matsuhashi (1983: 24 ff.) refer to the embedding of a sentence, still below the level of language, in its preceding context taking into account the type of speech act the sentence is to represent. 4. Directing the proposition When directing their propositions, writers decide what should become the topic (theme) of their next sentence. At the text surface, this may show in the form of concepts for the respective arguments placed at the beginning the sentence (Cooper/ Matsuhashi 1983: 26 f.). 5. Connecting the proposition When connecting the proposition, the writer decides which information in the sentence to treat as new and which information to treat as given, i.e., decisions on functional sentence perspective are made. At the text surface, the results of these decisions show in the order in which concepts for arguments are arranged. The results of the steps ‘directing the proposition’ and ‘connecting the proposition’ correlate. The difference between them is described as follows by Cooper & Matsuhashi (1983: 23): “Unlike the just-previous decision about directing the proposition, a decision based solely on where the writer wants to go next in the discourse, this next decision [sc. connecting the proposition ] is determined altogether by the writer’s assessment of what the reader either already knows from some other source or will remember from previous propositions in the developing discourse. Directing the proposition, deciding on its theme, is a decision made only within the context of each clause or sentence. By contrast, connecting the proposition, deciding on what is given and what is new, takes account of what has preceded in the discourse.” Thus, connecting also involves decisions “to insure cohesion with the previous proposition and with the entire preceding text” (Cooper/ Matsuhashi 1983: 29). 120 Part III: Text Production 6. Wording the proposition It is not before this step that lexical choices are made, i.e., concrete lexical items are attributed to the predicate and arguments in the propositional frame (Cooper/ Matsuhashi 1983: 31 f.). 7. Presenting the proposition At this stage, the lexical items are transformed into grammatically correct sentences (Cooper/ Matsuhashi 1983: 32). 8. Storing the proposition The result of the steps 1 to 7, i.e., the complete plan for a sentence, is then stored in short-term memory. This completes sentence planning, and writing execution can begin. The plans for propositions which are longer than just a few words may have to be rehearsed in short-term memory (more precisely in the articulatory buffer) in order to avoid portions of it getting lost before transcription is complete (Cooper/ Matsuhasi 1983: 33). Writing execution, the transcription of the proposition, is a process that may happen more or less automatically in experienced writers but may require considerable effort on the part of beginners (Cooper/ Matsuhashi 1983: 33 ff.). In sum, Cooper & Matsuhashi’s writing process model specifies what happens at the transition from sentence planning to sentence realization, a portion of the writing process that was left unspecified by Hayes & Flower (1980) and Hayes (1996). Even in Cooper & Matsuhashi’s model, however, what is involved in the transformation of a propositional structure into a sentence that can be written down and read is still rather vague (cf. Günther 1993: 24 f.). This is an aspect of the writing process that Günther (1993) focuses on in his ‘ p hraseo riented p roduction s ystem’ (“ p hrasen o rientiertes P roduktions s ystem”), POPS. For the development of his POPS, Günther (1993) utilized models of speech production because these models are much more precise with regard to the specification of what happens between the stages of sentence planning and their linguistic realization than models of written text production. It can be hypothesized that certain aspects of speech production models are transferrable to writing. 6.4 Günther’s (1993) phrase-oriented production system (POPS) Models of speech production (e.g., by Garrett 1975; 1976; 1980; 1982a; 1982b; 1988; Herrmann 1985; and Levelt 1989) assume that, for each activated concept in a mental representation, a phrase is generated. These phrases are adapted to each other and expanded in such a manner that they fit together. 6 Writing (process) models 121 This allows an incremental production of sentences of which the speaker does not know in detail at the beginning how they will be structured in the end. Günther (1993: 49) transfers these assumptions to written text production in his p hraseo riented p roduction s ystem (POPS). POPS is intended to describe the cognitive processes that occur at the transition from sentence planning to sentence realization in an even more precise manner than Cooper & Matsuhashi’s writing process model. In contrast to many other writing process models (cf. the ones introduced in this chapter), Günther’s POPS does not assume that mental representations created during the writing process take the form of propositions but rather the shape of mental models (Günther 1993: 82). 1 According to POPS, text production can be subdivided into the following six subprocesses, which may occur sequentially, in parallel and also interactively (Günther 1993: 82 ff.): a) knowledge activation b) knowledge selection c) knowledge linearization d) phrase generation e) writing program generation f) graphomotor realization The processes a) to d) involve semantic and syntactic processing of contents. Günther refers to them as central writing processes . The processes e) and f), which Günther (1993: 83) calls peripheral writing processes , involve the graphomotor execution of the results obtained in the central writing processes. Günther describes these six processes as follows: a) Knowledge activation Roughly speaking, knowledge activation involves the activation of those concepts in the writer’s long-term memory that may be verbalized in the text to be produced and inhibiting those other concepts which will not be needed. Factors that have an impact on knowledge activation are, for example, the text topic, the genre, the writer’s goals and attitudes, the writer’s assumptions concerning the target group, matters of style and the text produced so far (Günther 1993: 83). In contrast to other scholars, Günther does not assume that the knowledge activated takes the form of propositional structures but is rather used to create a mental model that is then verbalized bit by bit while being adapted in a dynamic manner to the requirements of each step in the production process (for a justification, see Günther 1993: 84 and also FN 1). 1 Investigations such as the ones by Keseling, Wrobel & Rau (1987) lead us to assume that mental representations may also already include genre-specific fixed formulations. Taking this into account, it seems legitimate to assume that mental representations may take the shape of both mental models and propositions as well as fixed formulations. 122 Part III: Text Production b) Knowledge selection In the process of knowledge selection, those concepts and the relations between them are selected from the mental model generated that will be verbalized in the next sentence. In this manner, a mental sentence model is created for each sentence to be verbalized whose concepts and concept relations are being kept activated (Günther 1993: 85). c) Knowledge linearization Knowledge linearization involves arranging the concepts and concept relations contained in the mental sentence model in the order in which they will be verbalized. Factors which have an influence on knowledge linearization according to Günther (1993: 85) are: temporal and causal relations between concepts, the minimization of cognitive effort in the planning and production phase, topicalization of concepts (i.e., their fronting as a thematic point of departure in a sentence). The temporal execution of knowledge selection and knowledge linearization are hypothesized as follows by Günther (1993: 85 f; my transl.; cf. also 137): “The beginning and the end of knowledge selection and knowledge linearization processes correlate with the beginning and the ending of the respective verbal unit, i.e., the sentence beginning and the sentence ending. The beginning of these processes at the sentence beginning binds the writer’s cognitive capacity: The more concepts to be selected and linearized, the more complex and time-consuming these processes will become. Both knowledge selection and knowledge linearization are started at the beginning of a sentence and executed as far as possible. However, they are not completed at the beginning of the sentence in order to keep the writer able to interpret and monitor revision processes and phrase generation processes that may become necessary. The completion of knowledge selection and linearization at the end of the sentence leads to a temporal cognitive relief in the writer, which allows faster text production.” d) Phrase generation In the process of phrase generation, the concepts and concept relations contained in the mental sentence model are transformed into phrases (nominal phrases, verbal phrases, prepositional phrases) in a sequential manner. Günther assumes that concepts also include linguistic information, for example, about the parts of speech in which the verbal representations of the concepts may occur. This information is used for their transformation into phrases. The order in which this happens conforms more or less to the results of knowledge linearization. As for the temporal aspects, Günther assumes that at the beginning of a phrase, “the immediate and far-reaching initiation of the process of phrase generation binds cognitive capacity, whose amount also depends on the number of modifiers 6 Writing (process) models 123 to be activated for each concept and makes temporal requirements on planning. At the end of each phrase, phrase generation is completed, which leads to a temporal cognitive relief in the writer and thus accelerated text production.” (Günther 1993: 86 f.; my transl.) e) Writing program generation Writing program generation involves the development of graphomotor programs for writing down the words in each phrase in a sequential order. At this stage, it becomes relevant whether a text will be written down by hand, sprayed on a wall or typed using a keyboard since each manner of production involves different motions and groups of muscles. This stage does not yet involve concrete programs of muscle activation (Günther 1993: 88). f) Graphomotor realization This is the stage at which programs of muscle activation are developed (Günther 1993: 88 f.). To test whether the hypotheses derived from his POPS, which concern temporal aspects of the writing process, actually represent cognitive reality, Günther conducted two experiments. In the first experiment, eight secretaries of the University of Bielefeld were required to describe three models made of building bricks. They had to do this in such a manner that their descriptions would enable another person to reconstruct the models (Günther 1993: 111 ff.). In the second experiment, eleven students of the University of Bielefeld had to complete the same assignment as the secretaries; in addition, each of them had to write six more texts: three descriptions of different dot patterns and three verbalizations of what was depicted in different sequences of pictures (for more information on the assignments, see Günther 1993: 146 ff.). The texts had to be typed on a computer using a keystroke-logging software. The participants, who were all used to working on a computer, were not allowed to take notes. In this manner, all pauses in the writing process could be determined exactly (Günther 1993: 114 f.). An analysis of pauses in relation to the positions at which they occurred in the writing process confirmed Günther’s hypotheses. The central writing processes, knowledge activation, knowledge selection, knowledge linearization and phrase generation, bind cognitive capacity leading to longer pauses at the beginning of these units and faster text production at their end. Günther (1993: 179; my transl.) summarizes his results as follows: “Both at the sentence level and at the phrase level, there is a linear relationship between the pause lengths at the beginning of these production units and their semantic complexity (number of object concepts, relational concepts and concept modifiers). Pause lengths reflect the cognitive effort involved in performing central writing processes. At the phrase level, the procedures ‘generating a noun phrase’, 124 Part III: Text Production ‘generating a prepositional phrase’ and ‘generating a verbal phrase’ turned out to be psychologically relevant, i.e., at the beginning of these phrases, the expected increase in pause lengths could be observed whereas at the end of these phrases shorter pauses occurred. Completion of central writing processes at the end of phrases and sentences leads to cognitive relief at these positions which shows in an accelerated text production and reduced pause lengths. […] Below the phrase level, the integration of new concepts into the text turned out to be another timeconsuming process. The activation and execution of peripheral writing processes mainly depends on the number of letters a word possesses.” 6.5 Bereiter & Scardamalia’s (1987) models of beginners’ and advanced writers’ composing processes Due to fundamental differences between the writing processes of inexperienced writers as compared with advanced writers, Bereiter & Scardamalia (1987) felt the need to develop two different models of composing processes, the knowledge-telling model depicted in Fig. 6-3, that represents the composing processes of inexperienced writers, and the knowledge-transforming model depicted in Fig. 6-4, that represents the composing processes of advanced writers. According to the knowledge-telling model, writing tasks are handled “in ways that make maximum use of already existing cognitive structures and minimize the extent of novel problems that must be solved” (Bereiter/ Scardamalia 1987: 5). By contrast, the knowledge-transforming model “makes writing a task that keeps growing in complexity to match the expanding competence of the writer” (Bereiter/ Scardamalia 1987: 5). Knowledge transformers may not only achieve higher levels of literary quality but also have the opportunity to derive cognitive benefits from the writing process itself. Knowledge transforming involves deliberate, strategic control over parts of the writing process that remain unmonitored in knowledge telling (Bereiter/ Scardamalia 1987: 5 f.). In the following, both models will be described in more detail. Knowledge tellers more or less simply write down what comes to their minds in an associative process induced by a writing assignment or a self-chosen writing project. In this process, their mental representation of the writing task and then of the text written so far serve as cues for the retrieval of content (content knowledge) and discourse or genre schemata (discourse knowledge). The content retrieved will then be presented following the discourse schemata if found to be appropriate by the writer. Knowledge telling processes are characterized by an absence or scarcity of planning and goal-setting processes (Bereiter/ Scardamalia 1987: 5 ff.). Knowledge tellers typically start writing without longer initial planning pauses; the think-aloud protocols of their composing processes contain little more than what also appears in the texts they 6 Writing (process) models 125 produce, and their revision processes are more or less confined to a cosmetic level without any structural modifications (Bereiter/ Scardamalia 1987: 13 ff.). Fig. 6-3: Structure of the knowledge-telling model (Bereiter/ Scardamalia 1987: 8, Fig. 1.1.) 2 By contrast, knowledge transforming is a problem-solving process in which goal-setting and planning play an important part as depicted in Fig. 6-4. Content knowledge and discourse knowledge are drawn on in the content problem space and the rhetorical problem space respectively to solve problems of knowledge and belief on the one hand and problems of how to best achieve the goals the writers have set for themselves on the other hand. As a consequence, content is not simply verbalized as it comes to the writer’s mind but transformed in such a manner that it fulfils the goals pursued by the writer for the audience to be addressed (Bereiter/ Scardamalia 1987: 10 f.). Knowledge telling may form part of a knowledge-transforming approach, as depicted in Fig. 6-4, but is embedded in a problem-solving process (Bereiter/ Scardamalia 1987: 11). The arrows connecting the two problem spaces in Fig. 6-4 indicate that the output of one problem space may serve as input for the other. Bereiter & Scar- 2 © 1987 Lawrence Erlbaum Associates, Inc.; republished with permission; permission conveyed through Copyright Clearance Center, Inc. 126 Part III: Text Production damalia illustrate this using the example of a writer who is working on the clarity of his text in the rhetorical problem space and realizes that he needs to define a central concept that he uses. This is a content problem that is passed on to the content problem space. Solving the content problem might lead him to the decision that the concept in question is not really central for his line of argumentation and needs to be replaced by another one. This replacement then necessitates a reorganization of the text, a problem that is passed back to the rhetorical problem space, etc. (Bereiter/ Scardamalia 1987: 11). Thus, “the distinctive capabilities of the knowledge-transforming model lie in formulating and solving problems and doing so in ways that allow a two-way interaction between continuously developing knowledge and continuously developing text” (Bereiter/ Scardamalia 1987: 12). Fig. 6-4: Structure of the knowledge-transforming model (Bereiter/ Scardamalia 1987: 12, Fig. 1.2.) 3 3 © 1987 Lawrence Erlbaum Associates, Inc.; republished with permission; permission conveyed through Copyright Clearance Center, Inc. 6 Writing (process) models 127 6.6 An instruction-oriented writing process model In the following, I will introduce my own instruction-oriented writing process model that represents an updated version of my 2002 model (Göpferich 2002: 249 ff.). It is based on the writing models discussed in the previous sections and, in addition, takes account of Beaufort’s (2007: 19) and Kruse & Chitez’ (2012: 63) models of knowledge domains that writers draw on. While integrating certain components of these earlier models which I deem useful, it also includes additions and modifications the reasons of which will be explained. As a model with a didactic or instruction-oriented function, it is intended to visualize the complexity of writing processes depicting their subprocesses and the factors that have an influence on them with the goal of assisting writing instructors, and also writers themselves, in locating deficient phases or components in their students’, or their own, writing processes and suggesting measures to overcome the weaknesses involved. The designation ‘ writing’ process model may induce the reader to believe that it models only text production in the verbal mode. Although it is correct that it focuses on verbal text production, it can also be applied to the exteriorization of knowledge in other semiotic systems and combinations of semiotic systems, such as text and pictures (cf. also Hayes [1996: 12], who also includes non-verbal representations in his model). Like Hayes in his 1996 model, I make a distinction between those factors “outside the writer’s skin” (Hayes/ Flower 1980: 11) that influence the performance of a writing task, the ‘task environment’, and those factors that determine it from within the writer, the ‘individual’. For the reasons specified in Section 6.2.4, motivational influences occur in both components of my model: in the component ‘individual’ as ‘attitudes and predispositions’, and in the component ‘task environment’ as ‘incentives’; they also occur underneath the monitor, on whose operation they have an influence. 6.6.1 The task environment The task environment includes ‘external sources of information’, which comprise the ‘assignment data’, which we already encountered in the Karlsruhe comprehensibility concept and therefore need not be explained here (see Section 4.3.1); ‘informants’, i.e., persons who may provide information relevant for the writing task; ‘documents’ as sources of information; and ‘interim versions’, i.e., drafts, outlines, mind maps, etc., in which ideas can be put down for cognitive relief and future orientation. The interim versions also include the ‘text produced so far’ until the writer considers it to be completed. In addition, the ‘task environment’ also comprises ‘text production aids’, such as the ones listed in the respective box in Fig. 6-5; the ‘incentives’ already mentioned; and artefacts as well as options regulating the division of work and collaboration (‘division of work’). In contrast to Hayes (1996), I have not subdivided the 128 Part III: Text Production component ‘task environment’ in my model into a ‘social environment’ and a ‘physical environment’ for several reasons: 1. The writer’s social and cultural background, which Hayes (1996: 5) assigns to the ‘social environment’, should not play a role in professional text production in contrast to writers’ private communication. Professional text production should be guided by the text assignment and the audience and not by the personal background and socialization of the writers themselves. Furthermore, the writers’ social and cultural backgrounds are ingrained in their long-term memories and therefore should be attributed to the ‘individual’. 2. The ‘audience’ is a central external factor to which text production must be geared. What is as important for text production as the audience, however, are the purpose of the text and how the sender wants to be seen. These three factors taken together are needed for a fully specified text production assignment. Purpose and sender, however, do not occur at all in Hayes’ 1996 model 4 , perhaps because they can neither be assigned to the ‘social environment’ nor to the ‘physical environment’, which is another argument against this dichotomy, from which I refrained in my model. 3. When writers, especially professional writers, such as technical communicators or journalists, are confronted with a new text production task, they cannot be assumed to possess the entire knowledge that is required for the job. They must therefore fill knowledge gaps in their long-term memories by research in external resources. These external resources may be both documents and informants, which do not occur in Hayes’ model. If we tried to integrate them into his model, documents would have to be attributed to the ‘physical environment’ and informants to the ‘social environment’. If we take into account, however, that informants need not be asked face-to-face but could also be requested for information via e-mail, for example, the question arises whether an e-mail reply should be considered a document and thus be attributed to the ‘physical environment’ or as a product of social interaction and thus be attributed to the ‘social environment’. This problem does not arise if we refrain from using the dichotomy ‘physical environment’ vs. ‘social environment’ and integrate the external sources of information as depicted in Fig. 6-5. 4 Hayes & Flower’s (1980) pioneering model was more explicit in this respect: it contained the component ‘writing assignment’ comprising the ‘audience’, the ‘topic’ and ‘motivating cues’. 6 Writing (process) models 129 Fig. 6-5: Instruction-oriented writing process model (cf. Göpferich 2002: 250) 130 Part III: Text Production 4. Hayes’ ‘social environment’ also comprises ‘collaborators’. Collaboration in text production undoubtedly has an impact on processes. It may take two different forms: a) collaboration in the sense of writing a text in an interactive manner with the collaborating writers communicating with each other and b) collaboration in the sense of splitting up a complex text into portions with only one person in charge of each portion. Collaboration in the second sense involves coordination and standardization if a consistent result is to be obtained. Standardization can, for example, be achieved, through style guides which all writers involved in a complex project must follow. Style guides, however, do not form part of the ‘social environment’, the component in which ‘collaborators’ appear, but should be attributed to the ‘physical environment’. My model takes account of problems involved in collaboration by including style guides in the ‘external sources of information’ and the ‘division of work’ in the ‘task environment’. 6.6.2 The Individual In my model, the component ‘individual’ is subdivided into the ‘long-term memory’ and the ‘cognitive processes’. 6.6.2.1 The long-term memory The long-term memory contains the writer’s entire knowledge. Writing competence is a complex competence that draws on different types of knowledge. The number and granularity of the knowledge types that are distinguished in the different writing competence models suggested to date depends on the purpose for which these models were designed (Kruse/ Jakobs 1999; Kruse/ Chitez 2012: 63; Beaufort 2007: 17 ff.; Göpferich 2014). In his model, Hayes differentiates between the following types of knowledge: topic knowledge; audience knowledge; linguistic knowledge, which includes, among other things, the writer’s lexicon and knowledge of grammar (Hayes 1996: 24); genre knowledge; and task schemas (see Section 6.2.3). Kruse & Chitez (2012: 63) and Beaufort (2007: 19) use different terminologies and subcategorize the knowledge types that are relevant for writing competence in different manners. Beaufort (2007: 221) explains how she operationalized the definitions of the knowledge domains between which she distinguished in her ethnographic study of a student’s (Tim’s) writing competence development by means of the following table: 6 Writing (process) models 131 Tab. 6-1: Defining features and examples of Beaufort’s five knowledge domains of writing expertise (Beaufort 2007: 221) Defining Features Example 5 Discourse community knowledge Knowledge of overarching goals for communication; underlying values; and metadiscourses of the discipline “Although your purpose is usually to discuss the past, some historians who probably are more comfortable being historians...would like to get at the future...” Tim, 2nd year Subject matter knowledge Knowledge of specific topics, central concepts, and appropriate frames of analysis for documents. Also, critical thinking skills to apply, manipulate, and draw from subject matter knowledge for rhetorical purposes. “I missed some historical points... [ that ] inevitably happens when you don’t study one thing deep enough...” Tim, year 2 “Your hypothesis is interesting and sophisticated. The logic with which you apply it to the readings is sometimes faulty.” professor’s comment to Tim, 1st year Genre knowledge Knowledge of standard genres used in the discipline and features of those genres: rhetorical aims, appropriate content; structure and linguistic features. “There’s so may different kinds of historical writing.... there’s the textbook, there’s the Shrewbury type paper which just focuses on one little document and squeezes as much blood as it can out of that...” Tim, 2nd year Rhetorical knowledge Knowledge of the immediate rhetorical situations: needs of a specific audience and specific purpose(s) for a single text “It’s aiming at popular audience like you know an intllectual [ sic ] audience I guess...when I was writing it, I was thinking of a New York Times Book review.” Tim, 4th year Writing process knowledge Knowledge of how to get discipline-specific writing tasks accomplished (metaknowledge of cognitive processes in composing and phases of writing projects) “I don’t have my idea...So I get my fifteen sources on the table and go through them.” Tim, 4th year 5 In this column, Beaufort (2007: 221) provides examples from texts written and answers provided by a single student, Tim, whose college writing in a freshman writing course, a first-year history course and an engineering course Beaufort analyzed. 132 Part III: Text Production Tab. 6-2, which juxtaposes Hayes’, Beaufort’s and Kruse & Chitez’ knowledge classifications, shows that there is considerable overlap between their classifications but that there are also individual knowledge types which occur in only one of their classifications. Tab. 6-2: Knowledge classifications according to Hayes (1996), Beaufort (2007) and Kruse & Chitez (2012) 6 Hayes (1996) Beaufort (2007) Kruse/ Chitez (2012) topic knowledge subject-matter knowledge knowledge audience knowledge rhetorical knowledge communication linguistic knowledge language genre knowledge genre knowledge genre task schemas writing process knowledge process discourse community knowledge media The knowledge type Kruse & Chitez (2012: 63) add is media knowledge. What is specific about Beaufort’s writing competence model is the prominent role that she assigns to “discourse community knowledge”. Without this knowledge, Beaufort asserts, students just write for getting grades, for “doing school” (Beaufort 2007: 144). What students need to grasp, however, is the real social context for writing in their disciplines and later professions (cf. also Paltridge/ Woodrow 2012: 89). For this reason, she pleads for creating an awareness in students of the discourse communities they are expected to write for in their writing assignments (not only in disciplinary writing courses, but even in general writing courses), for writing-intensive seminars in the disciplines and a “clear progression of increasingly challenging writing tasks” (Beaufort 2007: 147, 189 f.). These requirements are deemed justified since they involve realistic writing assignments, i.e., writing assignments that are relevant for the next phase of students’ academic development, for their personal lives and/ or for their future careers. Assignments which do not fulfil these requirements can be assumed to have a negative impact on students’ motivation to complete these assignments and thus lead to poorer performance than writing assignments that commit them. The knowledge types distinguished in my instruction-oriented writing process model represent a synthesis of Hayes’ (1996), Beaufort’s (2007: 19) and Kruse & Chitez’ (2012: 63) categorisation but also comprise additional knowledge types. Furthermore, the category “linguistic knowledge” or “language” is con- 6 The concepts underlying the knowledge types juxtaposed in this table are not absolutely identical. The authors do not provide exact definitions, so that the terms can only be considered rough equivalents. 6 Writing (process) models 133 sidered too narrow and has therefore been extended into the category ‘semiotic knowledge’, encompassing both knowledge about the use of language and about the use of non-linguistic signs, such as illustrations. Kruse & Chitez’ category “media” has found its way into my model in the categories “semiotic knowledge” and “knowledge about hardware and software utilization”. The resulting classification takes account of the linguistic (semiotic), cognitive, social/ rhetorical, conventional (genre), processand media-related as well as disciplinary aspects of writing. Some of the knowledge types require further explanations, which will be provided in the following. a) Subject-matter knowledge With regard to subject-matter knowledge, Alexander (2003) introduces a distinction not reflected by Beaufort (2007) and Kruse/ Chitez (2012). Alexander subdivides this category of knowledge into two more specific types: domain knowledge and subject-matter knowledge in a narrower sense. In Alexander’s terminology, domain knowledge refers to the breadth of knowledge somebody possesses in a field (e.g., history) whereas subject-matter knowledge refers to the depth of knowledge somebody possesses about a specific topic within the domain (e.g., a specific historical event). As will be explained in Section 7.6, this distinction may have didactical implications. The category ‘subject matter knowledge’ in my model encompasses both types of knowledge. b) Task schemas Examples of task schemas are ‘stored writing plans’ for various genres, information structuring methods, such as the creation of mind maps and network representations (see Section 3.4), but also research strategies needed by writers whenever the knowledge stored in their long-term memories is insufficient for accomplishing their writing task. Research strategies applied to acquire knowledge (in an efficient manner) are decisive for the quality of the text product. If the knowledge required cannot be accessed due to, for example, deficient research strategies, even ideal writing processes in a narrower sense will not lead to the desired result. Due to their importance for text production, they are explicitly mentioned in my writing process model as an example of task schemas. The examples of task schemas provided above illustrate that task schemas may have different degrees of complexity: schemas for performing particular steps, such as the structuring of information, may be embedded in more complex schemas, such as the ‘stored writing plans’ for the production of a specific genre. 134 Part III: Text Production c) Audience knowledge Audience knowledge is composed of different subtypes of knowledge, such as knowledge about what the audience already knows, knowledge about their prejudices, their habits, their cultural background, their values, etc. Characteristics of the audience that may be relevant for text production have already been mentioned in Section 4.3.1. What is of particular importance for text production in multilingual societies and for cross-cultural communication is the audience’s cultural background with their culture-specific expectations including genre conventions. d) Semiotic knowledge Texts frequently contain nonverbal representations and may appear in different media (on paper, on CD, on the Internet). Proficient writers do not only need linguistic knowledge but also media knowledge, knowledge about the use of nonverbal or multi-modal representations and the requirements they have to fulfil, knowledge that helps them to select the most appropriate type of representation (language, pictures, movies or a combination of them) to convey a particular type of knowledge. For these reasons, Hayes’ category ‘linguistic knowledge’ has been replaced by ‘semiotic knowledge’ in my model. e) (Genre) convention knowledge Hayes’ category ‘genre knowledge’ has been extended to ‘(genre) convention knowledge’ in my model. The reason for this is that genre conventions are only one type of conventions which have to be taken into account in text production. Further conventions are conventions for non-verbal representations, which may vary considerably from culture to culture and therefore have to be taken into account when producing texts for an international audience or for cross-cultural communication (see Göpferich 2002: Section 7.2.1; and Wang 2000). In addition to the types of knowledge Hayes, Beaufort and Kruse & Chitez included in their models, my model also contains the following ones: f) Knowledge about cultural differences Text production for a globalized market or multilingual communities does not only require an awareness of the cultural differences mentioned under c), but also knowledge of other cultural specificities such as different value systems. If these are unknown to the writer, this may lead him to project his own expectations and values to his target-culture audience and thus may have negative effects on communication (see Göpferich 2002: Chapter 7). 6 Writing (process) models 135 g) Knowledge about legal requirements, style guides as well as organizational processes Some genres, such as operating instructions, have to fulfil specific requirements, for example, due to product liability regulations (cf. Göpferich 1998: Chapter 11). Other texts have to be written and formatted in accordance with company style guides or even following the rules of a controlled language (cf. Göpferich 2002: Chapter 7.3). In some cases, the terminology to be used in a document including abbreviations and brand names needs to be standardized first in all languages required before it can be used in the production of the source text and its translations. Knowledge, for example, of organizational processes, need not be stored entirely in writers’ long-term memories. However, writers must at least know that such regulations exist and where they can look them up (legal texts, style guides, quality assurance procedures, etc.; cf. also item h)). Due to the relevance of this type of knowledge for text production in specific contexts, this knowledge category has been added to my model. h) Knowledge about knowledge deficits A prerequisite for the decision to research external resources is that writers are aware of the fact that they have a knowledge deficit. That a writer lacks knowledge in certain areas may itself be knowledge, knowledge about knowledge deficits, stored in their long-term memories. Therefore, the knowledge type ‘knowledge about knowledge deficits’ has been integrated into my model. However, the awareness that additional knowledge is needed may also be the outcome of reflection processes during text production. Especially authors who write for audiences from different cultures should be made aware of the fact that conventions, values, expectations, etc. may vary from culture to culture in order to raise their sensitivity that they may have knowledge gaps with regard to these differences and in order to increase their readiness to research additional information. i) Knowledge about hardware and software utilization Today, texts are almost exclusively produced in electronic environments. The tools used may range from scanners via word processors up to content management systems (see Göpferich 1998: Chapter 12). The knowledge that is required to select appropriate tools and to utilize them in an efficient manner is covered in the category ‘knowledge about hardware and software utilization’. j) Knowledge about quality criteria During text production, writers permanently have to evaluate their sources of information, their writing plans, formulations that come to their minds, etc., in order to decide whether to use or to reject them. For this assessment, crite- 136 Part III: Text Production ria are needed. Of these criteria at least those should be stored in the writer’s long-term memory that can be applied in general (see the text production guiding features in the Karlsruhe comprehensibility concept in Section 4.3.2). In addition to these, there may be task-specific criteria which writers have to develop in the course of the writing process. Such quality criteria may be stored as subschemas (e.g., for revision) in task schemas (see item b)). k) Attitudes and predispositions Although, strictly speaking, attitudes and predispositions cannot be termed ‘knowledge’, they are ingrained in a writer’s personality and have therefore been attributed to ‘long-term memory’. 6.6.2.2 Cognitive processes In his 1996 model, Hayes makes a distinction between ‘working memory’ and ‘cognitive processes’. On the one hand, he asserts that working memory is the place in which all non-automated, i.e. conscious, processes, occur. On the other hand, he explains: “In the writing model, I represent planning and decision making as a part of the reflection process rather than as part of the working memory.” (Hayes 1996: 8) This is contradictory insofar as planning and decision making may occur consciously and therefore need to be attributed, at least in part, to working memory. Due to this contradiction and the fact that, for the purposes of writing instruction, it is more important to differentiate between different types of cognitive processes than between different stores within working memory, of which the writer is unaware anyhow, I will specify different types of cognitive processes, which may take place either consciously and thus use working-memory capacity, which is limited, or unconsciously. For the specification of cognitive writing processes, I will draw on Hayes & Flower’s 1980 model, which is more explicit in this respect than Hayes’ 1996 model and which includes the component ‘monitor’, which I also deem relevant. Furthermore, I will draw on Günther’s POPS, which, in addition, takes account of unconscious processes and thus contributes to a specification of processes at the interface of sentence planning and sentence generation. As illustrated in Fig. 6-5, I differentiate between the following four types of cognitive processes: ‘information retrieval’ and ‘reflection’, which take place in the conception phase, and ‘text production’ and ‘evaluation’, which are processes of the exteriorization phase. In the conception phase, the assignment data are used to determine, among other things, the text production guiding features (see Fig. 4-2, especially arrow  ). In the following, these four cognitive processes and their control (‘monitor’) will be illustrated. 6 Writing (process) models 137 6.6.2.2.1 Informatio n retrieval Information retrieval involves recalling information from long-term memory that may be relevant for the writing task as well as completing this knowledge by research in external resources or by inferencing. The knowledge collected in this manner may not only include subject-matter knowledge but also all other types of knowledge indicated in Fig. 6-5 that may be relevant for the text production task. What should ideally guide information retrieval are the assignment data (see ‘Task environment’ in Fig. 6-5). 6.6.2.2.2 Reflection Competent writers do not directly use the knowledge retrieved, at least not in the production of the actual text, but rather in interim versions such as collections of key words, network representations, etc., which may serve as planning aids. Before they start writing, they engage in reflection processes (‘reflection’), in which they screen and transform the knowledge retrieved to adapt it to the text production task, a process that Bereiter & Scardamalia (1987) take account of in their knowledge transforming model. It is these screening and transformation processes which are insufficient in the writing processes of inexperienced writers and need to be intensified. Possible subprocesses of ‘reflection’ are: • planning, • goal setting, • problem solving, • decision making, • inferencing and • organizing (structuring). These subprocesses need not take place independently but may also interact with some of them including others as their components. A planning process, for example, may be composed of a goal-setting process, a problem-solving process and an organizing process. According to Hayes & Nash (1996: 43 ff.), planning processes can again be subdivided into the types illustrated in Fig. 6-6. Fig. 6-6: Planning processes according to Hayes & Nash (1996: 45) 138 Part III: Text Production ‘Process planning’ refers to the manner in which the writer intends to tackle the writing task. Typical reflections that Hayes & Nash (1996: 43) provide as examples of process planning are: “I’m going to write this tomorrow” and “First, I’ll read the text trough, then, I’ll edit it.” ‘Text planning’ does not refer to how writers go about the task of organizing their text production but focuses on the text itself: its contents, its style, its desired effect on the audience: “[T]ext planning is focused on what is being written - its content, its form, its impact on the audience (e.g. ‘I’ll tell them about my job’, ‘Let’s organize this as pros and cons’, ‘I’ll make them [the audience] think I’m one of them.’).” (Hayes/ Nash 1996: 43) ‘Text planning’ can again be subdivided into planning at the level of mental representations (‘abstract planning’) and planning at the linguistic level (‘language planning’), the result of which is the text (Hayes/ Nash 1996: 43). Since language planning already involves the transformation of mental representations into perceivable signs, it is not considered a planning process in my model and thus not attributed to the phase ‘reflection’ but already to the phase ‘text production’. ‘Abstract planning’ can again be subclassified into ‘content planning’ and ‘non-content planning’. ‘Content planning’, as the designation already suggests, refers to the contents to be exteriorized in the writing process; using the terminology introduced in my Karlsruhe comprehensibility concept, it could be called planning of the mental denotation model . ‘Noncontent planning’ refers to the adaptation of the contents to the specific requirements which can be derived from the assignment data, i.e., the adaptation of the mental denotation model to the needs of the target audience and the purpose of the text (Hayes/ Nash 1996: 44). As pointed out by Bereiter & Scardamalia (1987) and Carey et al. (1989: 17), it is the non-content planning that inexperienced writers neglect in the first place. The output of planning processes may be verbal and non-verbal exteriorizations, such as collections of key words, sketches, mind maps and network representations. 6.6.2.2.3 Text production During ‘text production’ the results of the conception phase are transformed into the signs which make up the text to be produced. This corresponds to the exteriorization phase in my Karlsruhe comprehensibility concept (see arrow  in Fig. 4-2). In this phase, writers need to fall back on the assignment data wherever the text production guiding features leave them leeway (see arrow  in Fig. 4-2). To model the subprocesses of ‘text production’, I integrate Günther’s phrase-oriented production system (POPS) into my model. It specifies the cognitive processes at the transition of sentence planning and sentence realization (see Section 6.4). Especially the subprocesses of knowledge selec- 6 Writing (process) models 139 tion and knowledge linearization turn out to be deficient in inexperienced writers and thus deserve special attention in writing instruction. 6.6.2.2.4 Evaluation ‘Evaluation’ is another integral component of the writing process, which may have a considerable impact on the quality of the text product. It involves critical assessment of exteriorized interim versions and the preliminary final text. Evaluation subprocesses are editing and revision processes (for the difference between them, see Section 6.1.3). Depending on the types of modification made in these processes and their causes, they can be further subclassified. Types of editing and revision processes are, for example, additions and deletions (inserting a missing letter or word or deleting them when they appear twice), stylistic improvements and the restructuring of information (see, e.g., Baurmann/ Ludwig 1985; and Becker-Mrotzek 1992). Editing and revision processes may be triggered by formulations over which writers ‘stumble’ when re-reading their texts but also by discrepancies they perceive between the text they have produced so far and their goals. In my model, ‘reflection’ and ‘evaluation’ are similar processes: they use the same criteria (cf. the six dimensions of the Karlsruhe comprehensibility concept in Section 4.3). What distinguishes them is that, with regard to the text product, ‘reflection’ is a prospective process referring to content or text that has not yet been exteriorized, whereas ‘evaluation’ is a retrospective one referring to stretches of text that have already been transcribed. 6.6.2.2.5 Sequencing, interconnection and monitoring of the processes In accordance with Hayes & Flower (1980), it is not assumed that the different cognitive processes occur in a strictly sequential order in the sense that ‘information retrieval’ would have to be completed before ‘reflection’ could begin and that ‘reflection’ would have to be finished before ‘text production’ could take place, etc. It is rather assumed that, at the beginning, ‘information retrieval’ processes dominate, then ‘reflection’ processes, after them ‘text production’ processes and finally ‘evaluation’ processes. Each of these processes, however, may be interrupted by any of the other processes at any time: a planning process may be interrupted by an information retrieval process because the writer realizes that some information is missing while planning; a text production process may be interrupted by an evaluation process because the formulation the writer is about to write down sounds awkward. This may trigger further information retrieval processes, etc. The sequencing and prioritization of processes is effected in my model by the ‘monitor’ located at the centre of the ‘cognitive processes’. It corresponds to the monitor in Hayes & Flower’s 1980 model. The control function of this monitor is assumed to be highly affected by the writers’ motivation. Their desire or reluctance to write and their 140 Part III: Text Production perseverance, for example, will determine whether they will put up with a result that is less than optimal or whether they will continue optimizing their text and apply very strict evaluation criteria until they are perfectly happy with it. (For the sequencing of writing processes, see also Section 6.1.4.) 6.6.3 Concluding remarks The empirical relevance of the cognitive processes and subprocesses distinguished in the instruction-oriented writing model could be shown, for example, in a text production dialogue of two students who were required to formulate verbal instructions for the construction of a Lego car on the basis of the Lego parts themselves and the pictorial construction plan that comes with them. While accomplishing this task, the students had to think aloud. Already in the 22 first lines of their dialogue protocol, all cognitive processes and subprocesses introduced in the model could be identified except for the subprocesses ‘inferencing’ and ‘revision’ and the subprocesses of ‘text production’. The cognitive relevance of the latter, which are generally not accessible for verbalisation, has already been supported by Günther’s (1993) pause analysis. Revision processes could not be detected because the students were still at the beginning of their writing process where revision processes are rare if they occur at all; their relevance for writing processes, however, seems to be indubitable (Göpferich 2002: 259 ff.). The didactical implications of the model will be discussed in Chapter 8. 7 Writing competence development models How does writing competence develop? And what are the factors that have an impact on writing competence development? Being able to answer these questions is of importance when it comes to the design of writing curricula and course syllabi, especially at the interfaces of our educational systems, for example, between secondary and tertiary education. At these interfaces students who look back on various educational biographies, sometimes from different disciplines and different countries, may come together and need to be picked up where they are in their individual writing developments. Only if we know in what order subcomponents of writing competence develop, how and by what factors this development is induced, how much time must be devoted to practicing the various subcompetencies and how they are interconnected, are we able to design our courses and programs in such a manner that students will reap the maximum benefit from them. Insight into writing competence development and its different stages is also important for writing assessment. One of the first models of writing competence development was suggested by Bereiter (1980). It will be introduced in Section 7.1. The five stages of writing development that he distinguishes focus on the writing skills development that can be observed during primary and secondary education. A broader perspective is taken by Kellogg (2008). With his macro-stage model of writing competence development, presented in Section 7.2, he has provided a framework into which not only Bereiter’s five stages of writing development can be integrated but also other models and theories of writing competence development, such as McCutchen’s (1996) capacity theory of writing development, outlined in Section 7.3, and dynamic systems theory, presented in Section 7.4. Whereas Bereiter’s five stage model, Kellogg’s macro-stage model, McCutchen’s capacity theory of writing and dynamic systems theory explain writing development from a process-oriented cognitive perspective, Steinhoff (2007) and Pohl (2007) take corpus-linguistic product-oriented approaches, which will be presented in Section 7.5. Steinhoff compares formulations used in term papers by university students at different stages of their academic socialization with their equivalents in articles composed by scholars and typical formulations used in journalistic texts. His model of writing competence development is inspired by Piaget’s (1972) model of cognitive development. Pohl’s study, whose findings complement Steinhoff’s, is primarily based on analyses of the term papers and theses composed by three students of the University of Giessen at different stages of their university training. The increasing relevance that has been attributed to personal characteristics, especially motivational factors, for writing competence development is reflected in Alexander’s (2003) Model of Domain Learning, introduced in Sec- 142 Part III: Text Production tion 7.6, and even more so in Bronfenbrenner & Morris’ (2006) bioecological model of human development, presented in Section 7.7, which has been adopted in writing research (e.g., Driscoll 2009; Wells 2011; Driscoll/ Wells 2012). Section 7.8 summarizes the progress in the modelling of writing competence development that is reflected in the models. 7.1 Writing development stages according to Bereiter (1980) Bereiter (1980) distinguishes five stages of writing development, whose relation he describes as illustrated in Fig. 7-1. He points out that writers need not go through these stages in the order indicated, which is assumed to result from the American educational system, but that “it is quite conceivable that, with a different sort of educational experience, children might go through different stages with a different kind of order” (Bereiter 1980: 82). The stages result from the extent to which “six different systems of knowledge or skills” that may be identified in mature writing have been integrated by a writer. These systems of knowledge or skills are: “fluency in producing written language, fluency in generating ideas, mastery of written conventions, social cognition (appearing as ability to take account of the reader), literary appreciation and discrimination, and reflective thought” (Bereiter 1980: 82). Bereiter assumes that these skills cannot be integrated all at once due to information-processing capacity limitations. Instead, writers “integrate what skills they can in order to achieve a functional skill of a higher order. As this higher skill becomes automatized, they integrate another skill with it to achieve yet a higher order skill, and so on” (Bereiter 1980: 83). This then results in the hierarchic skill systems integration illustrated in Fig. 7- 1. Associative writing combines the two skill systems of fluency in generating ideas and fluency in producing written language. It corresponds to Bereiter & Scardamalia’s (1987) “knowledge telling” and “consists essentially of writing down whatever comes to mind” (Bereiter 1980: 83). At the next stage, performative writing , writing fulfils the requirement of conformity with the conventions of style and mechanics. At this stage, the additional skill system of the mastery of written conventions has been integrated. As Bereiter (1980: 85) points out, traditional school writing instruction is an effort to shape students’ associative writing into conformity with these conventions. The stage of communicative writing has been achieved by writers when they are able to take their audience into account, i.e., when the skill of social cognition has been integrated. At earlier egocentric writing stages, writers are not completely unaware of reader-related requirements; they simply do not yet have enough working memory capacity left to systematically keep their readers’ requirements in mind. Unified writing is the stage when “writing becomes a productive craft and not merely an instrumental skill” (Bereiter 1980: 87). Writers 7 Writing competence development models 143 develop their personal style and viewpoint and apply stricter quality criteria to their own texts, which they may have acquired in reading. The skill system integrated at this stage is literary appreciation and discrimination. At the final stage, epistemic writing , writing becomes a means of generating new knowledge, it is “no longer merely a product of thought but becomes an integral part of thought” (Bereiter 1980: 88). Fig. 7-1: Model of skill systems integration in writing development (Bereiter 1980: 81, Fig. 4.1) 1 7.2 Kellogg’s (2008) macro-stages of writing competence development Kellogg subdivides the cognitive development of writing competence, or “writing skill”, into the three macro-stages specified in Fig. 7-2. It depicts “[w]riting skill [ ... ] as continuously improving as a function of practice, as is typical for perceptual-motor and cognitive skills in general” (Kellogg 2008: 4). The first two of these three stages, knowledge telling and knowledge transforming, were suggested by Bereiter & Scardamalia (1987; see Section 6.5). They differentiate between these two stages mainly by drawing on the intensity with which ideas that come to the writer’s mind while composing are transformed to meet the specific requirements that the text to be produced has to fulfil. Knowledge telling is a beginner’s stage of writing competence characterized by writers simply telling what they know about a topic without selecting from 1 © 1980 Lawrence Erlbaum Associates, Inc.; republished with permission; permission conveyed through Copyright Clearance Center, Inc. 144 Part III: Text Production what comes to their minds or adapting it to a specific text function or audience. According to Kellogg (2008: 4), children usually need 10 years of writing instruction to move on to the next stage, knowledge transforming. Fig. 7-2: Macro-stages in the cognitive development of writing skill (Kellogg 2008: 4, Figure 1) The stage of knowledge transforming is typically mastered by advanced highschool and college students. At this stage, writers ‘transform’ what comes to their minds in order to make it fulfil specific text functions. This is also the stage at which most writers get stuck in their writing competence development. Only writers aspiring to become skilled professional writers may achieve a third stage of writing competence development, knowledge crafting, a stage that Kellogg (2006; 2008) added to Bereiter & Scardamalia’s knowledge telling and knowledge transforming by drawing on the work of Walter Ong (1978; cf. Kellogg 2008: 4 ff.). In addition to the criteria that Bereiter & Scardamalia used for differentiating between knowledge telling and knowledge transformation, Kellogg introduces the number of perspectives from which writers are able to critically assess what they intend to write or have written so far. He distinguishes between a) the author representation, i.e., the mental denotation model of what the author wants to express), b) the text representation, i.e., what the text actually says, and c) the reader representation, i.e., the manner in which a prospective reader will interpret the text. 7 Writing competence development models 145 At the initial stage of knowledge telling, writers focus on their thoughts, the author representation, and still have difficulty to check whether the text they have composed really says what they wanted to express, i.e., whether the text representation matches the author representation (Kellogg 2008: 6). This difficulty is due to limited working memory capacity (Kellogg speaks of the “central executive”, which forms a specific component within working memory). In order to be able to compare the author representation with the text representation, the relevant aspects of both representations have to be kept active in working memory in parallel. This is only possible when a certain number of subprocesses of writing have become automatized thus releasing workingmemory capacity that will then be available for higher-order processes, such as comparing the author and text representations. Progression from one stage to the next is achieved by gradual automatization of less complex processes, an assumption that also underlies Bereiter’s (1980) model. In children, for example, the basic mechanical skills of handwriting and spelling deplete the limited resources of their working memories and thus constrain their ability to generate language fluently. It is not before these basic mechanical skills have been mastered to a level at which they are performed in a more or less automatized manner that cognitive capacity in their central executives is freed for higher-order processes (Kellogg 2008: 13). “Unless children develop sufficient fluency in handwriting (or typing) before the age of 12 or so, then their subsequent development of writing skill is weakened substantially.” (Kellogg 2008: 13) At the stage of knowledge transforming, enough working-memory capacity has been released for writers to be able to compare the text representation with their author representation and to make adaptations where discrepancies are spotted. According to Kellogg (2008: 6 f.), knowledge transforming also “involves changing what the author wants to say as a result of generating the text. It implies an interaction between the author’s representation of ideas and the text representation itself. What the author says feeds back on what the author knows in a way not observed in knowledge-telling. Reviewing the text or even ideas still in the writer’s mind can trigger additional planning and additional language generation.” What Kellogg refers to here is the epistemic function of writing. Expressing their ideas makes writers aware of imprecision and gaps in their knowledge, which leads to a higher processing depth. Kellogg (2008: 7) further specifies knowledge transformation as follows: “In reading the text, the author builds a representation of what it actually says. At times such reviewing may lead to a state of dissonance between what the text says and what the author actually meant, but it can also become an occasion for rethinking afresh the author’s ideas (Hayes, 2004). During knowledge-transforming, the act of writing becomes a way of actively constituting knowledge representations 146 Part III: Text Production in long-term memory (Galbraith, 1999) rather than simply retrieving them as in knowledge-telling.” At the stage of knowledge crafting, all three perspectives mentioned above come into play: the author representation, the text representation and the reader representation. Expert writers are able to consider the relevant aspects of all three representations in their working memories and compare them for differences from which adaptations should ensue. This is only possible for writers who have automatized many of the cognitive processes involved in reflection, text production and evaluation (see Fig. 6-5). Limited working memory capacity is the reason why less proficient writers make less profound and global revisions than experienced writers. While professional writers make deep structural and semantic changes taking into account both their line of argumentation and their readership, novices more or less confine themselves to changes in the vocabulary they use (Kellogg 2008: 8; cf. Sommers 1980: 382 ff.). The progression from knowledge telling to knowledge crafting takes more than two decades (Kellogg 2008), which sounds extremely long. We must not forget, however, that speech acquisition is something that happens more or less automatically in human development whereas the acquisition of writing skills is something that must be learned as a cultural practice. Comparing speech acquisition with the acquisition of writing skills, Kellogg (2008: 2) states: “Learning how to write a coherent, effective text is a difficult and protracted achievement of cognitive development that contrasts sharply with the acquisition of speech. By the age of 5, spoken language is normally highly developed with a working vocabulary of several thousand words and an ability to comprehend and produce grammatical sentences. Although the specific contribution of a genetic predisposition for language learning is unsettled, it is apparent that speech acquisition is a natural part of early human development. Literacy, on the other hand, is a purely cultural achievement that may never be learned at all. Reading and writing are partly mediated by the phonological speech system, but an independent orthographic system must also be learned.” From Kellogg’s model, it becomes obvious that the degree of text comprehensibility that writers can attain in their texts depends on the stage of writing competence development they have achieved. The early stages of writing competence development are characterized by egocentrism. 