Reflections upon Genre Encounters between Literature, Knowledge, and Emerging Communicative Conventions Jan Engberg / Carmen Daniela Maier Ole Togeby (eds.) Reflections upon Genre Europäische Studien zur Textlinguistik herausgegeben von Kirsten Adamzik (Genf) Martine Dalmas (Paris) Jan Engberg (Aarhus) Wolf-Dieter Krause (Potsdam) Arne Ziegler (Graz) Band 13 Jan Engberg / Carmen Daniela Maier Ole Togeby (eds.) Reflections upon Genre Encounters between Literature, Knowledge, and Emerging Communicative Conventions 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. © 2014 · 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 978-3-8233-6817-5 Table of Contents Jan Engberg / Carmen Daniela Maier / Ole Togeby Introduction ............................................................................................................. 7 Svend Østergaard / Peer F. Bundgaard The Double Feedback Loop and the Parameter Theory of Text Genres ....................................................................... 17 Simon Borchmann The Perilous Life of a Linguistic Genre Convention A Study of News Analysis within the Danish System of Newspaper Genres ............................................................................................ 45 Carmen Daniela Maier / Jan Engberg Tendencies of Multimodal Gradations in Academic Genres Network ............................................................................ 113 Ole Togeby A Model of Text Types and Genres .................................................................. 147 Peter Widell The Literary Text: Four Parameters .................................................................. 177 Bionotes ................................................................................................................. 229 Jan Engberg / Carmen Daniela Maier / Ole Togeby Introduction The aim of this book is to contribute to the research field of genre in novel ways. The increasingly wide range of publications about genre definitely suggests that research in genre represents a dynamic field in constant expansion, just like is the case with the texts used in contemporary communication and studied through the concept of genre. This book is meant to provide some proves of the potential force of the concept of genre for exploring a wide spectrum of communicative aspects of our world. It does so by presenting tendencies in approaching these aspects from various perspectives upon genre. It is worth emphasizing from the very beginning that genre is considered here as being a culturally, socially, linguistically and multimodally significant entity. Only by combining all of these viewpoints may this research object be properly described and characterised. Therefore, the present book has been conceived and should be considered as an interdisciplinary endeavor in the quest of a fuller explanation than is traditionally the case in disciplinary studies of genre. The book is the result of a series of discussions about genre that took place at Aarhus University in Denmark. It is meant to reflect the spirit of these discussions. The group was convened in 2008 by Ole Togeby and has met with regular intervals since then. It combines researchers with backgrounds in philosophy, linguistics, semiotics, literary studies, cognition, multimodality, translation and professional text production. The variety of perspectives, theories and basic positions represented in the group guaranteed long, animated and interesting discussions over the years. Topics that have kept resurfacing include the characteristics of literary texts, non-literary features underlying classifications of texts into literary genres, relevance and viability of basic positions on the emergence of meaning in communication like the ones proposed by Wittgenstein, Searle, Grice and others, the importance of philosophical thinking in studying textual communication, the necessity or impossibility of breaking with systemic-structuralist thinking in genre analysis, and the nature and importance of communicative functions for the study and classification of genres, to mention only the most important ones. The reader will find diverse, often cognate, sometimes antagonistic answers to and positions on these topics in the chapters. The group was not convened in order to write a book; our discussions were meant to inform each other about the variety of viewpoints and approaches to genres that each of us has been confronted with over the years. 8 Jan Engberg / Carmen Daniela Maier / Ole Togeby We have used our stimulating and challenging discussions to clarify differences. In this connection, giving place to individuality was one of our priorities during the discussions. And therefore, this book is meant to reflect individualities engaged in converging interactions. However, during the clarifying process, we have experienced how interactions became possible and consensus points started to emerge. Hence the idea to publish the papers together instead of doing that in our different research communities. Chapters of the book The chapters oscilate in their interests between theory building and descriptions of concrete (instances of) genres and generic developments. Individuality emerges in the specific disciplinary “voice” of each chapter, while the content-related differences include the various purposes, conceptualizations, analytical parameters, data and contexts. Collectively, the chapters engage in processes of interpretations and explanations which are intersecting each other on various levels and from different standpoints. The first article in our collection (The Double Feedback Loop and the Parameter Theory of Text Genres) is written by Østergaard and Bundgaard. It is interested in generic development and has a double scope. First, the dynamics inherent in the emergence of genres are considered. The propagated view is that genres emerge relative to two sets of constraints, which are captured in a suggested double feedback loop model for the dynamics of genres. On the one hand, (text) genres emerge as a variation of already existing genres. On the other hand, genres develop as a response to the negative constraints or positive affordances of given situations: that is, either the exigencies of the situation or the new resources available in a situation. Accordingly, the article starts out with a characterization of situations and of the dynamic relation between situational constraints/ affordances and genres. The main claim is that situations and genres stand in a relation of mutual scaffolding to each other so that the existence of a genre is not simply caused by the exigencies present in a given situation, but, once emerged, also feeds back into the situation, further stabilizing or consolidating it: hence, the use of the term “feedback loop”. This argument is followed by a more detailed discussion of the dynamics of genres with a particular focus on the first feedback loop: the way genres develop as deviations from existing genres and then stabilize as proper genres with a normative import. The second scope of the article consists in developing a typological apparatus consistent with the dynamic approach to the emergence of genres in the form of a parameter theory. Genres are considered as governed by parameters external to them and intrinsic to the situations they are dynamically related to. Genres should thus be understood not simply in terms of inherent textual or formal traits, but also relative to a cer- Introduction 9 tain set of situational parameters and relative to the degree to which they are governed by them. The following two chapters are also interested in the factors influencing generic development, and they also treat many of the same factors as the one introduced in the first article. But they investigate the process relative to two concrete emerging genres. However, the primary aim of Borchmann’s chapter (The Perilous Life of a Linguistic Genre Convention: A Study of News Analysis within the Danish System of Newspaper Genres) is still theoretical: to present a linguistic text analysis that differs from standard text linguistic approaches by being informative with regard to the linguistic choices and textual organisation that characterise a text as a social act. The analysis is exemplified by using texts of the relatively new Danish journalistic genre nyhedsanalyse (news analysis). The secondary, empirical aim of the article is to present a corpusbased, linguistic analysis of central elements of the genre nyhedsanalyse within the Danish system of newspaper genres. Text linguistics is based on the assumption that linguistic structures extend beyond the sentence border and comprise parts of texts (cohesion) or even whole texts (macro structures). In the article it is argued that the concepts cohesion and macrostructure are incompatible with the basic assumptions of natural language philosophy, and that they prevent an informative analysis of texts as social acts. Alternatively, it is suggested that linguistic choices and textual organisation are guided by social motives (Miller 1984), and that social motives, themselves, are the result of a multitude of factors, among others: constitutional, political, governmental, ideological, educational, professional, ethical, technological, economical, and commercial. These dependency relations appear as linguistic and organisational regularities in naturally occuring social acts constituted by the same social motives. Consequently, the regularities should not be accounted for as manifestations of linguistic rules, but as manifestation of social conventions. Furthermore, an account of the linguistic and organisational regularities presupposes an account of the social motive and the factors related to the motive, and vice versa. As Miller’s description of the dependency relations remains abstract on the textual levels, it is supplemented by Swales’s move-structure analysis. Swales’s analysis, on the other hand, is supplemented by focusing on the developmental history of the genre system, as illustrated by Bazerman, by incorporating psychological and, in particular, cognitive factors, as suggested by Bathia, and by integrating ideas and linguistics descriptions of the Danish school of functional linguistics. This theoretical framework forms the basis of an analysis of a corpus consisting of 43 news analysis texts, of which 27 are descriptive. The empirical analysis focuses upon establishing the move structure of this sub-genre. Elements of the move structure are subjected to an information structural analysis, yielding the result that information structural 10 Jan Engberg / Carmen Daniela Maier / Ole Togeby characteristics of the moves support a determination of framing as a typified social action within the news analysis genre. The central claim of the article is that the linguistic convention that governs the step ‘framing an event’ is regulative rather than constitutive, and that the convention, as a manifestation of a social motive that is the result of a multitude of interdependent factors, lives a perilous life. The theoretical point is that this precarious status as a temporary, constantly-threatened, regulative rule, does not make the convention a less appropriate object of linguistic description. A modern, linguistically-based text description should comprise this object. Firstly, the descriptions are more informative than the structures hitherto described by text linguistics. Secondly, as historical norms, they are a testimony to the development and change of language use. Thirdly, the descriptions contribute to language users’ awareness of the origin of standards, their understanding of how their own language use is related to standards, their ability to analyse a situation of conflicting standards, and their awareness of a priority of standards within a given form of life. The other chapter aimed at describing a concrete genre has the title Tendencies of Multimodal Gradations in Academic Genres Network. The authors Maier and Engberg depart from the mainly monomodal approaches adopted in the other chapters and focus on the multimodal aspects of texts belonging to the academic genre network: traditional research articles and two relatively new academic genres, visual essays and video essays, which have appeared in two international peer-reviewed journals, Visual Communication and Audiovisual Thinking. The shift to a multimodal descriptive framework is realized through selecting a social semiotic perspective upon genre. It allows researchers to explore other semiotic modes than language, together with their meaning-making interplay. In fact, identifying the role of the semiotic interplay in creating and multiplying meaning at various levels of a generic structure is one of the main concerns of this chapter. The multimodal analysis is undertaken in several descriptive layers, employing Bateman’s (2009) concepts that label three distinctive resources of structuring multimodal texts: the text-flow, the page-flow and the image-flow. As these genres are supposed to facilitate the communication of academic knowledge, aspects related to intrageneric and intergeneric intertextuality are also examined in order to explain in a more nuanced manner the differences between these academic genres. For the same reason, the relation between tacit and explicit knowledge has been comparatively revealed in each of the three genres. Although this research endeavor is still in an incipient phase, it has been possible to show that the dynamic network formed by these genres is characterized today by the existence of various forms of multimodal gradations - transitional and intermediate - due to the overlapping and constantly evolving generic forms. According to Maier and Engberg, this tendency might be enhanced in the years to come due to the ever-developing technological tools and to the multi- Introduction 11 literate generations of students that will continue to join the academia (cf. the idea of the perilous life of generic regularities introduced in Borchmann’s article). Therefore, the analysis of these evolving genre forms becomes correspondingly more important both for those creating these genres and for those supposed to use them. The last two chapters are again more focused upon theory building, although without leaving the field of empirical testing and documentation. In his chapter A model of text types and genres Togeby proposes a model of genre and text types, unifying theories about speech acts (Searle 2010, Habermas 1971), theories about text types that deal with English for Special Purposes (Swales 1990), relevance theory (Grice 1975, Sperber and Wilson 1986), textlinguistics (Beaugrande and Dressler 1981) and literary genre theory (Bachtin 1986, Richard Walsh 2007, James Phelan 2007). Terminologically, a distinction is made between a speech act that is an utterance of a single sentence, and a text (act) that is the utterance of several (written or spoken) sentences combined into one complete intentional unit. Analytically, one can discriminate between levels of acts with layers of conditions of satisfaction: reference and predication acts (with felicity and truth conditions), speech act (with direction of fit conditions), verbal macro act (with expediency conditions) and text acts (with situational fit conditions). Texts containing several coherent sentences are subdivided into three main types: practical texts, factual prose and literature, primarily based on their dominant function, as a social act in relation to the addressees, as a representation of some state of affaires, or as a remarkable form that is an end in itself, respectively. They tend to be distributed differently on many parameters such as number of sentences, dialogue or monologue, fleeting or permanent, used up or reusable, one-to-one or one-to-many, addressed or to the public. These parameters are, however, not criterial for belonging to one genre or the other, but only conditions for ‘texthood’. Text types (and as an overlapping synonym in this chapter genres) are defined as historically evolved types of text with collectively accepted standards for efficient fit between the text internal subject matter and linguistic form and the text external physical, situational and social conditions. So a theory of genre and text types involves a typification of speech acts, a typification of communicative situations, a typification of linguistic macro acts, and a typification af subject matters with which the text deals. Linguistic macro acts are described by their dominant type of information, their types of reference, their types of predicates and their types of connections between adjacent propositions. The last chapter of the book (The literary text: Four Parameters) by Widell is an attempt to shed light on the concept of the literary text using Grice’s communication-based theory of language. The theoretical reflections are not aimed at establishing a genreor text type-taxonomy as is a case in some of the other chapters of this volume. The presented exploration is instead an 12 Jan Engberg / Carmen Daniela Maier / Ole Togeby attempt to establish a parameter theory of literary prose in the form of a universal, restricted set of four parameters characterizing every literary text. The four parameters are, respectively, figurativity, tropicity, fictionality and degree of (literary) realism. On the basis of Grice’s theory it is possible both to give a precise definition of the literary text and to highlight numerous subordinate conceptual distinctions in genre theory - and actually in the whole field of rhetoric and stylistics in general - in the light of a comprehensive, unifying and, in that sense, explanatory theory. The secret object of the theory is the ability of humans to use a special kind of inferences - Grice calls them implicatures - initiated by breaches of the standards of non-literal prose. It is here that the backdrop for identifying the four above-mentioned parameters can be found. The article is not only an attempt at defining and characterizing the literary text in terms of Gricean theory. It also tries to establish a new look on the parameters of figurativity, tropicity, fictionality and degree of realism. An important prerequisite here is a robust adherence to philosophical realism, that is to the thesis that the world and the things in the world exists independently of our speaking or thinking about them. As to figurativity, the article is an attempt at showing that all figures in a literal text are instances of kinds of identical or varied repetitions. Figurativity cannot constitute meaning, but it can be an indicator of finding loosely connected meanings in a literary text. As to tropicity, the article is a showdown with the prevailing cognitive linguistic theories of the day and especially with their insistence on the view that metaphors create new meanings. In reply to that, the present article tries to pay respect both to the textual marker and to the literal paraphrase of the trope in an effort to establish a modern version of the classical theory of tropes dating back to Aristotle. As to fictionality, the article is a showdown with both the prevailing semantic theories of fictionality - where fiction is thought of as a kind of reference into a thought image or more abstractly a possible world - and with the speech act oriented theories of fictionality according to which fictionality is established via a sort of direct or indirect illocutionary act or a kindred kind of conventional act. Instead, the article intends to found fictionality in our ability to fake actions in general and not just speech acts. As to degree of realism, the article is a showdown with the idea that literary realistic texts speak directly about the world, in contradistinction to ideas or phantasies. As a fictional text, the literary realistic text cannot speak directly about the world because in fiction there is no world. Instead, we have to look at other texts. Literary realism is a mimicry of non-literary texts as was already seen by Aristotle. The result of the article’s discussion is that if you have on the one hand Grice in your pocket and on the other hand a thorough understanding of the concept of (figurative) repetition, and if you furthermore have a robust con- Introduction 13 cept of realism, then the literary text will - abstractly characterized - reveal itself through the scrutiny of the theory as a tight system of identical and - especially - varied repetitions at different levels of reality as a vibrant means for satisfying your aesthetic pleasure. Perspectives and synergies in the book Held together by family resemblances, the chapters complement each other in painting a multifacetted picture of the concept of genre. It is important to see the relations between the chapters exactly as family resemblances, i.e., as traits that are shared by some, but not all of the members of the group. The relations are visible as overlapping interests between the chapters, although they only partially overlap in some areas. As a first example, from the point of view of topic, the chapters span from having a main focus upon basic aspects of one group of genres, i.e. literary texts (Widell), over systems of analytical parameters intended to help categorising (systems of) genres (Bundgaard/ Østergaard, Togeby) to studies focusing upon the development and distinction of specific genres (Borchmann, Maier/ Engberg). All chapters are interested in the historical emergence and evolution of genres and the parameters, social conditions and constraints influencing the process. But where Widell has a main focus upon the field of literary texts, such texts are only one genre among others covered in the chapters by Togeby and by Bundgaard/ Østergaard, who use them as examples in the description of the parameters. In all cases, general characteristics of literary genres rather than traditional literary research are focused. In the chapters by Borchmann and by Maier/ Engberg, the classification and characterisation of literary texts play no role. Instead, these chapters treat texts from the more information-functional fields of newspapers and academic writing. Apart from these partial, scalar overlaps concerning the treated objects, actual synergies are visible along a number of dimensions across the chapters. These dimensions signal routes along which the discussions leading to this book have been developing. In the following, we will highlight some of them. One such dimension is the principle of describing and characterising genres on the basis of parameters. It is a common feature of the chapters by Bundgaard/ Østergaard, Togeby and Widell. The main point here is that due to the characteristics of the objects TEXT and GENRE , the difference between specific types is rather one of gradation along a number of parameters than one of presence or absence of characteristics in an Aristotelian system of classification. As is stated in all three chapters, all texts and genres may generally be characterised by the features of all the parameters suggested. However, the overlapping interests in the field of setting up relevant parameters also show one of the other characteristics of this book: due to the fact that each author has a different disciplinary background, a different theoretical standpoint 14 Jan Engberg / Carmen Daniela Maier / Ole Togeby (e.g., different position concerning the importance of text internal and situational factors) and writes about a different object, the chosen sets of parameters are different - but they overlap enough for them to supplement each other. Another dimension concerns the basic nature of the concept of genre. This is a common feature of the chapters by Borchmann and Maier/ Engberg. Both chapters are based upon the idea that genres are mental categories applied by text producers as guidelines in their text production in predominantly information-functional situations. The chapters differ in the methods used for describing the characteristics of the investigated genres. Maier/ Engberg use methodologies developed in the more structuralistically oriented field of system-functional linguistics, introducing theory-based systems of descriptors guiding the analysis (e.g., a system of multimodal interactions in the social semiotic process). Borchmann, on the other hand, follows a methodology in his analysis above sentence level, based upon Swales’ method of move structure analysis, which lends more importance to the empirical basis and the context in building its descriptive categories. However, the basic agreement upon the (cognitive) nature of genres makes the two chapters complement each other in showing (compatible) features of processes of historical development. A dimension which has played a central role in many of the discussions over the years concerns the basic theories of meaning creation. In his chapter, Widell builds his argumentation upon Grice’s concept of implicatures based upon universal communicative maxims as the main mechanism of meaning creation in conversation and also includes Searle’s speech act oriented approach. Togeby also applies these two thinkers as basis for his chapter, when treating the issue of meaning creation. On this topic, Maier/ Engberg draw upon the structure-oriented approach of social semiotics (Kress, van Leeuwen, Bateman). All of these approaches emphasise the social side of the meaning creation process. In some of the parameters presented in the chapter by Bundgaard/ Østergaard, Grice and Searle play a central role. But apart from them, authors more oriented towards cognitive approaches like Talmy, Langacker and Tomasello are introduced and applied in this chapter. Finally, cognitive approaches also play a central role in the chapter by Borchmann. The cognitive approaches focus more upon the importance of (individual) mental processes for the emergence of meaning. Despite the differences in these focal points, the present chapters and especially the chapter by Bundgaard/ Østergaard show that truth may well be found in a complementary description, including both socially and cognitively oriented approaches. Another dimension recurring in the chapters is the question of how to describe the relation of fit between texts and the situation in which they occur. Togeby bases the idea of fit upon Grice’s (1975) general cooperative principle and thus introduces a general principle of succesful communication. Bundgaard/ Østergaard are mostly prone to Bitzer’s (1968) rhetorical ap- Introduction 15 proach to the fit between text and situation, emphasising the guiding role of the situation. Borchmann favors Miller’s (1984) response to this approach, emphasising the role and importance of conventions due to the nature of situations as social constructs and the dynamic relation between texts and situations. And Maier/ Engberg draw upon a similar approach, the sociocognitive approach by Berkenkotter/ Huckin (1995), which also starts from the idea of a dynamic relation between texts and situations, but emphasises the importance of (individual) cognition. The importance of the dimension stems from the fact that all of the authors consider the function of linking text design to situation as a central aspect of the concept of genre. The four approaches are different answers to the central question. And they have emerged as reactions to each other. Thus the four chapters may be seen as complementary answers to the question of fit, each drawing upon the stronger parts of the chosen basic approach. On a final note, in this book, we hope to assist our readers in their own endeavors to demystify genre issues that they may have to cope with both as researchers and educators. Aiming to “record” our own dialogues, this book is also meant to encourage an inner dialogue which each reader could engage in while being confronted with the viewpoints presented here. Svend Østergaard / Peer F. Bundgaard The Double Feedback Loop and the Parameter Theory of Text Genres 1 Situations By a linguistic text we understand a series of utterances expressed to others or exchanged with others in a given situation, at a given time, relative to a given object (or with a given purpose). The social world is constituted by a limited set of recurrent types of situations in which people gather and express themselves about objects, states of affairs, or purposes proper to these situations. Situations are, in different ways and for different reasons, platforms for recurrent kinds of internally structured semiotic interaction. The different situations proper to the social world call for different kinds of actions and behavior and also different kinds of texts in the above sense. When people get together at religious meetings, in schools and academies, in assemblies or societies to decide upon legal and social matters, in public spaces to commemorate or celebrate such and such past or present event, they exchange or express specific kinds of texts bound to those situations, informed by the objects dealt with or the goals pursued in them. Thus, in decisional contexts, just to name one of these situations, people’s deliberation is based on principles established in law books or on prevailing customs, often recollected and made available in exemplary narratives (myths, legends, songs, and so on). To summarize: by situation we mean a recurrent social situation that gives rise to recurrent utterances with identifiable characteristics. That is, then, the main claim: genres are recurrent ways of using language that emerge according to the constraints inherent in situations. It follows from this that by properly understanding and characterizing the nature and requirements of the typical situations of the social world we also gain insight into the kind of texts or genres they call for. The present approach therefore has a double scope. On the one hand, it aims at laying bare characteristic aspects of genres and typologizing them in terms of those situational ‘parameters’ that motivate their existence; on the other hand, it considers genres as text types that co-emerge with and, therefore, shape the situations in which they are used. For this reason, our investigation is rather an inquiry into the phenomenological, social, or intersubjective constraints under which genres come about than an attempt at defining formal traits which constitute one genre and distinguish it from others. In other words, the parameters by way of Svend Østergaard / Peer F. Bundgaard 18 which we try to differentiate between and categorize genres capture relevant aspects of the situations from which genres have evolved. What counts, then, as a situation? We can best elucidate our notion of situation by means of examples. An epitome of and an exclusively human kind of social situation involves transmission of knowledge writ large: i.e. theoretical knowledge (laid down, first, in mythological-religious cosmologies, next, in science, and then preserved in text books, manuals, and so on) and practical knowledge (how to do what and when, captured originally in rituals and ritual texts, now cashed out in, for example, instruction books), but also deontic knowledge (what to do and what not to as explained by myth, religion, and law, but also, say, by exemplary tales, songs, and narratives). Already at this point it is important to emphasize that the relation between situations and genres is by no means one-way or simple. It is, in other words, not always the case that a genre develops and stabilizes as a straightforward response to a requirement proper to a situation. As shown by cases where the situation is one of theoretical or deontic knowledge transmission, text and situation codetermine each other, in that the stabilization of the genre is concurrent with the constitution of the institutions and the social agents entitled to make use of it. Situations are, of course, not describable only in terms of the object of communication and transmission. Humans gather for many other reasons, and situations are therefore also suitably characterized in terms of their eliciting conditions: for example, the significant temporal events that may cause both intense and more and more stylized ritual, linguistic, and textual behavior, for example in the shape of invocatory, evocatory , and convocatory commemorations and celebrations of both calendar events (solstice, first plowing, first harvesting, and so on) and individual and social events (birth, death, passage to adulthood, significant achievements). In these cases it is the situations that cause the genres to exist. Situations are also determined by their social structure or intersubjective power relations. The semiotic interactions that take place in them and, thus, genres, as the stylized outcome of these, are therefore amenable to description in terms of the different relevant aspects of such social relations; certain genres are monological, as we shall call them below, in that they do not call for or imply any reply (e.g., verdicts, sermons, speeches, laws, narratives), whereas others are dialogical by nature, meaning that it is a property inherent in them that they enable or are susceptible to replies, responses, objections, and so on (op-eds, letters to the editor, editorials, scientific papers, applications). 1 Yet 1 It is, of course, always possible to respond to or contest any text. If Mr. Jones receives a letter from the army convoking him to the draft, he may very well respond by sending back a letter telling the authorities in question that he does not feel like it, or what about next week. But it is just not appropriate for reasons pertaining to the monological character of that subgenre of letters. The Double Feedback Loop and the Parameter Theory of Text Genres 19 other genres hinge substantially on the role incarnated by the speaker or writer, either as a singular (charismatic) individual (presidents’, prime ministers’, kings’ and queens’ speeches) or as the representative of an institutionally warranted function (pastors and sermons, judges and verdicts, teachers and school programs). Also, in these examples, the situation and the genre seem to co-determine each other; for instance, in the State of the Union Address the act of presenting the address is dependent on the genre and the genre is dependent on the situation. It follows from this that to look at genres in a synchronic cut - i.e. with a view to capturing such and such formal traits of a given text type - is in a way to miss some essential aspects of genres. The main problem with such an approach is, of course, that genres develop under social and cultural pressures. As is manifest from the above, the notion of genre that we marshal is inseparable from the organization of social life. This way of looking at genres as determined by the societal development is not new. Here and to different degrees we follow paths opened by Bitzer (1968), Miller (1984), Bazerman (1994, 2012), and Bawarshi (2000). This type of approach is also represented in other contributions to this volume (Togeby, Engberg and Maier; cf. also Borchmann’s thorough discussion). Our take on this is particularly close to Bawarshi’s suggestion that “writing is not a social act simply because it takes place in some social context; it is social because it is at work in shaping the very context within which it functions” (Bawarshi and Reif 2010, p. 70). We agree with Bawarshi notably as regards the conception of genres as emergent phenomena which stand in a relation of mutual scaffolding with the situations from which and in which they emerge. As regards Bitzer (1968), we take over his notions of “situation” and “exigency”, notably that a situation is a natural context of persons, events, objects and relations possessing an exigency which invites utterances [i.e. a given text type or genre]. Just like Bitzer, we consider genres as text types that emerge as “fitting responses” to the demands of a certain situation. However, whether, as Bitzer claims, situations are materially recurring complex wholes, displaying systematic defects (thus calling for “fitting responses” or genres as textual “amendments” of these defects) or social constructs, as Miller would have it, is not essential for our purpose, even though we are not particularly sympathetic to Miller’s constructivism (see Borchmann in this volume for a discussion of these issues). The exact ontology of the situation is not important here. What counts is that generic social situations exist, constructed or not. Similarly, the dynamic interaction between text type and situation is more complex and less obviously one-way than the notion of “exigency” suggests. Legal discourse and legal discursive praxis developed (in the shape of laws and verdicts) from decisional situations, but probably in ways that did not immediately in any way fill in any gap; we are more likely dealing with a text type (the judicial genre) which initially is embedded in exemplary narratives (legends) and chronicles of past events. Furthermore, the individual and social responses to these events as Svend Østergaard / Peer F. Bundgaard 20 well as the use of these text types in situations which, in virtue of their decisional character, are different from other contexts of use, may lead to a progressive stylization of the text type which in turn feeds back to the situation, operating a gradual situational stylization (with distribution of prototypical roles, institutions, and a more and more immutable script). This is, indeed, why we characterize the relation between situations and genres in terms of a “feedback loop”. Finally, we would like to stress that the relation between situations and genres is not simply negative. Genres do not only come about as responses to given “exigencies”, “needs”, “demands” in a situation; they also develop as a positive result of technological advances - as is well illustrated by the invention of writing techniques (see also Engberg and Maier in this volume). In this text, however, and somewhat following the traditional procedure, we focus on the negative situational constraints on the emergence of genres. If genres are types of language use that typically emerge in connection to a situation, then it follows that by properly understanding and characterizing the nature and requirements of the typical situations of the social world we also gain insight into the kind of texts or genres they call for and the way in which the stabilization feeds back into the situation. There are, for example, strong reasons to believe that the existence of writing and, thereby, the possibility of having external representations of knowledge and information has had a huge impact on social and cultural organization. So, on the one hand, the existing text types have decisive import on the political and institutional organization of society; on the other hand, changes in economic and political conditions demand new or refined versions of existing genres. This is, by the way, yet another reason why a given genre should not be approached synchronically, i.e. independently of the genres that have preceded it. This dynamics between the production of texts and the politico-economic organization of society has gone on for the last 3000 years, and the text types we have today are therefore continuously connected to the very first appearance of genres; each new genre is, as it were, a bifurcation of an already existing genre. We will now turn to a more detailed inquiry into the social dynamics of the emergence of genres. 2 The dynamics and the emergence of genres So how do genres more specifically come about? Consider the first written texts. The development of writing itself is obviously bound to social exigencies. Bazerman (2012, p. 380) argues that the driving force of the invention of writing was the need to have an external representation of information, which basically amounted to different sorts of agricultural records. A further refinement of the writing system is connected to political expansion with the need to communicate reliably over large distances, the need to make financial The Double Feedback Loop and the Parameter Theory of Text Genres 21 reports, and so on and so forth. In the old Mesopotamia trade and administration became too big a load for the human memory system, and the need for a permanent and dependable way to represent transactions and record administrative decisions led to the development of a writing system (Robinson 2003). However, once developed, writing rather quickly became a technology for the economic and administrative power. It became possible to expand the use of writing beyond those problems which it was originally intended to solve. The administrative power is now in a position to make records of taxes, of information related to control of the population, and so on and so forth. We can therefore see the development of genres in the light of the text types that emerged along with the writing system which originally had a rather unitary form: the administrative record which can be considered the oldest and the most permanent genre throughout the history of writing. Even today it is probably the most extensive genre in terms of the number of texts produced. This simple example of a genre illustrates the underlying assumption in this paper; namely that any new genre is developed in a dynamic feedback loop between already existing genres and societal need. However, we here use the word need in its broadest sense so as to include not only practical needs like making financial accounts, but also the need to manipulate, to create political and ideological cohesion, to establish and reinforce interpersonal rapports, and so on. Also for oral genres - whatever the first ones were, probably mythological narratives compressing different moral, didactic, and cosmological purposes - we propose that, in principle, their emergence can be ideally described by a bifurcation graph with a source genre. Each node in the graph represents the unfolding of a new genre either as a response to a sociocultural need or as an intrinsic development of an already existing genre or both. A case where the bifurcation is punctual is in the State of the Union Address, which George Washington was the first to make in 1790. That this is indeed a bifurcation of already existing genres can be seen from the fact that Washington copied the form of the speech from the British King’s Speech to parliament; see Bawarshi (2000, p. 341) for details. Why this is a new genre appears from the fact that the State of the Union Address is delivered by the president on his own behalf, whereas the King’s Speech is made on behalf of the government. Moreover, the King’s Speech, contrary to the State of the Union Address, opens the legislative period of the parliament and is somewhat determined by that. A case like this, where a genre emerges as a result of a punctual event, is nevertheless rather exceptional. As a rule, genres develop from fluid interactive dynamics stretched over time. Another factor which makes it complicated to track down the exact emergence of a genre is what Bazerman (1994) calls the system of genres; a given genre might only exist due to its position in a more comprehensive web of text types. Bazerman illustrates this point in a detailed study of the patent genre. More generally he argues that the legal system con- Svend Østergaard / Peer F. Bundgaard 22 sists of a system of interrelated genres. The examiners, the judges, the legislators, the defenders, the opponents, and so forth all produce documents in different but mutually dependent genres; for instance, psychiatric reports prepared as part of sentencing a criminal is a genre - different from an academic article in a psychiatric journal - which only exists in relation to the other documents prepared in legal procedures. All this supports our view of genres as elements of a complex dynamic system. When we talk of genres in terms of a bifurcation graph system, we therefore simply mean to stress that in our view no genre develops independently of already existing genres. Shortly, we shall come to grips with the specific components of the relation of co-scaffolding that are relevant for a description of text types. Before that, we will devote a subsection to a crucial aspect of our argument, namely the one concerning the dynamics of genres. 2.1 The dynamics of genres: the double feedback loop As to the dynamics, it consists of two loops: one that is intrinsic to written texts, and one that is embedded in the politico-economic dynamics of society. Fig. 1: The double feedback loop model for the emergence of genres 2.1.1 The textual loop Concerning the intrinsic part, we follow the assumptions from usage-based linguistics, which state that the form of language develops from the use of language. In the same way, the use of texts in given situations causes the Modification Genre Texts Situations Propagation Propagation Response The Double Feedback Loop and the Parameter Theory of Text Genres 23 emergence of stabilized types of texts (genres) related to types of situations. Part of the driving force behind this statement is what is known as the alignment principle: a writer’s output relative to a given situation is determined by previous input; see Garrod and Doherty (1994) for experimental results that confirm this principle on the level of language production. This is one of the reasons why genres develop from other genres; they are output alignments to previous inputs which transform or amend the latter in order to make them meet the requirements of a new situation. A case in point was the abovementioned State of the Union Address, where George Washington, facing an entirely new communicational demand, initially looked for sufficiently cognate predecessors he could adapt to his present needs (here, rather paradoxically, a King’s Speech to the parliament served as a convenient template). This illustrates two things: a. the drive to align with previous forms in any linguistic production; but, as we have just seen, b. this drive is also in a dynamic equilibrium with the drive to make variations. Another example is the news article analyzed in Borchmann in this volume. Besides the traditional news articles, which are communicative and have trivial frames, and the explanatory news articles, we have what is termed a news analysis whose frames are informative. In our approach we consider the news analysis as a slight form variation of already existing journalistic forms (in this case, the variation consists in the introduction of an informative framing to the news report). The motivation for this variation is undoubtedly very complex, but two factors seem to be decisive: the growing complexity of the political and financial decisionmaking, which requires an explicatory form, and the growing competition between the news media, which requires a more glaring journalistic form (for details and references, see Borchmann in this volume). This process illustrates the response arrow in fig. 1. The news analysis with an informative frame is a response to a complex media situation. However, we, as well as Borchmann, would also emphasize the emergence of the news analysis as a genre. This relates to the arrow we have called propagation and is explained in terms of the alignment principle. When texts are produced and endowed with a given relevant characteristic (e.g., informative framing), then this is imitated by other writers in the field, and this form in turn propagates in the community until it stabilizes as an acknowledged form. At this point, we have a new genre. If the use of texts leads to a stabilization of types of texts, i.e. genres, then, on the other hand, the new stabilized forms work as norms for the production of texts in a kind of bootstrapping process. This relates to the arrow we have called constraint in fig. 1, meaning that the emergent genre constrains the text production once it is established. Søllinge (1999) (see Borchmann in this volume for more details) provides a good example of this, when she describes the transformation of the canonical newspaper text from being a communicational platform of a political party to becoming an independent witness and commentator of political and societal events, and she further shows how this Svend Østergaard / Peer F. Bundgaard 24 new form now requires both formal training and textbooks. In other words, the emergent new form becomes a norm for how future journalists should write their newspaper articles. This is, then, our textual feedback loop which captures the dynamics between text writing and genre; certain text types come about (related to some sort of new societal need/ resource), then they propagate in the community until they are recognized as a genre, and in turn the genre now serves as a constraint; i.e. it becomes a norm for any future instantiations of that kind of text. 2.1.2 The societal loop The other dynamic loop that determines genres is the one between the societal institutions and texts or, put in another way, between the (experiential) situations in society and texts that respond to the situations. This dynamics is also a feedback loop, but it differs from the textual one in that we are not dealing with a case of emergence. The text does not emerge from the situation; it is a response to it. And the situation neither emerges from nor is constructed by the texts; it is modified by them. The dynamics of this relation is not only complex, it also comprises various different aspects related to the content/ nature of genres, on the one hand, and their form, on the other. Let us first consider the feedback loop from the point of view of the content or nature of a given genre. It seems safe to say that law texts as a genre have developed out of a need to enhance both the accessibility, the homogeneity and general applicability of the rules governing or establishing social facts, the right and wrong social behavior, and the sanctions of the latter. Now, law texts do not develop in a void or in some societal interregnum; on the contrary, they cannot stabilize as genres, unless they co-develop with the social institutions in charge of their interpretation and application. This simple example illustrates what we mean by the notion of feedback loop (or co-scaffolding); the development of the text (and thereby the genre) feeds back into the situation from which it originates, not simply by adding a hitherto nonexistent text type to it, but because it co-appears with the institutions and the social roles which make use of it (the same goes for religious texts and churches, scholarly texts and academia, and so on). This is, thus, a key property of the dynamic relation between texts and situations, but as suggested it is not the only one. Yet other aspects concern the relation between, say, the exigencies of a situation and the form of a text. Our take on this is similar to the one we find in cognitive linguistics (Talmy 2000) regarding the relation between the linguistic form and the experiential world. The assumption is here that the form expresses a motivated conceptualization of basic experiences, so that - to take a very trivial example - the focus of attention in a referent scene is systematically specified by the grammatical subject in a sentence. To be able to exist, any social world, whether human or animal, has to rest on recurrent patterns of behavior and situations. It is gen- The Double Feedback Loop and the Parameter Theory of Text Genres 25 erally assumed in cognitive linguistics that a recurrent pattern is subject to a conceptualization which takes the form of a schematic representation or an idealized cognitive model of the situation (Johnson 1987, Lakoff 1987). This, then, transposes to texts and situations; insofar as there is a textual response to the situation the form of the text will itself be motivated by the situation. Instructions could serve as an example. Here we typically find passive forms: “The asparagus are cleaned,” the purpose of which is to delete any reference to a specific agent, and the use of the passive participle: “… having fried the onions they are to be …,” the purpose of which is to mark the full scanning of a previous process before the onset of a new one. This motivated relation between situation and text is discussed in detail in Togeby (this volume). Here it will suffice to mention that motivation does not mean that the form is determined, i.e. that the requirements proper to a given situation trigger the emergence of one and only one text type. There is a one-to-many relation between a situation and the types of texts that may be satisfying responses to the situation. 2 However, due to the intrinsic feedback loop between text types and text production the types will in a given cultural context stabilize in one type. That is, again, when a genre emerges. The relation of a text’s form to the situation also concerns the question of whether texts are understood bottom-up, i.e. in virtue of their linguistic form, or top-down, which would imply that we need to know the situation and, thereby, the genre beforehand in order to understand the text. Now, since we live in a dynamic flow where we are constantly concerned with predicting what happens next, we are, as a rule, already in a situation where we know the genre before we actually approach the text. It is indeed difficult to imagine encounters with texts outside a (contextual or discursive) framework that predetermines their genre. On the other hand, we dislike confusion, so we want texts that refer to the same situation type to have the same overall form (or to display certain recurrent features). This, together with the alignment principle discussed above, is why genres are characterized by possessing certain (prototypical) local characteristics, which can differ from one cultural context to the next. Take obituaries. Danish obituaries are typically weakly codified. It is rarely explicitly mentioned that the person they are devoted to has actually passed away, but they typically do have the initial form: “Axel Bolvig was a man who …” X was a man is the only linguistic form that tells us that the text is about a deceased. The rest of the text could be a celebration of the birthday of a still living person. In America, for example in The New York Times, we find a more rigid construction used for all obituaries: A. [NAME], [FOREMOST QUALITIES], has died [LOC]. B. [AGE]. C. [CAUSE of 2 Spanish, for example, favors the use of 1st person plural in recipes, instead of the passive form, whereas French has a faible for the infinitive, both with a view to producing exactly the same generalizing effect. Svend Østergaard / Peer F. Bundgaard 26 DEATH], confirmed by [CLOSE FAMILY MEMBER/ SPOKESMAN]. D. [BIOGRAPHY]. As in the following: Gil McDougald, the Yankees’ versatile All-Star infielder who played on five World Series championship teams but was remembered as well for a single atbat resulting in one of baseball’s most frightening moments, died Sunday at his home in Wall Township, N.J. He was 82. The cause was prostate cancer, his son Tod said. [Followed by an account of the important moments of his life] (The New York Times, November 29, 2010). Similarly, but somewhat less costly, The Guardian has defined its own fixed incipit for obituaries: [NAME], [who has died aged X], [VERBAL PHRASE], followed by [BIOGRAPHY]. As in the following from November 29, 2010: Maurice Murphy, who has died aged 75, was the leading British orchestral trumpet player of his generation. Now, even though local specifications of the genre (here the obituary) seem to distribute on a continuum from the less parsimonious, most costly use of them (The New York Times) to minimal local encoding (The Guardian) and up to the limit case where only a loose form like X was a man provides the relevant frame, the fixed character of the templates used, whether costly or thrifty, arguably transforms them into discursive constructions with a fixed meaning: whatever text which follows the present incipit counts as an obituary and has no other relevance than to commemorate the deceased person and the significant moments of his life. A genre is thus a conceptualization of types of situations in which the texts are used and the genre stabilizes in a form that is motivated by the situation; i.e. the relation between a genre’s form and the situation is motivated as suggested above, but it is not deterministic. What is then the external driving force for the production of a text as a response to the situation and, thereby, (in the textual loop) the emergence of a genre? There is no general answer to this question. The genre of patent presented in Bazerman (2012) is illustrative as regards social dynamics. As noted by Bazerman, the patent genre develops through time. In Renaissance England the patent was a privilege given by the king. In its origin there was an economic drive behind the new genre; for instance, someone could be asked to invest in the production of some needed commodity and in return be given a patent, meaning that nobody else was allowed to produce the same commodity. However, in England the genre was inseparably tied to the concept of royal favor, which opened for abuse, so that, in the end, someone, favored by the king, could get a patent for, say, playing cards, meaning that nobody else would be allowed to produce playing cards. This then became a hindrance for economic development contrary to the original intention behind the patent system. This development might illustrate the societal loop in fig. 1; the Renaissance patent was a response to an economic need, but since it was conceptualized as a royal favor, it was distrib- The Double Feedback Loop and the Parameter Theory of Text Genres 27 uted independently of any economic consideration and thereby worked against its original intention; i.e. it modified the economic situation. This changed after the English civil wars (1640-1660) and the execution of King Charles I where all forms of monopoly were abandoned, except in the cases where an inventor had produced something that could be for the benefit of the general economy. Bazerman shows how patents even today are variations of these earlier forms, and this again illustrates the point that there exists an intrinsic dynamics in the development of genres: text forms that address new situations emerge from other genres as copies, variations, inversions, and so on. As regards the external dynamics, the case of patents shows how this genre first worked as a way to express political power and then became an instrument in the deployment of economic growth. However, and we insist on this, the relation between societal need and text forms that address this need is not unidirectional; i.e. it is imprecise to say that the need causes the genre to emerge. Rather, need and text form stand in a relation of mutual scaffolding; they influence each other so that in a given socioeconomic situation already existing genres might be recycled, as in the case of the State of the Union Address. Furthermore, in many cases it is not clear what the need is, or rather the need could exist at another level than the one where the genre emerges. For instance, it seems plausible that emails developed as a response to the need for fast two-way communication; that, however, was already taken care of by phone calls, so clearly emails as a genre (what characterizes this genre is its very short format) are rather the result of technological progress, which initially was not intended for everyday communication, but for military use only. This is indeed trivial, but the interesting thing is that it strongly modifies the situation of two-way communication in the sense that hardly anybody writes letters anymore. 3 Genres and parameters As mentioned in the beginning, we want to describe genres in terms of parameters. One question is, then, how the dynamic account of genre development fits into the more structural parameter analysis. When a given system develops according to a dynamic model it is because of a change in one or more parameters; in an economic model, for example, a parameter could be inflation or deflation, in a physical model it could be time, pressure, etc. These are not the parameters we have in mind. Our parameters are not internal to the system but are external conditions for communication. Here is an example of an external constraint on text types, namely what we will call the “temporal scope” of communication. In oral communication the message only exists in the moment of speech, except, of course, for cases where it is kept in the memory of the receiver. Similarly, only those within hearing distance of the speaker have access to the message. In this sense, such a type of communi- Svend Østergaard / Peer F. Bundgaard 28 cation has a very limited temporal scope, meaning that its duration is not everlasting. With the advent of writing systems this external condition for communication changes, but it is still so that the default situation of oral speech prevails in certain text types. For instance, a written exam paper only exists within the temporal scope of its evaluation, and then it can be thrown away, whereas Newton’s Philosophiae Naturalis Principia Mathematica has an indefinitely long temporal scope. Another example of a parameter which is both external to and characteristic of text types is “accessibility”. If the president of the United States writes a diary, this is a text with restricted accessibility to only one person, whereas the State of the Union Address is accessible to everybody. In other words, the parameters represent general conditions for the dynamics of text production, and in that regard they can be used to outline a typology of text types. It follows from this that we are not bothered by any attempt to present a logical description of genres. Instead of considering necessary and sufficient conditions for membership in a given genre, we will consider the determinant parameters in the development of genres. There are two important properties of parameters in our use of the term. First, and as already mentioned, parameters are not intrinsic features of genres as text types; rather they pertain to those situations in which genres are used and through which they serve an optimization function in the dynamic process we outlined above; for instance, the restricted temporal scope of an exam paper is not inherent in the produced text but in the situation (notice that the very same text can be recycled as a scientific paper and, thereby, occupy a different position on the temporal scope parameter). The next important property of parameters in contradistinction to a feature-based determination of genres is that texts can manifest the parameter or be determined by it in various degrees. In other words, a parameter is not (necessarily) simply an either-or two-pole affair; text types may distribute on a continuum that leads from one end of the parameter pole to the other. In the following we shall propose a list of parameters, which externally control the nature of a situation and which thus appear as possible dimensions of genres. Notice that the examples of genres given to illustrate the impact of these parameters cut through levels of categorization, so that a text type of a basic level (say a “letter”) is, concerning the communicative situation, determined in the same way as the subordinate category: letters sent to the tax authorities, or a letter sent as an application for a job. However, it also sets up differences at the same level of categorization, so that the subcategory of response letters to job applications or authorities’ response letters to questions are determined differently in the communicative situation than letters asking questions to a tax accountant. In other words, the impact of the parameters does not follow general categorization principles. The Double Feedback Loop and the Parameter Theory of Text Genres 29 3.1 The parameters Parameters are factors of relevance which define the aspect under which a situation is attended to. As such, they have import on the types of texts produced in given situations. As we have already seen, situations can, for example, be determined with respect to the events that take place in them: births, anniversaries, religious passages, death, and so on. Each of these events may call for specific types of texts (e.g., celebratory speeches, commemorative discourses, eulogies, and obituaries). Even though parameters are extrinsic to texts and text types, they have decisive import on the generation and classification of such text types. Genres can therefore be classified in accordance with the way they are controlled by the situational parameters we will now proceed to describe. Type of exchange: monological or dialogical exchange One parameter is the degree to which a text is a piece of one-way communication or whether it partakes in a dialogical exchange. In oral communication verbal interaction is dialogical by nature, but it can be a special property of a communicative situation that a person is monologically in charge of the verbal interaction and the rest of the participants appear as addressees in different respects. This will be the case, when the monolog is related to a power or to the enforcement of a power with which the hearer is not expected to interact. The existence of writing changes this dramatically, partly because of the trivial fact that it is now possible to communicate over time and space, but also because the ensuing multiplicity of text genres do not simply fall into one or the other of the two types of communicative situations. What counts here is, then, the degree to which a text is meant to be part of a dialogical exchange: one-way communication versus two-way communication, but considered as a graded scale. In one end of the continuum we find the diary (eminently monological, in that the addressee coincides with the utterer), followed by verdicts, statesmen’s speeches to the nation, speeches tout court, literary writing, and even some types of letters as mentioned above. Intermediary text genres may be religious invocations (prayers) and some kinds of letters whose temporally delayed and insecure response character makes them differ clearly from verbal interaction. At the other end of the continuum we find genres that intrinsically call for response, in that they are themselves responses to states of affairs, claims, and previously advanced opinions (letters to the editor, op-eds, editorials), ending with epitomes of two-way communicational text types (themselves resulting from technological changes in society): (many) emails, blogs, and even things like Facebook updates. As regards this parameter, the default case in oral communication is the dialogical one. The dynamics that makes it a determining factor is related to the existence of power relations in society. Also animals have power relations, but they are bound to the here and now. For humans a power declaration Svend Østergaard / Peer F. Bundgaard 30 extends beyond those who witness it. For instance, a law agreed upon in parliament extends beyond those present. The law is thus the prototypical monological discourse, and the prototypical situations connected to this parameter are the institutions supporting it: law courts, of course, but also the parliament. There is a difference between the discourse inside the parliament, which might be dialogical, and the outcome of these deliberations, which is monological. Our assumption is that the early genres, the administrative text, and primarily the narrative-mythical-philosophical-religious text are monological in nature. Two-way communicative genres seem indeed to require the existence of social spaces for deliberation and debate and, for the written genres, the emergence of democratic, liberal societies with freedom of speech and a culture of public debate. The intricate relation between monolog and dialog can be illustrated by the basic level category letter. For instance, you can write and ask the tax authorities a question and expect to get an answer; you can then respond to the answer, often because you find the latter unsatisfactory (all this is typically dialogical). However, in certain countries you can pay an amount of money and get a binding answer. This answer is less dialogical, since you cannot respond directly to it, but only indirectly by filing a complaint to the tax board of appeal, which then issues an answer that puts an end to all exchange. So even if we deal with the general category of letters, the institutional embedding makes us move from a dialogical to a monological value of the same genre. Accessibility The second parameter we will consider is accessibility, meaning how many in a community have access to the texts in a given genre. In one extreme of this parameter we find texts that are accessible to everybody (like the president’s speech to the nation), while in the other it is only accessible to one (for instance a letter). In-between these poles a text can be accessible to a limited number of people, like an administrative report or a samizdat. The general accessibility of written texts is, of course, one of the defining features of modern, open societies, so, in opposition to the oral case, this value is the prototypical feature of written text types. In the oral case, there is an overlap of this and the previous parameter. This is not the case for written texts, because a text can be publicly accessible and yet be part of a dialogical structure, for instance a reader’s letter to the editor or a political commentator’s column. In the case of oral communication, a leader of a tribe can speak to everybody, or he can speak only to the members of the ruling group without revealing to the general public what they are talking about. Even if the prototypical text is accessible to the public today, in the genesis of written genres arcane discourse has probably been the norm, mainly because, in general, people were not able to read. In Egypt the art of writing and reading was delegated to a small group: the scribes. The purpose of their writing was administrative and The Double Feedback Loop and the Parameter Theory of Text Genres 31 had no relevance for the public. Ornamental writing, which served to consolidate the power relations, was of course generally accessible to those who could read it,. So genres that were publicly accessible served the purpose of consolidating the religious, political, and economic power, or more generally they served to consolidate the ideological underpinning of the society. Therefore, the values of this parameter have probably been relevant from the very beginning of the emergence of genres; the texts that deal with the practical aspects of power administration have been inaccessible, except to a few persons, whereas the narrative-mythical-philosophical-religious text was accessible to everyone or, when written, those who could read: for instance The Epic of Gilgamesh, but also various inscriptions with a narrative-mythological content. With the invention of writing it became possible for the administrative power to reliably issue decrees to remoter parts of the empire. The prototypical situation that makes this parameter a decisive factor is one where administrative units are separated in space. In general, the text is a way to communicate exactly the same meaning to people separated in space. The driving force behind the emergence of newspapers, as described by Borchmann in this volume, was the need to communicate the same political message to as many as possible. This led to the omnibus press, but with the Internet we are in a situation where all members of a given society in principle can read the same news at the same time. This reduces the need for the traditional newspaper and, therefore, Søllinge’s (1999) and Schultz’ (2007) analyses of the news market as developing into more specialized forms are an eminent example of the relevance of this parameter (cf. Borchmann in this volume). Social role The third parameter is the social role ascribed to the writer of the text. The most essential elements in this parameter is whether a person expresses himself as the incarnation of a publicly known political, religious, or otherwise institutional function, or whether he expresses himself in the role of a private person. For instance, recently the Danish Minister of Finance sent an email to the Scandinavian Airlines staff; subsequently, he was asked if it had been written by the minister or on his own, personal account. In the latter case it would be an expression of a private opinion, in the former it would still be an expression of opinion, yet not a private one, but one that manifests or may manifest a government’s take on the issue in case. As