2 It is only with increasing writing competence that the perspective of the alter , i.e., the prospective reader, can be taken into account. Since this is only fully possible at the stage of knowledge crafting, being able to write in a comprehensible manner is an asset that is acquired late in the process of writing competence development. This does not mean, however, that a writer at an earlier stage of writing 2 Cf. also Bereiter (1980: 86), who speaks of “egocentric writing”. 7 Writing competence development models 147 competence development does not have any audience awareness at all. Kellogg (2008: 5) states: “ [ A ] 12 year [ sic ] old might be aware of the prospective reader, but this reader representation may be too unstable to hold in working memory. Although such a developing writer’s audience awareness might well guide, say, word choices in language generation at the moment of transcription, the reader representation would not be available for reviewing the text, if it cannot be maintained in working memory.” (cf. also Bereiter 1980: 86 f.) The total working-memory capacity available remains constant in all three stages. The differences between them come about by the use that is made of this limited working-memory capacity. With increasing writing practice, subprocesses become automatized thus releasing working-memory capacity for higher-order processes, such as the comparison of perspectives. With reference to Ericsson & Kintsch’s (1995) concept of “long-term working memory”, Kellogg (2008: 15), however, describes one manner in which the limited working-memory capacity can be extended. Long-term working memory refers to knowledge stored in a writer’s long-term memory to which, due to the frequent use that he or she has made of this knowledge, they have rapid access without having to download this knowledge into their working memories. In such cases, items activated in working memory function as retrieval cues for knowledge stored in long-term memory. In this manner, this knowledge can be used like knowledge active in working memory without it using working memory capacity (cf. also Kellogg 2001; McCutchen 2000: 14; 2011). In Section 8.5.1, we will come back to the implications that Kellogg’s model has for writing instruction. 7.3 McCutchen’s (1996) capacity theory of writing (development) In an effort to explain differences in writing processes that can be observed when comparing children or less experienced writers, on the one hand, with adults or more proficient writers, on the other hand, McCutchen (1996) developed a capacity theory of writing. At the core of her theory is human workingmemory capacity, which is limited. As illustrated by the writing process models discussed in Chapter 6, writing is an extremely complex process that involves a number of subtasks, such as planning, text generation and revision. Working memory is the space where all these processes have to be executed cognitively. In addition, it is the space where information that is needed for the processes to be executed has to be stored. According to capacity theories as they have been suggested for reading (cf., e.g., Just/ Carpenter 1992), “the limited resources of working memory necessitate trade-offs between processing and storage” (McCutchen 1996: 302). The basic assumption of the capacity theory of writing is that, in beginners, a large proportion of working-memory 148 Part III: Text Production capacity is needed for lower-order writing processes, such as the transcription of ideas, thus leaving little storage capacity for higher-order writing processes such as global planning and revision. As writing competence increases, more and more processes become automatized liberating an increasing proportion of working-memory capacity for higher-order processes and the storage of larger information units. What empirical evidence does McCutchen provide for this theory? In a survey of research on writing, she focuses “on ways in which capacity limitations contribute to individual and particularly to developmental differences in writing” (McCutchen 1996: 300). She finds evidence that effects of capacity limitations can be observed in all three subprocesses of writing between which Hayes & Flower (1980) differentiate in their early writing model: planning, translating (i.e., text generation and transcription) and revision. These three subprocesses do not seem to develop simultaneously; translating is the process to emerge first (Berninger/ Swanson 1994; McCutchen 1996: 306). This is the process to which writing novices more or less confine themselves according to Bereiter & Scardamalia’s knowledge-telling model. At this early stage, transcription may require so much effort “that some information is indeed lost from working memory as children transcribe their sentences” (McCutchen 1996: 307). According to McCutchen (1996: 309), this explains why children’s spoken texts are often perceived to be better than their written ones. However, McCutchen (1987) also found that children’s written essays were more coherent than their spoken ones. How can these seemingly contradictory findings be reconciled? McCutchen (1996: 309) argues that writing ideas down leads to a more stable representation of the evolving text in working memory, which becomes beneficial to text coherence “once the resource demands of transcription decrease to a manageable level”. Text generation, the second subtask of translating, has also been found to become more fluent with age (McCutchen et al. 1994) and thus requiring less working-memory capacity. Unlike transcription, however, “which may eventually exert only minimal resource demands (Bourdin and Fayol, 1993, 1994), text generation itself (spoken and written) continues to require resources within working memory, contributing to individual differences in writing across all age levels.” (McCutchen 1996: 314) Automatization of text generation and transcription processes contribute to linguistic fluency, which, as McCutchen (1996: 314) points out, “is important in its own right, but also in order to conserve resources for planning and reviewing”. Fluent language generation and extensive knowledge relevant to writing, such as topic knowledge and genre knowledge, are “two necessary (but not sufficient) components” that writing expertise depends on (McCutchen 2011: 61). She explains this as follows: 7 Writing competence development models 149 “Fluent language generation processes enable the developing writer to begin to manage the constraints imposed by working memory. The addition of extensive writing-relevant knowledge allows the writer to move beyond the constraints imposed by working memory (as traditionally defined) and take advantage of longterm memory resources by relying instead on long-term working memory (Ericsson & Kintsch, 1995, see also Kellogg 2001 for a discussion of LTWM [ sc. long-term working memory ] in writing).” With regard to planning, McCutchen (1996: 315) reviews studies which provide evidence of two phenomena: First, writers who have more workingmemory capacity at their disposal are better planners. Second, plans, such as outlines, prepared in advance, reduce resource demands during writing thus liberating working-memory resources for other processes. The latter finding has didactical implications: Providing students with schemata, or macrostructures, for genres they are supposed to write will allow them to devote more working-memory capacity to other subprocesses of writing and thus, ideally, to achieve better results at the corresponding levels. Limitations of working-memory capacity also have an influence on revision. Apart from reading and text generation, revision entails three subprocesses: problem detection, problem diagnosis and strategy selection (Flower et al. 1986; McCutchen 1996: 316). McCutchen (1996: 316) points out that “if problem detection itself makes large demands on resources (due perhaps to resources required by reading processes), text information crucial to successful diagnosis may be lost from working memory”. She also states that problem diagnosis is only possible if a representation of both the actual text (Kellogg’s text representation) and the intended text (Kellogg’s author representation) can be held in working memory in parallel to allow for a comparison (McCutchen 1996: 316 f.). In her literature review, she finds evidence that the three subprocesses of revision indeed “compete for limited working memory resources” (McCutchen 1996: 317). This also explains the finding described in Section 7.2 that skilled writers are able make more profound and global revisions whereas unskilled writers often confine themselves to revisions that are more narrow in scope and thus require less information to be held active in working memory. What is interesting from a didactical perspective is McCutchen’s (1996: 318) observation that “even though resource demands may lessen as writers gain fluency in some processes, such as transcription (Bourdin and Fayol, 1993, 1994), other processes, such as revision, may be so altered that they do not readily evolve into those used by skilled writers. Such an account may partially explain why knowledge-telling and other less-than-optimal writing strategies (e.g., Flower, 1979) persist well into adulthood for many writers.” For writing pedagogy, this implies that desired revision strategies involving a more global and profound perspective have to be practiced actively. The qualitative differences in certain processes indicate that “capacity limitations affect 150 Part III: Text Production not only the number of writing processes that the writer can manage simultaneously, but also the very nature of those processes” (McCutchen 1996: 320). As should have become apparent, Bereiter’s five-stage model, Kellogg’s macrostage model and McCutchen’s capacity theory of writing (development) are compatible with each other and can easily be combined. As we will see in Section 7.4, these models and theories are also compatible with dynamic systems theory. 7.4 Writing competence development from the perspective of dynamic systems theory Another paradigm in which writing competence development can be modelled is dynamic systems theory (DST; Thelen/ Smith 1994; Van Gelder 1989). 3 DST has found its way from mathematics into various other fields of research, such as developmental psychology (see Thelen/ Smith 1994), second language acquisition (e.g., de Bot/ Lowie/ Verspoor 2007; Verspoor/ Lowie/ van Dijk 2008), translation studies (Göpferich 2013) and tentatively also into writing skills development (e.g., Verspoor/ de Boot/ Lowie 2004). In DST, competences or skills are envisaged as dynamic systems, i.e., sets of variables that are interconnected and thus interact over time (de Bot/ Lowie/ Verspoor 2007: 8). With regard to writing competence, these variables can be regarded to represent writing subcompetencies that, in their entirety, make up writing competence. These individual subcompetencies may not develop at the same pace, nor will they always develop in a linear manner. Some may stagnate while others continue to develop. Certain subsystems (i.e., subcompetencies or clusters of subcompetencies) may be precursors of other subsystems in the developmental process. For a specific subcompetence to start developing, it may be necessary for other subcompetencies to have exceeded a certain threshold value. If we start from the cognitive capacity theory of writing (McCutchen 1996), which assumes limited working memory capacity, it appears plausible to assume that the value one variable takes, i.e., the cognitive effort required for a specific subprocess of writing, such as lexical choices, has an effect on all the other variables. For example, if one subcompetence reaches an advanced level, which may result in automatized performance of the respective tasks, working memory capacity is released. This capacity then becomes available for the application of other subcompetencies and their development, such as the capacity to take larger stretches of a text into account to create coherence, for which there may not have been enough cognitive resources left as long as other subprocesses still needed cognitive effort. 3 The outline of DST presented here is based on an earlier, more elaborated version in Göpferich (2013) introducing DST in the context of translation competence development. 7 Writing competence development models 151 DST further assumes that, in dynamic systems, repeller states and attractor states exist. These can best be explained by means of the metaphor of a plane with holes and bumps in it on which a ball rolls. The ball and its trajectory on the plane represent a specific competence and its development path. On its ways over the plane, the ball is attracted by the holes (attractor states) and repelled by the bumps (repeller states). To get it out of a hole, much energy is needed. Such holes can explain fossilization of certain errors or stages of development, as observed in second-language acquisition, which can only be overcome by intensive training. The development that a particular person undergoes, i.e., the trajectory of the ball, may be highly individual, yet attractor states account for the fact that certain subcompetencies seem to occur in a specific order or at a specific stage of competence development because the corresponding attractor states, in an evolutionary perspective, possess qualities that make the application of cognitive resources settle into an equilibrium at these stages. In DST, the fact that a certain competence stage has been achieved becomes visible by a relative stagnation in the development of the values the various variables that make up the system take at this stage, whereas a move out of the hole, a new competence development burst, shows in a large variability in the set of variables which form the competence system (Thelen/ Smith 1994: 97). What complicates the analysis of writing skills development in a DST approach is that dynamic systems are nested, i.e., every system is always part of a larger system. This nesting or embeddedness of competences makes it difficult to draw a border around the system an investigation focuses on, and drawing such a border always means ignoring factors beyond the border that may have effects relevant to the system in focus. Drawing on Bronfenbrenner & Morris’ (2006) bioecological model of human development, for example, Slomp (2012), Driscoll & Wells (2012) and Wardle & Roozen (2012) plead for analyses of writing competence development which do not only take into account contextual factors such as the type of writing instruction students were exposed to, but also characteristics of the individual, dispositions, such as values, self-efficacy, attribution and self-regulation. 7.5 Academic writing competence development from a corpuslinguistic product-oriented perspective The models and theories of general writing competence development discussed in the previous sections of this chapter have their foundations in cognitive psychology. They explain processes of writing competence development with reference to automatization or routinization of subprocesses of writing, the ensuing possibility of a reallocation of working-memory capacity to higher-order processes and the expansion of this capacity by the use of longterm working memory. 152 Part III: Text Production In contrast to these process-oriented approaches to explaining writing competence development, Steinhoff (2007) and Pohl (2007) take a product-oriented perspective and focus on the development of a specific form of writing competence: academic writing 4 competence. 7.5.1 Steinhoff’s (2007) stages of academic writing competence Steinhoff (2007) compares academic textual routines or procedures 5 (“wissenschaftliche Textprozeduren”) used in German term papers composed by university students at different stages of their academic socialization (student corpus) with their equivalents in articles composed by scholars (expert corpus) and functionally comparable routines and formulations in a corpus of journalistic texts. His student corpus comprises 296 term papers composed between 1992 and 2004 by 72 different students then enrolled either at the University of Bielefeld or at the University of Giessen. All of them were native speakers of German composing in their mother tongue. The term papers were classified into three groups: those composed between the first and the third semester, those written between the fourth and seventh semester and those composed from the eighth semester onwards. The disciplines covered were history (31% of the texts), literary studies (26% of the texts) and linguistics (16% of the texts). The expert corpus comprises 99 articles: 33 from the journal Geschichte und Gesellschaft [History and Society] , 33 from Sprachkunst [Linguistic Art] and 33 from the Zeitschrift für germanistische Linguistik [Journal of Germanic Linguistics] . The authors of all articles had to be native speakers of German. From each author only one article was included in the corpus. Apart from these criteria, the articles were selected randomly. The corpus of journalistic texts comprises articles from two German newspapers Mannheimer Morgen and Frankfurter Rundschau . This corpus was primarily used to test the hypothesis that, in their term 4 The English term academic writing has two German counterparts: wissenschaftliches Schreiben and akademisches Schreiben. Wissenschaftliches Schreiben refers to professional writing in academic disciplines and genres (e.g., articles in academic journals), whereas akademisches Schreiben is used as a more encompassing concept comprising also learner’s stages of academic writing including the composition of genres such as essays and term papers that learners write in preparation of more complex writing assignments. A back translation of wissenschaftliches Schreiben into scientific writing would be wrong because scientific writing only includes writing in the sciences and not, for example, in the humanities (cf. Jakobs 1999: 173; Girgensohn/ Sennewald 2012: 89 f.). 5 The term textual routines/ procedures refers to cognitive acts or moves, which may be closely connected with specific formulations. Cf. Steinhoff (2007: 431; my transl.): “Procedures are cognitive but also linguistic phenomena. For certain procedures preferred linguistic formulations exist. These formulations are what students encounter in specialized texts as signals of their being academic. They try to use them in a similar manner in their own texts in order to obtain similar effects.” 7 Writing competence development models 153 papers, students inexperienced in academic writing fall back on formulations which are typically found in journalistic texts (Steinhoff 2007: 153 ff.). Steinhoff derives the criteria according to which he analyses his corpora from the basic requirement of scientific research, which is the generation of knowledge. Two values that are linked to this requirement are intersubjectivity and originality. To fulfil these requirements, scholars make use of specific textual routines (“wissenschaftliche Textprozeduren”) in certain functional areas. Steinhoff’s (2007) analyses focus on aspects of language use and writing competence development in five such functional areas (Steinhoff 2007: 109 ff.): 1. author reference, i.e., whether, how and in what contexts authors refer to themselves; 2. intertextuality, i.e., to what extent and how authors refer to other texts, 3. concessive argumentation, i.e., whether authors anticipate and introduce criticism that could be levelled against their arguments and how they try to refute it (Steinhoff 2007: 329 ff.); 4. text criticism, i.e., whether and how authors refer to prior texts in an evaluative and not just descriptive manner, and 5. concept formation, i.e., whether authors just use concepts found in the literature or form concepts of their own and how concepts are introduced and defined. Steinhoff’s hypothesis is that, as students’ writing experience increases, the textual procedures and formulations they use in their term papers gradually approach those found in expert papers, i.e., students’ language use develops from context inadequacy to context adequacy, a hypothesis that he is able to confirm (Steinhoff 2007: 421). Steinhoff’s model of academic writing competence development is based on the following assumptions: 1. Writing competence development is as a life-long process in which specifically academic writing skills only start developing when students are being confronted with academic writing tasks in a specific domain at university. The cognitive, communicative and linguistic competences that students have acquired before entering university have to be expanded and specified for this purpose (Steinhoff 2007: 130 ff., 419). 2. Writing competence development is conceived of as a problem-solving process. Students required to compose their first term papers are confronted with new writing problems, which they learn to solve on the basis of their reading and composing experiences (Steinhoff 2007: 132 ff.). 3. Academic text competence is conceived of “as pragmatically constituted and stabilized intuitive knowledge” (Steinhoff 2007: 418; my transl.). 154 Part III: Text Production These assumptions also underlie Pohl’s (2007) model of academic writing competence development discussed in Section 7.5.2. As a heuristic model for the explanation of writing competence development, Steinhoff draws on Piaget’s (1972) theory of cognitive development. According to this theory, the driving force in the development of intelligence is the reestablishment of an equilibrium, whenever this has been destroyed, by processes of assimilation, i.e., the adaptation of the environment to the learner’s cognitive schemata, or accommodation, i.e., the adaptation of the learner’s cognitive schemata to the environment (Piaget 1972: 191; cf. Ortner 1995: 336). With regard to academic writing competence development, the search for this equilibrium means developing writing behaviour that is adequate in scientific contexts out of writing behaviour that has been acquired for other contexts and is found to no longer fulfil its function in a satisfactory manner in the new context. Steinhoff (2007) finds that, at the early stages of this development, assimilation processes dominate, i.e., processes in which known behaviour is used to solve new problems, whereas at more advanced stages, accommodation processes take over, i.e., processes in which new strategies are made use of. These stages will be described in more detail below. 6 The most important finding of Steinhoff’s corpus analyses is that the development of academic writing competence follows an interindividual pattern (Steinhoff 2007: 419), a result that is confirmed by Pohl’s analyses. Students’ initial term papers are characterized by context-inadequate language use. In the course of their academic writing socialization, this develops into contextadequate language use. For example, patterns of concessive argumentation are initially used by students with regard to their presentation; they try to defend weaknesses and omissions in their papers. At later developmental stages, they use this pattern with regard to the topic covered and their line of argumentation as in expert papers (Steinhoff 2007: 360). Depending on the individual writers and the requirements their texts have to fulfil, students’ writing competence development may either be a gradual one or take place in bursts and follows the following stages represented in Fig. 7-3. At the initial stage, two types of problem solving were found to dominate: transposition and imitation . Transposition is an assimilating strategy in which writers try to solve their problems of academic writing by resorting to means and measures that are untypical of academic writing and occur in genres they are more familiar with, for example, in journalistic (or popular-science) texts, 6 As Pohl (2007: 523) points out with reference to Karmiloff-Smith (1993), developmental processes may not only be initiated by a disequilibrium between the competences possessed and the competences required by new situations, as assumed by Piaget, they may also come about when the competences required have already been acquired but when their application can be optimized cognitively by “representational redescription”, i.e., “a process of ‘appropriating’ stable states to extract information they contain, which can then be used more flexibly for other purposes” (Karmiloff-Smith 1993: 25). 7 Writing competence development models 155 in genres that students had been required to compose at school and in oral communication. Evidence of this strategy can be found, for example, in procedures of intertextual reference and concept formation. Instead of precise references (e.g., Smith 1992: 25; and Myers 2001: 37 ), students at the initial stage use more general references, such as Scholars assume that 7 , as they can be found in journalistic articles. Instead of precise concept definitions, students use expressions, such as Everybody is talking about multiculture , which, as Steinhoff could show, can also rather be found in journalistic texts. Transpositions of patterns that students know from the compositions they had to write at school include the explicit expression of their opinions by formulations such as in my opinion and I feel that , the postponement of critical remarks (text criticism) to the concluding sections of their papers and as well as explicit reflections of their text production process (author reference), such as Since I could not find anything on this topic in the library, I decided to ... Examples of transpositions from oral communication are expressions such as I am not so sure whether , and but that is only half the truth (Steinhoff 2007: 139 ff.; 381, 422 f.). Fig. 7-3: The development of academic writing competence (adapted from Steinhoff 2007: 138) Imitation is an accommodating strategy in which students, unsuccessfully or only partially successfully, try to imitate academic language. Formulations that result from this strategy are characterized by, for example, extreme complexity, excessive nominalizations and the use of an extremely formal style, which often results in a lack of comprehensibility (Steinhoff 2007: 143 ff.; 423). The stages of transposition and imitation are followed by a stage of transformation , in which, also in an accommodating manner, the repertoire of academic formulations that students have at their disposal is expanded. Students begin to learn typical formulations of academic writing and the communicative functions for which they are used in academic contexts. The fact that, at this stage, they have not yet fully reached expert competence, shows in, for 7 Steinhoff (2007) provides examples from his German corpora. These have been replaced by fictive English equivalents in order not to exclude a non-German speaking audience. 156 Part III: Text Production example, wrong collocations, such as provide a survey over (instead of provide a survey of ) and consider as inappropriate (instead of consider [to be] inappropriate ), and an excessive and repetitive use of certain domain-specific formulations. Another characteristic of this stage is that students begin to recognize that certain general-language expressions are used for more specific purposes in academic texts (Steinhoff 2007: 146 ff.; 423 f.). Students’ writing competence development reaches its (preliminary) final stage, the stage of contextual adequacy , when they start to use academic language following domain-specific conventions and employing a differentiated academic vocabulary in the areas specified above. At this stage, they still have writing problems but, in contrast to the initial stage of academic writing competence development, they have learned to solve them in an academic manner (Steinhoff 2007: 149). According to Steinhoff, the attainment of this stage also indicates an expansion of students’ academic text competence as an analytical competence: “Writers learn to describe and classify their own actions as well as those of the authors they quote precisely.” (Steinhoff 2007: 424; cf. 148 f.) As Steinhoff (2007: 328, 425) points out, students’ writing competence development over these stages is also characterized by a change of roles. Students develop from reproducing learners via active thinkers to critical evaluators. Steinhoff illustrates this by, for example, the different use that students make of the pronoun I (author reference). At the initial stage, it takes the form of an unscientific narrator I (“Erzähler-Ich”) and then gradually turns into an author I (“Verfasser-Ich”) and then a researcher I (“Forscher-Ich”). It also shows in the types of intertextual references that are made. Academic writing novices start out reproducing other sources passively. At more advanced stages, they go on to reflect actively what they have read and finally move on to assessing and criticizing what they have read. At the beginning of their academic socialization, they learn procedures by means of which they prove that what they write is correct, e.g., by means of quotations. Towards the end of their academic socialization they acquire procedures by means of which new knowledge can be generated, e.g., concept formation procedures (Steinhoff 2007: 328, 425). 7.5.2 Pohl’s (2007) stages of academic writing competence Pohl (2007) analyzes the ontogenesis of academic writing competence primarily on the basis of a corpus of 12 German term papers and final theses that three different students (all native speakers of German) composed in the course of their academic training at the University of Giessen. From each of the three students, Pohl analyzed three term papers from different stages of their academic training and their final theses, the latter of which they had composed in their 10th to 15th semesters. The three students, all of whom studied German as a major in a university teacher training program, were 7 Writing competence development models 157 selected because they showed different developmental paths with regard to the quality of their papers: Nadine was a good writer from the beginning, Ansgar was among the weaker writers throughout his academic training and Sonja was a writer who began rather weakly but then improved considerably (Pohl 2007: 217 ff.). The papers and theses were analyzed with regard to 1. the specificity of their titles, 2. the types of their introductions, 3. the manner in which intertextual references were made, 4. their organization, line of argumentation and metacommunicative elements, and 5. the general academic language (“wissenschaftliche Alltagssprache” in the sense of Ehlich 1999) that the students used. The analyses of the general academic language used focused on lexicosyntagmatic features, such as the lexical density of the texts, the lexical density of noun phrases, the percentage of paratactic constructions, the number of words per subordinate clause and grammatical metaphors. For the analyses of introductions and the general academic language used, Pohl used a complementary corpus of experimentally elicited texts. These were parodies of an introduction to a scientific article which 56 participants, who were required to imagine being geneticists who had discovered a criminality gene that determined whether a person had an inclination to become a criminal, were required to compose on their fictive discovery (Pohl 2007: 277). Among the participants were 4 non-native speakers of German, 21 firstto third-semester students, 13 fourthto sixth-semester students, 15 seventhto tenth-semester students and 3 university teachers. In an additional third study, 85 participants had to fill in the blanks in the introductory part of an academic article using expressions ranging from formal written language (“gehobene Schriftsprachlichkeit”, 170) to general academic language. The group of participants comprised 8 non-native speakers of German, 17 first-year students, 21 second-year students, 14 third-year students, 11 fourth-year students, 7 fifth-year students and 7 university teachers (Pohl 2007: 179). In what follows, only a rough outline of the wealth of Pohl’s findings can be presented. In line with Steinhoff’s (2007) results, Pohl found, for example, that the numbers of blanks filled in correctly increased with the students’ number of semesters at university, with the non-German speaking students achieving the poorest results and the university teachers, the best (Pohl 2007: 180). Generally speaking, he found that the ontogenesis of academic writing competence can be modelled as a developmental process with three stages, as visualized in Fig. 7-4. At the first stage, students focus on the object or topic they deal with in their academic texts. In their introductions, for example, they describe what they cover in their papers (object dimension). Only gradually do they realize that the what cannot be dealt with in a manner independent of the academic dis- 158 Part III: Text Production course but has to be set in relation to what has been published about the topic previously. This is the stage at which students start embedding their object into the academic discourse (discourse dimension) referring to other researchers and their contributions to the description of the object at hand. At the third and final developmental stage, they realize that findings are also dependent on the methods that were employed to gain them. This is the stage at which an argumentative dimension shows in their papers. Pohl only analyzed papers and final theses composed in German and thus only provides German examples. His findings, however, can be assumed to be valid for the development of academic writing competence in other languages as well. To illustrate the three stages of academic writing competence development, first, examples of extracts of English theses (one fictive, two real) which represent the different stages will be presented. Subsequently, Pohl’s three stages will be described in more detail. Fig. 7-4: The ontogenesis of academic writing (adapted from Pohl 2007: 488, Fig. IV-1) [ 7-1 ] Focus on object dimension In the following, I would like to deal with the topic of comprehensibility first from a theoretical perspective and then provide an analysis of an extract of a contract. I will first ask the question what comprehensibility means. Then I will introduce the different dimensions of comprehensibility, which also played a role in my analyses. The results will be presented in the last part. [ 7-2 ] Focus on discourse dimension (from a Master’s thesis on strategic behaviour in translation problem solving) “A common belief among laymen with regard to translation is that people with bilingual competence are automatically able to translate adequately into their mother tongue (cf. Hönig/ Kußmaul 1982: 9). The assumption that translation competence is innate and automatically occurs and develops with bilingual competence is nowadays disproved in translation studies (cf. Lörscher 1991, Shreve 1997, Englund Dimitrova 2005). There is consensus about the fact that bilingual competence is necessary for the development of translation competence, but it is only one of several subcompetencies (cf. Göpferich 2008: 148). The questions of what exactly constitutes translation competence and how this competence develops over time 7 Writing competence development models 159 were some of the major questions in translation process research during the past 10 years (cf. Göpferich 2008: 143). In recent translation competence models, strategic competence has been considered one of the most important parts of translation competence (cf. PACTE 2005: 610). However, the degree to which professional translators and translation students proceed to translate in a strategic manner has not yet been intensively investigated. Drawing on the corpus of target texts and process-oriented data collected in the longitudinal study TransComp at the University of Graz, the present paper contributes to the comparative studies of semi-professional and professional translators in translation process research by comparing the problem-solving behavior of six translation students in their second year of translation studies and five professional translators with at least 10 years of translation experience.” [ 7-3 ] Focus on argumentative dimension (from a Bachelor’s thesis on connector usage) “In our global and increasingly interconnected world, the knowledge of foreign languages, especially English, is of growing importance. Therefore, making the imparting of English language skills more effective has become a focal point of interest in applied linguistic research in the last decades. With my bachelor thesis I seek to contribute to the investigation of Second Language Acquisition (SLA) by analyzing the use of connectors in essays written by German learners of English. Connectors or conjuncts have the function of ‘signal[ling] logical or semantic relations between units of discourse’ (Altenberg/ Tapper 2008: 80) and hence significantly add to the coherence and comprehensibility of a text, aspects of writing even advanced learners and native speakers typically struggle with. Since they are used to different extents and in different ways in English and German, it can be assumed that learners will have problems employing them, especially if transfer is at play. This hypothesis is supported by the fact that transfer from German has been shown to affect English text production by German learners of English with regard to connector usage (cf. Granger&Tyson 1996: 23) as well as other aspects of the language negatively (cf. Nesselhauf's analysis of the use of collocations; Nesselhauf 2004: 235). Interestingly, other studies have challenged the negative influence of transfer on connector usage in EFL writing, at least for learners from first language backgrounds other than German (cf. Altenberg/ Tapper 2008: 81). There are indeed reasons to believe that the influence of transfer in EFL writing is not as significant as hitherto assumed. This thesis examines these reasons. “ On the basis of his analyses and generalized observations, Pohl (2007: 489 ff.) characterizes the three stages. These characterizations can be summarized as follows: At the object-focused stage , students mainly use general-language expressions. Their texts may also contain expressions, often defective, which imitate academic language. These observations are in line with Steinhoff’s findings that, at the initial stage, students try to solve new writing problems by drawing on means and strategies they have acquired for other contexts and by imitating, often unsuccessfully, academic language. General academic language (in the sense of Ehlich 1999) is used rarely and where it occurs it is used in a descrip- 160 Part III: Text Production tive function with reference to the object and not in an epistemic function. Grammatical metaphors are rare. Students use complex sentences and closely stick to formulations they find in the literature. Intertextual references are characterized by the application of the macro-rules of omission and selection rather than generalization and construction (see Section 3.5). Students use quotations for poetical reasons or to spare themselves the trouble of paraphrasing. Texts composed at this stage contain no argumentation at the macroand meso-levels and only little argumentation at the micro-level, often of marginal importance. Text organization often takes the form of additive sequences, interspersed with explanations and interpretations. Metacommunication is limited to the announcement of different sections without any justification. Introductions take the shape of prefaces or descriptions of the object. If texts composed at this stage have a conclusion at all, it is an “evaluative appendix” (“evaluativer Appendix”, Pohl 2007: 490) comprising subjective comments and the description of the personal experiences the writers had when composing the text (Pohl 2007: 489 f., 493 f.). As Pohl (2007: 494 f.) aptly observes, students at this stage seem to write for the purpose of explaining things to themselves. At the discourse-oriented stage , writing is characterized by an increased level of idiomaticity. General academic language is used to an increasing extent as are grammatical metaphors. A tendency of creating hypercomplex sentence constructions can be observed. Source texts are referred to in a less literal manner. Although still absent at the macro-level, argumentation occurs at the mesolevel, at which different positions may be contrasted with each other and evaluated, which suggests that students at this stage, in contrast to the first, no longer believe in an objective truth (Pohl 2007: 506 ff.). Metacommunicative announcements are made to facilitate comprehension for the reader. Introductions are still object-focused but refer to prior findings and may contain first argumentative elements. At this stage, conclusions are still an exception or take the shape of an evaluative appendix (Pohl 2007: 490, 495 f.). At the argumentation-oriented stage , writers make elaborated and idiomatic use of a wide range of academic vocabulary and formulations. Scientific categories, methods and approaches are no longer represented as “given” but as conceived of by human beings and thus subject to potential criticism. Increasing conceptual complexity is reflected by a higher lexical density, especially in noun phrases. At the same time, the complexity of subordinated clauses decreases. The decisive epistemic concepts and their relations occur in positions appropriate from a communicative perspective. Writers succeed in consistently differentiating between their own ideas and ideas found in the literature. In references, the macro-rules of generalization and construction replace the formerly dominant macro-rules of omission and selection. Argumentative structures can be observed at all levels. Potential objections are anticipated and dispelled argumentatively in a reader-focused manner. Introductions are fully 7 Writing competence development models 161 developed including both discursive and argumentative elements. The argumentation leads to a real conclusion (Pohl 2007: 491, 496). 7.6 Alexander’s (2003) Model of Domain Learning What is specific about Alexander’s Model of Domain Learning (MDL, 2003) is that, in addition to cognitive aspects, it takes into account students’ motivation and affect, which “are significant contributors to the development of expertise, both in and out of school” (Alexander 2003: 19). Since, in addition, her model arose from contexts of educational practice (i.e., studies of student learning in academic domains, such as reading, history, physics, biology), it is particularly useful for the description and analysis of academic literacy development in educational contexts. The MDL includes three components that are assumed to play a role in expertise development in academic domains: knowledge, strategic processing and interest. Knowledge is subdivided into domain knowledge and subject-matter knowledge. Whereas the former refers to the breadth of knowledge somebody possesses in a field (e.g., history), the latter refers to the depth of knowledge somebody possesses about a specific topic within the domain (e.g., a specific historical event). With regard to strategic processing , Alexander distinguishes between surface-level processing strategies and deep-processing strategies: “Surface-level strategies (e.g., rereading or paraphrasing) are processes individuals use to make sense of the text. Deep-processing strategies, by comparison, involve delving into that text, as when students judge author credibility or form mental representations.” (Alexander 2003: 11) Interest is subdivided into individual interest and situational interest. Individual interest refers to the enduring interest that one has in a particular domain, whereas “situational interest is tied to the ‘here and now’. It is arousal or piquing of attention sparked by events or features of the environment. Because it is bound to the immediate situation, such interest is fleeting.” (Alexander 2003: 11) With reference to these three components, Alexander subdivides the developmental path from novice (or neophyte) to expert into three stages: acclimation, competence and proficiency/ expertise, which she characterizes as follows: Acclimation , the initial stage of expertise development, is characterized by students having limited and fragmented knowledge in the domain at hand. This refers both to domain and topic knowledge. Learners at this stage lack “principled knowledge” (Gelman/ Greeno 1989), “a cohesive and well-integrated body of domain knowledge” (Alexander 2003: 11) and have difficulty to discern the difference between accurate and inaccurate and relevant and tan- 162 Part III: Text Production gential information. They frequently use surface-level strategies, and their interest is rather situational than individual: “The seeds of individual interest, even if planted by meaningful and captivating instruction, have limited opportunity to take root. Thus, there is also an expected reliance on situational interest to maintain novices’ focus and spark their performance” (Alexander 2003: 11). Development from the acclimation to the competence stage is marked by quantitative and qualitative changes in individuals’ knowledge base. “Competent individuals not only demonstrate a foundational body of domain knowledge, but that knowledge is also more cohesive and principled in structure. Further, as the problems typifying an academic domain become increasingly more familiar, competent learners delve into such tasks by applying a mix of surface-level and deep-processing strategies. Moreover, these knowledge and strategy changes in competent learners are linked to increases in individuals’ personal interest in the domain and less dependence on situational features of the environment.” (Alexander 2003: 12) Movement on to the proficiency or expertise stage requires a synergy of all three components: a broad and deep knowledge base, deep-processing strategies and high individual interest. Those who have attained this third stage are not only problem solvers but also problem finders: “These experts are posing questions and instituting investigations that push the boundaries of the domain.” (Alexander 2003: 12) Tab. 7-1 provides an overview of the MDL’s three developmental stages and their characteristics with regard to knowledge, processing and interest. Tab. 7-1: The MDL’s developmental stages and their characteristics Acclimation Competence Proficiency/ Expertise Knowledge - subject-matter - domain limited or fragmented subject-matter and domain knowledge foundational body of domain knowledge that is more cohesive and principled in structure broad and deep knowledge base Strategic processing - surface-level - deep surface-level processing mix of surface-level and deep processing deep processing Interest - situational - individual situational interest individual interest, less dependence on situational interest high individual interest Against this background, it has to be questioned whether we can seriously expect our students to find their own research questions when they write their 7 Writing competence development models 163 first term papers at the beginning of their academic training, a phase when they are still in their acclimation stage or early competence stage (cf. also Kruse 2003: 24). At these stages, students still “require guidance in determining what content is central and what is peripheral, what information is accurate and well supported, and what information is inaccurate or unsubstantiated. These students also need explicit instruction on how to be strategic within a domain, since their strategic processing will often be ineffective and inefficient when left to their own devices” (Alexander 2003: 12). Furthermore, we have to take into account that few students intend to become domain experts (Alexander 2003: 12). A model that takes into account personal characteristics to an even higher extent is Bronfenbrenner & Morris’ (2006) bioecological model of human development. This model and its relevance for the analysis of writing competence development will be discussed in the next section. 7.7 The bioecological model of human development and its implications for the modelling of writing competence development Even more than Alexander’s Model of Domain Learning, Bronfenbrenner & Morris’ bioecological model of human development takes into account characteristics of the learning individual, but also its environment and the demands that the individual is exposed to. Their bioecological model of human development is composed of four components, Process, Person, Context and Time, which they term “defining properties” (794) and which interact dynamically. Process is at the core of the model and refers to the interactions between an organism (e.g., a writer) and its environment, the so-called “proximal processes”. Bronfenbrenner & Morris (2006) consider proximal processes “the primary mechanisms producing human development” (795). With regard to writing skill development, proximal processes can be defined as processes going on during the instances in which a person actually produces text(s). Without actively performing tasks related to the competence to be acquired no competence development can take place. For writing pedagogy, this means that students have to be engaged in writing processes. However, not any writing process will do. The extent to which proximal processes influence development is presumed to vary as a function of the characteristics of the other three components of the model, Person, the immediate and more remote Contexts and Time (Bronfenbrenner/ Morris 2006: 795). Within the component Person , Bronfenbrenner & Morris (2006: 795) distinguish three types of characteristics that they consider “most influential in shaping the course of future development through their capacity to affect the direction and power of proximal processes through the life course”. These 164 Part III: Text Production three types of characteristics involve dispositions , which include personal characteristics such as motivation and persistence; the bioecological resources of ability, experience, knowledge and skill; and demand characteristics, which encourage, or discourage, progressively more complex interaction with the environment (Bronfenbrenner/ Morris 2006: 796, 825). In recent years, especially student dispositions have attracted increased interest in writing research. They have been found to be critical to success in the transfer of learning (Driscoll 2011; Driscoll/ Wells 2012: 1). For example, Slomp (2012: 84) emphasized that “failure to consider the role that intrapersonal factors [ which include dispositions; SG ] play in the transfer process can cloud our ability to assess underlying barriers to transfer”. Following Slomp (2012), Driscoll & Wells (2012) have drawn on Bronfenbrenner & Morris’ (2006) bioecological model of human development to take students’ dispositions into account in the interpretation of their writing behaviour. In contrast to other researchers, Bronfenbrenner & Morris (2006: 810) do not consider dispositions “measures of developmental outcomes”, but as “precursors and producers of later development”. To their view, dispositions do not just interact with development but rather foster or prevent later development, i.e., they begin or prevent the start of processes or sustain them (795, 810; cf. Driscoll/ Wells 2012: 6). For example, dispositions such as curiosity and goal pursuit have a positive effect on development, whereas dispositions such as impulsiveness and distractibility have a negative impact (Bronfenbrenner/ Morris 2006: 810). According to Perkins et al. (2000), dispositions can change over time; they may be context-specific or broadly generalizable (Driscoll/ Wells 2012: 6). Driscoll & Wells (2012: 6) assert “that many different types of dispositions exist and that certain dispositions may be more or less prevalent within an individual learner”. In their dissertations on writing transfer (Driscoll 2009; Wells 2011), they found four dispositions that had an effect on their participants’ transfer of knowledge: value, self-efficacy, attribution and selfregulation. As they point out, all of them “fall under the umbrella of psychological theories of motivation” (Driscoll/ Wells 2012: 7). If students do not value what they are learning in a course or do not see how the knowledge and skills they acquire in a course can be useful for them in the future, they will not engage in the “mindful abstraction” relevant for transfer (Perkins/ Salomon n.y.). This is in line with the expectancy-value theory of motivation (Wigfield/ Eccles 2000; Eccles 2005), according to which students’ motivation, performance, persistence and choice-making are connected to the value they attribute to a particular assignment. One pedagogical implication of this is the need for a clear progression. Another implication is that teachers should explicitly explain the connections between assignments and their relevance for future tasks students will be faced with (Driscoll 2011: 14). Attribution refers to the causes which people consider the origin of their failure or success. According to Weiner’s (2010) attribution theory, people who 7 Writing competence development models 165 believe that the causes of their failure or success lie in themselves, in their abilities or efforts, have an internal locus of control. Individuals who attribute the causes of their failure or success to external factors beyond their control have an external locus of control. If they are not successful, low achievers tend to put the blame for this on somebody else or external circumstances. It is obvious that positive development can more likely be expected from students with an internal locus of control. The pedagogical implications this has is that teachers should give students the experience that they can decide their own fate. According to self-efficacy theory (Bandura 1977), people’s beliefs about their capabilities have an effect on the likelihood with which they take the steps necessary to achieve their goals. Students with high self-efficacy are better at self-regulating their own learning, they work harder and are more persistent than students with low self-efficacy (Driscoll/ Wells 2012: 8). Self-regulation includes the “ability to set reasonable goals, to choose to utilize strategies to achieve those goals, to self-evaluate progress, to manage the physical and social settings so that they serve to support and not distract from those goals, to practice effective time management, to reflect on the success of choices made and strategies used, to understand how performance leads to results, and to be able to make changes to any of these preceding actions to improve future performance” (Driscoll/ Wells 2012: 10). Zimmermann & Risemberg (1997: 76 f.) group self-regulatory processes into three groups: environmental processes, behavioural processes and personal processes. In environmental processes, writers adapt their physical or social environment to their needs, for example, by closing the window to screen out distracting sounds. Behavioural processes include the adaptive use of practices, such as keeping a record of the number of pages that were written per day and continuing or modifying this practice depending on its effectiveness. Personal processes “involve writers’ self-regulation of cognitive beliefs and affective states associated with writing” (Zimmermann/ Risemberg 1997: 76 f.). Optimally self-regulated writers can be assumed to attempt to use all three forms of self-regulation in conjunction with each other (Zimmermann/ Risemberg 1997: 78). According to Zimmermann (2002), self-regulation is not an inherent trait that a learner either has or does not have. It is “a process learners go through when they choose how they will adapt to new learning situations” (Driscoll/ Wells 2012: 9). Many students have not learned to self-regulate their studying well. Especially novices attempt to self-regulate their learning reactively; they “fail to set specific goals or to self-monitor systematically, and as a result, they tend to rely on comparisons with the performance of others to judge their learning effectiveness” (Zimmermann 2002: 69). From Driscoll & Wells’ description of how bad self-regulators go about the task of learning, implications can be derived on how to assist students in becoming better at self-regulating: They should be 166 Part III: Text Production supported 1. in setting themselves attainable goals, 2. in developing the selfreflective capacity to monitor their progress, e.g. by providing them with criteria for this, and 3. in developing the self-control to change their behaviour if their progress does not meet their expectations. These measures all involve metacognitive reflection about their writing and learning (cf. Driscoll 2011: 14) 8 . How self-regulation strategies can be taught will be further discussed in Chapter 8. As mentioned above, self-regulation and self-efficacy are related to each other: “Self-regulation is linked to self-efficacy theory because individuals with high selfefficacy tend to be good at setting reasonable and achievable goals for themselves. An efficacious learner has the predispositions, including motivation, to engage in self-regulation.” (Driscoll/ Wells 2012: 10) 9 Since we still know little about questions such as how and where dispositions are formed, how individual dispositions are related to each other and how teachers can understand, and adapt to, students dispositions in classrooms, Driscoll & Wells (2012: 11 f.) plead for more research on dispositions and their impact on the development of writing skills, for which they provide a list of potential research questions they consider worth looking into. After this illustration of the relevance of the component Person for competence development, the other three components of Bronfenbrenner & Morris’ model shall be described. Context refers to the features of the environment that either foster development or interfere with it. Among the latter are “the growing hecticness, instability, and chaos in the principal settings in which human competence and character are shaped” (Bronfenbrenner/ Morris 2006: 796). With regard to the development of academic literacy, it appears to be legitimate to question whether we grant our students sufficient time to complete their assignments in a manner most profitable for them. Do not time pressure and too many tasks assigned in parallel prevent them from deriving the maximum benefit (and pleasure) from the proximal processes involved in the completion of these tasks? With regard to Time , Bronfenbrenner & Morris (2006: 796) distinguish between Microtime, Mesotime and Macrotime: “ Microtime refers to continuity versus discontinuity in ongoing episodes of proximal processes. Mesotime is the periodicity of these episodes across broader time intervals, such as days and weeks. Finally, Macrotime focuses on the changing expectations and events in the larger society, both within and across generations, as they 8 The online document has no page numbers. The page numbers given refer to the pages of the printout. 