regards the origin of written genres, the first value of this parameter was predominant, since the first texts served the already mentioned administrative or political purposes; thus, writing would never be an expression of private opinion. This also goes for the narrative-mythical-philosophical-religious text whose origins would generally be some religious institution. Today, prototypical texts, which have no or little institutional basis, are found in newspapers and in fiction. So if the prime minister speaks at a press conference it is in the role of prime minister, but if I Svend Østergaard / Peer F. Bundgaard 32 write a letter to the editor of a newspaper, I do so as a mere citizen. However, this is not a matter of either or. Also here the institutional role can be more or less manifested. A textbook in physics used for teaching is written by a trained physicist; i.e. it is written by someone who is an acknowledged physicist. Editorials in newspapers represent the newspaper (even if the writer expresses personal opinions, this is done on behalf of the newspaper). We can again look at the analysis presented by Borchmann in this volume to illustrate the point. The professionalization of journalism in the beginning of the 20 th century represents a formalization of the journalist’s role, who now, like the physicist, writes on behalf of this new institution. The default case of oral communication is the non-institutionalized discourse. The dynamics that unfolds this parameter in the direction of more institutionalized discourse is the division of labor, the delegation of power, and the general compartmentalization of society. The prototypical situations where we meet the genres that express this tendency are thus the various institutions that maintain and control the circulation of opinions and information in a given society. No doubt the first text types were of this type. For a text to manifest the opposite value, it requires a discursive situation where the individual can be the locus of the discourse. We have quite early examples of this, like when merchants write about travel experiences and the like. Social function The fourth parameter we will call social function. In the theory of evolution there are different opinions on what was the determining function for the development of oral language. The theory we ascribe to here is the one presented in Tomasello (1999, 2008, 2009). Here the origin of communication is based on cooperative activity leading to the development of cooperative communication which, among other things, includes the Gricean notion of communicative intention (Grice 1957). This aspect of oral language would include basic speech acts like requests, giving information, and making obligatory statements. However, oral language has another function which is to secure the emotional bonds internally in the group, thus distinguishing it from other groups. This is done by narratives, gossip, anecdotal sharing of emotions, preferences, and so on. This kind of language activity is by some evolutionists compared to the grooming activity amongst non-human primates and is considered as the basic evolutionary drive for the development of language, because when group size becomes too large, grooming is no longer an efficient method to secure the social bonding between group members. Tomasello (2008), however, considers this aspect of oral language as a latecomer, since telling narratives requires a more sophisticated syntactic language structure - especially as regards the capacity of establishing internal relations between expressed events - than simply informing and requesting in relation to a cooperative activity. This is not the place to go into this discussion. We only use Tomasello’s distinction to set up two separate values for the The Double Feedback Loop and the Parameter Theory of Text Genres 33 function of language. We are especially interested in the notion of narratives as serving the function of sharing experience in a group and, thereby, securing a shared world view, a shared emotional response to external events, a shared set of values, and so on and so forth. Our purpose, as far as this parameter is concerned, is essentially to ground literature in a basic and necessary human activity. This activity is necessary because the development of the communicative competence also implies what Tomasello calls reciprocity; i.e. humans have to deal with a second order of representation of their activity. Humans have to think of how others think of their activity. This reciprocity is extended to the group level, so the group as a community has to think of how to represent their activity from an external perspective. This naturally leads to narratives, which then become the group’s way of representing itself (its values, its principles governing social behavior, and its cosmology). We see this as the origin of verbal art at large. When the group size grows, more urban communities or cities emerge and the social organization is no longer transparent for the individual. Hence, the conditions for narratives change. However, oral narratives continued to be the norm, but now as a local sharing of emotions, sometimes or perhaps often even with a subversive point, cf. jokes in The Soviet Union, but also narratives like The Canterbury Tales, Decameron, and so on. Literature as we know it today is a rather recent phenomenon, depending on the invention of printing, the extension of the literate class, and a specific socioeconomic organization where a few people are able to live on writing or have enough spare time to indulge in such activities. In short, we consider the notion of literature as an emergent phenomenon developed over thousands of years. In our account of the parameter social function we will at one extreme place literature at large (i.e. not only in the modern sense) with a huge emotional and identification enhancing function. At the other extreme we have instruction manuals which solely have a practical purpose for one individual. Of course, to consider “function” as a gradable parameter is in a sense a theoretical construct, but the idea is that a text can be more or less practical and, to some extent, have a character of sharing an emotion, establishing bonds between subjects, and participating in enhancing social cohesion. Consider, for instance, an ideological text issued by the communist party; this text both serves the function of presenting shared opinions for the members of the group and presenting a “manual” for how the group members should act in the world. Similarly, a novel written by an author endorsed by a given regime can serve or can be expected to serve the same educational function as a more flatfooted propagandistic text. In other words, social functions may be negative (when texts serve the purposes of oppressive regimes) and may operate at the larger level of society as such. The basic genres at the origin of the idealized bifurcation graph are textual versions of the extreme values of this parameter: the politico-economic administrative text as a genre with implications for the physical organization of Svend Østergaard / Peer F. Bundgaard 34 life and the narrative-mythical-philosophical-religious text with implications for the mental life of the participants in the community. Both aspects of this parameter are present in oral speech, which, on the one hand, is used by individuals to inform each other about potential problems, obstacles, or places for collecting food, etc. and, on the other hand, is used to gossip, tell stories, and so forth. The invention of a writing system just exploits this double structure of language, so that we in a way need two types of situations as the origin of two types of genres. For the practical use we can refer to the harbor as the locus of trade. Trade exploded after the Phoenicians invented the practical alphabet writing, which could be learned by every merchant and then used for making trading records and other types of documents. For the ideological use of language we can refer to the temple with its inscriptions and oral performance of texts. So it is the dynamics of economics as well as political power that this parameter unfolds. It is important to note that, according to our approach, any text has a use, so this parameter could also be described according to the extent to which the text has a direct implication for our bodily interaction with the environment or whether it is more conceptual. An instruction manual has a direct impact on behavior, a tourist guide has the same implications, a moral admonition less so, a chronicle in a newspaper even less so, a piece of literature almost none, and a scientific article none at all. Temporal scope The fifth parameter we will simply call temporal scope, whereby we mean scope of relevance. As we have already suggested, situations often take on a specific character relative to given events in time (which may simply cause these situations to exist): deaths, births, anniversaries of all sorts, and all other kinds of deviations from the expected take on a salient signification, which elicit general kinds of verbal responses. In other words, a whole series of text types are the verbal responses to these salient discontinuities in lived time and calendar time (obituaries, birthday speeches, commemorative texts, news reports, and so on and so forth). Such text types have a very limited scope, since they are only relevant at the time of the event giving rise to them. They obviously pertain to one end of the continuum, the parameter being clearly gradable. As a eulogy for his deceased friend, the bullfighter Ignacio Sánchez Mejías, Lorca’s poem “Lament for a dead bullfighter” has, of course, a strongly restricted temporal scope; as a poem, however, it has none. In the same vein, a news report is only relevant the day of the reported event, which is why we throw newspapers out when we have read them. If in the newspaper, however, there is an article on the worrisome situation of the Danish high schools, then we might cut out the article before throwing the newspaper out, because this might be of relevance in the near future, at least until these conditions have changed. A third possibility would be a novel about the Danish high school, written 20 years after the events described in the novel. This is possible be- The Double Feedback Loop and the Parameter Theory of Text Genres 35 cause the novel does not have to address the actual state of affairs; its time perspective is different from the newspaper article; it has an unbounded temporal scope. Thus, texts can display different varieties of temporal scope, extending from point event texts (obituaries and so on) to texts with a more or less wide scope encompassing, for instance, biographies, textbooks for teaching, chronicles, historical texts, etc., which we assume have a significance beyond the time of their production, and on to texts with a full temporal scope (religious texts, including Genesis and apocalypse). Probably, most texts have an intermediate scope. An exam paper has only a punctual scope related to the event of evaluation. Many administrative/ legal documents have a restricted and specific scope; this goes for contracts and the aforementioned patent genre, where the temporal scope of a document is restricted to seven years. A social worker’s report on the behavior and doings of a client is only relevant as long as the client is in the situation addressed by the report, etc. We consider the prototypical situation that is relevant for texts with a specific (but not a punctual) temporal scope to be the making of a contract. Here, following Searle, we can say: this text counts as Y in the period X, or in a more general version: this text counts as Y as long as the circumstance X prevails. This parameter also sets the crucial distinction between fiction at large, as unbounded in temporal perspective, and practical texts with bounded temporal scopes. It is important to note that the concept unbounded refers to the intention behind the text; for instance, a law text has a practical use, but its time perspective is intended as unbounded, regardless of the fact that historical changes will make any law obsolete at some point. The default value in oral communication is, of course, the bounded (in this case punctual) perspective, except if we take the memory of the receiver into consideration. If we look at written text types from the perspective of intended use, maybe only religious, literary, and scientific texts have a real unbounded perspective, because all other texts are historically intended to serve a specific function. However, one thing is the writer’s intention; another is the user’s intention. Since information can be stored (on microfilms or electronically), it is possible for a researcher to search in, for instance, old newspapers. The researcher’s intention is, of course, not the same as the original reader’s. It is a kind of meta-intention that in one way or another pursues a research purpose. So the situation that supports the unbounded scope is in any case one where some agents are addressing a mental problem. Cognitive accessibility The sixth parameter we call cognitive accessibility. This has to be distinguished from physical accessibility. Law texts epitomize the text types attached to the situations governed by one extreme of this parameter. They are physically accessible to everyone, but, cognitively and strictly speaking, they are only accessible to those who are endowed with the legal competence to interpret Svend Østergaard / Peer F. Bundgaard 36 them and the right to apply them. This parameter thus captures degrees of (expert) mediation. A less extreme case would be an academic textbook on algebra, which again is physically accessible (for example in public libraries), but only in very rare cases cognitively accessible without some expert’s helping mediation. Somewhere in the middle of the continuum, with one foot in each camp, as it were, we find religious texts which are cognitively accessible to anyone wanting to be instructed by their words (or be entertained by their astounding stories), but which, in certain contexts, require an officially qualified person’s mediation, interpretation, and application to take on their full dogmatic value. Text types that are (supposed to be) unambiguously situated by the fully accessible end of the continuum are instruction manuals at large, and quite close to them are all sorts of verbal art (narratives, dramas, and poems), however obscure and hermetic; they are cognitively accessible in the very same sense that paintings are visually accessible to the eye. Oral communication is by definition unmediated; yet, even in non-literate societies we have people with privileged access to knowledge, like, for instance, the shaman. We shall not speculate on the anthropological reasons for this phenomenon, but it has clearly passed on into the literate societies, where the sacred texts must be mediated by the priesthood. We therefore consider the basic situations of this parameter to be the temple (broadly understood) and the chair. In modern times (from 1100) this continued in the university chair with the teaching of, for instance, Roman law. In this respect, this parameter is closely connected with the parameter of social role. Today the cognitive inaccessibility of texts is, of course, not as much determined by religious discourse than by the complexity of society and the complexity of certain discourse types which require the intervention of experts at different degrees of mediation. Direction of fit Our final parameter aims at capturing those texts that are distinguished by the fact that they bring into existence the very things they are about, that is, texts which are constitutive. We here make a loose use of John Searle’s term of “direction of fit” (Searle 1998) to designate them; we thus say that they have a “world-to-mind direction of fit,” since, when used or produced, they cause a change in the world. Parade examples of such genres are, of course, judicial text types, oral as well as written, like, for example, laws (which bring into existence a certain number of crimes and misdemeanors), verdicts (which turn putative events into established acts), religious texts (which posit the extraterrestrial origin and finality of things terrestrial), and, perhaps surprisingly, all kinds of fictional discourse (which create the irreversible reality of [intentional] objects). There is, of course, no doubt that fictional objects such as Sherlock Holmes and Baker Street 221B do not exist in the same way that constituted crimes and misdemeanors do, but there is no doubt either that such entities have a manner of existence which makes it as true to say that The Double Feedback Loop and the Parameter Theory of Text Genres 37 Baker Street 221B is where Holmes lived as it is to say that the reality of this and that person’s crime was constituted by this and that verdict. Indeed, the family of constitutive texts is, in our view, far less heterogeneous than it may seem at first glance, and the presence of fictitious text types in this class is in our view a far more essential property of that genre than the post-Kantian characterization of it in terms of “disinterested”, “pastime”, or “end in itself” (here we diverge from Togeby’s claims to this effect in this volume). This is so for different reasons. Postor pseudo-Kantian characterizations of fiction in the above terms often rest on the assumption that literary art at large is “defunctionalized” in ways that no other genres are, insofar as it is not framed by institutional interests or driven by institutionally defined ends. Now, first, this is not an evident claim. Poets and authors have long since been appointed to the court, both literally and metaphorically; many regimes have had and still have officially acknowledged artists, capable of delivering the institutionally correct representations of their society; many regimes have recognized, and still do recognize, the subversive potential of certain texts - written for exactly that purpose - and thus their counter-institutional nature, hence submitting them to censorship, prosecution, and punishment. In this sense, literary art scores quite high on the social function parameter. Second, art as “disinterested” art (provided it actually does exist) is a relatively recent affair which, what is more, only captures a limited number of aspects of the artwork. Verbal art has developed, from the origin of human life and through all cultures, in contexts and from situations where it played a role analogous to the role played by ethics and law (see Hogan 2003); narratives, tales, and songs represent, in exemplary, easily memorized and smoothly transmissible ways, basic moral principles, epitomes of right and wrong actions, right and wrong regimes, and rules for the correct conduct of social life at both a local scale (the family) and a global scale (society and its power relations). Such narratives are, beyond any doubt, landmark texts for those who in given situations are to assess the ethical nature of a given action or attitude or to comprehend and solve a given social conflict (at both an individual and a collective level). In that sense, they further the cognitive access to or interpretability of intrinsically complex situations, involving social, moral, emotional, and existential conflicts. Strictly speaking, “direction of fit” is not a parameter, since it refers to a text-internal property. However, the constitutive texts are clearly related to a specific type of social situation, where one or more people are making a decision that has binding consequences for the community or parts of the community. We can call it a decisionmaking situation. In that way, we get a parameter with only two discrete values: whether the situation is one of decisionmaking or not. This explains why texts which have evolved from this original situation, such as literary artworks, still display this either-or property; either they are constitutive and thus fictitious, or they are not. Svend Østergaard / Peer F. Bundgaard 38 3.2 Categories of genres The collection of parameters constitutes a seven-dimensional space. A given text would then ideally be represented by a point, and a genre would make up a connected area in this space. We will now, briefly, develop a couple of consequences of this fact. Take, for instance, a textbook in physics written by Professor Spaulding. This is a monolog, physically accessible to everybody. It is written by Professor Spaulding in his social role as professor in physics, so it expresses an opinion that is supported by the scientific community. It has a practical, social function, and its time perspective is unbounded. It is not a constitutive text, and its cognitive accessibility is limited (the students become specialists because of the teaching). If we now change the value of the first parameter from monolog to dialog, we get something that is akin to the scientific textbook, but more in the direction of a scientific article, which is supposed to be commented upon by peers. Similarly, if we change the situation from not constitutive to constitutive, we no longer have the genre of scientific textbooks but rather something that could be characterized as law texts. Any genre should, in principle, be characterized by some values of the parameters, but no change of values necessarily corresponds to a known genre. It could, though, be a potential genre in the future. In fig. 2 we have illustrated the way in which genres can be typologized according to their position in parameter space: Fig. 2: Long dashes: Professor Spaulding’s textbook in physics. Short dashes: a scientific article. Full line: a law text Social role + Social function + Accessibility + Temporal Scope unbounded Constitutive Cognitive Accessibility + Dialogical Social role - Social function - Accessibility - Temporal Scope bounded Declarative Cognitive Accessibility - Monological The Double Feedback Loop and the Parameter Theory of Text Genres 39 As appears from the above examples, genres grouped together under one parameter may be distinguished from each other according to how high each of them scores on some other parameter. Jokes, fictional art, religious texts, and law texts are all constitutive, but while the cognitive accessibility of judicial texts is extremely restricted, jokes and novels are immediately accessible (religion lying somewhere in-between), and jokes and literary art, in turn, differ in terms of, for example, the monological or dialogical character of their situation of use. The way differences between genres can be accounted for or generated in terms of the controlling parameters is illustrated in fig. 3: Fig. 3: Long dashes: law texts. Small dashes: literature. Full line: jokes. All these text types are constitutive, but their ways part relative to other parameters such as Cognitive accessibility and Monological/ Dialogocial type of exchange Since parameters are situational, changes in society may cause a redistribution of genres in parameter space with an ensuing redetermination of their status. If, for example, a society changes from being secular to non-secular, religious texts will ultimately absorb law texts (on the constitutive side of the direction of fit parameter) and even scientific texts (on the non-constitutive side); if physics, as we know it, cannot explain extra-gravitational facts such as the flying of angels or the ascent of this and that person to Heaven, then physics would be considered as failing in ways religious texts do not (an attested example). Changes need not be that dramatic, but it can nevertheless be predicted, on the basis of our general assumption, that changes in a given societal situation may cause recategorizations of texts as regards their membership of such and such genre. Consider the following, real, example. A civil servant writes a brief to his minister and sends it to him by email. In Denmark, the so-called Freedom of Information Act (Offentlighedsloven) stipulates that such a written document is accessible, upon motivated demand, to anyone (usually journalists or members of the opposition) who wants to monitor or gain insight into the government’s paper trail. As such, it belongs to a genre of texts we could call “public documents”; as regards the parameter “physical accessibility”, it is therefore rather accessible. Now, if the act is amended (as will be the case in Denmark) and a new bill is passed, this kind of text will change status (and thus genre); emails or letters exchanged between the minister and his Constitutive Cognitive Accessibility + Monological Declarative Cognitive Accessibility - Dialogical Svend Østergaard / Peer F. Bundgaard 40 public servants will no longer be considered as publicly accessible documents, but as belonging to the genre “internal memos” or “confidential briefs” with ensuing restrictions on accessibility. They occupy a new point in genre space relative to the physical accessibility parameter, cf. fig. 4: Fig. 4: Changes in societal situations, for example as regards the accessibility of public documents, lead to recategorization of text genres. 4 Closing remarks We have thus identified seven parameters: (1) type of exchange, (2) accessibility, (3) social role, (4) social function, (5) temporal scope, (6) cognitive accessibility, and (7) direction of fit. We have chosen not to include an eighth parameter, even though it is easy to show that changes within this parameter have had and still have decisive import on the generation of text types. What we have in mind is the parameter “(degree of) technological advances”. One reason why we have chosen not to include it is that it is somehow too overarching (or too fundamental); the invention of writing launches the whole business, as it were, not only because it causes the generation of written genres, but also because it launches the whole feedback dynamics between text types and society (the latter is not the same once it disposes of administrational records, law books, and so on and so forth). However essential, though, for the development of genres at large, technological advances may have only limited import on the emergence of the prototypical members of the family of text types (which then would count as another reason not to include it in our The Freedom of Information Act at time t and t' Time t: Time t': Accessible text Unaccessible text Accessible text Unaccessible text Text A X Text A' X Text A = Text A' Text A: Public document Text A': Confidential / Internal memo The Double Feedback Loop and the Parameter Theory of Text Genres 41 list); with perhaps one exception, technological progress mainly or only leads to the development of new subgenres (which is, of course, in and of itself a phenomenon worth investigating): emails, text messages, tweets, Facebook updates, or (cf. Engberg and Maier in this volume) video-supported academic articles. 3 The exception we had in mind is the decisive influence which the invention of the mechanical, movable type printing press had on the development of the novel as a genre (or any long, written narrative); again in two interrelated respects. It becomes possible and therefore attractive to write rather long texts, because, on the one hand, it becomes incommensurably less cumbersome to have such texts massively reproduced (no need for scribes or monks) - in this sense, type printing is a technological device that furthers such, say, inventive or uncommitted verbal praxis (no one would have written the trivial cape and sword novels which got Don Quixote on the wrong track, if the author had had to get each and every exemplar of them transcribed by the local monk) - and, on the other hand, because these objects can now be easily distributed (no remote libraries with restricted access, just the bookshop down the corner), which furthers the creation of a market and a need. It is thus fair to speculate that the invention of printing may very well have had a greater effect on the development of literary art as a genre than sociopolitical changes. This again shows that it is imprecise to say simply that a genre develops because of a need; rather it is the technological evolution that makes it possible to articulate a need and, thereby, the development of the genre. This example aside, and the above subgenres notwithstanding, it is difficult to pinpoint cases where a technological advance (after the invention of writing) has generated a genuine text type. One likely candidate could be the difference between an agreement (verbal) and a contract (written), but in many societies and cultures this difference is shallow, since the former may be as legally binding as the latter. The above considerations suggest that developments in technology are aspects of the physical accessibility parameter, which would explain why it has import on the generation of subgenres, not genres proper. Finally, a somewhat embarrassing question looms large: is this it? Or should other parameters be integrated in our list? This is in fact not really a concern for us. Time will tell, as it were. What counts is that genres are functionally related to parameters which systematically capture relevant aspects of given situations. Genres develop as responses to these aspects, and once they 3 We know of at least one text message novel (Denmark) and one Twitter novel (France). The latter two are non-trivially related to the technology, in that they obey the formal constraints proper to the medium. It is not the case that an author has linearly published his otherwise normally written narrative on Twitter in chunks of 140 characters. Rather, all relevant information comes in structured segments of 140 characters. Svend Østergaard / Peer F. Bundgaard 42 are established they feed back into these situations (on a large scale, when law texts and constitutions are made; on a smaller, but still significant scale, when a given text changes from being a public document to a confidential memo). Our argument is not affected by the exact number of these parameters, which are obtained by empirical abstraction. We do think, however, that our model is correct; i.e. we think that we can suitably typologize any given genre in terms of our seven parameters. Now, we may be wrong. A rather simple litmus test can help us decide whether a new parameter should be recruited; if we cannot generate a well-known text type (which means, if we cannot distinguish between two acknowledged text types) in terms of the available parameters, then our list is incomplete. In that sense, the descriptive model in fig. 3 can serve as a tool for falsification; if any two distinct genres describe the same path through the seven parameters, then we will have to consider adding a parameter. Such an addition would, of course, have import on our understanding of the social world, in that our parameters capture essential aspects of a social ontology, but - and again - it would not affect our main claim, which is that there exists a dynamic relation between social situations and the emergence, stabilization, and change of text types, which in turn affects the social situation they stem from. References Bazerman, Charles. (1994) Systems of Genres and the Enactment of Social Intentions in Aviva Freedman and Peter Medway (eds.) Genre and the New Rhetoric, 79-101. London: Taylor & Francis. Bazerman, Charles. (2012) The orders of documents, the order of activity, and the orders of information in Archival Science no. 12, 377-388. Bawarshi, Anis. (2000) The Genre Function in College English, vol. 62, no. 3, 335- 360. Bawarshi, Anis and Reif, Mary Jo. (2010) Genre. An Introduction to History, Theory, Research, and Pedagogy, Parlor Press. Bitzer, Lloyd. (1968) The Rhetorical Situation in Philosophy & Rhetoric, vol. 1, no.1, 1-14. Garrod, Simon and Gwyneth Doherty. (1994). Conversation, co-ordination and convention: an empirical investigation of how groups establish linguistic conventions in Cognition, vol. 53, no.3, 181-215. Grice, H. Paul. (1957) Meaning in Philosophical Review 67: 377-388 Hogan, Patrick Colm. (2003) The Mind and Its Stories. Cambridge UK: Cambridge University Press. Johnson, Mark. (1987) The Body in the Mind: The Bodily Basis of Meaning, Imagination, and Reason. Chicago: University of Chicago Press. Lakoff, George. (1987) Women, Fire, and Dangerous Things. What Categories Reveal about Mind, Chicago: The University of Chicago Press. Miller, Carolyn. (1984) Genre as social action in Quarterly Journal of Speech, no. 70, 151-167. The Double Feedback Loop and the Parameter Theory of Text Genres 43 Robinson, Andrew. (2003) The Origins of Writing in David Crowley and Paul Heyer (eds.) Communication in History: Technology, Culture, Society, 27-33. Montreal: Allyn and Bacon. Schultz, Ida. (2007) Fra Partipresse over Omnibuspresse til Segmentpresse in Journalistica no.5, 5-26. Searle, John. (1998) Mind, Language and Society: Philosophy in the Real World, New York: Basic Books. Søllinge, Jette Drachmann. (1999) Historien om den politiske journalistik - Et drama i flere akter uden slutning in Erik Meier Carlsen, Peter Kjær and Ove K. Pedersen (eds.) Magt og fortælling, 76-95. Aarhus: Forlaget Ajour. Talmy, Leonard. (2000) Towards a Cognitive Semantics, Massachusetts: MIT Press. Tomasello, Michael. (1999) The Cultural Origins of Human Cognition. Harvard: Harvard University Press. Tomasello, Michael. (2008) Origins of Human Communication. Massachusetts: MIT Press. Tomasello, Michael. (2009) Why We Cooperate. Massachusetts: MIT Press. Simon Borchmann The Perilous Life of a Linguistic Genre Convention A Study of News Analysis within the Danish System of Newspaper Genres 1 Introduction New rhetoric genre theory stipulates substance-form fusions on a variety of levels, including culture, life forms, genre, strategy, speech acts, and language. As the relations between the levels are described as hierarchical, and as highlevel phenomena like culture and life forms are considered dynamic, formcontent fusions on the level of language should also be dynamic. However, within linguistics, the substance-form relations of language have been considered relatively stable; either as arbitrary, systematically-determined conventions, or as cognitively or communicatively-motivated structures. Therefore, there seems to be a disagreement between linguistics and rhetoric genre theorists with respect to the status of linguistic conventions. This disagreement is particularly apparent with regards to text linguistics, since this discipline is based on the assumption that such substance-form relations comprise parts of texts (cohesion) or even whole texts (macro structures). By assuming that such fusions are dynamic, new rhetoric genre theory can explain the emergence, change, and decay of genres. However, this remains very unspecific where language is concerned. Therefore, this disagreement confronts new rhetoric genre theory with the following questions: what is meant by substance-form fusions on the level of language? , how can they be simultaneously stable and dynamic? , and how do we account for them as dependent on the dynamics of high-level phenomena, such as culture, life forms, genre, and strategy? Text linguistics is also confronted with questions: how do we distinguish between different subtypes of descriptive texts or narrative texts? , and how do we explain that specific linguistic features appear in one subtype, and not in another? The aim of the study is to answer the abovementioned questions through an analysis of the relatively new journalistic genre, nyhedsanalyse (news analysis). The overall aim is to present a linguistic text description, which is able to distinguish between genres and explain linguistic regularities within a genre. The approach is similar to Swales’s genre analysis, though it differs from his work in three ways; firstly, by focusing on the developmental history of the genre system, as illustrated by Bazerman, secondly, by incorporating psychological and, in particular, cognitive factors, as suggested by Bathia, and third- Simon Borchmann 46 ly, by integrating ideas and linguistics descriptions of the Danish school of functional linguistics. 1 2 Genre and the new rhetoric The theoretical basis of the presented genre analysis is a critique of Bitzer’s concept, exigence, as an explanation of the relation between situation and rhetorical action. Bitzer’s idea can be summarised as such: whenever a rhetorical situation - defined as, “a natural context of persons, events, objects, relations, and an exigence which strongly invites utterance” (1968) - occurs, there is a motive for rhetorical action, and if the action is a “fitting response” to the situation, it will be constrained by the elements of the situation. This description of the relation between situation and action forms the basis of a genre theory; if situations have specific similarities, the actions motivated and constrained by these situations will also (provided they are fitting responses) display similarities. However, convincing as this account may seem from a functional viewpoint, the problem is that the social conventions involved in the operation of this apparently smooth-running rhetoric machinery are presupposed, or, in any case, omitted from the account. An informed and perceptive critique of this omission is provided by Miller (1984). Referring to Stebbins, Miller points out that, “what recurs cannot be a material configuration of objects, events and people, nor can it be a subjective configuration, a ‘perception’, for these, too, are unique from moment to moment and person to person” (1984: 156). According to Miller, situations are social constructs that are the results, not of perception, but of determination. Before we can act, we must interpret the indeterminate material environment, that is, determine the situation. Miller’s alternative is leaning on Schutz’s concept of type as the elements on which our stock of knowledge is based. Our “stock of knowledge is useful only insofar as it can be brought to bear on new experience: the new is made familiar through the recognition of relevant similarities; those similarities become constituted as a type” (ibid.). Thus, what recurs, according to Miller, is not the material situation, but our construal of the situation as type. Finally, almost echoing Wittgenstein’s private language argument, Miller argues that successful communication requires that the participants share common types, and this is (only) possible insofar as types are socially created. Miller’s alternative represents a very different conception of exigence. Whereas Bitzer’s exigence is described in materialistic terms as an “imperfection”, and as a “defect, an obstacle, something waiting to be done, a thing which is other than it should be” (Bitzer 1968: 6), Miller’s exigence is a social motive. The primary feature of exigence as a social motive is that it precedes 1 The research reported in this paper was carried out while Simon Borchmann was associate professor at Aarhus University. The Perilous Life of a Linguistic Genre Convention 47 the actual situation; it is an already-established social structure that determines the interpretation of the actual situation as an unsatisfied need, as well as the interpretation of the potential discourse as an attempt to satisfy the need. The second feature of exigence as a social motive is that the need, as well as the way of acting to satisfy the need, is mutually recognised and accepted by the participants in the discourse society. The third feature of exigence as a social motive is that it is historically and culturally determined. Social needs change, evolve and decay from time to time, and vary from culture to culture. The same applies to the ways of acting that are developed in order to satisfy these fluctuating and varying social requirements. Miller’s account implies a research strategic rationale for genre analysis: the object of the description must be the “conventions of discourse that a society establishes as ways of ‘acting together’”, and the principle of classification, thus, must be based on practice. As to the object, conventions show language users how to fuse together form and substance to create meaning. These fusions are made on a number of hierarchically-organised levels, listed in descending order: nature, culture, life form, genre, strategy, speech act, locution, language, and experience (Miller 1984: 162). The ‘fusion’ metaphor corresponds to the basic functional linguistic assumption that form carries meaning, and that form and substance - viewed from the side of meaning - are inseparable (Harder 1996). Thus, each level represents a fusion of substance and form, and the fusion on one level is the substance of a form on the level above. As to the principle of classification, the starting point must be the typologies of action, which affects the discourse community at a certain time in a certain culture. Thus, if we wish to determine newly-emerged ways of acting, e.g. news analysis, we can use neither historic, cultural or field-specific typologies, such as those found in Aristotle, nor general typologies based on theoretical principles of classification, such as the text typologies of text linguistics. At best, they will tell us very little about news analysis as a conventionalised social purpose. A supplement which might be useful in practice-based genre analysis is the notion of systems of genres. Of course, the text typologies constructed within text linguistic are systems. However, as Bazerman identifies, societies construct their own systems of complex located literate activity through typified actions. Systems of genres are “interrelated genres that interact with each other in specific settings” (Bazerman 1994: 97). “Specific settings” covers specialised discourse domains, such as patent litigation, court ruling, tax accountancy, scientific writing, schooling, and filling a vacancy. As to the relations, Bazerman focuses in particular on sequences of typified acts, related in such a way that “only a limited range of genres can follow each other in the specific settings”, for example, “a patent may not be issued unless there is an application. An infringement complaint cannot be filed unless there is a valid patent” (1994: 98). Such dependency relations also hold between some newspaper genres. However, as each newspaper comprises a range of different Simon Borchmann 48 genres, I suggest that we also characterise the system of newspaper genres as a set of simultaneous, typified actions, and that the relations between these simultaneous actions correspond to a division of labour. On a general level, there is division between descriptive, explanatory, argumentative, and entertaining genres. Specific differences regarding this relation will be described later in the analysis. This is the theoretical basis for the analysis. Besides being theoretically well informed and well founded, Miller’s approach also forms the basis for an explaination of the emergence of a new genre. However, the hierarchy of meaning presented by Miller remains abstract and calls for linguistic specification within the context of text description. 3 The text linguistic approach The text linguistic approach in this study is based on two fundamental assumptions of natural language philosophy: 1) that the act is the minimal unit of linguistic communication (Wittgenstein 1953, Austin 1962, Searle 2001(1965)), and, 2) that linguistic communication is inference based (Grice 1975, Levinson 1995). Whereas both speech act theory and the logics of conversation are often referred to within text linguistics (e.g. Ulbæk 2005, Vagle et al 1993, Bülow Møller 1989, de Beugrande & Dressler 1981, Togeby 1979, van Dijk 1977), the abovementioned assumptions are not fully integrated in the analytical framework. In this study, the text linguistic approach differs from standard text linguistic approaches, since it incorporates all the consequences of the abovementioned assumptions. These consequences are particularly significant with regards to two basic notions of the text linguistic curriculum: macrostructure and cohesion. Thus, it seems appropriate to take this as a point of departure in order to describe the text linguistic approach. 3.1 The component unit of the meaning of discourse: propositions or acts? The text linguistic notion macrostructure is based on the assumption that the component unit of the meaning of a discourse is the proposition, and that the meaning of a sentence corresponds to this unit (van Dijk 1977, 1995). This assumption is contested by Wittgenstein’s natural language philosophy and the speech act theory of Austin and Searle. According to Wittgenstein, a proposition is neither a sentence nor a move in a language game (1953: 11 (§22)). Even if a proposition is presented in a main clause structure, indicating an assertive act (‘es regnet’/ ‘it is raining’), Wittgenstein rejects that it serves to tell him something if he doesn’t know whether he has heard the beginning and end of the period (ibid.). In fact, it seems that Wittgenstein doubts that the proposition is a category in the analysis of natural language at The Perilous Life of a Linguistic Genre Convention 49 all (ibid. (§22-24)). Likewise, Austin points out that the traditional “statement” and “its traditional truth or falsity” is an “abstraction, an ideal” (1962: 147); “we use an over-simplified notion of correspondence with facts - oversimplified because essentially it brings in the illocutionary aspect” (1962: 145). Searle is also very clear on this point: the illocutionary act is the minimal unit of linguistic communication (2001(1965): 80) An assertion is an illocutionary act, but a proposition is not an act at all, although the act of expressing a proposition is a part of performing certain illocutionary acts. (...) for a large class of sentences used to perform illocutionary acts, we can say that for the purpose of our analysis that the sentence has two (not necessarily separate) parts, the proposition indicating element and the function indicating device. The function indicating device shows how the proposition is to be taken, (2001(1965): 84) Setting aside Austin’s and, in particular, Wittgenstein’s scepticism regarding the proposition as a category of natural language, we might say - following Searle’s analysis - that a sentence has a propositional content. Let’s even say that this content can be described in terms of abstract meaning (Thomas 1995) or locution. Nevertheless, for the proposition to contribute to the meaning of the discourse, it must form a part of a whole that constitutes an action. Thus, irrespective of how we formulate it, the following still holds: if a linguistic text description is based on the natural language philosophy of Austin and Searle or Wittgenstein, the component unit of the meaning of discourse is not the proposition, but the illocutionary act or the move in a language game within a form of life. Let us turn to the assumption that the meaning of a sentence corresponds to a proposition. It seems as though van Dijk, among others (e.g. Berge 2010), fails to distinguish between main clauses and subordinate clauses. The criterion of a main clause in Danish grammar is that it has an illocutionary frame (Hansen & Heltoft 2011). The illocutionary frame is the part of the sentence where the illocutionary force is indicated. This is exactly what differentiates the main clause from a subordinate clause (ibid.). Likewise in Swedish, Dutch, German, English and other languages: ‘What generally characterizes the distinction between main and subordinate clauses, (...) is the distinction between presence and absence of illocutionary force in the clause in question’ (Verstraete (forthcomming)). Clauses that form a part of whole clauses do carry propositional content (in the abstract sense), but they are incomplete as discourse units and, thus, cannot and do not (ceteris paribus) appear alone. As the material of text linguistic analysis is sequences of sentences, the following question might arise: can the propositional content of one sentence be related to the propositional content of another sentence, independent of the features of the text as a social act? That is, can the semantic and pragmatic processes run “parallel” to each other in the processing of sequences of sen- Simon Borchmann 50 tences, as suggested by van Dijk (1977, 1995)? According to Austin and Wittgenstein, the answer is no; we cannot derive truth conditions of a sentence independent of the use and/ or the features of the sentence as a social act. And, since in Searle’s account of the conditions of satisfaction, the illocutionary force indicating device can pose restrictions on the propositional content, according to Searle, the answer should also be no. It should be emphasised that this is not esoteric or irrelevant philosophy; even the simplest occurrences of natural language buttress the point. The most obvious examples that highlight the necessity of bringing in the use and features of the sentence as social act in order to establish coherence are sequences of sentences where speech acts are embedded in other speech acts (Borchmann 2010). But the same applies to sequences without embedded speech acts: Boil the broccoli in salted water until tender and leave it to drain. Chop it very finely. Heat the oil and fry the speck or bacon until brown. Add the garlic and fry for another minute. Add the broccoli and salt and stir in the milk. Cook the sauce stirring frequently until smooth. 2 Notice the definite form of the noun “sauce”, indicating that the author considers sauce accessible information on the focal point of the reference. However, unlike the other ingredients, up until that point, “sauce” had not been mentioned as an ingredient. Nevertheless, the author’s presupposition will not cause any problems, provided the reader has knowledge about cookery, i.e. that boiled, finely chopped broccoli, mixed with milk and heated, are about to become a sauce. Now, the point is that the transformation from a fresh head of broccoli and milk to sauce presupposes the intended perlocutionary effect of the illocutionary acts, indicated by the author’s use of the imperative mood; the illocutionary content is instructions to the reader in how to change the representation. Thus, the reader cannot establish a coherent propositional representation without bringing in the illocutionary content of each sentence (besides knowledge of cookery). As the linguistic approach in this study is based on the basic assumptions of natural language philosophy, it does not acknowledge the term macrostructure as defined, and continously modified, by van Dijk (1972, 1977, 1995), and it suggests that we refrain from using the term in the analysis of natural language use. However, it is important to note that, when it comes to the alternative analysis, there is also disagreement between the three philosophers mentioned above. The following two points of difference are of particular relevance to the conception of Miller’s hierarchy and the text linguistic approach: whereas Wittgenstein’s moves form a part of a language-game within a form of life (1953: 11e(§23)), Austin’s and Searle’s illocutionary acts are complete in the 2 Translation of a recipe from the Danish edition of the cookbook Passion for Pasta by Antonio Carluccio. The Perilous Life of a Linguistic Genre Convention 51 sense that they are “bound op with effects” and that one can “achieve” something with them (Austin 1962: 117); and, whereas Wittgenstein claims that there are “countless kinds” of language-games, and that “this multiplicity is not something fixed” (1953: 11e (§23)), Austin and Searle assume that there are a small number of general classes. As to the first point of disagreement, Miller’s hierarchy indicates that she prefers Wittgenstein’s account. In line with this, Miller claims that speech acts “exists only within a larger interpretative context” and, furthermore, that this context “itself is hierarchic”. Therefore, the hierarchical account seems to imply that even the level of nature must be involved if meaning is to be established. Nevertheless, Miller refers to the level of genre as “complete discourse types”. In order to clarify this, we need to distinguish between the hierarchy as an interpretive guideline for a single discourse (i.e. as an ontogenetic framework), and the hierarchy as an explanatory framework for the emergence and establishment of discourse as type (i.e. as a phylogenetic framework). When Miller refers to genres as complete discourse types, I assume it to be in the ontogenetic sense. Genres emerge from life forms, which emerge from cultures, which, in turn, emerge from human nature; thus, if you know the genre of a given text, you are already a part of the life form, the culture, and the nature, and you will have sufficient knowledge to interpret the strategy of the text. This is what is implied by Miller’s frequently-cited statement: “what we learn when we learn a genre, is not just a pattern of forms or even a method for achieving our own ends. We learn, more importantly, what ends we may have ... As a recurrent, significant action, a genre embodies an aspect of cultural rationality” (1984: 165). A strategy, on the other hand, might resemble the strategy used within another genre. Hence, it is incomplete as an interpretive guideline for the text. This becomes even more conspicuous when it comes to the interpretation of components on the level below strategy, e.g. the single speech acts that form a part of a strategy. Now, the nature, the culture, and the life form do determine the interpretation of the text, but only in the phylogenetic sense, i.e. through the genre as discourse type. That is, if we are to point out the level on which the meaning of a discourse is established in Miller’s hierarchy, it is the level of genre. Thus, strictly speaking, a genre is the minimal unit of linguistic communication in Miller’s account. This has partly something to do with the type of meaning described, and partly something to do with the types of discourse described. Now, Miller does not define discourse meaning, nor does she distinguish between types of meaning, but the type of discourse meaning Miller describes as complete is “conventionalised social purpose within the recurrent situation”. This type of meaning is only obtained on the level of genre, whereas, according to Miller, “idiosyncratic motives (intentions)” predominate on the level of speech act (and locution) (1984: 162). Simon Borchmann 52 There is a remarkable correlation between this idea and Grice’s logic of conversation. Although Grice defines meaning as, “A intended the utterance of x to produce some effect in an audience by means of the recognition of this intention” (1989 (1957)), he presupposes an accepted purpose or direction of the exchange when he formulates the principle of cooperation in his logic of conversation (1975). The accepted purpose determines how the quantity maxims are to be applied, and which implicatures are to be drawn. Hence, the accepted purpose is a prerequisite for the establishment of discourse meaning. Of course, there are a large number of exchanges where there is no consensus as to the purpose, where the purpose is unclear, where the purpose is negotiated, or where one part determines the purpose by force. Such exchanges seem to defy Miller’s concept of genre. It seems, thus, that the disagreement regarding the number of language games is based on a bias. Indeed, some speech acts occur within the context of a mutually-accepted, conventionalised, social purpose and are allocated specific interpretations according to this purpose. As to these speech acts, it seems very difficult to delimit the possible combinations, as well as the specific interpretations, particularly as these specific combinations and functions are dynamic. But other speech acts are not uttered within such contexts. In so far as these acts may, nevertheless, be associated with effects and contribute to the coordination of goals or the acceptance of a purpose, we must be able to assign meaning to them on a more general level. It seems reasonable to suggest that this more general level can be analysed in a limited set of speech act classes. This supposed disagreement can be resolved by identifying that genre theory and speech act theory have different objects, or rather that they are on different levels of abstraction in terms of language use and meaning (see also Togeby, this volume). However, the complex speech act, news analysis, is located precisely on Miller’s level of abstraction, and, therefore, the use of Miller’s theory does seem appropriate. The question remains how these complete discourse units are to be analysed linguistically. Before the analytical framework is outlined, some elaborative comments should be made on the assumption regarding inference-based linguistic communication. 3.2 Coherence: ‘semantic relations’ or intended inferences? Under the label cohesion, Halliday & Hasan introduce the idea of a basic, linguistic text constituent. The idea is described as follows: A text has texture, and this is what distinguishes it from something which is not a text. (1976: 2) cohesive ties between sentences stand out more clearly because they are the ONLY source of texture (1976: 9) The Perilous Life of a Linguistic Genre Convention 53 the concept of cohesion accounts for the essential semantic relations whereby any passage of speech or writing is enabled to function as at text. We can systemize this concept by clarifying a small number of distinct categories - reference, substitution, ellipsis, conjunction, and lexical cohesion (...) Each of these categories is represented in the text by particular features (...) which has in common the property of signalling that the interpretation of the passage in question depends on something else. If that ‘something else’ is verbally explicit, then there is cohesion. (1976: 13) Over the last thirty years, this idea has gone from strength to strength within linguistic text analysis in the Nordic countries, and has appeared in numerous textbooks (e.g. Ulbæk 2005, Lützen 2004, Møller, 2002, Vagle et al 1993, Bülow Møller 1989). It has also been written into the national curriculum for secondary schools. Thus, in the Danish Ministry of Education’s guide for English in secondary schools, one can read that, “cohesion is created through the use of words and word classes that bind text elements together” (Ministry of Education 2008). Even in current Academic English courses at Aarhus University, several hours are spent making cohesive ties verbally explicit in texts. This development is remarkable (and in many ways interesting from the perspective of sociology of scientific knowledge); for everyone who works with texts at a non-superficial level will realise that the idea is fatal to the understanding of what a text is and what characterises a well-functioning text. The problem is not that the text criterion disqualifies a number of wellfunctioning texts (Harder 1979, Brown & Yule 1983), nor that a number of texts that fully satisfy the criterion cause problems for language users (Haberlandt & Bingham 1978, Sanford & Garrod 1981, Brown & Yule 1983, Keenan et al 1984). The problem is not even that the idea is psychologically implausible, insofar as readers (and listeners) cannot maintain the chains of coreference implied by the term endophoric reference in memory (Brown & Yule 1983). Nor is it that the principle of substitution leads to absurdities (Ibid.). These are simply consequences of an overarching problem: that sequences of sentences are conceived of as a self-reliant linguistic structures, in which linguistic material in one sentence is related to linguistic material in another sentence, independent of a ‘mediating reader psychology’ that implies 1) knowledge about the world, 2) interests and needs, 3) limited memory capacities, 4) selective attention, 5) ability to infer, and, 6) ability to construct representations in a non-linguistic format. Hence, the model of communication implied by Halliday & Hasan’s account is a model in which understanding depends on the presence and recognition of certain linguistic features, and in which the recipient’s abilities - as far as the linguistic features are “the ONLY source of texture” - are limited to the recognition of these traits (this is why the relation needs to be semantic). This model prevents an adequate analysis of texts. In such a model, the focus is solely on the presence of certain linguis- Simon Borchmann 54 tic features, not on how they are used. Furthermore, linguistic forms that play a crucial role in the establishment of coherence are neglected. The most powerful opposition to the idea of cohesion originates in Grice’s logic of conversation (1975). Grice’s ambition was not to provide a model of communication; nevertheless, his description formed the basis for an alternative understanding of communication, i.e. communication as a cooperative enterprise, where the recipient is actively contributing to the establishment of meaning by using his knowledge about the situation and providing information to the text represented information through abductive, general or situation-dependent inferences. In contrast to Halliday & Hasan’s account, this account implies an active reader or listener. A similar idea was suggested within cognitive science. Here, the focus was on the recipient’s stored knowledge of the world as a prerequisite for language comprehension (Minsky 1975); in agreement with this, a text was conceived of as a set of instructions to the receivers on how to use their stored knowledge of the world in order build a coherent meaning representation. In other words, as formulated by Sanford & Garrod, coherence is in the mind of the reader, not in the text (Sanford & Garrod 1994). This recognition prompted comprehensive research on inferences in text processing. The starting point was that a large number of inferences can be made on the basis of a text, but that the capacity of the reader’s working memory is limited, so that readers can only make a very small number of inferences online. The initial approach was to try to determine which types of inferences were made online in general. Over the years, there has been an increased awareness of the factors that affect the likelihood of generating particular inferences. Among other factors, textual conditions (Sanford 1990, Whitney et al. 1992, Murray et al. 1993, Klin et al. 1999) have been emphasised. In line with this, a text linguistic analysis, based on the assumption that linguistic communication is inference-based, focuses on the textual conditions that contribute to the occasion of particular inferences. These are linguistic indications of an informational status, but also genre markers, and indications and markers for how the sender is related to the facts communicated (Borchmann 2005, 2011). 3.3 The analytic framework Even though Miller suggests a number of levels below genre, these levels remain unspecified, apart from four very general assumptions: 1) genre as meaningful action is rule governed, i.e. interpretable by means of conventions, 2) form shapes the response of the reader or listener to the substance, by providing instruction about how to perceive and interpret, 3) the substance of a form on a given level is formed on the lower level, hence, language is formed before it becomes the substance of locutions, and, 4) the substance of language is “drawn from our acting-together” (1984). These assumptions are The Perilous Life of a Linguistic Genre Convention 55 in agreement with the two fundamental assumptions of natural language philosophy, and they serve as guidelines for the establishment of the analytic framework. In this paper, I draw primarily on Swales’ genre analysis. Although there are disagreements between Miller’s new rhetoric analysis and the LSP (Language for Specific Purposes) tradition (regarding the need for a historical perspective and a broader socio-cultural context), there is agreement on two critical points. Firstly, it is also a social motive that constitutes a genre according to Swales; in his account, it is embedded in a communicative purpose. Certainly, Miller has expressed skepticism about the term purpose (2012), but this skepticism has been met by Askehave & Swales (2001). Secondly, the classification principle is practice-based, i.e. based on the knowledge of expert members of the parent discourse community - a rationale (1990: 58). Swales operates with three concepts, which are particularly informative in relation to the social conventions that constitute genre on Miller’s levels of strategy, speech act and language. These concepts are move structure, move (and step), and feature, respectively. Move structure is a purpose-oriented compositional analysis. The assumption that governs the analysis is that the organisation of the text is determined by the purpose that constitutes the genre’s rationale. The rationale is the expert’s knowledge about the purpose, and it translates into a rhetorical strategy that appears as a specific text organisation. The analysis is linguistically-based, but differs significantly from text linguistic composition analysis. Firstly, the component is not analysed in semantic categories, but in rhetorical, pragmatic categories. Secondly, the text is not analysed as a logical hierarchy (van Dijk 1977, 1995), but as a strategically-determined organisation. The analysis aims to uncover the rationale through the specific rhetorical functions controlling the organisation of the text. This approach results in a description that is informative as to the purpose-oriented use of language. While the semantic-based descriptions, such as those put forward by van Dijk’s, have analysed the meaning of a text in propositions, Swales analyses the meaning of a text in actions. These actions are called moves, and they are characterised by being part of an overall rhetorical strategy designed to achieve an institutionalised communicative purpose. Moves differ from illocutionary acts by being dependent; the functions they serve are assigned to them as parts of the overall strategy. Hence, there are no independent “conditions of satisfaction” (Searle 1969); they are dependant on the rationale and the strategy. Nevertheless, moves are distinct as parts of a rhetorical strategy. Firstly, they are assigned to rhetorical functions, and, secondly, regularities can be observed in the linguistic realisation. Features are general grammatical functions that serve specific rhetorical functions within a conventionalised communicative purpose. This relation corresponds to Miller’s idea that the substance of a form on a higher level in the hierarchy is formed on a lower level. Simon Borchmann 56 As to language, I will assume that we can account for the meaning of a sentence independent of the specific use. As this account is extremely abstract, I have called this meaning abstract meaning using a term from Thomas (1995). Abstract meaning is a theoretical reconstruction of a general linguistic competency (Harder 2010) and corresponds to general grammatical functions of a given language. Thus, abstract meaning serves as general instructions to the reader or listener, and, on the basis of these general instructions, the reader or listener engages in interpretive activities. The abstract meaning of a sentence comprises propositional instructions, but is not restricted to these; since it is drawn from our acting-together, it comprises illocutionary force indicators, indicators to the informational status of information bits, and instructions about the sender’s attitudes (Hansen & Heltoft 2011). All these types of instructions can serve as features indicating a move (or step). The analysis presented here operates with the concept move structure, move, and feature. The fundamental strength of these concepts is that they, in contrast to the ideas of macrostructure and cohesion, capture the properties of the text that change when the factors on the upper levels in the hierarchy change. They are, therefore, particularly suitable for making the connection between the upper and lower levels in Miller’s hierarchy. 4 The object 4.1 The emergence of news analysis In 1999, Danish journalistic expert and textbook writer, Mogens Meilby, drew attention to the emergence and advance of nyhedsanalysen (news analysis) in Danish news journalism during the nineties. He described it as a new addition to articles, which put news into perspective, and he characterised it as follows: “the content [of the articles] is completely or partly the journalist’s own analysis of the perspectives, causes or consequences of a piece of news” (Meilby 1999). Meilby’s observation is supported by a master study of the political news analysis in Danish newspapers, which shows that the number of articles designated nyhedsanalyse or analyse (analysis) increased six-fold from 1995 to 2005 (Pedersen et al 2006). Parallel to this elaboration of the journalistic genre system, a quantitative study of explanatory journalism in Denmark has shown an increased frequency of explanations in news articles during the last two decades (Svith 2011). In line with this, the birth of news analysis can be seen as a purpose-specific manifestation of a more general tendency to focus on explanation and, with it, interpretation Danish news journalism. This tendency is in line with a supranational tendency. According to McNair, there has been “columnary explosion” within political journalism in recent decades (McNair 2000). The journalistic form of the columnist is characterised by an “interpretive moment”, and, according to McNair, by virtue of The Perilous Life of a Linguistic Genre Convention 57 this quality, it is “able to penetrate the relentless flow of new information into the public sphere, giving insight and meaning to the events which it addresses” (ibid. 71-72). This journalistic form includes, among other genres, the analytical-advisory column (ibid. 64). 4.2 The background and explanation of the emergence As indicated above, the emergence of news analysis in Denmark can be explained on both a general, supranational level and on a specific, national level. On a general level, McNair points to the following three processes in order to explain the proliferation of political commentary in the late twentieth century: 1. the commercialisation of journalism, 2. the information overload associated with the revolution in communications technology over the last century and a half, and, 3. the emergence of political public relations as a science of news management (McNair 2000: 71). According to the ideal of the practitioners (Pedersen et al 2006), news analysis, as well as analysis, differs from commentary. Nevertheless, there is a strong resemblance between the genres when it comes to the interpretative moment. The demand for interpretation arises partly with the rapid flow of information facilitated by new communication technologies, and partly as a journalistic countermeasure to the spin generated by the managers of political public relations. The commercialisation of journalism makes it necessary for news producers to satisfy this demand: “If their use-value can be defined as ‘interpretation’, their exchange-value in a marketplace of many interpreters is predicated on their ability to interpret better than others” (McNair 2000: 64). As to the specific, national level, the emergence of news analysis can be explained partly against a background of the development of a politicallyindependent, professional, differentiated press, which is increasingly subject to the forces of the market (Thomsen 1991, Søllinge 1992, Søllinge 1999, Hjarvard 1999, Hjarvard 2009, Pedersen 2000, Lund 2002, Schultz 2007, Bro 2009), and partly by an increased complexity in the process of legislation (Søllinge 1999). The free press was introduced in Denmark in the constitution of 1849. After this, the press continued to develop as an organ for political parties: during the second half of the nineteenth century, a nationwide network of daily newspapers was established. The number of newspapers printed increased from 31 in 1847, with a circulation corresponding to 9% of households, to 137 in 1901, with a circulation reaching 100% of households (Søllinge 1999: 79). Each newspaper was advocating the views of its affiliated political party. The main genres were the editorial, the reports of parliamentary debates, and coverage of other political meetings. Since political debates were particularly heated during this time (owing to significant social and political tensions in the last quarter of the century), the reports were heavily biased in favour of the party, giving priority to the utterances of the fellow party members, and giving short shrift to their opponents. Thus, to inform Simon Borchmann 58 oneself of the overall, political picture, one had to read reports, as well as editorials, from at least four newspapers (ibid. 78). Around 1900, new journalistic tendencies appeared. Taking American and British journalism for a model the genres reportage and interview advanced, new topics (e.g. business and sports) were introduced, and the use of humaninterest criteria was promoted. Søllinge emphasises the transformation of content from opinion to news as the overriding feature of this development (see also Hjarvard 2009). This transformation is recognised as the point at which the political press was phased-out, and the modern omnibus press was born (Søllinge 1999). The most important consequence for political journalism was that political material ceased to be determined by the principles and interests of the political system itself; instead, it became subject to common, journalistic news criteria (Thomsen 1991, Hjarvard 2009, Søllinge 1999). The referent was gradually replaced by a reporter, who gathered, selected and combined information from a growing network of sources, which extended far beyond the press box (Bro 2009). This serge of information required an increased professionalism (ibid.), and, whereas journalism in the second half of the nineteenth century had been performed by men of letters, or by A-level graduates, it was now becoming a profession in its own right (Søllinge 1999). It was also during the birth of the omnibus press that expressions of individual ideals and principles appeared. In the first Danish journalism textbook, edited by a key figure in the development of the Danish, omnibus press, Henrik Cavling, and printed in 1928, one could read: “It is of the outmost importance that the press stands free to all sides and that the only consideration is the interest of the population. Consideration for private interests, whatever the reason, when they are contrary to the benefits of the public, is not consistent with upright journalism” (Source: Søllinge 1999). However, producing an omnibus newspaper required a far larger and more qualified staff. It also demanded more sophisticated printing machinery, because of the growing size and circulation involved. Consequently, economic pressures increased, and, since financial support from political parties was no longer sufficient, advertising revenues became even more significant, and the competition for readers increased. In other words, newspapers became subject to market forces (Hjarvard 2009). As a result, many newspapers were forced to close down; whereas the number of newspapers had grown to 159 in 1919, it reduced to 122 in 1938. Market competition subsided during the Second World War, due to paper rationalisation and a widespread thirst for news; however, after the German occupation, the competition revived, enhanced by new economic demands associated with the replacement of existing, run-down printing machinery (Søllinge 1999). At the same time, competition was intensified by the market principle, “the bigger the circulation, the more advertising revenues, and vice versa”. So newspapers continued to close on a local level; whereas each city used to have a newspaper for each major The Perilous Life of a Linguistic Genre Convention 59 political party (a relic from the former, political press), only one of them survived, reducing the number of newspapers from 111 in 1950, to 62 in 1970, and creating a local monopoly on news (ibid.). The surviving newspaper was typically the one with the broadest appeal to the readers. Therefore, this monopolisation was closely connected to the notion of de-politicisation. This is supported by statistics, which show that the correlation between the share of votes and newspapers sold for each of the four main political parties fell from a co-variation of 0.53 in 1953, to a co-variation of 0.10 in 1964 (Thomsen 1991: 84). Increased professionalism and the use of journalistic news criteria had a major impact on political journalism. A new genre - the political column - appeared, and newspapers invited representatives from different political backgrounds to write columns, so that most, if not all, political opinions were represented in one newspaper. These developments continued after 1970, and, as Søllinge writes, the angle of individual news pieces increasingly contradicted the traditional party lines from which the newspaper originated (ibid. 87, se also Thomsen 1991). This was another instance that reflected general, nationwide de-politicisation of the press. Thus, at this developmental stage, the Danish press can be described as a politically-independent, professional news media, partly striving for objectivity and straight reporting, supplied with factual interpretation and allowing political movements and agents to express themselves on their own terms (Thomsen 1991), partly subject to market forces. The development from a political press to an omnibus press has been described, within a more general model of the media, as a transformation from a political institution to a cultural institution; the role of latter being to educate the public and disseminate social norms and public debate (Hjarvard 1999: 32-33) Yet, however worthwhile it may seem to educate the public (and to make a profit in the process), the development of media moves beyond this stage. Media historians have observed two, fundamental shifts. The first regards professional ideals. In the late sixties, the strive for objectivity, straight reporting and political versatility was partially superseded by a desire for a more engaged, critical and investigative journalism (Thomsen 1991). This change might be understood within the framework of a more general model of the media. According to several Danish media researchers, by detaching themselves from the political system, the Danish media became an institution in itself, interacting with other institutions and serving their audience according to their own logic (Hjarvard 2009). The second change concerns the different types of newspaper. Søllinge points to a differentiation caused by increasing demands on the editorial resources, since the popular newspapers, as well as the national morning newspapers, were gaining ground outside the capital. The popular newspapers were concentrating more on entertainment, the large, national omnibus newspapers were becoming increasingly extensive, with a growing number of sections, the smaller national newspapers were Simon Borchmann 60 focussing on particular areas (e.g. quality or business), and the local newspapers were concentrating more on local issues. As Søllinge indicates, this change is a result of competition. In line with this, and with reference to Bourdieu, some media sociologists have emphasised that the relative autonomy of the press cannot be determined solely in relation to the political field; the economic field and the field of media consumption, among others, must be considered (Schultz 2007). Thus, as the successor of the omnibus press, Schultz suggests “segmentpressen” (segment press), indicating a marketoriented, life-style political press, rather than the press as a politicallyindependent media institution (Schultz 2007). Regardless of which account represents the highest explanatory value, there is widespread agreement that, from 1970 onwards, the press has been ever more subject to market forces, and, thus, ever more influenced by readers as consumers, rather than readers as citizens. Against this background, the emergence of news analysis in the nineties can be explained as an attempt to provide a critical press, as well as to satisfy the needs of a particular group of media consumers, who were occasioned by the development of the political system in Denmark. With the abolishment of the bicameral system in 1953, the Folketing (Danish Parliament) became the only legislating authority. As a consequence of this, the strategic aspects of the political process were exposed. Parallel with this, the complexity of the political process increased considerably, due to the increased involvement of the organisation in the legislative process, and a significant increase in the number of legislative initiatives, not least in continuation of the membership of the EF in 1973. To ensure the understanding of each, individual event, background information and interpretation was required. A critical press, subject to the forces of the market, operating within a niche of a national omnibus readership, has to respond to such a requirement. The emergence of news analysis can be seen as such a response. To add a twist to this development, it should be noted that the report from parliament, which was still featuring 1965 (although in a reduced version), had disappeared entirely by 1991 (Søllinge 1992). 4.3 A text theoretical point The point of the elaborated description above is, of course, that the development of the newspaper genre system, including the emergence, change, and decay of the genres described, is the result of a collection of interdependent factors: constitutional, political, governmental, ideological, educational, professional, ethical, technological, economical, commercial, and others. It is remarkable that this understanding is fully and naturally integrated into the accounts of media researchers and historians, whereas it is completely absent in the concept of macro-structural text linguistics because of its basis in generative grammar or formal semantics. From a functional viewpoint, however, The Perilous Life of a Linguistic Genre Convention 61 the emergence of news analysis confirms the assumption of a correlation between genres and specific needs, arising with the development of society. 4.4 The practitioner’s view of the genre According to the journalistic textbook author, Mogens Meilby, the purpose of news analysis is, “to give the reader explanation and context - to help him understand and evaluate information should it become too comprehensive, too complex and too hard to grasp”. This provision is very wide and allows for a variety of motives and functions. Meilby distinguishes the following subtypes: descriptive, predictive, judgemental, and advisory. The first two belong to news journalism, while many professionals consider the last two to be part of opinion. A study has shown that the descriptive and predictive analyses are the most frequent (53% and 27% respectively), but the proportion of judgemental and advisory are rising slightly, from 1995 to 2000 and from 2000 to 2005 (Pedersen et al. 2006). In the Danish newspaper, Politiken, the genre marker news analysis is used solely for news journalism. The new analyses of news journalism are described as follows in the internal genre catalogue: Whereas a news article focuses on answering the questions who, what, where and when, a news analysis seeks to answer the questions: why and what now? The basis of the analysis is still the facts. However, the analysis explains the causes and / or motives behind the news. As the news analysis is largely based on the writer’s interpretation and assessment of the essential elements, it is particularly important that the approach aims at objectivity and fairness. Sources (usually many) are quoted throughout the analysis. The news analysis focuses particularly on dilemmas and possible further developments. (Source: Zabel 2007) The division between news journalism and opinion is clearly indicated by ‘facts’, ‘objectivity’, ‘fairness’ and ‘sources (usually many)’. Notice also that news analysis is described from a division of labour between the news journalistic genres ‘news’ and ‘news analysis’. Politiken allows for another type of analyses. These belong to opinion and are described in the following way: The analysis occurs only on the opinion pages. It belongs to the same family as the news analysis, but differs in several respects. First, the requirement for timeliness is different. The analysis is not necessarily linked to a news item. Secondly, it puts much more emphasis on the writer’s personal assessment of causal relationships than in the news analysis. And thirdly, it is expected that the writer draws a conclusion on the analysis. As with news analysis, it is important that the approach aims at objectivity and fairness and that sources which are contrary to the analysis and conclusion are quoted in good faith. (Source: Zabel 2007) Simon Borchmann 62 What characterises these analyses as opinion are personal assessments and conclusions. It is worth noting that, according to professionals, the standards that apply to news journalism also hold for these analyses. This could be viewed as indication that these analyses are borderline cases. The newspaper, Berlingske, appears to use a similar distinction between news analysis and analysis. The newspaper, Information, has been using both terms, though apparently not entirely systematically; thus, some descriptive news analyses bear the genre marker analysis. 5 Methodology The framework of the linguistic text analysis presented here is a movestructure analysis, following Swales (1990: 140ff), but supplementing this with psychological, particularly cognitive, factors, as suggested by Bathia (1993: 14), and by considering the genre as a part of a genre system, as suggested by Bazerman (1994). Because of the focus on specific linguistic genre conventions, some aspects of the analysis and specific steps were given priority. The analysis consisted of three parts. The first part concentrated on establishing an overview of the recurring moves and patterns of moves in news analysis texts, irrespective of eventual developments in the continuation of the emergence of the genre. This part comprised a preliminary determination of the communicative purpose (se above), a selection of a corpus, an identification of the genre-specific moves, a determination of the status of the moves relative to the purpose, an identification of systematic variations, and an identification of patterns. The second part was restricted to the two common opening moves of news analysis, aiming at establishing them as distinctive of news analysis within the system of newspaper genres. The third part concentrated on describing the linguistic realisation of one of the steps in the opening move, with a view to identifying a distinctive linguistic genre convention. The analysis in the first part was based on a corpus consisting of 43 news analysis texts. The two criteria in the composition of the corpus were, 1) that the texts should appear in one of the national, segmented, omnibus newspapers printed and uploaded to the homepage of the newspaper, and, 2) that they should be marked either with the genre marker “nyhedsanalyse” (news analysis) or with the genre marker “analyse” (analysis - this marker is used in some newspapers). Within these criteria, the aim was to make the selection as varied as possible. The corpus was, thus, composed of texts written by experts as well as journalists, by news correspondents as well as political commentators. The texts spanned the period 1997 to 2011, and dealt with cultural material, as well as sport and politics. They were selected from the newspapers Berlingske and Information, as well as Politiken. Of course, a far larger corpus, systematically divided into subject, type of author, year, paper, etc., would have been preferable. However, since the purpose of my analysis was neither The Perilous Life of a Linguistic Genre Convention 63 to describe a possible development of the genre, nor to present a detailed description and explanation of each variant, but, instead, to establish common moves and steps of texts labelled news analysis (or analysis), with a view to describing a specific linguistic feature of a single step, I take such a corpus to be entirely adequate. The second part of my analysis concentrated on the two opening moves common to the six types of news analysis observed in the first phase. This analysis was partially theoretical, drawing on schema theory, textunderstanding theory, and introducing distinctions. The empirical basis was the corpus, but it also involved a selection of headings of news paragraphs, which were produced by the news bureau, Ritzau, and which occurred in the newspaper, Information, on January 18 th and 19 th 2012. The third part was analysis of the linguistic features of framing as a step in the opening move of news analysis texts. This analysis was based on the same corpus as the global move-structure analysis. 6 Findings 6.1 Types and distribution On the basis of a move-structure analysis of the corpus, six types of news analysis can be distinguished: descriptive, evaluative, predictive, judgemental, directing, and consequence-oriented. 40 of the 43 texts correspond to one of these types. The final three are mixed; two of them are a combination of descriptive and judgemental, and one of them is a combination of descriptive and evaluative. One the basis of the analysis the 43 texts are divided among the aforementioned types in the following way: Descriptive Evaluative Predictive Judgemental Directing Consequence oriented Mixed 27 3 2 3 4 1 3 Tab. 1: Distribution of types As the corpus was designed to be as broad as possible, this distribution does not, of course, say anything about which distribution characterises a specific period, newspaper, type of author, subject, etc. However, the distribution of the corpus does correspond to the result of Pedersen et al with respect to the dominance of descriptive news analysis (2006). The descriptive analysis, thus, is by far the most frequent. Simon Borchmann 64 6.2 The prototypical descriptive news analysis The prototypical move-structure of the descriptive news analysis is as follows (optional moves are in italics; i, ii designate fixed order; a, b, c designate variable order): 1) OFFERING AN INTERPRETATION OF AN EVENT i) Framing the event ii) Summarising the interpretation of the event 2) ESTABLISHING AN EXIGENCY a) Identifying the event b) Presuming the exigency c) Supporting the exigency by settings 3) ELABORATING THE SETTINGS OF THE EVENT BY SUPPLEMENTARY FACTS 4) EXPLAINING THE EVENT 5) PUTTING THE EVENT INTO PERSPECTIVE 6) SUMMARISING ASSESSMENT OF THE INTERPRETATION An example is the following (obligatory moves in bold, optional in italics): Part Example Step Move Heading Regeringen spiller højt spil med ghettoplan i 1 Lead Regeringen må finde penge til sin ghettoplan uden satspuljemidler, efter at VKO i går lod oppositionen spænde ben for sig i forhandlingerne. Derved blegner Løkke Rasmussens glansnummer, hvad først og fremmest DF kan glæde sig over ii Text body De brede forhandlinger om satspuljemidlerne på det sociale område er slut, a 2 men VKO’s ordførere og socialminister Benedikte Kiær (K) måtte gå fra bordet velvidende, at der manglede 80 millioner kroner til finansieringen af den højt profilerede ghettoplan. b Og det bare ét døgn, efter at fire ministre med Kiær i spidsen fremlagde planen, som skulle være et af Løkke-regeringens hovednumre i år. c Siden nytår har VK krydset klinger internt og med Justitsministeriets embedsmænd, men nu var der endelig enighed om et kompromis. Ghettoplanen tilgodeså både Venstres ønske om stramninger, Konservatives nye hensyntagen til den borgerlige anstændighed - partiets egen ghettoplan var bemærkelsesværdig mild - og var spiselig for DF. Finansieringen manglede blot at falde på plads. I VK’s fælles ghettoplan var der lagt op til, at 120 millioner kroner skulle findes i satspuljen, og socialministeren medbragte således et detaljeret forhandlingsudspil til mødet med forligskredsen, der strækker sig fra DF til SF. 3 The Perilous Life of a Linguistic Genre Convention 65 Der var dog den krølle, at størstedelen af de millioner, regeringen foreslog uddelt til kommunale ghettoprojekter, i ghettoplanen var bundet sammen med krav om sanktioner. Hvis kommunerne ville have andel i pengene, skulle de samtidig gøre brug af de udvidede muligheder for økonomisk straf til ghettobeboerne i form af reduktion af boligtilskud, forældrepålæg og straf til de forældre, som ikke ville overholde kravet om obligatorisk brug af daginstitution. Socialministeren kunne have sagt sig selv, at hverken Radikale, Socialdemokraterne eller SF vil kunne acceptere det oplæg. Mens Radikale helst ser ghettoproblematikken løst med pædagoger, har S-SF deres eget strammerimage og meningsmålinger at tilgodese. Man kan næppe heller forestille sig, at S-SF uden videre ville forære regeringen bred opbakning til en konkurrerende ghettoplan. Belejligt Om finansieringsforviklingerne skyldes dårligt forog fodarbejde eller ønsket om udadtil at signalere samarbejdsvilje over for oppositionen, er uvist. Statsminister Løkke Rasmussen styrker sin påstand om, at S-SF reelt ikke ønsker stramninger. Når det kommer til stykket, tør SF’s formand, Villy Søvndal, ikke bede nogen skride ad helvede til. 4 Tilbage står, at ghettoplanen nu må realiseres med penge fra finansloven, der ikke bliver vedtaget under samme konsensusvilkår som satsforliget. Her handler det om at få Dansk Folkeparti (DF) med på vognen, hvilket næppe bliver noget problem, idet ghettostramningerne vil være en fornøjelse for DF at finansiere. I sidste ende er det sandsynligvis DF, der får mest ud af regeringens fremgangsmåde. De bevarer status som de sande strammere, og under finanslovsforhandlingerne vil DF kunne kræve noget til gengæld for at finde de fornødne 80 millioner til at redde den forkromede plan og Lars Løkke Rasmussens ansigt. Hans tre politiske glansnumre risikerer alle sammen at strande som forkølede forsøg: Vækstforum kickstartede snarere kritik fra erhvervslivet end vækst i Danmark, 360-graders eftersynet af folkeskolen stødte på grund med en ydmyget undervisningsminister og en forbitret lærerstand. Og i går kunne vælgerne så læse, at finansieringen af ghettoplanen smuldrede. 5 Uanset årsagen til fremgangsmåden i ghettoforhandlingerne må det betragtes som højt spil, at regeringen igen lader sine strategiske vanskeligheder udstille. 6 Tab. 2: Move-structure of a prototypical descriptive news analysis, original text in Danish Simon Borchmann 66 Part Example Step Move Heading The government is playing for high stakes with the ghetto plan i 1 Lead The government must find money for its ghetto plan without adjustment pool funds after VKO yesterday let the opposition trip up in negotiations. This leaves Løkke Rasmussen’s gloss number lacklustre, which particularly DF can be happy with ii Text body The broad negotiations on adjustment pool funds for social services are over, a 2 but VKO’s spokesmen and Minister of Social Affairs Benedikte Kiær (K) had to leave the table knowing that 80 million kroner needed to finance the high-profile ghetto plan were missing. b And this is just one day after four ministers headed by Kiær presented the plan, which was meant to be one of the Løkke government’s main numbers this year. c Since New Year VK have crossed blades internally and with Justice Department officials, but now, finally, there was agreement on a compromise. The Ghetto Plan benefitted both Venstre’s desire for austerity, Konservative’s new penchant for the bourgeois decency - the party’s own ghetto plan was remarkably mild - and was palatable to DF. Only funding needed to fall into place. In VK’s common ghetto plan the idea was to find 120 million kroner in the adjustment pool, and thus the Minister of Social Affairs brought a detailed negotiation proposal to the meeting with the conciliation circle, which extends from DF to SF. 3 However, there was the crease setback that the majority of the millions, the government proposed be distributed to municipal ghetto projects, in the ghetto plan were tied together with demands for sanctions. If the municipalities wanted a share of the money, they also had to make use of the expanded opportunities for economic punishment of ghetto residents in terms of reduction of rent allowance, parental orders and penalties for parents who would not comply with the requirement for mandatory use of daycare. It should have been evident to the Minister of Social Affairs that neither Radikale nor Socialdemokraterne nor SF would accept the paper. Whereas Radikale prefer the ghetto problem to be solved by teachers, the S-SF have their own hard-liner image and opinion polls to accommodate. One can hardly imagine that the S-SF would readily give the government broad support for a competing ghetto plan. Conveniently Whether the complications of the funding were caused by poor preliminary work and footwork or the desire to signal willingness to cooperate with the opposition is uncertain. Prime Minister Løkke Rasmussen strengthens his claim that S-SF in actual fact do not want tightening. When it comes to the crunch, SF’s president, Søvndal does not dare to ask anyone to go to hell. 4 The Perilous Life of a Linguistic Genre Convention 67 The fact remains that the ghetto plan must now be realized with money from the budget bill, which will not be adopted under the same conditions as the adjustment pool. Here the crucial step is to get Danish People’s Party (DF) on the bandwagon, which hardly will be a problem as ghetto constraints will be a pleasure for DF to finance. Ultimately, it is probably DF which makes the most of the government’s approach. They retain the status as the true hard-liner, and during budget negotiations, DF could require something in return in order to find the necessary 80 million to save the grand plan and Lars Løkke Rasmussen’s face. All of his three political luster numbers are at risk of being wrecked: Vækstforum sparked criticism from the business world rather than growth in Denmark, the 360-degree vision of the school ran aground with a humbled minister of education and an embittered teaching profession. And, then, yesterday the voters could read that the financing of the ghetto plan crumbled. 5 Whatever the reason for the course of action in the ghetto negotiations it must be considered high stakes that the government again exhibits its strategic difficulties. 6 Tab. 3: Move-Structure of a prototypical descriptive news analysis, English translation of the Danish original This move-structure can be seen as a rhetorically-effective way of fulfilling the purpose of the descriptive news analysis, and, since this structure characterised 22 of the 27 descriptive news analyses in the corpus, the structure is presumed to be - and stated as being - the prototypical move-structure of the descriptive news analysis. The structure of the remaining five descriptive news analyses deviates in varying degrees from this structure, typically by being either source-governed or case-governed. Interestingly, four of these analyses were either written early in the period 1997-2011 (1997 and 2002), or they were written by experts about cultural subjects. Therefore, the deviations could be explained with reference to period or subject. Due to the composition of the corpus, however, no such hypotheses can be tested. 6.3 Move structures of the other types of news analyses The move-structures of the five remaining variants were the following: Evaluating 1) OFFERING AN INTERPRETATION OF AN EVENT 2) ESTABLISHING AN EXIGENCY 3) EXPLAINING THE EVENT 4) EVALUATING THE ACTIONS INVOLVED IN THE EVENT Simon Borchmann 68 5) SUPPORTING THE EVALUATION BY SUPPLEMENTARY FACTS OR ASSESMENTS 6) CONDENSED EVALUATION Predictive 1) OFFERING AN INTERPRETATION OF AN EVENT 2) ESTABLISHING AN EXIGENCY 3) ELABORATING THE SETTINGS OF THE EVENTS BY SUPPLEMENTARY FACTS 4) OFFERING FACTS SUPPORTING A PREDICTION 5) RESTATING THE PREDICTION Judgemental 1) OFFERING AN INTERPRETATION OF AN EVENT 2) ESTABLISHING AN EXIGENCY 3) SUPPORTING THE EXIGENCY BY SUPPLEMENTARY FACTS 4) COMPARISION OF FACTS 5) JUDGEMENT Directive 1) OFFERING AN INTERPRETATION OF AN EVENT 2) ESTABLISHING AN EXIGENCY 3) SUPPORTING THE EXIGENCY BY SUPPLEMENTARY FACTS 4) EXPLANING THE EVENT 5) PUTTING THE EVENT INTO PERSPECTIVE 6) DIRECTION Consequence oriented 1) OFFERING AN INTERPRETATION OF AN EVENT 2) ESTABLISHING AN EXIGENCY 3) SUPPORTING THE EXIGENCY BY SUPPLEMENTARY FACTS 4) CALLING ATTENTION TO THE IMPORTANT ASPECTS OF THE EVENT 5) PRESUMING THE CONSEQUENCES OF THE EVENT Whereas the move-structure of the descriptive analysis is empirically relatively well founded, the five structures presented above are based on very few occurrences. Therefore, these move-structures are not stated as prototypical. Within this analysis, they serve partly as a refinement of Meilby’s typology, partly as contradistinctions to the structure of the descriptive news analysis, and partly as a basis for stating common moves of news analysis as a genre. The Perilous Life of a Linguistic Genre Convention 69 Local patterns of moves in the text body may be repeated. This is particularly characteristic of the evaluating news analysis, i.e. “Explaining one action involved in the event”, “Evaluation”, “Explaining another action involved in the event”, “Evaluation”, etc. Optional moves like “Popularisation” (reiterations in figurative or metaphorical language, by analogy or by example, and marked by linguistic features, such as ‘in others words’, ‘in short’ or ‘that is’) appear relatively frequently in continuation of the moves “Elaborating the settings of the event by supplementary facts” and “Explaining the event”. Likewise, a number of moves, as well as steps, comprise elaborations in continuation of general information. Besides these local, optional patterns, no common pattern of steps was observed in the moves realised by the text body. Since the move-structure analysis serves as a basis for describing the linguistic realisation of a particular step in the opening move of the news analysis in general, I will concentrate on the two common moves of particular relevance to this realisation. 6.4 The common opening move: Offering an interpretation of an event As shown by the move-structure analysis, the two first moves are common to all six types of news analysis. Within the system of newspaper genres and their developmental history, these moves can be accounted for as essential for the purpose of news analysis. The first move, “Offering an interpretation of an event”, consists of two steps: “Framing the event” 3 and “Summarising the interpretation of the event”. The function of the first step is two-fold: cognitive and rhetorical. The cognitive function is to activate an overall frame of interpretation. The empirical basis for the attribution of this function is partly the observed relation between the frame represented by this step and the interpretative activities represented by the subsequent moves of news analyses, and partly the position of the part realising this step. As to the relation, the frame that can be derived from the specific linguistic choices characterising the heading is a prerequisite for the subsequent interpretative activities, whether it is an explanation, an evaluation, a prediction, a substantiated direction, or a judgement. Thus, the frame serves as a basis for attributing qualities, attitudes, wishes and motives to events and actions, for connecting events to other events or actions in the past or in the future (e.g. as predictions), and for assigning value to events and actions (Bartlett 1950 (1932), Minsky 1975, Schank & Abelson 1977, Entman 1993, Lakoff 2004). In other words, the game frame activated through the heading in the example above is a prerequisite for interpreting the results of a political negotiation as a play for high stakes (i.e. a strategic move), and, furthermore, it connects this move to the 3 The concept of framing, as used in this analysis, will be described in 6. Simon Borchmann 70 presumed image strategy of another political party. This relation is characteristic of all the texts in the corpus. As to the position, the activation of the frame that forms a basis for the interpretation of the text must be achieved in advance (Bransford & Johnson 1972, Rumelhart 1980, Gernsbacher 1996). Thus, the observation of the consistent realisation of the framing of the event with the first part of the text supports the assignment of this function to this part. One might object that these characteristics are based on basic, cognitive principles, which are common to most expository texts, including news articles. However, the function of the step does differ from the function of the corresponding parts of news articles, namely with regards to the rhetorical function. The rhetorical function of “Framing the event” is to offer an interpretation of an event. The empirical foundation for the assignment of this function is partly an observed characteristic of the informational status assigned to the central event, and partly an observed characteristic regarding the information value of the frames related to the step. The informational status of the central event in the news analysis differs from the informational status of the central event in the news article (paragraph or article with by-line). Whereas the central event is presented in the heading of news articles (se tab. 4), the central event is typically presented in the text body of news analyses. Furthermore, where the central event is always presented as new information in news articles (often in sentences with a broad focus), it is very often presented as presupposed information in news analyses. For example (central event in italics): 1) Heading Dansk folkeparti danser stammedans Framing the event Text body Der er i høj grad tale om en rituel stammedans, når Dansk Folkeparti lader op til finanslovsforhandlingerne med krav på udlændingeområdet. Central event presupposed Heading Dansk Folkeparti dance tribal dance Framing the event Text body It is very much a ritual tribal dance, when Dansk Folkeparti lets up to budget negotiations by making demands within the area of immigration . Central event presupposed The Perilous Life of a Linguistic Genre Convention 71 2) Heading Tættere på demokrati? Framing the event Text body Meget tyder på, at løsladelsen af Aung San Suu Kyi er en del af en større plan for at rehabilitere militær-styrets image og give efter i begrænset omfang for omverdenens pres. Central event presupposed Heading Closer to democracy? Framing the event Text body Evidence suggests that the release of Aung San Suu Kyi is part of a larger plan to rehabilitate the military regime’s image and to some extent yield to international pressure Central event presupposed 3) Heading Skift åbner konservativ fløjkrig Framing the event Text body Fyringen af Lene Espersen havde intet med politiske holdninger at gøre. Central event presupposed Heading Replacement opens war between conservative wings Framing the event Text body The dismissal of Lene Espersen had nothing to do with political views. Central event presupposed 4) Heading Fogh og Bendtsen på taktisk tilbagetog Framing the event Text body Ved at droppe tanken om at gribe ind over kommunerne har både Anders Fogh Rasmussen og Bendt Bendtsen erkendt, at det ville være en for stor foræring til Helle Thorning-Schmidt og resten af oppositionen at stå fast på, at indgåede aftaler skal overholdes. Central event presupposed Heading Fogh and Bendtsen on tactical retreat Framing the event Simon Borchmann 72 Text body By dropping the idea of taking action against municipalities, both Anders Fogh Rasmussen and Bendt Bendtsen have acknowledged that it would be too great a gift to Helle Thorning- Schmidt and the rest of the opposition to insist that agreements must be respected. Central event presupposed This observed difference in information structure is a remarkably-clear, linguistic indication of the division of labour between news analysis and news articles in the system of newspaper genres; whereas the news article offers news in the form of events that fulfil the news criteria, news analysis offers interpretations of such events. Thus, it is not the event, but the framing of the event that makes the reading of the news analysis informative. The assumption that “Framing the event” has a rhetorical function is also supported by a feature that differentiates this step from the use of frames in news articles. This difference can be accounted for by distinguishing between two types of framing: informative framing and trivial framing. The distinction is gradual rather than absolute, and it determines the extent to which the writer’s framing of an event can, and will be, made the object of criticism. Informative framing is characterised by the presence of relevant alternatives to the frame used. The use of the specific frame, thus, constitutes an informative choice, and it will be marked (Levinson 1995) in so far as the reader is aware of the alternatives. Therefore, the presence of alternatives forms the basis of a critique. Trivial framing, on the other hand, cannot reasonably be made the object of critique. What characterises trivial framing is that the use of the frame has survived a pressure of selection exerted through feedback from the environment (Harder 2011: 314ff), and this is an indication of its general applicability; it can be used transversely for special interests, beliefs and ideals. In other words, the elements of the event correspond to the default values of the frame (Minsky 1975). Therefore, the use of the frame is not informative, and rather than being a part of what is communicated, it forms the basis for the communication. There might, of course, be alternatives to the trivial framing, but these will be marked, whereas the trivial is unmarked and, thus, all things being equal, intuitive, automatic and subconscious. Notice that trivial framing, as well as informative framing, allows the reader to supply information to the information presented; this is the basic function of frames. But whereas the inferences allowed for in trivial framing are inferences to the stereotypical situation, the inferences allowed for in informative framing are not (and should not be) for the stereotypical situation (Levinson 1995). It is also worth noting that this distinction does not coincide with the ‘metaphorical/ non-metaphorical’ distinction, since a lot of metaphorical language use represents trivial framing within a given culture (Lakoff & Johnson 1980). The Perilous Life of a Linguistic Genre Convention 73 On the basis of this distinction, the difference between the step “Framing the event” and the function of the corresponding part in news articles can be described as follows: whereas “Framing the event” is typically informative, the corresponding section in news articles typically represents a trivial framing. A comparison of a selection of headings from the corpus and a selection of headings from paragraphs in the Danish newspaper, Information, on November 18 th and 19 th , can serve as an illustration of the difference: Headings of news analyses Headings of news paragraphs USA’s troværdighed smuldrer The US’s credibility is eroding Rønns selvbillede krakelerer Rønn’s self-image crackles VKO svigter de unge - og velfærden VKO forsake the young - and welfare EU blev ikke mere populært The EU did not become more popular Japansk nyvalg skal redde økonomien Japanese new election to save the economy Problemet bliver at vinde - ikke at regere The problem will be to win - not to rule Banker skal give aktionærer mere magt Banks must give shareholders more power Hovedløs liberalisering var krisens moder Rash liberalization was the mother of the crisis Ny offentlighedslov begrænser åbenhed New act on Public Access limits transparency Saad Hariri kan kun fortsætte på Hizbollahs nåde Saad Hariri can only continue by the grace of Hezbollah Så blev gigantisk Lotto-pulje endelig udløst The giant Lotto pool was finally released Kvinde sigtet for pædofile overgreb Woman accused of paedophile offenses 11 danske film har været Oscar-nomineret 11 Danish films have been Oscar-nominated Pårørende vil stoppe dansk Breivik-teater Relatives want to stop Danish Breivik theatre Norsk PET-chef trækker sig omgående Norwegian PET chief withdraws immediately LA vil stemme nej til finansloven LA will vote against budget Dansk præst har deltaget aktivt i krig Danish priest has actively participated in war Mange besætningsmedlemmer hjalp ved forlis Many crew members helped at the accident To dræbt ved Israels luftangreb i Gaza Two killed in Israeli airstrike in Gaza V: Regeringen har løjet sig til magten V: The government has lied its way to power Tab 4: Comparision of headings: informative framing vs. trivial framing I take it that the conclusive difference can be acknowledged intuitively, even without knowledge of the events described; it is quite easy for somebody to disagree with the statements in the left-hand column, but rather difficult to disagree with the statements in the right-hand column. This assumption is Simon Borchmann 74 based on our pragma-linguistic knowledge; the typical use of combinations of linguistic expressions. Therefore, most of the differences can be traced back to the specific collocations: claims, such as, the “credibility” of a state is “eroding”, or a “self-image crackles”, or a government “forsakes” political areas, etc., leave plenty of room for interpretation and, thus, are a matter of judgement. But whether or not a “lottery pool” has been “paid out”, a person has been “charged with” a criminal offence, a “film” has been “nominated for an Oscar”, etc., is a matter of fact; it leaves no room for interpretation. This difference is a perfect reflection of the division of labour between news articles and news analyses. The news analysis headings represent interpretations, the acceptability of which can be discussed; sometimes even without the possibility to decide whether it is an acceptable interpretation or not. Conversely, the news paragraph headings represent statements that are typically verifiable, and, in so far as we regard them as factual statements, we will accept them by default. It is also worth nothing that, when an informative framing occurs in the news paragraph - for example, “V: The government has lied to get to the power” -, it is embedded, and the source is referred to. In this way, what is stated in the heading is that representatives from the political party, Venstre, have claimed that the government has lied to gain power. And this statement is verifiable. Conversely, the writer of the directive news analysis can present an utterance like, “Banks should transfer more power to the shareholders” as his or her own. As to the predictive news analysis, it should be mentioned that the predictions are typically verifiable, e.g. “No job release scheme negotiations before election”. However, at the time, it is stated that they cannot be verified and, thus, they have the status of interpretations, which may or may not be considered acceptable. I claim that this difference is an observable norm: all thing being equal, the framing of the event in a news analysis is (and should be) informative; all things being equal, the framing of the event in news paragraphs is (and should be) trivial. There are, of course, deviations from this norm, but this applies to all norms. Besides reflecting the division of labour within the present system of newspaper genres, the informational characteristic of the step can be described as a necessary consequence of the developmental history of the genre. News analysis emerged, among other things, as a response to an increased complexity of the political process and the professional ideal of critical journalism. Now, to serve as a social act responding to these needs, the writer of the news analysis text has to address a complex situation and attempt to go behind the immediate, available facts. The characteristic of a complex situation is that there is no satisfactory overall interpretation, but only a set of local or competing, alternative interpretations. Thus, if a news analysis text serves its purpose as a social act, the framing of the event will be informative. However, this characteristic of the step “Framing the event” does not distinguish the news analysis from opinion articles, e.g. editorials and columns. The Perilous Life of a Linguistic Genre Convention 75 Within these genres, the framing of events is also typically informative; otherwise, persuasion would be unnecessary. On the other hand, there seems to be a difference regarding the linguistic realisation. This difference will be described in the third part of the analysis. The step “Summarising the interpretation of the event” is simply a short presentation of the interpretation offered by the writer of the text. As a summarising act, the step is neither distinctive nor essential for the purpose of the news analysis. It can be seen as a general principle for organising newspaper texts, presumably originating from the days of the telegraph, where the line of communication could be broken at any moment (Bro 2009). Nevertheless, the step does serve a function in the performance of the move “Offering an interpretation of an event”. Due to the limited space allocated to the realisation of the step “Framing an event”, the linguistic cues to the interpretation are not always sufficient to serve as an instruction, selecting the frame relevant to the interpretation. Thus, the step “Summarising the interpretation” serves as an elaboration of the interpretive instruction carried by the step “Framing the event”. The allocation of this function to the lead of an article is in agreement with journalistic textbooks; thus, one of the fundamental functions of the lead is to identify the angle (Andersson-EK et al 2001). 6.5 The common second opening move: Establishing an exigency As with the first move, the second move of the news analysis, “Establishing an exigency”, is essential to the purpose of news analysis and can be explained in terms of the division of labour that characterises the Danish system of newspaper genres. In Denmark, news analysis was developed not as an addition to opinion journalism, but as an addition to news journalism. As indicated above, the practitioners actually view the distinction as a virtue. However, taking news as a starting point, news analysis is characterised by the following flaw: that the event analysed in news analysis does not fulfil the news criteria; in other words, the event has either been presented in news articles, or it is of peripheral interest. Thus, taking up the event must be justified journalistically in some other way. This is done by designing an exigency taking the event as a starting point. There are some similarities between this concept and Bitzer’s exigence (and Hauser’s sophistication of the concept). An exigency represents a rhetorical opportunity and justifies the text as rhetorical act, in so far as the writer addresses the exigency with the text. However, the concept used here is not restricted to the rhetorical situation of the writer. On the basis of the corpus, I distinguish between two types of exigencies, which represent a rhetorical opportunity and justify news analysis as a rhetorical act: factual and interpretive. Factual exigencies refer to the events and people described. These exigencies are problems experienced by the people involved in the events, and they comprise impending assignments and decisions, sudden changes, negative conse- Simon Borchmann 76 quences of actions and events, illegalities, discrepancies between utterances, between actions or between utterances and actions, and other kinds of problems. For example: 5) På cirka dette tidspunkt sidste år nød den britiske liberale formand, Nick Clegg, en popularitet på højde med Winston Churchills efter sejren over Nazityskland. Meningsmålingerne dokumenterede det. Vælgerne elskede hans tjekkede, veltalende stil, hans løfter om en »ny politik« og hans friske ideer. Supporting the exigency by settings Men sikke en forskel et år og magt gør. Ved valget i torsdags fik Clegg og hans Liberaldemokrater ikke kun en blodtud af vælgerne, men noget nær en knockout. Presuming the exigency Deres brændende ønske om valgreform indførelsen af det såkaldte AV-valgsystem blev afvist med massive 69 procent af stemmerne. Og dertil mistede partiet hele 695 byrådsmedlemmer ved lokalvalget og gik i procent tilbage fra 24 procents tilslutning ved valget i 2007 til 15 procent ved torsdagens valg. Identifying the event At about this time last year the British Liberal leader, Nick Clegg, enjoyed a popularity on a par with Winston Churchill after the victory over Nazi Germany. The polls documented it. Voters loved his cool, eloquent style, his promises of a ‘new politics’ and his fresh ideas. Supporting the exigency by settings But what a difference a year and power make. In the election last Thursday Clegg and his Liberal Democrats got not only a bloody nose from the voters, but something close to a knockout. Presuming the exigency Their burning desire for electoral reform, the introduction of the so-called AV electoral system was rejected with a massive 69 percent of the votes. Furthermore, the party lost a total of 695 councillors in local elections and went back in percentage from the 24 percent approval in the election in 2007 to 15 percent at Thursday’s election. Identifying the event The event identified is an election. By focussing on the election’s negative consequences for a specific political party, an exigency is established. The exigency is supported by referring to an advantageous position that the political party previously enjoyed. By relating the event to other events and actions, the event forms part of a whole that represents an exigency, and that fulfils the The Perilous Life of a Linguistic Genre Convention 77 journalistic news criteria: drastic changes in the support of a government party is important. In this way, the news analyses are justified as rhetorical acts, without attempting to dissolve the exigency, but simply by establishing and exposing it. Therefore, these exigencies differ from rhetorical exigencies. An interpretative exigency is designed as a deficiency of interpretation, and, thus, constitutes a cognitive problem. These exigencies comprise three types of deficiency: 1) the rhetorical situation is designed as though public opinion is characterised by an established, but insufficient (or simply wrong), interpretation of an event. The writer presents his or her interpretation as an alternative, contrastive, sufficient interpretation. I call this deficiency false belief. 2) The rhetorical situation is designed as though public opinion is characterised by a small set of competing, mutually-exclusive interpretations of an event, and the writer supplies the reader with information representing a preference for one of the interpretations. I call this deficiency doubt. 3) The rhetorical situation is designed as though the set of interpretations of an event is not clearly limited. In other words, there is neither an established interpretation nor a small set of mutually-exclusive alternatives. The writer presents an interpretation of the event. I call this deficiency lack of knowledge. The following is an example of false belief (the event is indicated with italics): 6) Finanskrisen var ikke forårsaget af grådighed. Presuming the exigency Embedded: Identifying the event Siden udbruddet af finanskrisen er der alligevel skrevet artikel efter artikel om ‘grådighedens årti’, som om finanskrisen var en (pludselig) opblomstring af grådighed i den finansielle sektor. Men har vi mennesker ikke til stadighed været grådige, og vil vi ikke fortsat være det? Supporting the exigency by settings The financial crisis was not caused by greediness. Presuming the exigency Embedded: Identifying the event Since the outbreak of the financial crisis, article upon article have been written about the “decade of greed” as if the financial crisis was a (sudden) bloom of greed in the financial sector. But have we humans not always been greedy, and will we not remain so? Supporting the exigency by settings Simon Borchmann 78 The event is identified as the financial crisis. The writer implies that there is an established interpretation of this event, namely that is was caused by greediness, that is, a psychological framing of the event. By using a political frame, the writer interprets the event as politically determined, and this interpretation is presented as a contrastive alternative to the psychological. The following is an example of doubt: 7) Om Myanmar er tættere på demokrati efter Aung San Suu Kyis løsladelse i weekenden, er et af de store spørgsmål, som omverdenen stiller sig selv i disse dage. Presuming the exigency Embedded: Identifying the event For kun en uge siden konsoliderede militærjuntaens parti militærets magt. Landets første parlamentsvalg i 20 år blev afholdt, og militæret står til en valgsejr på over 80 procent. Supporting the exigency by settings Whether Myanmar is closer to democracy after Aung San Suu Kyi’s release this weekend, is one of the major questions the surrounding world is asking itself these days. Presuming the exigency Embedded: Identifying the event Only a week ago the military junta party consolidated the military’s power. The country’s first elections in 20 years were held, and the military stands to an electoral victory of over 80 percent. Supporting the exigency by settings The event is identified as the release of Aung San Suu Kyi. The exigency is formed by doubting whether or not this event is a sign that Myanmar is getting closer to democracy. The following is an example of lack of knowledge: 8) Nationens politiske iagttagere sidder i disse timer og river hår ud af hovedbunden i hidtil forgæves forsøg på at uddrage en slags rationale af statsministerens determinerede nytårsudtalelse om regeringens påtænkte foranstaltninger mod den påståede voldsoptrapning i det danske samfund. Forundringen siden er ikke just blevet mindre, efter at den nyslåede justitsminister Frank Jensen har sat konkrete boller på knippelsuppen og bebudet diverse skærpelser af strafferammerne, herunder indespærring af mindreårige voldsmænd(drenge) i passende forbedringsanstalter. Presuming the exigency Embedded: Identifying the event The Perilous Life of a Linguistic Genre Convention 79 Lidet har det anfægtet de socialdemokratiske topministre at allerede eksisterende praksis, set i et forholdsvis kort tidsperspektiv, rent faktisk har betydet ganske betydelig skærpelse af straffene i de såkaldt meningsløse gadevoldssager. Lige så lidt har det åbenbart betydet at ministrenes umiddelbare bagland, som ifølge sagens natur er den radikale regeringspartner, ikke med et kvæk er blevet taget med på råd, inden hhv. statsminister og justitsminister udtalte sig på den samlede regerings vegne. Supporting the exigency by settings The nation’s political observers are in these hours tearing their hair out in up to now unsuccessful attempts to extract some kind of rationale from the Prime Minister’s determined New Year’s opinion on the government’s proposed measures against the alleged escalation of violence in the Danish society. The amazement has not exactly diminished after the newly knighted Minister of Justice Frank Jensen has made concrete preparations for a working-over by the truncheon and announced an extensive tightening of penalties, including incarceration of minor assailants (boys) in appropriate improvement institutions. Presuming the exigency Embedded: Identifying the event Little did it affect the social democratic top ministers that the already existing practice, seen in a relatively short time frame, actually meant quite a significant tightening of sanctions in the so-called senseless street violence cases. Apparently, it mattered so little to them that the ministers’ support base, which, in the nature of the case, is the Radical liberal government partner, was not consulted in the least before, respectively the Prime Minister and the Minister of Justice expressed their opinion on the overall government’s behalf. Supporting the exigency by settings The event is identified as the Prime Minister’s announcement of legislative initiatives against violence, including the Minister of Justice’s follow-up remarks on the tightening of punitive measures. The exigency is designed by stating that the political observer’s attempts to provide a rational explanation have been futile. This exigency, thus, constitutes the opportunity for the writer to present his interpretation as new information. Besides the abovementioned types of interpretive exigencies, there seems to be a borderline case. This is when the rhetorical situation is designed as if the use of the frame on the subject matter were established. The writer pre- Simon Borchmann 80 sents his or her interpretation as a default interpretation. I call these cases reiteration. The following is an example of reiteration: 9) Heading Fogh og Bendtsen på taktisk tilbagetog Framing the event Lead Sjældent har en valgkamp været kendetegnet ved så mange taktiske tilbagetog og frontafkortninger, som vi er vidne til nu. Summarising the interpretation of the event Text body Ved at droppe tanken om at gribe ind over kommunerne har både Anders Fogh Rasmussen og Bendt Bendtsen erkendt, at det ville være en for stor foræring til Helle Thorning- Schmidt og resten af oppositionen at stå fast på, at indgåede aftaler skal overholdes. Presuming the exigency Embedded: Identifying the event For få dage siden var det ellers parolen, at aftalen skulle overholdes, Supporting the exigency by settings Heading Fogh and Bendtsen on tactical retreat Framing the event Lead An election campaign has rarely been marked by so many tactical retreats and closure of flanks as we are witnessing now. Summarising the interpretation of the event Text body By dropping the idea of taking action against municipalities, both Anders Fogh Rasmussen and Bendt Bendtsen have acknowledged that it would be too great a gift to Helle Thorning- Schmidt and the rest of the opposition to insist that agreements must be respected. Presuming the exigency Embedded: Identifying the event A few days ago, however, the slogan was that the agreement had to be respected, Supporting the exigency by settings The exigency is, allegedly, that the government refrains from a planned intervention in the municipality budgets, due to their over-spending. This U-turn can hardly be regarded as a factual exigency. According to the distinction, the justification of news analysis as a rhetorical act must be that the U-turn de- The Perilous Life of a Linguistic Genre Convention 81 mands interpretation. Now, such events might be accepted as exigencies according to professional journalistic ideals. However, the use of the war frame to interpret this U-turn as a strategic move in election campaign warfare is presented as established. Notice that the linguistic cues, which identify the war frame, are presented as presupposed information in the lead. Thus, it is presupposed by the writer that election campaigns are always characterised by tactical retreats and the closing of flanks. It is also worth noting that the frame is a prerequisite for the understanding of the expression ‘foræring’ (present) as an open flank. Of course, the use of the war frame in the interpretation of political decisions and negotiations is well established in Danish political journalism. But this also means that the information value of the frame is severely restricted. In other words, the readers of news analyses are able to make such interpretations on their own if required. Therefore, reiteration is a borderline case of interpretive exigencies. Contrary to factual exigencies, the interpretive exigencies are addressed and more or less dissolved by the news analysis text as a rhetorical act. Thus, they correspond to rhetorical exigencies (Bitzer 1997). The writer justifies news analysis as a rhetorical act by designing an exigency. Both types of exigencies are legitimate according to the professional journalistic ideals observed by media researchers. Thus, both the exposure of factual exigencies and the dissolution of interpretive exigencies can be seen as attempts to fulfil the ideals of a more engaged, critical, and investigative journalism. 6.6 The structure and linguistic realisation of the move Establishing an exigency As shown above, establishing an exigency comprises three steps: “Identifying the event”, “Supporting the exigency by settings”, and “Presuming the exigency”. The order of the steps varies. But the functions and the linguistic realisation are distinctive, irrespective of the order. Identifying the event is a basic function of the news analysis text as a message; to serve as an exchange of information, the information must be about something, and this something must be identified (Clark & Havilland 1977, Gernsbacher 1996). This informational status corresponds to the (situational) topic of the text. As the exigency and the interpretation is starting from a particular event, and, as this central event has already been presented as new information in a news article, this event serves as the situational topic of the text. In agreement with this - and in accordance with the division of labour between news articles and news analyses - the event is very often presented as presupposed information. Likewise, the step identifying the event is often embedded in the step “Presuming the exigency” (see above). The linguistic features of the step “Identifying the event”, thus, are typically presupposition Simon Borchmann 82 triggers (Levinson 1983, Togeby 1993), with which the writer presupposes the accessibility of an actual event in memory (see example 1-4). As observed above, the event analysed in the news analysis does not fulfil the news criteria. Adopting the event as a topic is justified journalistically by designing an exigency starting from the event. Now, “Supporting the exigency by settings” serves an essential function in this design. By relating the topical event (only) temporally to other events and actions, the writer implicates (Grice 1975, Levinson 1995) notable motivational inferences, causal inferences, discordances or changes. The function, thus, is completed by relating the topical event to other events and by indicating the necessary (elaborative as well as bridging) textual inference (Sanford 1990). In agreement with this, this step is distinctly marked by the combination of temporal connectors and sentence adverbials, which mark sender relations and focus, and, thereby, indicate the necessary inference. The temporal connectors are typically temporal adverbs, and are very often positioned in the fundament (first position in the Danish sentence schema), but they may also be designated by changes in tense - from the present to the perfect, from the perfect to the past, from the present to the past (or vice versa, depending on the design and the order of the steps) - or combinations of temporal adverbs and tense shifting. Other ways of expressing temporal relations might also be used. The setting is typically past events or states, but it may also be subsequent events or states. The following are examples of temporal connectors (features underlined): 10) På cirka dette tidspunkt sidste år nød den britiske liberale formand, Nick Clegg, en popularitet på højde med Winston Churchills efter sejren over Nazityskland. Meningsmålingerne dokumenterede det. Vælgerne elskede hans tjekkede, veltalende stil, hans løfter om en »ny politik« og hans friske ideer. Supporting the exigency by settings Men sikke en forskel et år og magt gør. Ved valget i torsdags fik Clegg og hans Liberaldemokrater ikke kun en blodtud af vælgerne, men noget nær en knockout. Presuming the exigency Deres brændende ønske om valgreform indførelsen af det såkaldte AV-valgsystem blev afvist med massive 69 procent af stemmerne. Og dertil mistede partiet hele 695 byrådsmedlemmer ved lokalvalget og gik i procent tilbage fra 24 procents tilslutning ved valget i 2007 til 15 procent ved torsdagens valg. Identifying the event At about this time last year the British Liberal leader, Nick Clegg, enjoyed a popularity on a par with Winston Churchill after the victory over Nazi Germany. The polls documented it. Supporting the exigency by settings The Perilous Life of a Linguistic Genre Convention 83 Voters loved his cool, eloquent style, his promises of a ‘new politics’ and his fresh ideas. But what a difference a year and power make. In the election last Thursday Clegg and his Liberal Democrats got not only a bloody nose from the voters, but something close to a knockout. Presuming the exigency Their burning desire for electoral reform, the introduction of the so-called AV electoral system was rejected with a massive 69 percent of the votes. Furthermore, the party lost a total of 695 councilors in local elections and went back in percentage from the 24 percent approval in the election in 2007 to 15 percent at Thursday’s election. Identifying the event 11) Finanskrisen var ikke forårsaget af grådighed. Presuming the exigency Embedded: Identifying the event Siden udbruddet af finanskrisen er der alligevel skrevet artikel efter artikel om ‘grådighedens årti’, som om finanskrisen var en (pludselig) opblomstring af grådighed i den finansielle sektor. Men har vi mennesker ikke til stadighed været grådige, og vil vi ikke fortsat være det? Supporting the exigency by settings The financial crisis was not caused by greediness. Presuming the exigency Embedded: Identifying the event Since the outbreak of the financial crisis article after article have been written about “the decade of greed” as if the financial crisis was a (sudden) bloom of greed in the financial sector. But have we humans not always been greedy, and will we not remain so? Supporting the exigency by settings Simon Borchmann 84 12) Om Myanmar er tættere på demokrati efter Aung San Suu Kyis løsladelse i weekenden, er et af de store spørgsmål, som omverdenen stiller sig selv i disse dage. Presuming the exigency Embedded: Identifying the event For kun en uge siden konsoliderede militærjuntaens parti militærets magt. Landets første parlamentsvalg i 20 år blev afholdt, og militæret står til en valgsejr på over 80 procent. Supporting the exigency by settings Whether Myanmar is closer to democracy after Aung San Suu Kyi’s release this weekend, is one of the major questions the surrounding world is asking itself these days. Presuming the exigency Embedded: Identifying the event Only a week ago the military junta party consolidated the military’s power. The country’s first elections in 20 years were held, and the military stands to an electoral victory of over 80 percent. Supporting the exigency by settings 13) De brede forhandlinger om satspuljemidlerne på det sociale område er slut, Identifying the event men VKO’s ordførere og socialminister Benedikte Kiær (K) måtte gå fra bordet velvidende, at der manglede 80 millioner kroner til finansieringen af den højt profilerede ghettoplan. Presuming the exigency Og det bare ét døgn, efter at fire ministre med Kiær i spidsen fremlagde planen, som skulle være et af Løkke-regeringens hovednumre i år. Supporting the exigency by settings The broad negotiations on adjustment pool funds for social services are over, Identifying the event but VKO’s spokesmen and Minister of Social Affairs Benedikte Kiær (K) had to leave the table knowing that 80 million kroner needed to finance the high-profile ghetto plan were missing. Presuming the exigency And this is just one day after four ministers headed by Kiær presented the plan, which was meant to be one of the Løkke government’s main numbers this year. Supporting the exigency by settings The Perilous Life of a Linguistic Genre Convention 85 14) Listen af konservative formænd, der er blevet væltet, er lang, og fortællingen om fløjkrige veldokumenteret. Supporting the exigency by settings Men denne gang er forskellen bemærkelsesværdig; fyringen af Lene Espersen har intet med politisk uenighed at gøre. Presuming the exigency Hun har indtil for nylig haft gruppens ubetingede opbakning til trods for en nedtur, der trak ud i månedsvis. Supporting the exigency by settings I går tog Lene Espersen konsekvensen af meningsmålinger, som på dagen nåede et historisk lavpunkt med kun fem procents tilslutning. Identifying the event The list of conservative leaders who have been overthrown, is long and the story of wars between wings is well documented. Supporting the exigency by settings But this time the difference is remarkable; the dismissal of Lene Espersen has nothing to do with political differences. Presuming the exigency She has until recently had the group’s unconditional support despite a decline that dragged on for months. Supporting the exigency by settings Yesterday Lene Espersen took the consequence of the polls, which on the day had reached a historic low with only five percent popular support for the party. Identifying the event 15) SF-formand Villy Søvndal stod i går op til historien om, at 62 procent af hans egne lokalformænd er lodret imod regeringens indførsel af et nyt pointsystem for familiesammenføringer. Da dagen var omme, sad han alligevel med mandat fra sin landledelse til at acceptere et pointsystem efter et regeringsskifte. Supporting the exigency by settings Embedded: Identifying the event At Søvndal når til det slutresultat, er det synlige eksempel på, hvor meget SF har ændret sig - ikke bare politisk, men også organisatorisk under hans år som formand. I dag er det Villy Søvndal og ikke baglandet, der bestemmer. Presuming the exigency Simon Borchmann 86 SF leader Søvndal got up yesterday to the story that 62 percent of his own local leaders are diametrically opposed to the government’s introduction of a new scoring system for family reunification. When the day was over, he nevertheless had a mandate from his Executive Committee to accept a scoring system after a change of government. Supporting the exigency by settings Embedded: Identifying the event That Søvndal achieves this result is the visible example of how much SF has changed - not just politically but also organisationally during his years as leader of the party. Today it is Villy Søvndal and not the support base who is in charge Presuming the exigency It is quite remarkable that, typically, the relation is only marked as temporal, and that the more informative relation, therefore, is to be inferred by the reader. However, there are a few occurrences where the most informative relation is, in fact, explicitly marked. This is characteristic of one judgemental and one mixed judgemental and descriptive news analysis (features underlined): 16) Det var ikke kun regeringen, der i går blev ramt af endnu en sag om misinformation af Folketinget i forbindelse med en ny lov om hjemsendelse af afghanske asylbørn. Det var i særdeleshed integrationsminister Birthe Rønn Hornbech (V), Presuming the exigency Embedded: Identifying the event fordi hun som uddannet jurist har gjort det til sit politiske adelsmærke at kræve høj standard for lovgivningsarbejdet. Supporting the exigency by settings It was not just the government which yesterday was hit by yet another case of misinformation of Folketinget in connection with a new act on repatriation of Afghan asylum children. It was particularly Minister of Integration Birthe Rønn Hornbech (V), Presuming the exigency Embedded: Identifying the event because she as a trained lawyer has made it her political hallmark to require high standards of legislative work. Supporting the exigency by settings The Perilous Life of a Linguistic Genre Convention 87 Examples of adverbials marking sender relations and focus follow below. A supplementary observation, which supports the account of the function of the move, “Establishing an exigency”, is that the inferences necessary to the establishment of the exigency are clearly indicated by the relevance structure of the sentences in the step. To illustrate this point, the relevance structure is annotated in the following examples (focus in bold, other features underlined): 17) Nationens politiske iagttagere sidder i disse timer og river hår ud af hovedbunden i hidtil forgæves forsøg på at uddrage en slags rationale af statsministerens determinerede nytårsudtalelse om regeringens påtænkte foranstaltninger mod den påståede voldsoptrapning i det danske samfund. Forundringen siden er ikke just blevet mindre, efter at den nyslåede justitsminister Frank Jensen har sat konkrete boller på knippelsuppen og bebudet diverse skærpelser af strafferammerne, herunder indespærring af mindreårige voldsmænd(drenge) i passende forbedringsanstalter. Presuming the exigency Embedded: Identifying the event Lidet har det anfægtet de socialdemokratiske topministre at allerede eksisterende praksis, set i et forholdsvis kort tidsperspektiv, rent faktisk har betydet ganske betydelig skærpelse af straffene i de såkaldt meningsløse gadevoldssager. Lige så lidt har det åbenbart betydet at ministrenes umiddelbare bagland, som ifølge sagens natur er den radikale regeringspartner, ikke med et kvæk er blevet taget med på råd, inden hhv. statsminister og justitsminister udtalte sig på den samlede regerings vegne. Supporting the exigency by settings The nation’s political observers are in these hours tearing their hair out in up to now unsuccessful attempts to extract some kind of rationale from the Prime Minister’s determined New Year’s opinion on the government’s proposed measures against the alleged escalation of violence in the Danish society. The amazement is not just diminishing after the newly knighted Minister of Justice Frank Jensen has made concrete preparations for a working-over by the truncheon and announced an extensive tightening of penalties, including incarceration of minor assailants (boys) in appropriate improvement institutions. Presuming the exigency Embedded: Identifying the event Simon Borchmann 88 Little did it affect the social democratic top ministers that the already existing practice, seen in a relatively short time frame, actually meant quite a significant tightening of sanctions in the so-called senseless street violence cases. Apparently, it mattered so little to them that the ministers’ support base, which, in the nature of the case, is the Radical liberal government partner, were not consulted in the least before, respectively the Prime Minister and the Minister of Justice expressed their opinion on the overall government’s behalf. Supporting the exigency by settings 18) Finanskrisen var ikke forårsaget af grådighed. Presuming the exigency Embedded: Identifying the event Siden udbruddet af finanskrisen er der alligevel skrevet artikel efter artikel om ‘grådighedens årti’, som om finanskrisen var en (pludselig) opblomstring af grådighed i den finansielle sektor. Men har vi mennesker ikke til stadighed været grådige, og vil vi ikke fortsat være det? Supporting the exigency by settings The financial crisis was not caused by greediness. Presuming the exigency Embedded: Identifying the event Since the outbreak of the financial crisis article upon article have nevertheless been written about “the decade of greed” as if the financial crisis was a (sudden) bloom of greed in the financial sector. But have we humans not always been greedy, and will we not remain so? Supporting the exigency by settings 19) Ved at droppe tanken om at gribe ind over kommunerne har både Anders Fogh Rasmussen og Bendt Bendtsen erkendt, at det ville være en for stor foræring til Helle Thorning-Schmidt og resten af oppositionen at stå fast på, at indgåede aftaler skal overholdes. Presuming the exigency Embedded: Identifying the event The Perilous Life of a Linguistic Genre Convention 89 For få dage siden var det ellers parolen, at aftalen skulle overholdes, Supporting the exigency by settings By dropping the idea of taking action against municipalities, both Anders Fogh Rasmussen and Bendt Bendtsen have acknowledged that it would be too great a gift to Helle Thorning-Schmidt and the rest of the opposition to insist that agreements must be respected. Presuming the exigency Embedded: Identifying the event But a few days ago the slogan was that the agreement had to be respected, Supporting the exigency by settings 20) Om Myanmar er tættere på demokrati efter Aung San Suu Kyis løsladelse i weekenden, er et af de store spørgsmål, som omverdenen stiller sig selv i disse dage. Presuming the exigency Embedded: Identifying the event For kun en uge siden konsoliderede militærjuntaens parti militærets magt. Landets første parlamentsvalg i 20 år blev afholdt, og militæret står til en valgsejr på over 80 procent. Supporting the exigency by settings Whether Myanmar is closer to democracy after Aung San Suu Kyi’s release this weekend, is one of the major questions the surrounding world is asking itself these days. Presuming the exigency Embedded: Identifying the event Only a week ago the military junta party consolidated the military’s power. The country’s first elections in 20 years were held, and the military stands to an electoral victory of over 80 percent. Supporting the exigency by settings 21) De brede forhandlinger om satspuljemidlerne på det sociale område er slut, Identifying the event men VKO’s ordførere og socialminister Benedikte Kiær (K) måtte gå fra bordet velvidende, at der manglede 80 millioner kroner til finansieringen af den højt profilerede ghettoplan. Presuming the exigency Simon Borchmann 90 Og det bare ét døgn, efter at fire ministre med Kiær i spidsen fremlagde planen, som skulle være et af Løkke-regeringens hovednumre i år. Supporting the exigency by settings The broad negotiations on adjustment pool funds for social services are over, Identifying the event but VKO’s spokesmen and Minister of Social Affairs Benedikte Kiær (K) had to leave the table knowing that 80 million kroner needed to finance the high-profile ghetto plan were missing. Presuming the exigency And this is just one day after four ministers headed by Kiær presented the plan, which was meant to be one of the Løkke government’s main numbers this year. Supporting the exigency by settings 22) SF-formand Villy Søvndal stod i går op til historien om, at 62 procent af hans egne lokalformænd er lodret imod regeringens indførsel af et nyt pointsystem for familiesammenføringer. Da dagen var omme, sad han alligevel med mandat fra sin landledelse til at acceptere et pointsystem efter et regeringsskifte. Supporting the exigency by settings Embedded: Identifying the event At Søvndal når til det slutresultat, er det synlige eksempel på, hvor meget har ændret sig - ikke bare politisk, men også organisatorisk under hans år som formand. I dag er det Villy Søvndal og ikke baglandet, der bestemmer. Presuming the exigency SF leader Søvndal got up yesterday to the story that 62 percent of his own local leaders are diametrically opposed to the government’s introduction of a new scoring system for family reunification. When the day was over, he nevertheless had a mandate from his Executive Committee to accept a scoring system after a change of government. Supporting the exigency by settings Embedded: Identifying the event That Søvndal achieves this result is the visible example of how much SF has changed - not just politically but also organisationally during his years as leader of the party. Today it is Villy Søvndal and not the support base who is in charge. Presuming the exigency The Perilous Life of a Linguistic Genre Convention 91 Referring to the account of the move, “Establishing an exigency”, the function of the step, “Presuming the exigency”, is to justify the news analysis as a rhetorical act. This function is completed by identifying a factual exigency or an interpretive exigency. The linguistic features of the step depend on the type of exigency. Thus, each of the three interpretive exigencies is distinctly marked by grammatical functions. The linguistic features of the false belief design are partly markers of contrast, e.g. negations, adversative conjunctions, written language indicators of contrastive stress (for example, marked constructions), and partly words or collocations expressing irrealis, scepticism or reservations (features underlined): 23) Finanskrisen var ikke forårsaget af grådighed. Siden udbruddet af finanskrisen er der alligevel skrevet artikel efter artikel om ‘grådighedens årti’, som om finanskrisen var en (pludselig) opblomstring af grådighed i den finansielle sektor. The financial crisis was not caused by greediness. Since the outbreak of the financial crisis article upon article have nevertheless been written about “the decade of greed” as if the financial crisis was a (sudden) bloom of greed in the financial sector. But have we humans not always been greedy, and will we not remain so? 24) Hvis en regeringsleders evne til at regere var betinget af, at 90 folketingsmedlemmer havde den samme tilgang til den økonomiske politik, ville Danmark være et uregerligt land. Det er det ikke, det har det ikke været i mange år, og det behøver det absolut heller ikke blive for Helle Thorning-Schmidt (S), selv om Radikale Venstre skulle indgå et efterlønsforlig med VKO. (...) Skulle det alligevel ske, vil det måske nok være en ydmygende start på en regeringsperiode, men spørgsmålet er, om det vil være det katastrofescenario, som rigtigt mange af næsten lige så mange forskellige grunde har forsøgt at fremmane de seneste uger. If a government leader’s ability to govern were conditioned by 90 MPs having the same approach to the economic policy, Denmark would be an ungovernable country. It is not, it has not been for many years and it absolutely does not have to be so for Helle Thorning-Schmidt (S), even if Radikale Venstre were to enter into a pension benefit compromise with VKO. (...) In the unlikely event, it will probably be a humiliating start to a reign, but the question is whether it will be the disaster scenario that a lot of people, Simon Borchmann 92 because of almost as many different reasons, have attempted to evoke the past weeks. 