9 In a meta-analysis of studies investigating the impact of self-efficacy on writing processes and writing products, Pajares (2003) comes to the conclusion that high selfefficacy has a positive effect both on writing motivation and writing products. 7 Writing competence development models 167 affect and are affected by, processes and outcomes of human development over the life course.” With regard to text competence development, microtime can be interpreted to refer to questions such as whether a writer can complete an assignment on end without interruptions by other tasks and without time pressure or not; mesotime can be interpreted to refer to the time between different writing sessions for the same writing assignment or the continuity or discontinuity with which students are involved in completing different writing assignments in the course of their programs, and macrotime can be interpreted to refer to a person’s entire writing biography. Confronting students with a continuous sequence of writing assignments which build on each other with regard to their difficulty and complexity both within courses (mesotime) and across the curriculum and beyond (macrotime) can be assumed to be more favorable to writing skills development than just intensive ‘doses’ of writing instruction at specific points in their academic training (cf. Bronfenbrenner/ Morris 2006: 798). 7.8 Summary In this chapter, three types of approaches towards modelling writing competence development have been discussed: process-oriented approaches from the perspective of the cognitive sciences with reference to limited workingmemory capacity, product-oriented approaches from the perspective of linguistics, and approaches which focus on personal characteristics of the writer. The three types of models complement each other and, as has been illustrated in this chapter and will be discussed in more detail in Chapter 8, all have implications for writing instruction. From the cognitive-science approaches, didactical strategies can be derived which promote the automatization of lower-level processes before students are confronted with increasingly complex tasks. The complexity reduction that ensues from this can also be fostered by allowing students, in situations of “cognitive apprenticeship” (Kellogg 2008), to ‘observe’ problem-solving strategies as performed by experts before they embark on solving such problems themselves (cf. also Braaksma et al.’s [ 2004 ] concepts of “observational learning” and “learning by doing”, which will be discussed in Chapter 8). Based on Steinhoff’s and Pohl’s findings, training material can be developed which, first, illustrates in a contrastive manner what distinguishes academic writing from popular-science writing and other genres that students are more familiar with, and then includes exercises in which students have to adapt typical formulations found in beginners’ term papers to the conventions of academic writing. As Steinhoff (2007: 430) recommends, such training material ideally does not only focus on the differences between scientific texts and 168 Part III: Text Production other less academic genres but also addresses the functions that are fulfilled by these differences and provides (extracts of) model texts. Steinhoff (2007: 429) terms this approach a “model-oriented approach to the teaching of academic writing” (“modellorientierte wissenschaftliche Schreibdidaktik”) and justifies it in his “reflections on an academic writing pedagogy” (Steinhoff 2007: 425 ff., my transl.). Furthermore, the missing, inadequate and finally adequate use of typical academic textual procedures can be used as an indicator of the stage of academic writing competence development that a writer has attained (Steinhoff 2007: 421). Approaches focusing on motivational aspects emphasize the importance of writing assignments that students deem relevant for their personal development and future career. The implications of all three types of models will be discussed in more detail in Chapter 8. Part IV Writing Instruction 8 Best-practice approaches to writing instruction This chapter provides a survey of best-practice approaches to writing instruction. Although it will not be confined to the teaching of academic writing, special emphasis will be placed on this type of written academic communication. It is particularly interesting from the perspective of cognitive development due to its outstanding relevance for deep learning and academic socialization. The concept of ‘deep learning’ is used here to refer to learning in a manner that the knowledge acquired is reflected and can be transferred to new problem-solving contexts. The opposite of deep learning is surface learning, which, in its worst case, may be learning by heart just for the purpose of reproduction in an exam. Ideally, the writing skills that students acquire and the insights they gain from completing an assignment can be transferred to problem-solving processes involved in future assignments, in other courses and their future lives. Therefore, in accordance with Beaufort (2007: 149 ff.), approaches to writing instruction will be pleaded for “that aid transfer of learning for writers”. How transfer can be fostered in writing instruction will be focused on in Section 8.2. Lea & Street (1998: 157) distinguish between three models of educational research into student writing in higher education: a) the study skills model, b) the academic socialization model and c) the academic literacies model, which, as depicted in Fig. 8.1, are not mutually exclusive but encompass one another. Fig. 8-1: Three models of educational research into student writing in higher education In this chapter, didactical strategies pertaining to each of the three models will be discussed. The study skills model focuses on a “set of atomised skills which students have to learn and which are then transferable to other contexts” (Lea/ Street 1998: 158; Lea 2004: 741). It emphasizes aspects of writing, such as spelling, punctuation, grammar and genre conventions, that can be learned more or less 172 Part IV: Writing Instruction independently from the disciplines, takes a deficit perspective, i.e., tries to “fix” students’ writing problems, and “conceptualizes student writing as technical and instrumental” (Lea/ Street 1998: 158). Its sources lie in behavioural psychology. The academic socialization model underlies writing-in-the-disciplines (WID) approaches. Following this model, the students gradually have to get accultured to the ways of thinking of their specific disciplines. This is no longer possible in general writing courses but has to happen in discipline-specific courses, such as writing-intensive seminars (see Sections 8.6 and 8.7). The sources of this model lie in social psychology, in anthropology and in constructivist education (Lea/ Street 1998: 158). This model, too, has certain shortcomings, which Lea & Street (1998: 159) specify as follows: “Although more sensitive both to the student as learner and to the cultural context, the approach could nevertheless be criticised on a number of grounds. It appears to assume that the academy is a relatively homogeneous culture, whose norms and practices have simply to be learned to provide access to the whole institution. Even though at some level disciplinary and departmental differences may be acknowledged, institutional practices, including processes of change and the exercise of power, do not seem to be sufficiently theorised. Similarly, despite the fact that contextual factors in student writing are recognized as important (Hounsell, 1998; Taylor et al., 1988), this approach tends to treat writing as a transparent medium of representation and so fails to address the deep language, literacy and discourse issues involved in the institutional production and representation of meaning.” The academic literacies model sees literacies as social practices and views “student writing and learning as issues at the level of epistemology and identities rather than skill or socialization” (Lea/ Street 1998: 159). It is embedded in academic discourses and power relationships and requires students to switch practices between disciplines and different settings. What the academic socialization and the academic literacies models share is that “both conceptualize learning and writing in higher education as inseparable” (Lea 2004: 743). However, while the former “focuses on the writing requirements of particular disciplinary and subject areas”, the latter “is more appropriate for the increasing number of interdisciplinary and multidisciplinary contexts, where students are undertaking courses based on a complex mix of disciplinary conventions and knowledge bases. Such courses are particularly common in professional and vocational areas of study, at both undergraduate and postgraduate level, although they are also popular in more academic areas, where they tend to be conceptualized as fields of study, for example, environmental studies and human sciences.” (Lea 2004: 743) Following the academic literacies approach, students must be drawn into the process of knowledge construction (Lea 2004: 754). The growing body of research based on the academic literacies model suggests that “one explanation for problems in student writing might be the gaps between academic staff 8 Best-practice approaches to writing instruction 173 expectations and student interpretations of what is involved in student writing” (Lea/ Street 1998: 160 with further references). Many problems seem to arise because teachers do not make their requirements explicit enough, or are not able to do so, for students to interpret what is expected of them when completing a writing assignment. As Lea & Street (1998) found in interviews with staff and students at two British universities, orientation students are provided with, for example, in the form of departmental style guides, is often limited to formal aspects of writing; orientation concerning aspects such as content and line of argumentation is lacking. The genres that students have to compose at university differ depending on disciplines and on the country of their academic socialization. In the German tradition, the term paper (Seminararbeit) is the most relevant academic genre that students are required to compose in the humanities and the social sciences (Steinhoff 2007; on the history of this genre in Germany, see Pohl (2009). It represents “the didactical correspondent of the academic article” (Ehlich 2003: 20, my transl.). In the Anglo-Saxon tradition, essays are more popular than term papers. Composing term papers requires the full range of competencies that scholars need to write academic articles: the ability to search and select the relevant literature, to specify research questions, to familiarize themselves with the state of the art, to create intertextual references, etc. (Ehlich 2003: 20). Essays, in contrast, are shorter genres that may need a more rigorous argumentation than term papers but allow students to take a more personal perspective without the obligation to systematically discuss prior publications on the topic in question (Kruse/ Chitez 2012: 67). The competence to write term papers has long been expected to develop automatically as a by-product of academic socialization. Even if it does, this is definitely not the most efficient way to acquire it (cf. also Beaufort 2007). In 2003, Björk et al. (2003a: 9) pointed out that European higher education seemed to be “at the historical point where the pedagogy of trial-and-error” was found inadequate and its replacement with “a pedagogy of instruction” was called for. In the United States, the insight that academic literacy development needs to be fostered led to the establishment of writing centres and writing programs more than 30 years ago. Today, more than 90% of the universities in the USA have writing centres (Grimm 1996: 523). In Germany and other European countries, in contrast, this is a fairly recent insight (cf. Girgensohn/ Sennewald 2012: 84; and the articles in Björk et al. 2003b). The first German writing centre was founded in 1993 at the University of Bielefeld (Frank/ Haake/ Tente 2003); writing centres at the University of Bochum (1997), at the University of Education in Freiburg and the European University Viadrina in 174 Part IV: Writing Instruction Frankfurt/ Oder followed. 1 Whereas only few writing centres existed at German universities until 2011, the situation has changed since then. This change has been initiated by the German Federal Ministry of Education and Research through its Quality Pact for Teaching (Qualitätspakt Lehre) , in which German universities could apply for funding to improve the quality of teaching and to provide better orientation for students. 2 For this purpose, approximately 2,000,000,000 € have been devoted to German universities for the time span between 2011 and 2020 (BMBF 2012). This was urgently needed in the wake of the Bologna Reform because degree programs had become shorter and schedules tighter for students and thus teaching had to become more efficient. In addition, student intake in universities has been increasing and students have become more heterogeneous with regard to their educational biographies. 3 These changes have called for adaptations in curricula and teaching methods, compensatory courses, student support services, etc. (see also Russell 2003: V ff.). A number of German universities have tackled these challenges by establishing writing centres and developing programs in their teaching centres which prepare teachers of all faculties and domains for more writing-intensive approaches in their seminars and thus for supporting students in the development of their academic literacy. Teachers of many disciplines, however, have to be convinced of the usefulness of such writing-intensive seminars for the academic development for their students in their specific disciplines first before they are prepared to offer such seminars. They ask for empirical evidence of the role of writing for learning. Langer & Applebee (1987) present encouraging findings from an investigation of the effect of different types of writing (such as note-taking, summarizing and argumentative essays) on learning. 4 Generally speaking, they found that “activities involving writing [ ... ] lead to better learning than activities involving reading and studying only” (Langer/ Applebee 1987: 135). More specifically, they found, first, that the more profoundly content from texts read is manipulated in writing, the more likely it is to be recalled and understood. Second, they found that writing only has a positive effect on learning for those bits of information of a text that have to be focused on in writing. Third, they observed that the type of writing assignment (note-taking, comprehension questions, summarizing tasks, analytic writing tasks) has an effect on processing depth. Analytic tasks lead to deeper processing than more reproducing tasks. And finally, they found that writing has no 1 For a history of writing pedagogy in German-speaking countries, see Girgensohn & Sennewald (2012: 84 ff.). 2 Similar programs had been launched before in other European countries, such as the Netherlands (Kramer/ van Kruiningen/ Padmos 2003) and Norway (Dysthe 2003). 3 For the situation in the United Kingdom, see Scott & Coate (2003). 4 For a research overview of studies that have addressed either this question or, more generally, the question of the effect of writing on reading comprehension or recall, see Langer & Applebee (1987: 91 ff.). 8 Best-practice approaches to writing instruction 175 major effect at all if the content that has to be written about is familiar and relationships between concepts and ideas are well understood (Langer/ Applebee 1987: 130 ff.; see also Applebee 1984). The implications of these findings are that “ [ d ] ifferent types of written tasks promote different kinds of learning, and choosing among them will depend upon the teacher’s goals for a particular lesson or a particular course” (Langer/ Applebee 1987: 131 f.). This leads us to the question of how academic literacy, especially writing, can and should be taught and how teachers can be prepared to support their students’ writing competence development, or, more generally, literacy development, in the best possible manner. These questions will be focused on in this chapter. They can be answered at three different levels: a macro-level, a meso-level and a microlevel. At the macroor university-wide level, the question has to be answered whether universities need didactical policies and central support structures that initiate measures to support students in the development of their literal competences and, if yes, what services they should provide. At the meso-level, decisions have to be made concerning the curricula of entire programs. Such decisions include whether literal competences should be taught in separate general or in domain-specific writing courses, whether their teaching should be integrated into existing domain-specific subject-matter courses, which requires a different conception of the teacher role (Langer/ Applebee 1987), or both, and how the transfer of the knowledge and skills acquired in one course can be fostered in such a manner that it can be adapted for future problemsolving processes. At the micro-level, course syllabi, teaching methods and assignments have to be devised. Before these three levels will be focused on in Sections 8.3 to 8.5, the theoretical framework for literacy pedagogy, or rather a multiliteracies pedagogy, will be outlined in Sections 8.1 and 8.2. 8.1 A theoretical framework for literacy pedagogy: multiliteracies In their “programmatic manifesto”, the New London Group 5 (1996: 63) provides “a theoretical overview of the current social context of learning and the consequences of social changes for the content (the ‘what’) and the form (the ‘how’) of literacy pedagogy” (see also Cope/ Kalantzis 2000). They argue that the linguistic and cultural diversity prevalent in all areas of life and more diversified communication channels and media call for a broader view of literacy, which they term multiliteracies . The concept of multiliteracies encom- 5 The New London Group is a team of scholars from different domains, who met in New London (USA) in 1994 to discuss the state of literacy pedagogy. Their main concerns included “the pedagogical tension between immersion and explicit models of teaching; the challenge of cultural and linguistic diversity; the newly prominent modes and technologies of communication; and changing usage in restructured workplaces” (The New London Group 1996: 62). 176 Part IV: Writing Instruction passes the ability to decode and encode messages in various media (spoken and written text, sound, picture, video, etc., as well as combinations of them) in and for diverse linguistic and cultural communities. A pedagogy of multiliteracies, they suggest, should comprise four components: Situated Practice, Overt Instruction, Critical Framing and Transformed Practice, which will be explained in the following and set in relation to requirements of a pedagogy of literacy suggested by other experts in the domain. 8.1.1 Situated Practice Acknowledging that the human mind is embodied, situated and social, the New London Group (1996: 84) argues that “if one of our pedagogical goals is a degree of mastery in practice, then immersion in a community of learners engaged in authentic versions of such practice” and guided by experts is necessary. 6 This type of immersion is called Situated Practice . The experts should be “people who have mastered certain practices”; they must at least be “expert novices, people who are experts at learning new domains in some depth” (The New London Group 1996: 85). Without expert guidance, situated practice would just be learning by doing, which, without corrective feedback, might lead to the routinization of inferior practices instead of reflected development. The practice that learners are engaged in has to be relevant for their future needs: “There is ample evidence that people do not learn anything well unless they are both motivated to learn and believe that they will be able to use and function with what they are learning in some way that is in their interest.” (The New London Group 1996: 85) For writing instruction, the following requirements can be derived from this: 1. Students should be engaged in writing projects that matter to them and are relevant for their future lives. In Bräuer & Schindler’s (2013: 12) terminology, writing assignments should be “authentic”, “situated”, i.e., related to the learning outcomes of the course in which they have to be tackled, and “learnercentred”, i.e., they should take into account what students can already do and what they need to learn next, and they should give students the leeway to find their own ways of solving problems. Since many students do not aspire to an academic career, they lack motivation to write academically. For them it may be more motivating to learn how to write journalistic genres, for which they need to acquire competencies, such as literature research and selection, summarizing, assessing, which are also relevant for academic writing (cf. Bräuer/ Schindler 2013: 32 f.). The requirement addressed here is in line with Beaufort’s (2007) requirement that writing tasks students have to fulfil need to be embedded in a discourse community. 6 Cf. also the concept of “communities of practice” (Lave/ Wenger 1991; Wenger 1998). 8 Best-practice approaches to writing instruction 177 2. It should not be expected that the version a student submits for the first time fulfils all requirements. It should rather be considered an interim version for which students obtain guidance for improvement. Writing-intensive seminars, in which the complex process of composing a specific text, such as a term paper, is broken down into smaller steps and in which feedback is provided on each step, can be considered to be more beneficial for students’ writing skills development than seminars in which a student obtains a grade (and, hopefully, also feedback) on the version submitted, but no chance to use this feedback for actual improvement (cf. Bräuer/ Schindler 2013: 21). In line with a socialconstructivist pedagogy of literacy, Bräuer & Schindler (2013) plead for writing-intensive seminars in which students are promoted in the development of their literal competences by means of complex writing arrangements (Aufgabenarrangements) . Bräuer & Schindler (2013: 34) define writing arrangements as didactically-motivated sequences of assignments related to the reception and production of texts whose overall objective goes beyond the objectives of the individual assignments and which, in the long run, guide learners on their way to attaining more complex competencies. Writing arrangements consist of assignments for auxiliary texts (Hilfstexte) , which students compose just for themselves (“writer-based prose”), and for transfer texts (Transfertexte) , which students write for a specific readership and purpose (“readerbased prose”; Bereiter/ Scardamalia 1987). Auxiliary texts may be bibliographies, summaries, extracts and comments. Whereas auxiliary texts help students to document or generate ideas and reflect on their own difficulties and insights, transfer texts represent preparatory material on which students may draw to compose the genres whose competent composition forms the ultimate goal of the writing arrangement, i.e., texts only become transfer texts when they are used for that purpose (Bräuer/ Schindler 2013: 34 f.). Such writing arrangements comprising subtasks with a clear progression lead to a complexity reduction, allow interim feedback and thus guided development without the learner losing the final, authentic goal out of sight. 8.1.2 Overt Instruction Overt instruction “includes centrally the sorts of collaborative efforts between teacher and student wherein the student is both allowed to accomplish a task more complex than they can accomplish on their own, and where they come to conscious awareness of the teacher’s representation and interpretation of the task and its relations to other aspects of what is being learned.” (The New London Group 1996: 86) Accomplishing a task collaboratively that is more complex than what the students could manage on their own distinguishes Overt Instruction from simple learning by doing. The teacher’s intervention provides the impetus for development. Overt instruction is always combined with the use of metalanguage 178 Part IV: Writing Instruction (The New London Group 1996: 86), which allows students to take a step back and reflect on what they are doing. Components of Overt Instruction are feedback and scaffolding. Scaffolding refers to material that supports learners on their way to solving a problem that would be too difficult for them to solve completely on their own. It sets free the learning potential that Vygotsky (1978) termed the “zone of proximal development”. 8.1.3 Critical Framing Critical Framing makes students aware of the scope of application of what they have learned from a specific assignment and thus prevents them from overgeneralizing what they have learned (cf. also Beaufort 2007: 177 ff.): “The goal of Critical Framing is to help learners frame their growing mastery in practice (from Situated Practice) and conscious control and understanding (from Overt Instruction) in relation to the historical, social, cultural, political, ideological, and value-centered relations of particular systems of knowledge and social practice. Here, crucially, the teacher must help learners to denaturalize and make strange again what they have learned and mastered.” (The New London Group 1996: 86) Critical framing can be supported by familiarizing students with constructs such as the Karlsruhe comprehensibility concept (see Section 4.3), which provides a framework from which genre-specific criteria for text assessment can be derived. This also includes assessment criteria for translations, which depend on the function of the target text. 8.1.4 Transformed Practice Situated Practice with Overt Instruction and Critical Framing should lead to reflective practice, practice in which students both apply and revise what they have learned resulting in Transformed Practice (The New London Group 1996: 87). In this manner, students learn to continuously adapt their behaviour to changing contexts. Bräuer & Schindler (2013: 25) formulate requirements for writing instruction which supports Transformed Practice: 1. Writing should be conceived as a reflective practice, as a method that accompanies the solution of complex working processes. 2. Writing should include peer feedback among the learners of a group to support the solution of the tasks learners are working on. 3. Peer tutoring should be available as an extracurricular form of writing support. Requirement 1 should be fulfilled by any writing arrangement. The objectives underlying the requirements 2 and 3 can also be achieved in alternative manners, for example by feedback from teachers or masters in the field, by observing masters performing a task (see Section 8.5.1.2) or by collaboration follow- 8 Best-practice approaches to writing instruction 179 ing the principle of cognitive apprenticeship (see Sections 8.5.1.3). An alternative designation for what the New London Group calls Transformed Practice is transfer . How transfer can be fostered by teachers will be focused on in the next section. 8.2 Teaching for transfer Transfer is the use of knowledge or skills learned in one context in another context and thus the central objective of formal education (Perkins/ Salomon n.y.: 1 7 ; Graves 2004: 439). Whatever students learn from writing assignments they are required to complete should be transferable to tasks they will have to accomplish in their future education as well as in their private and professional lives. When devising writing assignments, teachers should therefore have in mind for what contexts and future tasks the knowledge and skills students are supposed to acquire from the assignments can be used. Very often we do not get the transfer we want. According to Perkins & Salomon (1988; n.y.), we do not have to put up with this. In an effort to explain how transfer can be fostered, they contrast three theories of transfer: the “Bo Peep theory”, the “Lost Sheep theory” and the “Good Shepherd theory” (Perkins/ Salomon n.y.). The “Bo Peep theory”, named after the children’s rhyme “Leave them alone and they’ll come home / wagging their tales behind them.”, assumes that the knowledge and skills a person has learned anywhere will be transferred (will “come home” like Bo Peep’s lost sheep) to wherever they will be needed without any active fostering of the transfer by the teacher. Although this tacit theory is implicit in what happens in many classrooms, psychologists have found that “very commonly, transfer does not occur in the absence of blatant similarities between the originally learned task and the transfer task” (Perkins/ Salomon n.y.: 4). According to the rather pessimistic “Lost Sheep theory”, transfer, especially “far transfer” in contrast to “near transfer” 8 , is a lost case. If this theory were true, nothing much could be done to secure transfer. Empirical evidence, however, points to the contrary (Perkins/ Salomon n.y.: 4). 7 The online document has no page numbers. The page numbers given refer to the pages of the printout. 8 “Near transfer” and “far transfer” are two concepts coined by Perkins & Salomon (n.y.: 2) to refer to the transfer of knowledge and skills to contexts which are very similar to the context in which they were learned (near transfer) and transfer for which connections are made “to contexts that intuitively seem vastly different from the context of learning” (far transfer). They admit that “the far-near contrast is very fuzzy”. Nevertheless they believe that, although any precise differentiation is impossible, “the contrast proves useful in broad talk about transfer”. 180 Part IV: Writing Instruction In their “Good Shepherd theory”, Perkins & Salomon (n.y.) reconcile the findings that, in some cases, desired transfer does not occur, whereas, in other cases, even far transfer has been observed. According to this theory, transfer has to be fostered actively by teachers, the “good shepherds”. Depending on the type of transfer desired, two strategies can be used to achieve this: “bridging” and “hugging”. Bridging means “that the teacher helps the students to build a bridge from the context of learning to other contexts of potential application” (Perkins/ Salomon n.y.: 6). This can be achieved through questions calling for analogies and respects in which these analogies exist. With regard to the analogies, Perkins & Salomon (n.y.) make a distinction between “high-road transfer” and “low-road transfer”. High-road transfer requires “the mindful abstraction of a principle, the effortful search in one’s memory, the selection of the appropriate principle and, finally, its application to a new instance” (Perkins/ Salomon n.y.: 6). Lowroad transfer is a kind of automatic transfer, based on a skill that is “mastered near to automaticity by continuous repeated practice” and “applied to a new situation that is very similar to the situations in which it has been practiced in the past” (Perkins/ Salomon n.y.: 6 f.). High-road transfer can be both “forward reaching” and “backward reaching”. Perkins & Salomon (1988: 26) define these terms as follows: “In forward-reaching high road transfer, one learns something and abstracts it in preparation for applications elsewhere. [ ... ] In backward-reaching high road transfer, one finds oneself in a problem situation, abstracts key characteristics from the situation, and reaches backward into one’s experiences for matches.” Bridging is the strategy to be applied when high-road transfer is desired. The questions asked should “demand of the students mindful abstraction from the case at hand, rather than automatic reflective match-making characteristic of low-road transfer” (Perkins/ Salomon n.y.: 7). Perkins & Salomon illustrate the type of questions to be asked in bridging by means of the biology teacher who has discussed the human blood circulatory system and fosters bridging by asking students for other circulatory systems and their analogies to and differences from the blood circulatory system. Students come up with plumbing systems, electricity systems, oil pipelines, traffic, etc. Other bridging patterns that Perkins & Salomon suggest are 1. asking students to brainstorm how what they have just learned may apply in other situations, 2. asking students to explore in detail what analogies and contrasts exist between something they have just learned about and some related phenomenon and 3. asking students to take a metacognitive perspective on a problem by requiring them to specify how to best solve the problem at hand before actually beginning to solve it. They see the ultimate goal of teaching for transfer “in helping students to catch the spirit and craft of transfer themselves”, so that they no longer need teacher support for transfer. “Once their minds are set for transfer, they will begin 8 Best-practice approaches to writing instruction 181 spontaneously to look for the connections that earlier had to be provoked” (Perkins/ Salomon n.y.: 8). Where low-road transfer is sufficient, hugging can be applied. Hugging involves making “the learning situation more like the situations to which transfer is desired” because then “the automatic reflexive triggering of responses will frequently occur” (Perkins/ Salomon n.y.: 8). They illustrate the hugging strategy by means of the example of teaching students to write good topical sentences. This frequently happens by requiring students to circle topical sentences in a text or marking in a list of sentences those ones that could be used as topical sentences. These tasks, however, do not much resemble the task of having to write topical sentences oneself. Hugging can be achieved here by devising exercises in which students actually have to formulate topical sentences (Perkins/ Salomon n.y.: 8). Alternative hugging strategies that Perkins & Salomon (n.y.: 9) suggest are: 1. simulation games, 2. problem-based learning, i.e., learning in which students do not learn content by directly being presented with it but through tackling problems that demand the use of the content to be acquired, and 3. mental practice, e.g., preparing for a job interview by imagining what it would be like, i.e., making a kind of mental movie of the situation. How can this be transferred to writing instruction? Writing instruction in many institutions suffers from something that Perkins (1986), with regard to education in general, has called a “disconnected curriculum”: “The subject matters are taught in ways encapsulated from one another and sealed off from the lives students live outside of school [ and this also applies to university; SG ] , not to mention the lives they will live after they have completed their schooling [ and university education; SG ] . Partly, this is a problem of transfer. But partly, the disconnected curriculum reflects traditions that have become entrenched in educational practice: A good deal of what youngsters [ and this also applies to university students; SG ] learn does not connect very well to anything else but the class in which it is taught.” (Perkins/ Salomon n.y.: 9) How disconnected writing curricula can be transformed into connected ones, is one of Beaufort’s (2007) main concerns. Based on an ethnographic longitudinal case study of one student’s (Tim’s) writing biography, she makes a number of recommendations (Beaufort 2007: 151 f.), three of which I consider to be particularly valuable: 1. Writing assignments should become exercises of rhetorical problem solving, i.e., their completion should prepare students for text production tasks they will encounter in their subject-matter courses and later private and professional lives (see also Section 8.7). 2. There should be a clear progression in writing instruction across the curriculum. This means that devising writing course syllabi is not a matter that can be solved at the micro-level, i.e., the level of individual courses, but has to 182 Part IV: Writing Instruction be addressed at the level of university programs (meso-level) or even at the university level (macro-level). The measures to be taken at these levels will be addressed in the next sections. Beaufort (2007: 153; highlighting in the orig.) emphasizes the progression requirement by stating: “[D]epartment chairs, writing program administrators, and deans would greatly increase the ‘return on investment’ in writing instruction if they foster opportunities for faculty to create sequential, developmentally-sound writing sequences that extend across courses in a major. ” This involves that “we have to get over the notion that instructors are minidictators of their own private domains” (Smit 2004: 193). Without a clear progression both within and across courses, the effect of writing instruction could be nothing than what Bereiter & Scardamalia (1993: x) referred to as “increasingly fluent bad writing”. 3. Following the Good Shepherd theory, students “ need guidance to structure specific problems and learnings into more abstract principles that can be applied to new situations ” (Beaufort 2007: 151; highlighting in the orig.). This can happen through bridging strategies and with reference to the Karlsruhe comprehensibility concept (see Section 4.3). The Karlsruhe comprehensibility concept is particularly useful in raising students’ genre awareness since it emphasizes the dependence of textual features on the communicative functions of the texts in question. In addition to these recommendations, Driscoll (2011: 14) suggests that teachers should involve students in metacognitive reflections in various ways, for example, by requiring them to write a reflective comment on their writing and learning processes for each assignment they have to complete and a final reflective paper “that connects course concepts to their other coursework and anticipated writing needs”. Downs & Wardle (2007) and Wardle (2009) suggest to make writing the topic of a writing course itself. Recognizing that there is no such thing as a universal academic discourse or general writing skills, which are frequently named as the goals of first-year composition courses, they have replaced such first-year composition courses by an “Introduction to Writing Studies”, which “seeks [ ... ] to improve students’ understanding of writing, rhetoric, language, and literacy in a course that is topically oriented to reading and writing as scholarly inquiry and encouraging more realistic understandings of writing” (Downs/ Wardle 2007: 553). Addressing questions such as how writing works, how people use writing, what problems related to reading and writing occur and how they can be solved, students in such courses “read writing research, conduct reading and writing auto-ethnographies, identify writing-related problems that interest them, write reviews of the existing literature 8 Best-practice approaches to writing instruction 183 on their chosen problems, and conduct their own primary research, which they report orally and in writing” (Downs/ Wardle 2007: 558). In this manner, Downs & Wardle (2007: 559) expect students to learn that “writing is conventional and context-specific rather than governed by universal rules”. From students’ end-of-semester reflections, Downs & Wardle (2007: 572 f.) learned that an increased self-awareness about writing, improved reading skills and a new understanding of research writing as conversation were among the most prevalent outcomes of the course for their students. These are outcomes that can be assumed to be useful in subsequent courses. Selfreflection, explicit abstraction of principles and alertness to one’s context as proven means of facilitating transfer are fostered in such courses. Although Downs & Wardle (2007: 577) admit that longitudinal studies will be needed to show whether this approach really improves writing skills, the feedback they have received so far from their students is promising. Though I believe that their approach can be very useful for students in the humanities and the social sciences, my own experience in using it with engineering students makes me doubt whether it also works for them. They preferred writing on topics from their disciplines and were not interested at all in linguistic theory. I solved the problem by having them write about their disciplines but in genres such as popular-science articles that were comprehensible for a wider audience including myself. In this manner, I made sure that I myself was able to comment not only on the language, which is inseparable from the content, but also on the content, at least on its comprehensibility and plausibility. What may also play a role in transfer, are “threshold concepts”, i.e., ideas within disciplines “without which the learner cannot progress” (Meyer/ Land 2006a: 1). In composition studies, the relevance of threshold concepts has only recently been recognized (Adler-Kassner/ Majewski/ Koshnick (2012: 2) 9 . Examples of threshold concepts for composition studies and writing research are ‘genre’, ‘audience’, ‘purpose’ or ‘text function’ and ‘discourse community’. Threshold concepts are counter-intuitive to the learner, they exist often tacitly and are liminal, i.e., they are ‘portals’ through which learners must pass, which may be experienced as troublesome (Perkins 1986: 40; Meyer/ Land 2006a: 9 ff.) and may happen in iterative and recursive stages (Adler-Kassner/ Majewski/ Koshnick 2012: 2). Meyer & Land (2006b: 23) distinguish three stages: the preliminal stage, the liminal stage and the postliminal stage. In the preliminal stage, the learners’ tacit views get destroyed as they grapple with the new threshold concept. In the liminal stage, learners become aware of working with the threshold concept. In the postliminal stage, learners begin to think of the threshold concept like members of the respective scientific community, apply it in a discursive manner and use it in analyses. From these stages, the idea can be derived that students’ acculturation into their scientific commu- 9 The online document has no page numbers. The page numbers given refer to the pages of the printout. 184 Part IV: Writing Instruction nity can be measured by analysing whether they are still in the stage of tacit consciousness or already in the stage of discursive consciousness about threshold concepts, which is relevant for high-road transfer. Adler-Kassner, Majewski & Koshnick (2012: 8 ff.) illustrate such an analysis in an exemplary manner for three students. The relevance of threshold concepts for writing competence development can be illustrated by means of the following examples: In the early stages of their writing development, students tend to generalize rules about writing that they may have been given or may have developed themselves based on their experience. Typical examples of such rules are: Never use the first person. Avoid passive-voice constructions. Only when students recognize that texts come in various genres, i.e., when they learn the threshold concept ‘genre’, do they realize that there are no absolute, but genre-specific rules. Teachers should avoid giving students general rules but rather make them aware of the function of texts and their audiences’ requirements so that students can develop for themselves the criteria that a specific text they have to compose has to fulfil. Writing expertise does not show in being able to write texts of a specific genre or specific genres in an excellent manner but rather in writers’ flexibility to adapt to the requirements of genres they may never have written before but which they can compose nevertheless because they have learned to analyse their audience, functions, conventions, etc. and are then able to compose them accordingly (cf. Adler-Kassner/ Majewski/ Koshnick 2012: 3). As Adler-Kassner, Majewski & Koshnick (2012) illustrate, composition studies and history share threshold concepts, such as ‘audience’, ‘context’, ‘text purpose’, that students should be made aware of in order to foster transfer from writing courses to history courses. They conclude that “ [ t ] he ultimate goal of General Education courses [ ... ] might be reconsidered to focus on identifying, comprehending, and applying multiple threshold concepts across the curriculum” (Adler-Kassner/ Majewski/ Koshnick 2012: 13). Various other suggestions for encouraging students to reflect in the writing classroom can be found in Yancy (1998). For an example of an entire program based on the principle of writing for transfer, see Boone et al. (2012). Apart from a) a clear progression within and across courses, b) content and tasks that students experience as relevant for their future lives, and c) pedagogical strategies that foster transfer, there is another bundle of factors that have an impact on student writing transfer: student dispositions. They have not been widely addressed in writing transfer research but have received considerable attention in the fields of education and psychology (Driscoll/ Wells 2012). How they affect writing development has been looked into in Section 7.7 in connection with the bioecological model of human development. 8 Best-practice approaches to writing instruction 185 8.3 Approaches to writing instruction at the macro-level The previous sections of this chapter have outlined the theoretical framework into which literacy development can be embedded. Among other factors, they have emphasized the importance of situated practice and discourse community knowledge for literacy development in general and writing competence development in particular. Those who are most familiar with the practice in their field and possess discourse community knowledge with regard to the academic genres of their domain are subject-domain specialists publishing in their fields. Does this mean that such subject-domain specialists, i.e., physicists, biologists, historians, etc., are better teachers of academic literacies than those who have a background in linguistics, language teaching and related fields? In certain respects, the answer is yes, but in other respects, it is no. Starting from Beaufort’s (2007) model of writing competence, subject-domain specialists can be assumed to possess more subject-matter knowledge and discourse community knowledge in their specific disciplines than language specialists can be expected to posses in them. They can also be assumed to possess more genre knowledge with regard to the specific genres they produce frequently (cf. also C. Shanahan 2004: 90). Language specialists, on the other hand, possess more writing process knowledge and rhetorical knowledge. Some of them may also be proficient in a wider repertoire of genres than subject-domain specialists and, if they have focused on writing skills development in their training and professional practice, have more experience in teaching writing. This division of competences among the two groups seems to call for co-teaching approaches in which students could benefit from both sources of expertise. Such a conclusion, however, would ignore a number of aspects. First, being an expert in a domain does not automatically imply that the expert is able to pass his knowledge on. As we know from expertise research (Sternberg 1997; Ericsson/ Smith 1991: 2; Anderson 3 1990), growing competence in a domain goes hand in hand with automatization or proceduralization of knowledge, which leads to cognitive relief but also to the fact that some knowledge an expert uses for problem-solving is no longer available to conscious awareness and thus cannot be articulated and explained to learners, which, however, is a requirement for bringing it to the learners’ awareness. With regard to writing expertise, Russel (1995: 70) states (cf. also Polanyi 1966): “A discipline uses writing as a tool for pursuing some object. Writing is not the object of its activity. Thus, writing tends to become transparent, automatic, and beneath the level of conscious activity for those who are thoroughly socialized into it. [ ... ] As a result, experts may have difficulty explaining these operations to neophytes.” Second, developing literacy in one’s students requires a conception of one’s teacher’s role as a learning facilitator (Kember 1997). Many subject-domain 186 Part IV: Writing Instruction specialists complain about the poor writing skills of their students but do not consider it their job to assist students in developing their writing skills. If teams of teachers are formed in which the subject-domain specialist holds such a view, team teaching will not bring any benefit that could not also be achieved by a separation of subject-domain courses on the one hand and writing courses on the other. A prerequisite of being able to tap subject-domain specialists’ knowledge for students’ writing skills development is a change in teachers’ conceptions of their roles as teachers from a “teacher-centred/ content-oriented approach” to a “student-centred/ learning-oriented approach” (Kember 1997; Kröber 2010) and at the same time the willingness to teach writing-intensive seminars (for writing-intensive seminars, see Sections 8.6 and 8.7). To bring about this change, teaching centres and writing centres are required as central support structures at universities. Third, co-teaching is expensive. This raises the question whether there are less expensive alternatives to paying two teachers for one course. Craig (2013: 154 ff.) refers to “co-teaching” as “integration” and suggests three alternatives of interdisciplinary collaboration between subject-domain specialists on the one hand and writing and EFL/ ESL specialists 10 on the other hand: “consultation”, “coordination” and “adaptation”. One form of consultation is the employment of writing fellows in subjectdomain seminars (Bazerman et al. 2005: 110). Writing fellows are trained peer tutors enrolled in the same university program as the other participants of a course they accompany. Writing fellows attend the writing-intensive seminars themselves and instruct seminar participants on how to give peer feedback, they initiate discussions and are available to discuss assignments and to provide feedback on them (cf. Girgensohn/ Sennewald 2012: 93). Their prior training and acquired meta-competence concerning writing skills allows them to make the subject-domain teacher aware of challenges students are facing in the course and to elicit explanations from the subject-domain teacher that would otherwise remain implicit. Ideally, the same writing fellows accompany a specific seminar several times and train their successors when they leave university, so that a knowledge base is built up of typical challenges students encounter in the seminar and problem solutions that have proved to be effective and that can then be passed on. As for teacher training, teaching centres and writing centres are required which train these writing fellows. Another form of consultation Craig (2013) describes is practised when subject-domain specialists can consult writing specialists in teaching centres or writing centres to receive feedback on the assignments they intend to require their students to complete. For this purpose, a website with an upload option 10 Craig (2013) focuses on EFL/ ESL writing and therefore only mentions collaboration between subject-domain specialists and EFL/ ESL teachers. The four forms of collaboration she suggests, however, are also applicable to L1 writing contexts. 8 Best-practice approaches to writing instruction 187 for such assignments may be provided. Responses on assignments can, for example, first be given in a double-blind process by two student peer tutors and the aggregated feedback then be delivered by writing centre staff (Anson 2014). Ideally, subject-domain specialists on the one hand and language teachers and writing specialists on the other also serve each other as consultants. In this type of consultation, the language teachers and writing specialists provide advice on assignments and criteria of assessment that take into account students’ stage of development; the subject-domain specialists provide the language teachers and writing specialists with model texts which provide insight into what counts as successful writing in the respective discipline. Another model of interdisciplinary cooperation is coordination. Following this model, communication courses are linked to subject-domain courses and support students in their work on the assignments (either oral presentations or written assignments) they are expected to complete for their subject-domain courses. A weaker form of coordination is adaptation. Craig (2013: 157) uses this term to refer to forms of collaboration in which subject-domain courses and writing or EFL/ ESL courses are not directly linked to one another but in which writing instructors or EFL/ ESL teachers invite their students to work on their disciplinary writing projects in the writing or EFL/ ESL classroom. Apart from the support from the four collaborative approaches mentioned so far, students may also require writing support outside the regular course hours. As far as subject-domain-specific support is concerned, this needs to be provided by subject-domain specialists. Support on the writing process, rhetorical and language-related problems can be provided by language specialists and peer tutors, who, too, need special training, which can be provided in teaching centres. The extracurricular activity of peer tutoring (German Schreibberatung ) goes back to Kenneth Bruffee who pleaded for it in the 1970s (cf. Girgensohn/ Sennewald 2012: 81). Principles of peer tutoring are nondirectivity in the sense of Carl Rogers (1999), who introduced non-directive counselling into psychotherapy, and addressing higher-order concerns before lower-order concerns, i.e., content and organization should be addressed before language and formal aspects (cf. Girgensohn/ Sennewald 2012: 91). 11 All the measures mentioned so far have in common that they need to be supported by teaching and writing centres as central support structures at university level. Writing centres are also the institution which can cater to students’ writing skills at the study skills level in Lea & Street’s (1998) model. With regard to study skills, it has to be taken into account that students come with rather heterogeneous levels of language proficiency both in their L1 and even 11 Peer tutors must not be confused with writing coaches. Writing coaching has the purpose of gradually changing writers’ individual writing behaviours by familiarizing them with new writing strategies suited for their developmental stages and cognitive resources (Bräuer/ Schindler 2013: 29). 