25) Umiddelbart ligner det en degradering, når man forlader en regering for at tage plads blandt de almindelige folkevalgte og ansatte i et regeringsbærende parti som Venstre. Det gælder for tidligere skatteminister Kristian Jensen, der nu er formand for folketingsgruppen. Og det gælder Søs Marie Serup, der i dag har sidste arbejdsdag som Løkkes politiske rådgiver i Statsministeriet og i morgen sætter sig i en helt ny stilling som chef for Venstres nyoprettede Center for kommunikation og Politik. Men så enkelt er det ikke nødvendigvis. For Kristian Jensen og Søs Marie Serup får i samspil en helt afgørende rolle for udviklingen i dansk politik i den kommende tid. At first sight it looks like a demotion when leaving a government to take place among the common popularly selected and employees of a government-supporting party like Venstre. It applies to former Minister of Taxation Kristian Jensen, who is now chairman of the parliamentary group. And this applies to Søs Marie Serup, who today has her last working day as Løkke’s political adviser in the Prime Minister’s Department and tomorrow sits in a completely new position as head of Venstre’s newly created Centre for Communications and Policy. But it is not necessarily that simple. Because Kristian Jensen and Søs Marie Serup in interaction will get a crucial role in the development of Danish policy in the future. The linguistic features of the doubt design are interrogative signals (typically indicating yes-no questions), exclusive conjunctions and/ or words or collocations expressing doubt (features underlined). 26) Om Myanmar er tættere på demokrati efter Aung San Suu Kyis løsladelse i weekenden, er et af de store spørgsmål, som omverdenen stiller sig selv i disse dage. Whether Myanmar is closer to democracy after Aung San Suu Kyi’s release this weekend, is one of the major questions the surrounding world is asking itself these days. 27) Clintons besøg fik Kim Jong-il til at benåde de to journalister, der ulovligt havde krydset grænsen til Nordkorea. Den 36-årige Euna Lee og 32-årige Laura Ling var ellers blevet dømt til 12 års arbejdslejr for forseelsen, som de selv efterfølgende har indrømmet. Betyder det så, at den 68-årige diktator er blevet, lad os sige mere humanitær, på sine gamle dage? The Perilous Life of a Linguistic Genre Convention 93 Clinton’s visit made Kim Jong-il pardon the two journalists who had illegally crossed the border into North Korea. The 36-year-old Euna Lee and 32year-old Laura Ling had otherwise been sentenced to 12 years of hard labour for the offense, which they subsequently admitted. Does this mean that the 68-year-old dictator has become, let’s say, more humanitarian, in old age? The linguistic features of the lack of knowledge design is unmarked information structure, e.g. new focus sentences (often with extensive focus domains), wh-questions, and words or collocations expressing wonder or a lack of knowledge (focus in bold, other features underlined): 28) Nationens politiske iagttagere sidder i disse timer og river hår ud af hovedbunden i hidtil forgæves forsøg på at uddrage en slags rationale af statsministerens determinerede nytårsudtalelse om regeringens påtænkte foranstaltninger mod den påståede voldsoptrapning i det danske samfund. Forundringen siden er ikke just blevet mindre, efter at den nyslåede justitsminister Frank Jensen har sat konkrete boller på knippelsuppen og bebudet diverse skærpelser af strafferammerne, herunder indespærring af mindreårige voldsmænd(drenge) i passende forbedringsanstalter. The nation’s political observers are in these hours tearing their hair out in up to now unsuccessful attempts to extract some kind of rationale from the Prime Minister’s determined New Year’s opinion on the government’s proposed measures against the alleged escalation of violence in the Danish society. The amazement is not just diminishing after the newly knighted Minister of Justice Frank Jensen has made concrete preparations for a workingover by the truncheon and announced an extensive tightening of penalties, including incarceration of minor assailants (boys) in appropriate improvement institutions. 29) Hvor er logikken i, at et ludfattigt og isoleret militærdiktatur i årtier har turdet trodse verdens eneste militære supermagt og dennes allierede og ikke for længst er gået samme vej som i sin tid Østtyskland og de øvrige betonkommunistiske lande i den gamle østblok? Where is the logic in the fact that a desperately poor and isolated military dictatorship for decades has dared to defy the world’s only military superpower and its allies by not going down the same road as East Germany and the other hard-liner Communist countries in the old Eastern Bloc? 30) Det netop overståede folketingsvalg har affødt en lind strøm af analyser. Alligevel er der stadig emner, som mangler at blive belyst. Relationen mellem parti og vælger er et af de kapitler, hvis sider stadig er blanke. En lettere Simon Borchmann 94 malerisk titel til et sådant kapitel kunne være ‘Fra engangsknald til ægteskaber’. The recent general election has produced a steady stream of analyses. However, there are still issues which remain to be elucidated. The relationship between party and voter is one of the chapters whose pages are still blank. A more picturesque title for such a chapter could be ‘From one night stand to marriage’. Whereas each of the three interpretive exigencies are distinctly marked by specific linguistic features, the linguistic features of factual exigencies are more heterogeneous; that is, the common features of the factual exigencies are limited to rhetorical strategies that can be realised by different linguistic features. These strategies are intensification and accentuation. The former is carried out by using figurative language, spectacular comparisons, intensifying adjectives and constructions (not only A, but B), adverbs of grade, impressive figures, reiterations, and expressions carrying a negative value or evoking emotions. The latter is carried out simply by claiming importance or notability (features underlined). 31) Men sikke en forskel et år og magt gør. Ved valget i torsdags fik Clegg og hans Liberaldemokrater ikke kun en blodtud af vælgerne, men noget nær en knockout. But what a difference a year and power make. In the election last Thursday Clegg and his Liberal Democrats got not only a bloody nose from the voters, but something close to a knockout. 32) men VKO’s ordførere og socialminister Benedikte Kiær (K) måtte gå fra bordet velvidende, at der manglede 80 millioner kroner til finansieringen af den højt profilerede ghettoplan. but VKO’s spokesmen and Minister of Social Affairs Benedikte Kiær (K) had to leave the table knowing that 80 million kroner needed to finance the highprofile ghetto plan were missing. 33) I dag har statsminister Lars Løkke Rasmussen (V) endnu en afgørende chance for at lykkes. Today, Prime Minister Lars Løkke Rasmussen (V) has yet another vital chance to succeed. 34) At Søvndal når til det slutresultat, er det synlige eksempel på, hvor meget har ændret sig - ikke bare politisk, men også organisatorisk under hans år som formand. The Perilous Life of a Linguistic Genre Convention 95 That Søvndal achieves this result is the visible example of how much SF has changed - not just politically but also organisationally during his years as leader of the party. 35) Lene Espersens dage som partileder for Det Konservative Folkeparti er talte, og måske var det en idé, hvis partiet inden valget finder en anden til at stå i spidsen for det hårdt pressede regeringsparti. Det er essensen af det interview, som dagbladet Politiken bragte i går med den tidligere konservative Frederiksberg-borgmester og næstformand Mads Lebech, hvor han som den hidtil største profil i partiet, peger på de konservatives partiledelse dvs. Lene Espersen som hovedårsagen til partiets aktuelle vælgerkrise. Lene Espersen’s days as leader of Det Konservative Folkeparti are numbered, and wouldn’t it be an idea, before the election, to find another person to head the hard-pressed government party. This is the essence of the interview which the newspaper Politiken published yesterday, with the former Conservative Mayor of Frederiksberg and Vice-president Mads Lebech, in which he as the hitherto biggest profile in the party points to the leadership of Det Konservative Folkeparti, that is, Lene Espersen, as the main reason for the party’s current crisis. 36) Men denne gang er forskellen bemærkelsesværdig; fyringen af Lene Espersen har intet med politisk uenighed at gøre. But this time the difference is remarkable; the dismissal of Lene Espersen has nothing to do with political differences. 37) Hvad der vil gøre valgkampen særlig interessant er, at den vil vise, om den karismatiske regeringschef mener det alvorligt, når han siger, at han vil gå radikalt til værks i forsøget på at reformere verdens næststørste økonomi. What will make the election particularly interesting is that it will show whether the charismatic head of government is serious when he says that he will take drastic steps in an attempt to reform the world’s second largest economy. To sum up the analysis, I claim that the move-structure of the prototypical news analysis, irrespective of the type, is characterised by the two following moves in a fixed order: “Presenting an interpretation of an event”, followed by “Establishing an exigency”. The opening move comprises two steps in a fixed order; the second move comprises three steps in a varying order. Each of the steps is more or less distinctly marked linguistically. Both of these moves and the steps must be explained by referring to an existing system of newspaper genres, taking into account a developmental history driven by a collection of interdependent factors. Simon Borchmann 96 The following analysis is restricted to the linguistic realisation of the first step in the opening move of the prototypical news analysis. 7 The linguistic realisation of framing as a rhetorical act In the analysis presented above, the phenomenon ‘framing’ is determined as a step in a move; that is, a rhetorical act that contributes to the fulfilment of a conventionalised communicative purpose. This determination departs from the established accounts of framing within cognitive science and within political science-oriented communication and media research, as well as cognitive linguistics. Within cognitive science, frames (scripts, schemata or Memory Organisation Packets) have been described as memory structures (Minsky 1975, Schank & Abelsohn 1976, Rumelhart 1980, Schank 1999) and framing as mental processes, e.g. Minsky’s transframing or uniframing (1975). Though my determination comprises the mental aspect, it departs from this description by determining framing as an act. By doing so, it follows the political science-oriented communication and media research into framing (Kinder & Sanders 1990, Entman 1993, Kuypers 2002), since these disciplines operate with a wider use of the term frame. Frames are both “internal structures of the mind” and “devices embedded in political discourse” (Kinder & Sanders 1990), and, in virtue of the latter, they are described as “communicable” (Pan & Kosicki 1993: 55-57). This feature forms the basis of the conception of framing as an act. An example of a definition that reflects this conception is Kuypers; according to Kuypers, framing is, “the process whereby communicators act to construct a particular point of view that encourages the facts of a given situation to be viewed (or ignored) in a particular manner, with some facts more noticeable than others” (Kuypers 2002). As it appears, the mental aspect is preserved: that the point of view makes something more noticeable than others is clearly due to a matching-process (Minsky 1975). But it is equally clear that the viewpoint is constructed with an intention of communication, and that the process involves more than one participant. This interactional conception is characteristic of the use of the term framing within communication and media research, but also within cognitive linguistics (Lakoff 2008, 2004). Entman’s definition of framing makes it clear which aspects of the phenomenon these approaches have been focusing on: “To frame is to select some aspects of a perceived reality and make them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/ or treatment recommendation for the item described” (1993: 52). As the verb promote indicates, the focus is on the strategic use, in the sense that the framing serves the interests of the one who has framed the perceived reality. In line with this, Entman describes The Perilous Life of a Linguistic Genre Convention 97 framing as a projection of power (2007), and Lakoff draws attention to the conservative’s skilful use of framing to promote their values (2004). The presented analysis departs from this conception by determining framing not as a strategic act, but as a rhetorical act in the sense of Miller’s social motive or Swales’ communicative purpose; that is, a rule-governed social act, serving a purpose that is mutually acknowledged by the parts in the discourse. The basis for this determination is partly the observations of regularities described in the analysis, and partly an observed regularity in the linguistic realisation of the framing act. As claimed above, the frame activated by the framing act consistently serves as an overall frame in the interpretation of the events described, and the overall frame is consistently activated by the same part of the text, namely the heading, (that is, in the very beginning). Furthermore, the framing performed by the news analysis heading has a distinct informational characteristic compared to other genres of news journalism within the Danish system of newspaper genres: it is informative, whereas the framing of the heading in news articles is trivial. The final observation that supports the determination of framing as a conventionalised act within the genre news analysis is the following linguistic regularity: the pieces of information that are particularly decisive for the activation of the overall frame are typically realised as the focus of the sentence, and, thus, presented as new information. This observation is based on the schema theoretically-based analysis of information structure I have presented elsewhere (Borchmann 2005, 2007, 2009, 2011). In order to enable a critical examination of this observation, I shall now provide a brief description it. The central idea is that the task of a sentence in indicative mood is to satisfy an information need. From a schema theoretical viewpoint an information need correspond to an undetermined variable (or, to invoke the topical metaphor, an unfilled slot). The task is carried out by identifying a variable in a frame and determining it with a value. Thus, the information structure of a sentence comprises two sub-functions: the identification of the variable, the variable, and the determination of the variable, the value. For example: 38) Banker skal give aktionærer mere magt Banks should give shareholders more power With this sentence, the variable is identified by “magt” (power), and the value is determined by “mere” (more). In addition to these schema-related functions, the unique event (or element of the event) described must be recognised. This task is executed by the function I, which, in this context, we shall refer to as the recognition. In the example above, it is “Banker” (banks) that contribute to the recognition. So we have three functions: the identification of the variable (the variable), the determination of the variable (the value), and the recognition of the event described (the recognition). The information Simon Borchmann 98 structural correlation is as follows: the value coincides with the pragmalinguistic focus of the sentence, and the unique event (or element of the unique event) coincides with the pragma-linguistic topic of the sentence. Focus is the most informative part of information among the information represented with a sentence by a sender. Focus has three characteristics. Firstly, there are relevant alternatives to the focus. How many and what status they have depends on the focus type (Borchmann 2007). Secondly, the difference between the focus value and the alternative focus values is what determines the interpretation of the uttered sentence; hence, by focussing on a part of the information, the sender indicates that the eventual relevant forward textual inferences are related to this part, and must take this part as a starting point (Togeby 1993, Borchmann 2009). Thirdly, the focus is within the scope of negation, i.e. if one inserts a negation in the sentence, or says no to the uttered sentence, the negation affects the focused information (Heltoft 2003, Borchmann 2005); for example (focus in bold): 39) Ny offentlighedslov begrænser åbenhed New act on Public Access limits transparency Apart from specific constructions, focus is not coded in Danish grammar. Nevertheless, there are inter-subjectively-valid grammatical rules for determining the focus of a sentence (Heltoft 2003). Firstly, unless the main verb of the sentence is informative, focus is realised in the focus domain (Heltoft 2003), i.e. the post nuclear field in Danish sentence schema of an unmarked fundament (sentences with subject or situational adverbials in the fundament (first position in Danish sentence schema))(ibid.). Secondly, focus is typically realised by certain parts of the sentence (or parts of these parts), namely, predicative complement, noun phrases in indefinite form, adjectives, adverbs of grade, adverbs of manner, numerals and informative verbs and verb phrases (Borchmann 2011). Topic indicates what the utterance is meant to be about. Neither topic is coded in Danish. However, as with focus, there are inter-subjectively-valid grammatical rules for determining the topic of a sentence. Thus, the topic is typically realised by the subject of the sentence (whether in the fundament or the subject position). Alternatively, it might be realised by situational adverbials in the fundament, or the last position of the post nuclear field. This alternative, however, applies primarily to sentences with formal and preliminary subjects, but may also occur if the subject is a noun without an inflectional ending (naked form). It is also characteristic that the topic is realised either by names or definite descriptions, or by generic descriptions that specify more or less clearly-defined, identifiable sets. This means that, all things being equal, the informational correlate to the topic is presumed to be accessible. Finally, it is characteristic that the topic, as accessible information, is outside the scope of negation (Heltoft 2003); for example (topic underscored): The Perilous Life of a Linguistic Genre Convention 99 40) Ny offentlighedslov begrænser åbenhed New act on Public Access limits transparency The variable has not yet been an information structural correlate. Admittedly, there is still some work to be done regarding this correlation. Hence, this short presentation will be limited to a specific example of the realisation: in my corpus, the variable is often realised by a noun in generic form in the post nuclear field in Danish sentence schema (the variable in italics); for example: 41) Ny offentlighedslov begrænser åbenhed New act on Public Access limits transparency The variable is not always linguistically explicit. In these cases, it can be abstracted from the value; for example: 42) Bjarne Riis tilstår Bjarne Riis confess (CRIME <in danish ‘tilstår’ is restricted to crime>) The focus value, “tilstår”(confess), determines the variable “the claim of the accused”. This variable can be abstracted from the value confess (CRIME). Using this information structural analysis on the corpora, a regularity is revealed. Thus, the bits of information that are particularly decisive for the activation of the overall frame are typically not realised as topic, but as a focus (or as a variable within the scope of the negation). Furthermore, the topic typically represents a trivial framing. 4 The following is an example of this (topic underscored): 43) Regeringen spiller højt spil med ghettoplan The government plays high play with ghetto plan ‘The government is playing high stake with ghetto plan’ 44) Saad Hariri kan kun fortsætte på Hizbollahs nåde Saad Hariri can only continue on Hezbollah’s grace ‘Saad Hariri can only continue by the grace of Hezbollah’ 45) Japansk nyvalg skal redde økonomien Japanese new election must save the economy ‘Japanese new election to save the economy’ 4 Interestingly, this is contrary to what is assumed by Pan & Kosicki in their framing analysis. The analysis is based on van Dijk’s semantic text description. As a consequence of this semantic foundation, Pan & Kosciki point to designators as the default location of framing, e.g. ‘the Iraqi dictator’ (Pan & Kosciki 1993). Hence, if we apply this to news analysis, we are instructed to look precisely at the least informative part of the text with regards to framing. Simon Borchmann 100 46) Dansk Folkeparti danser stammedans Dansk Folkeparti dances tribaldance 47) EU blev ikke mere folkeligt The EU became not more popular ‘The EU did not become more popular’ However, there are 11 exceptions (out of 43) to this regularity in the corpus. It is interesting that seven of them activate the game or war frame, which is typical of the process-oriented political journalism; for example (presupposition underlined, variable/ frame in italics): 50) Den uerklærede krig mod Lene Espersen The undeclared war against Lene Espersen 51) Hvad bliver Nordkoreas næste træk? What is North Korea’s next move? 52) Da Bondam gik i seng med fjenden When Bondam went in bed with the enemy ‘The day Bondam went to bed with the enemy’ 53) Løkkes syge kamp Løkke’s sick battle Hence, all the exceptions, except for four, can be explained by the assumption that the political journalists or experts consider these framings of political affairs to be established in the public sphere and, hence, assume them as shared knowledge. This means that, in turn, these framings may be regarded as borderline cases with respect to informativity. In other words, these news analyses do not supply the readers with an interpretation that they could not have made on their own. Consequently, we may question to what extend these texts serve their purpose as news analyses. The observed regularity indicates that the framing act performed with the news analysis heading is governed by social conventions. Furthermore, the regularities can be described as functional, rhetorically motivated by reference to the needs that news analyses must satisfy, and by reference to the division of labour within the Danish system of newspaper genres: by presenting the frame-activating part of information as focus, the author indicates that there are alternatives to the chosen frame, and that the actual framing can be denied and, thus, is in dispute. These pragma-linguistic features are entirely consistent with the purpose of news analysis, which is to offer of an interpretation; an interpretation that the author can support, but that he or she, in principle, and contrary to the journalistic genres of opinion, has no political The Perilous Life of a Linguistic Genre Convention 101 interest in communicating. Hence, the information structural characteristics support the determination of framing as a typified social action within the news analysis genre. 8 The status of linguistic genre conventions: a conclusion In this analysis, I have shown that there is a set of norms (in a very broad sense) for news analysis as a social action. At the overall, textual level, the texts in the corpus are characterised by a typical organisation. The organisation realises a strategy that can be analysed as a sequence of actions. The first two actions (steps) in the sequence are common to all texts in the corpus. These actions are characterised by specific linguistic features. In the previous section, I highlighted the linguistic features of the act framing the event. For this action, the corpus shows a relatively uniform use of information structure, and the deviations are relatively systematic. This use of information structure is distinct for the news analysis within the system of newspaper genres. On this basis, I argue that the observed regularities represent a linguistic genre convention. Furthermore, the uniform use of information structure is consistent with the purpose described in sociology, media research, journalism, and with the observed division of labour in the Danish system of newspaper genres. Hence, the convention is claimed to be functionally motivated. Now, the question of particular relevance to the outlined problem is as follows: what is the status of a convention of this kind? This question will finally be answered in terms of Miller’s hierarchy. When answering the question, I will distinguish between the hierarchy as an interpretive guideline for a single discourse (i.e. as an ontogenetic framework), and the hierarchy as an explanatory framework for the emergence and establishment of discourse as type (i.e. as phylogenetic framework). First the ontogenetic determination is presented, then the phylogenetic. 8.1 The interpretive dependency of the linguistic genre convention Up until this point, my analysis has been ‘top down’, so I shall now present an analysis that operates from the ‘bottom up’. At the lowest point of the hierarchy, experience is substance of the linguistic form. It is crucial that this substance is not the material world, or perceptions thereof, but substance drawn from our “acting-together”, as stated by Miller. The substance of the information structure formed is precisely substance in this interactional sense. Thus, it is a basic experience of communicative interaction that there is asymmetry between the sender and the receiver regarding their knowledge; that some information is more important than other information, and that the memory system of the receiver is capacity limited. This explains why some language codes’ informational statuses, e.g. the statuses accessible in memory Simon Borchmann 102 (definite form), are not accessible in memory (indefinite form). Although a number of more specific informational statuses are not coded in modern Danish, there are inter-subjectively-valid grammatical rules for determining them (Heltoft 2003). Thus, the described information structure indicators are part of the system of language. Indeed, the coding of these is relatively open, but they nevertheless represent an input to the hierarchy. In Miller’s terms, the substance of a form on a higher level is pre-formed at a lower level. At this point, there is no significant discrepancy between Miller’s hierarchy and (functional) linguistics. At the next level, language is the substance of the locutionary form. According to Miller, it is constitutive rules that tell us how to fuse substance and form to make meaning, and regulative rules that tell us how the fusion itself is to be interpreted within its context. Regarding the constitutive rules, what makes something recognised and accepted as a locutionary act is that the sender follows the rule of language. A key rule for the formation of a locutionary act is that it constitutes a discourse unit, and the criterion for this is that it comprises an illocutionary force indicator (Heltoft & Hansen 2011). In this connection, it should be noted that the news analysis heading is very often realised as a main clause, although it is a convention to use short forms as headings in a number of newspaper genres. The main clause in Danish enables the realisation of information structure indicators, and, thus, it can be seen as an indication that the informational status of the bits of information plays a role. Regarding the regulative rules, the hierarchical account contributes an important point: how the information structural indicators of the locution are interpreted is regulated by rules on the level of the speech act. A helsætning (main clause) realises an illocutionary frame and a focus domain. The illocutionary frame and focus domain can be used for topic and focus indication, but this option is not necessarily utilised; it depends on the class of speech act involved. In the linguistic descriptions of natural language information structure, the realisation of topic and focus has typically been associated with the sentence. However, not all sentences have a topic and a focus; topic and focus are primarily associated with assertive speech acts. Thus, directives very often do not have a topic and a focus. This also applies to requests for information; it is limited to the identification of a variable (so that the receiver can supply the value). Likewise, declaratives do not carry topic or focus; it is pointless to argue that ‘I’ realises the topic of the sentence ‘I hereby declare you man and wife’, nor is there a basis for determining a focus of this utterance. While expressives can be connected with the use of information structural indicators, they are not typically used to indicate topic or focus. The characteristic of an expressive is that it is sender-oriented; information is not organised for the sake of the receiver, but on the basis of the sender’s representation and in order to express what the sender feels and deems important. Thus, the fundament (first position in Danish sentence schema) of a sentence that serves as The Perilous Life of a Linguistic Genre Convention 103 an expressive is interpreted as a starting point (i.e. the first in the mind of the sender) rather than a topic. So the basis for interpreting the information structural indications as indications of topic and focus is that the act performed by the use of the sentence belongs to the class of assertive speech acts. Framing the event belongs to this class. At the next level, there are regulative rules for the interpretation of speech acts. Again, it seems to me that Miller’s hierarchy contributes an important point, this time regarding the meaning potential of linguistic indicators; the assertion-specific functions, topic and focus, can be given a more specific function when the assertive is a part of a genre-specific strategy. In the analysis, I have argued that the topic is used for recognition of an event through trivial framing, while the determination of the variable by focus is used for an informative framing of the event. With regard to the dependency of the interpretation, two aspects are important to highlight. Firstly, the strategy-specific interpretation is not contrary to the coded function; it is rather a more specific interpretation of the general function. Each time a higher level is reached (until the level of genre), the interpretation gets more specific. Secondly, the strategy-specific interpretation does not depend on the unique situation; it is secured by the genre. This also means that there are prerequisites for the strategy-specific interpretation. If topic and focus are to be assigned to the strategy-specific functions, the following must be satisfied: 1) the genre must be determined as news analysis through a genre marker or indicator, and, 2) the topic and focus indicators should be realised in the same place in which framing the action is usually realised, i.e. in the heading. If these conditions are fulfilled, the interpretation is legitimised. In this way, the linguistic genre convention forms the basis of a heuristic, which the reader may use in the understanding; focus indicates the overall frame of interpretation by default. The interpretation of the strategy is governed by rules on the level of genre. On this level, the genre-specific interpretation of the frame indicator is that it represents a relevant interpretation of a number of possible alternative interpretations. According to the hierarchy, the interpretation of the genre is determined by the rules at the level of life form; these rules correspond to the principles of the division of labour between genres in the life form, which we might call news production and consumption. The interpretation of this life form is governed by the rules on the level of culture, etc. However, since Miller points to the level of the genre as the level where motives become conventionalised social purposes, the genre seems to serve as the extreme specification of the interpretation. Hence, I do not elaborate on how the genre-specific interpretation depends on the rules at these levels. The hierarchy modelling the interpretation of information structure indicators of framing the event is as follows: Simon Borchmann 104 Level Meaning Nature interpretation Culture marketplace of many interpreters Life form news production and consumption Genre one relevant interpretation of a set of possible alternative interpretations of an actual event Strategy recognition of an event (by trivial framing), informative framing of an event Speech act topic, focus Locution information structure indication Language informational status Experience asymmetry of knowledge, saliency, memory capacity limits Tab. 5: A model of the contextual dependency of the enrichment of the content of a linguistic form 8.2 The stability of linguistic genre conventions On the basis on the hierarchy as a model of the genre’s dependency (phylogenetic), we can ask the following questions: how stable is this linguistic genre convention, and why? What ensures the stability and what threatens it? Initially, it should be noted that the convention, as a linguistic convention, depends on the stability of the coding that language makes available for the specific use. I assume that there are inter-subjectively-valid grammatical rules for determining topic and focus, and these are relatively stable. Thus, in modern Danish, declarative sentences have a fundament and a focus domain (Heltoft 2003). Having established this, we can now turn to the genre-specific use of the language that linguistic genre convention represents. As a social convention, the survival of the linguistic genre convention depends on repetition, i.e. framing of the event by use of the heading in news analyses must be informative, the information decisive of the framing must be presented as new information, and the topic realised must represent a trivial framing. This is a prerequisite for the reader to incorporate the convention into his or her genre-specific competence and use it as a heuristic. On the other hand, the repetition is dependent on the fact that the writers of news analyses have success when following the convention. The writer’s experience of success is determined by feedback from different sources. Thus, the repetition depends of positive feedback and/ or the absence of negative feedback. From a functional viewpoint, the feedback mechanisms can be identified as the basis for the stability of a functionally-motivated convention (Harder 2010). Conversely, dysfunctional regularities might be explained by a short circuit of the feedback mechanism. This account is based on a type of “evolutionary rationality”, which, roughly described, means that, all things being equal, we only preserve the conventions that we find appropriate. This The Perilous Life of a Linguistic Genre Convention 105 also means that, all else being equal, the existing conventions must be appropriate, or else they wouldn’t have been preserved. This roughly-formulated functional approach and the related assumptions of stability are supported by several observations. At lower levels, we can observe a functionally-motivated information structure regularity in the corpus. At higher levels, there is the correlation between the need for interpretation of news and the emergence of the genre, the observed division of labour between genres, verbally explicit rules in order to maintain the division of labour, the growth of the genre, and the positive feedback received by Politiken (according to Seidenfaden, Pedersen 2006). However, there are a number of factors that militate against the functional account and the related assumption of stability. Firstly, there is the observed deviation in the corpus. Regarding the presupposition of the war and game frame, the use of them is considered established in the public sphere because of the frequent use. Thus, the use of the war and the game frame represents a threat to the linguistic genre convention. Over the last 10 years, the use of these frames within political news journalism has been criticised (Kock 2002). However, it has not affected the frequency of their use (Pedersen et al 2006, Kock 2002). On the basis of the functional explanation, we can ask if the feedback mechanism is short-circuited. Attempting to answer this question, it becomes very clear that the feedback phenomenon is more complicated than I have described above. There is not only a number of different sources of feedback, but these sources are probably also prioritised differently, and the priorities are determined by interests that do not necessarily coincide with the common interests of the discourse community. For example, if the competition is tough and the financial circumstances doubtful, it would be natural to prioritise feedback from sources that are closely related to turnover, e.g. sales and reader response. Consequently, the sources that evaluate news analysis from a deliberative ideal, claiming to represent public interest, could be given low priority. Furthermore, the use of the war or game frame, or rhetoric, which are criticised by experts of democracy, might even generate positive feedback, e.g. a huge number of comments on a discussion thread. Thus, the feedback can be cloudy. Another factor that contributes to a nuanced understanding of the feedback phenomenon is that the negative effects only appear in the longer term. Among other things, this could lead to a cynicism and apathy towards politics, which, according to Kock, is a consequence of the use of the war and game frame. Therefore, it is justified to claim that feedback can be cloudy and lazy, and the prioritisation of sources of feedback can be determined by interests that are not necessarily in line with the common interests of the discourse community. The four remaining deviations cannot be explained by the use of the war and game frame. However, they can be construed as threats to the linguistic genre convention. Two of the analyses are judgemental. This subtype is con- Simon Borchmann 106 sidered by the profession to belong to opinion rather than news journalism. Within the opinion genre, it is perfectly legitimate to have certain interests associated with a particular interpretation; there is not the same demand for objectivity, balance and neutrality in the use of linguistic resources, including presupposition triggers. Thus, the growth in proportion of this subtype of analysis represents a threat to the genre convention. This tendency is observed in a master study (Pedersen et al 2006). One of the other two deviations derives from an analysis written by a PhD. student within political science. As far as experts have limited knowledge of the system of newspaper genres, they can be said to represent a threat to the linguistic genre convention, as well as other conventions of news analysis. In support of this, Pedersen et al observe that the proportion of texts deviating from the standards of news journalism by being subjective and biased, increases with the proportion of analyses written by experts (2006). As I have tried to show by elaborating on the deviations, the following applies: in the same way as the regularity can be explained with reference to factors in the broader context, so the deviations may be seen as determined by factors in the broader context. As described in the background and explanation of the emergence of the genre, there are several factors that contribute to the emergence of the genre. But it is not all of these factors whose impact serves the purpose mutually recognised by the members of the discourse community. 8.3 The linguistic status of the linguistic genre convention The linguistic status can initially be defined by comparing the status of the linguistic genre convention to the original object of text linguistics. While the rules described within text linguistics are, allegedly, linguistic rules and parts of the system of language (e.g. van Dijk 1972: 7, Halliday & Hasan 1976: 5), the linguistic genre conventions belong entirely to use. It is a specific use of a general coded function, which forms part of the system of language. The fusions of substance and form that are the substance of a form on the level of locution can thus be considered stable; it is the genre-specific use that is dynamic. The genre-specific use is nonetheless a regularity, and, so far, the linguistic genre convention can be said to be periodically stable. Its survival depends on its being motivated by a recognised, communicative purpose; this relation contributes to the likelihood that compliance with the convention generates positive feedback and counteracts negative feedback. However, as a genrespecific convention, it is under constant threat from the affects of a combination of interdependent factors that determine the genre’s emergence, change and decay: constitutional, political, governmental, ideological, educational, professional, ethical, technological, economical, commercial, and others. Feedback can be cloudy and lazy; sources of feedback, which aim to ensure its The Perilous Life of a Linguistic Genre Convention 107 survival, may be subject to prioritisation determined by the interests that are contrary to the common interests of the discourse society, and to deviate from conventions might even generate positive feedback in the short term. The linguistic genre convention, thus, lives a perilous life. The convention applies to an act. Following it contributes to the successful performance of the act. The act, however, is only a part of a strategy. And there is no independent illocutionary indicator of this act; it is based on the determination of the (overall) speech act realising the strategy. Furthermore, the disclosed strategy is only a stereotype. Thus, it is not the realisation of this strategy that constitutes the overall speech act news analysis; this act is constituted by the use of the illocutionary indicator news analysis or analysis within the framework of the system of newspaper genres. A news analysis is a news analysis even though the writer presupposes an informative frame with the heading. Indeed, it is a prerequisite for breaking the convention that the text is constituted as a news analysis. Thus, the linguistic genre convention is surely a regulative rule. The status of the linguistic genre conventions can be further constrained by comparing them with Levinson’s three layers of meaning. On the basis of on Grice’s concept of general conversational implicatures, Levinson designates a layer of meaning (utterance-type meaning) between conventional implicatures (coded meaning) and particularised conversational implicatures. The account of this layer is based on the characteristic of generalised, conversational implicatures; in other words, that they are based on Grice’s maxims. From the maxims, Levinson formulates a small number of guidelines for making inferences. The guidelines are called heuristics and are described as, “assumptions that can be used to enrich the encoded content in predictable ways” (1995: 96). The linguistic genre conventions resemble Levinson’s heuristics, since they allow the receiver to enrich the encoded content in predictable ways. Hence, they differ from particularised conversational implicatures. Likewise, they resemble the heuristics by being default rules. However, insofar as they apply only to specific genres, they are not general. The methodological point is that this precarious status as a temporary, constantly-threatened, regulative rule, in an area between general heuristics and particularised conversational implicatures, does not make the rule a less appropriate object of linguistic description. A modern, linguistically-based text description should comprise this object. Firstly, the descriptions are more informative than the structures hitherto described by text linguistics. Thus, they make it possible to distinguish between types and subtypes of texts, and they represent the differences that make a difference in language use. Secondly, as historical norms, they are a testimony to the development and change of language use. Thirdly, the descriptions contribute to language users’ awareness of the origin of standards, their understanding of how their own language use is related to standards, their ability to analyse a situation of conflict- Simon Borchmann 108 ing standards, and their awareness of a priority of standards within a given form of life. References Andersson-Ek, Per, Andréasson, Kenth and Edwardson, Åke. (2001) Skriv journalistisk - lær at skrive. Valby: Borgen. Askehave, Inger and Swales, John. (2001) Genre Identification and Communicative Purpose: A Problem and a Possible Solution in Applied Linguistics, vol. 22 no. 2, 195-212. Austin, John L. (1962) How to do things with words. Oxford: Clarendon. Bartlett, Frederick C. (1950 (1932)) Remembering. Cambridge. Bhatia, Vijay K. (1993) Analysing Genre: Language Use in Professional Settings. Harlow: Longman. Bazerman, Charles. (1994) Systems of Genres and the Enactment of Social Intentions in Aviva Freedman and Peter Medway (ed.) Genre and the New Rhetoric, 79-101. London: Taylor & Francis. Berge, Kjell Lars. (2010) Tekst og ytringhandlingsteori. Skandinaviske Sprogstudier , vol.1, 93-110, http: / / ojs.statsbiblioteket.dk. De Beaugrande, Robert and Dressler, Wolfgang. (1981) Introduction to text linguistics. New York: Longman. Bitzer, Lloyd. (1968) The Rhetorical Situation. Philosophy & Rhetoric, vol. 1, 1-14. Borchmann, Simon. (2005) Funktionel tekstteori og fiktivt fortællende tekster med refleksiv funktion. Frederiksberg: Roskilde Universitetsforlag. Borchmann, Simon. (2007) Fokus i dansk skriftsproglig sprogbrug in Peter Widell and Ulf Dalvad Berthelsen (eds.) 11. Møde om Udforskningen af Dansk Sprog, 64-75. Borchmann, Simon. (2009) Relevansstrukturen som den fundamentale funktion i sprogforståelsen - skitse til en teori in Inger Schoonderbeek and Peter Widell (eds.) 12. Møde om Udforskningen af Dansk Sprog, 79-100. Borchmann, Simon. (2011) Sproglige faktorer i tekstforståelsen. Fra det skrevne til det mente in Lars Heltoft (ed.) Sprogiagttagelse og tekstlæsning, 57-129. EMU Danmarks Undervisningsportal. http: / / fou.emu.dk/ offentlig_show_projekt.do? id=120087 Bransford, John D. and Johnson, Merieta K. (1972) Contextual prerequisites for understanding: Some investigators of comprehension and recall in Journal of Verbal Learning and Verbal Behavior, vol 11, no. 6, 717-726. Bro, Peter. (2009) Menneskelig kommunikation - Fra telegrafisternes til journalisternes tidsalder. Odense: Syddansk Universitetsforlag. Brown, Gillian and Yule, George. (1983) Discourse analysis. Cambridge: Cambridge University Press. Bulow-Møller, Anne Marie. (1989) The Textlinguistic Omnibus. København: Nyt Nordisk Forlag. Clark, Herbert H. and Havilland, Susan E. (1977) Comprehension and the givennew contract in Roy Freedle (ed.) Discourse Production and Comprehension. Hillsdale, 1-40. NJ: Lawrence Erlbaum. The Perilous Life of a Linguistic Genre Convention 109 Van Dijk, Teun. (1972) Some Aspects of Textgrammars. Haag: Mouton. Van Dijk, Teun. (1977) Text and context: explorations in the semantics and pragmatics of discourse. London: Longman. Van Dijk, Teun. (1995) Discourse semantics and ideology in Discourse & Society, vol. 6, 243-289. Entman, Robert. (1993) Framing: Toward a Clarification of a Fractured Paradigm. Journal of Communication, vol.43, 51-58. Entman, Robert. (2007) Projections of power: framing news, public opinion and U.S. foreign policy. Chicago: Chigago University Press. Gernsbacher, Morton Ann. (1996) The Structure-Building Framework: What It Is, What It Might Also Be, And Why in Bruce K. Britton and Arthur C. Graesser (eds.) Models of Understanding Text, 289-311. London: Lawrence Erlbaum. Grice, Paul. (1989) Studies in the Way of Words. Cambridge, Massachusetts, London, England: Harvard University Press. Grice, Paul. (2001 (1975)) Logic and Conversation in C. Henriksen (ed.) Can you reach the salt? Frederiksberg: Roskilde Universitetsforlag. Harder, Peter. (1996) Functional semantics: a theory of meaning, structure, and tense in English. Berlin: Mouton de Gryuter. Harder, Peter. (2010) Meaning in Mind and Society. Berlin/ New York: De Gruyter Mouton. Haberlandt, Karl and Bingham, Geoffrey. (1978) Verbs Contribute to the Coherence of Brief narratives: Reading Related and Unrelated Sentence Triples in Journal of Verbal Learning and Verbal Behavior, no. 17, 419-425. Halliday, Michael A.K and Hasan, Ruqaiya. (1976) Cohesion in English. London: Longman. Hansen, Erik and Heltoft, Lars . (2011) Grammatik over det danske sprog. Det Danske Sprogog Litteraturselskab, Syddansk Universitetsforlag. Heltoft, Lars. (2003) Iconic and categorial focus systems in Scandinavian in Jorunn Hetland and Valeria Molnár (eds.) Structures of Focus and Grammatical Relations, 43-37. Tübingen: Max Niemeyer Verlag. Hjarvard, Stig. (1999) Politik som mediemontage. Om mediernes forandring af den politiske kommunikation in Jørgen Goul Andersen et.al.: Den demokratiske udfordring, 30-51. København: Hans Reitzels Forlag. Hjarvard, Stig. (2009) Samfundets medialisering - En teori om mediernes forandring af samfund og kultur in Nordicom-Information, no. 31, 5-35. Keenan, Janice, Baillet, Susan D. and Brown, Polly (1984) The Effect of Causal Cohesion on Comprehension and Memory in Journal of Verbal Learning and Verbal Behavior, no. 23, 115-126. Kinder, Donald R. and Sanders, Lynn M. (1990). Mimicking political debate with survey questions: The case of white opinion on affirmative action for blacks in Social Cognition, no. 8, 73-103. Klin, Celia M., Guzmán, Alexandira E. and Levine, William H. (1999) Prevalence and Persistence of Predictive Inferences in Journal of Memory and Language, no.40, 593-604. Kock, Christian. (2002) Ud af hvalfiskens bug in Christian Kock (ed.) Forstå verden, 9-88. Frederiksberg: Samfundslitteratur. Simon Borchmann 110 Kuypers, Jim. (2002) Press Bias and Politics: How the Media Frame Controversial Issues.Westport Conn.: Praeger. Lakoff, George and Johnson, Mark. (1980) Metaphors we live by. Chicago: The University of Chicago Press. Lakoff, George. (2004) Don’t think of an elephant! : know your values and frame the debate. White River Junction, Vt.: Chelsea Green Pub. Co. Lakoff, George (2008) The political mind: why you can’t understand 21st-century politics with an 18th-century brain. New York: Viking. Levinson, Septhen C. (1983) Pragmatics. Cambridge: Cambridge University Press. Levinson, Stephen C. (1995) Three levels of meaning in F. Palmer (ed.) Grammar and Meaning: Essays in honour of Sir John Lyons, 90-115. Cambridge: Cambridge University Press. Loftager, Jørgen. (2007) Demokratisk journalistik - mellem politisk magt og afmagt in Journalistica, vol. 4, 33-54. Lund, Anker Brink. (2002) Den redigerende magt - nyhedsinstitutionens politiske indflydelse. Århus: Aarhus Universitetsforlag. Lützen, Peter Heller. (2004) Sproglig analyse og relevans. Dansklærerforeningen. McNair, Brian. (2000) Journalism and democracy: an evaluation of the political public sphere. London/ New York: Routledge. Meilby, Mogens. (2004) Nyhedsanalyse vinder langsomt frem. Århus: Update. www.update.dk. Meilby, Mogens. (2006) Journalistikkens grundtrin. Århus: Ajour. Miller, Carolyn. (1984) Genre as social action in Quarterly Journal of Speech, no. 70, 151-167. Miller, Carolyn. (2012) Genre emergence and evolution. Paper presented at the Rhetoric Society of America conference Re/ framing Identification. 25.05.12. Minsky, Marvin (1975) A Framework for Representing Knowledge in Patrick H. Winston (ed.) The Psychology of Computer Vision, 211-277. McGraw-Hill. Murray, John D., Klin, Celia M. & Myers, Jerome L. (1993) Forward Inferences in Narrative Text in Journal of Memory and Language, no. 32, 464-473. Møller, Henrik. (2002) Grundbog i tekstlingvistik. Dansklærerforeningen. Pan, Zhongdang and Kosicki, Gerald M. (1993) Framing Analysis: An Approach to News Discourse in Political Communication, no. 10, 55-75. Pedersen, Martine Lind, Michelsen, Johnni and Bertelsen, Martin Leonhard. (2006) Den politiske analyse - En undersøgelse af genren politisk analyse i Politiken, Berlingske Tidende, Jyllands-Posten og Information. Master dissertation. Roskilde: Roskilde Universitetscenter. Rumelhart, David E. (1980) Schemata: the building blocks of cognition in Rand J. Spiro (ed.) Theoretical Issues in Reading Comprehension, 33-58. Hillsdale, NJ: Lawrence Erlbaum. Sanford, Anthony & Garrod, Simon C. (1981) Understanding Written Language. New York: John Wiley & Sons. Sanford, Anthony J. (1990) On the nature of text-driven inference in David A. Balota, G. B. Flores d´Arcais & Keith Rayner (ed.) Comprehension processes in reading, 515-535. Hilsdale, New Jersey: Lawrence Earlbaum. Sanford, Anthony & Garrod, Simon C. (1994) Selective Processing in Text Understanding in Morton Ann Gernsbacher (ed.) Handbook of psycholinguistics, 699-717. San Diego: Academic Press. The Perilous Life of a Linguistic Genre Convention 111 Schank, Roger C. and Abelson,Robert P. (1977) Scripts, Plans, Goals and Understanding: an Inquiry into Human Knowledge Structures. Hillsdale, N.J.: John Wiley. Schank, Roger C. (1999) Dynamic memory revisited. Cambridge: Cambridge University Press. Schultz, Ida. (2007) Fra Partipresse over Omnibuspresse til Segmentpresse in Journalistica, no.5, 5-26. Searle, John. (2001 (1965)) What is a Speech Act in Carol Henrisken (ed.) Can you reach the salt? Frederiksberg. Roskilde Universitetsforlag. Svith, Flemming. (2011) Nyhedsjournalistik - bedre end sit rygte? et sociologisk og diskursanalytisk studium af forklarende journalistik i danske aviser gennem 40 år. Ph.d.-dissertation. Århus: Politica. Swales, John, M. (1990) Genre analysis: English in academic and research settings. Cambridge: Cambridge University Press. Søllinge, Jette Drachmann. (1992) Strukturændringer i den politiske journalistik mellem 1965 og 1991: en indholdsanalyse af 10 danske dagblades indenrigspolitiske dækning med særligt henblik på finanslovforligene 1965 og 1991. København: Politiske studier. Søllinge, Jette Drachmann. (1999) Historien om den politiske journalistik - Et drama i flere akter uden slutning in Erik Meier Carlsen, Peter Kjær og Ove K. Pedersen (eds.) Magt og fortælling , 76-95. Århus: Forlaget Ajour. Thomas, Jenny. (1995) Meaning in interaction. Longman. Thomsen, Niels and Søllinge, Jette D. (1991) De danske aviser 1634-1991, vol. 3 1918-1991. Odense: Odense Universitetsforlag. Togeby, Ole. (1979) Men hva’ med kongen. Om tekstlingvistisk analyse in John E. Andersen & Lars Heltoft (eds.) Nydanske Studier & Almen kommunikationsteori, 10-11, 26-103. København: Akademisk Forlag. Togeby, Ole. (1993) PRAXT - pragmatisk tekstteori. Århus. Aarhus Universitetsforlag. Ulbæk, Ib. (2005) Sproglig tekstanalyse - introduktion til pragmatisk tekstlingvistik. Århus. Academia. Vagle, Wenche, Svennevig, Jan and Sandvik, Margareth. (1993) Tekst og Kontekst. Oslo. Landslaget for norskundervisning. Cappeln Akademisk Forlag. Verstraete, Jean-Christophe. (Forthcoming) Formal criteria for subordination in the Germanic languages: in search of an explanation in Bert Cornillie, Dagmar Divjak, Piet Mertens and Pierre Swiggers (eds.) Structuring linguistic form (Orbis supplementa). Whitney, Paul, Bill Ritchie and Robert S. Crane. (1992) The effect of foregrounding on readers use of predictive inferences in Memory and Cognition, no. 20, 424-432. Wittgenstein, Ludwig. (1953) Philosophical Investigations. Oxford: Basil Blackwell. Zabel, Louise. (2007) Politiken - Fra newspaper til viewspaper. MA dissertation. Roskilde: Rudar. http: / / rudar.ruc.dk. Carmen Daniela Maier / Jan Engberg Tendencies of Multimodal Gradations in Academic Genres Network 1 Introduction Traditionally, the field of academic communication of knowledge (i.e., of results of research processes) has been dominated by research articles, and these have mainly had the form of written texts. However, in recent years a certain development in this field has not gone unnoticed, so that other forms of presenting results are emerging. This is probably mainly due to changes in technological possibilities, although the fact that young people grow up with much more access to information in visual format may also play a role. In our case, we want to focus upon two such genres with the purpose of communicating knowledge alternatively: academic visual essays and academic video essays. The emergence of such alternatives is interesting from a number of points of view, including the empirical assessment of the actual importance of different communicative modes in different fields and situational settings, whether different communicative modes are equally efficient in creating informational and attitudinal impact on readers of academic texts, or whether introducing new modes or genres with normative standards different from traditional ways of presenting research results may be seen as more or less including from the point of view of giving different writers access to relevant fora of knowledge exchange. A case in point concerning the last issue is the development of the research article itself: back in the 1970s, much serious academic publishing happened in the format of prints of typewritten pages, as more professional typographic printing presupposed the inclusion of other professionals and was only considered relevant for major texts. But with the emergence of text processing software like Word, more professionally looking documents became accessible for everyone. With this accessibility also to some extent the standards of professional printing were introduced into the realms of academic writing, so that the typed documents of the style frequent in the 1970s would now not be acceptable in any official context. What has happened is that the change in technological possibilities has led to a broadening of the competences necessary for proper academic writing. The question is whether this makes it more or less easy for someone being introduced into academia to learn research article writing. In the context of introducing more and alternative communicative modes, as is the case in the two alternative genres studied in this chap- 114 Carmen Daniela Maier / Jan Engberg ter, it may be argued that on the one hand destabilizing the almost sacred position of the (written) research article is positive as it diminishes the gatekeeping role of traditional researchers trained to write a specific type of texts. On the other hand, introducing genres with existing quality standards from fields like art and film may also mean that even harsher limitations are set up concerning what is a good piece of communication and who can evaluate it. However, we will not primarily focus on any of these interesting aspects. Instead, in light of the topic of this volume, we will concentrate upon aspects of genre connected with the described situation. More concretely, we will look at three genres in academic discourse and assess characteristics of the multimodal meaning-making processes in each of them. From a theoretical point of view, we will depart from the assumption that genres are reflections of influential factors of situation types (situational dimension of genres, Bhatia 2004: 22-26) and concentrate upon the influence of technological advances for the development of genres and the variety of subgenres in the scientific field. The main general question to be investigated on the basis of our study of a special case is: what happens, if the general situation of a genre field is stable, but some of the situation’s central elements related to representation and dissemination are changing? The specific assumption to be investigated is that the development of the various academic genres is mainly governed by three factors: characteristics of the object of study (studies of legal problems will require less development of the genre than multimodality studies, cf. below), the possibilities offered by the technological development, and the expectations of the readers. Thus, we assume that within and across different domains, the range of possibilities will be exploited differently. We want to show some examples and look at some of the semiotic processes present in some academic genres. Are other types of meaning creation possible in multimodal academic genres, or is it merely a question of being able to treat different objects? Another question that has also been posed by Swales (2004: 27) should focus on how we discuss the process of emerging new subgenres in the academic genre network: do we discuss oppositional dichotomies or gradations? For the time being, the process will be explored in this study as a process of technology-mediated transition because we examine subgenres that have in common a general communicative purpose that remains valid independently of specific realizations conditioned by different technological affordances. When discussing the consequences of the intersection of genre and technology upon the genre of lecture, Myers argues that, due to the appearance of Power Point, “instead of my speaking with the aid of some visual device, the text is speaking with my aid” (Myers 2000: 184). We argue that, in the case of our data which will be presented in the next part of this chapter, the technology-mediated transition that links the analysed subgenres is expected to have similar effects. Tendencies of Multimodal Gradations in Academic Genres Network 115 2 Data The preliminary data is collected from two peer-reviewed international journals that belong to print medium and to multimedia: Visual Communication (research articles and video essays) and Audiovisual Thinking (academic visual essays). According to its editors, Visual Communication covers “what’s going on in the field of visual communication“, and “critically investigates how the social world is constructed, represented and contested in visual discourse“ (Visual Communication homepage, 2010). Meanwhile, “Audiovisual Thinking is the world’s first journal of academic videos about audiovisuality, communication and media”(Audiovisual thinking homepage, 2010). The editors foreground the fact that “in this journal, researchers disseminate their research about audiovisuality and audiovisual culture through the medium of video” (Audiovisual thinking homepage, 2010). The academic background of researchers publishing in these two journals varies from linguistics and mathemathics to design and film. This is interesting because it means that when developing the new genre, researchers do not necessarily draw upon exactly the same genre knowledge (as academic genres are not identical across disciplines, Hyland 2009: 12-13). This has been shown in a number of earlier studies of academic genres (Bazerman 1988, Hyland 2009, Laurén 1993). It is thus relevant to distinguish between a more general knowledge of academic genres, on the one hand, and genre knowledge more specific to the discipline of the writer, on the other. We propose to examine multimodal texts from a genre network belonging to the field of presenting and discussing research results: the academic research article, the academic visual essay and the academic video essay. In all these academic subgenres, several semiotic modes can be co-deployed. Although these subgenres are created in two different media (print and video), they are disseminated through the same media, the Internet, and they have in common the same communicative purpose: to communicate knowledge about semiotic modes and their relations. The research article published in scientific journals is “the primary genre for dissemination of new scientific knowledge” (Berkenkotter and Huckin 1995: 27). Its purpose is to present processes of thought and research in such a way that a discourse for scientific fact-creation emerges (Hyland 2009: 68). It is therefore one of the most central genres of academic discourse, from the point of view of importance as well as quantity. The general characteristics of research journal articles have been widely described (e.g., Bazerman 1988, Myers 1990, Swales 2004), so we will not go into any details here. This choice is also motivated by the fact that we want to focus upon the field of multimodal meaning-making and not pursue a more general genre description in this chapter. We have chosen to work with research articles from only one journal, viz. the journal Visual Communication. The rationale behind this choice is to keep the range of factors influencing the individual text as low as 116 Carmen Daniela Maier / Jan Engberg possible by looking at texts which have all been subject to the same set of editorial guidelines. Furthermore, we wanted to study a type of traditional research articles, where the interests of the authors and the topic of the texts would make it likely that multimodality plays a role in the formulation of the texts. Where it is our experience that, e.g., research articles in the field of law often do without any use of visual aids, it is to be expected that articles on visual communication will feel an imminent need to apply other modes than the written one within the framework of a printed research article. Concerning visual essays and video essays we have chosen a similar approach, i.e., we examine texts from only the two journals mentioned above. When defining the two genres, we therefore adopt the descriptions of the genres in these two journals. In the guidelines for visual essays published on Visual Communication’s homepage, it is stated that “a visual essay can focus on any social or political aspect of visual communication, it can be a response to the visual work of others, a commentary on visual processes, ideas, and so on” (Visual Communication homepage, 2013). As far as form is concerned, it is indicated that a visual essay can be between 6 and 12 pages, and it can contain a combination of (static) image and writing or it can consist only of visuals. On the homepage of Audiovisual Thinking there are enumerated the main characteristics of a video essay. Apart from having to be (audio) visual, the essay has to: “disseminate new observations, knowledge, insights or theories, thereby adding to the existing body of knowledge”, “ acknowledge previous knowledge, insights or theories, and build upon this existing body of knowledge”, “ credit all sources and references, be they visual, written or oral”, “be self-critical and self-reflective”, and “form a coherent piece of media, that is possible to store as one computer file which can be easily shared” (Audiovisual thinking homepage, 2013). It should be added here that we do not employ genre as a category that could help us to identify texts that have the same communicative purpose in a certain context. We explore texts that have already been labeled as belonging to certain (sub)genres in order to identify their specific meaning-making structures. This, however, means that we do not know in advance whether the emerging genres are actually homogenous and stable categories. In accordance with the approach by Paltridge (1997: 53-56; 106f.) and by Borchmann (this volume) we see genres as basically dynamic categories of the prototypical kind (cf. also Engberg 2003). This means that instances of a genre may present to a higher or lower degree the traits present in the (mental and discursive) prototype of the genre, and that in this way new genres may emerge if ‘deviations’ from the prototype begin to stabilize and thus lay the foundations for a different prototype. In our case, we can see that the emerging genres exist as a (discursive) category, as indicated by the genres’ description made by the journal editors. But it is still possible and interesting to investigate whether individual instances of the genre are actually closer to the genre from which the emerging genre has developed. Due to lack of sufficient empirical evidence we will Tendencies of Multimodal Gradations in Academic Genres Network 117 not go deeper into this classification in this chapter. But as a first step we will try to classify the investigated texts accordingly. Borrowing two terms from the evolutionary taxonomy used in natural sciences, we classify them as transitional and intermediate forms. Transitional are those forms that do not present a substantial number of traits that the established genre does not possess as well. They are thus closer to the prototype of the established genre than to that of the emerging genre. Intermediate forms, on the other hand, are those forms that although retaining some of the characteristics of the established genre, actually do have a significant number of unique traits not connected to that genre. They are thus closer to the prototype of the emerging genre as described by the journal editors than to that of the established genre. In the conclusion of this chapter, we will return to the implications of this for the genres. With consequences for our analysis we consider the investigated subgenres as belonging to the genre network available to the members of the research community. Following Swales, we adopt the network concept because (h)aving a network frame allows us, by tracing intertextual links and other kinds of intertextualization, to place individual genres within a heuristically valuable wider context (Swales 2004: 23). Seen from a genre perspective, Bakhtin’s (1986: 91) concept of intertextuality as it has been developed in Fairclough’s typology of manifest and constitutive intertextualities is considered useful here in order to discuss the traces of various texts and their roles in the structure of these academic subgenres. Fairclough (1992: 117) considers that in the case of manifest intertextuality, “specific other texts are overtly drawn upon in a text”, while in the case of constitutive intertextuality or interdiscursivity, the echoes of the other integrated texts are more difficult to trace. As Fairclough (1992: 104) explains, the latter term is introduced “to underline that the focus is on discourse conventions rather than other texts”. In the context of this study, the technological advances are seen as one of the situational elements that should be closely linked to both intertextuality phenomena that can take complex forms across media and semiotic modes. When analyzing aspects of intertextuality of advertising, both Cook (2001) and Kuppens (2009) find that there are two types of generic intertextuality: intrageneric, “containing the voice of another example of the same genre” (Cook 2001: 193), and intergeneric, “containing the voice of another genre” (Cook 2001: 194) and evoking knowledge of that other genre. All of the mentioned types of intertextuality will be of relevance to our analysis below. 