188 Part IV: Writing Instruction more so in their L2 (in many cases English if this is not the students’ native tongue). To derive the maximum benefit from disciplinary writing courses, they need to have automatized certain elementary skills: “If students are to become good writers, we need to help them become strategic, knowledgeable, and motivated writers who are not hampered by inefficient or faulty transcription and sentence construction skills.” (Graham/ MacArthur/ Fitzgerald 2007: 5) It is impossible to cater to the various language-related problems of students in disciplinary courses without boring those who already possess the competencies that others still need to develop and without running the risk of losing the actual domain-specific topic of the seminar out of sight if some students need substantial language support. Therefore, additional writing courses for academic writing in the L1 and, if English is not the L1, in English are needed. These courses can be differentiated following the linguistic proficiency levels students have achieved when starting these courses. These proficiency levels can be determined using placement tests. These courses need not necessarily be domain-specific and, for practical reasons, cannot be domain-specific in many cases. Such practical reasons may be groups that are becoming the smaller the more specific the topic covered is, and a lack of qualified teachers which combine language-teaching competence with subject-domain-specific competence. General courses are a viable option for teaching language skills if they also familiarize students with methods of how to increase their specialized vocabulary and how to familiarize themselves with subject-domain specific genre conventions, i.e., if they equip students with what they need for life-long learning. Those qualified for teaching such general courses are definitely not subjectdomain specialists but language-teaching specialists. Writing courses focusing on language-related issues, however, whether they are general or disciplinary ones, cannot replace writing-intensive seminars in the disciplines but can only complement them for reasons that will be provided in Section 8.6. What follows from the above line of argumentation is that a combination of different approaches towards the development of academic literacy among students is the best option. It includes: • general writing courses focusing on the development of language proficiency • teaching centres and writing centres which promote a student-centred and learning-oriented approach to writing among teachers, familiarize them with the concept of writing-intensive seminars, train both writing fellows and peer tutors and organize continuous communication among all those involved in fostering students’ literacy development across faculties 8 Best-practice approaches to writing instruction 189 • writing-intensive seminars by trained subject-domain specialists assisted by trained writing fellows • collaboration between subject-domain specialists on the one hand and language teachers and/ or writing instructors on the other following the models of integration (co-teaching), consultation, coordination or adaptation • peer tutoring as an extracurricular activity for student support upon request The above measures can be complemented by writing groups ( Schreibgruppen ; Girgensohn/ Sennewald 2012: 95). As the other constellations mentioned above, writing groups also follow the principle of collaborative learning. They are groups of students who work on similar writing tasks, such as their bachelor’s or master’s theses in a specific domain, and come together to discuss and solve problems collaboratively and to give each other feedback. Approaches to writing instruction and writing competence development which comprise all or only some of the measures and services mentioned above can be initiated by a single teacher but their successful implementation is a challenging endeavour which involves complex changes in the teaching philosophy of faculties, departments and their staff and, therefore, requires the cooperation of at least all those involved in a university program and, ideally, also of an entire university. The teaching staff and the university administration must be convinced of the fact that thinking skills are among the most important outcomes of university education, if not THE most important outcome, and that the most effective way to foster the development of thinking skills is through the development of writing skills (cf. also Banzer/ Kruse 2011: 9). To bring about these changes, support structures are needed. Teaching and writing centres 12 can be such support structures both for teacher, writing fellow and peer tutor training and for initiating the organizational changes needed. At the same time writing centres can provide direct support for students who seek to develop their writing skills outside the classroom. Writing centres can start with supplying general writing courses, peer tutor training and consultation hours for those who seek support for completing their writing tasks. Simultaneously, writing centres should initiate discussions in the faculties and degree programs on how they themselves can support the development of academic literacies in a more systematic and integrated manner. 12 Justus Liebig University in Giessen/ Germany has a more encompassing centre, the Centre for Competence Development ( ZfbK - Zentrum für fremdsprachliche und berufsfeldorientierte Kompetenzen) , which takes over the role of fostering innovation in teaching. It comprises a language centre, a centre for interdisciplinary skills, a writing centre, a teaching centre, a career centre and an alumni service with the teaching centre and the writing centre fostering initiatives in the fields of teaching for transfer and writingintensive teaching. 190 Part IV: Writing Instruction Writing centres should then also provide assistance to all those who are prepared to take steps towards such measures. Before changing curricula and assignments, agreement must be obtained on the purpose of the changes and the desired outcome. As Langer & Applebee (1987: 87) found: “It may be much more important to give teachers new frameworks for understanding what to count as learning than it is to give them new activities or curricula.” Although writing centres are usually found in universities, Bräuer (2000; 2003) also pleads for the establishment of similar services, so-called reading and writing centres (Schreib- und Lesezentren) , in secondary education. Collaboration between university writing centres and reading and writing centres at schools by, for example, having university peer tutors visit schools to work with school students, may provide students in teacher training programs with hands-on experience with their future audience and at the same time smoothen future university access for school students. Research on the impact of all these measures and their interplay is needed to optimize an institution’s literacy development policy on a regular basis wherever shortcomings are detected. How this can be implemented is addressed by Wardle & Roozen (2012) in their “ecological model of assessment”. They call it ecological because it “understands an individual’s writing abilities as developing across an expansive network that links together a broad range of literate experiences over lengthy periods of time” (Wardle/ Roozen 2012: 108) and takes into account students’ experiences both obtained in formal education and outside. The objective of an assessment following this model “is to offer students, teachers, departments, institutions and other stakeholders a fuller, richer account of the kinds of experiences with writing that are informing students’ growth as writers throughout their undergraduate years.” (Wardle/ Roozen 2012: 107) Such integrated literacy development policies require support across programs and faculties, they “require relationship-building among and across groups that often do not talk” and, as Wardle & Roozen (2012: 114) concede, will have to “be implemented in small pieces over some extended period of time”. I am convinced it is worth the efforts. 8.4 Approaches to writing instruction at the meso-level Section 8.3 has provided reasons for the establishment of teaching and writing centres as central support structures fostering the development of student academic literacy. This section addresses the question of departmentand program-wide policies of writing instruction, of how to integrate writing in- 8 Best-practice approaches to writing instruction 191 struction into individual degree programs. 13 Ideally, writing instruction at the post-secondary level should meet the following requirements: 1. Integration of content-and-language-integrated learning and literacy development This requirement is based on the insight that writing is not just a means of documenting knowledge but a means of learning (Langer/ Applebee 1987: 91 ff.; for a summary of their findings, see 132 f.) and of developing critical thinking skills (Beaufort 2007: 25). Critical thinking does not come automatically with the acquisition of knowledge, as Bereiter & Scardamalia (1993: 45) describe in a very illustrative manner: “If knowledge is just items in a mental filing cabinet, then it is easy to acknowledge that an expert must have a well-stocked filing cabinet, but that is like saying that a cook must have a well-stocked pantry. The pantry is not the cook, the filing cabinet is not the expert. What counts with cooks and experts is what they do with their material in their pantries or memory stores.” In the USA, such insights have led to two movements: the writing-in-thedisciplines movement (WID) and the writing-across-the-curriculum movement (WAC). Both movements resulted from the insight that writing is a means to learn. The WID movement took account of the fact that contentand-language-integrated learning and writing in the disciplines synergetically increase both domain-specific knowledge and domain-specific literacy. In addition, it also increases writing motivation (cf. Kellogg 2008: 15). The WAC movement is based on the additional insight that literacy development is a continuous process that needs to be fostered not just in a few individual courses but throughout a university program. As Craig (2013: 8) specifies, WAC is “oriented towards student learning and toward process and audience rather than toward an immediate focus on the correctness of grammatical and syntactical features. Writing-across-the-curriculum practice is student-centred, and thus tailored to the student’s developmental level and his ability to think critically. Teaching and classroom activities are active rather than lecture-based. Writing-across-thecurriculum pedagogy views student writing practice and revision as recursive and generative rather than a routine of drills. In WAC, writing and oral presentations are based on meaningful content and oriented toward the transformation of information into knowledge. Writing-across-the-curriculum practice stretches from the most basic writing-to-learn activity in the classroom to academic writing throughout the university and then Writing in the Disciplines with its authentic, situated, and complex communication objectives. Moreover, as communication demands on students evolve, writing-across-the-curriculum practice now includes not only 13 In 2003, Björk et al. (2003a: 13) deplored that there was no European example of a writing-in-the-disciplines program at that time. 192 Part IV: Writing Instruction writing but also oral presentation, the use of informational visuals, and an everincreasing array of multi-modal communication acts.” (Craig 2013: 9) 14 Integrating writing instruction into domain-specific courses has the advantage that, in this manner, it can be practiced continuously. General writing courses may lead to “massed practice”, which, as Kellogg points out, is less effective than “spaced practice”, which leads to long-term learning instead of just an immediate training performance, in which “the learning does not transfer as effectively to a different task in the future” (Kellogg 2008: 18). In sum, integrating writing instruction into discipline-specific courses has the advantage of the synergy effects mentioned above, higher motivation and also continuity in the development of writing skills. This does not only apply to writing in the L1 but also to proficient writing in the L2, which “cannot be achieved in a single course nor can the ability to write within the disciplines be mastered without practice that is situated within an authentic context” (Craig 2013: 4). Therefore Craig pleads for an increase in opportunities for writers to write in their L2, but not by offering them more discipline-independent L2 language courses in which writing is practised but by creating more opportunities for students to write extensively in their L2 in existing subject-matter courses following a process-oriented approach (Craig 2013: 7). This is only possible if disciplinary faculty, writing instructors and EFL/ ESL specialists cooperate and L2 writing instruction does not just take place in language or composition courses but is moved to an interdisciplinary space (Matsuda 1998; 1999; Matsuda/ Jablonski 2000). This interdisciplinary space could come about if, for example, disciplinary faculty sought advice on their writing assignment design from writing instructors and communicated the objectives of these assignments and their expectations concerning student outcomes to writing instructors. Writing instructors could then make sure that the assignments were formulated in a clear manner and, in their writing courses, support students in their work on these discipline-specific assignments instead of confronting students with more general assignments as they are common in general writing courses (Craig 2013: 8; see Section 8.3). Beaufort (2007) mentions additional advantages of an integrated approach. She emphasizes the central role that discourse community knowledge plays for writing expertise. Without this knowledge, Beaufort asserts, students just write for getting grades, for “doing school” (Beaufort 2007: 144). What they need to grasp in her opinion, however, is the real social context for writing in their disciplines and later professions (see also Kruse 2003: 27 f.). For this reason, she pleads for creating an awareness of the discourse communities students are expected to write for, which can ideally be achieved in writing-intensive seminars in the disciplines, but, to her mind, also in general writing courses if, in them, too, awareness of discourse communities and their conventions is 14 On the history of WAC, see Craig (2013: 13 f.). 8 Best-practice approaches to writing instruction 193 raised. She thus contradicts those in the field of writing instruction (Petraglia 1995) who hold that freshman writing courses, i.e. general writing courses, should be abolished altogether (Beaufort 2007: 6 f.): “Freshman writing, if taught with an eye toward transfer of learning and with an explicit acknowledgement of the context of freshman writing itself as a social practice, can set students on a course of life-long learning so that they know how to learn to become better and better writers in a variety of social contexts.” (Beaufort 2007: 7) From the perspective of the cognitive load involved in writing, “writing within the discipline of one’s major field has the added benefit of allowing the writer to free short-term working memory for the task by relying to some extent on long-term working memory” (Kellogg 2008: 15; for details, see Section 8.5.1). 2. Progression in the challenges involved in completing writing tasks There should be a progression both within the courses of a program and over the courses of the program not only with regard to the subject-domain knowledge and competencies to be acquired but also with regard to the challenges involved in the writing tasks that students are assigned (Beaufort 2007: 147, 153). Beaufort founds this requirement on the case study mentioned in which she analysed the college writing of one student, Tim, in a freshman writing course, a first-year history course and an engineering course. In this case study, Beaufort found that there was no clear progression in the writing tasks he had to accomplish in the course of his college training. This, among other reasons, led to ruptures and stagnation in Tim’s writing competence development due to a lack of orientation and guidance. Bräuer & Schindler (2013: 23; my transl.) specify the following requirements for the design of writing assignments that are integrated into the curriculum: “Writing assignments should increase in complexity over a specific period of a degree or training program. Writing arrangements should begin with less complex assignments of a similar type and then move on to more complex tasks which focus on different aspects of the rhetorical problems to solve. Each assignment within a writing arrangement should be motivated in itself so that writers can perceive their relevance for more complex tasks and authentic future problems they may have to solve.” Each course can be combined, for example, with a specific genre or specific genres which students are taught to compose. Popular-science genres may precede more academic genres. General writing courses at the beginning of the program are useful for student groups with heterogeneous language proficiencies. 194 Part IV: Writing Instruction 3. General writing courses focusing on language enhancement and/ or genres of general interest Requirement 1 was a plea for integrating writing instruction into writingintensive discipline-specific seminars. Beaufort (2007: 10 f.) summarizes that “research in composition studies and linguistic anthropology and literacy studies in the last 30 years has shown there is really no viable commodity called ‘general writing skills’ once one gets beyond the level of vocabulary, spelling, grammar and sentence syntax (and some would argue that even at the sentence level, writing is specific to particular discourse communities’ needs).” The fact that many students still have writing problems at the level of vocabulary, spelling, punctuation, grammar, sentence syntax and connecting sentences in both their L1 and often in English, if this is not their mother tongue in an English-medium instruction program (Göpferich 2014), is an argument in favour of general writing courses, in which these more elementary aspects of text production competence can be addressed depending on the specific needs of students which can be determined in placement tests (see requirement 2). 4. Definition of requirement standards for writing assignments and final theses and quality standards for their assessment Requirements that assignments and theses have to fulfil and quality standards that assignments at specific levels of a degree program and final theses should meet need to be developed and agreed upon among the teachers and examiners in a program. Such standards should replace the practice of each teacher having his/ her own system, which leads to confusion among students. Achieving consent on such standards is no easy endeavour. There are some more general criteria for term papers and theses, however, which many teachers might agree upon and which should be communicated to students. Büker (2003: 51) mentions the following: precise description of the research subject (topic, research question, objective), inclusion of all information necessary to enable the reader to verify the research findings, demonstration of the research method, clear distinction of the author’s own ideas from those found in the literature and linguistic accuracy (for further criteria, see Rienecker/ Stray Jörgensen 2003a: 66 ff.). For university teachers who still follow the traditional university pedagogy of apprenticeship and see writing as an art rather than as a craft, it will be difficult to specify their criteria, especially in the face of masses of students (Rienecker/ Stray Jörgensen 2003b: 107). Rienecker & Stray Jörgensen (2003b: 111) therefore recommend “that writing in the continental tradition 15 - if necessary or desirable at all - should not take place at least until later stages of study, when some sort of apprenticeship relation between teacher and student is a realistic possibility”. 15 For information on the difference between the continental traditon and the Anglo- American tradition, see Section 9.1. 8 Best-practice approaches to writing instruction 195 8.5 Approaches to writing instruction at the micro-level As Schindler & Zepter (2013: 205) state, two seemingly opposing tendencies can be observed in university literacy pedagogy in recent years. On the one hand, there are reflective approaches, which comprise exercises for creative writing and complex writing arrangements. 16 On the other hand, we find competence-oriented writing instruction, which focuses on, develops and assesses specific components of writing competence. 17 Whereas reflective approaches are always process-oriented, competence-oriented approaches may be both processand product-oriented. The two approaches address different functions of writing: Reflective approaches conceive of writing as an instrument of reflection, learning and expressing oneself. In competence-oriented approaches, writing is primarily a means of communication and documentation following conventionalized domain-specific patterns, which have to be internalized by the learner (Schindler/ Zepter 2013: 205). What learners need is a combination of both. Thus, the two approaches should be considered as complementary rather than as opposing. For reasons of complexity reduction, complex writing tasks can be subdivided into smaller subtasks thereby separating knowledge generation processes (reflective components) from processes of genreand audience-specific adaptation (competence-oriented approaches; cf. also Girgensohn/ Sennewald 2012: 31). Writing intervention studies have shown that students’ writing development can be enhanced by a variety of practices. The largest meta-analysis of experimental and quasi-experimental studies on writing intervention to date that I know of was conducted by Graham & Perin (2007). Although the participants in these studies were adolescents from grades 4 to 12 in the United States, it can at least be assumed that some of their findings can be transferred to university students in general. From their findings, Graham & Perin (2007) derived 10 recommendations for writing instruction. Six out of these 10 recommendations will be presented in the following in the order of their effect sizes; 16 This tendency in writing pedagogy, termed expressionist rhetoric , dates back to the 1960s (cf. Bräuer 1996: 108 ff.). Students were to generate their own ideas and gain insights through writing before being confronted with the rigid structures and conventions of academic genres. Before being able to express their ideas for others in a comprehensible manner , they had to be given the chance to find out for themselves what they actually thought and wanted to say (cf. Elbow 1998; Girgensohn/ Sennewald 2012: 79 f.). 17 Schindler & Zepter (2013) designate these two approaches „prozessorientiert“ (processoriented) and „kompetenzorientiert” (competence-oriented). Their term „prozessorientiert” is misleading because it does not refer to the opposite of product-oriented but rather to approaches which induce students to reflect on the topics to be written about and the processes involved in completing the writing assignment. Therefore, I call them reflective approaches . The approaches that Schindler & Zepter term „prozessorientiert” and „kompetenzorientiert” can both be process-oriented in the sense that a complex writing process is split up into subtasks. 196 Part IV: Writing Instruction the other four with smaller effect sizes will not be discussed because Graham & Perin (2007: 467) cautioned themselves that the average weighted effect sizes for these other treatments had to be interpreted with care because there was considerable variation in the comparison conditions (Graham/ Perin 2007: 466 f.). The six most recommendable teaching strategies were the following: 1. The teaching of strategies for planning, revising and editing obtained the largest mean weighted effect size (0.70-1.02). 18 Within this category of teaching methods, the S elf- R egulated S trategy D evelopment (SRSD) model developed by Harris & Graham (1996) proved particularly useful. Its objective is to intensify students’ self-regulation in writing. Pritchard & Honeycutt (2007: 34) compare self-regulation in writing to metacognition in reading: Self-regulation in writing involves monitoring one’s comprehension when writing as well as applying specific strategies to complete an assignment. Following the SRSD model, students are taught to set goals; self-record their progress; use mnemonic strategies specific to the task, such as “SCAN each sentence” (Does it make S ense? Is it C onnected? Can I A dd more? N ote errors! ; Graham/ MacArthur 1988); utilize selfinstructions for strategy induction; and self-evaluate their progress (Zimmermann/ Risemberg 1997: 93). 2. The teaching of strategies and procedures for summarizing reading material turned out to be another effective measure (mean weighted effect size: 0.82). 3. Instructional arrangements in which adolescents (peers) collaborate to plan, draft, revise and edit their texts also achieved a high mean weighted effect size (0.75). 4. Setting specific goals, such as identifying the purpose of an assignment (e.g., to persuade) and characterizing the final product (e.g., address both pros and cons), is another recommendable strategy (mean weighted effect size: 0.70). 5. The use of word processors had a positive impact on the quality of writing (mean weighted effect size: 0.55). 6. Exercises in which simpler sentences had to be combined to form more complex ones also led to an increase in text quality (mean weighted effect size: 0.50). In addition to these recommendations, Graham, MacArthur & Fitzgerald (2007: 6) recommend creating a supportive writing environment in which a process-oriented approach is valued, involving students in research activities focusing on the writing process, and providing good models of writing to in- 18 With reference to Lipsey & Wilson (2001), Graham & Perin (2007: 459) state that “a widely used rule of thumb is that an effect size of 0.20 is small, 0.50 is medium, and 0.80 is large”. 8 Best-practice approaches to writing instruction 197 crease students’ knowledge. Furthermore, a study of master teachers reveals that they employ a variety of procedures to enhance students’ motivation. One of the most powerful motivators for learning to write better is engaging students in the completion of writing assignments which create in them the urge to solve problems which are meaningful to them (cf. Bazerman 1997; Beaufort 2007: 206). In addition, as Graves (2004: 447 f.) states, we know at least the following about motivation and engagement: “We know that students must develop positive attitudes about themselves as learners, about their ability to succeed in school, and about the instructional goals that they, their teachers, and their schools set. We know that success is crucial, that if students are going to learn effectively they need to succeed at the vast majority of learning tasks they undertake. We know that students can attribute their successes and failures to factors that are beyond their control, such as ability or luck, or to factors under their control, such as effort and perseverance; and that we want to do everything possible to help students attribute their success or failure to factors that are under their control. We know that students who repeatedly fail are apt to fall into a passive failure syndrome (Johnston & Winograd, 1985), a syndrome in which they repeatedly attribute their failure to forces beyond their control and, as a consequence, stop trying.” Summarizing some important findings from research on how to motivate students, Graves (2004: 448) recommends the following: “- Demonstrate deep concern for students. - Do everything possible to ensure students’ success. - Scaffold students’ learning. - Present appropriate challenges. - Support risk taking and help students realize that failure will sometimes occur. - Encourage students to attribute their success to their efforts, and realize that additional effort can help avoid failures. - Encourage cooperative learning and discourage competition. - Favor depth of coverage over breadth of coverage. - Communicate to students that many academic tasks require and deserve intense attention and effort.” 8.5.1 Process-oriented approaches Stephen North considers the primary objective of writing centres and writing counsellors “to produce better writers, not better writing” (1995: 69). This documents his emphasis on a process-oriented approach but at the same time neglects the quality of the products, which are of equal importance. That writers have become better must show in their writing, otherwise it is worthless. Therefore, writing centres, and writing instruction in general, have the job to produce better writing through better writers. In process-oriented approaches, the focus of writing instruction is shifted from the final product to all phases of the composing process and their intermediate 198 Part IV: Writing Instruction results. They emphasize feedback that is important for improving selfregulation (Zimmermann/ Risemberg 1997: 77). Kellogg (2008: 14) describes the “most important constraint on developing from knowledge-telling to knowledge-transforming, and possibly then on to knowledge-crafting” to be “the limitations of the central executive component of working memory” and asks the question: “In what ways can the educational process aid the functioning of working memory in the service of writing skills? ” The didactical approaches that he suggests can be described by the following keywords: • complexity reduction and • automatization of subprocesses through deliberate practice. Both measures, which will be explained in more detail below, will lead to cognitive relief and thus allow writers to focus on those subtasks of the writing process in which their competencies still need to be developed, as well as their coordination. The burden that the execution of tasks places on working memory depends on a number of factors, which have to be taken into account when devising assignments for writing instruction. Kellogg (2008: 14 f.) discusses the following ones: 1. The maturation of the executive component of working memory The executive component of working memory is supported by the frontal lobe regions of the brain, which mature slowly. As a consequence, children and younger teenagers cannot be expected to achieve advanced stages of writing competence due to the fact that their brains are not yet fully developed. Kellogg (2008: 15) states: “High-resolution structural magnetic resonance images reveal a high degree of frontal development in young adults, 23-30 years of age, compared with 12-16-year olds (Sowell, Thompson, Hilmes, Jernigan, & Toga, 1999). These regions quite possibly are needed (1) to maintain simultaneous representations of the author’s ideas, the text written, and the perspective of the imagined reader and (2) to coordinate interactions among planning, generating, and reviewing. The slow maturation of the central executive component of working memory stresses the absolute necessity of reducing burden placed on it by writing processes.” 2. Domain-specific knowledge stored in long-term memory Domain-specific knowledge for the writing task that the writer already has can be accessed more easily. Ericsson & Kintsch (1995) called this more easily accessible knowledge “long-term working memory”. Accessing knowledge from there is cognitively less laborious than having to maintain it in an active state in short-term working memory and thus frees short-term working memory capacity for other tasks (Kellogg 2001; 2008: 15). For writing instruction, this implies, for example, that having students write about topics they are well 8 Best-practice approaches to writing instruction 199 familiar with and/ or asking them to write texts of genres for which they have stored typical macro-structures in their long-term memories leads to cognitive relief in these areas and allows them to focus on other subprocesses and possibly also their coordination. Familiarity with the topic must not lead to boredom, however, from which a lack of motivation may ensue. 3. The degree of automatization of subprocesses By splitting up complex writing tasks into individual steps, such as planning, generating and revising, these individual steps can be trained intensively and thus automatized to some degree so that when they have to be used in more complex tasks later on they will bind less working-memory capacity (Kellogg 2008: 16). From factor 1, it follows that the complexity and type of writing assignments must be adapted to the age group for which they are devised. This factor can be neglected in university writing instruction, which addresses students whose brains are already fully developed. Commenting on factors 2 and 3 above, Kellogg (2008: 17) states that deliberate practice is what is needed to achieve both increased automaticity through the conversion of declarative knowledge into procedural knowledge (Anderson 1983) and the possibility of knowledge retrieval from long-term working memory as opposed to computation in working memory (Logan 1988). According to Kellogg (2008: 17 ff.), the deliberate practice required may either take the form of learning by doing (practice makes perfect), or learning by observing a mentor (cognitive apprenticeship) or a combination of both. Practice, however, only makes perfect, if feedback is provided since otherwise there is the risk that inferior behaviour gets automatized. This feedback can also be obtained from collaborative work on assignments (collaborative learning). Another option, that Kellogg does not mention, is learning by observing somebody else doing the task who need not be an expert (Braaksma et al. 2004). In all these options, learning by reflecting (metacognition) plays a role. These options with their advantages and disadvantages will be discussed in the following subsections. 200 Part IV: Writing Instruction 8.5.1.1 Learning by doing For learning by doing, exercises are required that involve deliberate practice. Such exercises should fulfil the following requirements (cf. Kellogg 2008: 17): 1. Their execution should involve effort but be within the students’ “zone of proximal development” (Vygotsky 1978). Exercises which are perceived as too easy, i.e., do not involve effort, may lead to a loss of motivation. Only if they involve effort, will they lead to an expansion of the level of skills that students already have. In Vygotsky’s (1978) terms, the exercises should be within the students’ “zone of proximal development”, i.e., they should demand slightly more than what the learners can already do. Having students work on assignments which are challenging to them and on which they are provided with individually tailored profound feedback will be more beneficial for their competence development than having them do large numbers of assignments which are perceived as easy, on which they get no or only superficial feedback and that they may find boring (see item 2). 2. They should be motivating. Ideally, assignments should be related to the discipline students are studying, their special interests and/ or the field of professional activity they are qualifying for. Students will perceive assignments that fulfil these requirements as relevant for their future lives, which increases their motivation to give their best when working on them (cf. Johnstone/ Ashbaugh/ Warfield 2002). In contrast, assignments that are perceived as irrelevant or boring by students may lead to results that do not reflect their actual competence level but rather their lack of motivation. If this is the case, working on them will be less beneficial for students’ cognitive development than working on assignments which motivate them to give their best. Furthermore, feedback on the completion of unmotivating assignments may be less useful for the receiver and a waste of effort for the giver because it may address weaknesses that students are already aware of but did not avoid due to a lack of motivation, instead of a lack of competence, and thus reduced effort invested in task completion. 3. Feedback should be provided on them. Feedback is a prerequisite for learning and future error avoidance. Without feedback, weaknesses may go unnoticed and thus be perpetuated by routinization. Feedback is an important component of collaborative learning, which will be discussed in Sections 8.6.1.5 and 8.8. 8 Best-practice approaches to writing instruction 201 4. Exercises should be practiced repeatedly. Repetition is a requirement for the development of routinized and finally automatized behaviour. Automatization of subprocesses is again a requirement to be fulfilled for the release of working-memory capacity for the execution of more complex processes and for taking into account more factors in problem solving and decision making. Whenever tasks are complex and students are still unfamiliar with a large number of their subtasks, complexity reduction may be a didactically useful strategy. Complexity reduction refers to the subdivision of a complex writing task into smaller subtasks and allowing students to focus on one of these subtasks at a time (Göpferich 2002: 264 ff.; cf. Kellogg 2008: 15 f.). Ideally, students obtain feedback on the results of each of these subtasks and may revise them, if necessary several times, before moving on to the next one. The smaller subtasks may be steps in a chain of actions that have to be performed to solve a complex writing task, such as composing a term paper, or reflective texts on certain of these steps. The chain of actions that lead to a term paper includes many obstacles over which students may stumble. If serious flaws occur at a certain point in this chain of actions, motivation may suffer and subsequent steps may be flawed as well thus reducing the effectiveness of the learning process, leading to the routinization of wrong practices and to valuable time of the students to be wasted. The following example illustrates how this can be avoided. Instead of having students write entire term papers on their own at the end of the term without getting interim feedback, the composition of a term paper can be prepared continuously in writing-intensive seminars by splitting it up into smaller assignments such as the following: 1. Find a question that is related to the topic(s) that we have covered in the seminar, that you would like to see answered and that you feel you could answer by you using methods whose application you have learned so far. 2. In a brainstorming session, write down aspects relevant to answering this question. 3. Transform the result of your brainstorming session into a mind map. 4. Systematically search for the literature related to your research question, select the three most important references and document your research question and the relevant titles you have found in a bibliography summarizing the contents of the three most important titles from the perspective of your research question in 200 words each. 5. Sketch out a hierarchical structure of your term paper. 202 Part IV: Writing Instruction 6. Write an outline of your term paper including your research question, a literature review, the methods of data collection and analysis you will use (if applicable), and your hypotheses (if applicable). Confine yourself to three pages. 7. Write a first draft of your term paper and get feedback on it (either by fellow students, peer tutors or your supervisor). 8. Write a revised draft of your term paper and get feedback on it again. 9. Write the final draft of your term paper. Ideally, students get peer and/ or teacher feedback on each step. This approach has the following advantages: Students get a feeling of the relevance of each step in the composition process and learn that writing involves more than writing down sentences in bursts. Procedural errors can be corrected at an early point in the composition process. This gives students the feeling of being on the right track throughout their composition process, increases motivation and helps them to avoid wasting time and to automatize flawed procedures. Being able to focus on one step at a time leads to cognitive relief and allows students to focus on the task at hand thus leading to more profound cognitive processing in connection with the task and ideally to the automatization of these processes (cf. Kellogg 2008: 16). It helps students to avoid procrastination. What evidence is there that the method outlined above is effective? Kellogg (1990) found that preparing a macrostructure prior to writing the first draft leads to relatively more time being devoted to sentence generation and cohesive links. Kellogg also found, however, that “the benefits of outlining were substantially reduced when writers had already developed their thinking about the specific topic, knowledge that could be retrieved from long-term working memory rather than computed and stored in working memory during composition” (Kellogg 2008: 16; Kellogg 1990). As to reviewing processes, Chanquoy (2001) reported beneficial effects of delayed reviewing. Kellogg (2008: 16) hypothesizes that “ [ t ] he time delay could facilitate the construction of a reader representation that accurately captures what the text literally says as the students re-read what they had written earlier.” Whereas the above procedure helps students in the organization of the writing process and the selection and structuring of their ideas, complexity reduction can also be achieved on a linguistic level leading to increased linguistic fluency in the text generation phase and thus again a reduction of cognitive load. Complexity reduction on a linguistic level can be achieved by transformation exercises, such as turning passive-voice into active-voice constructions, transforming direct speech into indirect speech or optimizing passages with regard to their topic-comment structures. 8 Best-practice approaches to writing instruction 203 In summary, measures of complexity reduction ideally lead to increased attention to the processes to be practiced and thus contribute to their automatization, not necessarily their complete automatization but at least reduced effort in their execution (Kellogg 2008: 16 f.; cf. McCutchen 1988). 8.5.1.2 Learning by observing Cognitive relief cannot only be achieved by subdividing complex tasks into smaller subtasks and having students practice them step by step on a learningby-doing plus feedback basis as described in Section 8.5.1.1, it can also be achieved by means of “observational learning”, i.e., learning by observing others who execute a task (Braaksma et al. 2004; 2006). The cognitive relief involved in observational learning is assumed to allow the observer to focus on reflective activities concerning the writing process rather than on writing itself; in other words, cognitive effort shifts from executing writing tasks to learning about their execution (Braaksma et al. 2006: 35; Couzijn 1999; Rijlaarsdam/ Couzijn 2000a, 2000b). The learning activities involved in observational learning are observation, analysis, evaluation and reflection (Braaksma et al. 2006: 33). By these self-regulatory and metacognitive activities, knowledge about writing is assumed to be constructed (cf. Graham/ Harris 1994). A requirement for this type of self-regulated learning is, however, “that learners have reached developmental levels sufficient to be able to step back and consider their own cognitive processes as objects of thought and reflection” (Brown 1987: 68; cf. Braaksma et al. 2006: 35). Observational learning was found to influence writing processes in a manner different from learning by doing and to lead to better writing products. The differences between learning to write by writing and learning to write by observing others write found on the process level in subsequent writing tasks are summarized as follows by Braaksma et al. (2006: 35 f.; cf. also Braaksma et al. 2004): 1. Observational learning led to “relatively more metacognitive activities (goal-orientation and re-reading) at the start and relatively more executional activities (writing and re-reading) in the second part of the writing process” (Braaksma et al. 2006: 35) than learning-by-doing. 2. Observational learners showed more planning activities than those who learned by doing. 3. In the middle and at the end of the writing process, writers who learned by observation were involved in more meta-analyzing activities indicating monitoring and regulating processes than writers who learned by doing. 4. Some of the activities of those who learned by observing changed over time whereas those who learned by doing showed a more homogeneous process throughout. 204 Part IV: Writing Instruction 5. Among the participants in the observational learning condition more variances occurred than among those in the control condition, indicating more heterogeneous processes. The finding that observational learning also leads to better writing products is supported by a number of studies (e.g., Couzijn 1995, 1999; Zimmermann/ Kitsantas 2002; Braaksma et al. 2006). Zimmermann & Kitsantas (2002) investigated whether the type of model observed has an effect on the learning outcome. In their study, they used three different models: no model, a mastery model and a coping model. In the ‘no model’-condition, students were simply required to study nine problems on an overhead projector. In the ‘mastery model’-condition, the participants observed an experimenter solving the same nine problems without errors on the overhead projector. In the ‘coping model’-condition, the participants observed the experimenter solving the nine problems making errors in the beginning and then gradually reducing them. After the observation phase, the participants had to solve 12 similar problems individually. Half of the participants received feedback after each step, the other half did not. Feedback was limited to the steps and strategies they performed properly. Observing a model was found to lead to better results than having no model. The students in the coping-model condition outperformed the students in the mastery-model condition. Feedback was found to have a significant positive effect as well. The implications for writing pedagogy that this finding has are that confronting students with writing processes that contain typical problem-solving processes can be assumed to be more effective than confronting them with more fluent writing processes, and that feedback is important for competence development. Braaksma, Rijlaarsdam & Van den Bergh (2002) found that observing poorer models is not necessarily more rewarding for the observers than observing better models. What type of model to observe best rather depends on the observer’s competence level. They found that “weak students (low aptitude score) learned more when focusing their observations on weak models, whereas good students (high aptitude score) profited more from focusing on good models. For students with a medium aptitude score no differences were found” (Braaksma/ Rijlaarsdam/ Van den Bergh 2006: 37). In a study in which they tried to pinpoint the mental activities responsible for better writing products after observational learning, Braaksma et al. (2001) found that “evaluation and elaboration activities are important for the effectiveness of observational learning. Students who correctly evaluated peer models and commented correctly on the products the models produced were better in writing argumentative texts.” (Braaksma/ Rijlaarsdam/ Van den Bergh 2006: 36) Drawing on these findings, Braaksma, Rijlaarsdam & Van den Bergh (2006) conducted a case study in which the learning activities of six students when processing observation tasks were analyzed in detail. To get access to their 8 Best-practice approaches to writing instruction 205 learning activities, the students were required to think aloud during the observation tasks. The observers were found to be strongly engaged in metacognitive activities: “They observed the models’ writing, identified and conceptualized the writing strategies, evaluated the performance of the models and reflected explicitly on the observed performances. The performance of these activities suggests that observers internalized, applied, and developed criteria for effective writing. Furthermore, observers also performed (mental) executional processes and during the observations they compared their own (covert) performance with the models’ actual performances.” (Braaksma/ Rijlaarsdam/ Van den Bergh 2006: 53) The study also corroborated the earlier finding (Braaksma et al. 2001) that evaluation and elaboration are activities during the observation process that are beneficial to learning. 8.5.1.3 Learning by cognitive apprenticeship A type of learning that combines a special variant of observational learning with learning by doing is cognitive apprenticeship. The learning-by-observing component of cognitive apprenticeship is characterized by the fact that the learner observes an experienced mentor or at least a mentor who is more experienced than the learner. In the learning-by-doing component, the mentor involves the learner in guided participation in a task and helps the learner to work through the task, for example, by focusing his attention on a manageable subgoal. In this manner, again a reduction of task complexity is achieved. The task that the learners are expected to perform should again be in their “zone of proximal development” (Vygotsky 1978). With reference to Schunk & Zimmermann (1997), Kellogg (2008: 19) asserts that deliberate practice and cognitive apprenticeship can readily be integrated in writer training. Such an integration takes place in Schunk & Zimmerman’s (1997) four-step training with the explicit goal of self-regulation. The four steps are (cf. Kellogg 2008: 19 f.): 1. The leaner observes the model’s action. 2. The learner tries to emulate the model’s behaviour. 3. The learner deliberately practices to gradually reduce the cognitive demands involved in the processes underlying performance and thus frees executive attention to control cognitive processes. 4. With additional practice, the learner is able to adapt performance to changes in internal and external conditions and thus to achieve the selfregulation characteristic of expert performance. With regard to cognitive apprenticeship, it is worth noting that most of the recent world chess champions were at one time tutored by chess masters (Ericsson/ Krampe/ Tesch-Römer 1993; Ericsson/ Charness 1997: 22). It is also 206 Part IV: Writing Instruction interesting to note how Benjamin Franklin, in his autobiography, describes how he tried to learn to write in a clear and logical manner. He states that he would read through a passage in a good book with the goal of understanding it rather than memorizing it. He would then try to reproduce its structure and content. After that, he would compare his reproduction with the original and spot differences. Franklin argued that the repeated application of this cycle of study, reproduction and comparing his text to a well-structured original fostered his ability to organize his thoughts for speaking and writing (Ericsson/ Charness 1997: 21). Translation in the functionalist sense from a well-known L2 (in most cases English) into the L1 might be a good exercise for improving one’s writing skills as well. It may even be better than Franklin’s method because translation does not allow the translator to use the same lexical and sometimes also grammatico-syntactical inventory as the source text, which Franklin’s method does, and thus forces the translator into deeper processing, which can be assumed to have a positive effect on competence development because it intensifies deliberate practice and fosters flexibility. The effectiveness of cognitive apprenticeship is supported by results documented in Hillocks (1986; 1995). Situations for cognitive apprenticeship can also be created by peer tutors, who provide extracurricular writing support, and writing fellows, who provide writing support in subject-matter courses. 8.5.1.4 Learning by reflecting A form of learning that is fostered by learning by observing and cognitive apprenticeship is learning by reflecting. As described in the previous sections, one of the effects of observing instead of performing oneself is that cognitive capacity, normally needed for executing a task, is liberated in the observer and can be used for reflecting on what is being observed. Thus, learning by observing automatically becomes a form of learning by reflecting. Reflection is also a central component of writing diaries, working journals and portfolios. Diaries and working journals can be kept to document ideas, to develop drafts and strategies, to comment on them and to evaluate them, in other words, to take notes of everything that comes to one’s mind and may be relevant for a writing assignment to be completed. By putting these ideas down, the writer makes sure that they will not be forgotten (Bräuer 2000: 16 ff.). A portfolio is “a purposeful collection of student work that exhibits the student’s efforts, progress and achievements in one or more areas. The collection must include student participation in selecting contents, the criteria for selection, the criteria for judging merit, and evidence of student self-reflection” (Paulson/ Paulson/ Mayer 1991: 60). From this follows that portfolios do not only comprise finished products, i.e., texts on whose quality product-oriented approaches to writing instruction 8 Best-practice approaches to writing instruction 207 focus, they also illustrate the paths that led to these products (i.e. processes) and contain self-assessments of the owner’s own achievements (cf. Ballweg et al. 2013: 189). Ideally, portfolios are not written in isolation but result from a dialogue with persons who give feedback. Such persons may be either peers or teachers. Feedback receiver and feedback giver engage in what Bruffee (1984) termed “collaborative learning” (see Section 8.5.1.5). In tertiary education, collaborative learning constellations may take the following forms: a) peer - peer in curricular activities, b) peer - peer in extracurricular activities (e.g. writing groups) c) peer - trained peer (writing fellow) in curricular activities, d) peer - trained peer in extracurricular activities (e.g. writing tutoring) and e) student - teacher In all five constellations, portfolios may serve an important reflective function. For those who receive training in providing writing support (trained peers and teachers who offer writing-intensive seminars), portfolios can be documents in which they reflect on their conceptions as teachers. For those who learn writing, portfolios may be documents in which they create a self-awareness of their progress. By exchanging their portfolios or parts of their portfolios with their students and vice versa, teachers may get feedback on whether their approaches to teaching are effective. Students may learn from their teachers what they have mastered so far and what still needs to be improved. 8.5.1.5 Collaborative learning Collaborative learning is a concept that has been widely discussed in the USA since the 1980s. Its origins, however, date back to the 1950s in Great Britain (Bruffee 1984: 636). Typical forms of collaborative learning are peer tutoring, peer criticism (or peer evaluation) and classroom group work. As the examples illustrate, collaborative learning may take the form of extracurricular activities but can also be integrated into classroom work, for example, when students learn to provide feedback on other students’ essays and the teacher evaluates both the essays themselves and the critical responses written on them by peers (Bruffee 1984: 638). What all forms of collaborative learning have in common is that they do not primarily (but also) change what students learn but the social context in which they learn it (Bruffee 1984: 638). Bruffee (1984: 638) describes its effect as follows: “Students’ work tended to improve when they got help from peers: peers offering help, furthermore, learned from the students they helped and from the activity of helping itself. Collaborative learning, it seemed, harnessed the powerful educative force of peer influence that had been—and largely still is—ignored and hence wasted by traditional forms of education.” What is the rationale of collaborative learning? For Bruffee (1984: 639), reflective thought is “conversation that takes place within us”. He asserts that “ [ w ] e 208 Part IV: Writing Instruction can think because we can talk, and we think in ways we have learned to talk” (Bruffee 1984: 640). As a consequence, “ [ t ] o think well as individuals we must learn to think well collectively—that is, we must learn to converse well. The first steps to learning to think better, therefore, are learning to converse better and learning to establish and maintain the sorts of social context, the sort of community life, that foster the sorts of conversation members of the community value.” (Bruffee 1984: 640) Bruffee (1984: 641) goes on to conclude that “ [ i ] f thought is internalized conversation, then writing is internalized conversation externalized”. For the writing teacher, this has the following implications: “The inference writing teachers should make from this line of reasoning is that our task must involve engaging students in conversation among themselves at as many points in both the writing and the reading process as possible, and that we should contrive to ensure that students’ conversation about what they read and write is similar in as many ways as possible to the way we would like them eventually to read and write. The way they talk with each other determines the way they will think and the way they will write.” (Bruffee 1984: 642) Since academic literacy involves learning the conventions of a knowledge community of which students are not yet members of, the question arises that Bruffee (1984: 644) rhetorically asks himself: “ [ H ] ow can student peers, who are not members of the knowledge communities they hope to enter, who lack the knowledge that constitutes those communities, help other students enter them? ” or, put differently, “Isn’t collaborative learning the blind leading the blind? ” (Bruffee 1984: 646). Bruffee (1984: 644) answers this question as follows: “Students are especially likely to be able to master that discourse collaboratively if their conversation is structured indirectly by the task or problem that a member of that new community (the teacher) has judiciously designed. To the conversation between peer tutors and their tutees in writing, for example, the tutee brings knowledge of the subject to be written about and knowledge of the assignment. The tutor brings sensitivity to the needs and feelings of peers and knowledge of the conventions of discourse and of standard written English [ or the respective language of instruction; SG ] . And the conversation is structured in part by the demands of the teacher’s assignment and in part by the formal conventions of the communities the teacher represents, the conversation of academic discourse and standard English [ or the respective language of instruction; SG ] . Such conversation among students can break down, of course, if any of these elements is not present. It can proceed again if the person responsible for providing the missing element, usually but not always the teacher, is flexible enough to adjust his or her contribution accordingly.” This line of argumentation is supported by another argument: In line with Thomas Kuhn (1970), Bruffee (1984: 646) asserts that knowledge is not something given in an objective manner, “information impressed upon the individ- 8 Best-practice approaches to writing instruction 209 ual mind by some outside source” but “an artifact created by a community of knowledgeable peers constituted by the language of that community”. As a consequence, “learning is a social process and not an individual process” (Bruffee 1984: 646). Bruffee (1984: 646 f.) goes on to argue: “To learn is to work collaboratively to establish and maintain knowledge among a community of knowledgeable peers through the process that Richard Rorty calls ‘socially justifying belief.’ We socially justify belief when we explain to others why one way of understanding how the world hangs together seems to us preferable to other ways of understanding it. We establish knowledge or justify belief collaboratively by challenging other’s biases and presuppositions; by negotiating collectively toward new paradigms of perception, thought, feeling, and expression; and by joining larger, more experienced communities of knowledgeable peers through assenting to those communities’ interests, values, language, and paradigms of perception and thought. If we accept this concept of knowledge and learning even partially and tentatively, it is possible to see collaborative learning as a model of the way that even the most sophisticated scientific knowledge is established and maintained. Knowledge is the product of human beings in a state of continual negotiation or conversation. Education is not a process of assimilating ‘the truth’ but, as Rorty has put it, a process of learning to ‘take a hand in what is going on’ by joining ‘the conversation of mankind.’ Collaborative learning is an arena in which students can negotiate their way into that conversation.” If we accept this, collaborative learning is not just a manner of learning the known but a manner, the manner in which knowledge is generated, changes and grows (Bruffee 1984: 647). It is the method by means of which students are initiated into scientific communities. Whether forms of collaborative learning are accepted by students, however, may depend on the cultures in which they have been socialized (see the research overview in Hu/ Ren 2012). In teacher-centred cultures, students may consider their peers to be insufficiently qualified to criticize their drafts (Paulus 1999). Furthermore, students from collectivist societies may refrain from criticising their fellow students and expressing disagreement for reasons of harmony within the group (Triandis 1995). China is both a teacher-centred and collectivist society so that peer feedback could be assumed to be the wrong pedagogical strategy to be employed for Chinese students. However, in a study in which 116 junior English language majors at a major Chinese university were asked about their experience with peer and teacher feedback and their evaluation of it in a written questionnaire, Hu & Ren (2012) obtained more differentiated results on students’ feedback preferences. Although more than a third of their participants preferred teacher feedback alone, two thirds of their participants valued both teacher and peer feedback. Avoidance of criticism to maintain interpersonal harmony was not a big issue in their study (Hu/ Ren 2012: 83). They also found that “good experience with peer feedback can help students to perceive peer review more positively”, from which they draw the 210 Part IV: Writing Instruction conclusion that “providing opportunities for students to engage in well prepared and carefully implemented peer review tasks can be a useful avenue to pursue in developing a favourable attitude towards this pedagogical activity” (Hu/ Ren 2012: 84). Due to the observed prevalent belief in the superiority of teacher feedback over peer feedback among students, they also recommend raising students’ awareness of the benefits of peer feedback and its potential to complement teacher feedback in a useful manner as well as the fact that useful peer feedback need not be restricted to lower-level problems, as assumed by many of their participants, but that students are also able to address higherlevel issues in their fellows’ writing (Hu/ Ren 2012: 84). 19 8.5.2 Product-oriented approaches Product-oriented approaches in writing instruction typically use models or prototypes, for example, in the form of typical examples of the genres which students are supposed to learn to write. The use of model texts in writing instruction, however, has been discussed controversially (for an overview, see Macbeth 2010). Especially students who have to complete a specific writing task, such as writing an essay, a review or a term paper, for the first time, ask for model texts in order to develop a clearer picture of what is expected of them. For such novices, models are useful. However, the very advantage of models, i.e., the fact that they represent something more complex in simplified form and thus make it more accessible from a pedagogical perspective, at the same time constitutes a disadvantage in the form of “false provisions” (Macbeth 2010) they make. Students tend to overgeneralize what they find in models. To overcome this disadvantage and disclose the false provisions that models make, students need to be familiarized with their “shortcomings” and learn to adapt what they find in models to the various contexts in which and for which they have to compose texts. Overgeneralizing what they find in models, or “imitation” in Steinhoff’s (2007) terms, may help students to produce first drafts for an assignment. These first drafts may be useful steps on the way to good practice, for which processes of “transformation” (Steinhoff 2007) are needed. To make students’ aware of the limitations of models, student conferences and other forms of feedback, which allow students to share the experience of an expert writer, can be useful (cf. Macbeth 2010). 8.6 Writing-intensive seminars and the role of the teacher Writing-intensive seminars are domain-specific seminars in which the acquisition of domain knowledge is combined with the acquisition of the text compe- 19 For a comparison of peer feedback, writing instructor feedback and content instructor feedback, see Patchan, Charney & Schunn (2009). 8 Best-practice approaches to writing instruction 211 tence that is needed to read, write and learn from texts of the domain in question. In such seminars, writing assignments are used as a pedagogical means of getting students actively involved in knowledge construction. The number of such seminars, which are still rare in German universities, should be increased given the paradox situation pointed out by Jetton & Alexander (2004: 15) that just at a time when the linguistic demands of text-based learning are increasing, academic support for students in the form of explicit literacy instruction is diminishing. Writing-intensive seminars are based on the insight of writing research that the acquisition of literacy has not been completed when students enter university but must be fostered throughout their university training and education in order for students to continue developing their literacy in a domainand task-specific manner (Steinhoff 2007; Pohl 2007; Beaufort 2007). For this reason, general writing courses and peer tutoring, though useful for the development of certain aspects of writing competence, cannot replace domain-specific courses in which writing forms a central component (cf. also Banzer/ Kruse 2011: 8). For this type of subjectand literacy-integrated instruction, writing competence models which model writing competence as sets of linguistic competencies are too restricted in scope. Models such as the ones by Beaufort (2007) and Kruse/ Chitez (2012) take into account that the development of literacy needs to go hand in hand with an acculturation into the discourse community in question. Against this background and due to the epistemic or knowledge-constructing function of writing (cf. Galbraith 1999), fostering literacy within the domain is expected to not only lead to an improvement of literacy itself but also to a more profound engagement with the knowledge of the domain and thus to domain-specific learning (NSSE 2013), an improvement of knowledge transfer competence (Robertson/ Taczak/ Yancey 2014) and reflective competence (Hillocks 1995). A prerequisite for literacy development in students’ subject-domain courses is an adequate teacher‘s conception of teaching. This conception must be characterized by “a shift from teaching to learning“ (Kember 1997; Kröber 2010). Instead of considering students’ writing skills as defective, a view that is frequently held in the faculties, teachers must take a developmental perspective. They need to develop teaching and writing arrangements which meet students at the developmental stage they have achieved and support them in the further development of their skills taking into account all knowledge domains in Beaufort’s model, i.e., subject-matter knowledge, writing process knowledge, rhetorical knowledge, genre knowledge and discourse community knowledge (Bean 2001; Gottschalk/ Hjortshoj 2004; Bräuer/ Schindler 2013). Writingintensive teaching has been integrated optimally into domain-specific seminars only when the teacher’s conception of teaching has shifted on Kember’s (1997: 264) continuum from a teacher-/ content-centred approach, also termed “transmission approach“ (Wade/ Moje 2000), to a learner-/ learning-processcentred approach, also designated “transactional” or “participatory approach to subject-matter learning” (Alvermann 2002; Wade/ Moje 2000). Teachers 212 Part IV: Writing Instruction who follow the wide-spread first type of approach convey ‘authorized’ or ‘official’ knowledge to students (Wade/ Moje 2000: 611). In a learning-centred approach, however, students have a much more active role: “Through this approach, teachers facilitate domain learning by providing opportunities for adolescents to engage in classroom structures that promote interaction and increase the likelihood that students will begin to construct the knowledge for themselves. When teachers employ this approach, they provide students with a variety of texts, including linear or traditional texts, nonlinear or hypertexts, and the texts of peer-led discussions to explore the content of a particular domain. By exploring a wide range of texts, students are able to search for important information within the domain, in order to construct a mental model of the concepts (Stahl et al., 1995). 20 In using the participatory approach, teachers also ask students to construct their knowledge of the subject matter by writing a variety of forms that include both their own sociocultural experiences and more sanctioned texts of the classroom (Wade & Moje, 2000).” (Jetton/ Alexander 2004: 29) According to Jetton & Alexander (2004: 29), the two approaches are not dichotomous choices, both can be effectively combined in the classroom (cf. also Moore 1996). They provide the following example of a situation where teacher-focused components are needed: “We also believe that a classroom in which the participatory approach is the sole pedagogy could potentially be just as detrimental to student learning. Sometimes students need the teacher, as a more knowledgeable other, to provide a minilecture on a particular principle within the domain (Vygotsky, 1934/ 1986). In our observations of classroom discussions, we found that when students within a peer-led discussion reached an impasse in their knowledge, the discussion became unproductive and frustrating to them. The teacher would then ask students to pause their discussions, and she would provide a short lecture about the content (Alexander & Jetton, 2003; Jetton, 1994). In this way she could elaborate on an important concept that would make it easier for the students to go back to their peer-led discussion and construct knowledge in a more participatory way.” (Jetton/ Alexander 2004: 30). Teaching writing-intensive seminars requires special qualification and the teachers’ conviction that their teaching can actually contribute to the development of their students’ literacy. They must be able to take over the role of a writing facilitator in a professional manner instead of considering themselves and behaving as “a sage on the stage”. Instead of telling their students how to complete a writing assignment, they rather should provide scaffolding for students to explore their own ways. Research on reflective practice has shown that these objectives, i.e., a shift to a learning-centred approach among teachers combined with a more engaged and reflective attitude among students, can best be achieved by portfolio work (Szczycrba/ Gotzen 2012; Bräuer 2014). 20 Reference not included in Jetton &Alexander (2004: 29). 8 Best-practice approaches to writing instruction 213 Teaching writing-intensive seminars takes a „careful orchestration of content and pedagogical knowledge“ (Jetton/ Alexander 2004: 16, 25). 8.7 Assignments for writing courses and writing-intensive seminars Academic writing courses may have different underlying values and objectives. Beaufort (2012: 2) 21 lists the following ones: expressivist goals (self-expression and finding one’s own voice), critical theory/ cultural studies goals (critique of social hierarchies and cultural hegemonies), democratic/ rhetorical goals (informed participation in civic issues), pragmatic goals (successful written expression in academic and work contexts), aesthetic goals (appreciation of language), and process goals (managing complex writing tasks). If we want to equip our students with competences that they can transfer, writing courses with pragmatic objectives and/ or process goals are the ones that are most relevant to the widest range of students, whereas writing courses with, for example, expressivist and aesthetic functions are relevant only for students in a limited number of programs such as literary studies. Nevertheless, these types of course have been dominant in university writing instruction. Bergmann & Zepernick (2007: 129) found that students considered writing in English classes “personal and expressive rather than academic and professional” and therefore questioned the relevance of what they learned in these classes to other disciplinary contexts. In accordance with Beaufort (2012: 3), I therefore plead for writing courses with pragmatic objectives in the sense specified above. What are useful writing assignments for writing courses with a pragmatic objective? In answering this question, at least two aspects need to be addressed: topic and genre. In writing-intensive seminars, all writing assignments should be related to the topic of the seminar, which is always a subject-matter seminar. Writing courses, too, should have a single topic or a limited number of topics that are relevant to their student audience. A single topic has the advantage that students can acquire more profound knowledge on this topic and also compose more profound texts on it thus acquiring skills that can be transferred much better in academic and professional contexts than simply writing superficial texts on various topics (Beaufort 2012: 4 f.). Genres that are particularly useful in writing courses and writing-intensive seminars, at least in the humanities, are rhetorical analyses of nonfiction (or pragmatic texts) and literature reviews of a body of research that addresses a 21 The online document has no page numbers. The page numbers given refer to the pages of the printout. 214 Part IV: Writing Instruction question relevant to the students and provides contradictory results. The literature review can also be replaced by the comparison of different articles on a controversial issue, a task that can also be assigned to prepare students for more encompassing literature reviews. Why are these assignments useful? The literature review trains students in how to analyze texts from a rhetorical perspective and thus teaches them critical reading skills, which can be transferred into all areas of life and work. Literature reviews are used in most academic disciplines both as autonomous texts and parts of research reports. By composing literature reviews, students learn how to summarize and synthesize a number of texts, how to create intertextuality and also to understand the rhetorical moves that characterize this genre (Beaufort 2012: 8). These genres thus have the potential of good antecedent genres (Devitt 2008) on which transfer for future writing tasks can be based. They allow students to acquire and practice the entire range of skills that are also needed for longer academic texts, such as research articles and theses, by composing relatively short texts. In this chapter, a number of additional and more specific criteria for effective writing assignments have been mentioned already, some of which occur in Langer & Applebee’s (1987: 141 ff.) list of criteria for effective writing tasks and in T. Shanahan’s (2004: 66 ff.) “guidelines for using writing well to stimulate student learning”. The following list represents a synthesis of those guidelines for effective writing that are based on empirical research: 1. Students should be given frequent opportunities to engage in writing for the following reason: “Writing to learn is a skill, and like any skill, it is best developed through practice. Writing should be a regular part of activities in all content-area classes, so that students’ experience with writing is sufficient to allow them to understand how to use writing effectively to accomplish learning.” (T. Shanahan 2004: 67) 2. The texts to be composed must have a clear purpose that is relevant for students. Teachers should explain what they expect of their students, for example, by means of an adequate amount of scaffolding. 3. Assignments should be within the zone of the students’ proximal development. 4. Writing assignments should allow students room to have something to say of their own and to take a perspective. Requiring students to take a perspective or position does not allow them to simply reproduce a selection of information from a text, but induces them to translate what they find in a text for a specific purpose and thus to gain a deeper understanding of the text. It induces them to connect ideas to personal experience, to evaluate, compare and criticize (T. Shanahan 2004: 68). 5. Students should be given opportunities to engage in extended writing. Studies that compared extended essay writing with briefer forms of writing 8 Best-practice approaches to writing instruction 215 showed that extended writing is more effective for learning although students also benefit from shorter forms such as writing summaries, notes or answers of the length of a paragraph (T. Shanahan 2004: 67). “Extended writing creates the ground for extended learning. Such writing allows for a wider choice of approaches (personal, analytical, critical), and it requires a more thorough coordination of information from the source text, presentation, or discussion.” (T. Shanahan 2004: 67). 6. Writing is one way of encouraging deep thinking and may yield even more profound results when combined with other activities such as reading, speaking, listening and viewing: “Having students write and rewrite about some aspect of a concept is valuable. Having them do this in the context of reading, interacting with peers, hearing presentations, trying to apply the information in a lab, viewing a videotape, comparing two articles, and so on, is much more powerful.” (T. Shanahan 2004: 69) Therefore, students should work collaboratively and writing should be made one of many ways of thinking about a subject. Externalizing their ideas first in speaking with somebody may be easier for students than writing them down directly. Collaborative writing tasks can support this (T. Shanahan 2004: 70). 7. With students’ progress, assignments must change. 8. Writing arrangements should include revision loops. Revision in the sense of not primarily correcting linguistic errors but of revisiting and rethinking the original composition and line of argumentation leads to refining one’s thinking. For this purpose, feedback is important (T. Shanahan 2004: 68). 9. Writing assignments should require analysis and synthesis. Writing assignments can engage students in analysis, for example, by requiring them to compare certain concepts or characters presented in a text. Syntheses are obtained by requiring students to combine information from more than one text or source (T. Shanahan 2004: 71). 10. The texts written should be evaluated. With reference to a professional development project that Howard (1983) conducted with 32 high-school teachers, in which they focused on writing to learn, Shanahan describes how grading texts can be both beneficial to the teachers and the students. The teachers can find out how much the students actually understood and thus get feedback on the effectiveness of their pedagogical approach. By providing students with the criteria of evaluation and examples of texts that fully meet the requirements and of texts that only meet them in certain respects, students learn in what respects they can improve (T. Shanahan 2004: 72). Writing is generally assumed to play a minor role in the natural sciences and writing in science courses has not been the focus of literacy research (C. Sha- 216 Part IV: Writing Instruction nahan 2004: 88). The few studies conducted, however, show that even in the sciences, reading and writing have a positive effect on learning. For example, Hand (1999), Prain & Hand (1999) and Hand & Prain (2002) have found that “students benefit from instruction that involves them in writing for a variety of purposes to a variety of audiences, and in a variety of genres” (C. Shanahan 2004: 88). Against this background, C. Shanahan (2004: 89) complains that “ [ p ] robably one of the biggest drawbacks to traditional science instruction is that students are not taught to write.” According to Yore, Hand & Prain (1999), writing instruction in the sciences should focus on writing to learn. For this purpose, they suggest two instructional practices: a) reaction papers and b) collaborative essays. Reaction papers involve reading and writing, e.g. reading an article and writing a summary or reflection on it. To prepare them for these tasks, students should be taught summarizing skills, such as deleting redundancies, identifying relevant ideas and synthesizing ideas from different sources in a coherent manner, but also reflecting including evaluation and questioning of ideas presented and methods used (C. Shanahan 2004: 89). Reflection can also be intensified by requiring students to write collaboratively. Collaborative essays, for example, can be written by teams of students on the basis of research they have done. Another team of students can then take over the function of peer reviewers for these essays. Another suggestion C. Shanahan (2004: 89) makes for writing in the sciences is providing students with “skeletons” for writing different genres. Such skeletons illustrate the typical macrostructures of the respective genres and may also include typical formulations (textual routines). Writing curricula which follow these suggestions contribute to building students’ critical thinking skills, rhetorical skills and academic writing skills and foster learning of transfer to other contexts for writing (Beaufort 2012: 9). In parallel, critical reading should be fostered because reading skills are central for recognizing discrepancies between text meaning and author meaning, which leads to revisions (MacArthur 2007). 8.8 Giving feedback Feedback has been mentioned as an essential component of effective processoriented writing instruction. Giving effective and constructive feedback, however, is something that needs to be taught. For this purpose, both teachers and students need criteria, or even better, frameworks, such as the Karlsruhe comprehensibility concept (see Section 4.3), from which they can derive adequate criteria for the assessment of each specific assignment. Targeted strategy instruction, i.e., instruction in which students are provided with assessment rubrics such as the “Six Traits” (ideas and content, organization, voice, word choice, sentence fluency, and conventions) developed by Spandel & Stiggings (1997) can also help in the feedback phase (Pritchard/ Honeycutt 2007: 36 f.). 8 Best-practice approaches to writing instruction 217 Fig. 8-2 shows the feedback rules that I give my students in preparation of their first feedback assignments. Feedback Rules: How to give (written) feedback in a constructive manner and how to react to it Giving feedback: 1. Summarize what you have understood to be the objective, main content and line of argumentation of the text in no more than five to ten lines of text. This will show the person whom you give feedback that you have read the text and whether you have understood it in the sense the author intended. If this deviates from what the author intended to express, it gives the author an idea of how the text can be misunderstood. 2. Read the text first for feedback on the content level and only then for feedback on formal and linguistic aspects, such as orthography, grammar, punctuation, idiomaticity, etc. This always means that you have to read the text at least twice. (For longer texts this is not always possible in practice.) 3. Make I statements and be as precise in your criticism as possible, for example: Avoid: This is incomprehensible. Use instead: I find it difficult to follow you here because I do not understand why you think that … 4. On the content level, as opposed to the formal level, do not correct ‘errors’ or make suggestions for alternative formulations but describe why you feel something needs to be changed. This can frequently be done by asking questions. Taking feedback: 1. Thank the person who gives you feedback assuming that the person does it to help you and not to bring you down. 2. Take into account that giving feedback is hard work. Therefore people usually only provide feedback if they feel that what they provide feedback on is worth it. 3. Read the feedback (or listen to it) and do not defend yourself or start arguing. If we assume that persons who give feedback follow the above rules, they express their personal opinions. You are free to decide whether you accept their suggestions because you find them useful or ignore them because you feel they are not justified. 4. Never forget that you have the responsibility for what you write, not the people who have given feedback on earlier versions of it. Fig. 8-2: Rules for giving and taking feedback Types of feedback that have been found to be particularly effective in L2 writing instruction will be addressed in Section 9.6. Teacher feedback should be formulated from the perspective of a reader and not a judge. It should be formulated in a constructive manner and at the end summarize the most important revisions that should be made for the next draft (Craig 2013: 105 ff.; cf. also Scott/ Coate 2003). Part V L1 vs. L2 Writing 9 Writing in the L1 vs. writing in the L2 1 This chapter addresses the specificities of composing in an L2 as compared with composing in an L1. In most studies conducted to date, the L2 under investigation has been English. English is the lingua franca in an increasing number of disciplines and gains in importance in tertiary education. Even in countries where English is no official language, such as Germany, Englishmedium instruction has started to replace programs taught in the national language(s) (cf. also Björkmann 2013). Against this background, the question arises whether the requirement to write - as well as to comprehend and to think - in English as a foreign or second language leads to semantic and argumentative losses and hampers the epistemic function of writing and thus, ultimately, cognitive development resulting from writing. Another question worth answering in this context is whether better L2 texts arise from formulating one’s ideas in one’s L1 first and then translating them into the L2 or whether translation should be circumvented altogether by writing directly in the L2. These questions will also be addressed in this chapter, although final answers cannot always be provided. When answering these questions, numerous factors have to be taken into account, among them L1 and L2 proficiency levels, L1 writing ability, the type of the writer’s bilinguality and the level of translation competence that the person writing in the L2, or translating into the L2, has acquired. These factors will be considered in the sections to follow. 9.1 Differences between composing in the L1 and the L2 The few existing studies which have compared L1 and L2 text production have indicated that L2 text production processes, aside from the additional lexical and grammatical challenges associated with foreign language production in general, are strikingly similar to L1 text production processes. This suggests that there is a “composing competence” (Krapels 1990) which exists across languages and is at least partially independent of L2 language proficiency (cf., e.g., Cumming 1989; Hirose/ Sasaki 1994: 216 ff.; Sasaki 2000). As Arndt (1987: 259) points out: “It is the constraints of the composing activity, or of the discourse type, which creates problems for students writing in L2, not simply difficulties with the mechanics of the foreign language.” However, Silva (1992) observed in a study in which he surveyed university students about their own L2 writing processes that exactly these difficulties with lexis and grammar, as well as interference between the L1 and L2, are so cognitively demanding that 1 Sections 9.1 and 9.5 of this chapter are based on Göpferich/ Nelezen (2013; 2014). 222 Part V: L1 vs. L2 Writing not only the form but also the content of L2 written work, and thus the epistemic function of writing, suffer. This leads to texts that are “less sophisticated” and express the ideas of the writer less effectively (Silva 1992: 33). Devine, Railey & Bischoff (1993) came to a similar conclusion from the results of their study comparing the written compositions of 20 first-year college students in the United States, half of whom had English as their L1 and half as their L2. These subjects were further required to complete a questionnaire addressing their writing processes in order to investigate the metacognitive writing models used for L1 and L2 composition. The students writing in their L2 reported having to omit certain content from their texts when they felt they did not possess the linguistic means to express this content correctly, a problem the L1 writers did not have. Unsurprisingly, the L1 essays were also rated more highly than their L2 counterparts (see also the literature review by Cumming 2001). Such findings support the assumption that the epistemic benefits of writing are less pronounced when this writing takes place in the L2. They also warrant the assumption that the epistemic function of writing can only be fully exploited, both in the L1 and in the L2, if students have achieved a certain minimum fluency with regard to lower-order processes, for example, at the lexical and grammatical levels, when writing in the respective language (Cumming 1989: 126). Thus, offering students compensatory writing courses, both in the L1 and the L2, which address lexical and grammatical decisions can be assumed to be useful for preparing them to derive the maximum benefit from discipline-specific courses, where writing, either in the L1 or the L2, is used as a means of more profound reflection. Several studies have established a correlation between the level of L2 proficiency and the varying amounts of attention given to different aspects of the writing process. From an analysis of English and French texts produced by native English-speaking university students while thinking aloud, Whalen & Menard (1995) found that L2 writers with insufficient L2 competence tend to neglect important macro-level writing processes, including planning, evaluation and revision, in order to focus on lower-level processes. Schoonen et al. (2003) provide further support for this finding from their study in which 281 8th-grade pupils composed texts in both their L1 and L2, the quality of which was then compared with their overall language competency: “The L2 writer may be so much involved in these kinds of ‘lower-order’ problems of word finding and grammatical structures that they may require too much conscious attention, leaving little or no working memory capacity free to attend to higher-level or strategic aspects of writing, such as organizing the text properly or trying to convince the reader of the validity of a certain view. The discourse and metacognitive knowledge that L2 writers are able to exploit in their L1 writing may remain unused, or underused, in their L2 writing.” (Schoonen et al. 2003: 171) Roca de Larios, Manchón & Murphy (2006) arrived at a similar interpretation after analyzing the L1 and L2 (English) texts and accompanying think-aloud 9 Writing in the L1 vs. writing in the L2 223 protocols (TAPs) of 21 Spanish-speaking subjects who were separated into three groups based on their levels of English proficiency: “In L2 writing […] the patterns emerging from the data indicate that the lower the proficiency level of the writer, the more he or she engages in compensating for interlanguage deficits vis-à-vis ideational or textual occupations” (Roca de Larios/ Manchón/ Murphy 2006: 110). In accordance with this, Roca de Larios, Murphy & Manchón (1999) also found language proficiency-dependent differences in the frequencies with which different types of restructuring, i.e. searches for alternative syntactic plans for linguistic, ideational or textual reasons, are used in formulation processes. In their study, more language-proficient writers used restructuring more for stylistic, ideational and textual reasons whereas less proficient writers needed to resort to restructuring more for compensatory purposes due to a lack of linguistic resources in the L2. Such results warrant the assumption that L2 writing processes only strongly resemble L1 writing processes after a certain L2 competence threshold level has been reached (cf. Cumming 1989: 126; Sasaki/ Hirose 1996: 156; Roca/ Murphy/ Manchón 1999; Kohro 2009: 16). Below this threshold level, students’ L2 proficiency seems to have a major impact on their L2 writing ability (Hirose/ Sasaki 1994: 217). If this competence threshold level has been reached in the L2, however, L1 and L2 writing processes seem to resemble each other more at the language-remote levels of idea generation, planning and revising than at the language-close levels of lexical and grammatical encoding (cf. the overview in Roca de Larios/ Muphy/ Manchón 1999: 29 f.). In this connection, Sasaki (2004) brought up the question whether the writing processes of highly proficient L2 writers ultimately resemble those of highly proficient L1 writers or whether L2 writing processes generally, even at the highest competence levels, differ from the respective processes in the L1, at least with regard to certain features. More research is needed to answer this question. What has to be taken into account in many of the above-mentioned studies is that the writers’ language proficiency might have been confounded to a certain extent with their writing expertise. Cumming (1989) was among the first to consider writing expertise and L2 language proficiency separately. In a study with 23 participants who had different combinations of writing expertise and L2 proficiency for which they were controlled, Cumming (1989: 123) came to the conclusion that writing expertise is independent from language proficiency, again once the latter has reached a certain threshold level. 2 He found that “writing expertise is a central cognitive ability-with second-language proficiency adding to it, facilitating it in a new domain, and possibly enhancing it”. Furthermore, he found that the average writers’ performance benefited 2 Cf. Sasaki & Hirose (1996) who found that L2 proficiency accounted for 52% of their participants’ L2 writing ability variance. 224 Part V: L1 vs. L2 Writing much more from higher ESL proficiency than that of participants with either high or low levels of writing expertise (Cumming 1989: 105). Differences observed between the writing processes of more and less proficient L2 writers that can also be observed between more and less proficient L1 writers include that more proficient writers compose longer and more complex texts and write faster and more fluently than novices (Sasaki 2000: 271, 282). More proficient writers also plan more at a global level whereas less proficient writers plan more at a local level (Sasaki 2000: 273 f., 278). In L2 writing, however, novices often stop to translate ideas they have generated from their L1 into their L2 (English), whereas expert L2 (English) writers rather stop to refine their English expressions (Sasaki 2000: 282). At Justus Liebig University, Giessen/ Germany (JLU), where I teach, as well as at many other universities, students pursuing degrees in the fields of English literature, culture and linguistics are generally required to write their term papers, even their first ones, and final theses in English. These students are thereby immediately confronted with two concurrent challenges: the challenge of academic writing, which itself requires the students to adapt to a specific form of discourse with which they are not yet familiar, neither in their L1 nor in English, and the very challenge of first having to do this in their L2. Against this background, the question arises as to whether the requirement of writing academic texts in the L2, before having mastered this skill in the L1, leads to such an increase in task complexity that it overburdens students, which could have consequences reaching beyond the poorer linguistic quality that L2 compositions inevitably display. Having students write term papers in their L2 may further result in a less profound analysis of the subject matter, not to mention a less profound treatment of the L2 literature associated with the subject matter. These potential consequences of requiring students to write academic texts in their L2 may, in turn, be detrimental to the epistemic function of writing. This fear is nourished by findings from studies comparing academic achievement of international students at US colleges and universities who had completed high school in their non-English first-language environments with that of L2 students studying in their L2 who had graduated from US high schools. These studies consistently show that the former outperformed the latter (Muchinsky/ Tangren 1999). As Leki, Cumming & Silva (2008: 19) conclude with reference to Bosher & Rowenkamp (1992) and Cumming (2001), “the best predictor of academic success in college for these students is the number of years spent in high school in L1 before immigration”. In accordance with Cummins’ (1981) interdependence hypothesis, it can be assumed that a cognitive-academic language proficiency 3 that has been ac- 3 Carson & Kuehn (1992: 159) term this general competence “generalized discourse competence“. It is “the ability to produce context-reduced academic prose in both L1 and L2 as a function of common underlying cognitive-academic language proficiency”. 9 Writing in the L1 vs. writing in the L2 225 quired in one language can be transferred to another if a threshold level of proficiency in that other language has been achieved. This, however, requires active practicing of writing academically in this other language, “adequate exposure” to it (Cummins 1981). At the same time, academic writing skills in the L1 may decline if writers exclusively write in their L2. This decline may occur in spite of the fact that the general-academic language proficiency is retained since the decline is limited to the language-close levels of writing (Carson/ Kuehn 1992: 163). Carson & Kuehn (1992: 176 f.) also assume a writing aptitude, which “imposes a ceiling on writing development in the L1”. This ceiling, they assume, affects writers in whichever language they compose. For writing skills transfer from the L1 to the L2 the following ensues from this: “Given the appropriate educational context, good L1 writers will be good L2 writers, but poor L1 writers may not rise above the level of their L1 abilities to become better L2 writers. If poor L1 writing results from lack of L1 educational experience and there is writing aptitude, then there is potential for good L2 writing to develop.” (Carson/ Kuehn 1992: 177) The potential risks for the individual academic development of students when exposed to an L2 for academic purposes before they have achieved a certain L2 proficiency level are also corroborated by studies which analyzed the effect of English-medium instruction in mathematics and engineering. They point to negative effects of abandoning instruction in the local languages in favour of English. One longitudinal study (Klaassen 2001), however, found that this negative effect of learning in English diminishes within one year (on the effects of English-medium instruction on learning, see the research overview in Björkman 2013: 22 ff.). We do not yet have contrastive studies which provide an answer on the question whether English academic text quality of, for example, German students who have acquired academic literacy in German first before learning to write academic texts in English exceeds that of German students whose academic writing socialization has taken place in English directly. An argument in favour of academic writing socialization in students’ L1 is that academic writing in the L1 might never be learned once academic writing in the L2 English has become the default (cf. Casanave 1998; Flowerdew 2000; Shi 2003) thus leading to domain losses in the non-English national languages (cf. also Tang 2012: 228). The differences between L1 and L2 writing addressed so far have developmental origins. Apart from such developmental differences, culture-specific differences have to be taken into account as well. Such culture-specific aspects were first addressed by Kaplan (1966), who established the field of contrastive rhetoric. Rhetoric, in this context, must not be misunderstood in the sense of Aristotle as a technique of persuasion, but rather refers to patterns of discourse 226 Part V: L1 vs. L2 Writing organization (Casanave 2004: 27), which may differ from culture to culture (cf. Hinkel 2002: 30 ff.; for a critical overview of differences found, see Göpferich 1995: Chapter 8). Contrastive genre analysis is one method to sensitize teachers and students for culture-specific differences in text organization, genre conventions, methods of creating coherence, etc. What is important for their own writing is “to avoid succumbing to stereotypes” (Casanave 2004: 49) and to make their decisions as authors dependent on the expectations of the audience they write, or translate, for (Casanave 2004: 44). 4 Rienecker & Stray Jörgensen (2003b: 104) point out important differences between the European continental tradition and the Anglo-American tradition of academic writing, which “spring from fundamentally different perceptions of the nature of science: The continental tradition emphasises science as thinking, while the Anglo-American tradition emphasises science as investigation and problem solving.” They argue that, due to these differences, approaches to writing instruction in the US cannot be transferred to Europe without adaptation. They also assert, however, that due to larger student intake and the transformation of German universities into mass universities, the apprenticeship model, in which professors pass on their knowledge to their students who “are supposed to sit at the feet of their masters and absorb their writing themes and styles” (108), is no longer feasible. Furthermore, they observe that, at the same time, the Anglo-American style of writing is adopted in Europe in an increasing number of disciplines thus leading to an assimilation of the situation in Europe to that in the US, even if this causes much concern among European researchers, especially in the humanities (Rienecker/ Stray Jörgensen 2003b: 108). The causality that Rienecker & Stray Jörgensen establish between the ‘massification’ of European universities and the adoption of Anglo-American writing conventions in Europe seems plausible although other, and perhaps more important, influences have to be taken into account as well. 9.2 Translating into the L2 vs. composing in the L2 Writing a text in one’s L1 first and then translating it into the L2 may be a means of circumventing cognitive overload that may occur when composing directly in the L2. This observation was made by Uzawa (1996) in a study in which 22 university students had to complete three writing tasks: one writing task in their L1 (Japanese), one in their L2 (English) and one translation of a completed L1 text into their L2. Their translations were higher in linguistic 4 On contrastive rhetoric, see also Kaplan (1967; 1980), Connor (1996), Hirose (2003) and the discussion in Casanave (2004: Chapter 2). Apart from contrastive rhetoric, culture-specific aspects of L2 writing have also been addressed in social process approaches (Bizzell 1982), in the transculturation model (Zamel 1996; 1997) and from the contact zones perspective (Pratt 1991). For a critical overview of these approaches, see Canagarajah (2002). 9 Writing in the L1 vs. writing in the L2 227 quality than their L2 texts, which the author attributes to the fact that the translation task relieved the subjects of extensive planning processes, resulting in more attention available for linguistic details. The subjects found the translation exercises more helpful than essay writing because they felt that translating forced them to use vocabulary that they would not have thought of when composing directly in their L2 (Uzawa 1996). Translation can thus be assumed to serve as a useful tool in writing pedagogy. Through translation, students can practice writing with a reduction in complexity, particularly on the macrolevel, as the source text already provides the contents to be composed in verbalized form, allowing the students to pay greater attention to subtleties on the micro-level they might have otherwise ignored. This advantage of translation is also corroborated by Kim (2011), who found that having her students translate from their L1 into their L2 enabled them to evaluate their L2 texts more critically. Furthermore, as Uzawa’s (1996) students confirmed, requiring students to render specific content provided in their L1 into their L2 may raise their awareness for lacunae in their L2 lexis and grammatical repertoire, a process referred to as noticing (see, e.g., Schmidt 1990; Qi/ Lapkin 2001). This noticing may then induce them to try and close these gaps, which might not be the case when they formulate in their L2 directly, where they could try to circumvent expressing ideas or even forgo generating ideas that they cannot express in their L2 (on the relevance of noticing for second-language learning, see also the research overview in Uzawa 1996: 272 f.). Translation and translation as a subprocess of L2 text production, however, may also have disadvantages as will become obvious in Sections 9.3 and 9.5. 9.3 Translating from the L1 as a subprocess of writing in the L2 Translating from the L1 has been seen rather negatively in most foreign language instruction paradigms from the humanistic approach until today and has more or less been banned from L2 teaching (see the research overview in Liu 2009: 2 ff.). Fries (1945), for example, recommended avoiding translation or seeking L2 lexical equivalents for L1 expressions until learners had established a direct connection between what they wanted to express and adequate formulations for it in the L2. He was convinced that “translation and ‘word equivalents,’ which seem to save time at the beginning, really cause delay in the long run and may if continued even set up such habits and confusion as to thwart any control of the new language” (Fries 1945: 6). Such negative attitudes towards translation in L2 learning and writing based mainly on the assumption that maximum exposure to the L2 was the best way to learn an L2 and that use of the L1 interferes negatively with the L2 may explain why translation had received little attention in L2 writing research until the 1980s and is still rarely addressed today (Liu 2009: 12; see, however, Lange 2012). The few empirical investigations of translation in L2 writing processes have revealed, 228 Part V: L1 vs. L2 Writing however, that translating from the L1 is a process that occurs naturally in L2 writing (see, e.g., Cumming 1989; Qi 1998; Roca/ Murphy/ Manchón 1999; Cohen/ Brooks-Carson 2001; Wang/ Wen 2002; Sasaki 2004; Liu 2009). Moreover, not only negative transfer from the L1 to the L2 composing processes has been observed but also positive transfer, especially at the stages prior to formulating the actual text, such as idea generation, organization and elaboration (Uzawa/ Cumming 1989; Cumming 1989; Kobayashi/ Rinnert 1992; Uzawa 1994; Woodall 2002). Liu (2009) was able to show that the functions for which the L1 is used in L2 composing processes depends on the writers’ L2 proficiency. In an experimental study, six native speakers of Chinese who had been educated in Taiwan, had learned English since the age of 12 and had not lived in other countries for more than a year had to compose a text in their L2 English while thinking aloud. The topic they had to write about was a comparison of American Christmas and Chinese New Year. Liu assumed that her participants had acquired knowledge about this topic basically in their L1 Chinese so that the writing assignment would evoke more L1 concepts and thus induce the participants to translate from their L1 into their L2 during the composing process (Liu 2009: 43). The participants were encouraged to take as much time as they needed to complete the assignment and were requested to verbalize anything that crossed their minds during the composing process in whatever language it occurred to them. Three of the six participants had TOEFL scores of 590 and above (high-proficiency group), the other three had TOEFL scores of 570 and less (low-proficiency group). Their think-aloud was transcribed into protocols and segmented into units. The units were classified into four categories: 1. thinking aloud in English only (L2 only), 2. thinking aloud in Chinese only or Chinese first and then translating into English (L1 only or L1 --> L2), 3. thinking aloud in English first and then repeating in Chinese (L2 --> L1) and 4. thinking aloud in unidentifiable chunks (Liu 2009: 45). In addition, she conducted cued retrospective interviews. In accordance with Hirose & Sasaki’s (1994) findings obtained from Japanese EFL writers and Wang & Wen’s (2002) results obtained from Chinese ESL/ EFL writers, Liu (2009) found that the low-proficiency group used their L1 significantly more during the L2 composing process than the highproficiency group. Furthermore, Liu observed that the low-proficiency group also relied more often on their L1 to reconfirm or monitor ideas expressed in their L2 than the high-proficiency group (Liu 2009: 54) and that the lowproficiency L2 writers translated significantly more at the syntactical level during the L2 composing process than at the semantic level, whereas more proficient L2 writers translated significantly more at the semantic level than at the syntactical level. In other words, low-proficiency L2 writers were more fixed on L1 syntactical structures whereas high-proficiency L2 writers more or less just retrieved concepts via their L1 and then went on composing directly 9 Writing in the L1 vs. writing in the L2 229 in their L2. The typical procedure of a low-proficiency L2 writer is reflected by the following statement: “Usually, I use Chinese to generate ideas, and if I like the idea, I will try to translate it into English... If I don’t use Chinese to lead the phrase or words, I’ll forget about what I want to say in English.” (Liu 2009: 68) Liu (2009: 68) explains this observation as follows: “This quote suggests that the L2 operation consumes too much cognitive energy and produces too much mental load for the unskilled writers to conceive of semantic formulations as well as to organize them with syntactic structures for textual production. Therefore, unskilled writers tend to rely on L1 to generate and form ideas in words and phrases. Once the idea has been well formulated semantically and has been represented by L1 syntactic structures, unskilled L2 writers may finally translate the L1 idea into L2 with L2 syntax. In other words, the unskilled L2 writers use L1 to take care of as many cognitive subprocesses as possible to reduce their mental loads. As a result, the L1-L2 code translation may take place at the level close to the textual output, i.e., the syntactic level. Since most of the semanticlevel concerns have been taken care of by L1, unskilled L2 writers may primarily pay attention to the use of L2 for the syntactic and lower level activities, such as orthography, grammar, equivalent lexical choices, and local changes. In a nutshell, skilled L2 writers tend to have more semantic transformation, whereas unskilled L2 writers tend to have more syntactic translation.” Liu (2009: 69) also observed that skilled L2 writers may resort to the strategies of unskilled writers whenever they encounter difficulties, and that unskilled writers make use of the strategies of skilled writers when they are capable of doing so. The participants of the high-proficiency group mainly used their L1 for higher-order processes such as planning, for controlling the incoming information and editing the written text whereas the intended meaning was expressed directly in the L2 (Liu 2009: 58 f.). Wang & Wen (2002) also analyzed L1 and L2 use in different subprocesses of writing by means of think-aloud protocols of 16 Chinese EFL students. They differentiated between process controlling, idea organizing, idea generation, task examining and text generation. For each subprocess they determined the ratio of the words in their participants’ think-aloud that were uttered in connection with this subprocess in each language and the entire number of words uttered in connection with this subprocess in both languages. In process controlling processes, L1 use dominated with on average 81.5%. For idea organization, their participants used their L1 to an extent of on average 70%, and for idea generation, to an extent of on average 61.5%. Task examination was carried out in the L1 to an extent of on average only 21%, and text generation, the most language-close process, to an extent of on average only 13.5% (Wang/ Wen 2002: 234). In line with Liu’s findings, Wang & Wen also found that the language of text generation depends on the writers’ L2 proficiency: “less proficient writers construct sentences through L1-to-L2 translation, while proficient writers generate text directly in L2” (Wang/ Wen 2002: 240). In the other subprocesses examined, the decline in the use of the L1 that could be observed 230 Part V: L1 vs. L2 Writing with increasing L2 proficiency was less salient among their participants (Wang/ Wen 2002: 241). With regard to these subprocesses, however, the question remains to be answered whether L1 use in these subprocesses declines to a more considerable extent as well once the participants have exceeded a certain L2 proficiency threshold level (Wang/ Wen 2002: 241). 5 In an effort to develop mental translation models for skilled and unskilled writers, Liu does not only draw on the findings summarized above but also on findings from psychoand neurolinguistics and makes reference to Wernicke’s and Broca’s areas. The resulting models, however, are highly speculative, as she admits herself (Liu 2009: 73), and encompass more than is necessary to explain her findings. Therefore, they will not be reproduced here nor explained in more detail. What is important about these models for future research are the levels at which translation takes place (lexical-conceptual vs. syntactical) depending on L2 proficiency. In a think-aloud pilot study in which a single participant had to complete two L2 writing tasks and two translations from her L1 (Chinese) into her L2 (English) as well as to solve two mathematical problems with one task in each set involving high cognitive demand and the other a lower one, Qi (1998: 423) found a positive correlation between L1 use and the task demands. Her data “strongly indicate that whenever the participant intuitively anticipated that the load of a task she faced would exceed the limit of her working memory span, she automatically switched to L1, her stronger language, to process the information in order to minimize the load to which the use of a weaker language might otherwise add.” (Qi 1998: 428) As factors that may influence switching from the L2 to the L1, Qi identified “an implicit need to encode efficiently a non-linguistic thought in the L1 to initiate a thinking episode; a need to facilitate the development of a thought; a need to verify lexical choices; and a need to avoid overloading the working memory.” (Qi 1998: 428 f.) In order to get more insight into the occurrence of translation in L2 composition, its positive and negative effects on the composition subprocesses and thus the role or desirability of translation subprocesses in L2 text composition and multiliteracies instruction, we need investigations which, in addition to the cognitive demand of the tasks, take the following variables into account (cf. also Hirose/ Sasaki 1994: 220): 5 Process-oriented longitudinal studies of writing skills development are a desideratum. Apart from Steinhoff’s (2007) and Pohl’s (2007) corpus-based longitudinal studies of academic writing skills development, one of the few writing skills development studies that have been conducted to date is the one by Sasaki (2004), which covers a time span of 3.5 years. 9 Writing in the L1 vs. writing in the L2 231 1. L2 language proficiency As Liu (2009) and other scholars (e.g., Hirose/ Sasaki 1994; Wang/ Wen 2002; Woodall 2002) 6 have been able to show, the role of translation in L2 text composition depends on the writers’ L2 proficiency. The more proficient the writer is in the L2, the less prominent the role of translation from the L1 into the L2 becomes during the composing process. A potential explanation for this could be that in the brains of bilinguals with low language proficiency the L1 and L2 lexica are stored independently whereas in the brains of highly proficient bilinguals this is not the case (Perani et al. 1996). Those writers whose L1 and L2 lexica are stored in the same cortical structures may be able to access L2 lexical entries directly via the concept whereas those L2 writers who have separate L1 and L2 mental lexica may only be able to access L2 lexical items via L1 lexical entries and thus through translation (Liu 2009: 24). If this assumption holds true, the type of the writer’s bilinguality should also be taken into account as a variable. 2. Type of bilinguality Whether the L1 and L2 mental lexica are stored independently or not has been assumed to be dependent not only on language proficiency but also on several other variables. One of them is the age at which the second language was learned with early bilingualism leading to the two lexica being stored in the same cerebral structures (Kim et al. 1997). Another variable that may have an influence on where the L1 and the L2 lexica are stored in the brain is the manner in which a language was acquired or learned. The L1, which is usually acquired informally in a natural manner, is stored in the subcortical structures, the L2, which is normally learned in a formal manner in classrooms at school, tends to be represented in the cerebral cortex (Paradis 1994; Fabbro/ Paradis 1995; Fabbro/ Peru/ Skrap 1997). 7 3. Composing competence (Krapels 1990) 8 As outlined in Section 9.1, a composing competence can be assumed which exists across languages and is at least partially independent of L2 language proficiency. Hirose & Sasaki (1994: 216 ff.), for example, found that L1 writing ability is highly correlated with L2 writing ability (cf. also Cumming 1989; Sasaki 2000). As a consequence, the amount of translation and the type of 6 Cf., however, Wang (2003) who found no significant correlation between language proficiency and language switching in L2 composition. 7 Cf. also Sasaki (2004), who could observe differences in L2 writing competence development between EFL and ESL writers. 8 With regard to academic writing, this composing competence has also been termed “cognitive-academic language proficiency” (Cummins 1981) and “generalized discourse competence” (Carson/ Kuehn 1992). 232 Part V: L1 vs. L2 Writing translation in L2 writing processes can also be assumed to be dependent on this general composing competence and thus on the level of writing skills in the L1. The more proficiently writers are able to compose in their L1, the more considerable positive transfer from the L1 to the L2 can be expected to be, especially at the higher, textual levels involving macro-structure, organization and coherence, which are more related to logical thinking than to languagespecific rules, in contrast to the lower levels of lexis and syntax, which are more language-dependent. 4. L1 language proficiency Since language proficiency has an effect on the amount of cognitive effort available for higher-order decisions in composing processes, L1 language proficiency, like L2 language proficiency, can also be assumed to have an impact on translation behaviour during L2 composition. 