118 Carmen Daniela Maier / Jan Engberg 3 Theoretical issues At the present stage of this exploratory study, we consider that the identification and the articulation of a common set of analytical categories of inquiry should be based on the following theoretical assumptions. 3.1 Genre theory Our work on genre originates from the line of genre research seeing genres as centered around conventionalized patterns of linguistic means (in a broad sense). The conventionalization is rooted in the fact that members of the same discourse community (here: researchers from the same field) perform text production tasks (like presenting research results for academic discussion) under similar situational conditions. Thus, we see a relevant distinction between actual, realized texts, on the one hand, and virtual genre patterns, on the other. This line of research is founded in works by Swales, Bhatia and Paltridge, but also of researchers from outside the Anglo-Saxon world of textual research. As a tendency, work on genres in this approach is rooted in the study of texts dominated by writing and thus by a linear organization of the information offered. Among other things, this has lead to the concept of move structure to be very central in works from the tradition of especially Swales (1990, 2004) and Bhatia (1993, 2008). The move structure is conceptualized as the strategically motivated sequence of textual chunks (‘moves’) in which the text is organized. Each move is seen as giving a different contribution to the realization of the overall communicative purpose of the text. And as communicative purpose is seen as one of the most influential factors in the situation leading to the emergence of a genre, the move structure is likewise traditionally seen as a very central part of the virtual pattern lying behind the genre. Already when used on written texts with their focus upon linearity, the concept of move structure is somewhat problematic, as it is well possible to have functions contributing to the overall function of a text and a genre which are distributed across the text and may not be ascribed to a specific section of the text. But this problem is aggravated as a consequence of leaving the linear written organization of printed text and entering the more non-linear organization of texts presented in other media like the video. Even though we expect to be able to preserve our basic assumptions about the nature of genres, a revised set of descriptive elements will be necessary in order to describe and compare the genres to be investigated here. Taking this need into consideration, we have found systemic functional linguistics’ (SFL) view on genre analysis to be a relevant source of inspiration. According to Halliday, language is organized around three main metafunctions that are “generated simultaneously and mapped onto one another” (1978: 112): the ideational one which is “the content function of language”, Tendencies of Multimodal Gradations in Academic Genres Network 119 the interpersonal one which is “the participatory function of language” and the textual one which focuses on text organization, enabling language to become “operational in a context of situation” (Halliday 1994: 27). As these metafunctions are associated with different types of meaning-making structures, they are also relevant resources for generalising across genres. Martin and Rose (2008: 259) explain that in the SFL’ view, genre coordinates “a complex interplay of complementary kinds of meaning (ideational, interpersonal and textual)”. Consequently, they define genre as “a configuration of meanings, realized through language and attendant modalities of communication” (Martin and Rose 2008: 20). This approach has a special relevance for the present study because it also acknowledges the role of other semiotic modalities than language in generic configuration. Furthermore, it is relevant because the focus upon genres as configurations of meanings makes it apt for comparing genres that are closely related concerning meaning but show differences in modes applied when creating the meaning. And this is very much the case with our three genres, which all convey academic knowledge, be it in formally different ways. 3.2 Multimodality issues In their approach to the above-mentioned metafunctions, Kress and Van Leeuwen emphasize that “any semiotic mode has to be able to represent aspects of the world”, “to project the relations between the producer of a (complex) sign, and the receiver/ reproducer of that sign”, and “to form texts, complexes of signs which cohere both internally with each other and externally with the context in and for which they were produced” (Kress and Van Leeuwen 2006: 42-43). Consequently, each metafunction provides a tool with which to approach genres that include other semiotic modes than language. Not only that contemporary genre networks continuously expand, but traditional written and spoken genres also include more semiotic combinations than before. Due to the appearance of several semiotic modes (language, images, music and sound) in the structure of old and new genres, the social semiotic approach of multimodality (Kress & Van Leeuwen 2001 and Kress 2010) that stems from SFL has become relevant in the exploration of these increasingly complex genres. When defining multimodality, Kress and Van Leeuwen explain that it is: The use of several semiotic modes in the design of a semiotic product or event, together with the particular way in which these modes are combined (Kress and Van Leeuwen 2001: 20). The concept of multimodality offers the possibility to address the semiotic modes’ interaction and combination within genres in order to identify the role of the semiotic interplay in creating and multiplying meaning at various levels of the generic structure. It might be necessary to add here that, similar to Kress, we use: 120 Carmen Daniela Maier / Jan Engberg The term “mode” for the culturally and socially produced resources for representation and “medium” as the culturally produced means for distribution of these representations-as-meanings, that is, as messages (Kress 2005: 8). As the semiotic modes enter continuously in dynamic relations with each other, “each mode is partial in relation to the whole of meaning” (Kress and Jewitt 2000: 3). O’Halloran elaborates on this issue, stressing that: It is not only the culmination of choices made across semiotic resources in their interaction with other resources that makes meaning, but also the temporal and spatial unfolding of those choices (O’Halloran 2004: 109). The transitions from static to dynamic multimodal texts, from one medium of production to another multiply the research challenges. Therefore, in order to articulate an analytical framework for the subgenres’ description and interpretation, the concepts of remediation (Bolter and Grusin 2002) and resemiotization (Iedema 2003) are also to be employed. In the glossary of their seminal study of remediation, Bolter and Grusin highlight that remediation refers to “the formal logic by which new media refashion prior media forms” (Bolter and Grusin 2002: 273). In the light of this definition, it is obvious that through the exploration of remediation processes, the understanding of genres as mediated constructs and the consequences of a genre’s transition from one medium to another can be relevantly addressed. The other concept, resemiotization, (i)s about how meaning making shifts from context to context, from practice to practice, or from one stage of a practice to the next (Iedema 2003: 41). Consequently, the manifested shifts of meaning making in multimodal genre transitions can also be explained through the discussion of resemiotizing processes. O’Halloran explicitly links these two processes with the multimodal understanding of genres: The semiotic landscape, the aggregate of semiotic choices materializing as objects and events which are recontextualized and resemioticized across different place and time scales, result in evolving but discernable patterns of semiosis called genres (O’Halloran 2009: 99). Bateman (2008 and 2009) addresses genres from a similar point of view, highlighting the genre continuum that appears due to the various multimodal structures made possible by the dynamic interplay of semiotic modes. In order to explore the tendencies in the gradations of contemporary academic genres from this multimodal perspective, we adopt Bateman’s conceptual framework as it allows us to explain genres in terms of meaning making flows (see section 4 in this chapter) which organize the generic configurations. Tendencies of Multimodal Gradations in Academic Genres Network 121 3.3 Knowledge communication approaches As it can be seen so far, the fact that we attempt to explore complex and dynamic genres implies the necessity to find theoretical support across broader research domains. Therefore, we consider that the results of the multimodal genre analyses can be better understood in the light of knowledge communication approaches focused on representation and communicative construction of knowledge. The idea, which sits well with the conceptualization of genres as virtual patterns underlying the production of actual texts (cf. section 3.1 above as well as in this volume Togeby and Borchmann), is to see the (conventionalized) configuration of meanings constituting the genre as residing in the minds of traditional users of the genre (Berkenkotter and Huckin 1995: 24). In our case, this means that academics writing in their field share some knowledge about how specific academic texts are normally written, a knowledge which is also an important factor in identifying them as members of the group of academics (Berkenkotter and Huckin 1995: 21-24; Hyland 2009: 54). The interesting aspect in the cases studied in this paper is that two of the investigated genres are fairly new and that they may be seen as derived from the more traditional academic genre of the research article. It is thus interesting to investigate overlaps between the configurations of meanings as well as the stability of conventions in the new genres. In the following we will first introduce the idea of knowledge communication as a framework of the questions posed in this paper. Subsequently, we shall give a brief overview over aspects of academic texts with relevance from a knowledge communication perspective. And finally, we will look into the aspect of genres seen as part of the knowledge of members of a discipline. We will start out by a statement to the nature of knowledge communication: Knowledge communication is strategic communication. As ‘strategic’ it is deliberately goal-oriented, the goal being the mediation of understanding across knowledge asymmetries. As ‘communication’ it is participative (interactive) and the communicative ‘positions’ converge on the (co-)construction of (specialized) knowledge. (Kastberg 2007: 8) For our purposes, the statement that the goal of knowledge communication is to mediate understanding across knowledge asymmetries is central: the study of knowledge communication is the study of how specialised knowledge generated by researchers is transferred to receivers so that they can understand and possess this knowledge, too. This framework of study makes two perspectives relevant, of which we will mainly treat one here, but also touch upon the other in our conclusion. The perspective to be briefly touched upon concerns the efficiency of a specific combination of modes and media in order to mediate understanding across knowledge asymmetries: are the traditional and the new ways of communicating research results equally efficient, or are some ways more efficient in some contexts than in others? 122 Carmen Daniela Maier / Jan Engberg The perspective to be mainly treated, on the other hand, concerns genre knowledge (understood as knowledge about the conventions behind a genre) as part of the specialised knowledge of researchers. One could say that we are here looking at disciplinary meta-knowledge underlying the communication of generated content-level knowledge of the discipline. In the framework of knowledge communication with its assumption of mediation of understanding as a participative process cited above, not only the content-level knowledge to be communicated is focused. The specialised disciplinary knowledge about typified forms of communication has relevance, too. For it is important for the process of co-constructing specialized knowledge that the participants also share knowledge about the traditions and concurring meanings of the communicative process. Genre knowledge may be seen as divided into a number of subtypes. For example, Beaufort (1999) distinguishes between four types of genre knowledge: Three types at the same level (subject matter knowledge, rhetorical knowledge and writing process knowledge) and one framing type (discourse community knowledge, a concept that Swales (2004) equates with Bordieu’s habitus (1990). Beaufort argues that these knowledge types or dimensions are “symbiotically related and nested one within another like a set of Russian dolls” (Beaufort, 1999: 177). Our focus here will be on the rhetorical knowledge, but taking into account its relations to the other subtypes. Knowledge in general is divided into explicit knowledge and tacit knowledge, and genre knowledge is often seen as dominantly tacit knowledge in the perspective of this divide (e.g., Beaufort 1999, Smart 2006). In Giddens’ (1986: 49) terms, explained by Nonaka and Royama when analyzing organizational knowledge creation the distinction is formulated as one between discoursive consciousness and practical consciousness: While the discursive consciousness gives us our rationalizations for actions and refers to more conscious and therefore more explicitly theoretical knowing, practical consciousness refers to the level of our lives that we do not really think about or theorize. In that sense, we can say that tacit knowledge is produced by our practical consciousness and explicit knowledge is produced by our discursive consciousness (Nonaka and Royama, 2003: 4). This distinction can be linked with Swales’ (1990: 84-92) discussion of the interaction between two types of knowledge which contribute to the production and recognition of genres, namely the mainly explicit prior knowledge of the world (content schemata) and the often tacit knowledge of prior texts (formal schemata). He suggests that it is not quite possible to understand their interaction or to separate them because genres „coalesce what is sayable with when and how it is sayable“(Swales 1990: 88). The fact that formal schemata are often tacit, but useful has been the driver behind numerous universities introducing courses in academic writing and numerous authors writing books on how to write, but also read and understand, e.g., research articles. It is an Tendencies of Multimodal Gradations in Academic Genres Network 123 example of attempts to make tacit knowledge explicit, with all the problems connected to such ventures. The main problem lies in the fact that it is difficult or maybe even not possible to make tacit knowledge fully explicit: While tacit knowledge can be possessed by itself, explicit knowledge must rely on being tacitly understood and applied. Hence all knowledge is either tacit or rooted in tacit knowledge a wholly explicit knowledge is unthinkable (Polanyi 1969: 144). But still it is probably the case that the more stabilised tacit knowledge is (in our case knowledge about conventional formal schemata), the better it will be possible to make parts of this knowledge explicit and thus available for instruction and the discursive consciousness. The categories mentioned so far have specific relevance for this study because in the present case the discourse community is in the process of acquiring or establishing new rhetorical knowledge about representation of given subject-matter knowledge due to the technological advancement. Knowledge about the object of study of the discourse community of researchers in the field of visual communication, viz. the interplay of semiotic modes, is communicated in specific patterns of structure in the emerging multimodal subgenres. An interesting aspect to be discussed below is the potential problem that arises from the specifics of the visual mode. This is connected to the distinction between tacit and explicit knowledge. The visual mode has as one of its main semiotic characteristics that representations generally have a greater interpretation potential than representations in the written mode. Thus, employing the visual mode as the dominant mode in an academic text to present research results of an abstract nature creates a challenge, as academic discourse is normally characterized by a high degree of precision which the larger interpretation potential will counteract. The interesting thing here is the emerging balance between ‘academic’ (with its focus upon precise expert knowledge) and ‘essay’ (with its focus upon a less precise and also aesthetically challenging presentation of knowledge). Returning to the tacit/ explicit knowledge categorization, we argue that the transition from research article to visual essay and video essay will probably imply either a shift of the balance towards the ‘essay’-aspect (and thus more importance on the production side to the tacit knowledge of creating aesthetic multimodal texts) or the emergence of (at least in initial phases) tacitly known procedures of achieving precision through conventionalisation. In both cases the consequence is a rise in importance of tacit knowledge elements for communication. A knowledge communication perspective can help us monitor the investigated genres concerning the gains and losses from going from articles to essays and from static images to moving ones from the point of view of the academic feature. Before we leave the knowledge communication perspective, however, we would like to take a brief look at the field of academic discourse, to which the 124 Carmen Daniela Maier / Jan Engberg genres investigated here belong, in order to highlight some of the characteristics of this discourse with relevance for our subsequent study. Hyland gives the following brief description of aspects of academic discourse: Academic discourse refers to the ways of thinking and using language which exist in the academy. Its significance, in large part, lies in the fact that complex social activities like educating students, demonstrating learning, disseminating ideas and constructing knowledge, rely on language to accomplish. Textbooks, essays, conference presentations, dissertations, lectures and research articles are central to the academic enterprise and are the very stuff of education and knowledge creation. But academic discourse does more than enable universities to get on with the business of teaching and research. It simultaneously constructs the social roles and relationships which create academics and students and which sustain the universities, the disciplines, and the creation of knowledge itself. (Hyland 2009: 1) What makes discourse academic is according to Hyland that it takes place in the settings of academia and that it contributes to a number of complex social activities typical of academia like educating students and disseminating ideas (cf. ‘the business of teaching and research’ in the citation above). At the same time academic discourse has a number of social implications concerning the social roles in academic settings, among other things because ‘the settings of academia’ also means ‘within the academic community’ (Hyland 2009: 48). Following the suggested distinction between social activities and social implications, we will in this chapter focus upon the first-mentioned social activities connected to ‘the business of teaching and research’ (hence our focus upon genres as complex interplay of meanings and our focus upon ideational meaning). In connection with the social implications, it should be highlighted that the possible roles of the researchers that emerge in these genres represent another challenging topic. Maier and Engberg (2013) have found that “the proliferation of new academic generic forms affects the range of available roles that can be taken by researchers who “write” and “read” the new multimodal academic texts”. Along these lines, it is a central characteristic of academic discourse in research articles to cite broadly the previous works by other researchers on which the presented work is based. In text linguistic terms this is a kind of intergeneric intertextuality (cf. section 2 above), as what is cited is normally other research articles or closely related genres. Apart from the ideational meaning created through this operation concerning the basis of the utterances made in the research article, the author also positions himself as part of the major context of academia and to some extent also assumes a role within more specific lines of the research in a field (by citing some rather than other authors). Due to the central and double role of this characteristic in academic discourse and especially in the writing of research articles, we will in our empirical analysis focus specifically upon how this is done in the new and emerging genres. Tendencies of Multimodal Gradations in Academic Genres Network 125 4 Interdisciplinary analytical framework Inevitably, in order to address the challenge of exploring these subgenres, it is necessary to combine these theoretical perspectives for devising a model of analysis with several layers of description (Bateman 2009: 60) that can highlight both their similarities and differences. Acknowledging the fact that the academic discourse in these subgenres is influenced by structures that are closely linked to the subgenres’ common communicative purpose of representing specific types of knowledge implies that it is necessary to explore what structures can be found and how these structures are influenced by the choice of semiotic modes. One way in which the subgenres’ similarities and differences can be examined is by finding common analytical categories that can be used for the respective subgenres across semiotic modes and media. These categories are supposed to address the meaning-making structural units of the subgenres in order to interpret their specific roles and their relations with the common communicative purpose. Consequently, it is at this analytical stage that the metafunctional approach to genre is to be employed. As already mentioned, according to the SFL perspective, genre is a configuration of meaning, “a resource for co-ordinating communication across modalities in multimodal texts” (Martin and Rose 2006: 234). Certainly, defining meaning-making structural units, identifying their characteristics and formalizing the respective characteristics in each subgenre are necessary stages in the process of genre analysis, but they cannot underpin the specificity of these subgenres alone. Therefore, before focusing on the meaning-making structural units, it may be necessary to direct our attention to the ways in which the subgenres employ specific strategies in combining semiotic modes with communicative purposes. Bateman (2008 and 2009) finds three distinctive resources: the text-flow, the page-flow and the imageflow. In the text-flow, “the visual line of the developing text provides a basic one-dimensional organizational scheme”, and “the spatial nature of the page is not made to carry significant meanings in its own right” (Bateman 2009: 61). This does not exclude the incorporation of visual elements like tables and diagrams. The page-flow can incorporate the text-flow and “adds the possibility of spatially-signaled rhetorical relations supporting the communicative intentions” (Bateman 2009: 61). The image-flow “is used to organize sequences of graphical elements rather than the text organized by text-flow”, and these sequences have “a very specific range of additional meanings over and above those in the contributing images” (Bateman 2009: 61). We adopt these concepts because we assume that each of the three genres is dominated by one of these flows. If making sense of the subgenres’ specific strategies in combining semiotic modes with communicative purposes requires the employment of these three types of flows then, afterwards, at the level of each one of them it is necessary to address the interplay of the semiot- 126 Carmen Daniela Maier / Jan Engberg ic modes. This interplay is to be explored in terms of the three metafunctions. Our exploration draws on a considerable body of work in multimodality, that following Halliday’s metafunctional framework, addresses and identifies the meaning-making multimodal relations at the level of each metafunction. Among the researchers who focus on multimodal relations in various genres from advertisments to news websites are Lim 2004, Macken-Horarik 2004, Martinec and Salway 2005, Van Leeuwen 2005, Kong 2006, Unsworth 2006, Knox 2007 and Royce 2007. In this chapter, we focus on the multimodal relations established at the ideational level, and with some exceptions, our exploration is based on the categorization of multimodal interactions suggested by Martinec and Salway 2005, Van Leeuwen 2005, and Unsworth 2006. The main types of relations employed in this multimodal analysis are concurrence, complementarity and connection. As for concurrence, two main subtypes have been identified: elaboration through specification, explanation and similarity, and elaboration through overview and detail. Complementarity is realized through extension that can take two forms: augmentation and contrast. Finally, the ideational connection is established through circumstantial enhancement: temporal, spatial and causal. The types of each relation will be identified in the course of the following analysis. Concurrence Elaboration (between text and image) Specification Explanation Similarity Elaboration (between images) Overview Detail Complementarity Extension Augmentation Contrast Connection Enhancement Temporal Spatial Causal Tab. 1: Categorization of multimodal interactions based on Martinec and Salway 2005, Van Leeuwen 2005, and Unsworth 2006 5 Model of analysis Taking the approach suggested above and employing the flow concepts as starting points, we examine here the specific meaning-making multimodal interactions which characterize each subgenre and its transitional and intermediate forms. We also attempt to approach academic knowledge communication in these genres by examining their intertextual relations in terms of the intrageneric and intergeneric references existing in each of them. In this incipient phase of our research, we have decided to focus only on the ideational aspects instead of making a full-fledged multimodal analysis dealing with all Tendencies of Multimodal Gradations in Academic Genres Network 127 structural elements. Our choice is motivated by the fact that these genres are still genres “on the move”. 5.1 The text flow in the academic research article To explain the position of research articles in relation with the other two related genres, the text-flow concept has been chosen as it refers to “the sequentiality of text” (Bateman 2008: 175). The text-flow based model has been chosen because text with its “supporting typography” is the research article’s dominant feature, although this genre does include “transitional objects” that “make contact both with linguistic concerns (…) and with spatial concerns” (Bateman 2008: 100). Apart from tables and diagrams, diverse types of images in color or black and white accompany the articles’ text entering especially in relations of ideational concurrence with it. O’Halloran’s (2008) and White’s (2010) articles, Systemic functionalmultimodal discourse analysis (SF_MDA): constructing ideational meaning using language and visual imagery and Grabbing attention: the importance of modal density in advertising, display some of the traditional relations of ideational concurrence that can be established between texts and images in such a genre. For example, a series of relations of elaboration appear in both articles. Each of these elaborations predominantly communicates knowledge about the data and methodology employed by the author. When a screen capture of the software used in research is displayed in O’Halloran’s article, there is a relation of elaboration through explanation as the text clarifies the image. Fig. 1: Example of a page from O’Halloran’s research article 128 Carmen Daniela Maier / Jan Engberg Clusters of images displaying the analyzed ads enter also in several relations of elaboration with the texts. First, mainly relations of elaboration through specification are established with their captions as each image illustrates the text. For example, the visualization of the analyzed ad is labeled as: “Aids advertisement, Cleo magazine (Singapore), December 1998: 139”. Second, as the ads are manipulated through changes of color, size and/ or superimposed lines, the predominant relations become relations of elaboration through explanation because the main text clarifies them. For example, the previously mentioned ad, reduced in size, is accompanied by the caption “Generic elements” that is explained in the main text. Although not manipulated otherwise than in terms of size, several excerpts of journals’ homepages visualized in Knox’s article (2007), Visual-verbal communication on online newspaper home pages, enter in relations of elaboration through explanation because the accompanying texts reach another level of abstraction than simple descriptive labels. For example, an image composed of three separated excerpts from Bangok Post is accompanied by the explanation “Visual-verbal classification” (Knox 2007: 32) in the caption and further explained in the main text. In White’s article, similar situations as in O’Halloran’s article occur with the posters which represent his study’s data. However, both in White’s article, and in Martinec and Salway’s article (2005), A system of text-image relations in new (and old) media, it is the analytical results that are visualized in tables condensing the explanations from the texts. Apart from that, the drawings that are included in Martinec and Salway’s article enter in relations of elaboration through specification with the accompanying texts. For example, a drawing is accompanied by the following caption: “Example of complementary image-text relationship. Drawing after an advertisement in the Guardian, December 2004” (Marinec and Salway 2005: 345). Tendencies of Multimodal Gradations in Academic Genres Network 129 Fig. 2: Example of a page from Martinec and Salway’s research article It can be thus assumed that domain specific knowledge can be clearly communicated in research articles through these image-text relations to the members of the academic community who belong to the same research domain and who possess the necessary genre specific knowledge. There is a balanced link between tacit and explicit knowledge. When writing a research article, the researcher respects the genre’s rules established both by journal’s editors and through tradition, assuming that the readers are familiar with them, and introduces new domain specific knowledge in an explicit way. Before addressing the page-flow concept in order to discuss visual essays, it is necessary to present some aspects of transitional and intermediate forms of multimodal texts that have been labeled as visual essays while maintaining the text-flow of research articles as dominant structural principle. Goodrum and Hunt’s visual essay, Framing rural fashion: observations from Badminton Horse Trials (2011), retains the text-flow in its structure. The middle part of the visual essay seems in fact to be a conventional research article including the expected structural elements (abstract, key words, data information, theoretical framework, analysis, notes and references) without any kind of images interspersing the text. This textual part is preceded and followed by more than 20 one-page or double-page images in color. None of these images is accompanied by texts, and all of them illustrate the internationally renowned sporting event and its participants through close up or long shots taken from various view angles. In this transitional form, multimodal knowledge communication takes place through relations of elaboration through specification not at the level of the interplay of single images and text parts, but at the level of the interplay of the whole text with the sequence of images. However, what supports our understanding of this visual essay as a transitional 130 Carmen Daniela Maier / Jan Engberg form are the choice of ideational enhancement relations that are established inside the sequence of images included in the essay. These relations are spatial or temporal as the images qualify each other through circumstantial information. Such relations between images do not usually characterize the meaning-making strategies of research articles. In Jaworski’s visual essay (2011), Tracing Place, Locating Self: Embodiment and Remediation In/ Of tourist Spaces, texts and images also appear on separated pages. Three types of texts are intercalated with a wide range of color photos of various sizes and view angles, from various sources and displaying (people in) various tourist destinations: a front page with the usual information about author, abstract and key words, two pages in black and white with quotations from research work with a similar topic and two last pages that also seem to be taken from a conventional research article. The predominant type of multimodal relation through which knowledge is communicated is represented by complementarity, text and images extending each other’s meanings. The relations of extension through complementarity are not established at the level of a single image and part(s) of text, but at the level of all images and the whole text. Although still characterized by a text-flow structure and by spatially separated texts and images, this intermediate form is closer to the visual essays dominated by page-flow because we can witness a shift towards according meaning-making importance to the spatial placement of images in a page. In both visual essays, only knowledge about data is communicated through the sequences of images and through the interplay of these sequences with the texts. Such images sequences can be also characterized as imageflows although the temporal sequencing is not the predominant one; Bateman observes that image-flow elements can appear “alongside, and in combination with, text-flow elements” (Bateman 2008: 175). 5.2 The page flow in the academic visual essay In order to investigate how and what knowledge is communicated through the interplay of images and texts in the academic visual essay, we propose to use the page-flow concept as it refers to “spatial contiguity” (Bateman 2008: 175). Coming from the context of education, Hughes and Tolley approach the topic of visual essays when presenting them as strategies for engaging students through new literacies. They argue that a visual essay is “a text that relies more heavily on images with minimal print text” that “entails new forms of semiotic processing of the combinations” of semiotic modes (Hughes and Tolley 2010). Although, there can still be encountered relations of elaborations, images and text enter especially in relations of ideational complementarity in this genre. The focus of these relations is predominantly on communicating knowledge about research results and not only data. Tendencies of Multimodal Gradations in Academic Genres Network 131 In Sound and Vision (Van Leeuwen 2007), a black and white visual essay with the usual information about author, abstract and key words on its first page, the author makes direct reference to the meaning-making function of spatial organization: “Experimenting with page layout as a discursive mode, this visual essay offers a brief history of immersion”(Van Leeuwen 2007: 136). In this visual essay, the text is no longer dominant, and some parts of it are not even separated from images; on several pages, sentences or just words are superimposed on images. The greatest majority of multimodal relations in the essay are relations of extension through complementarity. For example, text is superimposed on a page that represents a detail from a photo of a sound study. This photo that visualizes sound vibrations can be also found on another page with inversed colors (black is white and white is black). On it, the superimposed text is represented by four quotations from other semioticians. Fig. 3: Example of a page from Van Leeuwen’s visual essay As opposed to Van Leeuwen’s visual essay, Design and the Aesthetics of Research (Roxburgh 2010), a 15 pages academic visual essay in color, replace the last traces of “traditional” academic argumentation still present in the first essay with a new approach where mathematical symbols add new meanings to the whole argumentation. Knowledge about research results is communicated in even more condensed multimodal relations in these pages. On the first page of the essay, under the label “abstract”, the author reveals not only information about the respective visual essay but also about his research concerns, namely “the aesthetics of research” (Roxburgh 2010: 425), and his research work in general. Overall, the relations of extension through 132 Carmen Daniela Maier / Jan Engberg complementarity also predominate in this visual essay when knowledge is communicated at the intersection between text and various visual elements. Fig. 4: Example of a page from Roxburgh’s visual essay Czakó’s visual essay (2011), Identity, lacks any kind of textual content - not even a list of references - apart from the first page of the essay that includes information about the author and his short and rather metaphorical clarifications about the topic of the essay. For example, “like a stone at the seaside, identity is shape by others” (Czakó 2011: 419). The rest of the visual essay includes a series of images that have rather loose relation of extension through complementarity with the short text, relations that might even be lost by an audience unfamiliar with Czakó’s work as graphic designer and photographer. The exploration of these visual essays highlights that the rather balanced link between tacit and explicit knowledge that can be discerned in the case of research articles starts to lose its equilibrium. The readers can no longer use their genre specific tacit knowledge in the usual way because they can neither expect nor predict with the same certainty the structural development of a multimodal text labeled as visual essay. The organizational integrity of the genre is no longer a priority for the researcher. More alternatives are open to the researchers who intend to communicate knowledge through this pagebased genre that is no longer restricted by the established constraints of the text-based research article. Furthermore, the domain specific explicit knowledge extends its boundaries incorporating knowledge that belongs to other research domains like mathematics, for example. Tendencies of Multimodal Gradations in Academic Genres Network 133 In the light of the above mentioned findings, we turn to Van Leeuwen’s characterization of genre in order to summarize our present understanding of the academic visual essay: “The text itself is no longer a staged, goal oriented process. It is an environment for such processes, and must be analysed as a kind of map, a spatial structure allowing a number of trajectories, or as the layout of a building, spatial structure designed to facilitate a range of specific activities” (Van Leeuwen 2005: 85). It is particularly interesting that we can employ almost the same characterization for academic video essays. However, it is important not to lose sight of the fact that we have to introduce a temporal dimension when exploring the academic video essay. As in the case of visual essays, there have also been found transitional and intermediate forms of video essays. For example, Endless Semiosis, Eriksson’s video essay (2010) consists of a series of still images in colors that fade in and out of the screen and that have superimposed texts. The transitional status of this multimodal text is given not only by the fact that it is just a succession of still images without any moving images; it is also reinforced by the fact that most of the images are labeled with words that paraphrase the respective images. According to Bateman, “whenever film employs text (…) placing text within the spatial array is a natural invocation of the resources of the pageflow” (Bateman 2008: 277). In this case, it cannot even be maintained that Endless Semiosis is a film as there are no moving images, so the connection with the page-flow is even stronger. The meaning of the words is similar to the content of the images and therefore the multimodal relations are represented by relations of ideational concurrence through similarity. Fig. 5: Example of a screen shot from Van Eriksson’s video essay These relations are also indirectly characterized as such in the text below the video’s screen, namely “the symmetrical relation between the written text and the visual text”. Relations of connection appear also at the level of image se- 134 Carmen Daniela Maier / Jan Engberg quencing when the clusters of images follow each other without a narrative connection. The relation is mainly formally brought to life as the same words appear as overlapping towards the end over different images. With Boogie Street (2010), Van Eriksson takes a step closer to what we assume that can be classified as a prototypical video essay. We consider it still an intermediate form because, once again, the still images predominate again although moving images are also introduced. In this video essay, the first part is represented by an alternating succession of still images. A close up shot of an eye is alternated with various striking shots taken from Google street. The ideational relation is one of connection through temporal enhancement as the two types of images following each other suggest a relation of simultaneity: the close up of the eyes are followed by various shots of what the eye presumably sees. At least this is the effect of this succession, although it is clear that the shots are from Google street. Additionally, two other types of relations are embedded in this visual alternation. In the close up shots of the eye, the eye seems to deteriorate. So, these shots enter into a new relation of temporal connection with each other. The clusters of Google street shots are also entering into a relation with each other, a relation of elaboration, as they function either as overview or as detail of each other. Fig. 6: Examples of screen shots from Van Eriksson’s video essay In the 2 nd part of the video essay, when the video’s shots start to be accompanied by superimposed texts, the ideational relations are either of neutrality or of elaboration through specification. For example, the information provided in the rhetorical questions and answers of the researcher and the information provided by the images on which they are superimposed exist in parallel. The captions with films’ titles anchor the meaning of the accompanying shots from the respective films entering therefore in ideational relations of elaboration through specification. The third part of the video remediates excerpts Tendencies of Multimodal Gradations in Academic Genres Network 135 from the feature films Blade Runner and Deja Vú dealing with the discussed topic. In the last part of the video, the researcher introduces shots in which he films himself. Fig. 7: Examples of screen shots from Van Eriksson’s video essay 5.3 The image flow in the academic video essay By applying the concept of image-flow which refers to “temporal sequentiality” (Bateman 2008: 175), the exploration of the prototypical academic video essay has been conducted in the present study in order to highlight the specific meaning-making relations between several semiotic modes that differentiate it from the academic research article and the academic visual essay. As the genre of video essay is borrowed from the context of film, we start by presenting how it has been discussed in the field of film theory. The genre of film essay was first conceptualized by the avant-garde filmmaker Hans Richter at the beginning of the forties. According to Alter, Richter proposed “a new genre of film that enables the filmmaker to make the ‘invisible’ world of thoughts and ideas visible on the screen” (Alter 2003: 13). Since then, both film critics and video essayists have continuously tried to capture and explain the essence of this “in-between genre” which cannot be crystallized in a formula like traditional film genres. The traditional coherence strategies like the voice-over commentaries and the succession of images are undermined by the complex mix of various kinds of both still and moving images belonging to various media and other film genres that are edited in loosely constrained ways. Instead of being discussed as a clearly defined genre, the video essay is sometimes referred to as “a particular method and approach” (Huber 2003: 93) . 1 The fact that video essays are difficult to be classified as manifestation of a prototypical generic form is also manifested in the variety of labels that are used by re- 136 Carmen Daniela Maier / Jan Engberg When appearing in academic context, this “method” of visualizing knowledge expands the range of possibilities for the representation, communication and reception of academic knowledge. The boundaries of both genre specific tacit knowledge and domain specific explicit knowledge are continuously moved in the complex construct of video essay, and this creative instability becomes itself a dynamic knowledge communication strategy. Additionally, although they belong to the same discourse community, the viewers might have “to guess” the researcher’s intentions by making use of various types of knowledge, becoming thus more involved in knowledge creation than if reading a conventional research article. Certainly, the chances for misunderstandings or controversies are greater especially if viewers do not have the same high level of visual or media literacy as the creators of the video essays, or if the range of domain specific “knowledges” integrated in the video essays is too wide. In Lindgren’s video essay, Geek Revenue (2011), this danger is almost avoided. The moving images of this video essay visualize well known aspects of the research process from books reading to taking notes or interviews. The author of the video essay appears either talking in the image or just as voiceover guiding the viewer through this “research journey”. The methodological approach of the domain specific research work is explicitly mentioned in the text accompanying the video: This video essay uses classical cultural theory as well as current internet research to address the relationship between the cultural industries and the increasingly active and tech-savvy audiences of the 21st century (Geek Revenue, 2011). The video’s images and the accompanying text are linked by elaboration through specification. This type of relationship of ideational concurrence appears also between the video’s images and the voiceover commentaries. When not taking interviews, the researcher addresses the viewers by maintaining eye contact with them during his explanations, minimizing in this way the distance between them. The relationships between images are kept at the level of ideational enhancement as the video essay seems to trace the researcher’s journey towards acquiring more domain specific knowledge. The viewers witness this journey that they also usually experience while doing research. Therefore, it can be assumed that when the viewers are given the witness’ role, the tacit knowledge that comes into play here is not limited to structuring a research genre, but it is extended to performing the research process. The author does not only explicitly communicate domain specific knowledge, but he also visualizes and verbalizes the process of acquiring it. This process is communicated in the video’s moving images through revealing the resemiotization of knowledge from talk to page. searchers who publish their work on AudioVisual Thinking’s homepage, the journal specialized in publishing video essays: visual text, video project, video text, experimental documentary, short film, video sketchbook, etc. Tendencies of Multimodal Gradations in Academic Genres Network 137 Fig. 8: Examples of screen shots from Simon Lindgren’s video essay Another video essay, Remediation (Joseph 2011), uses audio interviews that enter in relations of ideational complementarity with a series of moving images remediated from a wide range of sources of domain specific knowledge. There is no superimposed text that could disclose the images’ sources or clarify their relations with the accompanying audio interviews. These relations of complementarity through extension stretch the distance between the meanings realized in the texts and images to the point in which viewers might have difficulties to follow both of them simultaneously, namely to participate in this game of spotting (Joseph 2011). The author confesses in the accompanying text that is placed under the screen: Using found footage sourced from the internet the film sets out to playfully visualise the recorded audio interviews through analogy, allegory and visual rhetoric (Joseph 2011). In so doing, the author also engages the viewers in a process of questioning the creation process because the subject matter of the film is to look at the moral ambiguities of artists’ use of visual sampling and found footage in order to create new works (Joseph 2011). Fig. 9: Examples of screen shots from Miles Joseph’s video essay 138 Carmen Daniela Maier / Jan Engberg In the case of some visual and video essays in which the visual mode plays a predominant role and the verbal mode incorporates higher level of conceptualization reducing the text to minimal dimensions, a new strategy for communicating domain specific knowledge appears. A new relation of extension through complementation is established at the level of the whole multimodal text with an accompanying text that is supposed to clarify the research insights. For example, in the case of Roxburgh’s visual essay, that text is in fact an extended abstract that is displayed on the whole first page. This extended abstract clarifies the overall meanings of the next 14 pages. In the case of Eriksson’s video essay, Boogie Street, the text is placed below the screen on the journal’s homepage. Both meta texts communicate knowledge about the whole research work process. In such cases, the page-flow specific to visual essays and the image-flow specific to video essays are thus intertwined with text-flow at another level of multimodal structure that is extended beyond the usual frame of those genres. These meta texts are supposed to both add more specific domain knowledge, but also to control the video essay’s aesthetic meaning making potential in order to avoid the appearance of the misunderstandings or controversies that have been mentioned above. 5.4 Intrageneric and intergeneric intertextuality As one of the common purposes of these genres is to acknowledge previous knowledge, and as the authors are required to credit all sources and references, various forms of intertextuality are manifested in each of them. The different configurations of semiotic modes and media that characterize the structure of each of these genres influence the recurrent manifestations of intrageneric and intergeneric intertextuality in various ways. As expected, due to the well-established generic requirements, all the explored research articles contain intrageneric references. The references are made both verbally through quotes and visually through tables or diagrams retrieved from other research articles with similar research concerns. For the time being, intrageneric intertextuality is to be found only in research articles. This situation can be explained in two ways. First, through the rather limited number of academic visual and video essays that exists today. Cook also observes that since the early days of advertising, intrageneric intertextuality has increased its number of appearances as ads “have grown in quantity and salience, and accumulated their own history and tradition” (Cook 2001: 194). Second, through the fact that when either communicating explicit knowledge or implying tacit knowledge, the emphasis on personal creativity and original representations is much stronger than in the case of research articles. This can be linked with the fact that - as genres - the academic visual and video essays have their generic roots in traditional visual and video essays which put a strong emphasis on these two qualities: personal creativity and original representations. So, although academic visual and video essays are characterized by Tendencies of Multimodal Gradations in Academic Genres Network 139 complex visual associations among various types of still and moving images, they are not yet characterized by recurrent intrageneric intertextuality. As far as the intergeneric references are concerned, the research articles intertextually refer to other genres when presenting and analyzing specific sets of data. In accordance with the chosen data, the research articles include visualized references to advertisements (in O’Halloran 2008 and White 2010), posters (in White 2010), art works (in Thurlow and Aiello 2007), journal homepages (in Knox 2007), etc. In Martinec and Salway’s article (2005), the visualized representations of a CD-ROM encyclopedia for children, advertisements, homepages, and magazine pages are resemiotized appearing as drawings in the article due to copyright issues. The inclusion of data visualizations like homepages, visualized representations of CD-ROM, or mobile phone (in White 2010) is realized through processes of remediation. Not only data-related intergeneric intertextual references are remediated in the research articles. For example, O’Halloran (2008) includes a visualized reference to methodology by reproducing the screen capture of Adobe Premiere software (see image above). The unavoidable losses in knowledge communication about certain generic characteristics that are encumbered in remediation processes are minimized through verbal descriptions and clarifications that usually accompany the visual references either in captions or in the main text of the article. Such remediation processes, through which other genres appearing in other media are visually referred to, can also be encountered in visual and video essays. However, the descriptions and clarifications are either absent or scarce, as the existence of common tacit knowledge is more strongly assumed in both cases than in research articles. The situation is more complex in the case of video essays as this assumption is extended across even more various genres and media. Quoting Jenkins’ study of fan cultures, Verhoert also points out that the reception of a video essay should be regarded as a complex process because “the use of video forms fosters both multiple critical ‘rereading’ and ‘intertextual knowledge’” (Verwoert 2003: 28). From the point of view of a video essayist, Biemann highlights the fact that “the essay has always distinguished itself by a non-linear and non-logical movement of thought that draws on many different sources of knowledge. In the digital age, the genre experiences an even higher concentration” (Biemann 2003: 9). 6 Conclusions At the end of our chapter’s introduction, we have anticipated that the new technology-mediated gradations that link the analyzed genres give birth to new relations between researchers and their ways of communicating academic knowledge. Obviously, our preliminary analyses presented in this chapter, 140 Carmen Daniela Maier / Jan Engberg although focused only on some multimodal relations, have revealed how these new relations are manifested in clear tendencies emerging at least in the genres circulating today in certain discursive communities. Based on an interdisciplinary approach, our preliminary explorations of these genres have thus allowed us to discern the following specific tendencies. First, if in research articles the multimodal relations nuance the communication of knowledge about methodologies, in the visual and video essays, they predominantly communicate knowledge about the research results. Second, the appearance of an intensification of multimodal complementation relations in the form of multilayers of relations between the same verbal and visual elements both in visual and especially in video essays influences the processes of meaning-making. This multilayering makes it possible to have multiple interpretations. A third tendency consists in the appearance of additional textual explanations that attempt to enhance the communicative clarity and precision of the research work presented through a visual or a video essay. These metatexts extend the usual boundaries of these two genres. At the same time, neither in visual or video essays have we found any kind of intrageneric references. We have also revealed that the relation between tacit and explicit knowledge has become more nuanced and even more difficult to pinpoint. In our analyses we have investigated the ideational relations concerning the domain-specific knowledge of the treated field conveyed by the texts as well as the relation between tacit and explicit knowledge in the more genre-related knowledge behind creating texts of the genre. Especially the relations concerning the last type is of more general interest to the topic of this chapter. As the two emerging genres are not yet established, but investigating boundaries and possibilities, writers cannot rely upon tacit knowledge about norms and conventions held by the readers due to their belonging to the academic community. If the writers wanted to secure an optimal understanding of the content conveyed by their texts, a prototypical strategy would be to make the relevant knowledge on relations explicit. However, apart from including encompassing introductory texts or abstracts the authors of the investigated texts do not do so to any great extent. This is probably due to the fact that the two emerging genres have their traditionally non-academic genres like the video or the visual essay. Our analyses suggest that the rationale behind developing these emerging genres lies more in including personal creativity and aspects of original representation from the traditional video and visual essays in the field of academic communication and less in enhancing the traditional academic research article along the lines of its core characteristics. So they seem to lean more towards the ‘essay’ than towards the ‘academic’ aspect (cf. 3.3 above). Furthermore, according to our preliminary findings based on the range of multimodal texts chosen from the two international journals, the dynamic network formed by these genres is characterized by the existence of various Tendencies of Multimodal Gradations in Academic Genres Network 141 forms of multimodal gradations due to the overlapping and constantly evolving generic forms. For example, some multimodal texts are already labeled as academic visual essays while maintaining the main characteristics of research articles apart from intrageneric intertextuality. Similarly, we found multimodal texts labeled as video essays. However, they are structured as successive still images with superimposed texts, resembling in this way a visual essay. Above (p. xx) we suggested to distinguish in this context between transitional and intermediate forms, where the transitional forms are closer to the original forms than the intermediate ones, i.e., the transitional forms share more characteristic traits with the original forms than the intermediate ones. In the following figure we have schematically presented the distinction and placed the three investigated genres according to this criterion. Fig. 10: The representation of generic gradations in academic genres network We consider both new genres to have emerged from research articles for the reasons stated above. Visual essays are closer to research articles than video essays as they share more elements (higher importance of written text in visual essays than in most video essays, printed format (page flow) as opposed to screen format (image flow), to name only two differences). At the same time, the analysis has shown that the difference between transitional and intermediate forms also applies to the description of the texts within the two emerging genres. In conclusion, we consider that the exploration of the academic genre network is extremely relevant today, and we suggest that the issues discussed above have to be revisited systematically and recurrently because we are dealing with rapid processes of generic transformations. We expect that future explorations of the ongoing development of this academic genre network will definitely fill in more tendencies in the multimodal gradations that have been discussed in this chapter. The communication of academic knowledge will continue to evolve in ways that can be unforeseen at the present moment due to rapid technological developments, and therefore keeping pace with these developments might be one of researchers’ biggest challenges in years to come. 142 Carmen Daniela Maier / Jan Engberg The almost common by now multimodal resources facilitated by the online versions of prestigious peer-reviewed journals represent an obvious hint indicating the range of multimodal communicative competencies that are already expected from researchers belonging to various academic domains. Along with becoming literates in academic knowledge creation, we might all be expected in the future course of our academic careers to become routinized multiliterates, namely experts in resemiotization and remediation of the academic knowledge that we create in the context of our specific discourse communities. Simultaneously, if we recall Hyland’s understanding of academic discourse quoted in our presentation of knowledge communication approaches (Hyland 2009), it becomes obvious that the implications of the rapid technological developments reverberate both in processes of generic transformations and in our teaching activities too. Dealing on a daily basis with generations of students accustomed with searching, creating and sharing multimodal knowledge across language barriers, geographical borders and time zones means that we have to (re)position ourselves as part of this multiliterate community in order to be able to reach our students where they are now in terms of communicative expertise. References Alter, M. Nora. (2003) Memory Essays in Ursula Biemann (ed.) Stuff It: The Video Essay in the Digital Age, 12-23 . U. Zurich: Edition Voldemeer. Bakhtin, Mikhail. (1986) Speech Genres and Other Late Essays. Austin, Tx: University of Texas Press. Bateman, John. (2008) Multimodality and Genre. A Foundation for the Systematic Analysis of Multimodal Documents. New York: Palgrave Macmillan. Bateman, John. (2009) Discourse across Semiotic Modes in Jan Renkema (ed.) Discourse, of course. An Overview of Research in Discourse Studies, 55-67. John Benjamins. Bazerman, Charles. (1988) Shaping Written Knowledge: The Genre and Activity of the Experimental Articles in Science. Madison: University of Wisconsin Press. Beaufort, Anne. (1999) Writing in the Real World: Making the Transition from School to Work. New York: Teachers College Press. Berkenkotter, Carol and Huckin, N. Thomas. (1995) Genre Knowledge in Disciplinary Communication. Hillsdale, NJ: Erlbaum. Bhatia, K. Vijay. (1993) Analysing Genre: Language Use in Professional Settings. London: Longman. Bhatia, K. Vijay. (2008) Towards Critical Genre Analysis in Vijay K. Bhatia, John Flowerdew and Rodney Jones (eds.) Advances in Discourse Studies, 166-177. London: Routledge. Biemann, Uursula. (2003) The Video Essay in the Digital Age in Ursula Biemann (ed.) Stuff It. The Video Essay in the Digital Age, 8-12. U. Zurich: Edition Voldemeer. Tendencies of Multimodal Gradations in Academic Genres Network 143 Bolter, David and Grusin, Richard. (2002) Remediation: Understanding New Media. Cambridge: The MIT Press. Bourdieu, Pierre. (1990) The Logic of Practice. Cambridge: Polity Press. Cook, Guy. (2001) The Discourse of Advertising. Routledge. Czakó, Zsolt. (2011) Identity in Visual Communication , vol. 10, no.3, 419-432. Fairclough, Norman. (1992) Discourse and Social Change. Cambridge: Polity Press. Giddens, Anthony. (1986) The Constitution of Society. Berkeley, CA: University of California Press. Goodrum, Alison and Hunt, J. K. (2011) Framing rural fashion: observations from Badminton Horse Trials in Visual Communication , vol. 10, no.3, 287- 324. Halliday, A.K. Michael. (1978) Language as Social Semiotic: the social interpretation of language and meaning. Edward Arnold: London. Halliday, A.K. Michael. (1994). Introduction to Functional Grammar. London: Edward Arnold. Hyland, Ken. (2009). Academic Discourse: English in a Global Context. London: Continuum. Huber, Jörg. (2003) Video Essayism: On the Theory-Practice of the Transitional in Ursula Biemann (ed.) Stuff It: The Video Essay in the Digital Age, 92-98. Zurich: Edition Voldemeer. Hughes, Janette and Tolley, Sarah. (2010) Engaging Students through New Literacies: the good, bad and curriculum of visual essays in English in Education, vol. 44, no.1, 5-26. Iedema, Rick. (2003) Multimodality, Resemiotization: extending the analysis of discourse as multi-semiotic practice in Visual Communication , vol. 2 no. 1, 29-57. Jaworski, Adam. (2011) Tracing Place, Locating Self: Embodiment and Remediation In/ Of tourist Spaces, in Visual Communication , vol. 10, no.3, 349-364. Kastberg, Peter. (2007). Knowledge Communication: The emergence of a third order discipline in Claudia Villiger & Heidrun Gerzymisch-Arbogast (eds.), Kommunikation in Bewegung : Multimedialer und multilingualer Wissenstransfer in der Experten-Laien-Kommunikation. Festschrift für Annely Rothkegel, 7-24. Berlin: Lang. Kress, Gunther. (2005) Gains and Losses: New forms of texts, knowledge, and learning in Computers and Composition, vol. 22, 5-22 Kress, Gunther. (2010) Multimodality. A social semiotic approach to contemporary communication, London: Routledge. Kress, Gunther and Jewitt, Carey. (2000). Multimodal Literacy. New York: P. Lang. Kress, Gunther and Van Leeuwen, Theo. (2001) Multimodal Discourse: The modes and media of contemporary communication, London: Edward Arnold. Kress, Gunther and Van Leeuwen, Theo. (2006) Reading Images. The Grammar of Visual Design. London: Routledge. Kong, Kenneth. (2006) A Taxonomy of the Discourse Relations between Words and Visual. Information Design Journal, vol.14. no.3, 207-230. Knox, John. (2007) Visual-Verbal Communication on Online Newspaper Home Pages Visual Communication, vol. 6, no.1, 19-53. 144 Carmen Daniela Maier / Jan Engberg Kuppens, An. (2009) English in Advertising: Generic Intertextuality in a Globalizing Media Environment in Applied Linguistics, vol. 31, no.1, 115-135. Laurén, Christer. (1993). Fackspråk. Form, innehåll, funktion.: Lund. Lim, F. Victor. (2004). “Developing an Integrative Multi-semiotic Model” in Kay L. O’Halloran (ed.) Multimodal Discourse Analysis. Systemic-Functional Perspectives, 220-247. London: Continuum. Macken-Horarik, Mary. (2004) Interacting with the Multimodal Text: Reflections on image and verbiage in ArtExpress. Visual Communication, vol. 3, no.1, 5- 26. Maier, Carmen D. and Engberg, Jan. (2013) Tendencies in the Multimodal Evolution of Narrator’s Types and Roles in Research Genres in Maurizio Gotti and Carmen Sancho Guinda (eds.) Narratives in Academic and Professional Genres. Bern: Peter Lang AG, 149-175. Martin, Jim and Rose, David. (2008) Genre Relations: Mapping Culture. London: Equinox. Martinec, Radan and Salway, Andrew. (2005). A System of Text-image Relations in New (and Old) Media. Visual Communication, vol. 4, no.3, 337-371. Miller, R. Carolyn. (1984) Genre as Social Action. Quarterly Journal of Speech vol.70, 151-67. Myers, Greg. (2000) Powerpoints: Technology, Lectures and Changing genres. In Anna Trosborg (ed.) Analysing Professional Genres, 177-192. Amsterdam: John Benjamins. Nonaka, Ikujiro and Toyama, Ryoko. (2003). The Knowledge-creating Theory Revisited: knowledge creation as a synthesizing process. Knowledge Management Research & Practice, vol. 1, 2-10 O’Halloran, L. Kay. (2004). Visual Semiosis in Film” in Kay L. O’Halloran (ed.) Multimodal Discourse Analysis, 220-246. London: Continuum. O’Halloran, L. Kay. (2008) Systemic Functional-Multimodal Discourse Analysis (SF_MDA): constructing ideational meaning using language and visual imagery, in Visual Communication , vol.7, no. 4, 443-475. O’Halloran, L. Kay. (2009) Historical Changes in the Semiotic Landscape: From Calculation to Computation in Carey Jewitt (ed.), Handbook of Multimodal Analysis, 98-113. London: Routledge. Paltridge, Brian. (1997) Genre, Frames and Writing in Research Settings. Amsterdam and Philadelphia: John Benjamins. Polanyi, Michael. (1969) Knowing and Being. Chicago: University of Chicago Press. Royce, D. Terry. (2007). Multimodal Communicative Competence in Second Language Contexts in Royce, D. Terry and Wendy L. Bowcher (eds.) New Directions in the Analysis of the Multimodal Discourse, 361-391. London: Lawrence Erlbaum Associates. Roxburgh, Mark. (2010) Design and the Aesthetics of Research in Visual Communication, vol. 9, no. 4, 425-439. Smart, Graham. (2006) Writing the Economy: Activity, genre, and technology in the world of banking. London: Equinox Publishing. Swales, John. (1990) Genre Analysis. English in academic and research settings. Cambridge: Cambridge University Press. Tendencies of Multimodal Gradations in Academic Genres Network 145 Swales, John. (2004) Research Genres: Explorations and applications. Cambridge: Cambridge University Press. Thurlow, Crispin and Aiello, Giorgia. (2007) National Pride, Global Capital: a social semiotic analysis of transnational visual branding in the airline industry in Visual Communication, vol. 6, no.3, 305-344. Unsworth, Len. (2006) Towards a Metalanguage for Multiliteracies Education: Describing the meaningmaking resources of language-image interaction, English Teaching: Language and Critique, vol. 5, no.1, 55-76. Verwoert, Jan. (2003) Double Viewing: The Significance of the Pictorial Turn to the Critical Use of Visual Media in Video Art in Ursula Biemann (ed.) Stuff It. The Video Essay in the Digital Age, 24-34. U. Zurich: Edition Voldemeer. Van Leeuwen, Theo. (2005) Introducing Social Semiotics. Routledge. Van Leeuwen, Theo. (2007) Sound and Vision in Visual Communication, vol.6, no.2, 136-45. White, Paul. (2010) Grabbing Attention: the importance of modal density in advertising in Visual Communication, vol. 9 no. 4, 371-397. Video essays Eriksson, T. 2010. Endless semiosis. In Audiovisual Thinking, downloaded at http: / / www.audiovisualthinking.org/ dokument3/ edition01/ Eriksson, T. 2010. Boogie Street. In Audiovisual Thinking, downloaded at http: / / www.audiovisualthinking.org/ dokument3/ edition02/ Joseph, M. 2010. Remediation. In Audiovisual Thinking, downloaded at http: / / www.audiovisualthinking.org/ videos/ issue02/ remediation/ Lindgren, S. 2011. Geek Revenue. In Audiovisual Thinking, downloaded at http: / / www.audiovisualthinking.org/ videos/ geek/ Ole Togeby A Model of Text Types and Genres 1 Communicative act Research in text type and genre is often restricted to certain types of textual material, e.g. to the types of speech acts, to historical developed genres of literary texts, to linguistic features, moves and strategies of English for Specific Purposes, or to dependency of adjacency pairs in conversation. It is the intention in this article to develop a theory of all text types or genres and to elaborate on the relations between the different theoretical approaches and different material selection. It is taken for granted that a communicative act is an utterance that has linguistic form, is a representation of some state of affairs, and counts as a social act in relation to the other parties in the communication. The minimal communicative unit is a speech act performed by the uttering of one sentence with a truth value and one illocutionary force, but a text that consists of many concatenated sentences, each with its own truth value, also counts as one intentional and communicative unit. In this article it will be discussed how many sentences, each of which is a potential speech act, are composed into a textual whole that counts as a single communicative act. 2 Main text types The definition of the concept ‘text type’ is straight forward: A text type is a subcategory or subclass of texts. But what is a text? In everyday usage a ‘text’ denotes a unit of written language, and ‘an utterance’ normally denotes a unit of spoken language. But here the concept ‘text’ covers both written and spoken language; ‘text’ is defined in the following way: A ‘text’ (or strictly speaking a ‘text act’) is a communicative act that a sender performs when uttering a complete intentional unit of several written or spoken sentences delimited by silence or blank space, or by shift of sender. It is a criterion for textuality (among others) that a text consist of several sentences. One single uttered sentence counts as a speech act, and in a dialogue a remark need not contain more than one sentence or an elliptic sentence. A speech act, e.g. an oral promise, is defined in the following way: A speech act is a communicative act that a speaker performs when uttering a single sentence. 148 Ole Togeby The completeness of a text act means that it is a communicative occurrence which meets standards of textuality, viz cohesion, coherence, intentionality, acceptability, informativity, contextuality and intertextuality (Beaugrande & Dressler 1981). The concept of a ‘text’ will not be discussed further here, but many of these criteria for textuality will be dealt with in the following. If the concept ‘text’ is defined as an utterance consisting of several sentences, but with one communicative purpose, and delimited in time by shift of sender, quite dissimilar examples fulfill the criteria: a letter to the editor in a newspaper, a book about ornithology and a poem in a collection. They are prototypical examples of three main types of texts: P RACTICAL T EXTS that are normally conceived as acts in a social practise; they are often explicitly addressed to specific people having roles, rights and duties in institutions, organizations, society and culture; they are normally consisting of connected sentences on a page or two, and they have only one social purpose, e.g. a letter to the editor in a newspaper debate. F ACTUAL PROSE that is meant to be a true representation of some states of affairs talked about; it is most often written language (perhaps read aloud), not addressed to specific receivers, but is a piece of work exposed to the general public, often of book length, and only with the purpose of enlightenment of people being interested in the subject, e.g. a book about ornithology. L ITERATURE (imaginative) is a work of remarkable form produced by an artist; it is detached from specific organisational settings; a literary text is not read as a means to an end, but as an end in itself, being a ritual of pastime, play, or entertainment for the audience and an occasion for outlook on life and human affairs in general, e.g. a poem. Communicative acts are here classified in two separate classes: speech acts, which are the utterance one single sentence, and texts (text acts), which are utterance of several coherent sentences that make up a complete intentional unit. Speech acts are subdivided in speech act types, and texts (text acts) are subdivided in practical texts, factual prose and literature. Each of the three main types of text acts are subdivided in what in everyday language is called text types or genres, two terms that are taken in the following to be overlapping synonymous, and of which the term text types is taken to be the most encompassing and useful. A Model of Text Types and Genres 149 Fig. 1: Speech acts, text types and genres Notice that each of the four main types of communicative acts has its own dominant property and its own criterion for division into subcategories: for speech acts the criterion is the intention or illocutionary purpose; practical texts are subdivided according to the relations of power and knowledge established in the communicative situation and the organisation in which it is functioning; factual prose is subdivided according to the type of facts it is a representation of, and literature according to the type verbal form utilised in the text. Text types (practical texts, factual prose and literature) are not exclusive Aristotelian categories defined in terms of their necessary (defining) and sufficient (distinguishing) properties (genus et differentia) ordered in a taxonomy. They are radial categories with fuzzy edges and both central and peripheral examples, denoting cultural kinds belonging to different experiential domains and structured by chaining according to general properties (Lakoff 1987). A ‘letter to the editor’ is a more central example of the practical text type than a ‘label on a bottle’, and a ‘feature article’ is at the same time a ‘practical text’ and a piece of ‘factual prose’. As defined here, a text type is not a theoretical concept, but a concept according to which people change their behavior in producing and interpreting texts. When the three main subcategories of texts are radial categories and not Aristotelian categories the model is best illustrated, not by a tree diagram, but by an triangular area with radial structure: Dialogues Factual prose Imaginative literature Works Monologues text types Social practice Rhetorical acts Practical text types Speech acts Genres 150 Ole Togeby Fig. 2: Three types of text Cultural categories like ‘speech act type’, ‘text type’ and ‘genre’ do not denote objects in the world, external to human beings. They are defined jointly by the external physical world (the manifest forms of texts), human biology and human mind (the meaning of the text), and social interaction. 3 Dimensions of variation As radial categories the four main categories of communicative acts differ along many dimensions in the communicative situation and are kept together to one category by many family resemblances that are not shared by all four categories, but only of a couple of them. The dimensions are: oral or written, consisting of one sentence or of several sentences on a page or two, or in a big volume, a remark in an interactional dialogue or one coherent monologue, structured in time by interaction or by composition, a fleeting or permanent phenomenon, simultaneous or asynchronous processing by sender and receiver, used up or reuse again and again, one-to-one communication or oneto-many, addressed message or put forward to the public audience (for another proposal of genre dimensions or parameters, see Østergaard and Bundgaard in this volume). On these dimensions the four main types of communicative acts vary with speech acts in one side of a continuum, practical text types and factual prose in the middle and literature in the other side, as shown in the diagram: F ACTUAL P ROSE Representations of states of affairs A Model of Text Types and Genres 151 ORAL WRITTEN ONE SENTENCE BIG VOLUMES DIALOGUE MONOLOGUE INTERACTION speech practical factual litera- COMPOSITION FLEETING acts texts prose ture PERMANENT SIMULTANEOUS ASYNCHRONOUS USED UP REUSABLE ONE - TO - ONE ONE - TO - MANY ADDRESSED TO THE PUBLIC Tab. 1: Main text categories and general communicative conditions Here are examples of the four main types of communicative acts: Example 1: Speech act (Soya: To traade, Den skønne Fortid, 1943 p. 8 ) Example 1 is an utterance - delimited by shift of speaker - in the beginning of a play by the Danish author Soya. Here it is taken as an example of what is said by one of the characters in the play, Oda, and it is disregarded that it is in fact part of a literary play written by an author. Two women, Oda and her woman friend, enter the scene: Friend: Yeah jealous, he has always been. Oda: Yes but it is nothing compared to what he is now. Sometimes he behaves like a madman! [Two Women appear. The one pushing a pram.] There is a table here let us - [The long man turns his head ... and stares at them.] Friend: I do not understand that you still stand him. If I were you. 1\ Yes but it's nothing compared to what he is now. 2\ Sometimes he behaves like a madman! 1\ Johmen det er ingenting mod hva’ han er nu. 2\ Undertiden ter han sig saa man skulde tro han var sindssyg! 3\ Der er et Bord her - 4\ la’ os - 3\ There is a table here - 4\ let us - 152 Ole Togeby In example 1 there are four sentences, and four potential speech acts, viz 1) an estimation of how insane her husband is: Yes but it is nothing compared to what he is now, with 2) a statement of facts that serves as grounds for it: Sometimes he behaves like a madman! These two are separated from the next two sentences by seconds of silence. 3) Oda’s exclaiming that she has found a free table: There is a table here, and 4) a mutual request: let us - which is taken to be elliptical for let us sit down here. (The sentences 1-2 can also be analysed as one text consisting of two sentences, and 3-4 as another text consisting of two sentences, but the sentences are here taken as four speech acts.) The sentences 3 and 4 in example 1 can be seen as typical speech acts being oral units of one sentence, part of a dialogue, part of an interaction, fleeting, simultaneous, used up when it is uttered, one-to-one, and addressed (by imperative mode in 4). 3 and 4 have different illocutionary force and they function as different speech acts in the interaction, 3 as a statement (a constative speech act) and 4 as mutual request (a directive speech act). Example 2: Practical text (Information 16. februar 2010 p. 17, under the headline Letters to the editor) Example 2 is a typical example of a practical text; it is written in a newspaper, consists of many sentences, in a sort of slow dialogue (the example being a counterargument against something written in the paper some previous day), interaction, both semi-synchronous and permanent, used up during a couple of days, with actuality only a couple of days, a complex of one-to-one and one-to-many communication, and as a criterion of it’s type, addressed to both the Liberal tax spokesman and the readers of the actual edition of the paper in an attempt to convince them of a political issue. 1\ En anden mulighed! 2\ Henrik Moberg Jessen, København S 3\ Venstres skatteordfører udtaler i Information (den 15. februar), at danskerne ikke frygter skattestigninger mere, og at det skyldes “skattestoppets succes”. 4\ En anden mulighed kunne være, at danskerne mærker konsekvenserne af skattestoppet på plejehjemmene, sygehusene, skolerne, institutionerne og mange andre steder i Danmark. 1\ A second possibility! 2\ Henrik Moberg Jessen, Copenhagen S 3\ Liberal tax spokesman said in Information (February 15) that the Danes do not fear tax increases any more, and this is due to “the success of the tax freeze”. 4\ A second possibility could be that Danes feel the impact of the tax freeze in nursing homes, hospitals, schools, institutions and many other places in Denmark. A Model of Text Types and Genres 153 Example 3: Factual prose (Hans Hvass 1954: Fugle i farver. Politikens Forlag, p. 33 and 144) Example 3 is a typical piece of factual prose communicating facts about a bird species. It is written, part of a book, monologue, ordered in the text by composition (according to bird families), permanent, asynchronous, reusable year after year, one-to-many communication, and, as a criterion of it’s type, unaddressed and exposed to the public. 1\ Bohemian Waxwing, Bombycilla Garrulus 2\ 20-21 cm. 3\ Quite common, but very fickle winter visitor, which may appear in very large quantities. 4\ It is a nutrient deficiency on the breeding grounds in northern Scandinavia and perhaps also the weather conditions that forces them to come. 5\ Bohemian Waxwing get nutrition mostly from flesh-bearing berries and rowan. 6\ It has never bred in Denmark. 7\ Bohemian Waxwing can raise the top feathers straight up or lay them down almost completely as it suits the bird best .. 1\ Silkehale, Bombycilla gárrulus 2\ 20-21 cm. 3\ Ret almindelig, men meget ustadig vintergæst, der kan optræde i meget store mængder. 4\ Det er næringsmangel på ynglepladserne i det nordlige Skandinavien og måske også vejrforholdene, der tvinger dem hertil. 5\ Silkehalen lever mest af kødbær og rønnebær. 6\ Den har aldrig ynglet i Danmark. 7\ Silkehalen kan rejse hovedtoppen lige i vejret eller lægge den næsten helt ned, som det passer fuglen bedst. 154 Ole Togeby Example 4: Literature (Kumbel 1958: Gruk. 15. collection) Example 4 is a typical poem, written, part of a collection, monologue, composed, permanent in eternity, asynchronous, reusable, one-to-many and exposed to the public, sophisticated form, and, for the reader, an end in itself. 4 Perspectives on verbal social acts At the first level of division the main types of texts are practical texts, factual prose and literature. This is not a result of a scholarly categorization of a set of phenomena, but a psychological reality for the members of the speech community. Because of their, conscious or unconscious, categorization of a given text message they will participate in different types of social practice with different purposes and frames of interpretation, viz: practical acts, theoretical acts or rituals (Kant 1793, Habermas 1971, Berge 1988). In this field the terminological mess is enormous. Here I have chosen the three terms practical acts, theoretical acts and rituals, to denote approximately the same notions that are used by other authors under other names. They correspond respectively to what Kant calls, praktische Vernunft, theoretische Vernunft and Urtheilskraft; to what Berge calls strategiske, kommunikative handlinger and symbolske handlinger; to what Searle calls directive and com- 1\ DUER OG SLANGER 2\Udkast til begyndelse til bordtale. 3\(ikke holdt.) 4\ Folk burde være som skriften forlanger: blide som duer og kloge som slanger. 5\ Ak, mange ærede herrer og fruer er blide som slanger og kloge som duer. 1\ PIGEONS AND SERPENTS 2\ Draft of the beginning of an after-dinner speech. 3\ (not given.) 4\ People should be as Scripture demands: gentle as doves and wise as serpents. 5\ Alas, many honourable gentlemen and ladies are gentle as serpents and wise as doves. A Model of Text Types and Genres 155 missive (two types that here are united into the category regulatives), constative and expressive speech acts, and to what by Bühler is called Appel, Darstellung, Ausdrück. Receivers of text messages in a communicative situation will expect that Grice’s principle of cooperation is in force. It says (to the sender of the message): “Make your contribution such as it is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.” (Grice 1975). It means that it is necessary for cooperative communication that the subject matter (the topic) and the form of the text fit into the types of acts accessible to the role and the position of the sender in the communicative situation - like a key fits in a lock. All communicative acts involve the ideals of truth, justice and beauty, but in the various text types only one of the ideals is emphasized as the dominating purpose or direction of the talk exchange. In factual prose the main purpose is to find the truth about the subject talked about; in a practical text the main purpose is to make a social act that is legitimate and just at the stage of interaction, and the main purpose of literature is to produce a textual form that pleases the audience. The receivers may cooperate by tending to interpret subject matters and linguistic forms of the text as being means to achieve the conceived purpose of the exchange in the situation. The Gricean principle is in this way a formulation of collective intentionality or the rationality of cooperation, stating how purpose, situation, subject matter and textual form are expected to fit together in a talk exchange, i.e. in a communicative situation. Fig. 3: Types of acts F ACTUAL P ROSE Representations of states of affairs P RACTICAL A CT (justice) COMMUNICATIVE A CT (truth) R ITUAL A CT (the beautiful) 156 Ole Togeby Example 2 can be seen and analysed as an communicative act negotiating social values and as an example of a persuasive macro act performed through a communicative question on the dimension of social practice, as a contribution to a debate with symmetric relations among the participants in the political public. The prominent function of example 3 is that it should be a true constative account of what is known about the Bohemian waxwing, it is performed in a situation with complementary positions of enlightenment, the subject matter is natural science, and the form is an expository macro act. Finally example 4, is predominantly seen as a ritual expression (the poem of the day) of reflexions on a whole world view, and revealing the craftsmanship of the poet, performed as a piece of artwork in front of the public, and most importantly in a significant form that does not comply with Grice’s maxim of informativity (Grice 1975). In practical texts the positions in the communicative situation is the most important factor influencing the text, and the criterion on which texts belonging to the text type are subdivided. In factual prose it is the subject matter, and in a literary text the linguistic form. perspective text positions in the communicative situation subject matter or type of topic type of form example 2. (practical text) debate between equals social practice argumentative macro acts example 3. (factual prose) enlightenment of the learner by the teacher knowledge about theoretical issues expository macro acts example 4. (literature) artwork artist and public world view and human affaires figures of form: rhythm, rime etc. Tab. 2: Predominant function (shown in gray) of the three examples 5 Levels in making a whole text act A communicative act has many layers of conditions of satisfaction (Recanati 2004, Searle 1969, 2010). The condition of satisfaction of a speech act depends on it’s illocutionary force concerning the direction of fit, which can be regulative (commissives or directives), constative, expressive or declarative. In addition to the illocutionary force indicator a full blown speech act comprises a proposition consisting of a reference act and a predication act. The condition of satisfaction of these two acts are felicity and truth respectively. A Model of Text Types and Genres 157 When two or more sentences are combined and together constitute a coherent verbal sequence (paragraph), it has in my view an additional condition of satisfaction, viz. to achieve the communicative purpose of the sequence of sentences, a purpose that can be expository, explanatory, persuasive, evaluative, normative, or narrative. It is impossible to analyse the purpose of a whole paragraph or text as one illocutionary intention or as one speech act, because each of the sentences has its own illocutionary force and direction of fit. So example 2 sentence 3\: Liberal tax spokesman said in Information (February 15) that the Danes do not fear tax increases any more, and this is due to “the success of the tax freeze”, contains a quotation from the liberal tax spokesman and is as such a constative speech act, while 4\ A second possibility could be that Danes feel the impact of the tax freeze in nursing homes, hospitals, schools, institutions and many other places in Denmark is a description of the possible causes of what the liberal tax spokesman stated, and as such an explanation of the missing fear. That means that it is a constative speech act too. As part of an argumentative verbal macro act 4\ is the conclusion, and 3\ a premise for this conclusion. The overall communicative purpose of example 2 , the message of the whole text, is therefore expressed by sentence 4\, the other sentences being communicatively subordinated this text message. It is necessary as well to make a distinction between the communicative purpose of a paragraph and the social function of a complete text, that normally consists of several paragraphs and additional paratexts (Genette 1997). A text act is satisfied if the form and subject matter of the text fit into the structure of social relations in the communicative situation, like a key in a lock. The structure of the social situation is determined by the relations between the interlocutors as either symmetrical or complementary, or neither symmetrical nor complementary. Performing one social verbal act is in this way described as performing five partial acts: a reference act, a predication act, an illocutionary act, a (paragraphing) verbal macro act and a (situating) text act (Wentzel 1981). The minimal unit of verbal communication is a speech act performed by the utterance of a sentence. A sentence is made by performing a reference act and a predication act, which are satisfied if felicitous and true. The speech act is satisfied by its direction of fit. A paragraph made up of coherent combination of sentences is satisfied by the achievement of the communicative purpose, and a complete text is satisfied if it’s form and subject matter fit the social and institutional conditions of the situation. 158 Ole Togeby Level of form Level of act Condition of satisfaction text text act (situating) form and subject fit into situation (with symmetrical or complementary positions of the interlocutors sequence of sentences (paragraph) verbal macro act (paragraphing) expediency of communicative form (expository, explanatory, persuasive, evaluative, normative or narrative) sentence speech act (illocution) direction of fit (world-to-mind, mind-toworld, both, or none: regulative, expressive, constative or declarative types) noun phrases and predicates reference act predication act felicity truth Tab. 3: Levels and condition of satisfaction 6 The definition of the notion of text type (or genre) In the definition of ‘text type’ or ‘genre’ (which are taken to be synonymous) all three viewpoints has to be incorporated: THE SOCIAL PRACTICE under which it is uttered, the subject that it is about, and its linguistic form. A text type can be defined in the following way: A text type is a historical evolved type of text with collectively accepted standards for subject matters and textual forms by which the sender and the receiver most expediently may achieve their collaborative purpose in types of social practice. This definition presupposes that sender and receivers when cooperating have a common direction of their verbal exchange. Of course the purpose of every text is unique, but what sender and receivers do have in common in relation to a given text is a typification of social practices, the typification of discourse subjects, and a typification of textual form. In later versions of Searle’s speech act theory (Searle 1996, pp 24, 127) this common typification of social practice, subject and textual form are called the collective intentionality background of the interlocutors in a verbal exchange. To sum up: Communicative acts are at the first level divided into speech acts that are performed by uttering one sentence, and text acts performed by uttering several combined sentences in a monologue. Text acts are subdivided in three main types: practical texts, factual prose and literature differentiated on many dimensions, such as monologue or dialogue, means to an end or an end in itself, oral or written, long or short, fleeting or permanent, used up or reusable, addressed or to the general public. A Model of Text Types and Genres 159 At the same time any text can be described in three dimensions corresponding to its function in the type of situation in which it occurs, the subject it deals with and its typical form. It is necessary to make a distinction between four levels in verbal communication: the level of reference and predication, the sentence level, the paragraph level and the text level. The conditions of satisfaction are respectively: felicity and truth, direction of fit, expediency of the communicative form and contextuality. 7 Types of illocutionary force of an speech act A full blown speech act consists in an illocutionary force with conditions of satisfaction and a content that is normally a proposition with truth value. The speech act is made in accordance with constitutive rules of the form X (the linguistic form) counts as Y (the institutional fact resulting of one common interpretation by the interlocutors) in context C (which involves both the situation and the backgrounds of the interlocutors) (Searle 1969, 1975, 1976, 1979, 1996, 2010.) The direction of fit in a constative statement goes from mind to world, that is: it is the condition of satisfaction that the proposition is true about the world. In a regulation, e.g. a promise or an order, the state of affaires in the real world should be made corresponding to the propositional content of the speech act, and the direction of fit goes from mind to world. A declaration of an institutional fact, such as a birth certificate, has both directions of fit: mind-to-world and world-to-mind, and in an expression of mental states there is no direction of fit. mind to world no fit of the mind world to mind declaration of soc. fact regulation of behaviour no fit of the world constative speech act expression of mental states Tab. 4: Types of speech acts and their directions fit The class of regulations includes both what Searle calls directives (hortatory texts), and what he calls commissives; they are taken to belong to the same class because they share the property of solely mind-to-world direction of fit, in contrast to the other three speech act types. 160 Ole Togeby 8 Types of situations Cooperative communication implies that a text (with it’s form, subject and social purpose) fits into the type of interaction that takes place in the social institution in which it is performed. The typification of social situations in institutions and organisations is founded on the relations between the participants in these situations, relations that at an abstract level can be characterized as either symmetrical or complementary positioning. And the communication either deals with the knowledge or the power of the participants (Watzlawick et al. 1968, Wentzel 1984). That gives four types of social relations between interlocutors in institutional situations: 1. Symmetric knowledge in a DEBATE about evidence and facts, reasons and opinions, found in scientific and legal discourse in particular and in the cultural public in general, performed mainly by persuasive macro acts, e.g. example 2. 2. Symmetric power in NEGOTIATION about offer, evaluation and agreement in market institutions, performed mainly by evaluative and normative macro acts, e.g. example 1. 3. Complementary knowledge about facts and norms in ENLIGHTEN- MENT and teaching, found in colleges, schools and journalism, performed mainly by expository and explanatory macro acts, e.g. example 3. 4. Complementary power in GOVERNMENT , management and ruling found in state institutions and private firms, performed mainly by normative macro acts, e.g. a work instruction, an EU directive or a law. In addition to these four types of social relations in communication we find a fifth one, viz. 5. W ORKS OF ART in which the relation between the artist and the audience is based neither on power nor knowledge, but on quality and pleasure, and this relation is neither symmetrical nor complementary, but optional and often pretended. Artwork is found everywhere as islands of refuge in the stream of social life. Verbal art is most often performed by narrative, evaluative and expository macro acts, e.g. example 4. A Model of Text Types and Genres 161 Communicative Positioning Knowledge Fantasy and Pleasure Power Symmetric debate politics, science - argumentative and persuasive macro acts negotiation sales offer, agreement - evaluative and normative macro acts Play Refuge from social life art refuge and play - narrative and expository macro acts Complementary enlightenment teaching, journalism - expository and explanatory macro acts government law, rules, directive, regulation - normative macro acts Tab. 5: Types of positioning relations in communication 9 Types of discourse subjects In any society discourse subjects are typified according to the institutions and communication communities in which they are prominent and frequently used (Fairclough 2003, Swales 1990). In the public sphere of Danish postmodern society discourse communities held together by their common subject matters can be described as: everyday conversation religious discourse political discourse journalistic discourse management discourse market discourse educational discourse scientific discourse, subdivided into natural science social science arts philosophy 162 Ole Togeby For each discourse community conventions and norms have developed through history with respect to predominant illocutionary force, to the typical positioning of the interlocutors, and to the formal properties most often used, e.g. verbal macro acts, types of information, types of reference, types of predication, level of abstraction, types of connections between the sentences. It is an explicit norm that political discourse (e.g. a letter to the editor, example 2) should be dominated by constative speech acts, should be exposed in the section with letters to the editor in the newspaper of the day, and be performed in a situation in which the speaker has a relation of debate to the addressees, and should have the linguistic form as a persuasive, evaluative or sometimes normative verbal macro act with historical references, logical and syllogistic connexions, and both subjective and objective predications. In educational discourse about natural science (e.g. a popular science text about ornithology, example 3), on the other hand, it is expected that it is published in a book, that the illocutionary force is constative and the macro acts are expository, with generic references, no presuppositions, logical connexions and only objective predications. It is not a universal regularity that texts should be like that in such situations and speech communities, but it is a historical fact that the text types have developed like that in the history of the structural transformation of the public sphere (Habermas 1962). The historical development have formed our expectations to the patterns of predominant illocutionary force, positioning of the interlocutors, subjects and form of the text types in our world. In other historical eras with other economic structures, other social conditions, other institutions, and other ideologies, the available text types and textual forms were different. In classical rhetoric the main text types were types of speeches, viz: deliberative, forensic and epideictic. The deliberative (political) speech was performed in the Senate, it dealt with the future, had the purpose to exhort or dissuade legislation or political actions as good or unworthy, advantageous or disadvantageous. The forensic (judicial) speech was performed in the Forum (in court), they dealt with the past, had the purpose to accuse or defend in order to do justice to a defendant. The epideictic speech was performed at ceremonial occations, it dealt with the present and had the purpose to praise or blame virtues and vices (Burton 2007). 10 Types of textual form: verbal macro acts The formal variations of a whole text are multifarious; all of them cannot be dealt with here, but many of the features that can vary in a text can be summarized under the concept of ‘verbal macro act’. It is not just ‘style’ because besides all stylistic features of language it takes into account the composition of the sentences into a coherent sequences of sentences and paragraphs with a A Model of Text Types and Genres 163 textual aim. And it has many names: text form, form of presentation, writing act; here the term verbal macro act is chosen (Werlich 1975, 1976). Words that denote particular macro acts are often used when talking about texts and text types, especially in writing teaching and instruction, e.g. description, account, analysis, explanation, reasoning, argumentation, discussion, assessment, recommendation, comment, criticism, instruction, manual, directive, story, report. They are not denoting whole texts but paragraphs in a text, which is composed of several sequences of sentences linguistically formed as different macro acts. The number of names of macro acts is above 50, so the question is which main types of macro acts that are relevant in the description of text types and genres. Here it is suggested that the relevant major groups of verbal macro acts are: expository, explanatory, persuasive, evaluative, normative and narrative macro acts. Expository macro acts With expository macro acts the speaker takes a communicative position as one who knows something and enlighten the addressees who don’t know, but to whom it is relevant. Expository macro acts are considered most neutral, as merely forwarding information and facts. Examples of expository macro acts are: present, describe something seen or heard, take the minutes of a meeting, give a summary of something said or written, account for concepts, taxonomies, and meronymy; generalize (describe generically), document something, exemplify, inform, provide information, report, paraphrase, recount. Explanatory macro acts With explanatory macro acts the communicative positions are complementary, the speaker being not necessarily one who knows more than the addressees, but one who has understood the subject matter better, points out causes and effects, reasons ande consequences and can define it, make it plain, and enlightens or teaches the addressees by making the issue comprehensible for them too. Examples: explain, define, give reasons, indicate causes, processes and effects, analyse parts and aspects, prove by deduction, interpret the meaning of something verbal, reflect on, put into a perspective, reason about. Persuasive macro acts With persuasive macro acts the communicative positions of speaker and addressees are symmetrical with respect to power, both parts participating in a discussion having opposite viewpoints and opinions. Persuasive macro acts have all types of state of affaires as subjects, but predominantly logical and syllogistic connexions. 164 Ole Togeby Examples: argue a controversial stance, stating reasons for and against something, conclude something, express contradiction, discuss problematic issues, debate. Evaluative macro acts With evaluative macro acts the speaker is positioned as one who, facing the addressees in conditions of negotiation or management, is competent to assess and ascertain the amount of value of the case. Evaluative macro acts are texts where the predominant predicate type is subjective. Examples: assess, comment on, judge the value of, evaluate, weigh the pros and cons, express attitude and emotion, compare two things, interpret impact of something, criticize, estimate, express opinion, attitude, feelings and emotions, find faults with. Normative macro acts With normative macro acts the speaker is positioned as the ruler of a management situation. Normative macro acts deal with recipient behaviour, either specifically (instructions) or generically (directives, laws). Most often normative macro acts have a regulative illocutionary force, and no unstated types of information. Examples: instruct behaviour of the addressees, a recipe, order, prescribe, legislate, direct, control, rule, govern. Narrative macro acts In a narrative macro act the speaker is positioned complementary to the addressees with respect to knowledge, as one who says that something happened. It involves specific historical references (whether pretended like in fiction or not), temporal and causal connexions between the recounts of particulars of an act or occurrence or course of events, recounts that give addressees the opportunity to feel like being in the situation talked about (Walsh 2007, Phelan 2007). Examples: tell or narrate about episodes, incidents and events, report or recount on historic events, tell something fictitious, recall, witness or testify what happened. Again it is necessary to emphasize that the groups are not Aristotelian classes, but radial categories with fuzzy edges, central and peripheral examples denoting cultural kinds from different domains of social life. Many of the words used about verbal macro acts are not clearly about texts, but denote cognitive skills too, e.g. evaluate, criticize. The types of macro acts are best illustrated by a figure with 6 dimensions and neighbouring and overlapping examples of macro acts. A Model of Text Types and Genres 165 Fig. 4: Verbal macro acts To sum up: All speakers have a purpose for conveying information to the addressees in the given communicative situation, about a particular subject, in a certain linguistic form; so purpose in the social situation, subject matter, and linguistic form are dimensions of any text. Depending on which dimension is dominating texts are at the first level divided into three main types respectively: practical texts, factual prose and literature. The three main text types are subdivided, each by its own criterion: practical texts according to the discourse community and the social relations between the interlocutors in the communicative situation, factual prose according to the subject matter, and literature according to the form of the text. EKSPLANATORY EVALUATIVE law rule discussion instruction argumentation manual reasons appraisal evidence evaluation reasoning estimation explanation critique, criticism essay interpretation comment story news tale account narration report description statement 166 Ole Togeby Fig. 5: Synoptical model of the main text types, their predominant properties and their subdivisions 11 Types of information Formally verbal macro acts can vary on many dimensions. The most important will be mentioned here, viz: types of information, types of reference, types of predication, and types of connexions. A piece of information is here taken to be the meaning of what can be stated in one simple proposition having truth value. In the sentence An Aarhusian is driving on the highway E-45 with his wife two pieces of information independent of each other are conveyed: ‘An Aarhusian is driving on the highway E-45’ and ‘he is together with his wife’. F ACTUAL P ROSE Representations of states of affairs Natural science Social science Arts Philosophy A Model of Text Types and Genres 167 Example 5: A joke In this joke the various pieces of information are delivered in many different ways (Sperber & Wilson 1986, Youle 1996): Some are stated explicitly: ‘An Aarhusian is driving on the highway E-45’, ‘at E-45 north of Skanderborg there is a ghost driver’, ‘a ghost driver is a driver running in the wrong track’, ‘the Aarhusian says something’. Of the explicitly stated pieces of information some are necessary for cohesion and propulsion of the story: ‘An Aarhusian is driving on the highway E- 45’, while others are PARENTHETICAL because they are unnecessary for the final message of the text: ‘a ghost driver is a driver running in the wrong track’. This piece of information is just about the meaning of the words. A ghost driver is per definition a person running in the wrong track. Among the stated necessary pieces of information some are ASSERTED by the speaker as being true or false: ‘An Aarhusian is driving on the highway E- 45’, ‘the Aarhusian says something’, while others are just mentioned without any claim of truth: ‘there is a ghost driver on E-45’, ‘all the road users but one is driving in the wrong track’. The narrator need not believe neither that there is one nor that there are many ghost drivers on E-45, only that someone in the radio and the Aarhusian say so. In a story like “On the motorway” a lot of implicit information is conveyed too, e.g.: ‘motorway E-45 runs from Aarhus to Skanderborg’, ‘the Aarhusian has a wife’, ‘there is a radio in the car’, ‘the Aarhusian is answering to what is said in the radio’, ‘the Aarhusian is driving in the wrong track’. Some of the implicit information is PRESUPPOSED by manifest material in the text and supposed by the speaker to be given information because it is known and accepted by the addressees, e.g. the definite noun phrases his wife and the radio, presuppose undeniably in the text that ‘he has a wife’, ‘there is a radio’. Other pieces of information are conveyed effectively as new to the addressees as a CONVERSATIONAL IMPLICATURE (Grice 1975), e.g. the very point of the joke: ‘The Aarhusian is the ghost driver’, which is not stated but implicated as something the addressees have to infer on their own. Contrary to presupposi- 1\ På motorvejen 2\ En århusianer kører på motorvejen E45 med sin kone; 3\ pludseligt lyder det i radioen: 4\ - På E-45 er der nord for Skanderborg en spøgelsesbilist 5\der kører i det forkerte spor. 6\ Så siger århusianeren til sin kone: 7\ - Det gør de da alle sammen. 1 On the motorway 2\ An Aarhusian is driving on the highway E-45 with his wife; 3\ suddenly it sounds on the radio: 4\ - At E-45 north of Skanderborg there is a ghost driver 5\who is driving in the wrong track. 6\ Then the Aarhusian says to his wife: 7\ - Just one? All of them do. 168 Ole Togeby tions, implicatures may be cancelled by an explicit remark, e.g. the narrator could have continued: The Aarhusian was right; in fact they understated the problem in the radio. To sum up there are five different types of information, viz. parenthetical, mentioned, asserted, presupposed and implicated information: Fig. 6: Types of information If presupposed information is not known or accepted by the addressees (and the speaker knows that that’s the case), it is called bullying, e.g. When did you stop beating your wife (Harder & Kock 1976). Text types and verbal macro acts are characterized by the types of information which are criterial, frequently used, optional or not expected in such a text. In educational text with asymmetric knowledge (e.g. example 3) unstated information, whether given or new, is not expected, while parenthetical, mentioned and asserted information is frequently used, e.g.: Bohemian waxwing can raise the topfeathers... In a joke implicated information is criterial, and presupposed information often used, e.g.: ‘The Aarhusian is the ghost driver’. In a political debate (example 2) unstated types of information is not recommended, but bullying is in fact often used, e.g.: ‘what the Danes feel as the impact of the tax freeze, is cuts in nursing homes, purer conditions in the hospitals ...’. And so on. 12 Types of reference acts Sentences that can function as speech acts in a communicative situation consist of a reference act, that is felicitous or not, and a predication act, that is information stated unstated unnecessary necessary given new paranthetical mentioned asserted presupposed implicated without truth value with truth value A Model of Text Types and Genres 169 true or not. In the beginning of a text the writer has to refer to entities talked about by anchoring them to the communicative we-here-now situation. The form of the references is adapted to logical status of the situation talked about and the level of accessibility of the entities for the addressees in the communicative situation. Later in the text the writer can refer anaphoricly to the same entities as previously, but in the beginning the references constitute the logical relations between the situation talked about and the communicative situation. This relation is constituted in four different ways: as a deictic, historical, generic or fictitious reference. Deictic reference: The writer refers to an entity which is an element in the communicative situation by using words such as I, me, we, us and you, here and there, this, and that (connected with pointing), now and in a moment, words that shift their meaning depending on when, where, by whom and to whom they are uttered, e.g. (example 1) There is a table here let us. These deictic reference words constitute the situation talked about as overlapping with the communicative situation and consequently the entities as evident and immediately accessible. So called performatives utilise deictic references, e.g. I hereby declare this conference open. Historical reference: The writer refers to an entity historically when the anchoring is made to something already known by the addresses such as historical facts, dates of the calender, geographical places, e.g. (example 2) Liberal tax spokesman, February 15. The situation talked about and the communicative situation are separated in time and space, but connected by a well-known link. Generic reference: When the writer does not refer to a particular, but to the any instance of a category and at any time, the reference is generic, e.g. (example 3) Bohemian Waxwing get nutrition mostly from flesh-bearing berries and rowan. It has never bred in Denmark. The situation talked about comprises encyclopaedic knowledge about reality including state of affaires of the communicative situation. The text does not only describe facts, but prescribe as well how to talk about reality by using words. Fictitious reference: The writer’s reference to entities in the situation talked about are fictitious, if they are pretended. One way of pretending is to presuppose that the addressees know the persons and times talked about, although both interlocutors know that they don’t, e.g. The barque Charlotte was on her way from Marseilles to Athens, in grey weather, on a high sea, after three day’s heavy gale. (Isak Dinesen’s short story: The Sailor-Boy’s tale 1942). The readers do know 170 Ole Togeby about Marseilles and Athens, but they don’t know anything about any barque Charlotte, the author knows that, and the readers know that the author knows it. So the reference is pretended. Another way of pretending is telling about entities (persons and objects) in indefinite forms and not anchoring them sufficiently to the communicative situation, e.g. An Aarhusian is driving on the highway E-45 with his wife; suddenly it sounds on the radio... The readers know where, but not when, and the figures in the story are not anchored to anything they know about. The situation talked about is separated from the communicative situation, and there is not established a reliable link between them. Fig. 7: Types of reference 13 Types of predicates Predicates are expressed in a sentence as main verbs, predicative adjectives, adverbials, prepositions or predicative verbal nouns. Predicates are divided into two main types: subjective and objective predicates. Objective predicates deals with state of affaires in the situation mentioned, independently of who is observing it, and subjective predicates deal with how an observer experiences something. Objective predicates can be true or false because they are observer independent and about what the speaker knows, reference real pretended to the rhetorical act relative to the rhetorical situation specific general deictic historical generic fiction A Model of Text Types and Genres 171 thinks, believes, assumes, or guesses. Often the degree of epistemological certainty (underlined) is signalled: There is a table here. Liberal tax spokesman said in Information (February 15) that ... It is a nutrient deficiency on the breeding grounds in northern Scandinavia and perhaps also the weather conditions that forces them to come. A second possibility could be that ... Subjective predicates cannot be true or false, but are sincere or hypocritical, because they are observer dependent and about how the experiencer feels, estimates, evaluates, judges some state of affaire mentioned. Yes but it’s nothing compared to what he is now. like a madman! I fear tax increases ... I feel the impact of the tax freeze ... gentle as doves and wise as serpents the success of the tax freeze Subjective predications about third person experiences are uncertain objective statement and should be made probable by some evidence. the Danes do not fear tax increases any more, and this is due to “the success of the tax freeze”. The Danes feel the impact of the tax freeze on nursing homes, hospitals, schools, institutions and many other places in Denmark. It is a quite fixed norm that expository and explanatory macro acts utilise objective predicates, while evaluations and persuasions involve subjective predicates and narrations contain both types. 14 Types of textual connections In a coherent paragraph pieces of information from the sentences are connected by explicit conjunctions, e.g. and, but, explicit adverbials, e.g. then, of that reason, consequently, afterwards or without explicit marks indicating the type of semantic relation which nevertheless is implicated. It is in example 1 implicated that there is a semantic connexion between 1\ and 2\ and another semantic relation between 3\ and 4\, but none between 2\ and 3\. Example 1: 1\ Yes but it’s nothing compared to what he is now. 2 \ Sometimes he behaves like a madman! 3 \ There is a table here 4\ let us -. Between 1\ and 2\ the semantic relation is a relation of reasoning: the fact stated in 2\ is the reason for accepting the truth of 1\. The information given in 3\ makes it possible to issue the 172 Ole Togeby mutual request given in 4\. So the semantic relation between 1\ and 2\ is a FOR -connection, and the relation between 3\ and 4\ is a MAKE POSSIBLE connection. The term CONNECTION is denoting the semantic relations between the pieces of information given in two neighbouring sentences or clusters of sentences, independently of whether these relations are expressed explicitly or just implicated because of the expectation of unity of text meaning founded on the coherence of the propositional content of the concatenated sentences (Togeby 2003). There are many different types of connections: additive, alternative, causal, etc. Here they have a notation with an explicit textual marker of the relation in CAPITALS : AND -connection, OR -connection, THEREFORE connection and so on. Some connections are used generally because they are used in all types of macro acts, e.g. AND , OR , BUT , THAT , SAID X, others are only used in paragraphs about the situation mentioned: THEN , WHILE , AFTER , BECAUSE , THEREFORE , IN ORDER TO , and the rest only used about the communicative situation: ERGO , FOR , VIZ ., I . E ., E . G ., BRIEFLY , PRESUPOSED , MAKE POSSIBLE and PUNCH LINE . Of the general connections some are logical: and, or, but, and some establish a distinction between two logical levels, viz: communication and metacommunication (Watzlawick et al. 1968): THAT , SAID X . Of the connections about the situation mentioned some are temporal: THEN , WHILE , AFTER , and some are causal: BECAUSE , THEREFORE , IN ORDER TO . Of the connections about the communicative situation some are syllogistic: ERGO , FOR , and some are informational: VIZ ., I . E ., E . G ., BRIEFLY , PRESUPPOSED , MAKE POSSIBLE and PUNCH LINE . Examples (connections in capitals inside hatches: # CONNECTION #): 1\ Yes but it’s nothing compared to what he is now. # FOR # 2\ Sometimes he behaves like a madman! 3\ There is a table here # MAKE POSSIBLE # 4\ let us - 3\ Liberal tax spokesman said in Information (February 15) that the Danes do not fear tax increases any more, and this is due to “the success of the tax freeze”. # OR # 4\ A second possibility could be that Danes feel the impact of the tax freeze on nursing homes, hospitals, schools, institutions and many other places in Denmark. 2\ 20-21 cm. # AND # 3\ Quite common, but very fickle winter visitor, which may appear in very large quantities.# BECAUSE # 4\ It is a nutrient deficiency on the breeding grounds in northern Scandinavia and perhaps also the weather conditions that forces them to come. # AND # 5\ Bohemian Waxwing get nutrition mostly from flesh-bearing berries and rowan. # AND # 6\ It has never bred in Denmark. # AND # 7\ Bohemian Waxwing can raise the top feathers straight up or lay them down almost completely as it suits the bird best. A Model of Text Types and Genres 173 4\ People should be as Scripture demands: gentle as doves and wise as serpents. # BUT # 5\ Alas, many honourable gentlemen and ladies are gentle as serpents and wise as doves. 2\ An Aarhusian is driving on the highway E-45 with his wife; # THEN # 3\ suddenly it sounds on the radio: # THAT # 4\ - At E-45 north of Skanderborg there is a ghost driver # I . E .# 5\ who is driving in the wrong track. # THEN # 6\ Then the Aarhusian says to his wife: # THAT # 7\ - Just one? All of them do. The inventory of connections encompass approximately 22 types, perhaps a couple more: Logical Meta-communicative Temporal Causal Syllogistic Informational AND , OR , BUT , LIKE , EVEN THAT , SAID X THEN , WHILE , AFTER BECAUSE , THEREFORE , IN ORDER TO , ERGO , FOR VIZ , I . E ., E . G . BRIEFLY , PRESUPPOSED , MAKE POSSIBLE , PUNCH LINE and, too, also, but, even, however, never the less, in spite of, while, like : “...” (colon and quotation marks), this, that then, next, at that time, just then, at the same time, before, previously for, so, therefore, for that reason, by means of, with the help of ergo, consequently, so, for, therefore that is, namely, viz. that is, i.e. in short, e.g. for instance, presupposed Fig. 8: Types of connections The various types of connections are more frequent in some macro acts than in others: In narrative acts temporal and causal connection predominate, in connections general situational mentioned situation rhetorical situation logical metacommunicative temporal causal syllogistic informational 174 Ole Togeby expository acts informative connections, in explanatory causal connections, in persuasive syllogistic, in evaluative macro acts meta-communicative connections and in normative macro acts logical connections. 15 Literature, fiction, epic and narrative Many disciplines make use of the concept of text type or genre, but the disciplines and scholarly schools define it differently, especially notions such as literature, fiction and narrative are mixed up. In the proposed model these concepts can be defined in the following way: Literature is not (like all other main text types) for the reader a means to an end but an end in itself. Literature is made for disinterested pleasure of the audience (Kant 1793). Literature is exchanged in the “literary institution” that is neutral with respect to symmetry and to knowledge and power, but turns readers (with interests and goals of their own) into arbiters of taste with no external goals connected to reading the text. The literary references are often pretended, thereby transforming untrue statements from lies into fictional statements that are perceived as deliberate violation of Grice’s maxim of thruth (Grice 1975). Literature often flouts the maxims of relevance and informativity, too, by consistently over-structuring the form (e.g. by mime, repetition, rhyme, rhythm, metaphor and composition) (Kock 1979). Not all fictional references form part of imaginative literature. Smal fictive narratives are also used as descriptions of prototypical examples in one paragraph among many in a text otherwise dominated by factual prose. (For other criteria of fiction, se Peter Widell in this volume) Epic literary texts are texts mainly consisting in fictitious references, narrative verbal macro acts and a pointed ending (Aristotle 1992). Texts with such a punch-line changes readers from someone seeing the text as a source of knowledge (about the world talked about) to someone seeing it as a work exposed in the communicative situation, a work that is influencing and challenging them intellectually, emotionally and aesthetically and make them reflect on truth, goodness and the beauty in general (Phelan 2007). Many narrative macro acts do not deal with fictive matters, but are testimonies of historical events. Fictionality in a narrative is signalled by the pretended type of reference used in the macro acts. References Aristotle. (1992) Poetik. København: Hans Reitzel. Austin, John L. (1962) How to Do Things With Words. Oxford: Clarendon Press. Bachtin, Mikhail. (2000) The Problem of Speech Genres in David Duff (ed.) Modern Genre Theory, 82-97. Harlow: Longman. A Model of Text Types and Genres 175 Beaugrande, Robert de, and Dressler, Wolfgang (1972) 1981. Introduction to Text Linguistics London: Longman. Berge, Kjell Lars. (1988) Skolestilen som genre. Med påtvungen penn. LNU, Oslo: Cappelens Forlag. Burton, Gideon O. (2007) Silva Retoricae. http: / / rhetoric.byu.edu. Genette, Gérard. (1979) Introduction á l’architext in David Duff (ed.) 2000: Modern Genre Theory, 210-218, Harlow: Longman. Genette, Gerard. (1987) 1997 Paratexts. Thresholds of interpretation, translated by Jane E. Lewin, Cambridge: Cambridge University Press. Grice, Paul H. (1967) 1975 Logic and conversation in Peter Cole and Jerry Morgan 1975: Syntax and Semantics, vol. 3, 41-58. New York: Academic Press. Habermas, Jürgen. (1962) Strukturwandel der Öffentlichkeit. Untersuchungen zu einer Kategorie der bürgerlichen Gesellschaft (Habil.), Neuwied: Luchterhand. Habermas, Jürgen. (1971) Vorbereitende Bemerkungen zu einer Theorie der kommunikativen Kompetenz in Jürgen Habermas and Niklas Luhman, Theorie der Gesellschaft oder Sozialtechnologie - Was leistet die Systemforschung? 101-141. Frankfurt am Main. Harder, Peter and Kock, Christian. (1976) The Theory of Presupposition Failure in Travaux du cercle linguistique de Copenhague vol XVII. København: Akademisk Forlag. Kant, Immanuel. (1790) 1793. Kritik der Urteilskraft. Hrsg. Von Wilhelm Weischedel 1996. Frankfurt a.M.: Suhrkamp Verlag. Kock, Christian. (1979) Principles of temporal form in PTL. A Journal for Descriptive Poetics and Theory of Literature, no. 4, 267-284. The Hague: Mouton. Lakoff, George. (1987) Women, Fire, and Dangerous Things. What Categories Reveal about Mind, Chicago: The University of Chicago Press. Phelan, James. (2007) Experiencing Fiction - Judgements, Progressions, and the Rhetorical Theory of Narrative. Columbus: The Ohio State University Press. Recanati, François. (2004) Literal Meaning. Cambridge UK: Cambridge University Press Searle, John R. (1969) Speech acts. An essay in the philosophy of Language. Cambridge: Cambridge University Press. Searle, John R. (1975) The Logical Status of Fictional Discourse in New Literary History , vol. 6, no.2, 319-332. Searle, John R. (1976) A classification of illocutionary acts in Language in Society, vol 5,1-2 . Searle, John R. (1995) 1996 Construction of Social Reality, London: penguin Books Searle, John R. (2010) Making the social world. The Structure of human civilization. Oxford: Oxford University Press. Sperber, Dan and Wilson, Deirdre. (1986). Relevance. Communication and cognition, Oxford: Blackwell. Swales, John M. (1990) Genre Analysis. English in academic and research settings. Cambridge: Cambridge University Press. Togeby, Ole. (2003) Fungerer denne sætning? . Århus: Aarhus Universitetsforlag. Togeby, Ole. (2010) Handling, tekstualisering og tekst in Skandinaviske sprogstudier no. 1, http: / / ojs.statsbiblioteket.dk/ index.php/ sss/ issue/ current. Walsh, Richard. (2007) The Rhetoric of Fictionality. Narrative Theory and the Idea of Fiction. Columbus: The Ohio State University Press. 176 Ole Togeby Watzlawick, Paul, Beavin, Janet Helmick and Jackson, Don D. (1968) Pragmatics of Human Communication. A Study of Interactional Patterns, Pathologies and Paradoxes. London: Faber and Faber. Wentzel, Knud. (1981) Tekstens metode. Ét synspunkt på litterære og praktiske tekster: Gyldendal: København. Werlich, Egon. (1975) Typologie der Texte. Heidelberg: Quelle & Meyer. Werlich, Egon. (1976) A Text Grammar of English. UTB: Heidelberg. Yule, George. (1996) Pragmatics. Oxford: Oxford University Press. Peter Widell The Literary Text: Four Parameters 1 Introduction This article is an attempt to shed light on the concept of the literal text using the British-American language philosopher Herbert Paul Grice’s communication-based theory of language. 1 As will be seen, the following will not be a genre theory in the traditional sense of the word. I am not seeking a sort of genreor text type-taxonomy. The theory I will present should rather be called a parameter theory of the literary text. 2 To be more specific, it is my intention to establish a set of four parameters characterizing literary texts. These parameters I will call, respectively, figurativity, tropicity, fictionality and degree of realism (or sometimes, for short, realism). The reason for calling my theory a parameter theory is grounded in the fact that it is possible for each and every text segment, literary or non-literary, ranging from the size of single speech act to the size of a whole text, 3 to ask whether it is characterized by a 1 Cf. Grice (1959), reprinted in Grice (1989: 213-222); Grice (1975), reprinted in Grice (1989: 22-40). 2 Without being identical with Gérard Genette’s concept of mode, my parameter theory is inspired by Genette’s distinction between mode and genre where mode according to Genette is a term reserved for the more universal features of texts while genre is a more complex ususally referring to historically specific text types, cf. Genette (1992 (French 1977)). 3 My parameters should first and foremost be seen as analytical concepts, i.e. as precise instruments for characterizing any text segment in any text. In the following, I will be describing my parameters exclusively in this sense. In order to avoid misunderstandings and accusations of being naive, it should be pointed out that my concept of analyticity does not imply that we cannot find borderline cases, that is cases where it is difficult to determine whether we are dealing with, for instance, a piece of fictionality or not. Of course, we can. But it implies that (a) the concept itself will always be clear and distinct (as a necessary precondition for measuring borderline cases as borderline cases) and that (b) it will - despite occurrences of borderline cases - normally be possible to find distinct and clear core cases (cf. Grice & Strawson (1957); reprinted in Grice (1989: 196-212); cf. also Searle (1969: 4-12)). Even though the parameters should be seen mainly as analytical concepts, they could also be used as genre or text type concepts. But then they will be seen merely as prototypical concepts, that is as more quantitative and circumstantial means of characterizing a literary text as a whole - as when you are judging, for instance, Homer’s epic poem The Iliad to be a figurative text because of its recurrent meter, Peter Widell 178 value on one or more of the four parameters as, for instance, a certain rhyme on the figurativity parameter, a metonymy, a synecdoche or an irony on the tropicity parameter, or a pastiche, a satire or a sort of phantasy on the fictionality and degree of realism parameters. 4 As will be shown below, Grice’s communicative language theory involves a special approach to text description. 5 Thanks to Grice it is possible both to give a precise definition of the overall distinction between non-literary prose and literature, and to highlight the numerous subordinate conceptual distinctions in genre theory - and actually in the whole field of rhetoric and stylistics in general - in light of a comprehensive, unifying and, in that sense, explanatory theory. In what follows, I will first take a look on Grice’s theory. Then I will try to give a preliminary characterization of the literary text by the four text parameters. After that, I will examine each of the four parameters thoroughly trying to find their true nature. The upshot of the examination will be that, if you have Grice in your pocket and have a thorough understanding of the concept of repetition, and if you furthermore choose a robust concept of realism, then the literary text will reveal itself as a tight system of repetitions at different levels of reality as a vibrant means for your aesthetic pleasure. 2 Grice’s theory of implicature According to Grice, all human communication is subject to certain universal norms or - as he calls them - maxims. In his article “Conversation and Logic” 6 he tries to find out what these universal maxims consists of and which function they have in human communication. The point of departure for Grice is to state the fact that communication is impossible among persons Gustave Flaubert’s novel Madame Bovary (Flaubert 1859) to be a fictional and realistic text because of its overall fictionality and realism. 4 It is possible to refer to both the parameters and their values as styles or modes of presentation. Thus, it is all right to talk about a figurative or rhymed style, or a tropic or ironic style. Note that the distinction between the parameters and their values represents a clarification of the concept of style. Because of the status of my parameters it will be quite possible to talk not just about non-literal genres versus literary genres, but also about cross or hybrid genres between the non-literary and the literary as, for instance, faction, docudrama and auto-fiction. It will also be possible through my parameters to characterize non-literary texts, for instance political speeches, namely as more or less influenced by different literary styles (without, actually, being literary texts). Actually, it will even be possible to use my parameters to help pinpointing what we in normal non-literary prose intuitively identifies as good or bad writing. 