5. The type of cognitive operations in which translation is used In the think-aloud study with 16 Chinese EFL subjects already mentioned, Wang & Wen (2002) found that the further a cognitive process is away from textual output, i.e., the more it involves conceptual, organizational nonlinguistic operations, the more inclined writers are to use their L1. This suggests that the extent to which writers translate during L2 text composition also varies depending on the types of cognitive processes. 6. The language in which knowledge required for completing the writing assignment has been acquired or is provided as well as the language of the writing prompt In an empirical study involving 28 Chinese students who had to compose texts in their L2 English on topics involving either knowledge they had acquired in their L1 Chinese or their L2 English and in which they had to do their planning in Chinese, Friedlander (1990) found that students planned better and produced better contents when the language of knowledge acquisition and planning matched. Qi (1998: 430) hypothesizes that this might only be the case if the knowledge acquired in one language has never been reprocessed in the other. Wang & Wen’s (2002) findings referred to earlier suggest that the language of the prompt for the writing assignments to be completed also has an effect on L1 and L2 use in the composition process, an assumption that also needs to be looked into in more detail in future studies (Wang/ Wen 2002: 240, 244). 9 Writing in the L1 vs. writing in the L2 233 7. Translation competence Translation competence can also be assumed to have an effect on L2 text composition. The less translation competence L2 writers have acquired, the more prone they will be to interference from the L1 due to fixedness on the sourcetext surface structure and a lack of flexibility involving departure from surfacelevel expressions in favour of more language-distant representations. Writers with limited translation competence can thus be assumed to have more linguistically-determined or L1 language-related mental representations which may be in the way of finding idiomatic L2 expressions (Mandelblit 1995; Bayer-Hohenwarter 2012; Göpferich 2013). Highly proficient translators, on the other hand, can be assumed to generate mental representations of what has to be verbalised in the target text that are similar to the mental representations that writers have when composing in their L1. To take account of this, Liu (2009: 11) introduces a concept of translation that is wide enough to include the process of converting ideas into linguistic representations in the L1 or monolingual text composition. She defines translation as follows: “I define translation from a broader perspective. Translation in my research is not ‘the replacement of a representation of a text in one language by a representation of an equivalent text in a second language’ (Hartmann & Storck, 1972, p. 713) at the textual level. It includes the processes of informational or conceptual coding, decoding, and reformulating at the cognitive level. Therefore, the translation process in writing may apply to both monolinguals and multilinguals. For monolinguals, the L1 writing process involves translating conceptual representations into linguistic codes through reorganization, resynthesization, and reconstruction. For multilinguals, especially the bilinguals in this book [ i.e., native speakers of Chinese with English as their L2], the L2 writing process involves not only the writing process of the monolinguals as mentioned above but also the cognitive process of language switch. Under this definition, translation in L2 writing involves research areas such as L1 transfer (linguistic and rhetorical), the use of L1, and language switching (LS).” (Liu 2009: 11) 8. General intelligence L2 proficiency has been found to be related to general intelligence in a study involving EFL students who had learned English in formal education (Sasaki 1996). General intelligence could thus have an indirect (i.e. via language proficiency) or direct effect on translation behaviour during L2 text composition. 9. Metaknowledge of L2 writing Sasaki & Hirose (1996) found that metaknowledge of L2 expository writing together with L2 language proficiency and L1 writing ability made a significant contribution to explaining L2 writing ability variance among the participants in their study of EFL students’ expository writing. Thus, it can be hypothesized to also have an impact on L1-L2 translation in L2 writing. 234 Part V: L1 vs. L2 Writing 10. The oral vs. the written mode A distinction must also be made between translation in oral communication and in written communication. In oral communication, translation may have a negative impact because it reduces spontaneity and immediacy due to the fact that interim versions are generated before a formulation is found that can finally be uttered. In written communication, spontaneity and immediacy are less of an issue. 9.4 An explanatory model of EFL writing ability Sasaki & Hirose (1996) conducted a study to investigate the factors that influence EFL students’ expository writing in English. Their instruments included a standardized English proficiency test, expository writing tasks in the students’ L1 and L2, questionnaires eliciting the participants’ instructional and personal writing backgrounds, retrospective self-reports of the students’ L1 and L2 writing processes and a test of their metaknowledge of expository writing. In this study they found that “(a) students’ L2 proficiency, L1 writing ability, and metaknowledge were all significant in explaining the L2 writing ability variance; (b) among these 3 independent variables, L2 proficiency explained the largest proportion (52%) of the L2 writing ability variance, L1 writing ability the second largest (18%), and metaknowledge the smallest (11%); and (c) there were significant correlations among these independent variables. Qualitative analysis indicated that good writers were significantly different from weak writers in that good writers (a) paid more attention to overall organization while writing in L1 and L2; (b) wrote more fluently in L1 and L2; (c) exhibited greater confidence in L2 writing for academic purposes; and (d) had regularly written more than one English paragraph while in high school. There was no significant difference between good and weak writers for other writing strategies and experiences.” (Sasaki/ Hirose 1996: 137 f.) On the basis of these findings, they developed the model of EFL writing ability depicted in Fig. 9-1. In the diagram, squares indicate measured variables, circles unmeasured latent factors, and triangles, latent background factors. Unidirectional arrows indicate one-way causal relations. Uninterrupted lines are based on the findings of their study, broken lines, on speculation. Process features appear in broken line rectangles (Sasaki/ Hirose 1996: 161). 9 Writing in the L1 vs. writing in the L2 235 Fig. 9-1: Explanatory model of EFL writing ability (Sasaki/ Hirose 1996: 161) 9 9.5 Quality losses in L2 writing and the potential role of translation for writing instruction, text quality improvement and epistemic purposes In order to explore whether the ability of university-level English students to express themselves in written form in their L2 (English) differs from their ability to do so in their L1 (German), and, if this is the case, in which areas of the writing process these differences can be observed, a pilot study was conducted. For this study, six native-speaking German students from an advanced English linguistics seminar titled “Developing Writing Skills”, which was offered at the Department of English at the University of Giessen in the 2011/ 12 winter semester, volunteered to take part in an experiment after the seminar had ended. These six students had English as their major subject in either a master’s program or a high school teacher-training program. Their task during the experiment was based on an L2 assignment they had completed during the seminar in teams of two which required them to write a popular-science article based on one of the L2 term papers one of the two students had written in a previous semester. The goal of the article was to inform 12th-grade pupils (of about 18 years of age) about topics that are typically dealt with in university 9 © 1996 Language Learning Research Club, Univ. of Michigan; republished with permission of John Wiley & Sons; permission conveyed through Copyright Clearance Center, Inc. 236 Part V: L1 vs. L2 Writing English programs. They were explicitly instructed to write these texts so that they would be comprehensible for 12th graders and at the same time arouse interest in the topic (see assignment below). The challenge of this writing assignment was three-fold: First, the students had to reduce the length of what they had covered in their term papers to approx. 500 words and thus select only certain aspects of the term paper to be dealt with in the popular-science article. Second, they had to transform an academic text into a popular-science text. And third, they had to make sure that their popular-science text was selfcontained, i.e., met the readers’ expectations that were established by the author at the beginning of the text. To provide the students with a framework of assessment criteria, they were familiarized with the Karlsruhe comprehensibility concept (see Section 4.3) and provided a model text. After they had composed their first versions of the articles, each team of two had to exchange their article with another team and provide each other peer feedback, again based on the Karlsruhe comprehensibility concept, which, as they were informed, also formed the framework for the assessment of their final versions. On the basis of a) this peer feedback, b) the think-aloud of two students not attending the seminar who were required to reverbalize one article each in an optimizing manner while thinking aloud (on the method of optimizing reverbalization, see Göpferich 2006a, b, c) and c) the general remarks made by the teacher on how the articles could be improved, they were finally required to produce an optimized version of their articles to be submitted at the end of the semester as part of their portfolios, which were graded. This text production process, following the writing-isrewriting principle, should ideally have induced the students to give their very best and thus leave few aspects of the texts that would require further optimization in the subsequent translation experiment. During the translation experiment, the six students who volunteered to take part had to produce a German version of the English text they had composed, following the instructions below: 10 “Please produce a German version of the English article that you composed in the seminar “Developing Writing Skills”. The German version should have the same function and target group as the source text, i.e., it is supposed to provide students in their last year at school insights into topics that are dealt with in an English program at university. Please note that the text you are going to produce should not only be comprehensible but also arouse the students’ interest and motivate them to go on reading. Please also note that you do NOT have to produce a translation that is as literal as possible. If you have ideas as to how your German version can be improved with regard to, for example, its structure or individual formulations, please feel free to, and please do, put these ideas into practice. 10 The participants were provided with these instructions in German. 9 Writing in the L1 vs. writing in the L2 237 While composing your German text, please think aloud, i.e., please utter everything that comes to your mind while working on the text. There is no time limit, and you should not rush. Instead, please work for as long as you feel is necessary to produce a version with which you are absolutely satisfied. All the best of success and enjoy your work! It should be noted that the participants were intentionally not instructed to ‘translate’ the text, but to produce a German version, because the concept of translation might have falsely led them to assume that a literal translation was required and that defects in the source text would thus have to be taken over into the target text. What the participants were rather expected to do was to produce a functional translation, which allows for deviations from the source text if they contribute to making the target text more suitable for its function. The assumptions underlying these instructions were the following: The participants would experience cognitive relief due to the fact that a) they were allowed to use their L1, in which they would have a more differentiated repertoire of linguistic means available to them to express their ideas, and that b) the English text, due to its very existence in an externalized manner, would allow the participants to take a more critical stance towards the structure and line of argumentation of the text. If these assumptions held true, the German texts should have had a more logical structure and been more differentiated semantically than their English source texts. During the experiments, the participants’ think-aloud was recorded using the digital audio recording freeware Audacity and subsequently transcribed in XML following the modified guidelines of the Text Encoding Initiative documented in Göpferich (2010). Their writing processes were furthermore registered using the keystroke-logging software Translog 2006. Since the sample of this study was too small to allow for any generalization of results and since from the results numerous hypotheses can be derived which are worth investigating in future studies, the methods of analysis and findings will be described in detail in what follows. 9.5.1 Data analysis The data obtained from the experiment were analyzed at two levels. The first level of analysis concerned the differences in linguistic accuracy in the L1 and L2 texts. For this, the English source texts and the German target texts were marked according to the following error classification scheme: 238 Part V: L1 vs. L2 Writing Tab. 9-1: Error classification scheme Error category Description/ Example Formal errors punctuation missing or wrong punctuation mark; if both a comma at the beginning and at the end of an insertion are missing, this is counted as only one error; repeated comma errors are counted as repetition errors only if they have the same cause spelling spelling mistake which is not an obvious typo (e.g., Tauchen Sie das Gehäuse ihres Gerätes nie unter Wasser.) formatting line break where there should be none (the participants were not required to do any other formatting in the text); italics where there should be none and vice versa Lexical errors semantic use of words and phrases which do not express the intended meaning either denotatively (semantic: denotation) or connotatively (semantic: connotation). This category includes the use of a wrong register at word level (semantic: connotation). Note: Blendings and wrong illocutionary indicators are classified as separate categories. See also “collocation” and “modality/ illocution”. collocation wrong collocation (e.g., fast speed instead of high speed). This category only includes cases in which the meanings of the words used are appropriate but in which these words cannot be combined for other reasons, such as idiomaticity or convention. blending error caused by melding together parts of linguistic units or constructions which enter working memory simultaneously preposition use of a wrong preposition (see also “other grammar”) modality/ illocution wrong illocutionary indicator, such as sollte (recommendation) instead of muss (instruction). Strictly speaking, these errors could also have been classified as “idiomaticity/ genre” errors (genreconvention errors) because the illocutionary indicators to be used may be determined by the genre conventions (see Göpferich 1995: 308 ff.). redundancy superfluous repetition of meaning components, ideas, statements or words including tautologies (see also “text coherence”) Grammatical errors tense use of wrong tense aspect use of wrong aspect case, number, agreement use of wrong case or grammatical number, mostly after prepositions or in appositions; agreement error mood wrong mood, e.g., in indirect speech 9 Writing in the L1 vs. writing in the L2 239 voice active voice instead of passive voice and vice versa word form morphologically wrong word form, such as adjectives instead of adverbs and vice versa; wrong formation of past tense forms; use of expressions which do not exist; use of words from another language which cannot be expected to be understood in the target culture, etc. syntax syntactic error; constructions which are hard to understand due to their length, long parentheses, etc.; utterances which are grammatically correct but which would only make sense if additional lexical elements were inserted, e.g., The few utterances could not be categorized as either one of the two models. Comment: For this utterance to make sense it would have to be reformulated into: The few utterances could not be categorized as belonging to either one of the two models. valency missing actant; use of a lexical element that requires further specification; applicable also to nominalised verbs, e.g., Aufenthalt (from sich aufhalten in), where the indication of a place is required specifier (article or determiner) use of a determiner, e.g., an article, where there should be none; use of a definite article where an indefinite article should be used; etc. infinitive grammmatically wrong use of an infinitive construction (e.g., Das Wetter war zu schlecht, um schwimmen zu gehen.) secondary subjectivization use of verbs expressing human actions in connection with nonhuman subjects (possible in English but rare in German); e.g., This book describes (correct) vs. Dieses Buch beschreibt (wrong) other grammar other grammatical errors, such as the use of a prepositional phrase instead of a genitive and vice versa (e.g., von seinem Vater instead of seines Vaters) Text-level errors text coherence incoherent text segments, e.g., logically wrong connection of clauses and sentences by the use of semantically inappropriate conjunctions; use of wrong pronouns; sentence not related to its context; referent unclear; missing second part of correlative (two-part) conjunctions; repetition of a noun phrase where a pronoun would be sufficient sense incomprehensible or nonsensical section longer than a phrase (otherwise it is counted as a semantic error), contradictory statements implicitness too much information left implicit, e.g., author does not express something to which a conjunction, etc. refers (e.g., There are three types of birch trees. Therefore, I will describe only one. Here, therefore refers to a sentence that was left implicit, i.e., I cannot cover them all.) 240 Part V: L1 vs. L2 Writing functional sentence perspective (FSP) wrong topic-comment structure (theme/ rheme) rhetoric loss of communicative emphasis or effect (e.g., replacing a poem by a mere description of its content in translation); literal repetitions (see, however, “redundancy” and “text coherence”) Other idiomaticity/ genre unidiomatic expression which does not lead to a change of meaning but may make the text hard to understand and betray that it is a translation in a negative sense; use of expressions which do not conform to genre conventions (e.g., Das Bild ist kein Zufallstreffer. instead of Das Bild ist kein Schnappschuss. and Anfangend mit Namen as a title.) cultural specificity missing adaptation to the target culture or missing cultural neutralisation Every error identified in the English and German texts was highlighted and annotated with a corresponding error tag (for the entire annotated corpus, see Göpferich/ Nelezen 2012). To reduce subjectivity, three raters who were already familiar with the error classification scheme marked the texts separately; discrepancies among the raters’ marks were discussed and reconciled thereafter. In order to assure consistent error classification, errors were always categorized with the largest possible granularity. This means that the attempt was made to categorize errors with regard to the smallest linguistic unit involved in the error, or what could be considered the most elementary linguistic category. Once this was pinpointed, the errors were always classified under their primary cause and/ or most specific error type. For example, implicitness errors also tend to cause coherence problems and can thus also be considered coherence errors. The primary cause of this error type, however, is implicitness and not coherence; therefore, they were always marked as such and not as text coherence errors. In addition to the classification of error type, the process-oriented data collected (i.e., the think-aloud protocols and keystroke log files) were analyzed in order to determine whether or not the errors committed were reflected upon by the subjects. Errors were considered to be reflected upon when at least one of the following occurred: • at least one alternative was generated for the erroneous expression(s) as documented in the think-aloud protocols or the log files, • at least one pause of more than 5 seconds occurred in connection with the erroneous expression(s), or 9 Writing in the L1 vs. writing in the L2 241 • a problem was stated explicitly in the think-aloud. 11 The number of errors which were reflected upon by the subjects is of interest because it provides insight into the subjects’ own awareness of the problematic aspects of their texts. In cases in which a subject reflected on an error but was still unable to circumvent it, i.e., generate and subsequently choose a correct alternative, it can be assumed that the subject lacked competency in the area of text evaluation or optimization strategies but still displayed a certain degree of problem awareness. In contrast, when a subject made an error without reflecting on it, this indicates a lack of problem awareness. The latter shortcoming, i.e., a lack of problem awareness, is assumed to be more difficult to address didactically than lacking evaluation competence or optimization strategies. The total number of linguistic errors that occurred in the texts, while a useful means of comparison, cannot by itself serve as an exhaustive measure of overall text quality. For this reason, a qualitative comparison of the source and target texts was also conducted in which the improvements and deteriorations with respect to the above error scheme were identified and analyzed. The results of this comparison will be illustrated with examples in the subsequent sections. Finally, as a further text quality measure, the assessment scheme depicted in Tab. 9-2, which is based on the Karlsruhe comprehensibility concept, was implemented. 12 Tab. 9-2: Assessment scheme based on the Karlsruhe comprehensibility concept Criteria Communicative function To what extent does the article have a consistent communicative function? fully (4 points) - to a great extent (3 points) - to some extent (2 points) - to little extent (1 point) - not at all (0 points) Audience To what extent does the article take its audience’s requirements and interests into account? fully (4 points) - to a great extent (3 points) - to some extent (2 points) - to little extent (1 point) - not at all (0 points) 11 The alternatives generated or the problem stated explicitly must have referred directly to the respective error category to be considered reflected. 12 Four of the assessment criteria in this scheme were not applied because they were either irrelevant for the completion of the assignment (“Perceptibility”, “Media” and “Special Requirements”) or did not form part of the construct to be analyzed (“Correctness” at the content level). 242 Part V: L1 vs. L2 Writing Mental denotation model (content) Is the mental denotation model adequate (superfluous elements, relevant information missing)? fully (4 points) - a few superfluous/ missing elements (3 points) - many superfluous/ missing elements (2 points) - so many that comprehension greatly suffers (1 point) - so many that text is incomprehensible (0 points) Structure (argumentation) To what extent is the article structured logically at the macro-level? fully (4 points) - to a great extent (3 points) - to some extent (2 points) - to little extent (1 point) - not at all (0 points) Is the article self-contained (clear beginning and clear ending)? yes (2 points) - to some extent (1 point) - no (0 points) To what extent is the article structured logically at the micro-level (functional sentence perspective, connectors)? fully (4 points) - to a great extent (3 points) - to some extent (2 points) - to little extent (1 point) - not at all (0 points) Simplicity To what extent are the lexical elements and grammatical constructions adequate? fully (4 points) - to a great extent (3 points) - to some extent (2 points) - to little extent (1 point) - not at all (0 points) Motivation To what extent does the article attract and hold the audience’s attention (motivation)? adequately (4 points) - to a great extent (3 points) - somewhat (2 points) - vaguely (1 point) - not at all (0 points) Does the article have an attractive title? yes (2 points) - somewhat (1 point) - no (0 points) Correctness To what extent is the article correct contentwise? not applicable in the present study throughout (3 points) - contains minor errors (2 points) - contains major errors (1 point) - is completely wrong (0 points) To what extent is the article correct at the language level? 5 points (perfect) - ... - so poor that it is largely incomprehensible (0 points) Concision Are ideas expressed in a concise manner? concise text (4 points) - only few wordy expressions/ clumsy constructions (3 points) - many wordy expressions/ clumsy constructions (2 points) - very many (1 points) - so many that text is difficult to read (0 points) 9 Writing in the L1 vs. writing in the L2 243 Perceptibility Is the formatting and layout adequate? not applicable in the present study fully (2 points) - with a few exceptions (1 point) - with many exceptions (0 points) Mental convention model (style) Is the style adequate? fully (4 points) - to a great extent (3 point) - to some extent (2 points) - to little extent (1 point) - inadequate (0 point) Media Are the media selected adequately? not applicable in the present study fully (2 points) - with a few exceptions (1 point) - with many exceptions (0 points) Special requirements? To what extent have special requirements (legal, formal, etc.) been taken into account? not applicable in the present study fully (2 points) - with a few exceptions (1 point) - with many exceptions (0 points) 9.5.2 Results and discussion The contrastive analysis in which the source and target texts were assessed according to the Karlsruhe comprehensibility concept yielded the following results: No noteworthy difference in the scores of the English texts and the German texts could be observed for any of the subjects. Out of a maximum score of 45 points, the discrepancy between the source-textand target-text scores was only found to be between +4 and -2 points; in three cases, the text quality of the English texts was slightly better than that of their German counterparts, and in the three remaining cases, the opposite trend was observed (for the complete results, see Göpferich/ Nelezen 2012). This lack of significant change from the English to the German texts arose from the manner in which the subjects composed their German texts: Instead of attempting to make changes on a macro-level, the subjects primarily transferred the contents of the source texts into the target texts on a sentence-to-sentence basis and thus focused on the micro-level, i.e., the sentence level and level of neighbouring sentences. This behaviour is typical of translation novices and may be due to the students lacking translation competence. The changes made to the texts at the micro-level had little overall effect on macro-level issues such as the functional adequacy of the texts and their appropriateness for their audience. It should also be noted, however, that the similarity of the L1 and L2 texts on the macro-level may simply signal unawareness on the part of the subjects concerning the structural shortcomings of their texts, both in their L2 and in their L1. If the latter is the case, this would confirm Arnt’s (1987) assumption that 244 Part V: L1 vs. L2 Writing L1 and L2 writing processes are, on a global level, very similar. The data basis of the analyses of this study, however, is too small to draw such conclusions. Tab. 9-3 provides an overview of the linguistic errors made by every subject in their source and target texts, and further reports on the number of reflected errors in the German texts. The results in Tab. 9-3 show that, contrary to the assumptions stated in Section 9.5, more errors were made in the German texts (227 total errors) than in the English texts (186 total errors). If the error class “Other”, in which no difference was observed between the English and German texts, is disregarded, the number of errors in the German texts is actually higher in every category (formal errors, lexical errors and grammatical errors) with the exception of text-level errors, of which there were fewer in the German texts. Tab. 9-3: Number of errors found in the English (E) and German (G) texts and number of reflected errors (R) in the German texts (n = not applicable) FrSc InMa JeCr LaRe LaSe RiDö Totals Error Category E G R E G R E G R E G R E G R E G R E G R Formal errors 5 8 0 3 7 0 5 13 3 9 13 2 3 4 0 2 13 0 27 58 5 punctuation 4 6 0 0 3 0 4 10 2 7 8 1 3 3 0 2 11 0 20 41 3 spelling 1 2 0 3 4 0 1 3 1 2 5 1 0 1 0 0 2 0 7 17 2 formatting n n n n n n n n n n n n n n n n n n n n n Lexical errors 17 20 6 8 16 11 17 20 9 12 6 1 5 11 7 6 5 2 65 78 36 semantic 7 16 6 8 16 11 7 15 8 8 5 1 4 11 7 2 4 2 36 67 35 collocation 2 1 0 0 0 0 2 2 1 0 1 0 0 0 0 1 0 0 5 4 1 blending 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 preposition 3 0 0 0 0 0 3 2 0 2 0 0 1 0 0 2 1 0 11 3 0 modality/ illocution 1 1 0 0 0 0 1 0 0 0 0 0 0 0 0 1 0 0 3 1 0 redundancy 4 2 0 0 0 0 4 1 0 2 0 0 0 0 0 0 0 0 10 3 0 Grammatical errors 10 12 5 6 2 1 10 15 4 8 6 2 6 8 4 5 5 2 45 48 18 tense 1 0 0 2 1 1 1 1 0 1 1 0 1 3 2 0 0 0 6 6 3 aspect 1 0 0 0 0 0 1 0 0 0 0 0 2 0 0 1 0 0 5 0 0 case, number, agreement 1 4 2 1 0 0 1 3 1 1 2 1 0 0 0 0 2 1 4 11 5 mood 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 1 0 0 2 1 voice 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 0 0 1 0 0 word form 1 0 0 0 0 0 1 2 1 1 0 0 0 1 0 0 1 1 3 4 2 9 Writing in the L1 vs. writing in the L2 245 syntax 2 2 1 2 0 0 2 3 1 3 1 0 0 0 0 0 0 0 9 6 2 valency 0 0 0 0 0 0 0 0 0 0 0 0 0 2 1 0 0 0 0 2 1 specifier 4 3 1 1 0 0 4 2 1 2 1 1 3 1 0 2 1 0 16 8 3 infinitive 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 secondary subjectivization 0 2 1 0 1 0 0 1 0 0 1 0 0 0 0 0 0 0 0 5 1 other grammar 0 1 0 0 0 0 0 3 0 0 0 0 0 0 0 1 0 0 1 4 0 Text-level errors 6 10 9 2 1 0 6 9 1 21 11 3 8 7 1 2 1 0 45 39 14 text coherence 2 5 4 0 0 0 2 4 1 9 7 2 3 3 0 2 0 0 18 19 7 sense 0 1 1 1 0 0 0 0 0 8 2 0 1 1 0 0 0 0 10 4 1 implicitness 2 0 0 0 0 0 2 2 0 0 0 0 3 0 0 0 0 0 7 2 0 functional sentence perspective (FSP) 0 2 2 1 1 0 0 1 0 1 0 0 1 2 1 0 0 0 3 6 3 rhetoric 2 2 2 0 0 0 2 2 0 3 2 1 0 1 0 0 1 0 7 8 3 Other 0 1 1 2 1 0 0 2 1 1 0 0 0 0 0 1 0 0 4 4 2 idiomaticity/ genre conventions 0 1 1 1 0 0 0 2 1 1 0 0 0 0 0 1 0 0 3 3 2 cultural specificity 0 0 0 1 1 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 0 Totals 38 51 21 21 27 12 38 59 18 51 36 8 22 30 12 16 24 4 186 227 75 There are many possible reasons for such a result, two of which will be discussed here. First, the high number of errors in the German texts might have been caused by the translation task itself. Though the term ‘translation’ was deliberately avoided in the assignment, it is likely that many of the errors were caused by either L2 interference or a strong degree of fixedness on source-text formulations (for more about the phenomenon of fixedness in psychology, see Duncker 1945; for fixedness in translation, see Mandelblit 1995). This is also supported by the fact that students of translation tend to produce errors arising from interference and fixedness at the beginning of their translation training, errors which tend to occur in lesser frequency as translation competence develops. Considering that the subjects in this study had little or no experience in translation, it is reasonable to assume that their behaviour greatly resembled that of translation novices. In remaining as close to the source text as possible, the subjects may have been implementing a type of cognitive relief strategy: In order to save cognitive capacity for other processes (such as generating appro- 246 Part V: L1 vs. L2 Writing priate German renderings of English terms), they may have avoided diverting greatly from the source text, especially on the macro-level. Second, the higher number of errors in the German texts could have reasons independent of the translation task. It is reasonable to assume that the subjects may simply display certain shortcomings in the areas of, for example, German punctuation rules, word choice and grammar. One indication of this is that several subjects expressed their desire to use references such as a dictionary or the internet to assist them in finding appropriate expressions and synonyms, a desire which was not granted during this particular experiment. In light of these observations, an analysis focusing specifically on the L1 errors and their causes would be promising, especially in the paradigm of Situated or Embedded Cognition from the cognitive sciences, which postulates that an individual’s particular work environment and conditions have a decisive influence on his or her cognitive processes. As a consequence, taking subjects out of their usual work environments can be expected to have negative effects on the results of their work (cf. Hutchins 1995; Clark 1997, 2008; Clark and Chalmers 1998). If the high number of errors is due to the subjects’ not being allowed to use references, this would indicate that the subjects lack problem avoidance strategies by means of which many errors could have been avoided. Further interesting insights into L1 and L2 text composition processes can be gained when closely examining the results regarding the text-level errors, the only category in which an improvement from English to German was observed (45 errors in the English texts versus 39 errors in the German ones). When examining these errors for individual subjects, we see that four of the six subjects performed better in their German texts than in their English texts (see Tab. 9-3). This result supports the assumption stated in Section 9.5 that students, at least at the text level, are better able to express themselves in their L1 than in their L2 and seem to take a more critical stance towards their texts’ logical structure and argumentation. Another possible explanation for this result is that the subjects may have been able to improve upon the logical structure and argumentation of their German texts because both were already provided in the English texts, offering them a certain amount of cognitive relief and thus enabling them to analyze them critically when transferring them into German. The subjects did not, however, make improvements from English to German in every text-level subcategory. While a notable improvement can be witnessed in the subcategories of “sense” and “implicitness”, the point values for the subcategories “text coherence” and “rhetoric” differed by only one point. There was even an increase in functional sentence perspective (FSP) errors from the English to the German texts (3 errors in English, 6 errors in German). The lattermost result is likely due to the differences in English and German in inflectional morphology and hence the ways in which these two languages can obtain certain topic-comment structures. Whereas in English, 9 Writing in the L1 vs. writing in the L2 247 the S-V-O word order is relatively fixed, the German language allows for a greater degree of syntactic flexibility due to its rich inflectional morphology. Nonetheless, the subjects often seem to have simply imitated the word order used in their English source texts instead of finding an appropriate German alternative. The extent to which these German FSP errors were induced by fixedness on the English syntactic structures will be more closely addressed in Section 9.5.2.4. The results regarding the reflected errors in the German texts are also remarkable. Of the total number of formal errors made (e.g., punctuation, spelling), only 9% were reflected on by the subjects. In contrast, lexical errors were reflected on most frequently with 46%, followed by grammar errors with 38%. Finally, text-level errors were reflected on in 36% of all cases (though it should be noted that none of the implicitness errors were reflected on). The fact that the percentage of reflected errors was lower for errors on the text-level than for those in the categories of lexis and grammar could be due to a greater degree of fixedness on these latter categories. In other words, it appears that the subjects orientated themselves greatly on the surface structures of the English source texts when it came to lexis and grammar, and had to make a deliberate effort (i.e., had to reflect upon these structures) to free themselves from this fixedness and produce appropriate German expressions. For this to occur on the text-level, in contrast, a smaller degree of concerted effort seems to have been necessary. In these cases, the improvements on the text level likely occurred because the subjects were able to focus on the logical relations between the ideas in their texts, not simply the way they were explicitly expressed in English. The relative degrees of fixedness depending on error types will be discussed in greater detail in the sections to follow. The changes made on the text level are those which most greatly affect the structural and argumentative quality of the texts. For this reason, the analysis which follows will focus on these text-level errors in a contrastive manner, specifically the errors in the categories of text coherence, sense, implicitness and FSP. 13 9.5.2.1 Text coherence errors With 18 errors in English and 19 in German, text coherence errors form the largest subcategory among the text-level errors, accounting for 44% of all textlevel errors. 14 Though one might assume that the errors in English were simply 13 Because the types of rhetorical errors vary greatly, they will not be addressed here but can be found in Göpferich & Nelezen (2012). 14 Here it should also be noted that the other text-level errors, e.g., “sense” and “implicitness”, also carry negative effects on text coherence, but, as stated in Section 9.5.1, they were not classified as such because text coherence was not considered to be the primary error cause. 248 Part V: L1 vs. L2 Writing carried over into the German texts, only four of the 18 errors in the English texts were transferred to their German counterparts. This means that the subjects were able to avoid 14 of the 18 (English) errors in their German texts, a seemingly impressive improvement, but that they also made 15 new errors in their German texts that were not the result of direct transfer. Examples [9-1] and [9-1’], for example, illustrate errors which occurred in the same place in both the English source text and German target text, but the error types that occurred in these places differ: In the English text, the coherence error arises from the use of an illogical connector (adversative instead of additive), whereas in the German text, though a logical connector (sowohl - als auch) is used, the second part of this correlative conjunction (und instead of als auch) does not fit to the first one: 15 [9-1] By using contemporary modern [sic] elements like [sic] cubism and incorporating typical African elements, Douglas was able to create an unequal [sic] African American art style, [sic] that was able to address modern issues, but also represent the [sic] common identity. 16 (JeCr) [9-1’] Indem er Elemente des Kubismus und typisch Afro-Amerikanische [sic] Symbole in seiner Kunst vereinte, hat er einen neuen Afro-Amerikanischen [sic] Kunststil entwickelt, der sowohl moderne [sic] Themen portraitiert [sic] und die Identität der Afro-Amerikaner repräsentiert. In addition to illogical connectors and the inconsistent use of correlative conjunctions, the following text coherence error types were identified in the German and English texts: Reference to an antecedent by means of a noun that does fit: [9-2] Der Hauptgrund lag in der Versklavung von afrikanischen Amerikanern [sic] in der Vergangenheit [sic] und deren geringen [sic] Ansehen und [sic] Stellung in der damaligen [sic] Gesellschaft. Dieser Rassismus bestand nahe zu [sic] unverändert bis in die 1920er Jahre und führte zu einem zugespitzten rassistischem [sic] Denken und zu rassistisch motivierter Gewalt. (FrSc) 17 Use of a full form instead of a pro-form: [9-3] Wenn ein Mensch also von Leid getroffen wurde, suchte er in seinen Überzeugungen, die sich aus Ideal, Religion, Philosophie und eben der [sic] Kultur speisten, nach Antworten auf die „Warum? “-Frage des Leids und fand 15 The error in English could also be considered an inconsistent use of the correlative conjunction not only…, but also…, a likely assumption considering the similar error made in the corresponding German version. Even in this case, the text coherence error type in the English source text differs from the text coherence error type in the German version. 16 The errors illustrated in this section are italicized, while other errors that occurred in the shown excerpts are marked with [sic]. Since the errors marked with [sic] are not focused on here, it is not necessary for the reader to understand why they are marked. 17 For the complete English source texts and their translations, see Göpferich & Nelezen (2012). 9 Writing in the L1 vs. writing in the L2 249 [sic] je nach seinen Überzeugungen unterschiedliche Antworten auf die „Warum? “-Frage. (LaRe) Use of a pro-form whose referent is unclear: [9-4] Weiterhin [sic] war er ernsthaft daran interessiert [sic] die wichtigen afrikanisch-amerikanischen Fragen darzustellen und als „Schwarzer“ [sic] Künstler deren Erfahrungen auszudrücken. (FrSc) Missing coherence-generating element: [9-5] Wie auch immer du dich entscheidest, die Frage nach dem Leid ist für alle Menschen von höchster Bedeutung, daher ist es für dich persönlich wichtig, welche Antwort du dir selbst darauf gibst. (LaRe) In example [9-5], it is supposed to be emphasized that die Frage nach dem Leid (Engl. the meaning of suffering) is important for all people, which is why it is also important that the reader finds an answer for himself. Without the inclusion of auch (Engl. also), the emphasis rather falls on dir selbst (it is important which answer you find for yourself). Incorrect pro-form: [9-6] If the book is an aesthetic novel, then what is the function of the Gothic elements? It becomes clear when looking at both in close connection. (RiDö) Isolated sentence that has no identifiable relation to what is stated before and/ or afterwards: [9-7] Nevertheless, “The Yellow Wall-Paper” is still relevant today since it shows the danger of a declining mental state due to a [sic] wrong or no treatment at all. This danger is also further reinforced. Being nameless, the story’s protagonist can represent anyone. (LaSe) [9-7’] Dennoch ist diese Kurzgeschichte auch heute noch relevant, da sie sehr eindrucksvoll beschreibt, was passieren kann, wenn depressiv erkrankte [sic] Menschen nicht oder falsch behandelt werden. Ein weiterer Abschreckungseffekt wird auch dadurch erzielt, dass die Protagonistin von Perkin Gilmans Geschichte keinen Namen erhält und somit jeden Menschen wiederspiegeln [sic] kann. (LaSe) In the English text, the italicized sentence simply has no apparent connection to the sentence following it. Its German counterpart, in contrast, also has no apparent relevance to the preceding content because of the use of the denotatively incorrect noun Abschreckungseffekt (deterrent) in combination with ein weiterer (a further), falsely signalling that a deterrent was already referenced in the preceding text. Tab. 9-4 provides an overview of the absolute frequencies of each type of text coherence error in the English and German texts. The values in parenthesis in the English column refer to those errors out of the total number which were carried over from the English into the German texts. 250 Part V: L1 vs. L2 Writing Tab. 9-4: Frequencies of text coherence error types Text Coherence Error Type English German illogical connector 13 (3) 8 inconsistent use of a correlative conjunction - 1 reference to an antecedent that does not fit 1 3 use of a full form instead of a pro-form - 3 unclear referent 2 (1) 2 missing coherence-generating element - 1 incorrect pro-form 1 - isolated sentence 1 (1) 1 Totals 18 (5) 19 The low values concerning English text coherence errors carried over into the German texts and the high values concerning text coherence errors first committed in the German texts demonstrate the following: When it comes to verbalizing a text in another language, the relations of ideas that occur in the source text are not simply controlled by the formulations in the source text; instead, the ideas are newly set in relation to each other during reverbalization. The result is that the ability to create coherence in a target text is not as likely to be affected by fixedness as rather seems to be the case in other areas of text composition (e.g., grammar and lexis). However, the fact that so many text coherence errors still appeared in the German texts that are not the result of fixedness points to the need to address more thoroughly - in both the L2 and the L1 classroom - the methods of expressing logical relations between ideas. This can be achieved by using translation exercises that take a contrastive approach to creating coherent texts by means of, for example, connectors and pro-forms. For this purpose, writing instructors can create manipulated source texts with disturbed coherence, where the students’ task is then to identify and rectify these areas of disturbed coherence during the translation process. 9.5.2.2 Sense errors Sense errors occurred less often in the German texts than in their English counterparts (10 versus 4 respectively). It is interesting to note that only one subject (FrSc) made a sense error in her German text that did not occur in the same place in her English text: [9-8] By capturing the spirit of his time in his works [sic] Douglas was among the first African Americans to explore their history and generating [sic] a common pride in their heritage, for instance [sic] by using symbols that represented their existing political rights. (FrSc) 9 Writing in the L1 vs. writing in the L2 251 [9-8’] Als einer der ersten afrikanischen Amerikaner [sic] entdeckte er die eigene Geschichte neu und entwickelte [sic] einen gemeinsamen Stolz auf das afrikanische Erbe, zum Beispiel ihre existierenden, politischen Rechte symbolisiert wurden. The fact that the last portion of the sentence in [9-8’] fits neither syntactically nor contextually to what comes before it suggests that FrSc might have left it in by mistake; her keystroke log file reveals, however, that this information was in fact intentionally added on to this sentence, but that she struggled greatly in coming up with a German formulation with which she was satisfied. In comparing the different formulations FrSc considered, it can be assumed that she attempted to create as concise a sentence as possible but sacrificed all coherence to do so, which is probably a result of a lack of motivation to continue refining the sentence to express the intended meaning concisely. In three cases, the sense errors committed in the English source texts were simply reproduced in the target texts, while in three further cases, the English sentences that contained the sense errors were simply omitted from the German texts, signalling perhaps a lack of comprehension of the sentence on the part of the subjects themselves. Though they will not be discussed in greater detail here, these examples can be found in Göpferich & Nelezen (2012). The following example illustrates a case in which a subject was only first able to express herself clearly when writing in her L1 German: [9-9] In Lakoff’s linguistic model [sic], a Democratic President [sic] as a ‘Nurturant Parent’ generally includes the population into the realm of the national decisive agents [sic] and resorts to the first person plural pronoun we. (InMa) [9-9’] In Lakoffs linguistischem Modell [sic] räumt ein demokratischer Präsident als „Fürsorglicher Versorger“ den Bürgern des Landes entscheidende Mitspracherechte im nationalen Familienverbund ein und verdeutlicht dies durch die Benutzung des Pluralpronomens WE [wir]. The italicized portion of the sentence in example [9-9] is incomprehensible, especially to those without a background in Lakoff’s theory, for several reasons, including the polysemy of the word decisive, which can mean both a) ‘having the power or quality of deciding’ and b) ‘important’, ‘crucial’. The author’s think-aloud protocol quoted below shows that she becomes aware of this problem once she has to produce a German version, and is able to successfully correct it by selecting an expression that appropriately corresponds to the intended meaning, namely Mitspracherecht. „bürger in die: als decisive agents ob man die bürger als maßgebliche, naja, maßgeblich sind sie ja nicht, sie haben einfach nur, sie können mitreden und haben entscheidungsfreiheit oder zumindest ein mitspracherecht. als gleichwertig? und gleichwertig verantwortungsvolle, als mitspracheberechtigt? gibt es dieses wort? als mitspracheberechtigt in into the realm ob man die bürger als mitspracheberechtigt nein die bürger des landes oder gesteht ihnen mitspracherecht zu? räumt ihnen ein 252 Part V: L1 vs. L2 Writing räumt () den bürgern des landes ein entscheidendes mitspracherecht ein, ein“ (ll. 381-394) Later, the expression decisive agent occurs again in her English source text. This time, however, it refers to the president alone and should thus be rendered differently. Again, the author makes the right decision by rendering it this time as die entscheidende Autoritätsfigur. These two examples show that InMa was able to generate formulations in her German text that display a vast improvement in comprehensibility over her English version, pointing to the utility of translation to help students become aware of the important role their lexical choices have in text comprehensibility. That fewer sense errors were made in German than in English, and that many of the errors in English were identified by the subjects and subsequently corrected in the German texts, gives further support to the assumption stated in Section 9.5 that the subjects have a higher level of micro-level text-linguistic competence and are better able to differentiate semantically in their native tongue than in their L2. 9.5.2.3 Implicitness errors Implicitness errors occurred in English a total of seven times and in German only a total of two times. Remarkably, these two German errors were the result of their English counterparts being more or less directly transferred in the German texts, meaning the subjects were able to correct the five remaining source-text errors in their German texts. The two that recurred in the German texts were both made by the same subject, JeCr, who reported being raised bilingually. Interestingly, this particular subject committed the highest number of errors in her German text of all subjects, and also had the second-highest number of errors in her English text. JeCr wrote her English source text with another subject in the experiment, FrSc, who was able to avoid both implicitness errors in her German rendering of the same text. The following is an example of one such case: [9-10] Due to the past enslavement of African Americans, they were still believed to be uncultured and rough and were denied a cultural identity by White [sic] society. (JeCr and FrSc) [9-10a] Auf Grund der Versklavung der Afro-Amerikaner in der Vergangenheit [sic], [sic] wurden diese noch immer als unkulturell [sic] und ungebildet angesehen. Die weiße Bevölkerung glaubte, dass Afro-Amerikaner keine kulturelle Identität hätten. (JeCr) [9-10b] Die dominierende, [sic] weiße Gesellschaft hielt afrikanische Amerikaner [sic] für grob und kulturlos [sic], daher verweigerte sie ihnen die Anerkennung einer eigenen kulturellen Identität. Der Hauptgrund lag in der Versklavung von afrikanischen Amerikanern [sic] in der Vergangenheit 9 Writing in the L1 vs. writing in the L2 253 [sic] und deren geringen Ansehen [sic] und [sic] Stellung in der damaligen [sic] Gesellschaft. (FrSc) In example [9-10], a cause-and-effect relationship between the enslavement of African Americans and the white society’s subsequent opinion of them is established. Indeed, it could be argued that the relationship between the enslavement of African Americans and their reputation as “uncultured and rough” was actually the reverse of what is written in [9-10]: It was, among many other complex factors, these types of opinions about Africans in the United States that led to the social acceptance of their enslavement in the first place. Furthermore, the fact that white society had not substantially improved their view of African Americans at the beginning of the 20th century was not due to their enslavement itself but (at least partly) to the consequences of this enslavement. Certainly, former slaves and their children and grandchildren were by and large not given the opportunity to receive a high level of education, and, due to a different cultural mentality, did not fit into what were considered at that time to be cultural norms. These factors led to an African American generation that was, for the most part, viewed as uneducated and uncultured. In JeCr’s German rendering of the text in [9-10a], she kept the relation between the two ideas - slavery and the view that African Americans were uncultured and rough - a causal one, not making further thoughts about it during the experiment. FrSc, in contrast, became aware during reverbalization that there was a problem with the sentence (TAP ll. 680-690) and took measures to try and change it. Instead of establishing slavery the cause of this view, she made this view the cause of their being denied a cultural identity, a causality that is explicitly stated. Though her German text [9-10b] does not, in the end, ideally describe the role of slavery in this situation, both the improvements in her text and her considerations while thinking aloud signal her ability to approach writing in her L1 more critically than she did in her L2. Extract [9-11] exemplifies another case in which the author (LaSe) was, in her German text, successfully able to avoid making the implicitness error she made in her English text: [9-11] Though she is seriously ill, her husband and physician John does not trust her opinion and prescribes her a medication [sic] which insidiously worsens her condition. (LaSe) [9-11’] Obwohl sie ihrer Meinung nach äußerst krank ist, sind alle ihre Bemühungen [sic] ergebnislos. Sie wird von ihrem Mann, der zugleich auch ihr Arzt ist, einfach nicht wahrgenommen [sic]. In [9-11], it seems that the author wanted to express that the protagonist believed that she was seriously ill and made every effort to convince her husband of this, but that he, in spite of all her efforts, did not believe her. The conjunction though should thus not refer to the assertion that she was seriously ill but 254 Part V: L1 vs. L2 Writing to her efforts to convince her husband, an assertion left implicit in this sentence. In [9-11’], we see that the author is aware of the shortcomings of her expression of ideas in her English sentence and includes both ihrer Meinung nach (in her opinion) and Bemühungen (efforts) to make the relationship between the two statements more explicit; ideally, however, these efforts should have also been specified more closely (i.e., efforts to do what? ). The results concerning implicitness errors, i.e., that only one (bilingual) subject transferred these errors into her German text, while the rest of the subjects were able to avoid them, indicates that the cause of such errors may be the inability for the subjects to express themselves explicitly in a foreign language to the extent they can in their native tongue. As a type of avoidance strategy, perhaps, they may simply omit what they have difficulty expressing in their L2, negatively impacting the comprehensibility of these texts. This exclusion of content also has negative effects on the epistemic function of writing, as students do not practice expressing their ideas precisely and completely. 9.5.2.4 Errors concerning functional sentence perspective (FSP) Of the three FSP errors that occurred in the English texts, two were corrected in the German versions, whereas one was simply taken over in the German text. 18 This means that of the total six FSP errors in the German texts, five correspond with formulations in the English texts that are not erroneous. Consider first the two following errors that could be avoided in the German texts: [9-12] So decide for yourself which paradigm sounds more convincing to you, as they will be explained more in detail in this article. (LaRe) [9-12’] Diese beiden (Denk-)Muster werden im Folgenden genauer beschrieben, es ist an Dir [sic] zu entscheiden, welche der beiden Erklärungen dem Leiden mehr Sinn geben. [9-13] Women were often kept without love in a domestic sphere and not taken seriously by their husbands. The short story “The Yellow Wall-Paper” [sic] written in 1892 by Charlotte Perkins Gilman (1860-1935), an American writer, is an outstanding illustration of these attitudes and the treatment of 19th c. women. (LaSe) [9-13’] Oftmals wurden an Depressionen erkrankte Frauen von ihren Ehemännern nicht wahrgenommen und zu Hause einfach eingesperrt. Dies passiert auch in der Kurzgeschichte „The Yellow Wall-Paper“ (im Deutschen: „Die gelbe Tapete“) aus dem Jahr 1892, welche von der amerikanischen Schriftstellerin Charlotte Perkins Gilman verfasst wurde. 18 This error is not addressed here. A side-by-side comparison of all the text-level errors in English and German can be found in the data documentation in Göpferich & Nelezen (2012). 