5 In the following, I will - as a matter of convenience - refer to both texts and text segments as texts. 6 Grice (1975); reprinted in Grice 1989: 22-40. The Literary Text: Four Parameters 179 who directly opposes each other. This means, according to Grice, that persons who communicate must be obeying a principle that ensures that the speaker’s intention, in saying what he says, is understood by the hearer. This principle he calls the cooperative principle. 7 Grice realizes that this principle can only be complied with if the speaker tacitly follows exactly four rules or - as he, inspired by Immanuel Kant’s table of categories in Kant 2007 (1781): A80/ B106, calls them - four maxims for successful communication. The maxims - which Grice calls, respectively, again with a loan from Kant, the maxims of quantity, quality, relation and manner - are (here in a slightly modified form): 1. Do not say too little or too much! 2. Speak the truth (or at least do not say what you know is untrue)! 3. Be relevant when you speak! 4. Speak clearly and distinctly (i.e. grammatically and semantically correct)! Others have more tellingly called them the maxims of informativeness, truthfulness, relevance and correctness. 8 It is crucial to understand that these norms 7 The principle states: “Make your conversational contribution such as is required, at the stage at which it occurs, by the accepted purpose or direction of the talk exchange in which you are engaged.” (Grice 1989: 26). 8 Grice does not formulate this fourth maxim quite the same way as I do. He instead in his formulation talks about being orderly and understands it in a much looser sense. In opposition to Grice, I want to tighten the maxim in such a way that it will be strictly related to meaning, and meaning again strictly related to truth. Actually, I want to see (propositional) meaning in such a way that it is directly connected to the truth conditions of the corresponding sentence, where these truth conditions should again be understood as the situations in the world that will make the sentence true. Or to flesh it out by an example: When I use the English sentence “The snow is white”, I am using it according to the maxim of correctness if what I say is a grammatical English expression (which it, actually, is), and if it is used with a knowledge of how to use it to assert true propositions in situations where we have instances of white snow. Notice, that it does not mean that the sentence have to be true in the actual situation. There has been a lot of discussion about whether the four maxims are on the same level, and whether they should be applied in a certain order. Searle writes: “[…] the four are not on a par. For example, the requirement of truthfulness is indeed an internal constitutive rule of the notion of a statement. It is a constitutive rule of statement making that the statement commits the speaker to the truth of the expression expressed. […] But the other Gricean features are not like that. The standards of relevance, brevity, clarity, and so on, unlike truth, are not in that way internal to the notion of speech acts. They are all external constraints on the speech act coming from general principles of rationality and cooperation.” (Searle 2002 (1992): 185). I agree with Searle, but - as indicated in my definition of correctness above - only partly: for me, correctness is as constitutive of speech act meaning as truthfulness is of propositional truth. Peter Widell 180 are not just (collectively) universal. But that they are, actually, as I understand them, transcendentally co-present in each and every speech act. This transcendental argument has, of course, to be defended against any scepticist. Unfortunately we do not have enough space to defend it here. But we can at least show the presence of the maxims in an example chosen at random: if I, for example, utter the sentence “The earth is round” as an answer to a question from the hearer about whether the earth is round or not (in an everyday context; scientifically speaking it is, of course, not round), my speech is both informative - the hearer gets precisely the information he needs - truthful - the earth is round - relevant - my utterance corresponds precisely to what the hearer seeks - and grammatically and semantically or, what we could call it too, literally correct - the words I am using are the words you are supposed to use in English if you want to say that the earth is round in that language. Now, what is interesting and original about Grice’s theory is not that he was the first to find and justify these four - as we can call them - transcendental maxims governing understanding-oriented communication. Several other philosophers have discovered roughly the same maxims - or at least what can be cast in a similar form. The interesting and original aspect of Grice’s theory is rather his elucidation of how we use them when communicating. What he has discovered - and deserves special credit for - is that we can have rational reasons for not living up to them. There can be various reasons why the sender does not always live up to the four maxims. For example, he may lack the ability to live up to them, e.g. because he is a child, or because he is tired or drunk, or he may lack the will to do it, e.g. because he wants to cheat the person he speaks with. None of these reasons, however, are reasons that serve to facilitate communication. Cheating requires, indeed, that you are able to make rational planning. But the purpose here can only be non-communicative: when cheating, the goal is precisely to hide your intentions, not to bring them forth. Grice, however, discovers a kind of reason for not living up to the four maxims which not goes against communication but, actually, serves it and makes it more effective. It follows from my definition of correctness that complying with the maxim of truthfulness normally presupposes that the maxim of correctness is followed. As to informativeness and relevance, I consider these two maxims as related not so much to assertion and to other truth-evaluating acts, as to their instrumental embedding - giving us all sorts of perlocutionary acts (as the founding father of speech act theory, John L. Austin, calls them; cf. Austin 1962). That does not mean that they are of minor importance for conveying the content of what the speaker or author is trying to communicate to the hearer. But their function is merely regulative (cf. Searle (1969: 33-42) for a discussion of the distinction between constitutive and regulative rules). In the following discussion, I will - unless it is explicitly indicated otherwise - tacitly presuppose the use of my modified Gricean maxims. The Literary Text: Four Parameters 181 What kind of rational reason we are talking about can perhaps best be illustrated by an example borrowed not from Grice, but from the American philosopher and colleague John R. Searle: 9 when A in a restaurant wants the salt and turns to B with the words “Can you pass me the salt? ” in order to get B to pass him the salt, A deliberately breaches, according to Searle, at least two of the four Gricean maxims. First, it is not relevant in the situation to ask B about his ability to pass him the salt; here, the relevant thing to do is to request B to do what A wants him to do. Secondly, it is not informative, either, in the situation to ask about B’s abilities: to have the ability to pass the salt is indeed a prerequisite for passing the salt, but it is not a sufficient condition. Nevertheless, B will immediately, according to Searle, take what A says, as a request and henceforth as an example of successful communication. But how can he do that? How can he consider something which is basically flawed and defective successful? According to Searle (and Grice), the reason why A’s request is successful shall be found in the fact that A’s defective utterance prompts or initiates B to make a certain sequence of inferences. In the example, the sequence can, according to Searle, be seen as the following of five distinct inference steps: 1. By saying “Can you pass the salt? ” A is not living up to the maxims of relevance and information. 2. B assumes that A acts cooperatively (because A normally does so). 3. Therefore, A probably says more than what his words literally means. 4. A’s question can in the given situation be viewed as a question about whether B has the ability to pass him the salt. 5. Ergo: probably, A will ask B to pass him the salt. That means: through the inference steps in (3) and (5), B is able to transform A’s defective utterance to an utterance which - had it been explicitly formulated - would have been consistent with the cooperative principle and with Grice’s maxims. The reason for operating with defective utterance, is in the effective Gricean sense that they can be used by the speaker to initiate a particular type of situation-bound, implicit inference by the hearer, in the present example an inference that goes from A’s explicit remark “Can you pass me the salt? ” to his implicit request for the salt. This pattern is not only connected to indirect speech acts, as in our example (cf. note 7), but is quite general. It can be spelled out in the following five steps: 1. The speaker says something that does not meet the cooperative principle by deliberately breaching one or several maxims. 9 Cf. Searle (1975), reprinted in Searle (1979b: 30-57). The shown example is an example of a special sort of implicature, discovered, described and dubbed “indirect speech act” by Searle. Peter Widell 182 2. B assumes that A acts cooperatively (because A normally does so). 3. Therefore, the speaker (probably) means more than he literally says. 4. The literally said paired with the what is going on in the speech situation, makes it possible to understand what is the speaker says as an utterance which - had it been explicitly formulated - would have satisfied the cooperative principle. 5. Ergo: the speaker has (probably) meant utterance Y. The explanation of the fact that an utterance can be successful even if the words said do not meet the cooperative principle, is thus: the speaker intends with his defective utterance to initiate an implicit inference-pattern in the hearer. Such a speaker-induced inference-pattern has Grice - alluding to the concept of implication - dubbed an implicature. 10 The core of Grice’s theory and the source of its originality shall be found in the special double use of the cooperative principle found here: first, the principle is brought to use to identify a defect in the form of a violation, deliberately breach or flouting of one or more of the above mentioned maxims. Then, secondly, it is brought to use to produce the implicatures which, if they were explicitly formulated, would have satisfied the maxims. As to the effectiveness of the Gricean implicature, it can readily be seen from the example that implicatures helps speeding up communication: instead of doing everything explicitly in an attempt to live up to the cooperative principle, the speaker leaves it to the hearer to draw the appropriate inferences. And since thought is faster than words, there is a communicative profit to be gained by using implicatures. This profit is, according to Grice and Searle, the main reason why we use implicatures in the first place - and, in fact, implicatures occur in large numbers whenever we communicate. There is also a risk in using implicatures. When the speaker leaves it to the hearer to guess the intended meaning of his words, there is obviously a greater danger of communication failure than if the speaker is behaving in full compliance with the cooperative principle. This risk is outweighed, though, by the fact that there are other advantages by using implicatures - also besides the purely economic benefits. Many of these benefits are of a more specific psychological nature. 11 10 Grice distinguishes between two kinds of implicatures, conventional and conversational implicatures, and among the conversational implicatures again between particularized and generalized implicatures. Without going into discussion of these subtypes, I will confine myself in this paper to speak only about the most common subtype, the conversational and particularized implicature. While the name is somewhat prickly and quite opaque in the context of what we will be discussing in this paper, I will simply call this subtype an implicature. 11 To mention just one example of a more specific psychological advantage: it would be unbearable for the individual participants in a communicative engagement - The Literary Text: Four Parameters 183 Grice’s cooperative principle has also its aesthetic advantages. As we have seen, implicature-based speech requires work from the side of the hearer: the hearer has to infer what the sender can mean by the words he expresses, but to draw such inferences can be challenging and demands a sort of mastery, and mastery is, actually, aesthetically pleasing. But that is exactly where literature comes into play. In literature, it is precisely this possibility of using vague implicatures that are exploited for aesthetic purposes. 12 3 The literary text A literary text part is, to state my thesis briefly, a demonstration of four different ways of deliberately breaching the Gricean maxims. Generally seen, the definition is, as far as it goes, quite simple: while the non-literary prose text can be defined as a text that meets the maxims of informativeness, truthfulness, relevance and correctness, the literary text part can, conversely, be defined as a part where the speaker or writer deliberately violates one or more of these maxims in order to achieve an aesthetic advantage. The last words are crucial, because they point to a special form of these violations. Violations can also be found in non-literary prose-texts. But they are normally easily recovered by a few fixed inferences, as we have already seen demonstrated in our example with the salt. This is not so in the literary text. In literature, the texts and thus a disadvantage for the long-term maintenance of good communicative relations - if the participants strictly complied with the cooperative principle. It is important to have a certain space for negotiation among the conversation partners, a sort of interpretation gap where they can respect each other’s dignity as persons and maintain each other’s face as the American sociologist Erwin Goffman calls it, cf. Goffman (1955). But that is what implicatures can secure for us. Implicatures secure a sort of communicative politeness as Brown & Levinson has dubbed it (cf. Brown & Levinson 1986). This restraint on our communicative engagement Donald Davidson has tried to coin in a genuine communication ethical principle called the Charity Principle (cf. Davidson 1974). The principle sounds in all its simplicity (slightly modified compared to the original): remember always as hearer to be merciful to the speaker and try whenever you do not immediately understand his utterance to see it as an attempt to use implicatures. However simple this principle is, it has, actually, proved to be crucial for the long term insurance of the founding conditions of communication. 12 This aesthetic purpose should not be confused with aesthetic experience. Aesthetic judgment involves purposeful activity, but is in itself pointless - or without interest, as Kant says (Kant 1952 (German 1790)). It is a judgment of taste which only serves - as Kant also puts it - the disinterested wellbeing: “Does what I am doing, taste good, or does it taste bad? ” Or as Wittgenstein has expressed it - emphasizing the outer signs of the aesthetic judgment we all, experiencers as well as observers, share: “Do I feel like this”: “or like this”: (cf. Wittgenstein 1978 (1966): 4). . . . .. . . . . . Peter Widell 184 parts are dominated by implicatures that require a lot additional inference work from the side of the hearer - or in written texts: from the side of the reader. 13 In this respect, the literary implicatures can be characterized as more loose 14 than, for instance, the implicatures in indirect speech acts. They represent a considerably larger amount of “correct” interpretations, i.e. possible inferences from the side of the reader, and it is exactly here we can find what gives the literary text parts their aesthetic advantage or function, as we will call it. Since a literary text part cannot as easily be put to rest on the narrow path of non-literary prose, it can in a sense be said to operate parasitically on the non-literary prose text. 15 What lies in this metaphor is simply that the concept of the literary text part presupposes the concept of the ordinary non-literary prose text part (in order to deliberately breach with the non-literary prose text part’s standard conditions), while the concept of the non-literary prose text part, conversely, contains no concept of literary text part. This can be stated in the following radical thesis: whatever breaches with a non-literary prose text we will find in a text, this type of text will always represent a permanent methodical reduction of the text from the domain of the literally expressed to the domain of inference based instrumental action. This gives us the opportunity to see more clearly the connection between the implicatures of the literary text and its aesthetic function. Aesthetics is invariably linked with instrumental activities or results from instrumental activities - artifacts - as it is in painting or music. First and foremost: aesthetic reactions are - as Kant puts it - not reactions to the world directly, and - one might add - not, either, to the world-oriented functions of linguistic expressions. Instead, the aesthetic judgments are, in Kant’s words, reflective. 16 That is just another way of saying that the aesthetic judgment is a response to our instrumental activities dealing with things in the world. Furthermore, the judgment is always an emotional judgment. It concerns our feelings toward our activities. And this is where the concept of mastery, mentioned above, comes into play: if what we are doing involves a certain risk, then the skill to master it normally turns out to be a source of aesthetic pleasure. That means, in relation to texts, that we find aesthetic relevance not so much in the literal parts of the text where the cooperative princi- 13 Every text can be either an oral or a written text. It means that we can meet speakers and hearers or listeners as well as authors and readers. Since modern literary texts present themselves mostly in written form, I will mainly use the terms authors and readers in the following. It is important, though, to be aware of the fact that on the general level I am discussing in this article, it doesn’t matter which medium - oral or written - you choose. 14 Cf. Sperber & Wilson (1986). Cf. also Blakemore (1992: Ch. 7-9). 15 Cf. Austin (1962: 51-52 & 127). 16 Cf. Kant (1952 (German 1790): 18). The semantically based judgment is equivalent to Kant’s determinative judgment. The Literary Text: Four Parameters 185 ple is complied with, but in those parts where the principle is violated in various ways, and the capability to form and to draw inferences is challenged. Here, we will see that the principle can be challenged in two different dimensions of the text: in the textual expression - here we find the domain of the figurativity of the text - and in the textual implicature - here we find, respectively, the domain of tropicity, fictionality and realism in the text. The question of whether a text part is literary in the formal sense is not to be confused with the question of whether it is a text part with literary qualities, i.e. whether it is a good or a bad literary text part. But even if it does not belong to the concept of literature that it is good or bad, it remains, nevertheless, a feature of any literary text part that it is a potential object of aesthetic evaluation. But what does it mean? Do we have reliable standards of evaluation? Or are aesthetics just a matter of taste? Kant has convincingly tried to show that although it is meaningless to evaluate individual art works as good or bad based on some absolute scale - here we humans can only act recommendatory according to Kant if we want to influence others to a certain aesthetic attitude - the aesthetic judgment is, nevertheless, not subjective. It is, actually, as Kant asserts, possible as a generalization over form on the one hand and aesthetic appreciation on the other hand. It is here we find the judgement of the beautiful as a judgement of a free play between imagination and understanding 17 or - to skip the rather special Kantian terminology and choose more modern terms - between being challenged in our instrumental activities on the one hand and showing mastery of those instrumental activities on the other hand. 18 This also seems to hold in literature. Here, literarity seems to retain a certain aesthetic value across various literary orientations, schools, and epochs. This applies to figurativity to tropicity, to fictionality and to literary realism alike. 17 Cf. Kant (1952 (German 1790): 50). An example may illustrate what is meant by Kant’s formulation here: While a square or a circle immediately can be determined respectively as a square or as a circle via the simple concept of a square or a circle, foliage on a frame or a tapestry (Kant 1952 (German 1790): 60) is not immediately determinable through a simple concept. Here, the foliage rather gestalts itself as something with smooth edges without being a square or as something with curved lines without being a circle, i.e. as a foliage sensed as something that cannot be subordinated to a simple concept, but exists in a strange floating state between several possible simple concepts. This floating state is the state Kant is referring to when he talks about the existence of a free play between perception (belief) and concept. It is this state which, according to Kant, is the true source of aesthetic pleasure. 18 In these formulations I have been inspired by Wittgenstein who calls the aesthetic judgement: “[…] something like a gesture, accompanying a complicated activity.” (Wittgenstein 1966: § 11). Peter Widell 186 Below, we will take a closer look at these four parameters to see how they are able to characterize literary texts. As we will see, the parts all have their special internal structure and a special role to play. Although they are not directly constitutive of the literature as an aesthetic object - we cannot exclude that other things may have influence on what we find aesthetically commendable - they are, nevertheless, as close to being constitutive of an aesthetic text as we can get: they all give us opportunities to show excellence; and to show excellence is - that is what Kant tells us - aesthetically pleasing. 4 Figurativity By figurativity as a feature of a literary text we will understand the different kinds of repetitive shaping or patterning of the textual expression. Figurativity can always be reduced to patterns of simple or complex repetitions in the textual expression such as are known from e.g. rhyme and metre. The non-literary text must meet, as we have seen, the requirements of informativeness, truthfulness, relevance and grammatical and semantic correctness. This means that the textual expression of normal non-literary prose is more or less perceived as characterized by a lack of conspicuous repetitive shapes or patterns. Let me spell out what it means. If you are pre-occupied complying with the Gricean maxims - and this goes for the authors and readers alike - then repetitive shapes and patterns not grounded in the Gricean maxims will occur as a purely irrelevant and perhaps even disturbing feature of a certain text. We can, of course, always ornate a non-literary text with rhyme, meter etc., and by the feature of figurativity alone turn a text into a literary text. There have, actually, been times, for instance in the Baroque Period, where it was welcomed to ornate applications, recommendations and appointments - i.e. texts otherwise trying to live up to the Gricean maxims. But, it is exactly in this opportunity to meet all the Gricean maxims we find figurativity slightly different from the other literary text parameters. This parameter is not concerned with textual content, as the Gricean maxims normally are, but only with the figurative features of a text. Therefore this parameter is not constituted by the Gricean maxims as the other ones. Its presence in a text is only an indication of the possibility that we are standing in front of a text constituted by breaches of the Gricean maxims. 19 19 Actually, whether the text is just indicating or actually showing a deliberate breaching of the modified Gricean maxims, depends upon whether you choose Grice’s more relaxed definition of correctness, his manner maxim, or the more strict definition I am preferring according to which “correctness” is a matter of grammatical and semantic correctness alone. Grice would certainly have considered texts showing figures and patterns as a violation of his maxim of manner. The Literary Text: Four Parameters 187 Figurativity is not just a textual phenomenon. It is the hallmark of perception in general, reflecting our goals and interests in our actions. Since we always focus on something or other in perception, every perceptual act is, by its very nature, structured as a figure on a ground. Of course, we have most of the time to obtain complex information from our surroundings. But figures can be connected by being iteratively repeated. They can form repetitive patterns that again can be seen as figures - now complex ones - on certain backgrounds. When monitoring a room, for instance, the whole process can be seen as a repetitive pattern of simple perceptual acts where I first look at the table, then look on the things on the table, then at the lamp and so forth, until I have gathered sufficient information - e.g. satisfying the conditions for whatever I want to do next. The figures of figurativity are perceptually given figures. That does not, however, mean that repetitions have to be confined to the same perception situation. They can be distributed over many perception situations, only limited, actually, by the memory capacity of the perceiving subject. The simplest repetitive pattern you can envisage is, of course, two figures on a ground as exemplified in the two figures below: Fig. 1: Two figures on a ground Fig. 1 shows a noteworthy fact about perception, namely that each and every conceivable repetition is subject both to: (a) configurativity, the overall presence of two or several figures in time and space, (b) comparability (in terms of a possibility of similarity or difference) between every two figures in the repetitive pattern. Actually, these features will always show up whenever a perceiving subject moves through distinguishable, uniformly shaped (=comparability) mutually related (=configurativity) single instances. The repetition - with its two constitutive features - is so fundamental that we cannot go behind it or escape it: it is not empirical; it is, actually, transcendental. 20 In its most ab- It could also be argued, as I have already indicated that to ornate a text is to bring features to the text in question that hinders or delays an effective overall understanding of the content of the speech acts in the text. So in that sense figurativity could be seen as a violation of the maxim of relevance and could, therefore, be ranked as a parameter in line with the other parameters. Here we still have to emphasize that we are, then, only dealing with pure expression, not with content. 20 This should be emphasized against a merely psychological interpretation found for instance among the Gestalt Theorists (Wertheimer, Köhler, Kofka). They were the Peter Widell 188 stract form it is just our gaze into the world: focus number one, focus number two. In that sense, we are always confronted with three things: number, location and the question of similarity or not. On closer examination it turns out that we have not just a twofold distinction in (b), but a threefold distinction. If you compare Fig. 1 with the figure below: Fig. 2: Two figures on a ground you have not just a distinction between the similar and the different, but a distinction between (i) the (qualitatively) identical, (ii) the similar or the varied 21 and (iii) the different: in fig. 1 you have an identical repetition consisting of two identical figures, in fig. 2 you have a varied repetition consisting of two similar figures. The fact that you always can demonstrate a threefold distinction in your perception and action between the identical, the similar and the different must not be conflated with the question of when you are seeing things as identical, as just similar, or as completely different. The last question is an empirical-pragmatic question, not a metaphysical one. The question whether two forms or two colors are identical or just similar, is a question of exactness and a question of how you want to use the distinction. The same goes for the configurativity of the figures. Here, the topological and metrical relationships between the figures - are they overlapping or not, are they far from each other or close to each other? - are metaphysical relationships while questions about, for instance, the actual distance between the figures and the occurrence of deformations in relation to certain standard measures are purely empirical-pragmatic questions. 22 pioneers behind the idea of perception as a perceiving of figures governed by certain gestalt laws adding wholes to our perceptual inputs. This kind of psychologism represents an unacceptable kind of antirealism. Figures (on grounds) are not something added to the world by our perceptions. Figures are basically out there in the world. 21 Usually, you talk about identity and contrast (cf. e.g. Leech (1969: 65-66)). Here, contrast is subsumed under the category of the similar, not of the different. Contrast is not negation, but modification. 22 Depending on the situation, the actual distance between two or several figures can be felt as so great that the configurativity of the participating figures, actually, dissolves. Therefore, a hidden premise is usually attached to the concept of configu- The Literary Text: Four Parameters 189 As indicated above, complex patterns can be seen as repetitions of repetitions. They are iterative. Thus, a complex pattern such as the pattern in fig. 3 can, for instance, be seen as a simple repetition of two repetitions: Fig. 3: Simple repetition Instead of simple repetitions we also can have varied repetitions as in fig. 4: Fig. 4: Varied repetitions Here, the variation consists - in the presentation of slightly different ovals in the end of each member - of the two parts of the repetition. Now, what about texts? Since textual expressions are patterns, they exhibit the same characteristics as any other pattern. But textual expressions, of course, are special. The first thing to notice is that texts are man-made objects carrying meanings. Therefore, they usually consist of linear sequences of spoken or written signs. This puts some restraints on how the repetitive patterns may look like. For instance, they are not just patterns of acts of perceiving, they have a special shape dictated by the used material which in the case of spoken signs are sounds made by the human throat and in the case of written signs are graphic figures ultimately consisting of letters. Spoken sounds are the primary medium in language use. Since written signs partly represents oral sounds - the letter “a” represents a certain oral sound, the letter “b” another and so forth - the sound patterns of a text corresponds by and large to rativity, namely that the part figures in a configuration should not be located too far from each other. Or to use a technical term: they should bear the characteristic of contiguity, due to limitations in our perceptual apparatus. As to degree of similarity between figures, transformations such as inversion or litotic repetition are normally considered to preserve an underlying identity. Peter Widell 190 its graphic pattern (at least in phonetically based writing systems). Therefore, we find the same text patterns in both kinds of texts. But which repetitive shapes and patterns do we, actually, find in texts? Here, we must distinguish between the material the patterns are made of, and the patterns themselves. Concerning the material, we find sounds and pauses, we find different sound qualities, that is vowels and consonants, and we find relevant variation in stress, length and pitch. Concerning patterns, we find simple and varied repetitions in sound groups, as for instance end-rhymes, alliterations and assonances, and we find simple and varied, non-complex and complex repetitions of different stress/ length/ pitch-profiles, as for instance in iambic, trochaic and dactylic styles. 23 In classical stylistics, there has been put in great effort in establishing a taxonomy characterizing every possible repetitive pattern in a text. Below, is a text - some lines from the Danish poet Hans Adolf Brorson’s poem Her vil ties, her vil bies - presenting a little sample of those repetitive patterns - referred to by their technical names from classical stylistics: Fig. 5: Repetitive patterns 24 23 It is also possible to find semantically and pragmatically motivated repetitions, carrying highly abstract meanings, repetitions like words (as words tout court), sentences (as sentences tout court), paragraphs, bulleted lists etc. They are, of course, in itself manifestations of a figurative order in the text (and hence possible objects of aesthetic appreciation), but, normally, they are not sufficient salient to transform the whole text to be a literary text. 24 Poem in rough English translation: “Here will be silence; here will be waiting; here will be waiting, O soul so weak; surely, you will be gaining, but only by waiting; only by waiting for the summer to arrive.” Cf. Brorson (1765). Her vil ties, her vil bies Her vil bies, O svage Sind Vist skal du hente, kun ved at vente Kun ved at vente, vor Sommer ind The Literary Text: Four Parameters 191 All this seems also to be linked to the aesthetic judgment and our ability to feel aesthetic pleasure and pain. If we find patterns in our environment in relation to our wants and needs, we also find an opportunity to create challenging situations which require a certain amount of skill. It is precisely here we are able to apply Kant’s thinking from earlier about the relationship between the free play and the beauty to the sentence and to the text. But before we do this, I want to develop Kant’s theory a little bit further towards a more comprehensive phenomenology of the aesthetic sense revealing a certain dynamic: on the one hand we lack pattern and order in our existence, or it is too complex, which implies an inability to master the situation, and this leads - other things being equal - to worry and perhaps anxiety. On the other hand we have that the orderly and all too expected and monotonous, which - other things being equal - leads to boredom. And in the middle we have the skillful - not too dull, not too risky - mastery that creates exactly the Kantian beauty. 25 Beauty is, in that sense, in our more comprehensive phenomenology a balance point between boredom and anxiety. 26 It is important to notice that this balance point - together with the whole scale from the monotonous to the chaotic - always has to be considered relative to a certain culture, a certain group, a certain person. As we have already cited Kant for: aesthetics do not have any absolute standards. And this, indeed, also applies to sentences and texts. As a good illustration of a lack of an objective balance point of beauty concerning sentences and texts, is the observation that varying periods throughout history have had their own favored 25 Even if this dialectics between (a) monotony leading to boredom, (b) free play leading to beauty, and (c) the chaotic leading to worry and anxiety is not as explicitly formulated in Kant as it is here, something similar seems to underlie his theorizing: actually, in Kant (2005 (German 1790): 165), he talks about monotony and boredom; and in his theory of the sublime, partly inherited from Edmund Burke (1990 (1757)), he quite unmistakably approaches the borderline to the chaotic and the fearful (Kant 2005 (German 1790): 92-94). 26 These links between the cognitive and the emotive in the aesthetic judgement could, perhaps, be given an explanation in evolutionary terms. Before the invention of written language human memory was the only storage medium for knowledge. But, here, an artificially inflicted shaping of the material to be remembered, involved a cognitive-economic advantage and, thereby, an efficient means for survival. Concomitantly, these shapes and patterns enhanced the feeling of mastery that came with it. Memorizing was much easier which in itself was aesthetically pleasant. With the rise of written language, this aesthetic feeling of being worn by “another security” became much more conspicuous, and the new medium’s increased storage capacity allowed definitely greater degrees of craftsmanship. The oral presentation with its inability to create a distance between production time and presentation time of a text was replaced by a presentation that allowed infinite opportunities to refine the treatment. All these developments undoubtedly contributed to a sophistication of the aesthetic expression. Peter Widell 192 attitude to the patterning of sentences and texts, perhaps best exemplified by the Romantic breach with the classical form ideal, where the ambition was to try to test all patterns in the text against strict rules listed in various stylistics and rhetoric books. This was, in the Romantic period, felt as artificial, as overloaded, as mechanical, and - yes, ultimately - as a little bit too boring. Some uses of figures are worthwhile mentioning: for instance, it is possible to let parts of figures in a text represent focus points, which can be open to semantic interpretation. Thus, the Danish words “ties” and “bies” (twice) meaning, respectively, “silence“ and “waiting” (cf. fig. 5), can be used to elicit special non-semantic connotations or associations between the two words. But, it has to be remembered that such focus points have nothing to do with attending to the figurativity pattern itself. It would be to add something to the figurativity of the text: a looser reading breaching with the Gricean maxims. Often, rhymes go together with sentences, and - thereby - with what sentences represent. But, when looking for figurativity proper, we are not looking for meaning, but only for patterns in the text expression. Another use of figurativity is letting it imitate human bodily movement. This brings figurativity close to music and dance. Like music and dance, figurativity in texts can be seen as a means of expressing emotion. Frequent repetitions with short time intervals can express unrest, excitement, enthusiasm or joy, whereas frequent repetitions with longer time intervals can express rest, relaxation, sleepiness or mourning and so forth. Besides these feelings, a poem can also have its aesthetic point. Just as a certain dance can manifest itself as an eloquent expression of sorrow, so can the rhymes of a particular poem. In that sense, the figurative aspect of the poem can imitate or mime what the poem is about which is one of the reasons why you cannot so easily separate representation from represented in literature. 5 Tropicity By tropicity as a feature of a literary text we will understand the different kinds of tropes in the text - that is pragmatically triggered acknowledgements of repetitions in the world referred to by the textual trope, for instance Achilles and the lion in the well-known Aristotle-metaphor “The lion (e.g. Achilles) leapt”. 27 The repetitions in tropes are always - as part of the understanding of the metaphor - considered with regard to either similarity, contiguity or both. In fig. 5 we have seen different repetitions in the textual expression “translated” into a sort of painting, so to speak. In this section we will look upon its manifestation in relation to meaning contents. Here, meaning presents itself as a special kind of repetition. In that respect you can say that tropicity is a repe- 27 Aristotle Rhetoric 406b, in Aristotle (1984: 2243). The Literary Text: Four Parameters 193 tition of the figurative repetition, but now in the realm of semantics and - as we will see - even pragmatics. 28 To understand what that means, it is necessary to be familiar with the ordinary repetition of meaning content or, as we are going to call it, literal meaning. The hallmark of the repetition of meaning content in literal meaning shall be seen in a satisfactory first time application of the Gricean maxims. We have previously seen that these maxims ensure that our speech acts are informative, truthful, relevant and (grammatically and semantically) correct. Concerning literal meaning, it is important to notice that it is, in principle, possible to speak about things and situations in the world without presupposing any language based meaning at all. It is, actually, possible to establish ad hoc meaning or - as Grice calls it - occasion meaning. 29 But leaving this borderline case of meaning aside, we normally expect that people we discuss with will be able to identify the permanent meaning of the words said. We, actually, expect that our interlocutor is able to identify what is said as a repetition, namely a repetition of what happened in the baptizing situation where he for the first time learned how to use the words meaningfully. To know what the meaning of the sentence “The earth is round” is, is, then, basically to be able to assert a true proposition about the roundness of the earth in all situations (qualitatively) identical to the baptizing situation - that is, in all subsequent use situations. Without digging deeper into the question of what must be satisfied for such a repetition to take place - this, actually, requires a more comprehensive speech act theoretical approach 30 - it should be relatively obvious, that the maxim of semantic correctness - as an insurance of the baptizing situation’s repetition in the use situation - is complied with by the speaker, and thus also the maxim of truthfulness which in the first place gives the linguistic meaning a supporting platform: when saying “The earth is round”, what you do is, actually, repeating something you have said before. 31 But this repeating is also a repeating in the sense that the expression you truthfully 28 That means that the repetition cannot any longer be aggregated in simple perceptual acts, as it often can with respect to figurativity. 29 Cf. Grice (1959: 213-224). 30 I have tried to implement such an approach in Widell (2001; 2009; 2010), inspired partly by Saul Kripke who in Kripke (1981) coined the expression “baptizing situation”, and partly by Austin who in Austin (1979 (1950)) talks about truth as fixated by pairings of demonstrative and descriptive conventions. 31 Note, that this does not violate the Fregean Principe of Compositionality. Even if the sentence “The earth is round” is used with the exact wording for the first time, it still represents a repetition according to the Fregean Context Principle. Of course, it can, then, not be seen as an instance of an identical repetition, but only as an instance of a varied repetition. As a special case, a varied repetition always presents in any use situation the repetition of the logical form of the sentence, that is the form “R(A, B, C...)”. Peter Widell 194 used as an expression of the perception of a certain configuration of things in the world was used the same way as you did it in the baptizing situation. 32 This is precisely the relationship that forms the basis for understanding the creation of language meaning and the analyticity or literalness of concepts - e.g. the property of concepts that they can be related to each other without reference to empirical knowledge - such as the relation of ‘man’ (used in a baptizing situation) to ‘man’ (used in a corresponding use situation), the relation of ‘bachelor’ to ‘unmarried person’ (synonymy) and ‘monkey’ to ‘animal’ (hyperon my) in English. 33 That is, as we will see, also the basis for understanding the creation of tropes in texts. But what about the tropes? What do they consist in? And how are we going to characterize them? Literal texts are meaningful to the extent that they will double the perceptual and action-related event that has occurred in the baptizing situation. This is, however, not the case with tropes. The point of departure for understanding tropes is a deliberate breaching of or flouting this pattern. Here, we are exactly not invoking the internal relationship between baptizing situation and use situation but, instead, simply comparing - initiated by an understanding of the words of the trope - figures located entirely in the world. In a sense, the situation is the same as in our poem in fig. 5. But there is a difference: the figures we are, now, concentrating on have changed from a comparing of the letters and words in the poetic expression to a comparing of real figures out there in the world alluded to by the words we are using. To repeat fig. 1 here: Fig. 6: Two figures on a ground 32 This means that a sentence reflecting the structure of a proposition at the same time reflects the structure of perception. That is one of the reasons why Frege is founding his logic not just on the sentence, not even just on the sentence used in a certain situation, but on the thought covering both the structure of sentences and the structure of perception and - especially - action: “One brings about changes in the common outside world which perceived by another person, are supposed to induce him to apprehend a thought and take it to be true.” (Frege 1997b (German 1918-1919): 29). Here - in thought - we always find the following recurring structure: On the one hand a certain reference and on the other hand a certain way for this reference to be presented (Frege 1997a (German 1882)). 33 Cf. note 3. The Literary Text: Four Parameters 195 The figures should, now, be seen not as figures of expression in language, but as figures in the world. A popular example in the theoretical literature on metaphor can illustrate this. If Romeo says “Juliet is the sun” - that is, if Romeo use the well known metaphor from Shakespeare’s play Romeo and Juliet - he uses the words “Juliet” and “is” according to the semantic norms established in the baptizing situation for the correct use of the two words. Here, the use situation resumes - is a repetition of - the baptizing situation. But the word “sun” is used quite differently. It is, actually, seen as an invitation from the side of the author to his audience to watch the sun - the actual sun out there in the world - as repeated in the actual Juliet - the actual Juliet out there in the world. 34 It is exactly the nature of this repetition that, throughout history, has been the central topic of discussion in the illumination of the trope and its many different manifestations. 35 Let us first look at the so called master trope, the metaphor. 36 Later on we will turn to some of the other main tropes and their interrelations. As a starting point for such an examination, we can, roughly, say - and I think profitably - that the metaphor consists of three separate parts: a marker, a mechanism and a literal paraphrase. Here, we find, though, a divide concerning the importance, or even necessary existence, of each of the three parts between two different groups of metaphor theorists, which we, in the following, will call, respectively, the Communication Theorists, and the Cognitive Semanticists. While the Communication Theorists adhere to the conception that the three parts exist as analytically given, necessary parts of the concept of metaphor, the Cognitive Semanticists are of the opinion that only the mid part is necessary while the marker and the mechanism are dispensable parts superficially connected to the metaphorical expression. 34 It is important to notice that the fact that we have moved the repetition from the textual expression to the textual content, now, means that the second object of the comparison can be absent. The reader often has to imagine the second object (per implicature). That is not the case when we are dealing with figurativity. 35 The metaphor is, in the classical tradition from Aristotle to Quintilian, defined as an abbreviated comparison (elliptical simile). This means that it is presupposed in the classical tradition that the origin of the metaphor lies in the comparison (simile) of two concrete figures ( or two concrete configurations). That goes even for metaphorical expressions not directly expressing a comparison. In the metaphor “He was a pig” “a pig” is referring to a pig chosen at random. Still, the interpreter of the metaphorical expression has to imagine a concrete pig if he wants to establish the comparison relation (which is, of course, always possible: the interpreter can always refer back to the concrete pig of the baptizing situation for the word “pig”). 36 Since metaphor is considered to be the main trope, or the master trope, you often use the term “metaphor” instead of “trope” for all types of tropes. However, we would prefer to use the term “trope” here for the area as a whole, and then reserve the term metaphor exclusively for what in the tradition appears under that name. Peter Widell 196 The marker. Those who especially pay attention to the marker of the metaphor, the Communication Theorists, tend also to believe that metaphor should be seen as a special type of implicature. 37 Within the communicative paradigm, a metaphor is not a metaphor unless the author has the intention and the ability to create a marker by which the audience can identify this intention. Therefore, the main question for the Communication Theorists is: how do we, in general, recognize markers of metaphors in communicative engagements? How do we know, for example, that there is a metaphor hidden in “Juliet is the sun”? Here, the answer, according to the Communication Theorists, is straightforward: we know this because Grice’s maxim of correctness is deliberately breached or flouted by the speaker or writer. And so it always is, according to the Communication Theorists, when faced with metaphors: metaphors are simply marked, from the side of the author, as a special kind of violation of Grice’s maxim of correctness. 38 To give an example that will reveal this: Juliet in “Juliet is the sun” is not - and could not be - the sun in any meaningful sense of the word. Juliet is not a huge ball of burning gas in outer space. So the author is not using the word “sun” according to the semantic rules for the word. He has intentionally or deliberately breached the Gricean maxim of correctness. Or - to take another example - when the recently deceased Danish author Inger Christensen in her sonnet circle The Butterfly Valley (2004 (1991): 4) speaks about “the dense crimson hue of life”, we again have just another instance of flouted speech: life is not densely colored. So, the author is not using the words “dense”, “crimson” and “hue” according to a possible literal meaning of these words. She has intentionally or deliberately breached the Gricean maxim of correctness. The trope markers can be of different types. Here, metaphor is the most central one. But we can also have other types of tropes, types like metonymy, synecdoche, hyperbole and litotes, as well as types like irony, sarcasm, tautology and symbol. Together this family of tropes forms, according to the Communication Theorists, a beautiful system of breaches of the various Gricean maxims. 37 Adherents of the communicative metaphor theory are among others Grice (cf. Grice 1975), Searle (cf. Searle 1979a), Sperber & Wilson (cf. Sperber & Wilson 1983; 1986) and Glucksberg (2001). 38 The fact that metaphors represent a deliberate breaching of the maxim of correctness has led some scholars to abstain from using the term ‘implicature’ in connection with metaphors; reserving the concept implicature for post-propositional inferences, they instead see metaphors as a sort of explicatures (cf. for example Bach (2001: 253, note 3)). Grice himself (Grice 1989 (1975): 34) perceives of tropes, including metaphors, as implicatures. We will follow this practice. As we have chosen to consider implicature as a kind of umbrella concept covering all deliberate breaches of the modified Gricean maxims, we are protected against any inconsistency here. The Literary Text: Four Parameters 197 Metaphors are forming a special group of tropes together with metonymies, synecdoche’s, hyperboles, and litotes. What holds for the marker of the metaphor holds for the markers of all the other tropes too: they are all characterized by having a marker constituted by a deliberate flouting of the maxim of correctness. The things talked about are either too fragmented - that goes for metaphor and metonymy - too large - that goes for hyperboles - or too small - that goes for litotes. But we also have use for the other Gricean maxims. Contrary to the above group of tropes, irony and sarcasm have a marker constituted by a deliberate flouting of the maxim of truth. For tautology we have a marker constituted by a flouting of the maxim of informativeness and for symbol and allegory a marker formed as a flouting of the maxim of relevance. 39 Thus, all Gricean maxims come, actually, into full play in the shaping of the various tropes. The mechanism. The mechanism of the metaphor is the ball game for those who think that metaphor is a purely cognitive matter - that is for the Cognitive Semanticists. 40 They therefore attach minor importance to the communicative side of the metaphor. Actually, the metaphor itself has, according to the Cognitive Semanticists, nothing to do with texts at all. Metaphors, of course, could have a metaphorical expression in texts, so they recognize language markers, but they do not need to have any language marker. They need not even to be communicated: “The generalizations governing 39 As to symbol and allegory: when, for instance, in Hans Christian Andersen’s fairy tale “The Ugly Duckling” (Andersen 1844), we are reading about the duckling - on the basis of a metaphorical interpretation of the speech and behavior of the duckling (personification) - we are also reading about the general conditions of being a person misplaced or maladapted in the world. It’s important here to notice that this symbolic or allegorical interpretation covering the text as a whole is not prompted by any inconsistency or breach in the semantics of the text such as in metaphor and metonymy. To see the inconsistency you have to include the whole context: “Why are we being told this silly story about an ugly duckling which is, actually, a swan? It seems totally irrelevant to me.” Or does it? Perhaps Andersen has meant his fairy tale to be a lesson about life, about what is possible for a genius born under unfortunate circumstances. But this interpretation has its point of departure, not in a violation of the maxim of semantical correctness, but instead in a violation of the modified Gricean maxims, here relevance. This definition of symbol and allegory has side-effects on our conception of some metaphors: a trope like “He stabbed me in my back” is, just to mention an example, often - when no literal interpretation is possible in the situation - classified as a metaphor; but since this does not represent a flouting of the maxim of correctness, but rather of the maxim of relevance, it should, rather, be classified as a symbol. 40 Adherents of this cognitive conception of metaphor are among others Lakoff and Johnson (cf. Lakoff & Johnson 1980), Lakoff and Turner (cf. Lakoff & Turner 1988), Gibbs (cf. Gibbs 1994) and Turner and Fauconnier (cf. Turner & Fauconnier 2000). Peter Widell 198 poetic metaphorical expressions are not in language, but in thought […] the locus of metaphor is not in language at all […]” (Lakoff 1993: 203). Metaphors are, according to the Cognitive Semanticists, cognitive mappings from a source domain to a target domain (Lakoff & Turner 1988: 63), and they are, in this respect, extensively structured. “Juliet is the sun”, for instance, is not just a similarity between the sun and Juliet. It is a mapping determined by a vast underlying, cognitively structured system of correspondences between source and target domains. The sun is not just our conception of the sun. It also comprises our conception of the earth, the heaven, and all the other heavenly bodies: the moon, Mars, Venus etc. It is this system, and not only the sun itself that creates metaphorical meanings to our experiences. It brings, for instance, the sun to the centre of our planet system; it makes the sun a distant place; it gives us the pale moon at night; it gives us a dark side of the moon etc., furthering our thoughts with a plethora of opportunities to create still new correspondences to, in this case, the vast field of human relations. Finally, our thoughts are, according to the Cognitive Semanticists, founded on one huge coherent, hierarchical structured conceptual system of source domains mapped into target domains: “the Great Chain of Being” (cf. Lakoff & Turner 1988: 160-213). Contrary to the Communication Theorists, the Cognitive Semanticists put great effort in scrutinizing the way we are building larger systems of metaphors, and in that sense they have, without doubt, contributed to the illumination of how we build metaphors. The Communication Theorists have, however, in their attempt to maintain a conceptually necessary distinction between marker and mechanism, questioned whether the Cognitive Semanticists have the right conception of what they are doing in their examinations. For instance, the Cognitive Semanticists claim that their theory is a real, scientifically-based, psychological theory about a text and a communication independent basis for creation of metaphors. That is their major claim. But that calls for text and communication independent confirmation instances. However, according to the Communication Theorists, the Cognitive Semanticists have never really succeeded in establishing such confirmation instances. 41 Therefore, what the Cognitive Se- 41 Cf. Verveake & Kennedy (1996), Murphy (1996), McGlone (2001). I think this criticism is still forceful and sound. There have been resent attempts to show that time is structured as space, not by language, but by a deep language-independent metaphorical cognitive mapping (cf. Borodinsky 2000; Casasanto, Fotakopoulou & Boroditsky 2010; Merritt, Casasanto & Brannon 2010). These attempts are not convincing, simply because they are not as claimed attempts at examining the concept of time at all. Time is - and should be seen - as a metaphysical concept necessarily connected to space and to our acting and perceiving in the world. It is not an empirically concept, separately verifiable. It is transcendentally given - as a prerequisite for all cognition involving time. The Literary Text: Four Parameters 199 manticists are doing should rather, according to the Communication Theorists, be seen as a mixture of philosophical reconstruction and hermeneutical interpretation. 42 But that leaves us with the Communication Theorists interpretation of the metaphorical mechanism: What the Cognitive Semanticists consider as deep, language and communication independent psychological mental mappings from source areas to target areas could, according to the Communication Theorists, equally well be understood as a seeking for local similarities between things and processes in the world prompted by a breaking of a Gricean maxim. This interpretation is so much the better as it will, furthermore, confirm and preserve the Gricean realist presumptions of the concept of literal talk, namely that it consists of talk, not about what is going on in our heads, but instead about real things in the world laying out there for us to perceive and act on, and it will preserve the classical presumptions of how we understand metaphors, namely as local answers to questions of similarities between figures in the world revealing different sorts of simple and varied repetitions normally on the level of what can be expressed in a single sentence. 43 The (literal) paraphrase. According to the Communicative Theorists, a metaphor cannot exist as metaphor without being connected to a literal paraphrase. The Cognitive Semanticists on their side dismiss this presumption. For the Cognitive Semanticists, as for the theorists in the Romantic tradition, metaphors are not just reducible to their paraphrases. They have creative potential. The underlying mental mappings of source domains to target domains add, according to the Cognitive Semanticists, something to our understanding of things and events in the world that we do not have a grip on independently of the metaphor. There are - the Cognitive Semanticists admit - metaphor independent source domains that serve as points of departures for metaphor creation. A sentence like “The cat is on the mat” is, according to the Cognitive Semanticists, quite literal. But for metaphors, as for instance the metaphor “Juliet is the sun”, there are no literal interpretations, no paraphrases. And the same goes for all the areas of our experience where we are not able to refer directly to the physical world. This argument is, however, contested by the Communication Theorist: The distinction between literal and metaphorical meaning is not a contingent 42 As to the last point McGlone writes: “Lakoff and his colleagues base the metaphoric representation claim solely on intuitions about how certain idioms thematically cohere.” (McGlone (2001: 205)). 43 Allegories will, accordingly, be seen, not as extensive mental mappings lurking below the connected metaphorical expressions in a text, but just as series of individual similarities, connected to each other via the world based connectedness of things in the fields of experience referred to in “the source domain.” The journey in LIFE IS A JOURNEY is not something mental, but the actual connectedness of things - roads, landscapes, cars etc. - learned about during our upbringing and given (literal) labels in our language. Peter Widell 200 distinction between certain areas in our experience of the world, for instance the basic physical world alluded to in the example with the cat and the mat, and different, more or less non-tangible phenomenon. There are, of course, non-physical phenomena in the world, and there are lots of them - for instance, all kinds of thoughts and all kinds of social institutions. But according to the Communication Theorists they have nothing to do with nontranslatable, metaphorical mappings. All metaphors - also, for instance, our metaphor “Juliet is the sun” and some of the Cognitive Semanticists favoured metaphors like VALUE ORIENTATIONS ARE SPACES, LIFE IS A JOUR- NEY and TIME IS MONEY - are, according to the Communication Theorists, not creative devices, for there are for the Communication Theorist no such creative devices, but simply devices referring to a text and communication independent world - that is part of the realistic presuppositions in this paradigm - and are as a world revealed through paraphrases literary true of it. This dismissal of the concept of the creative metaphor brings the Communicative Theorists close to the classical theory of metaphor from Aristotle and Quintilian, according to which metaphors are, as we have seen, purely ornamental: according to the Communicative Theorists, metaphors and other sorts of tropes add nothing to our conception of the world. It gives us nothing to grasp what cannot be grasped through literal language. Juliet is still Juliet, the sun is still the sun, and Romeos feelings towards Juliet are still the feelings of Romeo as revealed by the metaphor independently of the metaphor used, and in principle totally describable in non-metaphorical terms. Therefore, we need the paraphrase as a necessary feature of our conception of metaphor - together with the marker and the mechanism. It is difficult not to find the Communication Theorists arguments against the Cognitive Semanticists compelling. Therefore, we have to adopt this tripartite distinction between marker, mechanism and paraphrase in metaphor. But, haven’t we thrown the baby out with the bathwater concerning the creativity of metaphor? Actually, not everything is said about the Communication Theorists’ conception of the paraphrase. The paraphrase is considered as a truth about the situation paraphrased. But that does not exclude the possibility of other true paraphrases. And that is, according to the Communication Theorists, the real - but also the only - secret, if there is any, about the talk about the creativity of the metaphor: Metaphor and other tropes allows us access to not just one, but a series of - and, supposedly, for some tropes in principle indefinitely many - literal interpretations. Or, as we will say, based on the fact that tropes are pragmatic and not semantic phenomena: In tropes, the speaker or author offers his audience a variety of inferential opportunities based on what is (faulty or not correctly) said. As with the other literary parameters, the metaphor, then, reveals itself as a loose 44 or open 45 implicature. 44 Cf. Sperber & Wilson (1983: 234-235). The Literary Text: Four Parameters 201 This means that the only, but important, thing that separates the metaphorical expression from the literal paraphrase is, according to the Communication Theorists, the amount of paraphrases allowed in the communicative exchange. There are no alternative types of cognition and no new entities in the world to be perceived by using metaphors. The understanding of the metaphor is, so far, quite literal. The fact that you, according to the Communication Theorists, can allow in principle indefinitely many interpretations, does not mean that anything can be a paraphrase as, for instance, Donald Davidson has faultily suggested. 46 Of course the marker has, in the midst of its defectiveness, set a framework for what can be a metaphorical interpretation and what cannot. Naturally, “Juliet is the sun” cannot be interpreted as a request to pay tax. Here, the interpretation opportunities are curtailed by the fact that the metaphorical expression lends some of its semantic content to the possible set of paraphrases. It is the same as saying that paraphrases are found among the restricted set of expressions that refer to a similarity between the two things the metaphorical expression refers to. Here, it is obviously the creative responsibility of the author to determine the guidelines for the metaphorical interpretation, while it is the responsibility of the hearer or reader to find the inferences which are the more striking ones among the many possible paraphrases. 47 We have seen that the way the repetition metaphor represents is decidedly different from the repetition the literal expression carries with it: While the repetition contained in the literal expression is internal to language, the repetition in metaphor and other tropes is external to language in the sense that the metaphor is a prompt to find a similarity between two or more things in the world. That means: In metaphor, the literalness degrades into something which equates baptizing with simple use, and that is precisely the claim made by the Communicative Theorists. Therefore, the metaphor looks like a predicate a class inclusion 48 , and yet not: Here, we have a repetition, but there is no requirement that the repetition must be identical to the thing repeated as in 45 Searle states that metaphors, in this context, are not creative helping us to see things we hitherto could not have seen but open ended, allowing several paraphrases (Searle 1979a: 115). 46 Davidson (1978). Although Davidson has a completely different philosophical basis than the adherents of the cognitive metaphor theory, his criticism is in part based on the same argument as the Cognitive semanticists, namely that the boundary between literal and metaphorical meaning is more or less non-existent. He can, therefore, be met with the same criticism as the Cognitive semanticists. 47 It is important to understand that metaphor and other tropes allow the hearer or reader to be more creative than the speaker or author. The speaker or author can lay eggs in the basket of interpretations of the metaphor which he did not even realize he has laid there but which are there for the receiver to enjoy. 48 Cf. Glucksberg (2001) who has exploited the concept of metaphor along these lines. Peter Widell 202 literal talk. It suffices that it is similar. The metaphor “My love is a rose” does not say that my beloved is a rose; instead it says: “Go and look at some rose and see what features in the rose you can reuse for describing your beloved” - but, this praxis of finding similarities has nothing to do with the semantics of the metaphorical expression. It is a purely pragmatic matter. If we scrutinize the matter a little bit more closely, we will find that choosing a metaphor is, actually, the same as choosing a higher level of abstraction. Romeo can describe Juliet in many ways: she is warm, she is sensitive, she is lovely, she is wonderful, she is life-giving, she is unattainable, but Romeo can also simply say that Juliet is the sun. By that Romeo gives a similar characterization of her, but in a much more abstract and at the same time looser sense: He can move in many possible directions. Nevertheless, it is the similarity between the literal and the metaphorical depiction that is central here. In this sense we find the metaphor in the field of text coherence: Metaphors create coherence in a text, but not in the same way as the literal expression. They do it in a more approximate manner - via similarities in the world. So if we are to find a cognitive creativity in metaphor, it must be this: Metaphor is an economic package full of possible paraphrases a package the readers could open and amuse themselves with. 49 But, here, we also find the aesthetic function of metaphor: It is always fascinating to open packages, especially if they present themselves with an evasive and - precisely for that reason - challenging content. In this, tropicity resembles figurativity: they both stimulate creativity and create human joy. As we have already seen, tropicity involves other tropes than just metaphors. Among other things, they also involve metonymy and synecdoche. Now, if you examine these two other tropes, an interesting thing shows up: The field of tropes reveals itself, in fact, to be a deeply seated, relatively simple field related to the ground structure of perceptual experience, actually, the structure we have already dealt with in the section about figurativity. Let us take a look at the following figure: 49 Here, you could say that metaphor meets literal talk two times when a metaphor is created: the first time is (a) when you are using literal words in an etiolated form to create a marker for the metaphor, and the second time is (b) when you are using literal words in a non-etiolated form to create paraphrases. It is important to emphasize that, if we still want to call a metaphor a metaphor, we must at least be able to produce more than one paraphrase. Sometimes we can’t. But, that is, normally, because the metaphor has been killed, as with idioms like “You are pulling my leg”, “Shall we call it a day” and “I am biting the bullet.” That metaphors are brought to death by fixating their meaning does not mean that we cannot revitalize them. But it usually requires special contexts to re-open them as metaphors - contexts of type (a) above. The Literary Text: Four Parameters 203 Fig. 7: Relations between two figures on a ground As you can see, fig. 7 looks like fig. 1 where we found an illustration of a repetition of a linguistic expression. In fig. 1 we found that we could always identify two relations, namely (a) configurativity (“C” in fig. 7), the overall presence of several expressions, and (b) comparability (in terms of a possibility of similarity or difference between the two figures; “≈” in fig. 7). But now the same two relations have, actually, shown up again in connection with the tropes, namely in the form of figures in the world, figures the linguistic expressions can refer to. This is, exactly, the starting point for formation of tropes generally: Two figures A and B are located in a field of perception. That allows either to make inquiries about (i) the degree to which A and B are similar, or to ask for ii) the spatial, temporal, colored etc. relation A and B form with each other or with the overall form. 50 Together with this simple understanding of the relationship in Fig. 7, a crucial juxtaposition of metaphor and metonymy/ synecdoche is pinpointed which has not in the same way - and not with the same radicalism - been realized within the rhetorical tradition from Aristotle to Quintilian. Now, with our new insight, we can see every trope as both a metaphor and a metonymy/ synecdoche. However, it is not always possible to see that it is so. Here, it 50 The relation in (ii) is so far the same metaphysical relation as a relation reflected in a proposition, namely that we will always be able to choose between portraying the referent of the proposition as a relation from part to whole, such as happens in all so-called “to be”-constructions - “The mailbox is red”, “There are pears on the tree this year” - or either from part to part in a configuration or from part to configuration or configuration to part as it occurs in all so-called “to have”construction - “Trees have finally produced pears this year.” These grounding relations and the causal matrix they are embedded in are described by me in Widell (1996). A B C ≈ Peter Widell 204 must further be required that the repetition that takes place of A in B, appears as sufficiently perceptually salient - i.e. perceptual in such a way that B can be seen as a sufficiently specific context for A. Let me explain what this means. In many tropical expressions it is difficult to read both tropes. It is, for example, not easy to see the metonymy in the abovementioned metaphor from Aristotle “The lion Achilles” and the metaphor in the metonymies “Can we eat a can of mackerel and tomato sauce today? ” or “The whole house was turned upside-down.” It is not because the ability to read both parts is not present, but rather that the configuration that constitutes the backdrop of tropical formation cannot form a sufficiently accurate picture of the context as we find them in the corresponding metonymies of the configurations. That does not mean it is impossible for us to do so. We can force them to appear. For instance, we can, for a decoding of the trope “The fog comes on little cat feet” (Carl Sandburg 1916), always ask what similarity “fog” can establish with some figure which could be part of the configuration referred to by the rest of the sentence. Or to take an example previously used: “the dense crimson hue of life” (Christensen 2004 (1991): 3). Here you, for an understanding of the trope, can place the referent of “dense crimson hue” both in a configuration relation to the rest of the sentence, and in a similarity relation to some figure that could enter into the configuration that the rest of the sentence refers to. With this explanation of the relationship between metaphor and metonymy/ synecdoche, and with the elucidation of the difficulties which usually means that we cannot see both the metaphor and the metonymy/ synecdoche in a repetition, we, now, cannot only say that the metaphor is more multifarious than metonymy and in that sense rightly bears the name of the “master trope”. We can also explain it: Metonymy/ synecdoche requires a concrete anchoring of repetition as context. In case you cannot find it for the metaphor, you cannot find the corresponding metonymy/ synecdoche either. 