9 Writing in the L1 vs. writing in the L2 255 In the English excerpts [9-12] and [9-13], information with relatively low communicative importance is presented at the prominent rheme position at the end of the sentence (see italicized text). In both German renderings, this information is now presented at an earlier position in the sentence, with in this article being rendered as im Folgenden, and is an outstanding illustration … quite smartly being replaced with the deictic expression dies to refer back to the entire preceding sentence and establish it as the topic of the new one. In both cases, the principle of theme-rheme is first instated in the German texts. The next case exemplifies an error (committed by LaSe) that, as mentioned in Section 9.5.2, first occurred in the German text due to the imitation of the English word order, though in German, the order in which the information about frequency and location is placed should have been adjusted to create the appropriate communicative dynamism. [9-14] Melancholia, burnout-syndrome [sic], depression - mental diseases seem to be increasingly common in today’s society. (LaSe) [9-14’] Melancholie, Burnout-Syndrom sowie Depressionen - physische [sic] Krankheiten treten immer häufiger in unserer Gesellschaft auf. The premise regarding the above error is that it is the result of interference from the English source text. In other words, the task of translating itself may have caused the subject to create a German text with syntactical structures the subject, in a free writing situation, would not have created. The following error also very likely occurred because of the translation task: [9-15] Due to the past enslavement of African Americans, they were still believed to be uncultured and rough and were denied a cultural identity by White [sic] society. (FrSc) [9-15’] Der Hauptgrund lag in der Versklavung von afrikanischen Amerikanern in der Vergangenheit und deren geringen [sic] Ansehen und [sic] Stellung in der damaligen Gesellschaft. The author’s TAP (ll. 640-687) reveals that she struggled greatly with finding an appropriate German equivalent for the modifier past, attempting several times in vain to incorporate a pre-nominal modifier before finally settling upon in der Vergangenheit, yet still apparently dissatisfied with the result of the sentence. Though semantically correct, the post-nominal position of in der Vergangenheit places undue weight on its communicative importance in the sentence; rather, the emphasis should be placed on Versklavung (enslavement) as could easily be achieved in English by the placement of past before enslavement. One possible solution would have been to express the enslavement in verbal form, such as in [9-15’’]: [9-15’’]Der Hauptgrund lag darin, dass die Afro-Amerikaner in der Vergangenheit versklavt worden waren … 256 Part V: L1 vs. L2 Writing Such errors could also be the result of fixedness on the wording in source texts, which, as mentioned in Section 9.5.2, often occurs among translation novices lacking the necessary translation competence. If this is the case, then half of the FSP errors in German (three of six) can be attributed to this cause. In contrast, the FSP error in example [9-16], which occurred in the first sentence of this particular text, cannot be ascribed to fixedness. [9-16] The United States of America is the land where revolutions are born. Aside from politically motivated ones …, important cultural revolutions derive [sic] their origin from America. (FrSc) [9-16’] Viele Revolutionen wurden in den Vereinigten Staaten von Amerika geboren. Eine dieser Revolutionen war in der Lage … Here, the subject did not imitate the word order in the English text but intentionally altered it in the target text. Because revolutions being born is the new information being introduced in this sentence - and is also taken up as the topic of the subsequent sentence in linear progression - it should have appeared, as it did in the English sentence, at the sentence-final rheme position. The TAP of the author of [4-5’] clearly reveals her motivation behind diverting from the word order in the source text: Though she generated alternatives that included the revolutions both at the beginning and at the end of this sentence, she was apparently preoccupied with other aspects of the sentence with which she was not satisfied, namely the verb phrase geboren werden and the relative pronoun wo: „the united states of america is the land where revolutions are born. OH, noch ein kurzer. würde aber ein bisschen umstellen. ich würde, glaube ich, schreiben ehm also statt in: in den vereinigten staaten von amerika werden revolutionen geboren, obwohl das sich auch nicht schlecht anhört, ich schreibe, glaube ich, beides mal hin. also in den vereinigten staaten von amerika werden ist das, die vereinigten, also übersetzt wäre es ja ganz genau die vereinigten staaten von amerika ist das land, wo revolutionen geboren werden. machen wir erst mal ups, da gehörst du nicht hin, das ist die vereinigten staaten von amerika ist ... ist das land, wo revolutionen geboren werden ... so. das mit dem wo gefällt mir nicht ... soh: , man könnte auch umstellen, in revo nicht die, man könnte ja auch die revolutionen nach vorne machen. viele revolutionen ehm wurden in den vereinigten staaten von amerika geboren? vielleicht finde ich noch etwas anderes für geboren. in den viele revolutionen wurden in den vereinigten staaten von amerika geboren. dann gebo: ren. geboren okay, das ist kürzer, ich habe diesen blöden, langen satz nicht drin. (ll. 197- 222) The author’s neglect of the requirements of this sentence’s communicative dynamism in this case might indicate a lack of awareness for such requirements. In general, the fact that the subjects do not seem to be aware of the principles underlying the arrangement of words from a communicative standpoint may be the cause of their lack of interference resistance in this area of text grammar. It can be assumed that such errors could be avoided with the 9 Writing in the L1 vs. writing in the L2 257 help of focused instruction concerning FSP. That topic-comment structures are not a part of the German school curriculum makes it all the more important to address FSP in writing classes at higher education institutions. This can be accomplished in a contrastive manner, with, for example, translation exercises in which target-text versions require an adaptation of word order due to the requirements of communicative dynamism. 19 9.5.3 Conclusion The comparison of the text-level writing competence of the six advanced English students revealed that, while fewer sense and implicitness errors were made when subjects wrote in their L1, there was no significant difference in the total number of text coherence errors between L2 and L1, and FSP errors occurred even more frequently in the L1 than in the L2. The lattermost result, as well as the overall result that more linguistic errors occurred in German than in English, can be primarily attributed to the translating task, i.e., many errors in German would likely have not occurred without the presence of an English source text on which it was based. It can be assumed that, due to the low level of translation competence among the subjects in this study, they lacked interference resistance and tended to display fixedness on the source text, phenomena which might not have occurred during a free text production task. Fixedness and interference thus represent disadvantages of translation tasks that must be taken into account when assessing text production competence in two languages on a contrastive basis using translation tasks as writing assignments. These disadvantages, however, may be overcome with an increasing degree of translation competence of the participants involved. Furthermore, it is important to note that during translation, the already-existing source text alleviates the cognitive demand of the task at hand because the content of the source text can be taken over into the target text. Because this decrease in cognitive demand can be assumed to free up working memory capacity that can then be dedicated to other text production processes (in line with McCutchen’s [1996] capacity theory of writing), this may also lead to target texts which, in certain aspects, are better in quality than those produced without a source text, as observed by Uzawa (1996; cf. Section 7.3). Aside from these methodological disadvantages for determining text production competence, translation tasks also have many advantages, especially when a product-oriented approach is combined with a process-oriented one. In many cases, process-oriented data make it possible to determine whether the subjects are aware of the errors and other shortcomings in their source texts, and, if so, whether they are able to improve upon these in their target texts. In the study presented here, for example, this was witnessed for implicitness and sense errors, indicating a greater degree of linguistic flexibility in their L1 and 19 On functional sentence perspective, see also Section 1.2.2.4. 258 Part V: L1 vs. L2 Writing a more restricted competence in their L2. In turn, this limited flexibility in expressing themselves in their L2 may prove to be harmful to the epistemic function of writing, as this limitation could lead to the fact that newly-formed knowledge is never fully - or only distortedly - expressed in written form, which is harmful to the development and refinement of knowledge. This study has also shown that translation can serve as a useful tool in the didactics of writing. It reduces task complexity on the macro-level, as the source text already provides the contents to be composed in verbalized form, allowing the students to pay greater attention to subtleties on the micro-level they might have otherwise ignored. On top of gaining a greater awareness of structural differences between their L1 and L2, students are also more sensitized to language-specific requirements such as those of sentence construction from a communicative standpoint. Specially tailored translation tasks can furthermore facilitate students’ awareness of language-specific coherencegenerating means; for example, having them translate using a source text in which connectors are systematically deleted would force them during their translation processes to think about the logical relations between two parts of a sentence or two sentences and how to express these appropriately in each language. In this way, students are prompted to express explicitly certain relationships in written form, which, during a free writing task, is significantly more difficult to establish and monitor. Despite the insights gained from this study, the limitation of its small sample size should be noted, which does not allow generalizations. The results, however, demonstrate the interesting insights that an analysis of text-level errors can bring to our knowledge about text composition competence in an L1 and L2, and lead us to the conclusion that a study with greater scope concentrating on the creation of text coherence in writing would bring even more promising insights into the development of writing competence. An especially interesting research question to explore would be whether or not writing competence correlates with the ability to receptively detect certain types of errors, and whether or not the quality of texts on a text-linguistic level, as was determined here by the errors on the text level, can serve as a reliable predictor of the general text production competence of a particular author. 9.6 L2 writing pedagogy From Cumming’s (1989) finding that writing expertise is language-independent once a certain minimum language proficiency has been achieved, the conclusion can be drawn that L2 writing instruction can follow the same recommendations that have been specified for L1 writing instruction in the previous chapter (Cumming 1989: 125). In addition, however, L2 language proficiency has to be fostered, a requirement that can also be formulated for many 9 Writing in the L1 vs. writing in the L2 259 students’ L1 language proficiency (Cumming 1989: 126). In L2 writing instruction, as in L1 writing instruction, emphasis has to be placed on meaning (cf. also the development of L2 writing instruction outlined in Hinkel 2002: 45 ff.). Diani (2012) suggests a combination of a genreand corpus-based approach to sensitize students for the specific requirements of academic writing. Genrebased lessons familiarize students with the typical structures of research papers and the “moves” in the sense of Swales (1990) as their basic elements. In corpus-based lessons students are then required to perform specified corpus analyses using tools such as WordSmith or the freeware AntConc and corpora such as the Michigan Corpus of Upper-level Student Papers (MICUSP) in order to find typical formulations with which moves and steps are realized (cf. also Macgilchrist 2010). Searches students are required to perform include finding out in what contexts personal pronouns such as I and we are used as well as which verbs of reporting are used in certain contexts, etc. A useful resource of phrases for academic writing is also the Academic Phrasebank (Morley 2014). Apart from the teaching methods that should be applied in L2 writing instruction, the question has to be addressed what norms academic texts should comply with. Do they have to comply with native-speaker standards and, if yes, of which variety of English? Or is communicative effectiveness for an international audience a requirement that has to be prioritized? As Björkman (2013: 201) points out, requirements formulated should comply with what is expected in the scientific community. Communicative effectiveness must be regarded as more important than perfectionism. A variable that has been found to have an impact on L2 writing instruction outcomes is the type of feedback provided for learners. On the basis of a research overview of studies investigating the impact of feedback in the form of reformulation and their own findings from a case study with two ESL learners with Mandarin as their L1, Qi & Lapkin (2001) recommend requiring students to compose texts in their L2, afterwards to confront them with an optimized reformulated version instead of just providing error correction, and finally to have them revise their original texts on the basis of this optimized version. Having to compose a text leads to the students noticing where they have problems, e.g., where they lack appropriate vocabulary. This noticing then raises their attentiveness to how these problems have been solved in the optimized versions and helps them to remember the respective linguistic means so that they can finally improve their original texts accordingly. They describe the advantages of providing reformulations over other types of feedback as follows: “For one thing, it [sc. reformulation] provides relevant TL structures to allow the learner to appropriate from them according to his/ her own needs and interests and to the context they themselves provided for a particular writing task. [...] Secondly, a text reformulated by a competent TL writer provides appropriate TL forms for 260 Part V: L1 vs. L2 Writing the given context. Without this, L2 learners are left to figure out the solutions themselves since they ‘have a smaller backlog of experience with English grammatical and rhetorical structure to fall back on, not having had the same exposure to those structures as native speakers have had (Leki, 1990; p. 58). Thirdly, this type of feedback provides a good balance between focus on form and focus on meaning since it exploits both the meaning-driven and the form-focused protential” (Qi/ Lapkin 2001: 282). Qi & Lapkin (2001: 294) also found, however, that the benefit students derived from the reformulated versions for their own textual improvements depended on whether they understood the reasons for the reformulated versions. Against the background of the students’ specific needs concerning linguistic accuracy, traditional peer tutoring approaches in writing centres with their maxim ‘higher-order concerns before lower-order concerns’ may need adaptation for L2 students (Williams/ Severino 2004). References Adler-Kassner, Linda/ Majewski, John/ Koshnick, Damian (2012): “The value of troublesome knowledge: Transfer and threshold concepts in writing and history.” Composition Forum 26 (2012). http: / / compositionforum.com/ issue/ 26/ troublesome-knowledge-threshold.php (last accessed: 17 August, 2014). Alexander, Patricia A. (2003): “The development of expertise: The journey from acclimation to proficiency.” Educational Researcher 32.8 (2013): 10-14. Alexander, Patricia A./ Jetton, Tamara L. (2003): “Learning from traditional and alternative texts: New conceptualizations for the information age.” Graesser, Arthur C./ Gernsbacher, Morton Ann/ Goldmann, Susan R., eds: Handbook of Discourse Processes . Mahwah (NJ): Lawrence Erlbaum, 199-241. Alverman, Donna E. (2002): “Effective literacy instruction for adolescents.” Journal of Literacy Research 34 (2002): 189-208. Anderson, John. R. (1976): Language, Memory and Thought . Hillsdale (NJ): Lawrence Erlbaum. — (1983): The Architecture of Cognition . Cambridge (MA): Harvard Univ. Press. — ( 3 1990): Cognitive Psychology and its Implications . New York: Freeman. Anson, Chris (2014): “When students teach teachers: Analysis of an assignment response experiment.” Paper presented at the EWCA Conference 2014 Let’s Peer across Borders . Frankfurt/ Oder, 19-22 July, 2014. Applebee, Arthur N. (1984): “Writing and reasoning.” Review of Educational Research 54.4 (1984): 577-596 (DOI: 10.3102/ 00346543054004577). Arbinger, Roland (1984): Gedächtnis . Darmstadt: Wiss. Buchgesellschaft. Arndt, Valerie (1987): “Six writers in search of texts: A protocol-based study of L1 and L2 writing.” ELT Journal 41.4 (1987): 257-267. Ausubel, David P. (1960): “The use of advance organizers in the learning and retention of meaningful verbal learning.” Journal of Educational Psychology 51 (1960): 267- 272. — (1968): Educational Psychology: A Cognitive View . New York, etc.: Holt, Rinehart and Winston. Baddeley, Alan D. (1986): Working Memory . Oxford: Oxford Univ. Press. —/ Lewis, Vivien J. (1981): “Inner active processing in reading: The inner voice, the inner ear and the inner eye.” Lesgold, Alan M./ Perfetti, Charles A., eds: Interactive Processes in Reading . Hillsdale (NJ): Lawrence Erlbaum, 107-129. Baker, Linda (1985): “How do we know when we don’t understand? Standards for evaluating text comprehension.” Forrest-Pressley, Donna-Lynn/ MacKinnon, G. E./ Waller, Thomas Gary, eds: Metacognition, Cognition, and Human Performance . Vol. 1 Theoretical Perspectives . Orlando, etc.: Academic Press, 155-205. — (1989): “Metacognition, comprehension monitoring, and the adult reader.” Educational Psychology Review 1.1 (1989): 3-38. 262 References Ballweg, Sandra, et al. (2013): „Schreibend lehren lernen.“ Bräuer/ Schindler (2013): 188-204. Bandura, Albert (1977): “Self-efficacy theory of behavioral change.” Psychological Review 84 (1977): 191-215. Banzer, Roman/ Kruse, Otto (2011): „Schreiben im Bachelor-Studium: Direktiven für Didaktik und Curriculumsentwicklung.“ G 4.8. Berendt, Brigitte, et al., eds: Neues Handbuch Hochschullehre . Berlin: Raabe, 1-37. Bartlett, Frederic Charles (1932): Remembering. A Study in Experimental and Social Psychology . Cambridge: Cambridge Univ. Press. Baurmann, Jürgen/ Ludwig, Otto (1985): „Texte überarbeiten. Zur Theorie und Praxis von Revisionen.” Boueke, Dietrich/ Hopster, Norbert, eds: Schreiben - Schreiben lernen. Rolf Sanner zum 65. Geburtstag . Tübingen: Narr, 254-276. Bayer-Hohenwarter, Gerrit (2012): Translatorische Kreativität: Definition, Messung, Entwicklung . (Studien zur Translation 8). Tübingen: Narr. Bazerman, Charles (1997): “The life of genre, the life in the classroom.” Bishop, Wendy/ Ostrom, Hans, eds: Genre and Writing: Issues, Arguments, Alternatives . Portsmouth (NH): Boynton/ Cook, 19-26. —, et al. (2005): Reference Guide to Writing Across the Curriculum . West Lafayette: Parlor. Bean, John C. (2001): Engaging Ideas. The Professor’s Guide to Integrating Writing, Critical Thinking, and Active Learning in the Classroom . San Francisco: John Wiley & Sons. Beaufort, Anne (2007): College Writing and Beyond: A New Framework for University Writing Instruction . Logan (UT): Utah State Univ. Press. — (2012): “College Writing and Beyond: Five years later.” Composition Forum 26 (2012). http: / / compositionforum.com/ issue/ 26/ college-writing-beyond.php (last accessed: 17 August, 2014). Beaugrande, Robert de/ Dressler, Wolfgang (1981): Introduction to Text Linguistics . London: Longman. Becker-Mrotzek, Michael (1992): „Wie entsteht eine Bedienungsanleitung? Eine empirisch-systematische Rekonstruktion des Schreibprozesses.” Krings/ Antos (1992): 257-280. Bereiter, Carl (1980): “Development in writing.” Gregg/ Steinberg (1980): 73-93. —/ Scardamalia, Marlene (1987): The Psychology of Written Composition . Hillsdale (NJ): Lawrence Erlbaum. —/ Scardamalia, Marlene (1993): Surpassing Ourselves: An Inquiry into the Nature and Implications of Expertise . Chicago: Open Court. Bergmann, Linda S./ Zepernick, Janet (2007): “Disciplinarity and transfer: Students’ perceptions of learning to write.” Writing Program Administration 31.1-2 (2007): 124-149. Berlyne, Daniel E. (1960): Conflict, Arousal, and Curiosity . New York: McGraw-Hill. — (1974): Konflikt, Erregung, Neugier . Stuttgart: Klett. References 263 Berninger, Virginia W./ Swanson, H. L. (1994): “Modifying Hayes and Flower’s model of skilled writing to explain beginning and developing writing.” Butterfield, Earl C., ed.: Children’s Writing: Towards a Process Theory of the Development of Skilled Writing . Vol. 2 of: Calson, Jerry S., ed.: Advances in Cognition and Educational Practice . Greenwich (CT): JAI Press, 57-81. —, et al. (1995): “Integrating lowand high-level skills in instructional protocols for writing disabilities.” Learning Disability Quarterly 18 (1995): 293-309. Biere, Bernd Ulrich (1989): Verständlich-Machen: Hermeneutische Tradition - Historische Praxis - Sprachtheoretische Begründung . (Reihe Germanistische Linguistik 92). Tübingen: Niemeyer. Bizzel, Patricia (1982): “Cognition, convention and certainty: what we need to know about writing.” PRE/ TEXT 3 (1982): 213-243. Björk, Lennart, et al. (2003a): “Teaching academic writing in European higher education: An introduction.” Björk, et al. (2003b): 1-15. —, et al., eds (2003b): Teaching Academic Writing in European Higher Education . (Studies in Writing 12). Dordrecht, Boston, London: Kluwer Academic Publishers. Björkmann, Beyza (2013): English as an Academic Lingua Franca: An Investigation of Form and Communicative Effectiveness . Bosten, Berlin: Walter de Gruyter. Blumenthal, Peter (1987): Sprachvergleich Deutsch-Französisch . (Romanische Arbeitshefte 29). Tübingen: Niemeyer. BMBF - Bundesministerium für Bildung und Forschung (2012): „Qualitätspakt Lehre - Die politische Zielsetzung.“ http: / / www.qualitaetspakt-lehre.de/ de/ 1294.php (last accessed: 17 August, 2014). Bohle, Ulrike (2013): „Schreibend Seminarinhalte vertiefen und reflektieren.“ Bräuer/ Schindler (2013): 176-187. Boone, Stephanie, et al. (2012): “Imagining a writing and rhetoric program based on principles of knowledge ‘transfer’: Dartmouth’s Institute for Writing and Rhetoric.” Composition Forum 26 (2012) http: / / compositionforum.com/ issue/ 26/ dartmouth.php (last accessed: 17 August, 2014). Bosher, Susan/ Rowenkamp, Jenise (1992): “Language Proficiency and Academic Success: The Refugee/ Immigrant in Higher Education.” (Eric Document ED 353 914). http: / / files.eric.ed.gov/ fulltext/ ED353914.pdf (last accessed: 17 August, 2014). Bot, Kees de/ Lowie, Wander/ Verspoor, Marjolijn (2007): “A Dynamic Systems Theory approach to second language acquisition.” Bilingualism: Language and Cognition 10.1 (2007): 7-21. Bourdin, Béatrice/ Fayol, Michel (1993): “Comparing speaking span and writing span: A working memory approach.” Paper presented at the Meeting of the European Association for Research in Learning and Instruction. Aix-en-Provence/ France, quoted in McCutchen (1996). —/ Fayol, Michel (1994): “Is written language production more difficult than oral language production? A working memory approach.” International Journal of Psychology 29 (1994): 591-620. 264 References Braaksma, Martine A. H./ Rijlaarsdam, Gert/ Van den Bergh, Huub (2002): “Observational learning and the effects of model-observer similarity.” Journal of Educational Psychology 94 (2002): 405-415. —, et al. (2001): “Effective learning activities in observation tasks when learning to write and read argumentative texts.” European Journal of Psychology of Education 16.1 (2001): 33-48. —, et al. (2004): “Observational learning and its effects on the orchestration of writing processes.” Cognition and Instruction 22.1 (2004): 1-36. — , et al. (2006): “What observational learning entails: A multiple case study.” L1 - Educational Studies in Language and Literature 6.1 (2006): 31-62. Bräuer, Gerd (1996): Warum Schreiben? Schreiben in den USA: Aspekte, Verbindungen, Tendenzen . Frankfurt/ M.: Lang. — (2000): Schreiben als reflexive Praxis: Tagebuch, Arbeitsjournal, Portfolio . Freiburg/ Breisgau: Fillibach. — (2003): “Centres for Writing & Reading - Bridging the gap between university and school education.” Björk, et al. (2003b): 135-150. — (2014): Das Portfolio als Reflexionsmedium für Lehrende und Studierende . (Kompetent lehren VI). Opladen, Toronto: Budrich. —/ Schindler, Kirsten (2013): „Authentische Schreibaufgaben - ein Konzept.“ Bräuer/ Schindler (2013): 12-63. —/ Schindler, Kirsten, eds (2013): Schreibarrangements für Schule, Hochschule und Beruf . Stuttgart: Fillibach bei Klett. Bransford, John D./ McCarrell, N. (1974): “A sketch of a cognitive approach to comprehension: Some thoughts about understanding what it means to comprehend.” Weimer, Walter B./ Palermo, David S., eds: Cognition and the Symbolic Process . New York, etc.: Lawrence Erlbaum, 189-229. Brinker, Klaus (1973): „Zum Textbegriff in der heutigen Linguistik.“ Sitta, Horst/ Brinker, Klaus, eds: Studien zur Texttheorie und zur deutschen Grammatik. Festgabe für Hans Glinz zum 60. Geburtstag . Düsseldorf: Schwann, 9-41. Bronfenbrenner, Urie/ Morris, Pamela A. (2006): “The bioecological model of human development.” Lerner, Richard M., ed.: Theoretical Models of Human Development . Vol. 1 of Lerner, Richard M./ Damon, William, eds: Handbook of Child Psychology . 6th ed. New York (NY): Wiley, 793-828. Brown, Ann L. (1987): “Metacognition, executive control, self-regulation, and other more mysterious mechanisms.” Weinert, Franz E./ Kluwe, Rainer H., eds: Metacognition, Motivation and Understanding . Hillsdale (NJ): Lawrence Erlbaum, 65- 116. Bruffee, Kenneth (1984): “Collaborative Learning and the ‘Conversation of Mankind’.” College English 46.7 (1984): 635-652. Büker, Stella (2003): “Teaching academic writing to international students: Individual tutoring as supplement to workshops.” Björk, et al. (2003b): 41-57. Byrne, Jody (2006): Technical Translation: Usability Strategies for Translating Technical Documentation . Dordrecht: Springer. References 265 Calfee, Robert C./ Greitz Miller, Roxanne (2007): “Best Practices in Writing Assessment.” Graham/ MacArthur/ Fitzgerald (2007): 265-286. Camp, Heather (2012): “The psychology of writing development - And its implications for assessment.” Assessing Writing 17.2 (2012): 92-105 (DOI: 10.1016/ j.asw.2012. 01.002). Camps, Anna/ Milian, Marta, eds (2000): Metalinguistic Activity in Learning to Write . (Studies in Writing 6). Amsterdam: Amsterdam Univ. Press. Canagarajah, Suresh (2002): “Multilingual writers and the academic community: Towards a critical relationship.” Journal of English for Academic Purposes 1.1 (2002): 29-44. Carey, Linda, et al. (1989): Differences in writers’ initial task representation . (Technical Report No. 35). Berkeley: Carnegy Mellon University, Center for the Study of Writing. Carson, Joan Eisterhold/ Kuehn, Phyllis A. (1992): “Evidence of transfer and loss in developing second language writers.” Language Learning 42.2 (1992): 157-182. Casanave, Christine (1998): “Transitions: The balancing act of bilingual academics.” Journal of Second Language Writing 7 (1998): 175-203. — (2004): Controversies in Second Language Writing: Dilemmas and Decisions in Research and Instruction . Ann Arbor: Univ. of Michigan Press. Chanquoy, Lucile (2001): “How to make it easier for children to revise their writing: A study of text revision from 3rd to 5th grades.” British Journal of Educational Psychology 71 (2001): 15-41. Chomsky, Noam (1969): Aspekte der Syntax-Theorie . Frankfurt: Suhrkamp. Christmann, Ursula (1989): Modelle der Textverarbeitung: Textbeschreibung als Textverstehen . (Arbeiten zur sozialwissenschaftlichen Psychologie 21). Münster: Aschendorff. Clark, Andy (1997): Being there. Putting Brain, Body, and World Together Again . Cambridge (MA): MIT Press. — (2008): Supersizing the Mind: Embodiment, Action, and Cognitive Extension . Oxford: Oxford University Press. —/ Chalmers, David J. (1998): “The Extended Mind.” Analysis 58.1 (1998): 7-19. Cohen, Andrew D./ Brooks-Carson, Amanda (2001): “Research on direct vs. translated writing: Students’ strategies and their results.” The Modern Language Journal 85 (2001): 169-188. Collins, Allan/ Brown, John Seely/ Larkin, Kathy M. (1980): “Inference in text understanding.” Spiro/ Bruce/ Brewer (1980): 385-407. Confais, Jean-Paul (1983): Grammaire explicative. Schwerpunkte der französischen Grammatik für Leistungskurs und Studium . 2nd ed. München: Hueber. Connor, Ulla (1996): Contrastive Rhetoric . Cambridge: Cambridge Univ. Press. Cooper, Charles R./ Matsuhashi, Ann (1983): “A theory of the writing process.” Martlew, Margret, ed.: The Psychology of Written Language . Chichester: Wiley, 3- 39. 266 References Cope, Bill/ Kalantzis, eds (2000): Multiliteracies: Literacy Learning and the Design of Social Futures . London, New York: Routledge. Couzijn, Michel (1995): Observation of Writing and Reading Activities. Effects on Learning and Transfer . Amsterdam: University of Amsterdam. Dordrecht: Dorfix [ PhD Dissertation ] . — (1999): “Learning to write by observation of writing and reading processes: Effects on learning and transfer.” Learning and Instruction 9.2 (1999): 109-142. Craig, Jennifer Lynn (2013): Integrating Writing Strategies in EFL/ ESL University Contexts: A Writing-Across-the-Curriculum Approach . New York, London: Routledge. Craik, Kenneth J. (1943): The Nature of Explanation . Cambridge: Cambridge Univ. Press. Crusan, Deborah (2010): “Assess thyself lest others assess thee.” Silva/ Matsuda (2010): 245-262. Cumming, Alister (1989): “Writing expertise and second-language proficiency.” Language Learning 39.1 (1989): 81-141. — (2001): “Learning to write in a second language: Two decades of research.” International Journal of English Studies 1.2 (2001): 1-23. Cummins, Jim (1981): “The role of primary language development in promoting educational success for language minority students.” Schooling and Language Minority Students: A Theoretical Framework. Los Angeles: California State Univ., Evaluation, Dissemination and Assessment Center, 3-49, quoted in Carson/ Kuehn (1992). Daneš, František (1974): “Functional sentence perspective and the organization of the text.” Daneš, František, ed.: Papers on Functional Sentence Perspective . Prag, Den Haag, Paris: Mouton & Co., 106-128. Deutsches Diabetes-Zentrum (DDZ) (2004): „Was ist Diabetes? - Grundlegende Informationen zum Diabetes.“ DDZ, ed.: Informationssystem zum Diabetes mellitus . http: / / www.diabetes.uni-duesseldorf.de/ wasistdiabetes/ (last accessed: 30 July, 2005). Devine, Joanne/ Railey, Kevin/ Bischoff, Philip (1993): “The implications of cognitive models in L1 and L2 writing.” Journal of Second Language Writing 2.3 (1993): 203- 225. Devitt, Amy J. (2008): Writing Genres . Carbondale (IL): Southern Illinois Univ. Press. Diani, Giuliana (2012): “Text and corpus work, EAP writing and language learners.” Tang (2012): 45-66. van Dijk, Teun A. (1977a): Text and Context. Explorations in the Semantics and Pragmatics of Discourse . (Longman Linguistic Library 21). London, New York: Longman. — (1977b): “Semantic macrostructures and knowledge frames in discourse comprehension.” Just/ Carpenter (1977): 3-32. — (1980a): Macrostructures. An Interdisciplinary Study of Global Structures in Discourse, Interaction, and Cognition . Hillsdale (NJ): Lawrence Erlbaum. — (1980b): Textwissenschaft. Eine interdisziplinäre Einführung . Translated by Christoph Sauer. Tübingen: Niemeyer. References 267 —/ Kintsch, Walter (1983): Strategies of Discourse Comprehension . New York: Academic Press. Doherty, Monika (2000): “Optimizing parsing and discourse linking by clefts.” Schmitt, Peter A., ed.: Paradigmenwechsel in der Translation. Festschrift für Albrecht Neubert zum 70. Geburtstag . Tübingen: Stauffenburg, 41-55. Downs, Douglas/ Wardle, Elizabeth (2007): “Teaching about writing, righting misconceptions: (Re)Envisioning ‘First-Year Composition’ as introduction to writing studies.” College Composition and Communication 58.4 (2007): 552-584. Driscoll, Dana Lynn (2009): Pedagogy of Transfer: Impacts of Student and Faculty Attitudes . Purdue University [Dissertation], quoted in Driscoll/ Wells (2012). — (2011): “Connected, disconnected, or uncertain: Student attitudes about future writing contexts and perceptions of transfer from first year writing to the disciplines.” Across the Disciplines 8.2 (2011). http: / / web.ebscohost.com/ ehost/ detail? vid=3&sid=d8abad2b-c886-4570-9ab5-1fa6f3c00f96%40sessionmgr4003 & hid4209&bdata=JnNpdGU9ZWhvc3QtbGl2ZQ%3d%3d#db=ehh&AN=90461232 (last accessed: 17 August, 2014). —/ Wells, Jennifer (2012): “Beyond knowledge and skills: Writing transfer and the role of student dispositions.” Composition Forum 26. http: / / compositionorum.com/ ssue/ 26/ beyond-knowledge-skills.php (last accessed: 17 August, 2014). Duchowski, Andrew T. (2003): Eye Tracking Methodology: Theory and Practice . London: Springer. Duncker, Karl (1945): “On problem-solving.” Psychological Monographs 58.5 (1945): 1-114. Dweck, Carol S. (1986): “Motivational processes affecting learning.” American Psychologist 41 (1986): 1040-1048. Dysthe, Olga (2003): “Writing at Norwegian universities in an international perspective.” Björk, et al. (2003b): 151-164. Eccles, Jacquelynne S. (2005): “Studying the development of learning task motivation.” Learning and Instruction 15.2 (2005): 161-171. Ehlich, Konrad (1999): „Alltägliche Wissenschaftssprache.“ Barkowski, Hans/ Wolff, Arnim, eds: Alternative Vermittlungsmethoden und Lernformen auf dem Prüfstand . Regensburg: Fachverband DaF, 1-30. — (2003): „Universitäre Textarten, universitäre Struktur.“ Ehlich/ Streets (2003): 13- 28. —/ Streets, Angelika, eds (2003): Wissenschaftlich schreiben - lehren und lernen . Berlin, New York: de Gruyter. Elbow, Peter (1998): Writing without Teachers . New York, Oxford: Oxford Univ. Press. Englund Dimitrova, Birgitta (2005): Expertise and Explicitation in the Translation Process . (Benjamins Translation Library 64). Amsterdam, Philadelphia: John Benjamins. Ericsson, K. Anders/ Charness, Neil (1997): “Cognitive and developmental factors in expert performance.” Feltovich/ Ford/ Hoffmann (1997): 3-41. 268 References —/ Kintsch, Walter (1995). “Long-term working memory.” Psychological Review 102 (1995): 211-245. —/ Krampe, Ralf Th./ Tesch-Römer, Clemens (1993): “The role of deliberate practice in the acquisition of expert performance.“ Psychological Review 100 (1993): 363-406. —/ Simon, Herbert A. ([1984] 1993): Protocol Analysis: Verbal Reports as Data . Rev. Ed. Cambridge (MA), London (England): MIT Press. —/ Smith, Jacqui (1991): “Prospects and limits of the empirical study of expertise: An introduction.” Ericsson, K. Anders/ Smith, Jacqui, eds: Towards a General Theory of Expertise. Prospects and Limits . Cambridge: Cambridge Univ. Press, 1-38. Esser, Jürgen (2009): Introduction to English Textlinguistics . (TELL 2). Frankfurt/ M.: Lang. European Commission Translation Service (n.y.): Fight the Fog: Write Clearly . http: / / www.eurosfaire.prd.fr/ bibliotheque/ pdf/ Write_EN_Fight-the-fog.pdf (last accessed: 19 October, 2014). Fabbro, Franco/ Paradis, Michel (1995): “Differential impairments in four multilingual patients with subcortical lesions.” Paradis, Michel, ed.: Aspects of Bilingual Aphasia . Oxford (UK): Pergamon Press, 139-176. —/ Peru, Andrea/ Skrap, Miran (1997): “Language disorder in bilingual patients after thalmic lesions.” Journal of Neurolinguistics 10 (1997): 347-367. Faerch, Claus/ Kasper, Gabriele (1983): “Plans and strategies in foreign language learning.” Faerch, Claus/ Kasper, Gabriele, eds: Strategies in Interlanguage Communication . London: Longman, 20-60. Feilke, Helmut/ Lehnen, Katrin, eds (2012): Schreib- und Textroutinen. Theorie, Erwerb und didaktisch-mediale Modellierung . Frankfurt/ M. etc.: Lang. Feltovich, Paul J./ Ford, Kenneth M./ Hoffmann, Robert R., eds (1997): Expertise in Context. Human and Machine . Melo Park, Cambridge (MA), London: AAAI Press/ MIT Press. Fillmore, Charles (1968): “The case for case.” Bach, Emmon/ Harms, Robert T., eds: Universals in Linguistic Theory . New York: Holt, Rinehatt & Winston, 1-88. Findler, Nicholas V., ed. (1979): Associative Networks: Representation and the Use of Knowledge by Computers . New York: Academic Press. Flavell, John H. (1976): “Metacognitive aspects of problem solving.” Resnick, Lauren B., ed.: The Nature of Intelligence. Hillsdale (NJ): Lawrence Erlbaum, 231-135. Flesch, Rudolf A. (1948): “A new readability yardstick.” Journal of Applied Psychology 32.2 (1948): 221-233. Flower, Linda (1979): “Writer-based prose: A cognitive basis for problems in writing.” College English 41.1 (1979): 19-37. —/ Hayes, John R. (1981): “A cognitive process theory of writing.” College Composition and Communication 39 (1981): 365-387. —, et al. (1986): “Detection, diagnosis and strategies of revision.” College Composition Communication 37 (1986): 16-55. Flowerdew, Lynne (2000): “Using a genre-based framework to teach organizational structure in academic writing.” ELT Journal 54 (2000): 369-378. References 269 Frank, Andrea/ Haacke, Stefanie/ Tente, Christina (2003): “Contacts - Conflicts - Cooperation: A report form the writing lab of the University of Bielefeld.” Björk, et al. (2003b): 165-174. Friedlander, Alexander (1990): “Composing in English: Effects of a first language on writing in English as a second language.” Kroll (1990): 109-125. Fries, Charles C. (1945): Teaching and Learning English as a Foreign Language . Ann Arbor: The Michigan Univ. Press. Gage, A. R. (1986): “Why write? ” Bartholomea, David/ Petrosky, Anthony R., eds: The Teaching of Writing . Chicago: National Society for the Study of Education, 8-29. Galbraith, David (1999): “Writing as a knowledge-constituting process.” Torrance, Mark/ Galbraith, David, eds: Studies in Writing. Vol. 4. Knowing What to Write: Conceptual Processes in Text Production . Amsterdam: Amsterdam Univ. Press, 139-160. Garrett, Merrill F. (1975): “The analysis of sentence production.” Bower, Gordon, ed.: The Psychology of Learning and Motivation . Vol. 9. New York: Academic Press, 133-177. — (1976): “Syntactic processes in sentence production.” Wales, J. R./ Walker Edward, eds: New Approaches to Language Mechanisms . Amsterdam: North-Holland, 231- 256. — (1980): “Levels of processing in sentence production.” Butterworth, Brian, ed.: Language Production. Speech and Talk . Vol. 1. London: Academic Press, 177-220. — (1982a): “Production of speech: Observations from normal and pathological language use.” Ellis, Andrew W., ed.: Normality and Pathology in Cognitive Functions . London: Academic Press, 19-76. — (1982b): “A perspective on research in language.” Mehler, Jacques/ Walker, E. C. T./ Garret, Merrill F., eds: Perspectives on Mental Representations . Hillsdale (NJ): Lawrence Erlbaum, 185-199. — (1988): “Processes in language production.” Newmeyer, Frederick, ed.: Psychological and Biological Aspects . Vol. III of: Linguistics: The Cambridge Survey . Cambridge: Cambridge Univ. Press, 69-96. Gelder, Tim Van (1989): “The dynamical hypothesis in cognitive science.” Behavioral and Brain Sciences 21 (1989): 615-656. Gelman, Rochel/ Greeno, James G. (1989): “On the nature of competence: Principles for understanding in a domain.” Resnick, Lauren G., eds: Knowing, Learning, and Instruction: Essays in Honor of Robert Glaser . Hillsdale (NJ): Erlbaum, 125-186. Girgensohn, Katrin/ Sennewald, Nadja (2012): Schreiben lehren, Schreiben lernen: Eine Einführung . Darmstadt: WBG. Göpferich, Susanne (1995): Textsorten in Naturwissenschaften und Technik. Pragmatische Typologie - Kontrastierung - Translation . (Forum für Fachsprachen-Forschung 27). Tübingen: Narr. — (1998): Interkulturelles Technical Writing: Fachliches adressatengerecht vermitteln. Ein Lehr- und Arbeitsbuch . (Forum für Fachsprachen-Forschung 40). Tübingen: Narr. 270 References — (2001): „Von Hamburg nach Karlsruhe: Ein kommunikationsorientierter Bezugsrahmen zur Bewertung der Verständlichkeit von Texten.“ Fachsprache/ International Journal of LSP 23.3-4 (2001): 117-138. — (2002): Textproduktion im Zeitalter der Globalisierung. Entwicklung einer Didaktik des Wissenstransfers . (Studien zur Translation 15). Tübingen: Stauffenburg. — (2005): Transkripte zur Optimierung des populärwissenschaftlichen Textes „Definition des Diabetes mellitus“ mit lautem Denken und Translog. http: / / www.susannegoepferich.de/ Anhang_Transkripte.pdf (last accessed: 17 August, 2014). — (2006a): “How successful is the mediation of specialized knowledge? - The use of thinking-aloud protocols and log files of reverbalization processes as a method in comprehensibility research.” HERMES 37 (2006): 67-93. — (2006b): “How comprehensible are popular science texts? The use of thinking-aloud protocols and log files of reverbalization processes in comprehensibility research.” Gotti, Maurizio/ Giannoni Davide Simone, eds: Trends in Specialized Discourse Analysis . Frankfurt/ M.: Lang, 221-246. — (2006c): “Popularization from a cognitive perspective - What thinking aloud and log files reveal about optimizing reverbalization processes.” Fachsprache/ International Journal of LSP 3-4 (2006): 128-154. — (2008): Translationsprozessforschung: Stand - Methoden - Perspektiven . (Studien zur Translation 4). Tübingen: Narr. — (2009): “Comprehensibility assessment using the Karlsruhe comprehensibility concept.” JosTrans - The Journal of Specialized Translation 11 (2009). — (2010): “Data documentation and data accessibility in translation process research.” The Translator 16.1 (2010): 93-124. — (2013): “Translation competence: Explaining development and stagnation from a Dynamic Systems Perspective.“ Ehrensberger-Dow, Maureen/ Göpferich, Susanne/ O’Brien, Sharon, eds: Interdisciplinarity in Translation and Interpreting Process Research . Special Issue of TARGET: International Journal of Translation Studies 25.1 (2013): 61-76. — (2014): “Methods of measuring students’ text production competence and its development in writing courses.“ Knorr, Dagmar/ Heine, Carmen/ Engberg, Jan, eds: Methods in Writing Process Research . Frankfurt/ M., etc.: Lang, 187-209. —/ Nelezen, Bridgit (2012): Data documentation for the article „Die Sprach(un)abhängigkeit von Textproduktionskompetenz: Translation als Werkzeug der Schreibprozessforschung und Schreibdidaktik“. http: / / www.susanne-goepferich.de/ Data_ Documentation_Writing_L1_L2 (last accessed: 17 August, 2014). —/ Nelezen, Bridgit (2013): „Die Sprach(un)abhängigkeit von Textproduktionskompetenz: Translation als Werkzeug der Schreibprozessforschung und Schreibdidaktik.” ZfAL - Zeitschrift für Angewandte Linguistik 58.1 (2013): 167-200. —/ Nelezen, Brigit (2014): “The language-(in)dependence of writing skills: Translation as a tool in writing process research and writing instruction.” Muñoz Martín, Ricardo, ed.: Minding Translation - Con la traducción en mente . Special Issue of MONTI 1 (2014): 117-149. References 271 Gottschalk, Katherine/ Hjorthoj, Keith (2004): The Elements of Teaching Writing. A Resource for Instructors in All Disciplines . New York: Bedford/ St. Martin’s. Gould, John D. (1980): “Experiments on composing letters: Some facts, some myths, and some observations.” Gregg/ Steinberg (1980): 97-128. Graesser, Arthur C. (1978): “How to catch a fish: The representation and memory of common procedures.” Discourse Processes 1 (1978): 72-89. Graham, Steve/ Harris, Karen R. (1994): “The role and development of self-regulation in the writing process.” Schunk, Dale H./ Zimmermann, Barry J., eds: Self-regulation of Learning and Performance: Issues and Educational Applications . Hillsdale (NJ): Lawrence Erlbaum, 203-228. —/ MacArthur, Charles A. (1988): “Improving learning disabled students’ skills at revising essays produced on a word processor: Self-instructional strategy training.” Journal of Special Education 22.2 (1988): 133-152. —/ MacArthur, Charles A./ Fitzgerald, Jill (2007): “Introduction Best Practices in Writing Instruction Now.” Graham/ MacArthur/ Fitzgerald (2007): 1-9. —/ MacArthur, Charles A./ Fitzgerald, Jill, eds (2007): Best Practices in Writing Instruction . New York, London: The Guildford Press. —/ Perin, Dolores (2007): “A meta-analysis of writing instruction for adolescent students.” Journal of Educational Psychology 99 (2007): 445-476. Graves, Michael F. (2004): “Theories and constructs that have made a significant difference in adolescent literacy.” Jetton/ Dole (2004): 433-452. Greenbaum, Sidney/ Quirk, Randolph (1999): A Student’s Grammar of the English Language . London: Longman. Gregg, Lee W./ Steinberg, Erwin R., eds (1980): Cognitive Processes in Writing . Hillsdale (NJ): Lawrence Erlbaum. Greimas, Algirdas Julien (1966): Sémantique structurale. Recherche de méthode . Paris: Larousse. Grieshammer, Ella/ Liebetanz, Franziska/ Peters, Nora/ Zegenhagen, Jana (2013): Zukunftsmodell Schreibberatung: Eine Anleitung zur Begleitung von Schreibenden im Studium . 2nd corr. ed. Baltmannsweiler: Schneider Verlag Hohengehren. Grimm, Nancy M. (1996): “Rearticulating the work of the writing centre.” College Composition and Communication 47 (1996): 485-523. Groeben, Norbert (1978): Die Verständlichkeit von Unterrichtstexten . 2nd, extended ed. Münster: Aschendorff. — (1982): Leserpsychologie: Textverständnis - Textverständlichkeit . Münster: Aschendorff. —/ Christmann, Ursula (1989): „Textoptimierung unter Verständlichkeitsperspektive.” Antos, Gerd/ Krings, Hans P., eds (1989): Textproduktion. Ein interdisziplinärer Forschungsüberblick. (Konzepte der Sprach- und Literaturwissenschaft 48). Tübingen: Niemeyer, 165-196. Günther, Udo (1993): Texte planen - Texte produzieren. Kognitive Prozesse der schriftlichen Textproduktion . (Psycholinguistische Studien). Opladen: Westdeutscher Verlag. 272 References Haberland, Hartmut/ Mortensen, Janus (2012): “Language variety, language hierarchy and language choice in the international university.” International Journal of the Sociology of Language 216 (2012): 1-6. Halliday, M. A. K./ Hasan, Ruqaiya (1976): Cohesion in English . London: Longman. Hamp-Lyons, Liz, ed. (1991): Assessing Second Language Writing in Academic Contexts . Westport (CT), London: Ablex Publishing. Hand, Brian (1999): “A writing in science framework designed to enhance science literacy.” International Journal of Science Education 21.10 (1999): 1021-1035. —/ Prain, Vaughan (2002): “Teachers implementing writing-to-learn strategies in junior secondary science: A case study.” Science Education 86.6 (2002): 737-755. Hansen, Gyde (2006): Erfolgreich übersetzen: Entdecken und Beheben von Störquellen . (Translationswissenschaft 2). Tübingen: Narr. Hanstein, Volker (1993): „Informationsverarbeitung aus Texten: Kognitionspsychologisches Modell als Hilfe bei der Textproduktion.” tekom-Nachrichten 15.2 (1993): 36-42. Harris, Karen/ Graham, Steve (1994): “Constructivism: Principles, paradigms, and integration.” The Journal of Special Education 28 (1994): 233-247. —/ Graham, Steve (1996): Making the Writing Process Work: Strategies for Composition and Self-regulation. 2nd ed. Cambridge: Brookline Books. Hartmann, Peter (1971): „Texte als linguistisches Objekt.“ Stempel, Wolf-Dieter, ed.: Beiträge zur Textlinguistik . München: Fink, 9-29. Hartmann, Reinhard R. K./ Storck, F. C. (1972): Dictionary of Language and Linguistics . Amsterdam: Applied Science, quoted in Liu (2009). Hayes, John R. (1996): “A new framework for understanding cognition and affect in writing.” Levy/ Ransdell (1996): 1-27. — (2004): “What triggers revision? ” Allal, Linda K./ Chanquoy, Lucile/ Largy, Pierre, eds: Revision of Written Language: Cognitive and Instructional Processes . Boston, Dordrecht, New York: Kluwer, 9-20. —/ Flower, Linda S. (1980): “Identifying the organization of writing processes.” Gregg/ Steinberg (1980): 3-30. —/ Nash, Jane Gradwohl (1996): “On the nature of planning in writing.” Levy/ Ransdell (1996): 29-55. —, et al. (1986): “If it’s clear to me, it must be clear to them.” Paper presented at the Conference College Composition and Communication . New Orleans (LA), quoted in Levy/ Ransdell (1996). Hein-Becker, Cathrin (2013): Connector Usage in Essays Written by German Learners of English: A Corpus-based Investigation . Bachelor Thesis for the program “Modern Languages, Cultures and Business Studies”. Giessen: Justus Liebig University, Department of English. Heinemann, Wolfgang/ Viehweger, Dieter (1991): Textlinguistik - Eine Einführung . (Reihe Germanistische Linguistik 115): Tübingen: Niemeyer. Heringer, Hans Jürgen (1979): „Verständlichkeit - ein genuiner Forschungsbereich der Linguistik? “ Zeitschrift für germanistische Linguistik 7 (1979): 255-278. References 273 — (1984): „Textverständlichkeit. Leitsätze und Leitfragen.“ LiLi Zeitschrift für Literaturwissenschaft und Linguistik 14.55 (1984): 57-70. Herrmann, Theo (1985): Allgemeine Sprachpsychologie. Grundlagen und Probleme . München, Wien, Baltimore: Urban und Schwarzenberg. Hillocks, George Jr. (1986): Research on Written Composition. New Directions for Teaching . Urbana (IL): National Conference on Research in English/ ERIC Clearinghouse on Reading and Communication Skills. — (1995): Teaching Writing as Reflective Practice . New York, London: Teacher College Press. Hinkel, Eli (2002): Second Language Writers’ Text: Linguistic and Rhetorical Features. London, New York: Routledge. Hirose, Keiko (2003): “Comparing L1 and L2 organizational patterns in the argumentative writing of Japanese EFL students.” Journal of Second Language Writing 12 (2003): 181-209. —/ Sasaki, Miyuki (1994): “Explanatory variables for Japanese students’ expository writing in English: An exploratory study.” Journal of Second Language Writing 3.3 (1994): 203-229. Hönig, Hans G. (1995): Konstruktives Übersetzen . (Studien zur Translation 1). Tübingen: Stauffenburg. —/ Kußmaul, Paul (1984): Strategie der Übersetzung. Ein Lehr- und Arbeitsbuch . (Tübinger Beiträge zur Linguistik 205). 2nd ed. Tübingen: Narr. Hörmann, Hans (1976): Meinen und Verstehen. Grundzüge einer psychologischen Semantik . Frankfurt/ M.: Suhrkamp. Holz-Mänttäti, Justa (1984): Translatorisches Handeln - Theorie und Methode . Helsinki: Suomalainen Tiedeakatemia. Hopf, Jennifer (2013): Solving Problems when Translating Operating Instructions: The Behavior of Professional Translators and Translation Students Compared. Master Thesis for the Program “Modern Languages, Cultures and Business”. Giessen: Justus Liebig University, Department of English. Hoppe-Graff, Siegfried (1984): „Verstehen als kognitiver Prozeß. Psychologische Ansätze und Beiträge zum Textverstehen.“ LiLi Zeitschrift für Literaturwissenschaft und Linguistik 14.55 (1984): 10-37. Hounsell, Dai (1988): “Towards an anatomy of academic discourse: meaning and context in the undergraduate essay.” Säljö, Roger, ed.: The Written World. Studies in Literate Thought and Action . Berlin: Springer, 161-177. Howard, James (1983): Writing to Learn . Washington (DC): Council of Basic Education. Hu, Guangwei/ Ren, Hongwei (2012): “The impact of experience and beliefs on Chinese EFL students writers’ feedback preferences: ” Tang (2012): 67-87. Hutchins, Edwin (1995): Cognition in the Wild . Cambridge: MIT. Isenberg, Horst (1977): „Text versus Satz.“ Daneš, František/ Viehweger, Dieter, eds: Probleme der Textgrammatik II . (studia grammatica XVIII). Berlin: Akademie-Verlag, 119-146. 274 References Jäger, Gert/ Müller, Dietrich (1982): „Kommunikative und maximale Äquivalenz von Texten.“ Jäger, Gert/ Neubert, Albrecht, eds: Äquivalenz der Translation . Leipzig: Verlag Enzyklopädie, 42-57. Jahr, Silke (1996): Das Verstehen von Fachtexten. Rezeption - Kognition - Applikation . (Forum für Fachsprachen-Forschung 34). Tübingen: Narr. Jakobs, Eva-Maria (1999): „Normen der Textgestaltung.“ Kruse/ Jakobs/ Ruhmann (1999): 171-190. Jakobson, Arnt Lykke (1999): “Logging target text production with Translog.“ Hansen, Gyde, ed.: Probing the Process in Translation: Methods and Results . (Copenhagen Studies in Language 24). Copenhagen: Samfundslitteratur, 9-20. Janich, Nina, ed. (2008): Textlinguistik. 15 Einführungen . (Narr Studienbücher). Tübingen: Narr. Jechle, Thomas (1992): Kommunikatives Schreiben. Prozeß und Entwicklung aus der Sicht kognitiver Schreibforschung . Tübingen: Narr. Jetton, Tamara L. (1994): Teachers’ and Students’ Understanding of Scientific Exposition: How Importance and Interest Influence what is Accessed and what is Discussed. Texas A&M University, College Station [Unpublished doctoral dissertation], quoted in Alexander/ Jetton (2003). —/ Alexander, Patricia A. (1997): “Instructional importance: What teachers value and what students learn.” Reading Research Quarterly 32 (1997): 290-308. —/ Alexander, Patricia A. (2004): “Domains, teaching, and literacy.” Jetton/ Dole (2004): 15-39. —/ Dole, Janice A., eds (2004): Adolescent Literacy Research and Practice . New York: Guilford. Johnson-Laird, Philip N. (1985): Mental Models. Towards a Cognitive Science of Language, Inference, and Consciousness . Cambridge (MA), etc.: Harvard Univ. Press. Johnston, Peter H./ Winograd Peter N. (1985): “Passive failure in reading.” Journal of Reading Behavior 17.4 (1985): 279-301. Johnstone, Karla M./ Ashbaugh, Hollis/ Warfield, Terry D. (2002): “Effects of repeated practice and contextual-writing experiences on college students’ writing skills.” Journal of Educational Psychology 94 (2002): 305-315. Just, Marcel A./ Carpenter, Patricia A., eds (1977): Cognitive Processes in Comprehension . Hillsdale (NJ): Lawrence Erlbaum. —/ Carpenter, Patricia A. (1992): “A capacity theory of comprehension: Individual differences in working memory.” Psychological Review 99 (1992): 122-149. Kaplan, Robert B. (1966): “Cultural thought patterns in intercultural education.” Language Learning 16 (1966): 1-20. — (1967): “Contrastive rhetoric and teaching of composition.” TESOL Quarterly 1 (1967): 10-16. — (1980): “Cultural thought patterns in intercultural education.” Croft, Kenneth, ed.: Readings on English as a Second Language . 2nd ed. Boston, Toronto: Little Brown and Company, 399-418. References 275 Karmiloff-Smith, Annette (1993): Beyond Modularity. A Developmental Perspective on Cognitive Science . Cambridge (MA), London: MIT Press. Kellogg, Ronald T. (1990): “Effectiveness of prewriting strategies as a function of task demands.” American Journal of Psychology 103 (1990): 327-342. — (2001): “Long-term working memory in text production.” Memory & Cognition 29 (2001): 43-52. — (2006): “Professional writing expertise.” Ericsson, K. Anders, et al., eds: The Cambridge Handbook of Expertise and Expert Performance . New York: Cambridge Univ. Press, 389-402. — (2008): “Training writing skills: A cognitive developmental perspective.” Journal of Writing Research 1.1 (2008): 1-26. Kember, David (1997): “A reconceptualisation of the research into university academics’ conceptions of teaching.” Learning and Instruction 7.3 (1997): 255-275. Keseling, Gisbert/ Wrobel, Arne/ Rau, Cornelia (1987): „Globale und lokale Planung beim Schreiben.” Unterrichtswissenschaft 15 (1987): 349-365. Kim, Eun-Young (2011): “Using translation exercises in the communicative EFL writing classroom.” ELT Journal 65.2 (2011): 154-160. Kim, Karl H. S., et al. (1997): “Distinct cortical areas associated with native and second languages.” Nature 388 (1997): 171-174. Kintsch, Walter (1974): The Representation of Meaning in Memory . Hillsdale (NJ): Lawrence Erlbaum. —/ van Dijk, Teun A. (1978): “Toward a model of text comprehension and production.” Psychological Review 85.5 (1978): 363-394. —/ Green, E. (1978): “The role of culture-specific schemata in the comprehension and recall of stories.” Discourse Processes 1 (1978): 1-13. —/ Keenan, J. M. (1973): “Reading rate and retention as a function of the number of propositions in the base structure of sentences.” Cognitive Psychology 5 (1973): 257-274. —, et al. (1975): “Comprehension and recall of text as a function of content variables.” Journal of Verbal Learning and Verbal Behavior 14.2 (1975): 196-214. Klaassen, Renate (2001): The International University Curriculum: Challenges in English-medium engineering education . Delft: Delft University of Technology [Unpublished PhD Thesis], quoted in Björkman (2013). Klare, George R. (1963): The Measurement of Readability . Ames (IA): Iowa State Univ. Press. Knorr, Dagmar/ Jakobs, Eva-Maria, eds (1997): Textproduktion in elektronischen Umgebungen . (Textproduktion und Medium 2). Frankfurt/ M. etc.: Lang. Kobayashi, Hiroe/ Rinnert, Carol (1992): “Effects of first language on second language writing: Translation vs. direct composition.” Language Learning 42.2 (1992): 183- 215. Kösler, Bertram (1992): Gebrauchsanleitungen richtig und sicher gestalten. Forschungsergebnisse für die Gestaltung von Gebrauchsanleitungen . 2nd ed. Wiesbaden: Forkel. 276 References Kohro, Yoshifumi (2009): “A contrastive study between L1 and L2 compositions: Focusing on global text structure, composition quality, and variables in L2 writing.” Dialogue 8 (2009): 1-19. Koller, Werner (1992): Einführung in die Übersetzungswissenschaft . (UTB 819). 4th ed. Heidelberg, Wiesbaden: Quelle & Meyer. Kramer, Femke/ van Kruiningen, Jacqueline/ Padmos, Henrike (2003): “Creating a basis for a faculty-oriented writing programme.” Björk, et al. (2003b): 185-194. Krapels, Alexandra Rowe (1990): “An overview of second language writing process research.” Kroll (1990): 37-56. Krause, Michael U. (1991): „Zur Nützlichkeit des Hamburger Verständlichkeitsmodells bei technischen Anleitungen.” Feldbusch, Elisabeth/ Pogarell, Reiner/ Weiß, Cornelia, eds: Neue Fragen der Linguistik. Akten des 25. Linguistischen Kolloquiums, Paderborn 1990 . Vol. 2: Innovation und Anwendung . (Linguistische Arbeiten 271). Tübingen: Niemeyer, 393-398. Krings, Hans P. (1986): Was in den Köpfen von Übersetzern vorgeht. Eine empirische Untersuchung zur Struktur des Übersetzungsprozesses bei fortgeschrittenen Französischlernern. Tübingen: Narr. — (2005): „Wege ins Labyrinth - Fragestellungen und Methoden der Übersetzungsprozessforschung im Überblick.” Meta 50.2 (2005): 342-358. —/ Antos, Gerd, eds (1992): Textproduktion: Neue Wege der Forschung . (Fokus Linguistisch-philologische Studien 7). Trier: Wissenschaftlicher Verlag Trier. Kröber, Edith (2010): Wirksamkeit hochschuldidaktischer Weiterbildung am Beispiel der Veränderung von Lehrkonzeptionen. Eine Evaluationsstudie . Dortmund: TU Dortmund [ PhD Dissertation ] . Kroll, Barbara, ed. (1990): Second Language Writing: Research Insights for the Classroom . New York: Cambridge Univ. Press. Kruse, Otto (2003): “Getting started: Academic writing in the first year of a university education.” Björk, et al. (2003b): 9-28. —/ Chitez, Madalina (2012): „Schreibkompetenz im Studium: Komponenten, Modelle und Assessment.“ Preußer/ Sennewald (2012): 57-83. —/ Jakobs, Eva-Maria (1999): „Schreiben lehren an der Hochschule. Ein Überblick.“ Kruse/ Jakobs/ Ruhmann (1999): 19-34. —/ Jakobs, Eva-Maria/ Ruhmann, Gabriela, eds (1999): Schlüsselkompetenz Schreiben. Konzepte, Methoden, Projekte für Schreibberatung und Schreibdidaktik an der Hochschule . Neuwied: Luchterhand. Kuhn, Deanna (1999): “A developmental model of critical thinking.” Educational Researcher 28.2 (1999): 16-25. Kuhn, Thomas (1970): The Structure of Scientific Revolutions . 2nd ed. Chicago: Univ. of Chicago Press. Lange, Ulrike (2012): „Strategien für das wissenschaftliche Schreiben in mehrsprachigen Umgebungen.“ Knorr, Dagmar/ Verhein-Jarren, Annette, eds: Schreiben unter Bedingungen von Mehrsprachigkeit . Frankfurt/ M.: Lang, 139-155. References 277 Langer, Inghard/ Tausch, Reinhard (1972): „Faktoren der sprachlichen Gestaltung von Wissensinformationen und ihre Auswirkungen auf die Verständnisleistungen von Schülern.“ Schule und Psychologie 18 (1972): 72-80. —/ Schulz von Thun, Friedemann/ Tausch, Reinhard (1993): Sich verständlich ausdrücken . 5th, improved ed. München, Basel: Ernst Reinhardt. —, et al. (1973): „Merkmale der Verständlichkeit schriftlicher Informations- und Lehrtexte.“ Zeitschrift für experimentelle und angewandte Psychologie 20.2 (1973): 269- 286. Langer, Judith A./ Applebee, Arthur N. (1987): How Writing Shapes Thinking . Urbana (IL): National Council of Teachers of English. Lave, Jean/ Wenger, Etienne (1991): Situated Learning: Legitimate Peripheral Participation . Cambridge: Cambridge Univ. Press. Lea, Mary R. (2004): “Academic literacies: a pedagogy for course design.” Studies in Higher Education 29.6 (2004): 739-756. —/ Street, Brian V. (1998): “Student writing in higher education: An academic literacies approach.” Studies in Higher Education 23.2 (1998): 157-172. —/ Street, Brian V. (2006): “The ‘academic literacies’ model: Theory and applications.” Theory and Practice 45.4 (2006): 368-377. Leijten, Mariëlle/ Van Waes Luuk (2013): “Keystroke logging in writing research: Using Inputlog to analyze and visualize writing processes.” Written Communication 30.3 (2013): 358-392. Leki, Ilona (1990): “Coaching from the margins: Issues in written response.” Kroll (1990): 57-68. —/ Cumming, Alister/ Silva, Tony (2008): A Synthesis of Research on Second Language Writing in English . New York: Routledge. Levelt, Willem J. M. (1989): Speaking. From Intention to Articulation . Cambridge (MA): MIT Press. Levy, C. Michael/ Ransdell, Sarah, eds (1996): The Science of Writing. Theories, Methods, Individual Differences, and Applications . Mahwah (NJ): Lawrence Erlbaum. Lillis, Theresa (2012): “English medium writing for academic purposes: Foundational categories, certainty and contingency.” Tang (2012): 235- 247. —/ Scott, Mary (2007): “Defining academic literacies research: Issues of epistemology, ideology and strategy.” Journal of Applied Linguistics 4.1 (2007): 5-32. Linke, Angelika/ Nussbaumer, Markus/ Portmann, Paul R. (1994): Studienbuch Linguistik . Tübingen: Niemeyer. Lipsey, Mark W./ Wilson, David B. (2001): Practical Meta-Analysis . Thousand Oaks (CA): Sage. Liu, Yichun (2009): Translation in Second Language Writing: Exploration of Cognitive Process [sic] of Translation . Saarbrücken: VDM. Logan, Gordon D. (1988): “Toward an instance theory of automatization.” Psychological Review 95 (1988): 492-527. Lux, Friedemann (1981): Text, Situation, Textsorte. Probleme der Textsortenanalyse, dargestellt am Beispiel der britischen Registerlinguistik. Mit einem Ausblick auf eine 278 References adäquate Textsortentheorie . (Tübingen Beiträge zur Linguistik. 172). Tübingen: Narr. MacArthur, Charles A. (2007): “Best practices in teaching evaluation and revision.” Graham/ MacArthur/ Fitzgerald (2007): 141-162. Macbeth, Karen P. (2010): “Deliberate false provisions: the use and usefulness of models in academic writing.” Journal of Second Language Writing 19 (2010): 33-48. Macgilchrist, Felicitas (2010): “Publish or perish? A genre-approach to getting published in leading English-language journals.” Girgensohn, Katrin, ed.: Kompetent zum Doktortitel. Konzepte zur Förderung Promovierender . Wiesbaden: Verlag für Sozialwissenschaften, 155-168. Mandelblit, Nili (1995): “The cognitive view of metaphor and its implications for translation theory.” Lewandowska-Tomaszcyk, Barbara/ Thelen, Marcel, eds: Translation and Meaning . Part 3. Maastricht: Hoogeschool Maastricht, 483-495. Manelis, Leon (1980): “Determinants of processing for a propositional structure.” Memory & Cognition 8 (1980): 49-57. Matsuda, Paul Kei (1998): “Situating ESL writing in a cross-disciplinary context.” Written Communication 15.1 (1998): 99-120. — (1999): “Composition studies and ESL writing: A disciplinary division of labor.” College Composition and Communication 50.4 (1999): 699-721. —/ Jablonski, Jeffrey (2000): “Beyond the L2 metaphor: Towards a mutually transformative model of ESL/ WAC collaboration.” http: / / wac.colostate.edu/ aw/ articles/ matsuda_jablonski2000.htm (last accessed: 17 August, 2014). Matsuhashi, Ann (1981): “Pausing and planning: The tempo of written discourse production.” Research in the Teaching of English 15 (1981): 113-134. McCutchen, Deborah (1987): “Children’s discourse skill: Form and modality requirements of schooled writing.” Discourse Processes 10 (1987): 267-286. — (1988): “Functional automaticity in children’s writing.” Written Communication 5 (1988): 306-325. — (1996): “A capacity theory of writing: Working memory in composition.” Educational Psychology Review 8.3 (1996): 299-325. — (2000): “Knowledge, processing, and working memory: Implications for a theory of writing.” Educational Psychologist 35.1 (2000): 13-23. — (2011): “From novice to expert: Implications of language skills and writing-relevant knowledge for memory during the development of writing skill.” Journal of Writing Research 3.1 (2011): 51-68. —, et al. (1994): “Individual differences in writing: Implications of translating fluency.” Journal of Educational Psychology 86 (1994): 256-266. McGinley, William/ Tierney, Robert J. (1989): “Traversing the topical landscape: Reading and writing as ways of knowing.” Written Communication 6.3 (1989): 243-269. Means, Linda (1999): “tfc-gen: Re: expanding text.” E-mails dated 12 and 14 June, 1999, sent to the international mailing list on technical communication tfc-gen@ listserver.tc-forum.org. References 279 Meyer, Bonnie J. F. (1975): The Organization of Prose and its Effects on Memory . Amsterdam: North Holland. Meyer, Jan H. F./ Land, Ray (2006a): “Threshold concepts and troublesome knowledge. An introduction.” Meyer/ Land (2006): 3-18. —/ Land, Ray (2006b): “Threshold concepts and troublesome knowledge: Issues of liminality.” Meyer/ Land (2006): 19-32. —/ Land, Ray, eds (2006): Overcoming Barriers to Student Understanding . London: Routledge. Michigan Corpus of Upper-level Student Papers (MICUSP). http: / / micusp.elicorpora. info/ (last accessed: 03 October, 2014). Miller, George. A. (1956): “The magical number seven, plus or minus two. Some limits on our capacity for processing information.” Psychological Review 2.63 (1956): 81- 97. Moore, D. W. (1996): “Contexts for literacy in secondary schools.” Leu, Donald J./ Kinzer, Charles K./ Hinchman, Kathleen A., eds: Literacies for the 21st Century: Research and Practice: Forty-fifth Yearbook of the National Reading Conference . Chicago: National Reading Conference, 15-64. Moore, Jessie (2012): “Mapping the questions: The state of writing-related transfer research.” Composition Forum 26. http: / / compositionforum.com/ issue/ 26/ mapquestions-transfer-research.php (last accessed: 17 August, 2014). Moreley, John (2014): Academic Phrasebank . University of Manchester. http: / / www. phrasebank.manchester.ac.uk (last accessed 01 October, 2014). Muchinsky, Dennis/ Tangren, Nancy (1999): “Immigrant student performance in academic intensive English program.” Harklau, Linda/ Losey, Kay/ Siegal, Meryl, eds: Generation 1.5 Meets College Composition . Mahwah (NJ): Lawrence Erlbaum, 211- 234. Neubert, Albrecht (1973): „Invarianz und Pragmatik.“ Neubert, Albrecht/ Kade, Otto, eds: Neue Beiträge zu Grundfragen der Übersetzungswissenschaft . Leipzig: Verlag Enzyklopädie, 13-25. The New London Group (Cazden, Courtney/ Cope, Bill/ Fairclough, Norman/ Gee, Jim, et al.) (1996): “A pedagogy of multiliteracies: Designing social futures.” Harvard Educational Review 66.1 (1996): 60-92. Nielsen, Jakob/ Pernice, Kara (2009): Eyetracking Web Usability . Berkeley: New Riders. Niess, Friedrich, et al. (1984): Kraftfahrzeugtechnik. Fachbuch für den Kfz-Mechaniker . Stuttgart: Klett. Nord, Christiane (1989): „Loyalität statt Treue. Vorschläge zu einer funktionalen Übersetzungstypologie.“ Lebende Sprachen 3 (1989): 100-105. — (1993): Einführung in das funktionale Übersetzen - am Beispiel von Titeln und Überschriften . (UTB für Wissenschaft/ Uni-Taschenbücher 1734). Tübingen, Basel: Francke. Norman, Donald A./ Rumelhart, David E., eds (1975): Explorations in Cognition . San Francisco: Freeman. 280 References North, Stephen (1995): “The idea of a writing center.” Murphy, Christina/ Law, Joe, eds: Landmark Essays on Writing Centers . Vol. 9. Davis: Hermagoras Press, 71-85. NSSE (2013): National Survey of Student Engagement . http: / / nsse.iub.edu/ html/ about.cfm (last accessed: 17 August, 2014). Ong, Walter (1978): “Literacy and orality in our times.“ Farell, Thomas J./ Soukup, Paul A., eds (2002): An Ong Reader. Challenges for further inquiry/ Walter J. Ong . Cresskill (NJ): Hampton Press, 465-478. Oomen, Ursula (1972): „Systemtheorie der Texte.“ Folia Linguistica 5.1-2 (1972): 12- 34. (also in Kallmeyer, Werner, et al., eds (1974): Lektürekolleg zur Textlinguistik . 2 vols. Frankfurt/ M.: Athenäum. Vol. 2, 47-70.) O’Regan, J. Kevin (1990): “Eye movements and reading.” Kowler, Eileen, ed.: Eye Movements and Their Role in Visual and Cognitive Processes . (Review of Oculomotor Research 4). Amsterdam, New York, Oxford: Elsevier, 395-453. Ortega, Lourdes/ Carson, Joan (2010): “Multicompetence, social context, and L2 writing research practice.” Silva/ Matsuda (2010): 48-71. Ortner, Hanspeter (1995): „Die Sprache als Produktivkraft. Das (epistemischheuristische) Schreiben aus der Sicht der Piagetschen Kognitionspsychologie.“ Baurmann, Jürgen/ Weingarten, Rüdiger, eds: Schreiben. Prozesse, Prozeduren und Produkte . Opladen: Westdeutscher Verlag, 320-342. Pajares, Frank (2003): “Self-efficacy beliefs, motivation, and achievement in writing: A review of the literature.” Reading & Writing Quarterly 19.2 (2993): 139-158. Paltridge, Brain/ Woodrow, Lindy (2012): “Thesis and dissertation writing: Moving beyond the text.” Tang (2012): 88-104. Pander Maat, Henk (1996): “Identifying and predicting reader problems in drug information texts.” Ensink, Titus/ Sauer, Christoph, eds: Researching Technical Documents . Groningen: Rijksuniversiteit, Department of Speech Communication, 17-47. Paradis, Michel (1994): “Neurolinguistic aspects of implicit and explicit memory: Implications for bilingualism and SLA.” Ellis, Nick, ed.: Implicit and Explicit Language Learning . London: Academic Press, 393-419. Patchan, Melissa M./ Charney, Davida/ Schunn, Christian D. (2009): “A validation study of students’ end comments: Comparing comments by students, a writing instructor, and a content instructor.” Journal of Writing Research 1.2 (2009): 124- 152. Paulson, Lewon F./ Paulson, Pearl R./ Mayer, Carol A. (1991): “What makes a portfolio a portfolio? Eight thoughtful guidelines will help educators encourage self-directed learning.” Educational Leadership 48.5 (1991): 60-63. Paulus, Trena M. (1999): “The effect of peer and teacher feedback on student writing.” Journal of Second Language Writing 8 (1999): 265-289. Perani, Daniela, et al. (1996): “Brain processing of native and foreign languages.” NeuroReport 7 (1996): 2439-2444. Perin, Dolores (2007): “Best practices in teaching writing to adolescents.” Graham/ MacArthur/ Fitzgerald (2007): 242-264. Perkins, David N. (1986): Knowledge as Design . Hillsdale (NJ): Lawrence Erlbaum. References 281 —/ Salomon, Gavriel (1988): “Teaching for transfer.” Educational Leadership 46.1 (1988): 22-32. —/ Salomon, Gavriel (n.y.): “The science and art of transfer.” http: / / learnweb.harvard. edu/ alps/ thinking/ docs/ trancost.htm (last accessed: 17 August, 2014). —, et al. (2000): “Intelligence in the wild: A dispositional view of intellectual traits.” Educational Psychology Review 12.3 (2000): 269-293. Petraglia, Joseph, ed. (1995): Reconceiving Writing, Rethinking Writing Instruction . Mahwah (NJ): Lawrence Erlbaum. Piaget, Jean (1972): Theorien und Methoden der modernen Erziehung . Wien, München, Zürich: Molden. Pohl, Thorsten (2007): Studien zur Ontogenese wissenschaftlichen Schreibens . (Reihe Germanistische Linguistik 271): Tübingen: Niemeyer. — (2009): Die studentische Hausarbeit. Rekonstruktion ihrer ideen- und institutionsgeschichtlichen Entstehung . Heidelberg: Synchron. Polanyi, Michael (1966): The Tacit Dimension . Garden City (NY): Doubleday & Company. Portmann-Tselikas, Paul R./ Schmölzer-Eibinger, Sabine (2008): “Textkompetenz.” Fremdsprache Deutsch (Themenheft Textkompetenz) 39 (2008): 5-16. Prain, Vaughan/ Hand, Brian (1999): “Students’ perception of writing for learning in secondary school science.” Science Education 83.2 (1999): 151-162. Pratt, Mary Louise (1991): “Arts of the contact zone.” Profession 91. New York: MLA, 33-40. http: / / www.jstor.org/ stable/ 25595469 (last accessed: 17 August, 2014). Preußer, Ulrike/ Sennewald, Nadja, eds (2012): Literale Kompetenzentwicklung an der Hochschule . Frankfurt/ M. etc.: Lang. Pritchard, Ruie J./ Honeycutt, Ronald L. (2007): “Best practices in implementing a process approach to teaching writing.“ Graham/ MacArthur/ Fitzgerald (2007): 28- 49. Qi, Donald S. (1998): “An inquiry into language-switching in second language composing processes.” The Canadian Modern Language Review 54.3 (1998): 413-435. —/ Lapkin, Sharon (2001): “Exploring the role of noticing in a three-stage second language writing task.” Journal of Second Language Writing 10 (2001): 277-303. Rayner, Keith (1978): “Eye movements in reading and information processing.” Psychological Bulletin 85 (1978): 618-660. — (1994): “Eye movements during skilled reading.” Ygge, Jan/ Lennerstrand, Gunnar, eds: Eye Movement in Reading . Tarrytown: Pergamon, 205-218. — (1998): “Eye movements in reading and information processing: 20 years of research.” Psychological Bulletin 124.3 (1998): 372-422. —, et al. (1989): “Eye movements and on-line language comprehension processes .” Language and Cognitive Processes 4.3-4 (1989): 21-49. Reiners, Ludwig (1963): Stilfibel. Der sichere Weg zum guten Deutsch . München: dtv. Reiß, Katharina (1983): Texttyp und Übersetzungsmethode. Der operative Text . 2nd ed. Heidelberg: Groos. 282 References —/ Vermeer, Hans J. (1984): Grundlegung einer allgemeinen Translationstheorie . (Linguistische Arbeiten 147). Tübingen: Niemeyer. Reynolds, Ralph E., et al. (1982): “Cultural schemata and reading comprehension.” Reading Research Quarterly 3 (1982): 353-366. Rickheit, Gert/ Kock, Horst (1983): “Inference processes in text comprehension.” Rickheit, Gert/ Kock, Horst, eds: Psycholinguistic Studies in Language Processing . Amsterdam: North-Holland, 182-206. —/ Strohner, Hans (1993): Grundlagen der kognitiven Sprachverarbeitung. Modelle, Methoden, Ergebnisse . (UTB 1735). Tübingen, Basel: Francke. Rienecker, Lotte/ Stray Jörgensen, Peter (2003a): “The genre in focus, not the writer: Using model examples in large-class workshops.” Björk, et al. (2003b): 59-74. —/ Stray Jörgensen, Peter (2003b): “The (im)possibilities in teaching university writing in the Anglo-American tradition when dealing with continental student writers.” Björk, et al. (2003b): 101-112. Rijlaarsdam, Gert/ Couzijn, Michel (2000a): “Stimulating awareness of learning in the writing curriculum.” Camps/ Milian (2000): 167-202. —/ Couzijn, Michel (2000b): “Writing and learning to write: a double challenge.” Simons, Robert-Jan/ van der Linden, Jos/ Duffy, Tom, eds: New Learning . Dordrecht: Kluwer Academic Publishers. Rijnders, J. (1993): De invloed van het opleidingsniveau van proefpersonen op de resultaten van een pretest. Pretest en effectmeting van de brochure Do You Know Do You Care? Universiteit Utrecht [Doctoraalscriptie], quoted in Pander Maat (1996). Robertson, Liane/ Taczak, Kara/ Yancey, Kathleen B. (2014): Writing Across Contexts: Transfer, Composition, and Cultures of Writing . Boulder (CO): Utah State University Press. Roca de Larios, Julio/ Manchón, Rosa M./ Murphy, Liz (2006): “Generating text in native and foreign language writing: A temporal analysis of problem-solving formulation processes.” The Modern Language Journal 90 (2006): 100-114. —/ Murphy, Liz/ Manchón, Rosa (1999): “The use of restructuring strategies in EFL writing: A study of Spanish learners of English as a foreign language.” Journal of Second Language Writing 8.1 (1999), 13-44. Rogers, Carl R. (1999): Die nicht-direktive Beratung. Counseling and Psychotherapy . Frankfurt/ M.: Fischer. Rogoff, Barbara (1990): Apprenticeship in Thinking: Cognitive Development in Social Context . New York: Oxford Univ. Press. Rooken, M. A. (1993): Pretesten bij hoger opgeleiden. De invloed van het opleidingsniveau van de proefpersonen op het effect van de plusen minmethode . Universiteit Utrecht [Doctoraalscriptie], quoted in Pander Maat (1996). Rorty, Richard (1979): Philosophy and the Mirror of Nature . Princeton: Princeton Univ. Press. Rose, Mike ([1984]/ 2009): Writer’s Block: The Cognitive Dimension . Carbondale: Southern Illinois Univ. Press. References 283 Rounsaville, Angela (2012): “Selecting genres for transfer: The role of uptake in students’ antecedent genre knowledge.” Composition Forum 26 (2012). http: / / compositionforum.com/ issue/ 26/ selecting-genres-uptake.php (last accessed: 17 August, 2014). Rubin, Jeffrey (1994): Handbook of Usability Testing: How to Plan, Design, and Conduct Effective Tests . New York, etc.: Wiley. Rumelhart, David E. (1980): “Schemata: The building blocks of cognition.” Spiro/ Bruce/ Brewer (1980): 33-58. Russell, David (1995): “Activity theory and its implications for writing instruction.” Petraglia (1995): 51-77. — (1997): “Rethinking genre in school and society: An activity theory analysis.” Written Communication 14.4 (1997): 504-554. — (2003): “Preface.” Björk, et al. (2003b): V-VIII. Sasaki, Miyuku (1996): Second Language Proficiency, Foreign Language Aptitude, and Intelligence.: Quantitative and Qualitative Analyses . New York: Lang. — (2000): “Toward an empirical model of EFL writing processes: An exploratory study.” Journal of Second Language Writing 9.3 (2000): 259-291. — (2004): “A multiple-data analysis of the 3.5-year development of EFL student writers.” Language Learning 54.3 (2004): 525-582. —/ Hirose, Keiko (1996): “Explanatory variables for EFL students’ expository writing.” Language Learning 46.1 (1996): 137-174. Schindler, Kirsten (2004): Adressatenorientierung beim Schreiben. Eine linguistische Analyse am Beispiel des Verfassens von Spielanleitungen, Bewerbungsbriefen und Absagebriefen . Frankfurt/ M.: Lang. —/ Zepter, Alexandra L. (2013): „Schreibarbeit mit Körperarbeit verbinden.“ Bräuer/ Schindler (2013): 205-220. Schmidt, Richard W. (1990): “The role of consciousness in second language learning.” Applied Linguistics 11.2 (1990): 129-158. Schmidt, Siegfried J. (1973): Texttheorie. Probleme einer Linguistik der sprachlichen Kommunikation . (UTB 202). München: Fink. Schmitt, Peter A. (1999): Translation und Technik . (Studien zur Translation 6). Tübingen: Stauffenburg. Schnotz, Wolfgang (1985a): Elementaristische und holistische Theorienansätze zum Textverstehen . (Forschungsbericht 35 des Deutschen Instituts für Fernstudien an der Universität Tübingen). Tübingen. — (1985b): Kohärenzbildung beim Aufbau von Wissensstrukturen mit Hilfe von Lehrtexten . (Forschungsbericht 39 des Deutschen Instituts für Fernstudien an der Universität Tübingen). Tübingen. — (2006): „Was geschieht im Kopf des Lesers? Mentale Konstruktionsprozesse beim Textverstehen aus der Sicht der Psychologie und der kognitiven Linguistik.“ Blühdorn, Hardarik/ Breindl, Eva/ Waßner, Ulrich H., eds: Text - Verstehen. Grammatik und darüber hinaus . Berlin, New York: de Gruyter, 222-238. 284 References Schoonen, Rob, et al. (2003): “First language and second language writing: The role of linguistic knowledge, speed of processing, and metacognitive knowledge.” Language Learning 53.1 (2003): 165-202. Schreiber, Michael (1993): Übersetzung und Bearbeitung: Zur Differenzierung und Abgrenzung des Übersetzungsbegriffs . (Tübinger Beiträge zur Linguistik 389). Tübingen: Narr. Schriver, Karen A. (1987): Teaching writers to anticipate the reader’s needs: Empirically based instruction . Pittsburgh (PA): Carnegie Mellon University [Unpublished Dissertation]. — (1989): “Evaluating text quality: The continuum from text-focused to readerfocused methods.” IEEE Transactions on Professional Communication 32.4 (1989): 238-255. Schubert, Christoph (2008): Englische Textlinguistik. Eine Einführung . (Grundlagen der Anglistik und Amerikanistik 30). Berlin: Erich Schmidt. Schulz von Thun, Friedemann (1974): „Verständlichkeit von Informationstexten: Messung, Verbesserung und Validierung.“ Zeitschrift für Sozialpsychologie 5 (1974): 124-132. Schunk, Dale H./ Zimmermann, Berry J. (1997): “Social origins of self-regulatory competence.” Educational Psychologist 32 (1997): 195-208. Schwarz, Monika (1992): Einführung in die Kognitive Linguistik . (UTB 1636). Tübingen: Francke. Scott, Mary/ Coate, Kelly (2003): “Rethinking feedback: Asymmetry in disguise.” Björk, et al. (2003b): 87-99. Searle, John Roger (1969): Speech Acts - An Essay on the Philosophy of Language . Cambridge, etc.: Cambridge Univ. Press. Selting, Margret, et. al. (1998): „Gesprächsanalytisches Transkriptionssystem GAT.“ Linguistische Berichte 173 (1998): 91-122. www.fbls.uni-hannover.de/ sdls/ schlobi/ schrift/ GAT/ gat.pdf (last accessed: 01 August, 2006). Shanahan, Cynthia (2004): “Teaching science through literacy.” Jetton/ Dole (2004): 75-93. Shanahan, Timothy (2004): “Overcoming the dominance of communication: Writing to think and to learn.” Jetton/ Dole (2004): 59-74. Shi, Ling (2003): “Writing in two cultures. Chinese professors return from the West.” Canadian Modern Language Review/ La Revue Canadienne des Languages Vivantes 59 (2003): 369-391. Shreve, Gregory M., et al. (1993): “Is there a special kind of ‘reading’ for translation? An empirical investigation of reading in the translation process.” Target 5.1 (1993): 21-41. Silva, Tony (1992): “L1 vs. L2 writing: ESL graduate students’ perceptions.” TESL Canada Journal 10.1 (1992): 27-47. —/ Brice, Colleen (2004): “Research in teaching writing.” Annual Review of Applied Linguistics 24 (2004): 70-106. References 285 —/ Matsuda, Paul Kei, eds (2010): Practicing Theory in Second Language Writing . West Lafayette (IN): Parlor Press. Slomp, David H. (2012): “Challenges in assessing the development of writing ability: Theories, Constructs and Methods.” Assessing Writing 17.2 (2012): 81-91. (DOI: 10.1016/ j.asw.2012.02.001). Smit, David W. (2004): The End of Composition Studies . Urbana (IL): Southern Illinois Univ. Press. Sommers, Nancy (1980): “Revision strategies of student writers and experienced writers.” College Composition and Communication 31 (1980): 378-387. Sowell, Elizabeth R., et al. (1999): “In vivo evidence for post-adolescent brain maturation in frontal and striatal regions.“ Nature Neuroscience 2.10 (1999): 859-861. Spandel, Vicki/ Stiggins, Rick J. (1997): Creative Writers: Linking Writing Assessment and Instruction . 2nd ed. New York: Longman. Spiro, Rand J./ Bruce, Bertram C./ Brewer, William F., eds (1980): Theoretical Issues in Reading Comprehension . Hillsdale (NJ): Lawrence Erlbaum. —, et al. (1987): Knowledge Acquisition for Application: Cognitive Flexibility and Transfer in Complex Content Domains . (Technical Report No. 409). Champaign: Univ. of Illinois, Center for the Study of Reading, quoted in McGinley/ Tierney (1989). Steinhoff, Torsten (2007): Wissenschaftliche Textkompetenz: Sprachgebrauch und Schreibentwicklung in wissenschaftlichen Texten von Studenten und Experten . (Reihe Germanistische Linguistik 280). Tübingen: Niemeyer. Sternberg, Robert J. (1997): “Cognitive conceptions of expertise.” Feltovich/ Ford/ Hoffmann (1997): 149-162. Swales, John M. (1990): Genre Analysis . Cambridge: Cambridge Univ. Press. Swaney, Joyce Hannah, et al. (1991): “Editing for comprehension: Improving the process through reading protocols.” Steinberg, Erwin R., ed.: Plain Language: Principles and Practice . Detroit: Wayne State Univ. Press. Szczyrba, Birgit/ Gotzen, Susanne, eds (2012): Das Lehrportfolio: Entwicklung, Dokumentation und Nachweis von Lehrkompetenz an Hochschulen . Berlin: LIT. Tang, Ramona (2012): “Two sides of the same coin: Challenges and opportunities for scholars from EFL backgrounds.” Tang (2012): 204-232. —, ed. (2012): Academic Writing in a Second or Foreign Language: Issues and Challenges Facing ESL/ EFL Academic Writers in Higher Education Contexts . London, New Delhi, New York, Sydney: Bloomsbury. Taylor, Gordon, et al. (1988): Literacy by Degrees. Milton Keynes: Society for Research in Higher Education/ Open University, quoted in Lea/ Street (1998). Thelen, Ester/ Smith, Linda B. (1994): A Dynamic Systems Approach to the Development of Cognition and Action . Cambridge (MA), London: MIT Press. Tierney, Robert J., et al. (1989): “The effects of reading and writing upon thinking critically.” Reading Research Quarterly 24.2 (1989): 134-173. Triandis, Harry C. (1995): Individualism and Collectivism . Boulder (CO): Westview. 286 References Ulmi, Marianne/ Bürki, Gisela/ Verhein, Annette/ Marti, Madeleine (2014): Textdiagnose und Schreibberatung. Fach- und Qualifizierungsarbeiten begleiten . Opladen, Toronto: Barbara Budrich. Uzawa, Kozue (1994): “Translation, L1 writing, and L2 writing of Japanese ESL learners.” Journal of the Canadian Association of Applied Linguistics 16.2 (1994): 119- 134. — (1996): “Second language learners’ processes in L1 writing, L2 writing, and translation from L1 into L2.” Journal of Second Language Writing 5.3 (1996): 271-294. —/ Cumming, Alister (1989): “Writing strategies in Japanese as a foreign language. Lowering or keeping up the standards.” The Canadian Modern Language Review 46.1 (1989): 178-194. Vermeer, Hans J. (1978): „Ein Rahmen für eine allgemeine Translationstheorie.” Lebende Sprachen 3 (1978): 99-102. Verspoor, Marjolijn/ de Bot, Kees/ Lowie, Wander (2004): “Dynamic Systems Theory and variation: a case study in L2 writing.” Aertsen, Hendrik, ed.: Words in Their Places. Amsterdam: VU, 407-421. —/ Lowie, Wander/ van Dijk, Marijn (2008): “Variability in second language development from a Dynamic Systems Perspective.” The Modern Language Journal 92.2 (2008): 214-231. Vitu, Françoise/ O’Regan, Kevin (1995): “A challenge to current theories of eye movements in reading.” Findaly, John M./ Walker, Robin/ Kentridge, Robert W., eds: Eye Movement Research: Mechanisms, Processes and Applications . Amsterdam, etc.: Elsevier, 381-392. Volkswagen (1988): Die Kupplung - ein Teil mit 7 Siegeln? (Druckschrift 000.7000. 30.00 [5/ 1988]). Volkswagen AG. Vygotsky, Lev S. (1978): Mind in Society: The Development of Higher Psychological Processes . Cambridge (MA): Harvard Univ. Press. — ( [ 1934 ] 1986): Thought and Language . Cambridge (MA): MIT Press. Wade, Suzanne E./ Moje, Elizabeth Birr (2000): “The role of text in classroom learning.” Kamil, Michael L./ Mosenthal, Peter B./ Pearson, P. David./ Barr, Rebecca, eds: Handbook of Reading Research . Vol. 3. Mahwah: Lawrence Erlbaum, 609-627. Wang, Lurong (2003). “Switching to first language among writers with differing second-language proficiency.” Journal of Second Language Writing 12 (2003): 347- 357. Wang, Yiqin (2000): Der Einfluß interkultureller Faktoren auf die Gestaltung der technischen Dokumentation. Empirische Untersuchung chinesischer und deutscher Kfz- Literatur . TU Berlin: Fachbereich Kommunikations- und Geschichtswissenschaften, Dezember 2000 [Dissertation]. Wang, Wewnyu/ Wen, Qiufang (2002): “L1 use in the L2 composing process: An exploratory study of 16 Chinese EFL writers.” Journal of Second language Writing 11 (2002): 225-246. Wardle, Elizabeth (2009): “‘Mutt genres’ and the goal of FYC: Can we help students write in the genres of the university? ” College Composition and Communication 60.4 (2009): 765-789. References 287 —/ Roozen, Kevin (2012): “Addressing the complexity of writing development: Toward an ecological model of assessment.” Assessing Writing 17.2 (2012): 106-119. (DOI: 10.1016/ j.asw.2012.01.001). Weiner, Irving B. (2010): “Attribution theory.” Weiner, Irving B./ Craighead, W. Edward, eds: The Corsini Encyclopedia of Psychology . Hoboken (NJ): John Wiley & Sons, 184-186. Weinrich, Harald (1976): Sprache in Texten . Stuttgart: Klett. Wells, Jennifer Holcomb Marie (2011): Millennials Strike Back: Students’ Reports of Knowledge Transfer from High School to College . Indiana: University of Pennsylvania [Dissertation], quoted in Driscoll/ Wells (2012). Wenger, Etienne (1998): Communities of Practice: Learning, Meaning and Identity . Cambridge: Cambridge Univ. Press. Whalen, Karen/ Menard, Nathan (1995): “L1 and L2 writers’ strategic and linguistic knowledge: A Model of multiple-level discourse processing.” Language Learning 45.3 (1995): 381-418. Wigfield, Allan/ Eccles, Jacquelynne S. (2000): “Expectancy-value theory of achievement motivation.” Contemporary Educational Psychology 25 (2000): 68-81. Williams, Jessica/ Severino, Carol (2004): “The writing center and second language writers.” Journal of Second Language Writing 13 (2004): 65-172. Wittgenstein, Ludwig (1953): Philosophical Investigations . New York: Macmillan. Woodall, Billy R. (2002): “Language-switching: Using the first language while writing in a second language.” Journal of Second Language Writing 11 (2002): 7-28. Wrobel, Arne (1988): „Pausen und Planungsprozesse. Realzeitverläufe beim Verfassen von Summaries, Wegbeschreibungen und Geschäftsbriefen.” Brandt, Wolfgang, ed.: Sprache in Vergangenheit und Gegenwart . Marburg: Hitzeroth, 196-207. — (1995): Schreiben als Handlung: Überlegungen und Untersuchungen zur Theorie der Textproduktion . (Reihe Germanistische Linguistik 158). Tübingen: Niemeyer. Yancy, Kathleen (1998): Reflection in the Writing Classroom . Logan (UT): Utah State Univ. Press. Yore, Larry D./ Hand, Brian/ Prain, Vaughan (1999): “Writing to learn science: Breakthroughs, barriers, and promises.” Paper presented at the International Conference of the Association for Educating Teachers in Science, January 1999, Austin, Texas, quoted in C. Shanahan (2004). Zamel, Vivian (1996): “Transcending boundaries: complicating the scene of teaching language.” College English 6.2 (1996): 1-11. — (1997): “Toward a model of transculturation.” TESOL Quarterly 31.2 (1997): 341- 351. ZfbK (2014): Zentrum für fremdsprachliche und berufsfeldorientierte Kompetenzen. Gießen: Justus-Liebig-Universität. [Homepage] https: / / www.uni-giessen.de/ cms/ fbz/ zentren/ zfbk (last accessed: 19 August, 2014). Zimmermann, Barry J. (2002): “Becoming a self-regulated learner: An overview.” Theory into Practice 42.2 (2002): 64-72. 288 References —/ Kitsantas, A. (2002): “Acquiring writing revision and self-regulatory skill through observation and emulation.” Journal of Educational Psychology 94 (2002): 660-667. —/ Risemberg, Rafael (1997): “Becoming a self-regulated writer: A social cognitive perspective.” Contemporary Educational Psychology 22 (1997): 73-101. Index abstract planning 138 abstraction 57 academic language 157 academic literacy/ literacies XVIII, XIX academic literacies model 171, 172 academic socialization XVIII, XIX, 156, 173 academic socialization model 171, 172 acceptability 5, 26 acclimation stage 161, 162, 163 accommodation 154, 155 acculturation 183, 211 adaptation 186, 187, 189 additive-elementaristic approaches 50 advance organizers 57, 63, 71, 72 adverbs see connectives aesthetic goals 213 AI see Artificial Intelligence Alexander’s Model of Domain Learning see writing competence development models, Model of Domain Learning ambiguity 74 anaphoric reference 8, 9 antecedent genres 58, 214 appreciation of language 213 apprenticeship 194 appropriateness 28 argument 35 argument embedding 36, 37, 40, 50 argument overlap 36, 37, 40, 50 argumentative dimension 158, 159, 160 articles 9 articulatory buffer 115, 120 Artificial Intelligence 42 aspect 10 assessment scheme 241-43 assessment window 53 assignment data 127 assimilation 154 association see principles of cognitive association associative writing 142 attitudes and predispositions 127, 136 attractor states 151 attribution 164 attribution theory 164 audience 128 author I 156 author reference 153, 155, 156 author representation 144, 149 author’s guidelines 61, 64 automatization 148, 151, 185, 198, 199, 203 auxiliary constructions 15 auxiliary constructions to achieve a marked word order 22-26 auxiliary constructions to achieve an unmarked word order 19-22 auxiliary texts 177 backward-reaching high-road transfer 180 Bandura’s self-efficacy theory see selfefficacy theory Beaufort’s model of knowledge domains 131 behavioural processes 165 beliefs and attitudes 117 Bereiter & Scardamalia’s models of composing processes see writing process models, by Bereiter & Scardamalia Bereiter’s five-stage model see writing competence development models, by Bereiter Bereiter’s writing competence development model see writing competence development models, by Bereiter bilinguality XVIII, 221, 231 bioecological model of human development see writing competence development models, bioecological model of human development BMBF XX Bo Peep theory 179 Bologna Reform XIX bottom-up processes 14, 33, 34, 39, 45, 46 bridging 180 Bronfenbrenner’s bioecological model of human development see writing competence development models, bioecological model of human development 290 Index capacity theory of writing see writing competence development models, by McCutchen case grammar 119 case roles 42 cataphoric reference 8 CBT see computer-based training central executive 115 central writing processes see writing processes, central chunking 39 chunks 38 clausal ellipsis see ellipsis clausal substitution see substitution cleft sentences see clefting constructions clefting constructions 15, 23-24 cloze procedures 78, 98 clutch see friction clutch cognitive apprenticeship see learning by cognitive apprenticeship cognitive constructivism 33 cognitive overload 226 cognitive processing 31 cognitive psychology 51, 151 cognitive relief 203, 237 cognitive sciences 60, 246 cognitive structure see structure cognitive writing processes 108, 109-11, 136-40 coherence 5, 11-26, 40, 41, 50, 107 cohesion 5, 6, 11, 26 collaboration 130, 189 collaborative essays 216 collaborative learning 199, 207-10 comment 14 communication theory 60 communicative competence 80 communicative dynamism 14 communicative effectiveness 259 communicative function 60, 61, 76 communicative quality of texts see quality of texts communicative writing 142 competence stage 161, 162, 163 competence-oriented approaches to writing instruction 195 completeness 76, 80 complexity reduction 195, 198, 201, 202, 203 composing competence 221, 231 composing medium 114 composition studies XIX comprehensibility 31, 32, 46, 52, 53, 54, 55, 57, 75, 78, see also text comprehensibility comprehensibility assessment 59, 75, 77-103 expert-judgement-focused methods 59 target-group-focused methods 77- 103 text-focused methods 77 comprehensibility concept by Groeben see Groeben’s comprehensibility concept comprehensibility dimensions 51, 52, 53, 54, 58, 60, 65-75 comprehension 33-34, 35, 40, 41, 46, 49-50, 56, 78, see also text comprehension conceptual 50 perceptual 50 pragmatic sense 50 referential 50 semantic 50 semantic sense 50 syntactical 50 comprehension evaluation 79, 80 comprehension monitoring 79 comprehension regulation 79 computer-based training 64 concept 45 concept formation 153, 155 conceptional phase 61 conceptual comprehension see comprehension, conceptual concessive argumentation 153, 154 concision 51, 53, 56, 64, 65, 66-68 conjunctions see connectives connecting the proposition 119 connectives 10 consecutio temporum see sequence of tenses consistency 74 construction 44 consultation 186, 187, 189 content knowledge 124, see also knowledge types content planning 138 content problem space 126 content structure 71 content-oriented approach see teachercentred approach Context (Bronfenbrenner) 163, 166 Index 291 context adequacy 153, 154, 156 context inadequacy 153, 154 contiguousness 18, 19 continental tradition 194 continuous theme see thematic progression, with a continuous theme contrast focus 15 contrastive approach XVIII contrastive genre analysis 226 contrastive rhetoric 225 converses 21 Cooper & Matsuhashi’s writing process model see writing process models, by Cooper & Matsuhashi coordinating conjunctions see coordinators coordination 130, 186, 187, 189 coordinators 10 co-reference 7 corpus-based approach to writing instruction 259 corpus-linguistic writing competence development model see writing competence development models, based on corpus linguistics correctness 59, 65, 68-69, 111 correlative conjunctions 248 cost/ benefit estimates 118 co-teaching 185, 186, 189 cotext 9 criteria of textuality see textuality Critical Framing 176, 178 critical theory goals 213 critical thinking 191 cultural studies goals 213 culture specificity 225 culture-specific differences 48 curiosity 164 curiosity arousal theory 57 cyclic processing see model of cyclic processing decision making 116, 137 declarative knowledge 199, see also knowledge types deep learning 171 deep processing 161, 162 deep structure 56 deep thinking 215 deficit perspective 172 definite articles see articles definition of quality standards 194 deletion 43 deliberate practice 198, 200 demand (Bronfenbrenner) 164 democratic goals 213 derived themes see thematic progression, with derived themes dimensions of comprehensibility see comprehensibility dimensions directing the proposition 119 directness 74 disconnected curriculum 181 discourse community knowledge 192, see also knowledge types discourse dimension 158, 160 discourse knowledge 124, see also knowledge types discourse plans 118 dispositions 164, 166 distractibility 164 division of work 127 documents 127 doing school 132, 192 domain knowledge 161, 162, see also knowledge types DST see dynamic systems theory dummy subject 19, 20 dynamic systems 150 dynamic systems theory 150-51 ecological model of assessment 190 editing 109, 111 educational psychology 59, 60, 65 effectiveness 28 efficiency 28 EFL writing model 234-35 egocentrism 146 Einstieg mit Erfolg see Starting with Success elaboration 228 ellipsis 9-10 embedded cognition 246 empirical inductive approach 51 emulation 205 endophoric reference 9 English-medium instruction 221, 225 engram 46 environmental processes 165 epistemic function of writing 145, 221, 222 epistemic writing 143 epistemology 172 equilibrium 151, 154 equivalence 66 error classification scheme 237-40 292 Index evaluation 139, 215, 222 evaluative appendix 160 existential presuppositions see presuppositions existential sentences 20 existential there 20 exophoric reference 9 expertise 161, 162 expert-judgement-focused methods see comprehensibility assessment, expert-judgement-focused methods explanatory model of EFL writing see EFL writing model explicit text reference see text reference expressivist goals 213 extended writing 214 exteriorization of comprehension 78-79 exteriorization phase 63 external consistency 80 external sources of information 127 extraposition 21-22 eye tracking 77, 101, 102 eye-mind span 102 far transfer 179 feedback 177, 178, 186, 200, 201, 202, 203, 204, 207, 209, 210, 215, 216-17, 236, 259 fixations 102 fixedness 245 fluency 142, 148, 222 focus 14 foregrounding 92 foreign language see L2 formulating a proposition 118 forward-reaching high-road transfer 180 frames 13-14 framing the proposition 119 freshman writing courses 193 friction clutch 62 FSP see functional sentence perspective functional sentence perspective 14-26, 107, 254 functional-communicative perspective see pragmatic perspective garden-path effect 25 general writing courses 188, 194, 211 generalization 44, 57 generating see idea generation genre-based approach to writing instruction 259 genre conventions 107, 114 genres 58, 184, 213 German Federal Ministry of Education and Research see BMBF goal pursuit 164 goal setting 109, 110, 137 goals 117 Good Shepherd theory 179, 180 Göpferich’s writing process model see writing process models, instructionoriented grammatical simplicity see simplicity, grammatical grammatical subject as ‘dummy’ see dummy subject granularity 240 graphomotor execution 121 graphomotor plan 118 graphomotor realization 121, 123 Groeben’s comprehensibility concept 51, 55-58, 59 group work 207 guiding features of text production 61- 65 Günther’s (1993) phrase-oriented production system see writing process models, phrase-oriented production system Hamburg comprehensibility concept 51-55, 59 Hayes & Flower’s writing process model see writing process models, by Hayes & Flower Hayes’ (1996) writing process model see writing process models, by Hayes hedging 63, 74 hermeneutic stylistics 55 higher-order concerns 187 high-road transfer 180 holistic approaches 50 hugging 180, 181 human memory see memory model hypertheme 17 idea generation 109, 110, 223, 228, 229 idea organization 229 identity 18 illogical connectors 248 imitation 154, 155 implicitness 252 impulsiveness 164 incentives 118, 127 Index 293 inclusion 18 indefinite articles see articles individual 113, 127, 130-36 individual interest 161, 162 inferences 13, 40, 41 inferencing 11, 39, 40, 41, 43, 46, 49, 50, 116, 137 inflectional morphology 15 informants 127 information clarity 80 information clustering maxim see maxim, information clustering information retrieval 137 information theory 56 informativity 5, 26 instruction-oriented writing process model see writing process models, instruction-oriented integrated approaches to writing instruction 191, 192, 211 integration 43, 186 integration of content-and-languageintegrated learning and literacy development see integrated approaches to writing instruction intelligence 233 intentionality 5, 26 interdependence hypothesis 224 interdisciplinary collaboration 186 interest 161 interference 221, 233, 245 interim versions 127 internal consistency 80 interpretation processes 114 intersubjectivity 153 intertextuality 5, 27, 153, 155, 156, 157, 160 intralingual translations 67 introductions 157 inversion 20-21 isotopic features see isotopy isotopy 4, 11 job applications 63 journalistic texts 154, 155, 176 Karlsruhe comprehensibility concept 51, 58-76, 98, 178, 182, 236, 243 Kellogg’s (2008) macro-stages of writing competence development see writing competence development models, by Kellogg kernel sentences 56 keystroke logging 77 knowledge 161, see also knowledge types knowledge activation 121, 123 knowledge classifications 132 knowledge crafting 144, 146 knowledge deixis 9 knowledge linearization 121, 122, 123 knowledge selection 121, 122, 123 knowledge telling 124, 125, 142, 143, 146 knowledge transforming 124, 126, 143, 144, 145 knowledge types 124, 130-36, 161, 162, 192, 199 L1 3 L2 3 L2 writing instruction 258 language enhancement 194 language planning 138 language proficiency 231, 232 leading-edge strategy 41 learner-centred approach 211 learning by cognitive apprenticeship 199, 205-6 learning by doing 177, 199-203 learning by observing 203-5 learning by reflection 199, 206-7 learning facilitator 185 learning-oriented approach see studentcentred approach legal requirements 61, 64 legibility 31, 75 legibility research 31, 50 level 1 verbalizations 78, 81 levels of comprehension see comprehension lexical presuppositions see presuppositions lexical simplicity see simplicity, lexical liminal stage 183 line of argumentation 157 linear progression see thematic progression, simple linear ~ linguistic simplicity see simplicity linguistics XIX, 60 literacy XVII, XVIII literacy development XIX literary appreciation and discrimination 142 literature reviews 213, 214 long-term learning 192 294 Index long-term memory 13, 33, 34, 39, 40, 41, 45, 46, 49, 108, 109, 117, 130-36 long-term working memory 147, 198, 199 Lost Sheep theory 179 lower-order concerns 187 low-road transfer 180 LSP research 60 macro-level see writing instruction at the macro-level macro-level structure 71 macro-propositions 44 macro-rules 43, 44 macro-structure 43 Macrotime 166 macro-typography 75 marked word order see word order, marked; auxiliary constructions to achieve a marked word order massed practice 192 massification 226 maxim 75, 80, 87, 91, 92, 95-98 information clustering 91 parallelism 92 McCutchen’s capacity theory of writing see writing competence development models, by McCutchen MDL see writing competence development models, Model of Domain Learning medium 61, 64 memory model 37-39 mental convention model 61, 63-64 mental denotation model 61-63, 66 mental models 35, 48-49, 50, 57, 60, 61, 110, 121 meso-level see writing instruction at the meso-level Mesotime 166 metacognition 79, 182, 196, 199 metacognitive competence 79, 81 metacognitive reading competence 79 meta-communication 157 meta-communicative competence 81 meta-communicative information 57 meta-communicative structure 71 meta-communicative text reference see text reference metaknowledge 233 metalanguage 177 metalinguistic competence 81 methods of comprehensibility assessment see comprehensibility assessment Michigan Corpus of Upper-level Student Papers 259 micro-level see writing instruction at the micro-level micro-level structure 71 micro-propositions 44 micro-structure 43 Microtime 166 micro-typography 75 MICUSP see Michigan Corpus of Upper-level Student Papers misunderstandings 46 model of cyclic processing 37, 40-41 model of EFL writing see EFL writing model models of educational research into student writing 171 models of knowledge domains 127 monitor 112 mother tongue see L1 motivation 51, 53, 57, 65, 69-70, 191, 192, 200 motivation/ affect 109, 113, 114, 117-18, 161 moves 259 multilinguality XVIII multiliteracy/ multiliteracies XVIII, 175-79 multi-modal units of communication 5 multi-storage model 38 narrator I 156 national discourses XIX native-speaker standards 259 near transfer 179 negative transfer 228 network models 35, 42 New London Group 175 nominal ellipsis see ellipsis nominal substitution see substitution nominalization 73, 93, 155 non-content planning 138 non-directivity 187 nonverbal representations 5, 32, 75, 127 noticing 227 object dimension 157, 158, 159 object structure 71 observational learning see learning by observing Index 295 observing see learning by observing ontogenesis of academic writing 156, 157 opposition 18 optimization 55 optimizing reverbalization 77, 79, 81, 101, 236 organization 109, 110, 137, 157, 228 organization - structure see structure originality 153 overgeneralization 45, 178, 210 Overt Instruction 176, 177-78 parallelism 25, see also maxim, parallelism parallelism maxim see maxim, parallelism parodies 157 parsing aids 24, 26 participants 79 participatory approach see learnercentred approach passive voice 21, 56, 73, 74, 184 peer criticism see peer evaluation peer evaluation 207 peer tutoring 178, 186, 187, 189, 207, 211 perceptibility 31, 59, 65, 74-75 perceptual comprehension see comprehension, perceptual performative writing 142 peripheral writing processes see writing processes, peripheral Person (Bronfenbrenner) 163 personal processes 165 phonological loop see phonological memory phonological memory 115 phrase generation 121, 122, 123 phrase-oriented production system see writing process models, phraseoriented production sytem physical environment 113, 114, 128 Piaget’s theory of cognitive development 154 placement tests 188 placing the proposition 119 planning 109, 110, 112, 114, 116, 137- 38, 149, 222, 223 plus/ minus-method 81 Pohl’s stages of academic writing competence see writing competence development models, by Pohl POPS see writing process models, phrase-oriented production system popular-science texts 53, 81, 154, 183, 193, 235 portfolios 206, 207 positive transfer 228 postliminal stage 183 postponement 21 pragmatic goals 213 pragmatic perspective 3 pragmatic presuppositions see presuppositions pragmatic sense comprehension see comprehension, pragmatic sense Prague structuralists 15 precision 111 predicator 35 predictability 78 predispositions 117 preliminal stage 183 présentateurs see presenters presenters 19, 20 presenting the proposition 120 presuppositions 11-13 principled knowledge 161, see also knowledge types principles of cognitive association 18- 19 problem detection 149 problem diagnosis 149 problem solving 109, 116, 137, 153 procedural knowledge 199, see also knowledge types proceduralization 185 Process (Bronfenbrenner) 163 process goals 213 process planning 138 process-oriented approaches to writing instruction 197-210 product-oriented approaches to writing instruction 210 proficiency stage 161, 162 pro-forms 8-9, 248, 249 programmatic manifesto (New London Group) 175 progression 132, 181, 182, 184, 193, 215 propositional cohesiveness 80 propositional density 37 propositional models 35-37, 42, 48, 49, 50 propositions 35, 40, 42, 43, 48, 49, 110, 121 protocol-aided revision 117 296 Index protocols of retrospective interviews 87 prototypes 210 proximal processes 163 pseudo-clefting constructions 24-26 psycho-linguistics 60 public school 48 purpose 60, 61 Qualitätspakt Lehre see Quality Pact for Teaching quality of texts 28 Quality Pact for Teaching 174 quality standards see definition of quality standards RE see Reading Ease reaction papers 216 readability 31, 32-33, 75 readability formulas 32, 77, 98 readability research 31, 33, 50, 52, 55, 56 reader representation 144 reader-based prose 177 reader-focused methods see comprehensibility assessment, target-groupfocused methods reader-text interaction 33, 50 reading 109, 111, 115 reading and writing centres 190 reading competence 80, 102 Reading Ease 32 reallocation of working-memory capacity 151 recall 56 reception economy 75 recurrence 6-7, 8 redundancy 68 referential comprehension see comprehension, referential reflection 114, 115, 116, 137, 139, 206, see also learning by reflection reflective approaches to writing instruction 195 reflective practice 178, 212 reflective thought 142 regressions 102 reinstatement 40, 41 reinterpretation 41 relational density 37 reorganization 41 repeller states 151 repetition 201, see also recurrence reproductions 78, 98 researcher I 156 resources (Bronfenbrenner) 164 retainability 78, 79 retrospective interviews see protocols of retrospective interviews reviewing see revision revision 109, 111, 112, 114, 148, 149, 222, 223 revision loops 215 rheme 14, 15 rhetoric XIX rhetorical analyses of nonfiction 213 rhetorical goals 213 rhetorical problem space 126 RIP see protocols of retrospective interviews roles 119 routinization 151, 176, 200, 201 saccades 102 scaffolding 178, 197, 212 schema 34, 45, 46, 47, 63, 71 schema theory 45-48, 60 schema-theoretical approaches see schema theory schema-theoretical models 35 screen recording 77 scripts 13-14 second language see L2 selection 43 self-efficacy 164, 166 self-efficacy theory 165 Self-Regulated Strategy Development 196 self-regulation 164, 165, 166, 196, 203, 205 semantic component see seme semantic comprehension see comprehension, semantic semantic macro-structures 35, 42-44 semantic memory 115 semantic networks 42 semantic presuppositions see presuppositions semantic sense comprehension see comprehension, semantic sense seme 4, 11 semiotics 60 sender 61 sensor systems technology 70 sensory storage 38 sentence planning 118, 120, 121 sentence realization 120, 121 sequence of tenses 10 Index 297 short-term buffer see short-term memory buffer short-term memory 38, 39, 120 short-term memory buffer 40, 41 similarity 18 simple linear progression see thematic progression, simple linear ~ simplicity 51, 53, 55, 65, 72-74 grammatical 73 lexical 73 situated cognition 246 Situated Practice 176-77 situational interest 161, 162 situationality 5, 27 Six Traits 216 skopos 66, 67, 76 skopos theory 67 slot 45 social cognition 142 social environment 113, 128 socialization 225 spaced practice 192 speech acquisition 146 speech production 120 spill-over effect 102 split rheme see thematic progression, with a split rheme SRSD see Self-Regulated Strategy Development stage model 112 stagnation 151 standardization 130 standards of textuality see textuality Starting with Success XX Steinhoff’s (2007) stages of academic writing competence see writing competence development models, by Steinhoff STM see short-term memory storing the proposition 120 strategic processing 161 strategy 75, 80, 87, 98 strategy selection 149 structural cohesiveness 80 structure 51, 53, 56, 57, 65, 70-72 structuring see organization student-centred approach 186, 188 study skills XVIII, XIX study skills model 171 style guides 64 stylistics 52, 60 subject-matter knowledge 161, 162, see also knowledge types subordinating conjunctions see subordinators subordinators 10 substitution 7-8 substitution by zero see ellipsis subsumption theory 56, 57 surface learning 171 surface-level processing 161, 162 syntactical comprehension see comprehension, syntactical TAP see think-aloud protocols target group 60, 61 target-group-focused methods see comprehensibility assessment, targetgroup-focused methods task environment 108, 109, 113, 127, 128, 129-30 task schemas 133 tautology 68 teacher roles 185, 210 teacher-centred approach 186, 211, 212 teaching centres 186, 188 term papers 152, 156, 173, 224 terminology science 60 text 3-5 text assessment scheme see assessment scheme text base 35, 36, 40, 41 text coherence 94 text competence XVII text comprehensibility 3, 15, 146 text comprehension 3, see also comprehension text constitution 3 text criticism 153, 155 text deixis 9 text delimitation 3 text generation 109, 110-11, 112, 114, 118, 148, 229 text interpretation 114, 115, 116 text linguistics XVII, 60 text optimization 59, 67 text planning 138 text produced so far 109, 114, 127 text production 3, 41, 59, 114, 115, 116, 138 text production aids 127 text reference 10 text representation 144, 145, 149 text typology 3 298 Index text-focused methods see comprehensibility assessment, text-focused methods text-grammatical perspective 3 text-in-function 4 text-in-situation 5 text-linguistic turn 3 text-semantic perspective 3 textual procedures 152, 153 textual routines 152 textuality 3, 5-28 thematic progression 16-18, 56 simple linear ~ 16 ~ with a continuous theme 16 ~ with derived themes 17 ~ with a split rheme 17 theme 14 theoretical deductive approach 51, 55 theory of cognitive development see Piaget’s theory of cognitive development theory of mental models see mental models think-aloud 77, 78, 102, 107, 111, 112, 117, 124, 236 think-aloud protocols 80, 87, 108 threshold concepts 183, 184 threshold level 223, 230 Time (Bronfenbrenner) 163, 166 titles 157 top-down processes 14, 33, 34, 39, 47, 63, 64 topic 14 transactional approach see learnercentred approach transcription 120, 148 transfer of learning 164, 171, 179-84 transfer texts 177 transformation 56, 155 transformational grammar 55 Transformed Practice 176, 178-79, see also transfer of learning translating see text generation translation XVIII, 221, 226, 227, 237, 245 translation competence 221, 233 Translog 84, 86, 87, 100 transmission approach see teachercentred approach transposition 154, 155 typography 31 ultra-short-term memory 38 unified writing 142 unmarked word order see word order, unmarked; auxiliary constructions to achieve an unmarked word order usability 75 usability testing 77 USTM see ultra-short-term memory value 164 verbal ellipsis see ellipsis verbal substitution see substitution visual-spatial sketchpad 115 WAC see writing across the curriculum Weiner’s attribution theory see attribution theory WID see writing in the disciplines WM see working memory word order marked 15 unmarked 15 wording the proposition 120 working journals 206 working memory 38, 39, 40, 113, 114- 15, 136, 147 writer-based prose 177 writing across the curriculum XIX, 191 writing aptitude 225 writing arrangements 177 writing assessment XVIII writing assignments 109, 113, 181, 194, 213-16 writing centres 173, 174, 186, 188, 189 writing competence 143, 150 writing competence development from the perspective of dynamic systems theory see writing competence development models, dynamic systems theory writing competence development models 141-68 based on corpus linguistics 151-61 bioecological model of human development 163-67 by Bereiter 142-43 by Kellogg 143-47 by McCutchen 147-50 by Pohl 156-61 by Steinhoff 152-56 dynamic systems theory 150-51 Model of Domain Learning 161-63 writing diaries 206 writing expertise 223 Index 299 writing fellows 186, 207 writing for thinking XVII-XVIII writing groups 189, 207 writing in the disciplines 172, 191 writing instruction XIX, 58, 107, 149, 163, 171-217, 258 at the macro-level XIX, 175, 182, 185-90 at the meso-level XIX, 175, 182, 190- 94 at the micro-level XIX, 175, 181, 195-210 writing model see writing process models writing pedagogy see writing instruction writing plans 109 writing process models 107-40 by Bereiter & Scardamalia 124-26 by Cooper & Matsuhashi 112, 118- 20 by Hayes 113-18 by Hayes & Flower 108-12 instruction-oriented 127-40 phrase-oriented production system 112, 120-24 writing processes central 121 peripheral 121 writing program generation 121, 123 writing programs 173 writing skill see writing competence writing tutoring 207 writing-intensive seminars 132, 172, 174, 177, 186, 188, 189, 192, 201, 207, 210-13 ZfbK XX, 189 zone of proximal development 178, 200, 205, 214 Narr Francke Attempto Verlag GmbH+Co. KG • Dischingerweg 5 • D-72070 Tübingen Tel. +49 (07071) 9797-0 • Fax +49 (07071) 97 97-11 • info@narr.de • www.narr.de JETZT BES TELLEN! JETZT BES TELLEN! Jan Engberg / Carmen Daniela Maier Ole Togeby (eds.) Reflecting upon Genres Encounters between Literature, Knowledge, and Emerging Communicative Conventions Europäische Studien zur Textlinguistik, Vol. 13 2014, 230 Seiten, €[D] 58,00 / SFr 74,70 ISBN 978-3-8233-6817-5 The present book departs from an observation made by a group of scholars at Aarhus University: On the one hand, the concept of genre is present in and pivotal for a number of different disciplines studying texts such as literary studies, analytical text linguistics and the investigation of text production in professional settings. On the other hand, and interestingly, each of these disciplines tends to have developed its own theoretical tools and basic assumptions, without taking into account the results and insights achieved in the neighbouring fields. The present work is intended to overcome this state of affairs. It is based upon a series of seminars involving scholars from the mentioned fields. Questions that emerged from the interdisciplinary discussions of the group and that are treated across the different contributions include the following: What is the relation between genre, text production and situation? To what extent is the situation or the function the overarching factor in characterising and distinguishing genres? How do genres develop and acquire new textual characteristics? How does the specificity of the represented genres surface in text and context? The result is an inquiry into problems with relevance across disciplines, where contributions from each field intend to also reflect aspects traditionally treated in the other fields. Narr Francke Attempto Verlag GmbH+Co. KG • Dischingerweg 5 • D-72070 Tübingen Tel. +49 (07071) 9797-0 • Fax +49 (07071) 97 97-11 • info@narr.de • www.narr.de JETZT BES TELLEN! JETZT BES TELLEN! Heike Ortner Text und Emotion Theorie, Methode und Anwendungsbeispiele emotionslinguistischer Textanalyse Europäische Studien zur Textlinguistik 15 2014, X, 485 Seiten €[D] 78,00 / SFR 101,40 ISBN 978-3-8233-6910-3 Der Zusammenhang zwischen Sprache und Emotion wird in der Linguistik aus vielen verschiedenen Perspektiven untersucht. Dieser Band dient als Zusammenfassung des aktuellen Forschungsstandes zu dem komplexen Thema. Berücksichtigt werden Erkenntnisse aus verschiedenen Teildisziplinen, z.B. Semiotik, Lexikologie, Pragmatik, Kognitive Linguistik und Textlinguistik. Im methodischen Teil wird gezeigt, wie eine emotionslinguistische Analyse emotive Strukturen in Texten offenlegt, wobei die vorgeschlagene Methode leicht an verschiedene Fragestellungen angepasst werden kann. Ihre Anwendung und empirische Überprüfung findet an sehr unterschiedlichen Textkorpora statt: Briefe von Franz Kafka als Beispiel für einen Individualstil sowie Nachrichtenartikel von verschiedenen Online-Plattformen als Beispiel für Medientexte. Forschende, Lehrende und Studierende finden hier sowohl einen umfassenden Überblick über theoretische Grundlagen als auch Anregungen für die Anwendung in der eigenen Forschung und Lehre. Europäische Studien zur Textlinguistik 16 How can text competence be fostered in a more efficient and effective manner? This is one of the central questions of this book, which combines the Anglo-American with the German discourse. The topics covered range from text linguistic foundations via text comprehension and comprehensibility to text production, writing skills development and writing in a second or foreign language. Students interested in writing research will be introduced to the pertinent models and theories. Writing instructors, writing centre staff and subject-domain teachers will find guidance on how to improve their assignments and feedback. University administrators and program coordinators can inform themselves about best-practice approaches to writing instruction and support at different levels ranging from individual courses to central support structures.