6 Fictionality By fictionality as a feature of a literary text we will understand the different kinds of pretended non-truths in the text, intended by the author and signalized pragmatically or semantically for an audience as so intended by him - where the pretence is a repetition or miming of other texts living up to the standards of the Gricean maxims. Figurativity, literal meaning, and tropes all build, as we have seen, on something very simple, namely repetitions of figures: While figurativity in written texts consists of repetitions of letters, letter groups, and other expressions, literal meaning consists of repetitions of situations - that is, repetition in use situations of what has been learnt in baptizing situations - and tropes The Literary Text: Four Parameters 205 in repeating one object (i.e. Juliet) “through” another object (i.e. the sun) (on the basis of common features). As we will see, the two last parameters, fictionality and literary realism, also build upon repetition. But, instead of attending to mostly repetitions of sentence parts or sentences, as in figurativity and tropicity, we will here predominantly attend to repetitions of whole texts (or at least more extensive text parts). It is a widespread opinion that fictionality refers to something imagined or to a world different from this world, called a merely possible world. 51 It is, however, an unfortunate way of characterizing fictionality. The reason why is that you cannot square it with the conditions of communication: if you want to get your message through to someone you are communicating with, then you have to operate via instances that are observable by you and the persons you are addressing: inner states must have outward criteria. 52 Otherwise they cannot exist from a communicative point of view. 53 That is, fictionality is - as one normally uses the concept - neither a concept linked to the imagination (if you understand it as a kind of inner movie), nor a semantic concept. It does not consist of swapping from reference to object in this world to objects in an imagined or fictional world different from the actual world as it is falsely envisaged from this semantic understanding of the concept. But how should we understand fictionality, then? When we speak, what we say may be true or false, and when we are talking falsely, we can do it in several ways. We can do it because we are ignorant. We can do it because we want to lie, perhaps by getting some benefit from letting someone believe it. Or we can say or write something untrue because we want to create fictionality, a piece of fictional prose. In creating fictionality we create a falsehood, but also something that is not a lie. We hide nothing. Quite the contrary: If we want to create fictionality, we have to show our intention - usually because we want a particular purpose accommodated with what we are saying, or, for short, because we have the meta-intention that our intention to say something false is apparent. It is important to note that this reflective element in fictionality always seems to be included in the creation of fictionality. In that respect, the question of whether a piece of text is literal or 51 Lewis (1978). 52 This is a hidden citation from Wittgenstein (2001 (1951): §580). 53 Wittgenstein (2001 (1951): §293). In this paragraph, Wittgenstein is presenting us for an analogy from which we can draw the appropriate conclusions. He asks us to imagine a situation where we all have a box in front of us in which only the owners of the respective boxes can see into his own box. When suddenly a person rises up and points into his box and says “beetle” what has he, then, communicated to all of us? The answer is of course nothing because there are no outward criteria for what he has pointed at. And that will be the same for inner experiences, mental images, private thinking and things like that. They do not exist independently of their public manifestations in perception and action. Peter Widell 206 fictional is totally up to the author. Exactly as it is when we are lying. Whether what is said is true or not is a matter of investigation, and here we are all participants. Not when we lie. And the same goes for fictionality: if you deliberately want to suspend the question of the truth of what you are saying, it is totally up to you to do it. 54 This can also be expressed by saying that fictionality is not so much a feature of the content of what is said or written, as it is a feature of the speaking and writing of it. You cannot always just by looking at what someone says or writes - the wording of what he is saying or writing - tell whether it is a piece of fictionality or not. In that respect fictionality is not a matter of syntax or semantics, but exclusively a matter of pragmatics. You have to consult the intention of the author (as it is revealed to you by the author) in order to know whether what you are reading, is fictional or not. 55 Most adherents of the pragmatic approach characterize fictionality in the following way: When an author writes something fictional, e.g. a novel or a short story, he is engaged in letting it be clear for the reader that he is just pretending to say or write what he is saying or writing. Fictionality is, then, for the pragmatically oriented analyst considered to be an ad hoc signalized pretended reference and truth. 56 Here “pretence” means that the author, as to the content, is doing exactly the same thing as he would have done if he actually referred and asserted something true, and that the only difference is that while reference and truth play a decisive role in a non-fictional text, this role is suspended in the corresponding fictional text. Or as the English 18th century poet Samuel Taylor Coleridge 57 has formulated it: fictionality is actually a 54 A common objection to this viewpoint is that there are a lot of examples where a text has been considered fictional by some and non-fictional and, actually, true for others. In this context many refers to the Bible as originally considered to be a book of telling the truth about how the world was created and ruled, but now mostly taken as a fictional work (or at least a book which requires a symbolic reading). However, this is not a counterexample to the view that fictionality is up to the writer to decide. The Bible is not a fictional work exactly because it has never been intended to be a fictional work. But, of course, you can always read the Bible as if it were a fictional work. 55 Perhaps it should be pinpointed that this is not a case of the intentional fallacy (Wimsatt & Beardsley 1954). It has nothing to do with the intentions concerning the content of what is said. It has only to do with the intention to communicate the ontological status of what is said, namely that it is a piece of fictionality. 56 Cf. Searle (1974-1975), Currie (1990), Walton (1990), Lamarque & Olsen (1994). The signalizing part of fictionality is put into parenthesis because it is normally not mentioned by adherents of the pragmatic approach - only implied. It is important, though, to emphasize - as, for instance, Searle does - that pretence cannot operate in a context consisting of hidden motives. If we detect such motives, we do not have fiction. We just have plain dissimulation. 57 Coleridge (1817: 6). Coleridge speaks in this connection about “poetic belief”. The Literary Text: Four Parameters 207 “suspension of disbelief”. This must not be misconceived, though, as a wish from the author that the reader in each and every respect should react on what he is reading, as if the text was not fictional. 58 This will not create fictionality. It will at most amount to a seduction of the reader. When dealing with fictionality, the reader should all the time have the opportunity to maintain meta-awareness about the fact that what he is reading is not something that is true, but something which at most could be true. In that sense a fictional text participates, as we have already mentioned, in a hollow, etiolated or parasitic form of communication. 59 Relating it to the Gricean maxims, fictionality is defined by flouting the maxim of truthfulness, and thereby also the maxims of informativeness and relevance - but, important to notice, not correctness: the words used in fictional speech acts retain their usual meaning. 60 Searle has - as we have seen - defined fictionality as a non-deceptive form of pretended reference and truth. But how does the author create the message: “What I say to you is just pretence (and not just meaningless rubbish)”? Here, Currie (1985), for instance, defines fictionality - along this kind of explicit indication - as a special kind of illocutionary act. But this seems a bad idea - as Currie himself later realizes (Currie 1990) - because this doubles up the number of types of illocutionary acts, so that for each and every literal illocutionary act there has to be a corresponding illocutionary act of fictionality. And that sounds implausible. Here, Searle defines the device thus: “Now, what makes fiction possible, I suggest, is a set of extra-linguistic, nonsemantic conventions that break the connection between words and the world […]” (Searle 1974-1975: 66). They are, as Searle wants to call them, “horizontal conventions” (Searle 1974-1975: 66). I think this characterization is wrong: sets of extra-linguistic, non-semantic conventions do not occur out of the blue. They stem from something simpler. The question is, however, how they are established. 58 Hearing or reading fiction is hearing or reading it as in discord with the modified Gricean maxims; nevertheless it is not forbidden to invest your feelings and emotions in what we are hearing or reading. And it is, actually, rather difficult to resist this investment. Not being able to resist investing our feelings and emotions in fictionality is a psychological feature which presumably has at least a part of its psychological explanation in the fact that we are thereby given the opportunity to practice empathic preferences for our fellow men, which probably again - in the light of evolutionary theory - could be seen as essential to the survival of the human race as a social species. 59 Cf. Austin (1962: 51). 60 The fictional text can, of course, be influenced by other literary parameters - for instance figurativity - which more or less can corrupt the meaningfulness of the text. Peter Widell 208 Since they, according to Searle, are conventions, they must have been established as conventions. But, how is it possible to establish a convention like the one Searle is talking about? Actually, Searle is telling us very little about it. Searle calls the conventions in question “extra-linguistic” and “non-semantic”. And, indeed, they must be located outside the pretended speech acts - and in that sense they must be extra-linguistic and non-semantic. But can they be conventions at all? Of course, you can always create conventions supporting language and institutions - also relating to fiction: when it is indicated at the front page of the book you hold in your hand that it is a novel or a collection of fairy tales, you are, for the interpretation of this indication, depending upon a convention governing the words “novel” and “fairy tale”. The words are shorthand versions of the assertions “The book in your hand is a novel” and “The book in your hand is a fairy tale collection” together with implications indicating that it is, therefore, fictional texts. And that is OK. 61 But, such paratextual indications, as Genette calls them 62 , are often not present in situations where we make judgments about the fictionality of a text. So what are we doing in these situations? Do we have a convention after all, but a different kind of convention? I do not think so. I think that we have and, actually, must have to define fictionality, intentions without conventions. What you hear from the author is not acceptable for you as a true description. On the other hand you cannot accept it as a lie. There are too many features in the context to indicate that it could be a lie either or - to put it in other words - it seems too be blatantly irrelevant as a lie. Therefore - and that is, in my opinion, the true specification of the marker of fictionality - a marker of fictionality is not a convention, and not just a violation of the maxim of truthfulness, because it is not a lie, but rather a potential candidate for being just an intentional violation of the Gricean maxim of relevance. 63 That is in a way a surprising result, because fictionality has almost exclusively been associated with the Gricean maxim of truthfulness. But, here we must be careful to distinguish between the marker and the content (corresponding to the “marker” and the “mechanism” of the metaphor): It’s the marker which is decisive for the definition of fictionality while there are, actually, no restrictions on the pretended content - at least, if you do not want to 61 The same goes for conventions integrated into the fictional text as, for example, standard phrases as “Once upon a time […].” Such standard phrases must not be taken to be constitutive of fictionality. 62 Cf. Grice (1997). 63 That is why fictional speech acts can, in principle, all be held true by an author, and still be fictional speech acts. The non-true character of speech acts in fictional literature is only an indication of the irrelevance of the text. It is not a constitutive element of it. The Literary Text: Four Parameters 209 restrict the content to just narrative fictionality, but want to recognize it for what it is, namely just a kind of fictionality. You can also pretend to perform orders, promises, announcements, exclamations etc. Here, you are not pretending to tell the truth. But the acts are, of course, still fictional. They are fictional orders, fictional promises, fictional announcements, fictional exclamations etc. It would be bizarre not to call them that. This means that fiction is a much more comprehensive category than a category just comprising narrative fictionality. Fictionality also includes what is going on in theatres and movies: here, we find fictionality just as well as in what we read when we are reading narrative fictionality. The only difference is that we, here, call the fictionality dramatic instead of narrative. However, in theatres and cinemas we can see some interesting things taking place on stage not so easily spotted in written literature: Here, we find that not all acts are speech acts as is the case in a book. Some of them, namely the lines, are what the characters say to each other; but others are not. For example, sitting in a sofa or mixing a drink. But they are fictional acts all the same. They are expressions of the fact that the actors are pretending to be the characters of the play. In a play everything is pretended - even the stage setting. But, it is here, I think, we shall find the core element in a proper definition of fiction. Fiction is pretended being, not just pretended saying. This should be understood in its most general form. Pretended being is not just confined to non-true propositions in a book. It is not just confined to literature, and to what is going on in theatres and cinemas. It is an essential ingredient in social games in general. What happens in theatres, cinemas and books is kindred to what is going on in nursery rooms or among animals at play. Here, we find a common effort to establish a fragile social, not yet conventional, bond between individuals. It is here we shall find the common Gricean framework, within which the author or the actor has the possibility to create, sovereignly, all the fictional acts and modes of being they want. Text fiction is just a smaller part of this larger concept of fictionality. Let me, then, in the light of these considerations on dramatic fictionality, try to spell out what fictionality of texts - according to my opinion - consists of. Instead of starting with semantics, we will now start our explanation with pragmatics 64 or, to be more precise, with the pragmatics of cheating. My strat- 64 Richard Walsh (2005) has tried to ground fictionality pragmatically, as we are trying to do it. Unfortunately, he thinks that he can do it without taking into account the meaning concept operative in Grice’s theory of implicature. Following Sperber & Wilson he tries to base the concept of implicature on only one maxim, the maxim of relevance. Briefly put, he is trying to skip the Gricean maxims of truthfulness and correctness as independent maxims carrying steady meaning contents. That is a strongly counterintuitive case of inflationism: of course, we can meet borderline cases. But we cannot doubt that “Juliet is the sun” represents a violation against the meaning rules of language (cf. note 3). Peter Widell 210 egy will be to establish a sequence of pretending scenarios beginning with the most primitive form, cheating, and ending up with the most sophisticated form, text fictionality: (1) The most primitive pre-form of text fictionality is cheating as we find it among higher animals, for instance in connection with hunting: Here, the prey is trying to cheat the predator by escape moves: the preys actions are not what they seem to be. This forms the basis for building (2) forms of “cheating” which are framed in double Gricean intentions of communicating that this “cheating” is a “cheating”; generally, such ad hoc frames constitutes forms of non-deceptive pretending. (3) On the basis of these kinds of double intentions more conventional sets of markers are, gradually, created as, for instance, a conventional marker for attending a play in the form of a certain building equipped with a stage and seats for the audience: the building with all its facilities simply counts as a non-deceptive place for performing non-deceptive pretended acts. (4) Under the scope of such markers, the whole stage setting and everything the actors are doing, including their lines, are by the author, the actors, and the audience, understood as pretended actions under instruction from the author; you could say, as Plato once did, that the author through his instructions to the actors are using them as a means of his addressing the audience with their pretended actions. (5) Included in the play we could now have a situation where one of the players is reading aloud from a book; let us say he is reading aloud from the telephone book; then we have a situation where an actor standing before the audience is pretending (before the audience) to tell a story to the by-standing characters in the play. (6) But, this situation can, again, lead to the following situation: let us say that the actor is, instead, reading a fairy tale, e.g. the fairy tale The Princess on the Pea by Hans Christian Andersen (cf. Andersen 1844); then he is, now, pretending (before the audience) that he is pretending (before the characters in the play) Another attempt at grounding fictionality pragmatically is Genette (1993 (1991): 30-53) who tries to analyse fictionality in terms of indirect speech acts. This is unfortunate. Even if we do not see a doubling of illocutionary speech acts as with Currie (1985), we have a major problem because we are violating the concept of an indirect speech act: Indirect speech acts always have a direct illocutionary act complying with the modified Gricean maxims. Here, we do not have a situation like this: The direct speech act cannot be a direct speech act in the Searlean sense because it is etiolated as such an act already from the start. It is, so to speak, born as a fictional act. Furthermore, even if this requirement is met, it presupposes that fictionality is a demand on each and every speech act all through the fictional text. But that sounds counterintuitive. The marker for fictionality is, normally, a marker set only once for a given text which is - in the vast majority of the cases - the text as a whole. Sometimes, it can, of course, take a while before the reader realizes that a given text is fictional. But if the text is once identified as fictional by its expressions, it is normal - as a rule of thumb - not to change opinion about the fictionality of the text. And if you do, you will probably not call what you are exposed to a text but, rather, a collage of texts, a fictional text and a literal text. The Literary Text: Four Parameters 211 to read this fairy tale. (7) Let us, now, say that the actor has been hired to entertain an audience in an evening arrangement at a public library reading aloud the same story he was reciting onstage, but now from a podium; then, his fellow actors have disappeared, but we still have a situation where an actor pretends to tell a story - now directly to the audience. But exactly this last situation is, I think, the secret key to the understanding of text fictionality. For even if we, in this situation, have two audiences fused together, which admittedly is a major transformation, there is nevertheless still a clear separation between the actual author and the actor telling the story before the audience. (7) And if the actor is the author himself reading his own text - let us say it is the fairy tale author Hans Christian Andersen reading his fairy tale “The Princess on the Pea” before an audience of children sitting around him - we have, actually a situation, where the author is reading a piece of literary fictionality before his audience. (8) Now, this piece of fiction can, easily, transform itself to a book. That will be the case when the author - e.g. Hans Christian Andersen - instead of telling the story to an audience of listeners, prefers to present it in written form to an audience of readers. Notice, that the distinction between author and actor from the stage is still preserved in the book: what has happened is just that the distinction has turned into a distinction between an author writing the text and a narrator pretending to read the text for the audience. 65 (9) This distinction can, now, be semantically conventionalized as it has happened with a marker as “novel” on the front page of a novel or the marker “Once upon a time […]” or “And they lived happily hereafter” as “internal” markers in a fairy tale. Actually, throughout history a lot of more or less conspicuous expressions have gradually evolved, characterizing the different styles in narrative fiction, and in fictionality generally, expressions which, in addition to their functioning inside the fictional text, also point to the text as more or less unmistakably marked piece of fictionality. 66 65 Perhaps it sounds strange to many to consider the relationship between author and narrator in a fictional narrative text as being parallel to the relationship between author and actor in a dramatic text. It perhaps makes sense - you could interpose - to say of a narrator reading his own text that he is an author instructing himself as narrator. But does it make sense when the author is addressing his audience as a writer? I think so. Of course, he is no longer instructing himself as a narrator in the case of a fictional narrative text. But there is still left the discrepancy between the situation of writing the work as an author and the situation of presenting the work to the audience when it is finished. Here, the parallel to the dramatic text is preserved: when addressing the actors, the person writing the dramatic text is exposing the same discrepancy between his role as a writer and his role as a person presenting the text. 66 It is important to emphasize that these styles do not constitute fictionality. The fictional form may facilitate these styles. And they can be important text features to be studied in their own right. They do not, however, define fictionality. Unfor- Peter Widell 212 But why do we have fictionality at all? What is fictionality for? Of course, it is mainly done for the aesthetic fun and pleasure of it, and, therefore, actually not done for anything special. That applies to the level of personal experience, and this is a reason we in a certain sense cannot ask behind this experience, as we have seen. But that does not mean that we cannot ask for a more objective, biological reason. Since fictionality is connected to social play, it offers - like play - opportunities to engage in human practice without being exposed to the restraints of real action. Truth is suspended, and with that the usual dangers and risks of life. This makes fictionality a perfect place for rehearsal, for training and for educating your skills. And this represents, undoubtedly, an evolutionary gain: the time used for education of your skills represents an enhancement of the adaptability in a complex environment of the species you belong to. And fiction, definitely, furthers this. 7 (Degree of) realism By (degree of) realism as a feature of fictionality in texts we will understand to what extend they are similar to non-fictive texts - that is texts living up to the standards of the Gricean maxims. It is well known that a fictional work can be subjected to thematic, symbolic or moral interpretation. The Ugly Duckling, for instance, is first and foremost a fictional story and a story loaded with tropes, especially anthropomorphisms; but it is also a story calling for a thematic or symbolic interpretation: The Ugly Duckling, is, really, we have learned, about how a born genius, in spite all odds, is bound to realize his potential as a genius (cf. note 36). It is, however, important to remember that such thematic or symbolic interpretations have nothing to do with the fictionality of the story. The question of the thematic or symbolic meaning of a fictional story is really a question added to the question of fiction, and belongs therefore to the area of tropicity: The story serves qua story as a marker of a trope, here a symbol or an allegory, calling for a paraphrase. Through the paraphrase the story is turning itself to a statement about reality. You could call this aspect of fictional literature an aspect of realism. But since this paraphrase is not a proper part of the fictional text it is a little bit misleading to associate this kind of realism with fictionality. And, usually, you do not talk about realism here. The real question of realism is found, not in the interpretation of the text, but in the very constitution of the fictional text as realistic. To this question we will now turn. tunately, many researchers have claimed that, partly following the claims of the Russian formalists. The Literary Text: Four Parameters 213 But before we do that, we first have to separate this kind of realism - we will call it literary realism 67 - from another concept of realism that we will call philosophical realism. Philosophical realism is the kind of realism we have talked about all along in the discussion above. It is simply the doctrine saying that the existence of the thing referred to is not created by the act of referring, but presupposed by it. Since fictionality is pretence, literary realism cannot be founded on this concept. It must be coined as a separate concept. But, why do we normally have a very strong feeling that certain fictional texts seem more realistic - more true to reality - than certain other texts? A question like that, of course, calls for an answer. 68 To answer the question we should perhaps take a closer look at our concept of fictionality. As we have seen, fiction takes its point of departure in 67 We should - perhaps more appropriate - have called literary realism fictional realism, because of its attachment exclusively to the parameter of fictionality. Then we will be able term-wise to distinguish between realism in texts - literary realism - and realism on stage - dramatic realism, both hyponyms of fictional realism. I choose, however, in this article to use the more common term literary realism. 68 It should be mentioned that we, in fictional literature, can meet not just pretended reference and truth and symbolic truth, but also, of course, real - that is: philosophical - reference and truth. When in Madame Bovary (Flaubert 1858) the narrator is talking about the city of Rouen and the village Yonville-l’A’bbaye we suppose it is the real city of Rouen and the real village of Yonville-l’A’bbaye he is talking about and not just fictional localities. And when we in the beginning of the second part of the novel read the extensive description of the town of Yonvillel’A’bbaye we presume that it is a fairly accurate description. Actually, most of the readers do not know, and live happily with that. They are not interested in historical or topological details of this region of France, and read it just as if it were a piece of fictionality (cf. note 60). If what we read is, in fact, true of the town of Yonville-l’A’bbaye, then it is reasonable to say that the fictional text is flavoured with reality and in that - philosophical - sense realistic (but irrelevant in relation to the aesthetic appraisal); and, in general, it is reasonable to say that the more a piece of literary prose refers to real places, real times and real things and persons, the more philosophical, but not necessarily literally realistic, it is. In that sense Winston Churchill’s autobiography My Early Life (Churchill 1874-1904) could be said to be more realistic than Sir Walter Scott’s historical novel Ivanhoe (Scott 1906 (1819)). Since it is not my concern to contribute to genre theory and to the prototypical concept of fictionality (as to prototypicality, cf. note 3) I will not go further into this question here. Likewise, I will not go into the related question of how many non-fictional elements we shall pass into a text and still call it fictional. Let me just say this: if segments of a mixed text is not true, then these parts have to be signalized as not true by the author. Otherwise they are not fictional. If the reader is deprived the means to evaluate whether a text is fictional or not, but the author knows they are not true, then, concerning these parts, the text is just a heap of lies. And if he is mocking in his signalizing or reluctant to signalize that he is doing it, he is just cheating. Peter Widell 214 feinting, and in feinting - as in the more evolved form: pretence - nothing is what it seems to be. But the thing that seems and the thing it is a seeming of are not totally disconnected. There is a likeness or similarity between what is feinted or pretended and the feinting or pretending of it. That’s why Plato and Aristotle talk about fictionality as a sort of mimicry, as mimesis. We have intentionally underplayed this mimicry element until now, because of its minor role in the definition of the concept of fictionality. For, although it is always present in fictionality, it is a relatively independent feature. It comes in degrees. But it is this we will identify with literary realism. In literary realism we always ask questions like: how true is the pretence to life, how true is it to reality? To clarify, more precisely, how this concept of literary realism evolves in human interaction and communication, it could be useful to take another look at our animals from last section. In our hunting example from earlier there isn’t any understanding between the hunted and the hunter about what the hunted is doing. If there were, the hunted would never escape its predator. But this means that there in feinting isn’t any basis to build mimicry upon. To build a platform for mimicry, we must have an agreement among the participants about what to compare. This can, actually, be the case, among some higher animals, for instance among young monkeys nipping each other during play. 69 Here, we find a mutual understanding and confidence between the two animals that the nip is not a bite. That is: Here we have a point of departure for comparing two things attentively: the nip and the bite. That is, actually, the birth of the concept of literary realism and any other form of fictitious realism: A situation of mimicry has evolved from a situation of a mere feinting. The rest of the story of literary realism and fictitious realism in general is just the story of how we can transform this fragile distance between the factual and the counterfactual - the distance from the nip to the bite in the same situation - to a distance between on the one hand a more freestanding etiolated example of a bite, and on the other hand bites in general. If we can imagine this transformation, we then have a backdrop for comparison which covers all fictional props and acts, including those we find on stage in theatres, in movies, and in books, and we have our measure of the degree of realism. We can now say that the more a fictional unit A - from a unit in a game among animals to a unit on the stage in a theatre or in a book - is similar to a unit outside the fictional situation B, known independently of the fictional unit A, the more realism we find in A with respect to B. That is, if we watch two persons onstage sitting on a sofa engaged in an exchange of words, this situation becomes more realistic, the more it resembles a (kind of) situation outside the theatre where two people are engaged in a similar exchange of words. 69 This example I have borrowed from Bateson (1972 (1955): 179). The Literary Text: Four Parameters 215 Concerning this similarity relation, it is important to see that it has nothing to do with reference. The realism here is not philosophical realism. It is fictitious realism: the situation on stage does not refer to a (kind of) situation outside the theatre. It is a relation of comparing between the two situations - like in painting. In painting a picture states nothing about what it depicts. It is just more or less similar to what it depicts. 70 The same is the case with a stage play or a novel. As to the stage setting and all the props in a play, we have a sort of picture or sculpture. Just as Leonardo da Vinci’s picture of Mona Lisa - irrespectively of whether it refers to Mona Lisa or not is a more or less realistic - of a young lady, so the stage setting in A Doll’s House by Henrik Ibsen from 1879 - as it is presented to the audience - forms part of a more or less realistic example of a house. The same goes for Nora’s leaving the house at the end of the play. Here again, we have just another example of a person leaving a house behind. And likewise when Nora says to Torvald: “As I am now, I am no wife for you”. These words, in their context, again form a more or less realistic example of a sentence, a sentence we as well can meet in a real life context. When you are attending a play sitting in a theatre, the furniture on stage is a part of the pretended setting on stage, and that they are more or less realistic examples of furniture, in the sense that they can look more or less like real furniture, which we can meet in almost every home. The same is true for the lines in a play. They are more or less realistic copies of speech acts from real life, and the more realistic they are, the more adequately embedded in the various situations on stage they are, in accordance with the Gricean maxims. When Nora, in act 2 of A Doll’s House, says to Mrs. Linde: “Let us sit down here on the sofa”, then Nora is referring to a (more or less reliable) sofa on stage. So besides the hollow use of speech acts on stage, there is a genuine use of them to connect to things and characters on stage contributing to the realism of the play. We can say that, besides miming situations outside theatre, the speech acts performed on stage have an internal reference to all that can be seen on stage by actors and audience (things referred to without having a prop on stage are, however, only referents of pretended referring acts). All this seems to change when we look at novels or other literary texts not playing a part in a play. Of course, we have parallels to what is happening on stage. What is true for a line on stage is, certainly, also true for a sentence delivered in a novel, a short story or a fairy tale. When Flaubert “paints” his portrait of Emma in Madame Bovary, we have the same opportunity to evaluate his work concerning its degree of realism. But if we ask what the elements 70 A picture may, of course, be used for referential purposes. I can, for instance, show a person a picture and say: “This is a portrait of Hans Christian Andersen.” Then this person knows what the famous fairy tale writer looks like. But, this reference is not due to the picture. It is due to my decision to let the picture be a part of my pointing gesture. Peter Widell 216 of the comparison are, we nevertheless get embarrassed. In the theatre, we have the setting which we can be put up against a situation in real life. And we have their internal reference to the props and persons on stage, securing the coherence of the performances there. But we don’t seem to have anything like that in a text. Here, all seems to shrink - the props on stage, the actors. All these kinds of referents seem to disappear. We still have, though, an element of comparison: we still have the text. In Madame Bovary we can still ask what the text looks like in the real world. Perhaps the answer looks a little bit odd since it cannot have anything to do with what the text is about, because it is fiction: it is about nothing. However, we still have an answer: We still have other speech acts. Here, we have something we can compare the speech acts in the book with. But that is exactly our situation: fictional realism is the degree to which a sequence of speech acts constituting a text is similar, or resembles, other sequences of speech acts constituting other texts. In Madame Bovary there are no sofas, chairs, walls, etc. to bring in comparison relationship with real sofas, chairs, walls in the world, and to secure the coherence of what is going on internally. Yet, we can still pretend to write a recipe in a book; namely, by miming a real recipe, and we can still pretend to write a diary; namely, by miming a real diary. And we can still pretend to write a sermon; namely, by miming a real sermon, and so on. That is, it is entirely in miming of speech acts from other texts that we find literary realism. Mimicry in fictional texts must not, though, be thought of as sheer copying. It is a highly varied repetition. What is mimed is neither the exact wording of other sentences, nor the content of the speech acts connected to them, but, rather, the way the speech acts are put together in a more circumstantial fashion. It is the style and the general circumstances of producing the speech act sequences in a text. A realistic text or text part is, actually, a kind of pastiche over other texts: when reading Madame Bovary we experience the realism of the psychological portrait of Emma, together with the realism of the scrupulous description of the houses, the villages, the landscapes, etc. That is, here we see Flaubert miming or imitating. He is miming or imitating modern research methods building on meticulous observation and the style of scientific texts being true to fact, consistent and exhaustive. That is why the term chosen for what Flaubert is doing in Madame Bovary, namely naturalistic realism, is a well-chosen term: the objects of comparison are topological descriptions, articles on engineering, psychiatric reports etcetera. Other sorts of realism can be found in saga-realism, historical realism, social realism, everyday realism and many other kinds of realism. The only requirement for fictional realism - at least as an analytical concept - is that the fictional text, in this respect, should mime or imitate a comparable more or less coherent non-fictional text-corpus or style of writing. To secure realism in fictional texts, it is important to notice though, that not every kind of mimicry will work. The texts mimed or imitated in a fic- The Literary Text: Four Parameters 217 tional text should comply with the Gricean maxims. They must - for their part - be real as texts, that is: informative, truthful, relevant and correct. 71 Integrating this in our considerations we are now able to present a reliable definition of literary realism: a fictional text - a novel, a short story, a fairy tale - is a realistic text - in the sense of literary realism - to the degree that it shows similarities with texts which lives up to the Gricean maxims of informativeness, truthfulness and relevance. 72 It should be remembered that since the text in literary realism is presupposed to be fictional it is, of course, as the fictional text complying with the maxim of correctness. As can be seen, this definition represents a confirmation of Plato’s and Aristotle’s theory of fictional realism: realism in fictional literature is mimesis. Literary realism is a kind of fictionality. This means that the parameter of realism is not an independent one. Facing this, it could be interesting to ask if it is possible to make fictional realism an independent parameter - or rather to see realism as an exemplification of a more general, wholly independent parameter. Is this possible? Since literary realism is a kind of comparison - we are comparing fictional with non-fictional texts - then a way to do this is to generalize this comparison to texts in general. It seems that we can do this. When evaluating texts as more or less realistic, we are, as we have seen, relating them to non-fictional standard texts: to recipes, to manuals, to contracts, to reports of different kinds, to stories, etc. each delivering their measure of centrality. But this measure is independent of whether the text is fictional or not. To put up classification standards is normally to create conditions for finding core cases and borderline cases, and this is the case for texts in general, too. So, a more general parameter for comparison of texts as instances of fictional realism seems to be text prototypicality: how close to a standard for a given type of text, S, is a given text, A. 73 This means that the headline for this section - if we want to level it with the headlines figurativity, tropicity and fictionality - perhaps, should have been not degree of realism, but rather text prototypicality. The concept of literary realism shares - as we have already hinted at - an important property with the tropes. First of all, we have a comparison or simi- 71 It is, of course, not possible to mime another fictional text. Miming a fictional text would be to pretend pretence. And that is still pretence: a group of actors engaged in a play embedded in a play are still engaged in playing. 72 By this definition we not only have classified degrees of realism, but also degrees of departures from realism: fairy tales, phantasy, science fiction, gothic, satire etc. Perhaps these departures from realism can be parameterized. I will not, though, go into this here. 73 If this wider concept of text prototypicality is applied to the fictional text then this implies a corruption of the modified maxim of correctness too. Grice’s original maxim of correctness, his maxim of manner, would, however, probably let such texts pass more easily. Peter Widell 218 larity relation (and a comparison to standards of the normal, the natural, the possible, the realistic) both in literary realism and in metaphors. And the marker for identifying, on the one hand, literary realism and, on the other hand, metaphor is also quite similar. In both cases we have a flouting of one or several of the Gricean maxims. This means that we must expect borderline cases where it is difficult to decide whether the text represents a breach of the Gricean maxim of correctness in the direction of tropicity or a breach of the Gricean maxims as a whole in the direction of a loss of reality. 74 However, we also have important divergences. Where the tropes are mostly (but not always) bound to the singular sentence, fictionality and literary realism, are mostly (but not exclusively) bound to whole texts. 8 Conclusion While tropes and fiction has been subjected to a substantial amount of attention during the last forty or fifty years, the question of literary realism has hardly been touched on, at least not in an appropriate manner. Probably, one of the reasons is that realism - in postmodern literary theory - has been considered a projective or constructed category. Here, realism is reduced to a sort of stylistic trick, as, for instance, with Roland Barthes who, in the headline of one of his essays, refers to realism as a means to create “the reality effect” (cf. Barthes 1989). This conception of realism I have tried to counter by showing how important philosophical realism is for obtaining a decent concept of literary realism. First, I have tried to show how important it, in the first place, is to separate philosophical realism from literary realism (instead of conflating them or neglect philosophical realism altogether). Secondly, I have also tried to make it clear that a proper concept of literary realism cannot be obtained without defining it in terms of philosophical realism. The best way to understand literary realism is by seeing it as a parasitic or etiolated variant of literal talk and writing that builds on philosophical realism. In fact, the assumption of realism in the philosophical sense has occupied us not just in connection with literary realism, but has, actually, accompanied us all through the investigation of the four parameters of figurativity, tropicity, fictionality and (literary) realism. Let us then sum up from our discussion how the relationship between philosophical realism - referred to below by the words “real” and “reality” - and realism in literature in general is to be understood mediated by the Gricean maxims: 74 Are the events occurring in Lewis Carroll’s Alice’s Adventures in Wonderland (Carroll 1865) metaphors, or are they merely effects of a defective miming of a text referring to reality? This is not always an easy question to answer. The Literary Text: Four Parameters 219 1. Figurativity: The figures in a literary part of a text are a real phenomenon. They lay out there in the world. They are not in any sense semiotic. They exist in their own right irrespective of whether they are identified by us or not. And because of this independency it is a VIO- LATION OF THE G RICEAN MAXIM OF RELEVANCE which makes us suspect that the text is etiolated or void, that is, unreal, not living up to the Gricean maxims. 2. Tropicity: The tropes in a literary part of a text are related to reality in the sense that they are dependent on the public world. When we meet a metaphor, we are identifying it because of its falling short of reality. What the metaphor is “saying” is something that reality - as it is engraved in the meanings of our words - forbids us to say. That is: the metaphor represents a VIOLATION OF THE G RICEAN MAXIMS OF COR- RECTNESS ( FOR METAPHORS ), TRUTH ( FOR IRONY , HYPERBOLE AND LI- TOTES ), INFORMATIVENESS ( FOR TAUTOLOGY AND SIMILE ) AND RELE- VANCE ( FOR SYMBOLS AND ALLEGORY ) in order to help us reinstalling reality through our inferential gifts. 3. Fictionality: The fictionality of a text - normally a whole text - is related to reality as suspense of that reality. In that way it re-affirms reality. Fictionality borrows from reality in order to let us play in a certain distance from reality. It represents a VIOLATION OF THE G RICEAN MAXIM OF TRUTH in order to establish its own fictional reality. 4. (Degrees of) realism: The realism of a text is not the same as a reality referred to. It is a reflected form of realism miming realistic texts, that is, texts complying with the Gricean maxims. That is, in a violation of the Gricean maxim of truth, derived from its being fictional, a new “reality” - a fictional reality - is put up. As we have seen, there is an intimate interplay between philosophical realism, literary realism and the Gricean maxims from (1) to (4). Here, the relationship between philosophical realism and literary realism is systematically established via a breach with or flouting of one or several of the Gricean maxims. We can illustrate this by listing some examples where the arrow is an implicature generated inference by the hearer or reader: Peter Widell 220 (1) “This is (probably) literature.” reality (in a philosophical sense) (2) unreal pictorial object flouting the maxim of correctness (metaphor) “Men think of sex constantly.” paraphrases relating to reality flouting the maxim of correctness (metaphor) Incorrect formulation of a proposition Juliet is my beloved” paraphrases relating literally to reality Juliet is the sun” “ “ The Literary Text: Four Parameters 221 flouting the maxim of relevance (symbol) pointless in context Now, we want peace” paraphrases relating to reality flouting the maxim of relevance (symbol) pointless in context [The Ugly Duckling] paraphrases relating to reality A genius always makes his name” Boys are boys” flouting the maxim of informativeness (tautology) Uninformative proposition paraphrases relating to reality Boys behave in their special way different from girls and grown-ups” “ “ “ “ Peter Widell 222 As can be seen, this way of characterizing realism refutes a crude projectivistic conception of realism according to which realism is merely a style. If realism HELMER goes into his room. The MAID ushers in Mrs. LINDE, who is in travelling dress, and shuts the door.] Mrs. Linde [in a dejected and timid voice]. “How do you do, Nora? ” Ibsen (1879) flouting all the maxims except correctness (dramatic realism) pretended reality (being) [ This is fictional”] (3) “The Nellie, a cruising yawl, swung to her anchor without a flutter of the sails, and was at rest. ...” Conrad (1899) flouting the maxim of truth (literary realism) pretended reality (being) [ This is fictional”] “The Nellie, a cruising yawl, swung to her anchor without a flutter of the sails, and was at rest. ...” Conrad (1899) flouting the maxim of truth (degree of realism) pretended reality [This is a literary realistic report] (4) “ “ The Literary Text: Four Parameters 223 were merely a style, it could be put at the same level as other sorts of styles. But that is not possible in our characterization of realism. Here, we find a hierarchical relationship between two sorts of realism, philosophical realism and literary realism. For a projectivist (1) - (4) are chimeras. Let us see why the projectivists cannot be right in rejecting (1) - (4)? Everything stands and falls with philosophical realism. That is, with the supposition that (1) the Gricean maxims are complied with, and that (2) the compliance with the Gricean maxim of truthfulness consists in letting the asserted propositions correspond to the language-independent facts. According to the projectivists this is not possible. We cannot - so the argument goes - compare language with reality without using language. All facts are language impregnated. Or as Jacques Derrida, provocatively, has stated it: “Il n’y a pas de hors-texte.” (Derrida 1976 (French 1967): 158). This is gibberish. Correspondence to facts in the world is, of course, not established via language, but via thought embedded in perception and action. It is the correspondence between thought and world in perception and action that counts. When I see a boiling kettle on the stove, it has nothing to do with the fact that I have formulated, or am able to formulate this fact in language (which I can if I have learned a language). The reason why I see the kettle on the stove is the fact that it is placed on the stove and I have the ability to see it, and this has nothing to do with my mastering of a language in which I can state this fact. Because we have a language-independent world, and we can perceive and act independently of the using of language, (1) - (4) is not a chimera. It is real. As to (1): language expressions are real physical phenomena in the world existing independently of whether they have been assigned a meaning or not; and as to (2): manmade more or less awkward physical pictures are real as starting points for making guesses as to what is communicated, and propositions referring to such physical pictures refer to something real, the same goes for paraphrases, which really and literally state facts about the world as implicature initiated interpretations of such pictures; as to (3) and (4): fictionality is something that, despite being not real, presupposes reality as a standard of comparison. All those distinctions are clear and simple distinctions indirectly confirming realism. If we choose to give up realism in favour of projectivism they will collapse: philosophical realism will be the same as literary realism as merely ways of putting words together; metaphor will appear as continuous with literal talk; and truth will be indistinguishable from what is going on in theatres. This I do not find satisfying. Choosing philosophical realism means that literature has to be thought of as parasitic on non-literary texts. That does not make literature merely a means of entertaining and leisure. Literature is - as it is the defined through Peter Widell 224 the four parameters we have attended to - in my judgment primarily aesthetically defined. To choose figurativity, tropicity, fiction and realism is to choose a path of aesthetic appreciation. That does not mean that it could not - at the same time - be, for instance, of cognitive importance. That is one of Kant’s achievements to have seen that. When choosing the path of aesthetics, one is choosing non-instrumentalism. But, one is not, thereby, abstaining from practicing cognitive skills. On the contrary: celebration of the beautiful and the sublime is, actually, at the same time a celebration of mastery - mastery of technical skills in dealing with nature and mastery of social skills in dealing with other persons, and here literature has its mission beyond furthering pleasure, too: The reading of literature is also an initiation into forms of life, which enhance the ability to make distinctions in perception and action and serves as an instrument for cultivating empathy towards other persons and furthers moralization - and through that real autonomy in the true Kantian sense of the word. References Andersen, Hans Christian. (1844) Nye Eventyr. Første Bind. Første Samling (English translation: Virginia Haviland, Hans Christian Andersen, Erik Christian Haugaard (1983) Hans Christian Andersen: The Complete Fairy Tales and Stories, 216-224. Anchor Folktale Library.) Aristotle. (1984) The Complete Works of Aristotle. Princeton, NY: Princeton University Press. Austin, John L. (1962) How to Do Things with Words. Oxford: Oxford University Press. Austin, John L. (1979) Philosophical Papers. Oxford: Oxford University Press. Bach, Kent. (2001) Speaking Loosely: Sentence Nonliterality in Peter French and Howard Wettstein (eds.) Midwest Studies in Philosophy, vol. 25: Figurative Language, 249-263. Blackwell Publishers. Barthes, Roland. (1989) The Reality Effect in Roland Barthes The Rustle of Language, 141-148. Berkeley: University of California Press. Bateson, Gregory. (1972) Steps to an Ecology of Mind. New York: Ballentine Books. Blakemore, Diane. (1992) Understanding Utterances. Oxford: Basil Blackwell. Boroditsky, Lera. (2000) Metaphorical Structuring: Understanding Time Trough Spatial Metaphors in Cognition no. 75, 1-28. Brorson, Hans Adolf (1765) Her vil ties her vil bies. Hans Adolf Brorson Svansang. http: / / adl.dk/ adl_pub/ vaerker/ cv/ e_vaerk/ e_vaerk.xsql? ff_id’34&id’8818&hist’ fmH&nnoc’adl_pub Brown, Penelope and Levinson, Stephen C. (1986) Politeness. Cambridge: Cambridge University Press. Burke, Edmund. (1990 (1757)) A Philosophical Inquiry into the Origin of our Ideas of the Sublime and the Beautiful. Oxford: Oxford University Press. The Literary Text: Four Parameters 225 Carroll, Lewis. (1865) Alice’s Adventures in Wonderland. http: / / www.gutenberg.org/ ebooks/ 11. Casasanto, Daniel, Fotakopoulou, Olga and Boroditsky, Lera. (2010) Space and Time in the Child’s Mind: Evidence for a Cross-Dimensional Asymmetry in Cognitive Science no. 34, 387-405. Christensen, Inger. (2004) Butterfly Valley: A Requiem (translated by Susanna Nied). Canada: Penguin Books. Churchill, Winston. (1996) My Early Life 1878-1906. Scribner. Coleridge, Samuel T. (1971) The Collected Works of Samuel Taylor Coleridge, Princeton, MA: Princeton University Press. Conrad, Joseph. (1899) Heart of Darkness. http: / / www.online-literature.com/ conrad/ heart_of_darkness/ Currie, Gregory. (1985) What Is Fiction? in Journal of Aesthetics and Art Criticism no. 43, 385-92. Currie, Gregory. (1990) The Nature of Fiction. Cambridge: Cambridge University Press. Davidson, Donald. (1974) “On the Very Idea of a Conceptual Scheme.” Proceedings and Addresses of the American Philosophical Association, no. 47, 5-20. Davidson, Donald. (1978) What Metaphors Mean in Critical Inquiry, no. 5, 31- 47. Davidson, Donald. (2001) Inquiries into Truth and Interpretation, Oxford: Clarendon Press, 2nd edition. Derrida, Jacques. (1976) Of Grammatology. Baltimore: Johns Hopkins University Press. Frege, Gottlob. (1997a) Sense and Reference in Peter Ludlow (ed.) Readings in the Philosophy of Language, 563-584. Cambridge, MA: A Bradford Book, The MIT Press. Frege, Gottlob. (1997b) The Thought in Peter Ludlow (ed.) Readings in the Philosophy of Language 9-30. Cambridge, MA: A Bradford Book, The MIT Press. Genette, Gerard. (1992) The Architext: An Introduction. Berkeley: University of California Press. Genette, Gerard. (1997) Paratexts: Tresholds of Interpretation. Cambridge: Cambridge University Press. Genette, Gerard. (1993) Fiction and Diction. Ithaka, NY: Cornell University Press. Gibbs, Raymond W. (1994) The Poetics of Mind: Figurative Thought, Language and Understanding. Cambridge: Cambridge University Press. Glucksberg, Sam. (2001) Understanding Figurative Language: From Metaphors to Idioms. Oxford: Oxford University Press. Goffman, Erwin. (1955) On Face-work: An Analysis of Ritual Elements in Social Interaction in Psychiatry, no.18, 213-231. Grice, Herbert P. and Strawson, Peter F. (1957) In Defense of a Dogma in Philosophical Review no. 65, 141-58. Grice, Herbert P. (1989) Studies in the Way of Words. Cambridge, MA: Harvard University Press. Ibsen, Henrik. (1879) A Doll’s House. http: / / www.gutenberg.org/ ebooks/ 2542 Kant, Immanuel. (1952) Critique of Judgment. Oxford: Clarendon Press. Kant, Immanuel. (2007) Critique of Pure Reason. Penguin Classics. Kripke, Saul. (1981) Naming and Necessity. Oxford: Blackwell. Peter Widell 226 Lewis, David K. (1978) Truth in Fiction in American Philosophical Quarterly, no.15, 37-45. Lakoff, George and Johnson, Mark. (1980) Metaphors we Live by. Chicago: University of Chicago Press. Lakoff, George and Turner, Mark. (1988) More Than Cool Reason: A Field Guide to Poetic Metaphor. Chicago: Chicago University Press. Lamarque, Peter and Olsen, Stein Haugom. (1994) Truth, Fiction, and Literature: A Philosophical Perspective. Oxford: Oxford University Press. Leech, Geoffrey N. (1969) A Linguistic Guide to English Poetry. London: Longman. McGlone, Matthew S. (2001) Concepts as Metaphors in Glucksberg, Sam (ed.) Understanding Figurative Language: From Metaphors to Idioms. Oxford: Oxford University Press. Merritt, Dustin J., Casasanto, Daniel and Brannon, Elizabeth M. (2010) “Do Monkeys Think in Metaphors? Representations of Space and Time in Monkeys and Humans.” Cognition, no.117, 191-202. Murphy, G. L. (1996) On Metaphoric Representation in Cognition, no. 60, 173- 186. Sandburg, Carl. (1916) Chicago Poems. New York: Henry Holt and Co. Scott, Walter. (2000) Ivanhoe. Penguin Classics. Searle, John R. (1969) Speech Acts: An Essay in the Philosophy of Language. Cambridge: Cambridge University Press. Searle, John R. (1971) A Taxonomy of Speech Acts in Gunderson, Keith (ed.) Language, Mind, and Knowledge, Minnesota Studies in the Philosophy of Science, vol. III, 344-369. Searle, John R. (1974-75) The Logical Status of Fictional Discourse in New Literary History, vol. vi, 319-332. Searle, John R. (1975) Indirect Speech Acts in Cole, Peter and Morgan, Jerry (eds.) Syntax and Semantics , vol. 3, Speech Acts. Amherst: Academic Press. Searle, John R. (1978) Literal Meaning in Erkenntnis, vol. 13, no. 1, 207-224. Searle, John R. (1979a) Metaphor in Searle, John R. Expression and Meaning, 76- 116. Cambridge: Cambridge University Press. Searle, John R. (1979b) Expression and Meaning. Cambridge: Cambridge University Press. Sperber, Dan and Wilson, Deirdre. (1983) Relevance. Oxford: Oxford University Press. Sperber, Dan and Wilson, Deirdre. (1986) Loose Talk in Proceedings of the Aristotelean Society 86. Turner, Mark and Fauconnier, Gilles. (2000) Metaphor, Metonymy and Binding in Barcelona, Antonio (ed.) Metonymy and Metaphor at the Crossroads, 264- 286. New York/ Berlin: Mouton De Gruyter. Verveake, J. and Kennedy, J.M. (1996) Metaphors in Language and Thought: Falsification and Multiple Meanings in Metaphor and Symbolic Activity 11, no. 4: 273-284. Walsh, Richard. (2007) The Rhetoric of Fictionality: Narrative Theory and the Idea of Fiction. Columbus: The Ohio State University Press. Walton, Kendall. (1990) Mimesis as Make-Believe: On the Foundations of the Representational Arts. Cambridge, MA: Harvard University Press. The Literary Text: Four Parameters 227 Widell, Peter. (1996) Aspektuelle verbalklasser og semantiske roller: Den dobbelte aspektkalkule in Odense Working Papers in Language and Communication no.10, 135-168. Widell, Peter. (2001) Illokutioner er assertiver: Et bidrag til belysning af talehandlingens struktur og taksonomi in Augias no. 56-59, 15-65. Widell, Peter. (2009) Logik, mening, handling og tale. MUDS - Møderne om Udforskningen af Dansk Sprog, vol. 12, 295-312. Widell, Peter. (2010) A Plea for a Simple Solution: What a Performative Theory of Illocutions Really Looks Like in Götzsche, Hans (ed.) Memory, Mind and Language, 296-308. Cambridge: Cambridge Scholars. Wimsatt, William K. and Beardsley, Monroe C. (1954) The Intentional Fallacy in The Verbal Icon: Studies in the Meaning of Poetry. Lexington: University of Kentucky Press. Wittgenstein, Ludwig. (1978 (1966)) Lectures and Conversations on Aesthetics, Psychoanalysis and Religion. Edited by Cyril Barrett. Oxford: Blackwell. Wittgenstein, Ludwig. (2001) Philosophical Investigations: The German Text with a Revised English Translation. New York & Oxford: Wiley-Blackwell. Bionotes Simon Borchmann is professor of Pragmatics and Textlinguistics at Roskilde University, Denmark. His doctoral dissertation presents a didactic model for describing, explaining and dealing with literary texts characterized by incoherence in the reading process. His major fields of interest are information structure, text understanding, experimental pragmatics, genre and linguistic and poetics. He is a member of the danish pragmatic circle and the co-editor of Skandinaviske Sprogstudier. Peer F. Bundgaard is associate professor at the Center for Semiotics, Aarhus University. He has published a series of articles on phenomenological linguistics and cognitive aesthetics in Journals such as Synthese, Phenomenology and the Cognitive Sciences, Semiotica and Cognitive Semiotics. He is the Editor-in- Chief of the journal Cognitive Semiotics (de Gruyter). Jan Engberg is professor of Knowledge Communication at the Department of Business Communication, School of Business and Social Sciences, University of Aarhus, Denmark. His main areas of research interest are the study of texts and genres in the academic field, cognitive aspects of domain specific discourse and the relations between specialised knowledge and text formulation. His research has been focused upon communication and translation in fields of academic communication like climate change communication and law and is published in journals related to specialised communication like Fachsprache, International Journal for the Semiotics of Law and Journal of Pragmatics. Finally, he is co-editor of the international journal Fachsprache and member of the editorial or advisory boards of a substantial number of international scholarly journals. Carmen Daniela Maier, PhD, is associate professor, member of the Knowledge Communication Research Group and of Center of Corporate Communication at the Department of Business Communication, School of Business and Social Sciences, Aarhus University, Denmark. Among her latest publications are the chapters “A multimodal analysis of the environment beat” in Critical Multimodal Studies (Routledge, 2013) and “Stretching the multimodal boundaries of professional communication” in The Routledge Handbook of Language and Professional Communication (Routledge, 2014). She is the co-editor of “Texts, Images and Interactions: A Reader in Multimo- 230 dality” (De Gruyter, 2014). Her research areas include multimodal discourse analysis, social semiotics, knowledge communication, environmental communication and corporate communication. Her main research goal is the theoretical and methodological development of the multimodal analysis of knowledge communication. Ole Togeby is professor of Danish Language at Aarhus University since 1991, leader of the Research program in Language and Linguistics at Faculty of Arts, Aarhus University. He has published books (in Danish) about sociolinguistics: Kvinder siger selvfølgelig mest - sagde manden (Women of course talk the most - said the man, 1985), about text theory: Praxt. Pragmatisk tekstteori (Praxt. Pragmatic text theory 1993), about Danish grammar: Fungerer denne sætning? (Does this sentence function? 2003), and about text types and Genre: Bland blot genrerne! (Do blend the genres! , in press) Peter Widell is associate professor of Danish Language at Department of Aesthetics and Communication, Faculty of Arts, Aarhus University, Denmark. He has published a series of articles mostly in Danish in areas such as philosophy of language, model theoretical semantics, pragmatics, text theory, genre theory, aesthetic philosophy and moral philosophy and psychology. He is a member of the Danish Pragmatic Circle and co-editor of Skandinaviske Sprogstudier. Svend Østergaard is cand.scient. and PhD in semiotics. He is currently associate professor at the Center for Semiotics, Institute for Aesthetics and Communication, Århus University. His current research interests focus on dynamic models of social interaction and especially on how language structure emerges as a result of interaction. Currently he is part of a group that studies the use of language in situations when people have to coordinate their behaviour in relation to solving a task. He has published a number of articles on dynamic semiotics and the books Mathematics of Meaning (1997) about the use of catastrophe theory and mathematics in the study of semantics, and Kognition og katastrofer (1998) about cognitive linguistics and its relation to the theory of dynamic models. 047211 Auslieferung Mai 2011.indd 6 18.05.11 10: 31 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 BESTELLEN! Magnus Pettersson Geschlechtsübergreifende Personenbezeichnungen Eine Referenz- und Relevanzanalyse an Texten Europäische Studien zur Textlinguistik, Band 11 2011, 222 Seiten, €[D] 58,00/ SFr 77,90 ISBN 978-3-8233-6623-2 Die vorliegende Studie nähert sich einem brisanten sprachpolitischen Thema - Personenbezeichnungen, die gleichzeitig auf Frauen und Männer referieren - erstmals aus textlinguistischer Sicht. Sie zeigt dabei, wie geschlechtsübergreifende Personenbezeichnungen in zeitgenössischen deutschen Texten, vor allem im feministischen Magazin Emma , benutzt werden. Anhand qualitativer Textanalysen werden Muster in der Variation zwischen Maskulinum, Beidbenennung und Neutralform festgestellt. Die Studie ist die erste textlinguistisch angelegte Arbeit, die sich diesem Thema jenseits der Frage der politischen Korrektheit der verschiedenen Benennungsstrategien widmet. 033411 Auslieferung April 2011 8 08.04.11 13: 52 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 BESTELLEN! Jakob Wüest Was Texte zusammenhält Zu einer Pragmatik des Textverstehens Europäische Studien zur Textlinguistik 12 2011, XII, 269 Seiten €[D] 68,00/ SFr 96,90 ISBN 978-3-8233-6642-3 Was macht aus einer Abfolge von Sätzen einen Text? Diese grundlegende Frage der Textlinguistik lässt sich nicht auf einer rein linguistischen Grundlage lösen. Das liegt daran, dass das Verstehen von Texten eine aktive Tätigkeit ist, die mehr als nur sprachliche Kenntnisse voraussetzt. An einen Text gehen wir dabei mit einer gewissen Erwartungshaltung heran, die das Verstehen steuert, aber auch im Verlauf der Lektüre modifiziert werden kann. Unsere grundsätzliche Annahme ist dabei, dass einem Text nicht nur eine bestimmte Kommunikationsabsicht zugrunde liegt, sondern dass auch dessen Sätze beziehungsweise dessen Sprechakte irgendwie untereinander zu einem Ganzen verbunden sind. Diese Verbindungen, die wir Konnektive nennen, werden aber häufig nicht sprachlich markiert. Aufgrund der Untersuchung zahlreicher Textsorten kommen wir zum Schluss, dass das Inventar der Konnektive durchaus begrenzt ist. Zudem ist ihre Verwendung auf bestimmte Textsorten beschränkt, so dass die Kenntnis der Textsorte es erlaubt, den Aufbau eines Textes ohne Mühe zu erkennen. Die vorliegende Studie beruht auf einer Synthese von Ansätzen, die vor allem aus der germanistischen Textlinguistik, der französischen analyse du discours und der Psycholinguistik stammen. Europäische Studien zur Textlinguistik 13 The presents 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.